We are at a threshold. The discipline of architecture has spent over forty years adrift in a self-indulgent pluralism, mistaking its own fragmentation for freedom and its disorientation for richness. This manifesto declares that the interregnum is over. A new epochal style has matured – Parametricism – and within it a decisive sub-style has crystallised that now constitutes the most sophisticated, and most consequential architecture of our century: Tectonism.

I declare Tectonism to be the most advanced architectural style of our time: the mature, self-conscious, technically empowered phase of Parametricism. It is the architectural articulation of the computational civilization now unfolding across the world, and in this sense historically necessary. It is the style adequate to the 21st century because it alone is capable of giving spatio-visual order to the unprecedented complexity of contemporary society. Tectonism is not one option among many. It is the rightful hegemonic style for the 21st century, and I proclaim it as such without apology.

Let us be clear about the historical stakes. Styles are not matters of taste; they are the means by which architecture absorbs and responds to societal transformation and re-coordinates the collective design effort of an entire epoch. Each epoch or socio-economic stage of society was – with some delay – eventually served by a unified epochal style. The Renaissance answered early capitalism, the Baroque responded to mercantilism and absolutism, Historicism answered the bourgeois nation-state, and Modernism facilitated Fordist mass society. The age of Modernism is over. Its heroic historical achievement must be acknowledged, but its recipes have become obsolete. Modernism belonged to the mechanical age, to Fordism, to the regime of industrial standardisation, with mechanical mass production delivering a universal consumption standard, and in the urban realm homogenisation via repetition and functional separation via zoning. It delivered standard solutions for a mass society organised around typified institutions, stable roles, and rigid functional divisions. But our world is no longer that world. Our society – the Post-Fordist network society – is defined by new, decisive socio-economic features: flexible specialisation, dense communication, perpetual re-association in fluid networks. The architecture it demands is not the serial monotony of Modernism, which expired with the Fordist mechanical mass production, nor the cacophony that erupted in Modernism’s wake. It is a complex, variegated, lawfully differentiated order. Parametricism delivers that order, and Tectonism is its most potent contemporary instrument. A unified style is not authoritarian monotony. It is the precondition for cumulative intelligence. It allows many authors to contribute to a shared language. It allows difference without chaos, variation without arbitrariness, freedom without noise. Nature is not monotonous because it is lawful; it is inexhaustibly rich because it is lawful. Tectonism aims at the same condition: an urban ecology of differentiated, correlated, mutually resonant forms, generated by intelligible principles and open to infinite variation.

Parametricism substituted the static object with the adaptive field, the isolated element with the scripted relation, the rigid type with the malleable topology. Yet the early phases of Parametricism—foldism, blobism, swarmism—were only the opening movements. They released architecture from the tyranny of the box, but they still had to acquire a deeper discipline. Tectonism is this discipline. It is the moment when Parametricism absorbs the intelligence of advanced engineering, structural optimisation, environmental performance, material computation, and robotic fabrication, and turns these logics into articulate architectural form.

We reject the comfortable consensus that stylistic pluralism is an inevitable condition in our time. During the 1970s and 1980s, in the period of transition from Fordism to Post-fordism, architectural Modernism experienced its crisis and demise. But Parametricism was not yet on the horizon. Transitional styles like postmodernism, deconstructivism emerged, competing with high-tech as the latest stage of modernism. The break-up of architectural culture into competing search trajectories was, at this moment, a natural and necessary development. But the moment where this was productive has long passed. Retro-styles like neo-historicism and minimalism (neo-modernism) also vied for attention. Collective brainstorming was necessary but cannot be a permanent condition. Postmodernism and Deconstructivism performed their transitional tasks, but they cannot guide us now. They broke the spell of modernist certainty, but they did not construct a new order. During the 1990s parametricism emerged when a whole generation of architects extended and transformed what deconstructivism had achieved via the utilisation of new computational design tools and processes. Since the early 2000s tectonism emerged when the computational empowerment encompassed also engineering and construction, with tectonism folding this back into the design process.

A viable, mature paradigm has evolved and proven itself for nearly a quarter of a century. However, parametricism’s and tectonism’s expansion stalled since the 2008 crash and adoption remains sluggish. This is the architectural expression of the general stagnation the advanced societies are experiencing (due to over-politicisation and statist overreach). Very restrictive, politicised planning prescriptions are slowing down development and disincentivising innovation. From our discipline’s perspective this continued failure to converge upon the new paradigm is not pluralist tolerance – it is intellectual confusion due to the immaturity of an undisciplined discourse. The unity and coherence of our cities is at stake. A built environment that cannot agree on its own language cannot orient those who must live within it. Pluralism is not a virtue when it produces dysfunction, illegibility and visual chaos. It is a pathology to be overcome.

We must also defend the very concept of style against those who imagine the concept to imply a superficiality that undermines the seriousness of the discipline. Style is not surface, fashion, or transient decoration. The concept of style(s) in the design disciplines plays the role the concept of paradigm(s) plays in the sciences. A style is the means by which the collective labour of architecture is coordinated into a coherent epochal project – a shared system of principles, methods, and values that lets thousands of designers, working independently across the globe, contribute to a single civilisational order rather than towards cacophony and mutual obstruction. Without a unifying style there can be no cumulative progress, only the churn of fashions. To name and canonise a style is therefore not vanity but a precondition of advance. Neither the parametricism manifesto put forward in 2008, nor the tectonism manifesto put forward here, is an invention or imposition; it is the self-description, recognition and consolidation of a sustained, worldwide avant-garde movement that had matured long before it was given its name. Tectonism is the sharpest edge of that movement now.

What, then, is Tectonism? Tectonism is the stylistic heightening of engineering and fabrication-based form-finding into a system of architectural articulation. Tectonism is not engineering. It is architecture making engineering logics speak. It does not abandon Parametricism – it rationalises and intensifies it. Where earlier sub-styles – Foldism, Blobism, Swarmism – explored innovative compositional logics, Tectonism embeds a series of technical rationalities that secure both greater efficiency, greater morphological rigour, while preserving sufficient degrees of freedom for the designer to exercise the discipline’s irreducible core competency: the framing of social communication through built form. With respect to this core competency of architecture, Tectonism’s versatile power of expression goes beyond anything hitherto available to architects.

Gothic vaults, Gaudí’s hanging models, Candela’s shells, Frei Otto’s tensile structures and material experiments were anticipations of Tectonism’s agenda. They discovered that form is not a platonic idea but that efficient form can emerge from the intricate interplay of forces. They are the true precursors of tectonism. But only today, with computational engineering, finite element analysis, topology optimisation, environmental simulation, parametric modelling, machine learning, robotic fabrication and 3D printing etc., can the pioneer’s intuition become a universal architectural paradigm. What was once exceptional craft and empirical genius can now become a systematic design research programme. Tectonism is therefore not nostalgia for craft. It is the radicalization of tectonics under conditions of computation.

Tectonism is also the true architecture of sustainability. Sustainability is not achieved by attaching smart machines to dumb boxes. It must shape the very morphology of architecture. Sun, wind, rain, thermal exposure, daylight, ventilation, drainage, material minimisation and fabrication constraints must become architectural drivers. The building’s orientation, massing, envelope and articulation must respond visibly to environmental conditions. A façade should translate and reveal solar exposure. A roof should register water and wind. A shading system should become an urban compass. Environmental rationality must become spatial and visual intelligence.

Tectonism delivers an aesthetic revolution. It brings forth a new aesthetic sensibility and thus a new conception of beauty: not the classical architectural beauty of symmetry and proportion, not the beauty of pure geometry, not the beauty of industrial repetition, but the beauty of organized complexity. It is the beauty of force flows, gradients, fibrous fields, intricate branching structures, reticulated shells, porous variegated skins, differentiated exoskeletons, cascading seams, adaptive apertures and environmental modulation. It is a beauty close to the beauty of organic nature, not because it imitates nature’s appearance, but because it shares nature’s mode of lawful formation. But beauty is never a free-standing value in its own right. It is perceived via intuitive, immediate subjective responses and signals beneficial high performance environments to those who acquired the up-to-date aesthetic sensibilities. Whenever the technological and/or social conditions of beneficial performance change, our aesthetic sensibilities must be brought in line with these conditions via aesthetic revolutions.

Engineering is responsible for the technical performance of the built environment. Architecture is responsible for its social performance. These two competencies must be sharply distinguished, yet the two must intensely collaborate. The engineer guarantees physical functioning; the architect orchestrates spatial and visual communication. But under Tectonism, the very same forms serve both purposes. Structural members, shells, ribs, exoskeletons, brise soleil structures, tessellations, apertures, gradients, joints, seams, fabrication traces become at once technically motivated and communicatively charged. They lead a double life: silently carrying forces, modulating climate, enabling fabrication, while visibly orienting users, characterising spaces, and articulating social situations. This is the core of Tectonism: the expressive utilisation of engineering logics. The building must become an information-rich interface. Its users must be able to read it, navigate it, infer its organisation, and intuitively understand its social offerings on the basis of the visual clues broadcast by architectural articulation. Architecture must once more become a legible ordering system for society. Architecture must speak. The ultimate task is semiological: to build up a new, versatile, expressive spatio-visual language for contemporary society. We must design not only spaces but systems of signification. The differences between programmes, atmospheres, degrees of privacy, types of encounter, modes of work etc. must find corresponding spatio-morphological differentiations. These differentiations must be systematic, learnable and intuitively navigable.

The designed environment is not merely shelter. It is a spatial ordering and communication system. It separates and connects, marks and differentiates, invites and inhibits, directs and gathers. It frames social interaction. It tells us where we are, who we are among, what kind of encounter is possible, what level of formality is expected, what kind of communication is solicited. This informs and empowers end-users who can read the social protocols that structure the event they are about to join and participate within. All participants are brought onto the same page, configured into the mutually complementary positions and social roles that allow an event to proceed smoothly and productively, achieving its social purpose. The more complex society becomes, the more articulate this environment must become. The crude neutrality of modernist space is therefore no longer progressive. It is an abdication. Minimalism is not refinement but retreat. Its silence is not profundity but refusal.

Here lies our central conviction, and our central polemic. The built environment is not merely a stack of technical solutions. It is a medium of communication. Every space, every modulated structural system, every differentiated envelope addresses its users, defining the social situation, telling them where they are, what is offered, and how to act, i.e. which protocols of interaction are governing the gathering. Architecture’s task is the ordering and framing of social interaction, and it discharges this task through spatial distribution and connection, as well as – crucially – via the legibility of its spatial forms read as designated social situations. The structural and the environmental morphologies that engineering produces are not obstacles to expression to be hidden or obfuscated. They are the raw material of expression. To deny them, to clad them over, to invent an arbitrary decorative layer on top of them, is both wasteful and would leave the communication with less credibility. The articulatory exploitation of technically motivated form is more economical, and – because the signifier is already an index of real functional motivators – it is a more trustworthy broadcast of information. Tectonic articulation transforms the indexical sign into a symbolic, deliberately communicative one. This is the instrumentality of ornament rightly understood: ornament is not the enemy of performance but a special category of it – communicative performance. Architectural theory had been reflecting on this task of expressive characterization since the 18^th Century, when the notions of ‘character’ and ‘expression’ were taken from the theatre and introduced into architectural theory by Germain Boffrand 1745v Livre d’architecture.

We therefore reject two opposing errors with equal force. We reject the Modernist taboo against ornament, the puritanism that strips the built environment to a mute, isotropic repetition and calls its poverty honesty. An undifferentiated order of sameness offers no navigational or orienting information; it abandons the user to disorientation. And we reject, equally, arbitrary ornamentation, as well as the hypostasis of structural expressionism – the fetishising of tectonic display as an end in itself, severed from the social function it ought to articulate. Tectonism subordinates the eloquence of structure and material articulation to the communication of social purpose. The building does not boast about its loads; it guided the users to their destination, tells them about the social situation they are about to join, and coordinates their interactions. This is empowering, and indeed an increasingly indispensable priming for participants, in order to have productive interaction experiences in an urban world where the panoply of communication situations are bewilderingly diverse.

This ambition and task does not stop at the building scale. It extends, necessarily, to the urban district and city scale. The contemporary metropolis is the engine of the knowledge economy, the dense synergy-cluster within which the productivity of each depends on legible, navigable proximity to all. Modernist urbanism – its zoning, its separation, its serial repetition – cannot serve our current lifes. It fragments what must be woven together and homogenises what must be differentiated. Tectonism answers with a complex, variegated urban order in which built forms register and respond to one another, establishing affiliations and continuities across the whole field. It establishes vectors of transformation that can be navigated intuitively just like a bacterium navigates along a nutrition gradient. Tectonism gives this urban order its articulate physiognomy: districts that read as coherent wholes despite their versatility, landmarks that orient, gradients that guide. Spatial topology in the service of social topology is, at the urban scale, nothing less than the recovery of legibility for cities that Modernism left mute and the laissez-faire aftermath left chaotic. A city that can be read is a city that can be lived.

This is sophisticated, exacting work, and architects cannot any longer rely on precedents. The raw, unedited imposition of pragmatic engineering priorities does not yield legibility – it yields visual chaos, an agglomeration of unreconciled concerns. Order must be imposed through a ruthless formalism: a disciplined regime of accentuation and suppression in which selected morphological features are heightened into a system of signification. The designer must orchestrate, select, edit. Tectonic articulation is the strategic détournement of an element’s technically induced morphology into the service of social meaning. It demands an intense and congenial collaboration between architect and engineer – a collaboration of distinct responsibilities discharged by the very same forms, the engineer guaranteeing their silent physical functioning, the architect orchestrating their eloquent social communication.

The discipline this demands is not improvised; it can be codified, as the author has done in his Manifesto of Parametricism in 2008, by positing both formal and functional heuristics that must be strictly adhered to if the power of parametricism is to be brought to bear. These heuristics are giving a decisive direction to the search for design solutions. However, they operate on a level of abstraction that leaves ample space for creative exploration. Indeed, the search space opened up by these heuristics is much larger than the combined space of all prior architectural styles – from the Renaissance to Modernism - put together. Tectonism is fully encompassed within these heuristic strictures. However, the search space and versatility opened up by tectonism is, again, much larger as all prior substyles pf parametricism – foldism, blobism, swarmism – put together.

Tectonism advances as a design research programme, governed – like every serious paradigm – by a hard core of principles that can be expressed by heuristics, both formal and functional heuristics, both expressed via a series of taboos and dogmas. We hold to its negative heuristics without exception and without nostalgic compromise: no rigid platonic primitives – no squares, cubes, pyramids, spheres – for these forms are inert and insensitive, incapable of inflection or nuanced relation; no simple repetition, that monotony which corresponds to the social homogenisation Modernism imposed; and no collage of unrelated elements delivering the chaos of garbage spill urbanisation. And we pursue its positive heuristics relentlessly: all forms parametrically malleable, all systems differentiated through gradients, all systems correlated into networks of mutual interdependence. These formal principles are paired with their functional counterparts, for function in our paradigm is not a fixed stereotype but a field of variable activity-scenarios in which everything communicates with everything else, instead of being separated via segregative zoning. This is not arbitrary doctrine. It is the operational definition of a style, and only by holding to it without backsliding can the new architecture be put fully to the test. Tectonism inherits this entire apparatus and intensifies it, binding the discipline of articulation to the shaping disciplines of structure, environmental engineering, and versatile, numerically controlled fabrication processes.

The conditions for this achievement are, for the first time in history, fully present. The computational revolution has placed sophisticated structural form-finding, environmental optimisation, and robotic fabrication into the hands of the designer at the earliest stages of conception. Physics engines, generative tools, and material-aware fabrication logics – robotic assembly, 3D printing – have expanded the disciplined universe of buildable form beyond anything available to prior generations. The visual trace of these processes – ‘faktura’ – returns as a positive, character-bearing quality. The result is a morphological and material repertoire that is at once richly diverse and deeply coherent, like the endless yet lawful forms of nature. This is not arbitrary novelty. It is rule-governed abundance.

Tectonism is already a worldwide movement – a collective avant-garde endeavour of many talents across many continents, advancing through research, teaching, and built work, fusing global best practice with regional materials and traditions. It is not the property of any single office. It is a research programme and a shared repertoire, capable of infinite local adaptation precisely because its general principles are abstract and open-ended. No region, culture, or subculture stands outside its reach, because differentiation and adaptation are its very essence.

I do not claim that the work is finished. The fully satisfactory built examples remain, as yet, relatively few and relatively modest in scale. But this is the condition of every style at its dawn, and it is no argument against the thesis – it is a summons. I make a wager on the long-term rationality of the discipline: on its capacity to recognise, through its own internal discourse, the superiority of this paradigm, and on the rationality of a wider society – of clients, of users – to accept the guidance an upgraded architectural expert discourse is able to give. The future of 21st century architecture will be decided not by the timid accumulation of more incompatible options, but by the courage to converge. The avant-garde must stop apologising for ambition. Architecture must again claim the right to lead, to formulate principles, to tease out and fight out the best shared way forward, to construct a world rather than merely accumulate commissions. We must replace the complacent relativism of today’s architectural culture with a new confidence in cumulative progress. We must converge around the only paradigm capable of absorbing contemporary complexity and turning it into articulate order. Tectonism is not a style among styles. It is the style that overcomes the exhaustion of styles as arbitrary appearances by restoring style as a design research programme, as a historical project, as a collective civilisational achievement.

This, then, is my manifesto. Architecture must once again become a unified, progressive, self-confident discipline, equal to the complexity of the civilisation it serves. It must abandon the false comfort of pluralism, reclaim the framing of social communication as its proper task, and harness the full computational intelligence of our age to design a productive, empowering and inspiring built environment that facilitates and further advances the next stage of our global civilisation. Tectonism is the name and programme of that ambition made concrete. I declare it the architecture of the 21st century, and I call upon the discipline to make it so.

Patrik Schumacher, London August 2026

Introduction

South Africa presents one of the clearest contemporary cases for thinking about the fragility of modern economic institutions. The peaceful transition from apartheid to liberal democracy in 1994 was an extraordinary political achievement. It appeared possible that South Africa could combine democratic inclusion with the sophisticated industrial, financial and administrative capacities already present in the country. The first post-apartheid governments did not abolish the market economy. They preserved private ownership, reintegrated South Africa into international markets after sanctions had been lifted and, particularly from the mid-1990s, adopted relatively orthodox economic policies. The post-1994 record therefore cannot be reduced to the proposition that political transition immediately produced economic decline.

South Africa performed comparatively well during much of the period from the mid-1990s to the global financial crisis. Growth accelerated particularly between 2004 and 2007, when real GDP expanded at roughly five per cent a year on average. Fixed investment also rose strongly: the South African Reserve Bank records gross fixed capital formation reaching 23.5 per cent of GDP in 2008. Thereafter, under president Jacob Zuma (2009–2018) the trajectory changed. Growth slowed sharply, investment weakened, unemployment remained extraordinarily high, major state-owned enterprises deteriorated, and infrastructure constraints increasingly became binding on the private economy.

The argument of this paper is that these developments can be interpreted through the concept of institutional regression: a partial movement away from impersonal, rule-bound competition and towards politically mediated allocation, patronage and rent-seeking, accompanied by the consumption of inherited physical and institutional capital. This is not offered as a monocausal explanation. South Africa has also been affected by the global financial crisis, commodity cycles, the pandemic, demographic pressures, the structural legacies of apartheid, educational weakness and global shifts in investment. The claim is narrower and more theoretically interesting: where politically mediated allocation expands, it can alter the selection mechanisms through which firms, public institutions and individuals succeed.

In my 2024 Substack essay, “Thesis of German Right Wing Populism - A Rationale for the Protection of High-performance National Cultures,” I argued that the Western European prosperity take-off was a historically improbable achievement dependent on an intricate constellation of technological, organisational, institutional and cultural preconditions (Schumacher 2024). South Africa suggests an important corollary: such achievements are not only difficult to create and transplant; they are reversible. A society may retain the outward architecture of modernity while weakening the mechanisms required for its reproduction.

1. The Historical Anomaly of Impersonal Competition

Modern market society is easily mistaken for a natural condition. Historically it is anything but. For most of human history, economic opportunity was embedded in status, kinship, clan, inherited privilege, personal loyalty and political patronage. Max Weber’s contrast between patrimonial and rational-legal authority captures one dimension of the transformation. Douglass North’s institutional economics captures another: modern growth depends on credible rules that shape expectations and reduce the importance of arbitrary personal discretion.

North, Wallis and Weingast provide an especially useful framework. Their distinction between limited-access orders and open-access orders turns attention away from the mere existence of markets or formal constitutions and towards the rules governing entry into organisations, economic competition and political power (North, Wallis and Weingast 2009). Open access is historically rare. It depends on impersonal rules and on the ability of individuals to form organisations without requiring discretionary permission from powerful patrons.

The rule of law and competitive markets – and I should add open discourses (Schumacher 2025) - are therefore mutually reinforcing components of a single institutional achievement: the replacement of personalised allocation by impersonal selection. Hayek’s account of competition as a discovery procedure identifies the informational advantage of such a system. Firms are tested by customers, prices and rival producers; capital can move away from persistently unsuccessful uses; and no political authority has to know in advance which entrepreneur, technology or organisation should succeed (Hayek 1945).

Patron-client systems select according to a different criterion. As politically allocated opportunities become more important, economic actors rationally invest more heavily in relationships with those who control access. The relevant question becomes not only what one can produce, but whom one knows, which classification one must satisfy, which official controls a licence or contract, and which political coalition can protect one’s position. Rent-seeking theory explains why the creation of politically allocable rents can divert resources away from productive entrepreneurship towards competition for privilege (Tullock 1967; Krueger 1974).

2. South Africa’s Liberal Interlude and the Post-2008 Break

The post-apartheid economy should be divided analytically into phases. The early ANC governments inherited severe inequalities and institutional distortions, but also a developed industrial base, deep capital markets and sophisticated professional sectors. The 1996 Growth, Employment and Redistribution programme marked a clear move towards fiscal restraint, openness and macroeconomic stability. This period did not eliminate unemployment or inequality, but it does demonstrate that ANC rule and relatively liberal economic policy were not initially incompatible.

The strongest growth phase came in the 2000s. Mahadea (2010) reports average real GDP growth of about five per cent during 2004-2007. Investment rose at the same time. By 2008, gross fixed capital formation had reached 23.5 per cent of GDP, compared with 15.2 per cent in 2002. The subsequent decline is striking: by 2024, gross fixed capital formation had fallen to 14.5 per cent of GDP (SARB 2025a). This is not a minor fluctuation but a large reduction in the share of national output devoted to building future productive capacity.

The growth record similarly deteriorated, with real GDP per capita falling. The IMF describes South Africa’s growth as having stagnated over the preceding decade and identifies governance, business regulation, electricity, logistics and labour-market rigidities among the structural constraints (IMF 2025a; IMF 2025b). The labour market is an especially severe indicator of failure to generate opportunities at scale. Statistics South Africa recorded an official unemployment rate of 31.4 per cent in the fourth quarter of 2025 and 32.7 per cent in the first quarter of 2026, corresponding to more than eight million unemployed people in Q1 2026. One should be cautious about comparing current official unemployment directly with older expanded measures, but the present magnitude is indisputably exceptional.

3. From Productive to Political Entrepreneurship

The political demand for economic transformation after apartheid was unavoidable and legitimate. A system that deliberately excluded the majority from ownership, skilled occupations and economic mobility could obviously not be maintained by the new government standing opposed to these inherited inequalities. The decisive institutional question is therefore not whether transformation was necessary, but how transformation was pursued.

Black Economic Empowerment, and later Broad-Based Black Economic Empowerment, increasingly made racial transformation criteria relevant to ownership, management, procurement and relations between firms and government. The academic literature is contested and should be represented as such. Southall (2007) treats BEE as an important element of the post-apartheid political settlement while documenting concerns about elite enrichment. Tangri and Southall (2008) argue that politically connected individuals disproportionately benefited from empowerment transactions and that the policy had done less for the mass of disadvantaged South Africans.

The libertarian critique developed here is institutional. Any system - regardless of the identity categories it employs - can strengthen patronage if political authorities acquire broad discretionary power over valuable economic opportunities. When procurement, licences, financing, appointments and regulatory advantages become contingent on politically administered criteria, economic actors have incentives to invest in political access. This need not mean that every beneficiary is corrupt or that every empowerment policy is illegitimate. It means that the mechanism creates additional margins on which political connection can acquire economic value.

The distinction between productive entrepreneurship and political entrepreneurship is therefore central. In an impersonal competitive order, an entrepreneur’s principal problem is to satisfy customers better than rivals. In a politically mediated order, an increasing share of entrepreneurial energy is devoted to satisfying gatekeepers. The social return on the second activity can be negative even when the private return is high.

4. Capital Consumption: Physical, Human and Institutional

An advanced economy lives partly from accumulated capital. The most visible form is physical: power stations, transmission lines, railways, ports, roads, water systems and machinery. But there is also human capital and institutional capital. A functioning utility embodies maintenance routines, engineering knowledge, procurement competence, internal controls, professional norms and networks of experienced personnel. None of these can be recreated instantly.

This creates a deceptive lag between institutional deterioration and visible collapse. A power station can continue producing after maintenance discipline weakens. A railway can continue moving freight while signalling systems and rolling stock deteriorate. A municipality can continue delivering water while engineering departments lose experienced staff. The inherited capital stock masks the deterioration until failures become nonlinear and cumulative.

South Africa’s investment data are consistent with this concern. The decline in gross fixed capital formation from 23.5 per cent of GDP in 2008 to 14.5 per cent in 2024 means that less of current output is being converted into future productive capacity. This figure alone cannot prove ‘capital consumption’ in the Austrian sense, but it strongly supports the narrower proposition that the economy’s investment effort has weakened substantially. Also, even this much reduced figure might in part represent the mere premature replacement of capital stock due to theft and vandalism, rather than real capital formation. The problems at Eskom - South Africa’s state-owned electricity company – are perhaps the clearest case of capital deterioration on post-apartheid South Africa. For decades South Africa possessed abundant and inexpensive electricity, based largely on on large coal-fired power stations constructed before 1994. From the late 2000s onward, inadequate investment and maintenance, management failures, corruption, procurement problems, and rapidly mounting debt contributed to declining reliability. Eskom can be regarded as institutional microcosm highlighting more general problems in South Africa’s state sector.

Eskom is the most obvious and dramatic illustration of capital consumption and capacity degeneration. It is particularly obvious in its devastation because electricity is foundational infrastructure for the whole economy. “Load-shedding”, scheduled electricity blackouts used to prevent collapse of the national grid, began in 2007-08 and ultimately became chronic. The Centre for Renewable and Sustainable Energy Studies at Stellenbosch University, using Eskom and regulator data, estimates that South Africa experienced 6,838 hours of national load-shedding in 2023 - roughly 78 per cent of the hours in the year. Eskom reported load-shedding on 329 days during its 2023/24 financial year.

The financial burden also became extraordinary. In the 2023 Budget, National Treasury announced an Eskom debt-relief package of R254 billion, explicitly recognising that the utility’s debt problem had become a sovereign fiscal issue (National Treasury 2023). The scale of the intervention is important to the paper’s thesis: monopoly failure can be sustained for long periods because losses are socialised through taxation, sovereign borrowing and state guarantees rather than immediately disciplining the organisation through exit or bankruptcy.

The economic cost radiates outward. Firms and households respond to unreliable electricity by purchasing generators, inverters, batteries and solar panels. These investments are rational from the perspective of the individual actor, but much of the capital is defensive: it duplicates infrastructure that the electricity system was supposed to supply. The same rand cannot simultaneously finance backup power and an additional production line.

Literal Capital Consumption: The Theft and Destruction of Infrastructure

In South Africa the concept of capital consumption is not only an abstract economic concept. One of its most striking manifestations is the literal physical destruction and removal of accumulated public capital. Electricity cables, transformers, substations, signalling equipment and railway infrastructure are repeatedly vandalised or stolen, often for the scrap value of their copper and other components. Eskom continues to identify theft and vandalism of transformers, mini-substations, pylons and electrical cables as a serious threat to electricity supply. The rail system has suffered the same phenomenon on an extraordinary scale. Transnet reported that the length of cable stolen annually from its network increased from approximately 120 kilometres in 2017 to 724 kilometres in 2021, while the number of incidents rose from fewer than 2,000 to nearly 4,500.

This represents capital consumption in its most literal form. Infrastructure accumulated through decades of investment is physically dismantled and converted into scrap. The private value recovered by the thief may represent only a minute fraction of the social value destroyed. A relatively small quantity of stolen copper can disable signalling, immobilise trains or interrupt electricity supply across an entire neighbourhood. Society subsequently has to devote new investment merely to restoring the productive capacity it already possessed.

Gross investment figures consequently understate the severity of the problem. If a state spends one billion rand replacing infrastructure that has been stolen or vandalised, the expenditure appears as new investment even though the society may merely have reconstructed yesterday’s capital stock. Resources that could have expanded productive capacity are instead consumed reproducing what already existed.

The process also generates a second-order cost. Infrastructure that cannot safely be left unguarded requires increasingly expensive protection. Security expenditure rises, repairs must be hardened against further theft, businesses construct alternative systems, and transport users shift from rail towards roads. Capital is therefore diverted simultaneously into replacing destroyed assets and defending the remaining stock.

This phenomenon reveals an elementary but often overlooked institutional precondition of advanced civilisation: the physical capital stock must be sufficiently secure that society can accumulate rather than repeatedly reconstruct it. Modern infrastructure presupposes a functioning system of property protection even when that property belongs to the state. Railway lines, electrical substations, telecommunications cables and water infrastructure extend across enormous territories and cannot economically be defended meter by meter. Their viability therefore depends upon a background social and legal order in which systematic predation is exceptional.

When this condition deteriorates, the economics of infrastructure itself changes. The railway becomes not merely a transport system but an immense inventory of exposed copper. The electricity distribution network becomes a dispersed stock of transformers, cables and metals available for extraction. An infrastructure network constructed to increase the productivity of society is partially transformed into a resource deposit to be mined by predatory actors.

This is institutional regression in an unusually literal sense. A modern capital-intensive system is being subjected to an older extractive logic in which accumulated productive assets are consumed for immediate appropriation. The contrast with productive entrepreneurship could hardly be sharper. The productive entrepreneur combines resources to create an asset whose future value exceeds its present inputs. The infrastructure thief performs the reverse operation: he destroys an asset with high future productive value in order to realise a small immediate scrap value. One creates capital. The other liquidates it.

Where theft reaches sufficient scale, economic development therefore runs backwards not only because insufficient new capital is being created but because yesterday’s capital is being physically dismantled faster than society can replace it. The South African rail and electricity networks thus provide perhaps the most tangible illustration of the wider thesis of this paper: civilisation is not simply inherited. Its accumulated capital has to be protected, maintained and reproduced. Once the institutional order can no longer reliably perform these elementary functions, the capital stock itself begins to disappear.

5. Cadre Deployment, State Capture and the Selection Problem

The same logic applies inside the state. Complex institutions depend on selection mechanisms that reward competence, preserve tacit knowledge and sustain professional norms. If political loyalty becomes an important appointment criterion, the selection environment changes. The ANC’s practice of cadre deployment is therefore significant not merely as partisan behaviour but as an institutional mechanism. Scholarship has associated cadre deployment with political interference, weakened accountability and governance problems, while recognising that public-sector failure has multiple causes (Mlambo, Zubane and Thusi 2022).

The state-capture period under Jacob Zuma intensified this problem. South African scholarship and official inquiries describe a process in which networks linked to political power sought influence over appointments, procurement and state-owned enterprises. Dassah (2018) analyses state capture as a governance problem in South Africa, while Mathebula and Masiya (2022) connect governance failure and capture to the performance of Eskom and Transnet. The Zondo Commission’s findings provide the most extensive official record of the phenomenon.

The relevance to economic performance lies not only in money stolen. Patronage changes what institutions select for. A technically complex organisation can survive political interference for a time because it inherits engineers, routines, plants, inventories and professional cultures built earlier. But if appointments, procurement and promotion systematically cease to reward competence and integrity, institutional capital is slowly consumed.

6. BEE Policies: From Economic Empowerment to Political Allocation

The evolution of Black Economic Empowerment provides perhaps the clearest illustration of the institutional mechanism explored in this paper. The objective of BEE was understandable and, after apartheid, politically unavoidable: to broaden black participation in an economy whose ownership, management and professional strata remained disproportionately white. The crucial question, however, is not the legitimacy of this objective but the institutional mechanism chosen to pursue it.

The distinctive achievement of the modern competitive order is the progressive replacement of status-based allocation by impersonal selection under general rules. In this respect BEE contains a paradox. Apartheid represented an extreme system of politically imposed racial status, excluding individuals from occupations, property and economic opportunities according to racial classification. The post-apartheid state dismantled this system but, in attempting to overcome its inherited consequences, retained racial status as an instrument of economic allocation, now operating in the opposite direction and for radically different purposes.

The original Black Economic Empowerment programme subsequently developed into the more comprehensive system of Broad-Based Black Economic Empowerment (B-BBEE), formalised principally through the Broad-Based Black Economic Empowerment Act of 2003 and subsequently strengthened and amended. Rather than functioning simply as a racial ownership quota, B-BBEE developed an elaborate scorecard through which companies are assessed according to such criteria as black ownership, management control, skills development, enterprise and supplier development and socioeconomic development. These ratings become particularly consequential for firms seeking government procurement, licences, concessions or business relationships with other firms whose own B-BBEE standing is affected by their procurement decisions.

B-BBEE therefore reaches beyond the direct relationship between the state and the individual firm. The regulatory incentives propagate through supply chains. A company seeking to improve its own rating has reason to prefer appropriately rated suppliers, which in turn have incentives to restructure ownership, management, procurement and training arrangements in accordance with politically determined transformation criteria. Political objectives thereby become embedded within private economic selection.

This does not make corruption inevitable. Nor should BEE itself simply be equated with corruption. The more important institutional argument is that the system increased the number and value of economic opportunities whose allocation depended upon politically defined classifications, regulatory compliance and administrative discretion.

The consequences were visible relatively early. Tangri and Southall’s study of BEE concluded in 2008 that the programme had generated controversy partly because its benefits had accrued disproportionately to politically connected individuals rather than to the mass of previously disadvantaged South Africans. Southall had already identified tensions between the broad developmental aspirations of empowerment and the emergence of a relatively narrow black capitalist elite.

This outcome was not entirely accidental to the institutional structure. Early empowerment transactions often required established corporations to transfer significant equity stakes to black investors who did not possess sufficient capital to purchase those stakes conventionally. Financing structures consequently became complex, while politically influential individuals could become particularly attractive commercial partners. Political connectivity itself could acquire economic value.

An emblematic new figure consequently entered South African political economy: the “tenderpreneur”— an entrepreneur whose competitive advantage derives substantially from access to government procurement and political networks. The term captures the distinction central to the argument of this paper. Productive entrepreneurship seeks profit by discovering what customers want and supplying it more effectively than competitors. Political entrepreneurship seeks opportunities generated or allocated by government.

Preferential procurement substantially increases the potential domain of the latter. Government procurement no longer evaluates potential suppliers exclusively according to price, quality, competence and reliability but simultaneously pursues politically prescribed transformation objectives. Whatever the legitimacy of those objectives, multiplying the criteria governing procurement necessarily creates a more complex administrative selection process and potentially increases official discretion.

The problem becomes considerably more serious when such discretion intersects with party-political patronage. The interaction between BEE procurement requirements, cadre deployment and ANC patronage networks therefore deserves particular attention. Political power influences appointments; politically appointed officials exercise discretion over procurement; procurement creates opportunities for politically connected businesses; successful beneficiaries acquire wealth and influence; and these resources can in turn reinforce the political networks through which the opportunities were originally allocated.

The potential cycle is therefore: political power → appointments → procurement discretion → politically connected businesses → private enrichment → political influence → political power. Patronage thereby acquires the possibility of becoming self-reproducing.

The phenomenon of B-BBEE “fronting” provides another indication of the distortions generated by the system. Businesses have sometimes constructed nominal black ownership, management or partnership arrangements in order to obtain empowerment credentials without transferring the substantive economic benefits that the legislation intended. The B-BBEE Commission itself recognises fronting, falsification and misrepresentation of empowerment credentials as enforcement problems.

Fronting is particularly revealing from an institutional perspective because it represents entrepreneurial adaptation to regulation. Once regulatory classifications acquire substantial economic value, entrepreneurs invest resources in satisfying, circumventing or manipulating those classifications. Human ingenuity has not disappeared; its direction has changed. This is the central economic problem. Corruption is usually conceived too narrowly as theft: public money disappears, contracts are inflated, officials receive bribes or politically connected individuals acquire unjustified wealth. These are serious losses, but they may be smaller than the dynamic losses generated by the corruption of the selection mechanism itself.

Once political access affects the expected return on economic activity, rational entrepreneurs invest in political access. Resources that under an impersonal competitive order might have been invested in products, technologies, skills, productivity improvements and cost reductions can instead be invested in relationships, compliance structures, politically advantageous partnerships and access to public procurement.

The entrepreneur increasingly confronts two possible routes to success: satisfying customers within an impersonal market or acquiring advantageous access to politically allocated opportunities. To the extent that the latter becomes more profitable, society selects for political rather than productive entrepreneurship.

This distinction allows the effects of BEE to be incorporated into the wider argument about institutional regression without making the simplistic claim that BEE “caused” South Africa’s decline. State capture, cadre deployment, corruption, deteriorating state-owned enterprises, labour-market regulation, educational failures, crime and other factors possess independent causal significance. BEE must be understood as one component within a larger institutional ecology.

Nevertheless, BEE is theoretically important because it illustrates a fundamental tension between status and contract. Apartheid itself represented an extreme system of status-based economic allocation. Race determined where individuals could live, which occupations they could enter, what property they could own and which economic opportunities were available to them. One possible post-apartheid trajectory would have been the complete replacement of this racialised allocation system by universal individual rights and unrestricted competition under general rules.

South Africa chose a different path. Racial restrictions against black South Africans were abolished, but racial classification was retained as an instrument for engineering the distribution of economic opportunities in favour of previously disadvantaged groups. There is no moral equivalence between these systems. Apartheid was a comprehensive system of racial domination and exclusion, whereas BEE is explicitly conceived as a remedial programme intended to overcome the consequences of that injustice. The institutional comparison is narrower but nevertheless significant: both depart from the principle that individuals should compete for economic opportunities without racial classification entering the selection process.

The post-apartheid state therefore confronted the consequences of an historical system of status by creating a new, remedial system of status rather than completing the transition from status to contract. This decision has consequences extending beyond racial preference itself because political allocation creates constituencies around its continuation. Individuals and businesses possessing valuable BEE positions, intermediary functions, procurement relationships and regulatory expertise acquire an economic interest in maintaining the regulatory system through which those advantages are generated. A policy initially justified as a transitional mechanism can thereby develop its own beneficiaries and institutional permanence.

South Africa’s deterioration in international measures of its business environment is consistent with, although it cannot by itself establish, this interpretation. In the World Bank’s former Doing Business rankings, South Africa fell from roughly the mid-thirties globally in the late 2000s to 84th of 190 economies in the final 2020 edition. The Heritage Foundation’s Index of Economic Freedom likewise records a deterioration from a score of 62.5 and 75th place in 2014 to 58.6 and 100th place in 2026.

Neither measure establishes BEE as the cause of this deterioration. The methodologies also change over time and rankings depend upon the performance and number of other countries. They should therefore be treated as corroborating indicators rather than causal evidence. Their direction is nevertheless significant when considered alongside declining fixed investment, infrastructure deterioration, state capture and the expansion of regulatory burdens. South Africa has not abolished capitalism. Private property, corporations, financial markets and entrepreneurship continue to exist, and important parts of its private economy remain highly sophisticated. This is precisely what makes the South African case theoretically interesting.

Institutional regression does not require the abolition of markets. A market economy can remain formally intact while an expanding layer of politically mediated allocation develops around it. Entrepreneurs continue competing, but an increasing proportion of entrepreneurial calculation becomes concerned with navigating political classifications, regulatory requirements and relationships with the state.

The distinction is therefore not simply between capitalism and socialism. It is between two different mechanisms operating within an ostensibly capitalist economy: impersonal market competition and personalised or politically mediated access. The danger arises when the second mechanism begins progressively to colonise the first. Seen from this perspective, the corruption associated with parts of the BEE system represents something more consequential than individual wrongdoing. It contributes to an institutional regression from a society in which economic selection is predominantly governed by performance under general rules towards one in which status, political access and administrative discretion once again become economically decisive.

The historical achievement of competitive capitalism under the rule of law consisted precisely in escaping this older patron-client logic. South Africa demonstrates how easily elements of that logic can re-emerge within the institutional shell of a modern market economy.

7. Why Did Apartheid South Africa Maintain a More Capable Economy and State Administration?

Why was the apartheid system—despite its racial discrimination, coercion, and policies fundamentally alien to a liberal order—able to sustain a more capable economy and state administration than the later post-apartheid regime?

The question is uncomfortable, but analytically important. It does not imply any moral equivalence between apartheid and the democratic order that replaced it. Apartheid was an unjust system of racial domination that denied fundamental political, economic, and civic rights to the majority of South Africans. The issue here is narrower: how could a deeply illiberal political order nevertheless preserve comparatively effective administration, infrastructure, and economic coordination, while a formally inclusive and democratic successor state subsequently suffered severe institutional deterioration? One possible answer lies in the distinction between exclusion from an institutional order and politicisation of the institutional order itself.

Apartheid imposed a rigid racial boundary around economic and political participation. The black majority faced restrictions on residence, movement, employment, property holding, education, political representation, and access to public services. In this respect, apartheid profoundly violated the principles of equal citizenship and open competition. Yet within much of the privileged institutional sphere reserved for whites, economic and administrative life could still operate according to relatively impersonal rules. Firms competed. Engineers were appointed because they possessed engineering expertise. Utilities were expected to maintain infrastructure. Public officials worked within bureaucratic hierarchies. Commercial relations were not generally mediated through an elaborate system of racial scoring affecting ownership, management, procurement, subcontracting, and supplier relationships at every level of economic activity.

The apartheid economy was therefore not liberal in the universal sense. But significant parts of its internal administrative and commercial order retained features normally associated with modern capitalist institutions: technical competence, professional hierarchy, relatively predictable rules, property rights, contractual enforcement, and bureaucratic continuity. This distinction helps explain an apparent paradox. A state may be profoundly unjust in determining who is admitted to an institutional order, while still maintaining relatively high levels of competence in the operation of that order among those admitted to it.

The post-apartheid settlement correctly abolished the racial caste system and established universal citizenship. This was an enormous liberal advance. The historical opportunity after 1994 was therefore to dismantle the racial boundary while extending the strongest elements of the inherited institutional order to the population as a whole.

In principle, this could have meant universalising equal legal rights while preserving and enlarging an impersonal system of professional administration, open competition, property rights, and technical competence. Instead, however, a different process gradually emerged. Increasingly, racial and political criteria entered directly into the internal mechanisms by which appointments, ownership, procurement, contracting, and advancement were determined.

8. Culture and the Fragility of the Inherited Order

We have raised the question: Why could an Illiberal Apartheid State nevertheless sustain a comparatively more capable Modern economy compared with the post-apartheid society?

Part of the answer developed in this paper concerns incentives: corruption, cadre deployment, B-BBEE, procurement preferences and the emergence of a political patronage economy. But incentives operate within cultures. Institutions cannot be reduced to legislation, organizational charts and physical assets. They also depend upon habits, expectations and behavioural norms that have accumulated historically.

The argument advanced here is therefore that South Africa inherited in 1994 not merely a stock of mines, factories, power stations, railways, universities, courts and administrative organizations, but a stock of institutional-cultural capital. This included professional disciplines, bureaucratic routines, technical knowledge, expectations concerning competence and responsibility, commercial practices, and norms concerning the distinction between public office and private interest. Such capital is much harder to transfer.

The cultural histories of the populations that came together within modern South Africa were significantly different. The Dutch and later British settlers arrived from societies that had already travelled a considerable distance along the historical transformation from kinship-based and patrimonial organization towards impersonal law, markets, corporations and professional administration. The indigenous African populations possessed their own complex and viable social institutions, but these had historically been organized to a greater extent around kinship, customary authority, lineage and local community. South Africa’s own post-apartheid government, in describing the historical institution of traditional leadership, has acknowledged the importance of tribal structures, patriarchy, ascriptive status and traditional authority in precolonial social organization. This distinction has nothing to do with innate racial capacities. It concerns different trajectories of institutional evolution.

What the Dutch Brought to the Cape:

When Dutch settlement began at the Cape in the seventeenth century, its settlers did not arrive institutionally empty-handed. They came from one of the most commercially advanced societies of early modern Europe. The Netherlands had developed sophisticated markets, private property, maritime commerce, banking and contractual institutions. Dutch commercial society routinely coordinated economic activity among strangers rather than exclusively among relatives. The corporation, commercial contract, credit relationship and legally transferable property had already greatly expanded the sphere of impersonal cooperation. This matters profoundly.

A modern market order is remarkable precisely because it permits individuals who have no family relationship, tribal affiliation or personal allegiance to cooperate. A merchant does not need to know the family of every customer. A lender does not need to belong to the borrower’s clan. A business owner can employ a stranger. Property can be bought from someone outside one’s community. Disputes can be submitted to courts according to general legal rules rather than negotiated exclusively through kinship groups.

The Dutch also brought Roman-Dutch law, which became one of the foundations of South African private law. Thus from the beginning European settlement carried with it a conception of property, contract and legally constituted economic relationships capable of extending beyond personal networks.

The Afrikaner society that subsequently emerged was not simply a transplanted Dutch society. Frontier conditions, farming, geographical dispersion, conflict and interaction with African societies produced a distinctive culture. Afrikaners developed particularly powerful forms of internal solidarity, reinforced by the Dutch Reformed Church and eventually by Afrikaner nationalism.

This is an important qualification because Afrikaners themselves were certainly capable of particularism. During the twentieth century, Afrikaner nationalism created dense political, cultural and economic networks. Party allegiance, ethnic solidarity and organized efforts to advance Afrikaners sometimes competed directly against universalistic meritocratic principles. The history therefore cannot be reduced to Europeans practising pure meritocracy while Africans practised patronage. The relevant distinction is one of degree, institutional development and competing principles.

The British Contribution: Administration, Professions and Commercial Modernity

British rule added another institutional layer. Britain brought administrative practices associated with an increasingly professional modern state as well as commercial institutions connected to what had become the world’s leading industrial economy. During the nineteenth and twentieth centuries, South Africa became integrated into British systems of finance, mining, engineering, law, accounting and international commerce.

The mineral revolution intensified this transformation. Modern deep-level mining could not be organized primarily through kinship relationships. It required extraordinarily complex organizations employing engineers, geologists, accountants, financiers, lawyers, technicians and managers. Railways had to run according to timetables. Mines had to satisfy engineering constraints. Banks had to maintain accounts. Electricity systems had to balance physical loads. Industrial capitalism therefore imposed a powerful discipline of its own: technical reality is indifferent to political identity. A turbine works or it does not. A railway delivers freight or it does not. An engineer understands the system or does not. A balance sheet eventually reconciles or exposes losses. The culture surrounding these institutions consequently attached authority increasingly to specialized knowledge, qualifications, organizational hierarchy and measurable performance. This was the institutional-cultural environment within which twentieth-century South Africa accumulated much of its physical capital.

From Kinship to the Impersonal Order

The deeper historical distinction can be expressed as the movement from particularistic to impersonal social organization. In a kinship-intensive order, the individual’s most reliable relationships are ordinarily with family, extended kin, lineage and local community. Obligations are reciprocal and personal. Assistance is provided because of who someone is in relation to oneself. Such arrangements are neither irrational nor uniquely African. They constituted the normal form of human social organization for most of history. European societies themselves were once organized much more extensively through kinship, dynasty, religious affiliation, patronage and personal loyalty. The emergence of modernity required a gradual expansion of a radically different principle: cooperation among strangers according to general rules.

Max Weber captured one dimension of this transformation in his distinction between patrimonial authority and rational-legal bureaucracy. In a patrimonial order, political authority and personal relationships overlap. Positions can become resources belonging effectively to those who control them. Offices generate opportunities for dependants, relatives and clients. Loyalty travels through persons. In an impersonal bureaucracy, by contrast, the office is theoretically independent of the person occupying it. The minister does not own the ministry. The procurement officer does not legitimately distribute contracts among relatives. The successful operation of such a system depends upon more than formal regulations. It requires individuals to have internalized the legitimacy of these distinctions.

Indigenous African Institutions:

Precolonial black South African societies developed sophisticated institutions suited to their own social and economic circumstances, but their organizing principles differed substantially from those of industrial north-western Europe. Kinship, lineage, extended family, customary law and traditional leadership occupied important positions. Political authority was frequently embedded in relationships between chiefs, communities, families and councils rather than separated into the impersonal organizations characteristic of the modern bureaucratic state. Traditional authority continues to occupy a constitutionally recognized position in contemporary South Africa, illustrating the persistence as well as adaptation of these institutions.

It would be misleading, however, to imagine a single unchanging “black culture.” Zulu, Xhosa, Sotho, Tswana, Venda and other societies possessed different histories and institutions. Christianity, urbanization, migrant labour, industrial employment and education transformed them dramatically during the nineteenth and twentieth centuries.

Nor should modern ANC patronage simply be labelled a survival of tribal society. Contemporary patronage is created partly by modern political incentives: control over public appointments, procurement budgets and state-owned companies can produce patron-client networks in almost any culture. Nevertheless, older social structures can matter because institutional modernization involves changing the radius of social obligation—from obligations predominantly structured through known persons towards systems in which strangers are routinely treated according to abstract rules.

That transition took centuries in Europe. There was never a plausible reason to assume that it could simply be bypassed elsewhere by writing a constitution.

At this point, however, the argument turns back upon apartheid itself. The apartheid state could point to highly competent institutions dominated by whites, but apartheid systematically obstructed the process through which black South Africans might have accumulated equivalent institutional experience.

From Inclusion to Replacement:

The liberal solution to the injustice inherited in 1994 would have been universalization. Remove every racial barrier. Open every occupation. Permit unrestricted upward mobility. And above all, preserve institutional knowledge by ensuring continuity between experienced incumbents and the new generation entering the organizations. Such a process would have transformed the racial composition of South African institutions while attempting to preserve their institutional capital.

Instead, transformation increasingly became about changing the institurtions’ demographic composition according to politically determined targets. Once demographic representation becomes an objective independent of the available distribution of skills and experience, the speed of personnel replacement can exceed the speed at which institutional competence can be reproduced.

The danger becomes still greater when demographic transformation combines with cadre deployment. Political loyalty now enters alongside racial identity as a criterion affecting appointment. Research on contemporary South African public administration identifies an enduring conflict between cadre deployment and meritocratic professionalization.

The Return of the Patron-Client Principle:

This provides one possible cultural dimension of South Africa’s regression. The transformation should not be understood crudely as a return of “tribalism.” Something more complex occurred. Traditional particularistic social obligations encountered a modern state controlling enormous resources. The result could be a modern patron-client system. The political party controls appointments. Officials control contracts. Businesspeople cultivate political relationships. Those obtaining positions acquire obligations toward the networks that helped them acquire those positions.

9. Apartheid, White Labour Protectionism and the Market

There is a further historical irony that complicates the commonplace identification of apartheid with capitalist interests. In important respects, particularly in the labour market, apartheid represented not the unrestricted operation of capitalism but its political suppression in favour of racially protected white labour.

The roots of this arrangement predated the formal establishment of apartheid in 1948. White workers had long feared competition from black workers who could perform jobs at lower wages, and organised white labour sought political protection against precisely the labour-market substitution that an unrestricted capitalist economy would have encouraged. The alliance between Afrikaner nationalism and sections of organised white labour was already visible in the National Party–Labour Party Pact government of 1924, and the broader tradition of what historians have termed “white labourism” combined working-class economic protection with racial exclusion.

The industrial colour bar and subsequent job-reservation system can therefore be understood partly as a form of racial labour protectionism. Certain occupations were reserved for whites, wage structures protected white workers from lower-paid competition, and black occupational advancement was administratively constrained. These arrangements protected the economic position of white workers precisely by preventing employers from freely selecting workers according to productivity and cost. This produced an important conflict between political apartheid and the logic of the market.

Private employers frequently had a commercial incentive to circumvent, relax, or oppose racial job restrictions. Where a capable black worker could perform a task that legislation, union agreements, or established practice reserved for a more expensive white employee, the employer’s profit motive pointed toward racial integration. Economic historians have shown that the actual boundaries of job reservation were repeatedly contested between employers and white trade unions, and that shortages of suitably skilled white workers contributed to the gradual erosion of occupational restrictions.

This is particularly revealing from a libertarian perspective. Racial segregation was not continuously reproduced by voluntary market exchange. On the contrary, market exchange repeatedly generated pressures against segregation. Employers wanted access to a larger labour pool; firms wanted to employ workers capable of performing the required tasks; black workers sought access to better occupations; and mutually advantageous transactions therefore tended to cross the racial boundaries imposed by political institutions. The state had to prevent these transactions.

In this respect, apartheid illustrates a general proposition associated with classical-liberal economics: discrimination can certainly exist in markets, but systematic exclusion becomes much easier to sustain when the state prevents competitors from profiting by disregarding it. An employer who refuses to employ a productive worker because of race bears a competitive cost if another employer is free to hire that worker. A statutory colour bar removes this competitive mechanism by prohibiting or restricting precisely the transaction that would otherwise undermine discrimination.

The contradiction became increasingly severe as the South African economy industrialised. Modern industry required skilled and semi-skilled labour faster than the white population could supply it. Employers therefore faced growing economic pressure to train, promote, and employ black workers in occupations from which racial regulation sought to exclude them. Research on apartheid labour markets finds that shortages and changing skill patterns contributed materially to the relaxation of job reservation.

There is consequently a striking inversion of the familiar picture of apartheid as the straightforward political expression of capitalist interests. Parts of South African capital undoubtedly benefited from apartheid, not least from the availability and political control of low-paid black labour, and major businesses frequently accommodated themselves to the regime. The claim should therefore not be exaggerated into a general opposition between “capitalism” and “apartheid.”

But within the labour market there was nonetheless a recurrent structural conflict. White organised labour had an interest in restricting competition; employers had an interest in expanding the pool of usable labour. The apartheid state frequently intervened on the side of the former by enforcing racial occupational boundaries.

This helps explain another aspect of the institutional paradox examined in this chapter. Apartheid’s racial system represented a politically imposed distortion of an economy that otherwise contained strong market institutions. The discriminatory boundary continually required reinforcement because ordinary economic incentives tended to erode it.

This differs in an interesting way from the later B-BBEE system. Under apartheid, a profit-seeking employer might find himself wanting to violate racial regulation in order to employ the worker he considered most productive. Under B-BBEE and related procurement regimes, by contrast, government regulation can make racial classification itself economically valuable to the firm. The entrepreneur therefore has an incentive not merely to endure the political classification but to incorporate it into corporate strategy—through ownership structures, procurement relationships, partnerships, and access to state contracts.

The comparison is not morally symmetrical. Apartheid denied opportunities to the majority in order to protect a racial minority, whereas B-BBEE is intended to remedy the inherited consequences of that exclusion. The institutional comparison concerns something different: the relation between political racial classification and market incentives.

Under the apartheid colour bar, market incentives frequently pulled against racial regulation.

Under a sufficiently pervasive preference and procurement regime, market actors can instead be induced to compete through the racial-regulatory system itself.

That distinction reinforces the broader argument of this chapter. Apartheid erected an extraordinarily unjust political barrier around large portions of the economy, but the competitive mechanism operating behind and against that barrier retained a degree of autonomy. The later system risks making political classification endogenous to economic competition itself.

The libertarian lesson is therefore especially revealing: one of the forces steadily undermining apartheid’s labour-market segregation was precisely the capitalist search for productive labour. The employer who wanted to hire the capable black worker was not necessarily motivated by egalitarian conviction. He did not need to be. The price system supplied its own incentive to disregard a politically imposed racial boundary.

10. From Ruling Elite to Refugees: Trump and the Afrikaner Controversy

An extraordinary measure of South Africa’s changed international standing came in February 2025, when President Donald Trump issued an executive order directing the United States government to facilitate the resettlement of Afrikaners claiming racial discrimination in South Africa. The historical reversal was striking. Thirty years after the end of apartheid, members of the population that had constituted South Africa’s politically dominant racial minority were being designated by an American administration as potential refugees from racial discrimination under the successor regime.

The episode should be approached cautiously. Some of the rhetoric surrounding it—particularly claims that South Africa is experiencing a state-sponsored “white genocide” or a systematic programme of confiscating white-owned farms—goes considerably beyond what the available evidence establishes. Nevertheless, Trump’s intervention did not arise from nothing. It brought together several genuine and long-running controversies within post-apartheid South Africa: racially differentiated economic policies, land reform and property rights, violent attacks on farmers, inflammatory racial rhetoric, and a broader deterioration in relations between Pretoria and Washington.

The Unresolved Land Question: Land ownership was inevitably one of the most difficult questions inherited from apartheid. The 1994 constitutional settlement simultaneously protected property rights and committed the new political order to land restitution and reform. Progress proved much slower than many ANC supporters expected, while agricultural land remained disproportionately white-owned. The resulting frustration created increasing political pressure for more radical redistribution. The Economic Freedom Fighters in particular campaigned for expropriation without compensation, while the ANC itself eventually moved in the same direction, although with considerably more qualifications.

The controversy culminated in President Cyril Ramaphosa’s signing of the Expropriation Act in January 2025. The legislation replaced the apartheid-era Expropriation Act of 1975 and permits nil compensation in specified circumstances where this is considered just and equitable. This immediately attracted Trump’s attention. His executive order of 7 February 2025 presented the legislation much more dramatically, describing the South African government as having adopted a law enabling it to seize the agricultural property of ethnic-minority Afrikaners without compensation.

That description requires qualification. The Expropriation Act is not textually a law for confiscating white-owned farms. Its provisions are not formally restricted to a particular racial group, and nil compensation is permitted only under specified circumstances. Nor had South Africa embarked upon a general programme of confiscating white farms under the legislation.

Nevertheless, the underlying property-rights controversy is real. Critics argue that allowing the state to acquire property for nil compensation weakens the security of ownership and increases political discretion. The Democratic Alliance, despite participating with the ANC in the Government of National Unity, has challenged the legislation, as have other civil-society organisations.

The controversy therefore illustrates precisely the distinction important to the wider argument of this paper. The strongest claims made in American political rhetoric need not be accepted in order to recognise a genuine deterioration in confidence concerning the stability and impersonality of South Africa’s property-rights regime.

From BEE to Claims of Racial Discrimination: The Trump administration did not view the land question in isolation. It incorporated it into a broader interpretation of South Africa’s post-apartheid racial policies. Black Economic Empowerment, employment equity and related measures explicitly differentiate among citizens according to racial categories inherited from the apartheid system. Their purpose is remedial: to overcome the economic consequences of centuries of racial discrimination and, in particular, the institutionalised exclusions of apartheid.

The American administration interpreted these policies from a different normative starting point. Rather than emphasising historical redress, it emphasised equal treatment of individuals irrespective of race. From this perspective, policies granting economic preferences according to racial classification constituted discrimination against those excluded from the preferred categories. The February 2025 executive order accordingly referred not merely to land but to government policies allegedly undermining equal opportunity in employment, education and business.

This is important because the refugee policy cannot be understood simply as a response to farm murders. It emerged from a broader diagnosis: that Afrikaners and other white South Africans had changed from being the beneficiaries of a racial state to being a politically disfavoured minority within another system in which racial classification continued to have economic consequences.

The interpretation is contentious. Post-apartheid racial preferences plainly cannot be equated morally or institutionally with apartheid. Apartheid denied fundamental political, residential and economic rights to the majority population and enforced racial domination throughout society. BEE and employment-equity policies operate within a constitutional democracy and are explicitly justified as remedial measures intended to overcome that historical injustice.

Yet the controversy reinforces the institutional paradox identified in the preceding chapter. South Africa abolished an extreme system of racial status without altogether abolishing racial status as a category of economic governance. The direction, objectives and magnitude of discrimination changed fundamentally, but race remained relevant to the allocation of economic opportunities. The Trump administration effectively internationalised the resulting dispute.

Farm Attacks and the Failure of Public Security:

A second strand of the controversy concerns attacks on South African farms. South Africa suffers extraordinarily high levels of violent crime. Farmers and farm workers operate in isolated environments that can make them especially vulnerable to robbery and violent attack, and some farm murders have involved extreme brutality.

Afrikaner organisations, most prominently AfriForum, have for years argued that government has failed adequately to recognise and combat the problem. Farm attacks consequently became an international political issue well before Trump’s second presidency. Trump himself first intervened publicly in August 2018, during his first administration, when he instructed Secretary of State Mike Pompeo to investigate South African land seizures, expropriation and the killing of farmers.

The central empirical dispute is not whether farm attacks occur. They plainly do. The question is whether they constitute evidence of a specifically racial campaign against white farmers or are principally one manifestation of South Africa’s much wider violent-crime epidemic. The evidence does not establish a government-directed campaign to murder white South Africans, and the much stronger description of the situation as a “white genocide” is not supported by the available evidence. A serious argument about the state’s failure to provide security does not require this claim.

Indeed, from the perspective of the present paper, the more significant point is different. The state’s inability to guarantee physical security has contributed to precisely the private substitution mechanism examined elsewhere in this essay. South Africa has developed one of the world’s largest private-security sectors because citizens who lack confidence in public policing purchase protection privately. Farm attacks therefore belong within the wider story of declining confidence in state capacity, even if they cannot legitimately be characterised as evidence of genocide.

The controversy has been intensified enormously by the continued public performance of the struggle song Dubul’ ibhunu, commonly translated as “Kill the Boer” or “Shoot the Boer.” The song emerged from the anti-apartheid struggle and its contemporary defenders insist that it must be interpreted within this historical and political tradition rather than as a literal instruction to murder white farmers. South African courts have also considered the historical context important when assessing the song.

Yet its continued performance at large political rallies, particularly by Economic Freedom Fighters leader Julius Malema, inevitably carries a radically different meaning for many Afrikaners. The spectacle of thousands of political activists chanting words translatable as “kill the Boer” while South African farmers are in fact sometimes murdered provides extraordinarily powerful material for those arguing that Afrikaners face racial hostility.

The distinction between symbolic revolutionary rhetoric and literal incitement may be important legally and historically, but it is much harder to sustain politically once such imagery circulates internationally detached from its South African historical context. For the Trump administration and parts of the American political right, the song became visual evidence connecting land redistribution, racial preferences and farm attacks into a single narrative of persecution. Whether or not this synthesis is justified empirically, the South African political system itself supplied much of the material from which it was constructed.

Afrikaner applicants began arriving in the United States through the refugee programme, even as the Trump administration severely restricted refugee admissions from many other parts of the world. Whatever judgment is ultimately made about the justification for this preferential treatment, the historical reversal is remarkable. The Afrikaners had constructed and governed the apartheid state. In 1994 they surrendered their monopoly of political power in exchange for a constitutional settlement founded upon universal citizenship, democratic government, property rights and the rule of law. Three decades later, the government of the world’s most powerful Western democracy was offering members of this formerly dominant minority refugee protection from the political order that succeeded apartheid. This does not establish that the Trump administration’s interpretation of South Africa is correct. It does establish that something extraordinary has happened to international perceptions of South Africa.

The strongest interpretation should be resisted. There is no persuasive evidence of a coordinated South African government programme to kill white citizens. There has been no general confiscation of white-owned farms under the Expropriation Act. Claims of a “white genocide” therefore go considerably beyond the evidence. But rejecting these claims does not require dismissing the underlying phenomena.

South Africa maintains extensive race-conscious economic policies. Its new expropriation legislation has generated genuine controversy concerning property rights. Farm attacks are real within a wider environment of extreme violent crime. Radical political actors continue to employ inflammatory racial rhetoric. And substantial parts of the white population perceive their prospects as increasingly constrained by racial preference policies and deteriorating public institutions. The appropriate conclusion is therefore neither the Trump administration’s strongest persecution narrative nor the opposite assertion that the controversy is wholly invented. Rather, the episode provides another indicator of institutional deterioration.

A successful liberal post-apartheid order would ideally have made racial classification progressively less important as the inherited inequalities of apartheid diminished. It would have strengthened universal citizenship, secure property rights, impersonal economic competition and confidence in the state’s capacity to protect all citizens. Instead, racial classification remains embedded in significant areas of economic policy more than three decades after apartheid; property rights have become newly controversial; the state struggles to guarantee physical security; and racial political rhetoric remains sufficiently inflammatory that an external power can plausibly present members of the country’s historically dominant population as a vulnerable minority. The Trump refugee policy should therefore enter the analysis not as proof of white persecution but as a symptom of South Africa’s failure fully to consolidate an impersonal post-racial liberal order.

It also reinforces the central thesis of this paper. Institutional regression does not require the formal abandonment of constitutional democracy. South Africa retains elections, courts, constitutional rights, private property and sophisticated markets. Yet within this modern institutional framework, status-based allocation, patronage, insecurity and political discretion have expanded sufficiently to undermine confidence in the universality and predictability of the rules.

The extraordinary spectacle of white South Africans applying for refugee status in the United States is therefore less important as evidence for any particular claim made by Donald Trump than as an historical marker. It reveals how far South Africa has travelled from the optimism of 1994—and how fragile the achievement of an impersonal order based upon universal rights, competitive markets and the rule of law can prove to be.

11. Anti-immigrant Violence: Economic Competition Becomes Group Conflict

One of the most disturbing manifestations of South Africa’s post-apartheid social deterioration has been the recurrent eruption of xenophobic violence directed overwhelmingly by black South Africans against other black Africans.

The phenomenon complicates any interpretation of contemporary South African conflict primarily through the country’s historical black-white divide. The principal targets have instead been Zimbabweans, Mozambicans, Malawians, Somalis, Ethiopians and migrants from elsewhere on the African continent. The operative distinction is therefore not primarily racial but national: South African versus foreign African.

The violence also exposes an important political-economic dimension of South Africa’s institutional trajectory. In an expanding market economy, competition among individuals and firms is predominantly channelled through productive activity. New entrants simultaneously compete, produce, invest and consume. Economic interaction is therefore potentially positive-sum.

Under conditions of prolonged stagnation, mass unemployment, inadequate public services and politically mediated distribution, however, competition can increasingly be perceived as a struggle between groups over a fixed stock of opportunities.

The question changes from: How can more opportunities be created? to: Who is entitled to the opportunities that remain? South Africa’s xenophobic politics can partly be understood as this transformation of economic competition into distributive group conflict.

The decisive event occurred in May 2008. Violence beginning in Alexandra township in Johannesburg spread through Gauteng and subsequently into other parts of South Africa. Foreign nationals were assaulted, their homes and businesses looted or burned, and people were driven from communities in which they had been living. At least 62 people were killed, hundreds were injured and tens of thousands were displaced or left the country. The victims were overwhelmingly African.

The violence was therefore sometimes described as “black-on-black” conflict, although this formulation obscures the more important distinction. The attackers did not principally select victims because they were black. They attacked them because they were perceived to be foreigners. Indeed, some South African citizens were themselves killed after being mistaken for foreigners. Language, accent, appearance and perceived national origin could become criteria by which mobs distinguished insiders from outsiders. The episode revealed how rapidly the universalistic language of post-apartheid citizenship and African solidarity could break down under conditions of perceived scarcity.

The timing was not accidental. South Africa is by far the most developed large economy in southern Africa and consequently acts as a powerful migration magnet. Economic and political crises in neighbouring countries therefore generate migration towards South Africa. Zimbabwe was particularly important. The economic collapse associated with Robert Mugabe’s later rule, hyperinflation, political repression and declining employment drove large numbers of Zimbabweans southwards. Mozambique, Malawi and other poorer neighbouring countries similarly supplied migrants seeking employment and commercial opportunities.

From a regional perspective this migration was entirely predictable. Capital and labour move towards higher returns. The same economic asymmetry that makes South Africa attractive to international investors makes it attractive to African workers. Yet the receiving society was itself failing to generate sufficient opportunities for its existing population.

South Africa consequently combined two conditions that are politically explosive: regional economic superiority and domestic economic failure. It was prosperous enough relative to its neighbours to attract migrants, while simultaneously generating too little growth and employment to satisfy the expectations of its own citizens.

Mass Unemployment and the Political Economy of Resentment:

This context is essential. South Africa’s unemployment rate has for years remained at levels that would constitute a national emergency in most advanced economies. Youth unemployment is still more extreme. In such circumstances, even economically productive migrants can readily become politically framed as competitors for scarce jobs.

The same mechanism operates in informal markets. Somali, Ethiopian and other migrant entrepreneurs became highly visible in township retailing, particularly through small grocery shops commonly known as spaza shops. Their commercial success sometimes generated resentment among local competitors. Migrants were accused of accepting lower wages, operating businesses on smaller margins and displacing South Africans from employment and trading opportunities.

From a market perspective, these behaviours represent competition. Lower prices benefit consumers. Successful immigrant businesses create economic activity. Migrants earn income and subsequently spend it. Businesses employ workers and purchase supplies. But this positive-sum interpretation requires a social environment in which competitive success retains legitimacy. Where opportunities are perceived as fixed, the successful outsider can instead appear to have taken something that belonged to an insider. Competition becomes interpreted as displacement.

From Market Competition to Entitlement:

This transition is particularly important for the argument of this paper. A competitive market does not ask who deserves a particular economic opportunity according to group membership. It asks whether individuals can persuade customers, employers, investors or contractual partners voluntarily to transact with them. The principle is impersonal.

Nationality, ethnicity and political status become economically irrelevant unless they affect the individual’s capacity to perform. A distributive political order operates differently. Once economic opportunities are conceptualised as a stock to be allocated among politically recognised groups, the identity of the claimant becomes central. Who has priority? Who was historically disadvantaged? Who belongs? Who is entitled? Who has waited longest? Who is consuming resources intended for somebody else?

South Africa’s political economy already answers many economic questions through categories of group entitlement. BEE and employment-equity policies explicitly distribute advantages according to historically defined racial categories. Welfare and public services necessarily distinguish between different classes of claimants. Political rhetoric frequently interprets economic outcomes through collective identities. Anti-immigrant politics introduces another category into this distributive system: citizen versus foreigner. The underlying principle becomes South Africans first.

The Paradox of African Nationalism:

This development is particularly striking given the history of the ANC. The anti-apartheid struggle depended substantially upon solidarity from other African countries. Neighbouring states provided sanctuary, political support and, in some cases, bases for ANC activists despite considerable political and military costs. The ideological language of the liberation movement was correspondingly internationalist and pan-African. Yet post-apartheid South Africa has repeatedly produced intense popular hostility towards migrants from precisely the African continent that supported its liberation.

The contradiction is revealing. Political ideology can proclaim solidarity at the national level while everyday competition for jobs, housing, trading opportunities and public services produces very different incentives locally. Material conditions can overwhelm ideological fraternity.

The 2008 violence was initially interpreted as an extraordinary eruption. However, t did not remain exceptional. Further significant outbreaks occurred in 2015, particularly around Durban and Johannesburg. Foreign-owned businesses were attacked and migrants again fled affected communities. This became a recurring pattern. Another major outbreak followed in 2019.

By then the phenomenon had acquired international consequences. Nigerians and other African nationals were among those attacked, producing anger across the continent and damaging South Africa’s diplomatic standing. Retaliatory demonstrations and attacks against South African commercial interests occurred elsewhere in Africa. A country whose corporations had expanded successfully across the continent was confronting the political consequences of hostility towards Africans who had moved in the opposite direction. The asymmetry was difficult to ignore: South African capital was expected to be welcomed throughout Africa while African labour was increasingly unwelcome in South Africa.

A further stage appeared during the 2020s. Anti-immigrant mobilisation became more organised. Operation Dudula transformed grievances that had previously erupted principally through episodic township violence into a sustained political movement. Its activists demanded stronger immigration enforcement, the removal of undocumented migrants and priority for South African citizens in employment and public services. The name itself—dudula, approximately “push out” or “drive away”—captures the movement’s political logic.

Foreign nationals have been targeted at businesses, workplaces and public facilities. Activists have demanded documentation and attempted to prevent suspected migrants from accessing services. Particularly revealing have been attempts to exclude foreign nationals from public healthcare facilities. The argument is explicitly distributive: medicines, doctors, hospital beds and public resources are scarce; therefore foreigners should not consume resources that belong to South African citizens. This is precisely the politics of scarcity.

The State Loses Its Monopoly over Immigration Enforcement:

There is another institutional dimension. Immigration control is conventionally among the clearest functions of the sovereign state. Government determines who may enter, who may remain and under what conditions non-citizens may work. Yet where citizens perceive government as incapable of controlling borders or enforcing immigration law, private political movements begin attempting to perform this function themselves. Activists demand identity documents. Groups inspect businesses. Migrants are told to leave communities. Access to public facilities is policed informally. Whatever one’s view of South Africa’s immigration policy, this represents another manifestation of weakening state authority. The issue is no longer merely whether immigration is too high or too low. The institutional question is: Who decides? When political movements begin enforcing their own conception of immigration law on streets, workplaces and clinics, the state’s monopoly over legitimate enforcement has begun to erode.

By 2026 anti-immigrant mobilisation had again intensified substantially. Movements including Operation Dudula and March and March organised campaigns demanding stronger enforcement and the prioritisation of South African citizens. Large numbers of migrants subsequently left or were repatriated, while government enforcement also increased. The economic consequences revealed an important contradiction within the anti-immigrant argument. Industries employing substantial numbers of migrants began experiencing labour shortages.

Manufacturing businesses in Newcastle, for example, reported significant losses of migrant workers, including workers possessing skills that could not immediately be replaced locally. This illustrates a general economic principle. The migrant appears politically as a competitor for employment. But to the employer he is also labour. To the consumer he contributes to production. To the landlord he is a tenant.To businesses he is a customer. To the entrepreneur he may be a supplier or employee. Removing the migrant therefore does not simply transfer an existing job to a citizen. Economic relationships reorganise. Businesses may contract. Prices may rise. Investment may decline. Some jobs may disappear altogether. The assumption of a fixed quantity of employment—the “lump of labour” intuition—is therefore misleading.

Immigration and the Positive-Sum Order:

The distinction leads back to the central theoretical argument of this paper. Markets transform strangers into collaborators without requiring them to share ethnicity, nationality, religion or political identity. The employer does not need to regard the migrant as a compatriot. The customer does not need to know the shopkeeper’s ancestry. The supplier does not need to share the buyer’s culture. They require only sufficiently predictable rules of property and contract to cooperate. This is one of the great civilisational achievements of the impersonal market order. It enormously enlarges the circle of potential cooperation.

The stranger ceases to be principally a rival for tribal resources and becomes a potential customer, employee, employer, investor, supplier or contractual partner. Economic growth reinforces this process because expanding opportunities reduce the plausibility of zero-sum interpretations. Institutional and economic regression can push in the opposite direction. As growth disappears, unemployment remains extreme and public services deteriorate, individuals increasingly perceive society as a contest over a fixed stock of resources. The stranger once again becomes a rival.

Patronage and the Politics of Insiders:

This is where xenophobia connects to the patron-client analysis developed earlier. Patronage necessarily distinguishes insiders from outsiders. The patron possesses resources and distributes them among clients. The political struggle therefore concerns membership in the network entitled to receive benefits. Once this logic expands, political competition increasingly takes the form of competing claims to insider status.

BEE asks which historically defined racial categories should receive preference. Cadre deployment asks which political networks should control appointments. Local patronage asks which supporters should receive contracts or positions. Anti-immigrant politics asks which national group should receive jobs and services. These phenomena are not identical and should not be morally or causally conflated.

But they share an institutional tendency: economic opportunities become interpreted increasingly through the identity of the claimant rather than through impersonal competitive selection. The foreign migrant represents the ultimate outsider because he possesses the weakest claim within a nationally bounded distributive system.

It would nevertheless be wrong to claim that South Africa’s economic deterioration mechanically caused xenophobic violence. Hostility towards migrants existed before the economic deterioration examined in this paper. National identity, policing failures, misinformation, criminal networks, political mobilisation and local conflicts all contribute independently. Nor should every objection to uncontrolled immigration be labelled xenophobic. A sovereign state can legitimately determine immigration rules, enforce borders and distinguish citizens from non-citizens in access to particular political rights or publicly financed benefits.

The relevant distinction is between immigration policy administered through general law and collective punishment or violence directed against individuals because of their foreign origin. The latter represents institutional failure rather than immigration control. The significance of xenophobic violence for this argument is therefore diagnostic. It reveals what can happen when mass unemployment, weak state capacity, migration pressure and distributive politics interact.

Two Responses to Institutional Failure:

South Africa now displays two radically different responses to deteriorating institutions. The first is distributive exclusion. When opportunities are scarce, groups attempt to exclude competitors: foreigners from jobs, foreign traders from markets, migrants from public services, outsiders from access to scarce resources. This is fundamentally zero-sum.

The second response is institutional exit and reconstruction. Households install solar panels. Businesses provide their own security. Residential communities manage infrastructure. Private schools and hospitals substitute for public provision. And, at the most radical end of this spectrum, communities such as Orania attempt to assemble an increasingly comprehensive alternative institutional environment.

The contrast is fundamental. One strategy responds to scarcity by asking: Whom can we exclude from the existing stock? The other asks: What can we build ourselves? The first directs entrepreneurial and political energy towards redistribution. The second directs it towards production and institution-building.

From Scarcity Politics Back to Open Competition:

South Africa’s xenophobic episodes therefore provide another warning about the fragility of the impersonal competitive order. Economic competition is peaceful only within an institutional and cultural framework that legitimises the gains of strangers.

The market asks individuals to accept a remarkable proposition: somebody with whom they share no kinship, ethnicity, nationality or political affiliation may legitimately outperform them economically if that person better satisfies the voluntary choices of customers or employers. This proposition is historically unusual.

It requires confidence that the rules are general, that opportunities remain open and that today’s competitive loss does not imply permanent exclusion. When that confidence disappears, competition can revert towards older forms of group conflict. South Africa’s anti-immigrant violence is therefore not merely an immigration issue. It is a manifestation of the deeper institutional problem examined throughout this paper: the displacement of impersonal, positive-sum competition by political struggles over status, entitlement and access.

The tragedy is especially acute because migrants from poorer African countries come to South Africa for precisely the reason that millions of people throughout history have migrated towards successful societies: they seek access to a more productive economic order. If the receiving society responds by treating their productive participation primarily as a claim upon a fixed stock of resources, it demonstrates how far confidence in that expanding order has already deteriorated. The ultimate antidote to xenophobic scarcity politics is therefore not merely tolerance. It is the restoration of a sufficiently dynamic, open and impersonal economic order in which newcomers are principally perceived not as claimants upon a shrinking inheritance but as additional participants in the creation of wealth.

12. Patronage, Spontaneous Privatization, and ‘White Flight’

The economic damage of corruption does not primarily lie in redirecting public funds into private pockets. Its deeper effect is to alter the payoff matrix facing everyone else. The classic rent-seeking literature shows that when government creates valuable discretionary privileges, resources will be expended to capture them (Tullock 1967; Krueger 1974). Olson’s theory of distributional coalitions similarly explains how organised groups can accumulate privileges that reduce dynamism over time (Olson 1982).

The mechanism is self-reinforcing. More politically allocated resources increase the value of political connections. Higher returns to connections encourage greater investment in influence. Those who succeed at acquiring influence then have incentives to expand the domain of political allocation. Political office itself becomes an economic asset. Competition does not disappear; it migrates from the market for serving customers toward the market for gaining access to patrons. This is why corruption and patronage can persist even when almost everyone recognises their aggregate cost. Each participant can be responding rationally to an incentive structure that is collectively destructive. Institutional reform therefore requires changing the rules of selection, not merely exhorting individuals to become more virtuous.

South Africa’s distinctive response to institutional failure has been widespread substitution. In security this process long predates the electricity crisis. Research documents extensive interaction between private security firms, neighbourhood organisations and public police, making South Africa an important case of plural or hybrid security provision. Electricity has followed a similar path. Eskom-supplied electricity demand is falling by about three per cent in 2024 partly because of increased private-sector generation uptake. The key institutional fact is that users did not simply wait for the monopoly provider to recover; they invested in exit options. The same logic appears in private schooling, healthcare, gated residential developments and business districts that supplement municipal services. This produces an unusual form of privatisation without a comprehensive privatisation programme. The legal and fiscal state remains extensive, but its effective domain contracts where citizens cease to rely on it.

This should not be romanticised. Private substitution is expensive and unequal. The affluent can purchase parallel systems; the poor often cannot. A household that pays taxes and then pays again for security, electricity backup, education or healthcare bears a duplication cost. Yet the process reveals the resilience of decentralised provision and the limits of formal monopoly when exit becomes technologically and socially possible.

Physical emigration is another form of exit, but the evidence requires caution. South Africa has experienced substantial outward migration of citizens and persistent concern about the loss of skilled professionals. Statistics South Africa’s Migration Profile documents emigration as a significant component of the country’s migration system and notes rising numbers of South Africans studying abroad. “White flight” representing a skill exodus is real, although official emigration measurement is incomplete. The theoretical point does not depend on a precise statistical measure of emigration. In any jurisdiction, mobile human and financial capital will tend to respond to deteriorating security, infrastructure, taxation or institutional predictability. Exit can discipline governments, but it can also weaken the tax base and reduce the stock of scarce skills. This creates another possible feedback loop: deteriorating institutions encourage the departure of those best able to leave, which can make institutional repair more difficult, and induce a further skill exodus.

The libertarian significance of the South African case extends beyond the familiar observation that governments can fail. The deeper lesson concerns institutional competition. Where consumers can exit, failing organisations lose resources. Where provision is monopolised and funded coercively, failure can persist because the organisation’s revenue is partly insulated from user satisfaction.

South Africa nonetheless demonstrates that exit reappears when technology and private coordination make it possible. Residential associations and business improvement districts supply local collective goods. These are examples of polycentric governance: overlapping providers rather than a single sovereign supplier (Ostrom 2010).

The important caveat is that spontaneous privatisation under conditions of state failure is not equivalent to an orderly libertarian reform. It can be wasteful, unequal and legally insecure. A more attractive libertarian programme would remove barriers to competitive provision deliberately, reduce duplication by lowering taxes as public responsibilities recede, establish secure property rights and contractual rules, and permit infrastructure markets to emerge before state failure forces emergency substitution.

13. Orania: Institutional Exit, Competitive Governance and Order from Below

One of the most unusual responses to South Africa’s institutional trajectory can be observed in Orania, the small Afrikaner settlement in the Northern Cape established in 1991. Orania is usually discussed as an experiment in Afrikaner nationalism or separatism. Its inhabitants explicitly seek greater Afrikaner self-determination, and the community’s longer-term political horizon is connected to the Volkstaat tradition: the idea that Afrikaners should concentrate territorially in order eventually to govern themselves.

These ideological origins cannot be ignored. Yet for the purposes of the present argument, Orania is interesting for another reason. It represents an unusually advanced experiment in institutional exit. The conventional affluent South African response to declining state capacity is functional substitution. If municipal services become unreliable, gated communities and neighbourhood associations assume some of their functions. Orania extends this logic from individual services towards the construction of an entire community. Instead of separately substituting private solutions for particular failures of the state, it attempts to assemble land, infrastructure, businesses, schools, community organisations and local governance within a territorially concentrated institutional ecosystem. Its importance therefore lies not only in what its inhabitants are trying to preserve, but in how they are attempting to preserve it.

Exit Rather Than Capture:

Albert Hirschman’s classic distinction between exit and voice provides a useful starting point. When individuals are dissatisfied with an institution, they can attempt to change it politically—voice—or withdraw from it and choose an alternative—exit. Markets institutionalise exit continuously. Consumers dissatisfied with one supplier can switch to another. Political systems normally rely much more heavily upon voice because citizens cannot readily change governments or jurisdictions.

South Africa increasingly displays hybrid forms of exit. The taxpayer remains formally inside the state but progressively exits from particular services. He ceases relying upon the police but remains subject to taxation. He disconnects partially from Eskom but remains connected to the national political system. He sends his children to private school while continuing to finance public education.

Orania radicalises this process by territorialising it. Its strategy is not principally to acquire control over the South African state and redirect its resources towards Afrikaner interests. Instead, its inhabitants seek to reduce their dependence upon that state by accumulating the institutional capacities necessary for self-government.

This distinction is particularly important in the context of the patronage argument developed earlier in this paper. One response to a political system organised increasingly around distributive competition is to construct a sufficiently powerful political coalition to capture a larger share of the distribution. Another is to withdraw as far as possible from distributive politics altogether. Orania aspires to the latter. Its underlying proposition is effectively: rather than compete for control of the state, build institutions that reduce the importance of controlling the state. This makes it interesting from a libertarian perspective even though its underlying nationalism is not itself libertarian.

Own Land, Own Labour, Own Institutions:

Orania summarises its philosophy through three interconnected principles: eie grond, eie arbeid, eie instellings—own land, own labour, own institutions. Each contains an institutional insight.

Own land provides the territorial basis of governance. A community dispersed throughout a much larger political jurisdiction cannot easily establish alternative institutions because its members remain embedded within the institutions and infrastructure of the surrounding society. Territorial concentration allows rules, services and social norms to operate together.

Own institutions means that political aspiration must be supported by institutional capacity. Schools, businesses, administrative structures, infrastructure, financial arrangements and community organisations must exist before meaningful autonomy can exist.

This reverses the conventional nationalist sequence. The conventional separatist movement demands political sovereignty first and assumes that institutions can subsequently be constructed. Orania’s strategy is closer to: institution-building → economic viability → demographic concentration → increasing practical autonomy → potentially greater political autonomy. Sovereignty, on this interpretation, is not primarily declared. It is accumulated.

The third principle, own labour, is perhaps the most revealing.

Historically, affluent white South African communities depended extensively upon lower-paid black labour for domestic work, construction, agriculture and municipal services. Orania regards such dependence as inconsistent with genuine self-determination. A community cannot plausibly claim to be autonomous while depending upon an external population to perform the work required for its everyday reproduction. Residents therefore undertake forms of manual and service labour that affluent white South Africans elsewhere have traditionally outsourced. Whatever one thinks about the ethnonationalist motivation, the institutional principle is significant: autonomy requires bearing the full cost of one’s own social order. This is the opposite of rent-seeking. Instead of attempting to externalise costs onto politically weaker groups or the general taxpayer, a genuinely self-governing community must internalise the costs of maintaining itself.

Orania as a Proprietary Community:

Orania also possesses features of what might be lead us to talk here about a proprietary community, in the sense of Spencer H. MacCallum’s ‘The Art of Community’. MacCallum used “proprietary community” (MacCallum 1970) for communities in which common land and facilities are privately owned and the proprietor supplies community services contractually, i.e. a community in which a substantial part of the territory, infrastructure, or collective services is owned or controlled through private property arrangements rather than administered in the ordinary way by a municipal government. The idea extends to the concept of a privately developed and administered town. Orania has some striking similarities. Its land was originally acquired privately in 1991; much of its development has been deliberately organized through private/community institutions; residents are subject to admission arrangements; it provides or organizes many local services; and the community has developed a considerable degree of economic and institutional autonomy. However, it is not a Free Private City (Gebel 20218) in Titus Gebel’s sense because Orania is still within the Republic of South Africa and subject to South African law.

However, Orania – via the extensive use of condominium-style residence management regulates access like a proprietary community. Residence is not simply determined by purchasing any available house and thereby acquiring an unconditional right to participate in the community. Prospective residents are subject to admission procedures and are expected to subscribe to the community’s Afrikaans language, Afrikaner cultural identity and self-determination project. This is obviously the most controversial aspect of Orania.

Critics argue that cultural selection functions in practice as racial exclusion and that Orania reproduces segregation within post-apartheid South Africa. Supporters respond that voluntary cultural association on privately controlled territory is fundamentally different from apartheid because Orania claims no authority over outsiders and does not seek to compel other ethnic groups to live according to Afrikaner rules. For the present argument, however, another institutional feature is important: entry into the community is selective while exit remains voluntary.

This resembles other forms of private association. Firms select employees. Clubs select members. Universities select students. Residential associations impose covenants. Religious communities establish membership requirements. Individuals who reject these conditions remain free to organise themselves differently elsewhere. The difficult question is how far this associational principle can legitimately extend once an organisation begins performing functions normally associated with territorial government. Orania therefore sits at an intellectually interesting boundary between private association and political jurisdiction. It is precisely this ambiguity that makes it useful as an experiment in alternative governance.

The High-Trust Hypothesis:

There is another possible explanation for Orania’s performance that deserves serious investigation. Its relative institutional effectiveness may have less to do with Afrikaner ethnicity as such than with the consequences of self-selection and high social trust.

Individuals who move to Orania are not a random sample of Afrikaners. They voluntarily enter a community requiring unusually strong commitment to a common project. They accept cultural expectations, perform their own labour, participate in local institutions and knowingly relocate to an isolated settlement whose principal attraction is precisely this institutional experiment. The selection effect is therefore enormous. Orania might consequently be producing a high-trust environment through several mutually reinforcing mechanisms: voluntary entry; shared norms; reputational accountability; small institutional scale; relatively easy identification of free-riding; local responsibility for collective goods; and the possibility of exit.

These conditions reduce some of the principal-agent problems characteristic of large bureaucratic states. An official administering a national department is remote from millions of taxpayers and service users. A local administrator in a small community encounters the people affected by his decisions repeatedly. Poor performance becomes visible. Reputation matters. Information travels quickly. This does not eliminate bad governance, factionalism or incompetence. Small communities can suffer all three. But it alters the feedback mechanism.

Local Knowledge and Competitive Governance:

This brings Orania surprisingly close to Hayekian arguments concerning local knowledge. Large political systems necessarily make decisions using aggregated information. Central authorities cannot possess detailed knowledge of the circumstances, preferences and opportunities dispersed among millions of individuals. Markets address this knowledge problem through decentralised decision-making and price signals. Local governance potentially captures some of the same advantage. Decisions are made closer to the individuals possessing relevant knowledge and bearing the consequences. Orania adds another mechanism: jurisdictional competition.

If living conditions deteriorate sufficiently, residents can leave. If the community becomes attractive, outsiders who satisfy its membership conditions seek to enter. Capital follows similar signals. This creates something analogous to a market test of governance.

The comparison should not be exaggerated. Moving town is vastly more costly than changing supermarkets. Governments possess coercive powers firms do not. And Orania remains embedded within South Africa’s larger legal and economic order. Nevertheless, the underlying principle resembles Charles Tiebout’s famous model of individuals “voting with their feet” among local jurisdictions.

Instead of all citizens attempting to impose one preferred institutional arrangement upon an entire country through national elections, different communities can potentially offer different packages of rules, services and obligations. Individuals then sort themselves among them. The result is not a single optimal institutional model but competition among institutional models.

From Monopoly Governance to Polycentric Order:

Elinor Ostrom’s work provides another useful theoretical framework. Ostrom challenged the assumption that collective goods must be supplied either by a central state or by a single privatised provider. Real societies frequently develop polycentric systems in which households, firms, associations, municipalities and other organisations overlap in supplying governance functions. Contemporary South Africa increasingly exhibits precisely such arrangements.

A gated estate employs security guards while remaining within the jurisdiction of the South African Police Service. A private school operates alongside a public education system. A business improvement district supplements municipal services. Orania can be interpreted as an attempt to bundle many of these arrangements territorially. It therefore represents not the complete disappearance of the state but the emergence of a nested governance system. South African sovereignty remains outside it. Inside that framework, however, a growing range of functions is organised locally. This distinction matters because institutional evolution rarely proceeds through clean replacement. New institutions emerge alongside old ones, gradually absorbing functions as they demonstrate greater competence. The process can be evolutionary rather than revolutionary.

A Competitive Alternative to Patronage

This brings Orania directly into the central argument of this paper.

Patronage and competitive governance represent opposite responses to institutional failure. Under patronage, individuals seek access to the political centre because the political centre controls valuable resources. Under competitive governance, individuals seek to reduce the number of resources controlled by the political centre by constructing alternatives. The first strategy increases the value of political connections. The second decreases it. The first produces incentives to capture institutions. The second produces incentives to build institutions. The first asks: Who controls the allocation? The second asks: Can we produce an alternative? This is why Orania is more theoretically interesting than its small population might suggest. It constitutes an experiment in changing the direction of entrepreneurial effort from rent-seeking towards institution-building.

The Ethnicity Problem

There remains, however, an unavoidable question. Is Orania successful because it is Afrikaner, or because it possesses institutional characteristics that could in principle be reproduced by communities with completely different ethnic compositions? This distinction is decisive. If its effectiveness depends fundamentally upon ethnic homogeneity, then Orania principally supports an ethnonationalist interpretation of social order: culturally homogeneous populations possess advantages in trust, solidarity and collective action that heterogeneous societies cannot easily reproduce.

But there is another possibility. Ethnic-cultural homogeneity may simply be Orania’s particular mechanism for achieving a more general set of institutional conditions: high trust, shared norms, voluntary self-selection, reputational accountability, local knowledge, strong property rights, internalisation of costs and meaningful exit. If these are the operative mechanisms, there is nothing inherently Afrikaner about them.

A Chinese entrepreneurial community, an Indian technology enclave, an international university town, a religious community, a privately developed city or a deliberately cosmopolitan charter city might theoretically produce similar institutional advantages through entirely different mechanisms of selection and cohesion. The hypothesis should therefore be tested rather than assumed. The intellectually interesting question is not: Does Afrikaner ethnicity produce superior governance? It is: Which institutional characteristics of Orania allow collective goods to be produced effectively, and which of those characteristics are transferable beyond the particular cultural community in which they emerged?

This distinction is particularly important in relation to the argument I developed in “Thesis of German Right Wing Populism – A Rationale for the Protection of High-performance National Cultures.” The relevant object of preservation is not ethnicity as such but the cultural and institutional patterns capable of sustaining advanced social cooperation and productive performance. From this perspective, Orania should be judged by performance rather than identity.

The Experimental Test

Orania therefore presents an empirical opportunity. If its institutional model is genuinely superior in relevant respects, this should become observable. Does its population grow voluntarily? Does capital flow into the community? Are businesses created? Do property values rise? Does infrastructure improve? Are services reliable? Is crime comparatively low? Are schools effective? Can the community finance the infrastructure required by a growing population? Does it generate sufficient productivity to support itself without hidden subsidies from the surrounding state? Can it attract skilled individuals rather than merely ideological adherents? And, perhaps most importantly, does institutional performance persist as the community grows?

The last question is crucial. Governance that works among several thousand highly self-selected people may cease working among tens or hundreds of thousands. Informal trust mechanisms become weaker as populations increase. Personal reputation becomes less effective. Administrative complexity grows. Orania’s real test is therefore not survival. It is scalability.

Exit as Discovery

The appropriate libertarian attitude towards Orania is therefore neither romanticisation nor dismissal. There is no need to decide in advance whether its social model ought to be universally adopted. Indeed, that would contradict the very principle that makes the experiment interesting. The point of institutional competition is that we do not know in advance which institutional arrangements will work best. Different communities should be allowed, within the constraints imposed by the rights of others, to experiment with different arrangements.

Successful institutions attract people and capital. Unsuccessful institutions lose them. Institutional diversity thereby becomes a discovery process analogous to competition within markets. This is perhaps Orania’s deepest significance. South Africa’s institutional regression has encouraged many citizens to retreat partially from the state. Orania transforms that defensive retreat into a constructive project of institution-building.

Its inhabitants are not merely purchasing generators because Eskom failed or guards because policing failed. They are attempting to answer a more radical question: How much of a functioning social order can a community reconstruct for itself? The answer remains uncertain. But the experiment itself demonstrates the possibility of responding to institutional deterioration not solely through political struggle for control of the existing state but through the creation of competing institutions.

Orania as Institutional Entrepreneurship

Seen in this light, Orania can be understood as an example of institutional entrepreneurship. The entrepreneur normally combines resources to create a new product or organisation within an existing institutional framework. The institutional entrepreneur attempts something more fundamental: to create or modify the framework within which social cooperation occurs. Land, capital, labour, rules, infrastructure and community are assembled into an alternative governance arrangement. The product being offered is, in effect, institutional quality.

Residents choose whether to buy into it with their lives and capital. Businesses choose whether to invest within it. Potential migrants decide whether its opportunities compensate for its restrictions. Existing residents decide whether to remain. This does not make Orania a pure market institution. It possesses collective rules and political structures and remains dependent upon the larger South African constitutional and economic order. But neither is it simply another municipality. It occupies the space between market, community and state. Precisely this institutional ambiguity makes it relevant to a society in which the conventional boundary between those spheres is already being renegotiated.

From Retreat to Reconstruction

South Africa’s private enclaves are sometimes interpreted principally as symptoms of social fragmentation: wealthy citizens withdraw behind walls while the public realm deteriorates around them. There is considerable truth in this criticism. Exit can weaken voice. Citizens who privately replace public services may have less incentive to improve the institutions serving everyone else. Private substitution can therefore coexist with—and perhaps even accelerate—the deterioration of the common public realm.

Orania forces the argument one stage further, however. At some point, exit ceases to be merely retreat. When private security, energy, education, infrastructure, businesses, community organisations and governance are recombined territorially, exit becomes reconstruction. That possibility deserves attention far beyond the particular Afrikaner politics of Orania.

The central problem confronting a deteriorating institutional order is ultimately not how citizens can escape its failures individually. It is how functioning institutions can emerge to replace those that no longer perform. South Africa may therefore be generating, unintentionally and from below, experiments in the reconstruction of institutional order.

Orania is currently the most radical of them. Its ultimate significance will depend not upon whether it preserves Afrikaner identity but upon whether it demonstrates something more general: that communities possessing sufficient social capital, local knowledge, responsibility for costs and freedom to experiment can construct institutions capable of outperforming centralised political provision. If so, Orania would represent not merely a refuge from South Africa’s institutional regression, but a small experiment in the competitive discovery of alternative forms of governance.

14. Institutional Regression, Exit and the Reconstruction of an Open Order

South Africa’s post-apartheid history should not be read as a simple morality tale. It is neither the story of an inevitably doomed African state nor the straightforward destruction of a successful economy at the moment majority rule arrived in 1994. The first decade of democratic government contradicts such an interpretation. The ANC initially preserved important elements of the inherited market order, pursued relatively orthodox macroeconomic policies and presided for a period over expanding investment and economic growth. Nor can South Africa’s subsequent deterioration plausibly be reduced to a single policy such as BEE, a single president such as Jacob Zuma, or a single phenomenon such as corruption.

The argument developed in this paper is more structural. South Africa illustrates the reversibility of institutional modernity. The institutions characteristic of an advanced market economy can remain formally intact while the mechanisms operating through them progressively change. Private property survives. Corporations continue trading. Banks function. Elections take place. Courts continue adjudicating disputes. Universities operate. Government departments, municipalities and state-owned enterprises retain their legal identities.

Yet beneath this architecture, or rather facade, the criteria determining access to opportunities can shift. Competition can give way incrementally to political allocation. Merit can become entangled with political loyalty. Rights can become increasingly dependent upon administrative discretion. Entrepreneurial effort can migrate from discovering what customers want towards discovering what political gatekeepers require. Contracts can become vehicles of patronage. Political connections can become economic assets. None of these changes requires the formal abolition of capitalism. A society can regress institutionally while retaining the outward appearance of a modern market economy.

The Historical Anomaly of the Open Order

This is the wider historical significance of the South African case. Modern people easily assume that impersonal economic competition under general rules represents the normal condition of society. Historically, the opposite is closer to the truth. For most of human history, access to economic opportunity was embedded in relationships of kinship, clan, inherited status, political allegiance and personal patronage. The individual advanced through networks of protection and obligation. Economic and political power were intertwined.

The emergence of an order in which strangers could transact under relatively general rules, acquire property independently of political favour and compete for opportunities without belonging to the same family, tribe, religion or patronage network was an extraordinary historical achievement. North, Wallis and Weingast’s distinction between limited-access and open-access orders captures an important part of this transformation. Open access to economic and organisational competition is not humanity’s default condition. It is a comparatively rare institutional accomplishment.

As I argued in my earlier essay, “Thesis of German Right Wing Populism – A Rationale for the Protection of High-performance National Cultures,” the prosperity take-off achieved first in Western Europe depended upon the cumulative co-evolution of an intricate network of cultural, technological, organisational and institutional preconditions. The South African case adds an important corollary: what was historically difficult to construct can also be comparatively easy to dissipate.

Institutions have to be reproduced. Capital can be Consumed. This principle applies to more than physical infrastructure. A power station is capital. But so is the engineering organisation capable of maintaining it. A railway is capital. But so are the routines, professional norms, technical knowledge and managerial structures necessary to operate it. A constitutional provision protecting property is an institution. But so is the expectation, embedded across millions of decisions, that property rights will in fact remain predictable.

Civilisation therefore possesses stocks of physical, human, social and institutional capital. All can be consumed or dissipated. This explains why institutional deterioration can remain invisible for surprisingly long periods. Infrastructure continues functioning after maintenance deteriorates. Organisations continue operating after experienced personnel disappear. Investment continues for some time after regulatory uncertainty increases. Inherited capital conceals the deterioration of the mechanisms required to reproduce it. Eventually, however, the accumulated stock becomes insufficient.

The failures then appear suddenly even though their causes have been accumulating for years. Eskom provided perhaps the clearest South African illustration of this process.

From Productive Entrepreneurship to Political Entrepreneurship

The distinction between productive and political entrepreneurship provides a connecting mechanism through much of the paper. All institutional systems generate incentives. Entrepreneurial intelligence does not disappear when institutions deteriorate. It changes direction.Where profits depend principally upon satisfying customers, entrepreneurs invest in products, technologies, organisational efficiencies and cost reduction.

Where substantial profits can be acquired through government contracts, licences, regulatory classifications and political relationships, entrepreneurs rationally invest in those instead. This is the deeper problem with corruption. The immediate theft is visible. The opportunity cost is not. A corrupt procurement contract costs the taxpayer whatever was stolen or overcharged. But a political-economic system in which thousands of capable people discover that relationships generate higher returns than innovation imposes a much larger dynamic cost. It changes what society selects for.

The interaction between BEE, preferential procurement, cadre deployment and patronage is significant in this respect. BEE should not simply be equated with corruption. Its historical motivation was the understandable objective of integrating a population deliberately excluded from economic opportunities under apartheid.

But politically administered preferences create opportunities for political entrepreneurship. The deeper issue is not who received the rents. It is the creation of rents worth competing for.

From Status to Contract—and Back?

The post-apartheid settlement consequently contains a profound paradox. Apartheid represented an extreme system of status. Race determined political rights, residence, employment and access to economic opportunities. The liberal alternative would have completed the transition from status to contract: universal individual rights, general rules and unrestricted competition irrespective of racial identity.

Post-apartheid South Africa instead pursued a hybrid arrangement. Apartheid’s exclusionary racial restrictions were abolished, but racial classification was retained as an instrument of remedial economic allocation. Race remained relevant to economic selection. More than three decades after apartheid, South African businesses still calculate the racial composition of ownership, management, employment, procurement and supplier relationships. A category that might have been expected progressively to lose political significance instead became embedded within a permanent regulatory architecture.

The controversy surrounding President Trump’s decision to create a refugee pathway for Afrikaners illustrates how far this tension has travelled internationally. The Trump administration’s strongest claims concerning white persecution and farm confiscation go beyond what the evidence establishes. There is no persuasive evidence of a state-sponsored “white genocide,” nor has South Africa implemented a general confiscation of white-owned farms. Yet the controversy could not have arisen in a vacuum. Race-conscious economic regulation remains extensive. Property rights have become newly contested through the Expropriation Act. Violent crime remains extreme. Farm attacks are real. Inflammatory racial rhetoric persists. The extraordinary spectacle of members of South Africa’s formerly dominant racial minority seeking refugee status in the United States therefore serves as a marker of the unresolved transition from racial status towards universal citizenship.

The Politics of Scarcity

The xenophobic violence directed against African migrants reveals another dimension of the same institutional problem. The attacks beginning dramatically in 2008 and recurring subsequently are not primarily an extension of South Africa’s historic white-black conflict. They are conflicts between insiders and outsiders. This matters theoretically. In a dynamic market order, the stranger is simultaneously competitor and collaborator. Competition is therefore embedded within a potentially expanding system of cooperation. Under conditions of stagnation and mass unemployment, the same stranger can instead be perceived as consuming part of a fixed stock. Politics then asks not how the stock can be expanded but who is entitled to it. This is the transition from positive-sum competition towards zero-sum distributive conflict. The immigrant becomes the ultimate outsider to the distributive coalition.

Two Responses to Institutional Failure

Here the South African experience presents a revealing contrast. One response to scarcity and institutional deterioration is exclusion. Foreigners are excluded from jobs, expelled from markets. Migrants are prevented from accessing services. Groups struggle politically to reserve existing resources for themselves.

But South Africa simultaneously displays a radically different response. Businesses increasingly construct infrastructure around failing public provision. And at the most radical edge of this process, Orania attempts to assemble many of these functions territorially into an increasingly comprehensive alternative institutional environment. Here the question is not: Who can we exclude from what still exists? It is: What can we build ourselves?

This distinction may be one of the most important conclusions to emerge from the South African case. Both responses originate partly in dissatisfaction with existing institutions. But they direct human energy in opposite directions. Scarcity politics searches for outsiders to exclude. Institutional entrepreneurship searches for alternatives to construct. The first directs effort towards redistribution. The second directs effort towards production. The first increases the political value of controlling existing institutions. The second reduces dependence upon them.

Orania Beyond Afrikaner Nationalism

This is what makes Orania more interesting than its tiny scale might initially suggest. Its ethnonationalist origins are undeniable. The community explicitly seeks Afrikaner self-determination. But the theoretically interesting question is whether its institutional performance depends upon Afrikaner ethnicity as such.

There is an alternative hypothesis. Perhaps Orania’s relative effectiveness derives from more general institutional characteristics: voluntary self-selection, high levels of social trust, strong common norms, local knowledge, reputational accountability, internalisation of costs, secure property arrangements, institutional experimentation, and meaningful possibilities of exit.

If these are the operative mechanisms, then the lesson is not ethnonationalist. It is institutional. The same principles might emerge in communities organised around entirely different identities or purposes: entrepreneurial cities, university towns, privately developed communities, religious settlements, charter cities or deliberately cosmopolitan enclaves. The relevant test is performance: Do people voluntarily move there? Does capital follow? Are businesses created? Can the institutional system scale? Can it maintain openness to innovation? Does it generate productivity without imposing costs upon outsiders?

Orania should therefore be treated as an experiment rather than a model to be accepted or rejected in advance. This is Hayek’s discovery procedure applied to governance.

Exit Is Not Enough

A libertarian analysis must nevertheless recognise a serious difficulty. Exit can itself accelerate institutional fragmentation. When affluent citizens abandon public schools, public hospitals, public policing and public electricity, they reduce their dependence upon state institutions—but potentially also their incentive to improve them. The most capable citizens can retreat into private provision while those without sufficient resources remain trapped within deteriorating public systems. Private exit can therefore coexist with public decline.

This is Hirschman’s dilemma. Exit may protect the individual while weakening voice. The South African question is consequently not merely whether private substitutes can emerge. They clearly can. The deeper question is whether institutional exit can eventually become institutional reconstruction. This occurs when alternative providers cease merely compensating defensively for isolated state failures and begin assembling durable systems of governance capable of generating their own institutional capital. At this point exit becomes entrepreneurial.

South Africa may consequently be evolving towards something more complicated than either state collapse or conventional privatisation. Ostrom’s concept of polycentric governance offers a useful framework. Governance need not be supplied by one monopolistic centre. Multiple institutions can overlap. The emerging order is neither purely public nor purely private. It is polycentric. South Africa’s institutional deterioration may therefore be producing, unintentionally, experiments in governance that would never have emerged under a competent monopoly provider.

This does not redeem the deterioration that generated them. The resources devoted to generators, walls, guards and duplicated infrastructure represent enormous opportunity costs. But institutional failure can nevertheless become a discovery process if it permits alternative arrangements to emerge.

The Choice: Capture, Exclusion or Construction

The South African experience ultimately reveals three responses to declining institutional performance. The first is capture. Acquire political power and use it to distribute opportunities among clients. The second is exclusion. When resources become scarce, redefine the circle of people entitled to compete for them. The third is construction. Create alternative institutions and expand the stock of opportunities.

These responses are not unique to South Africa. They are universal possibilities confronting societies under institutional stress. But South Africa displays all three with unusual clarity. State capture and tenderpreneurship exemplify the first. Xenophobic mobilisation exemplifies the second. Private electricity, private security, residential self-provision and, in its most comprehensive form, Orania exemplify the third.

The distinction between them is fundamentally evolutionary. Every institutional order selects behaviours. A patronage system selects for connections. A scarcity system selects for political mobilisation around insider status. An open competitive order selects for productive performance. The question facing South Africa is therefore ultimately not simply which party governs or which policies are adopted. It is: What forms of behaviour does the institutional system reward?

The Reversibility of Regression

There is finally a reason not to end this analysis pessimistically. If modernity is reversible, regression is reversible too. Institutions are human constructions. South Africa still possesses extraordinary assets: sophisticated businesses, financial markets, universities, professional expertise, substantial infrastructure, an independent civil society, competitive elections and important elements of constitutional legality. Its private economy repeatedly demonstrates considerable adaptive capacity.

Eskom’s recent partial improvements demonstrate that even severely degraded public institutions can recover when incentives, management and operational discipline improve. The task is therefore not to reconstruct civilisation from zero. It is to restore the selection mechanisms that allow existing capacities to flourish. It means restoring competence as the principal criterion of institutional selection. It means strengthening property rights and predictable general rules. It means reducing regulatory barriers to entrepreneurship and investment. And it means permitting institutional alternatives to compete wherever monopolistic public provision fails.

15. Conclusion: Civilisation as a Continuous Achievement

South Africa’s post-apartheid history warns against one of modern society’s most complacent assumptions: that once the institutions of modernity have been established, they become permanent. They do not. Physical capital depreciates unless maintained. Human capital disappears unless reproduced. Institutional knowledge vanishes unless transmitted. Professional norms weaken unless enforced. Competitive markets contract when political allocation expands. South Africa is certainly not the only example of this kind of regression. Venezuela is another stark case, among many others, including the experiences of economic decline in most of the former European colonies after independence, the then so called Third World. South Africa’s decline is more recent and stands out because South Africa was always considered to be part of the First World.

However, the process is not irreversible. As Javier Milei has demonstrated in Argentinia, the reversal of this politically induced process of decline and regression is possible when the political direction is turned around once more, changing incentives, and initiating a positive feedback loop towards economic growth sustained by markets, libertarian institutions and discourses.

South Africa shows how private citizens, firms and markets can self-organize and reconstruct alternatives from below, exemplifying a successful spontaneous privatization process. The very weakness and declining capacity of the state gives – inadvertently – room to this emergent private order.

The broader lesson is not unique to South Africa. Impersonal competition under the rule of law is a rare and precious achievement. Patron-client systems are older, easier to reproduce and capable of re-emerging inside formally modern institutions. The task of a liberal society is not only to create markets and legal rules once, but continuously to preserve the conditions under which impersonal selection defeats political favour. Advanced, modern civilization is not an inheritance that can simply be taken for granted and passively consumed. It is a complex, improbable, process with many intricate, necessary institutional and cultural ingredients that must all be continuously reproduced. Such an intricate order – involving markets, the rule of law and the active monitoring and steering via open discourses - is inherently fragile and vulnerable to political disruptions, leading to historical regression. South Africa offers a stark and recent example of this tragic possibility.

The deepest lesson of South Africa is therefore the same lesson with which this inquiry began. Civilisation is not an inheritance that can simply be possessed. It is an activity. Infrastructure must be maintained. Knowledge must be transmitted. Professional competence must be selected. Trust must be reproduced. Rules must remain credible. Capital must be replenished.

Productive cooperation among strangers must continuously allowed to outperform the older temptations of status, patronage and group conflict. The historically exceptional achievement of the modern open order was to replace a large part of the struggle for political favour with impersonal competition under general rules. South Africa demonstrates how easily that achievement can begin to unravel. But it also demonstrates something else. When established institutions deteriorate, individuals do not simply stop cooperating. They adapt. They exit. They experiment. They build.

The critical distinction is between adaptations that intensify the struggle over a shrinking stock and adaptations that create new institutional capacity. Operation Dudula and Orania, interpreted at this level of abstraction, represent almost opposite answers to the same underlying anxiety. One says: There is not enough. Keep the outsider out. The other says: The existing institutions no longer suffice. Build alternatives.

The future of South Africa—and perhaps the more general future of liberal civilisation—depends upon which of these logics prevails. The route out of institutional regression cannot ultimately be the ever finer political allocation of scarcity among competing groups. It must be the reconstruction of an open order in which individuals once again have stronger incentives to produce than to capture, to cooperate than to exclude, and to build new opportunities rather than fight over inherited ones. The most precious institutional achievement is therefore not prosperity itself. It is the open competitive process capable of continually recreating prosperity.

References

Dassah, M.O. (2018). ‘Theoretical analysis of state capture and its manifestation as a governance problem in South Africa.’ Journal for Transdisciplinary Research in Southern Africa 14(1): a473.

Gebel, Titus (2018). Free Private Cities: Making Governments Compete for You. Aquila Urbis.

Giliomee, Hermann (2003). The Afrikaners: Biography of a People. Cape Town: Tafelberg; Charlottesville: University of Virginia Press.

Hayek, F.A. (1945). ‘The Use of Knowledge in Society.’ American Economic Review 35(4): 519-530.

IMF (2025a). South Africa: 2024 Article IV Consultation. IMF Country Report No. 25/28.

IMF (2025b). South Africa: Selected Issues. IMF Country Report No. 25/29.

Krueger, A.O. (1974). ‘The Political Economy of the Rent-Seeking Society.’ American Economic Review 64(3): 291-303.

MacCallum, Spencer H. (1970). The Art of Community. Menlo Park, CA: Institute for Humane Studies.

Mamdani, Mahmood (1996). Citizen and Subject: Contemporary Africa and the Legacy of Late Colonialism. Princeton, NJ: Princeton University Press.

Mathebula, N.E. and Masiya, T. (2022). ‘Governance Issues and State Capture at Eskom and Transnet: A Kingdon’s Multiple Streams Theory for Policy Studies.’ African Journal of Governance and Development 11(1.1): 146-163.

Mlambo, V.H., Zubane, S.P. and Thusi, X. (2022). ‘Interrogating Governance Challenges and Cadre Deployment in South Africa.’ Cihan University-Erbil Journal of Humanities and Social Sciences 6(1): 11-18.

National Treasury (2023). Budget Review 2023. Pretoria: Republic of South Africa.

North, D.C. (1990). Institutions, Institutional Change and Economic Performance. Cambridge: Cambridge University Press.

North, D.C., Wallis, J.J. and Weingast, B.R. (2009). Violence and Social Orders. Cambridge: Cambridge University Press.

Olson, M. (1982). The Rise and Decline of Nations. New Haven: Yale University Press.

Oomen, Barbara (2005). Chiefs in South Africa: Law, Power & Culture in the Post-Apartheid Era. Oxford/James Currey; New York: Palgrave Macmillan.

Ostrom, E. (2010). ‘Beyond Markets and States: Polycentric Governance of Complex Economic Systems.’ American Economic Review 100(3): 641-672.

Schumacher, P. (2024). ‘Thesis of German Right Wing Populism - A Rationale for the Protection of High-performance National Cultures.’ Substack.

Seekings, J. and Nattrass, N. (2005). Class, Race, and Inequality in South Africa. New Haven: Yale University Press.

Southall, R. (2007). ‘Ten Propositions about Black Economic Empowerment in South Africa.’ Review of African Political Economy 34(111): 67-84.

Tangri, R. and Southall, R. (2008). ‘The Politics of Black Economic Empowerment in South Africa.’ Journal of Southern African Studies 34(3): 699-716.

Tullock, G. (1967). ‘The Welfare Costs of Tariffs, Monopolies, and Theft.’ Western Economic Journal 5(3): 224-232.

Weber, M. (1978). Economy and Society. Berkeley: University of California Press.

Note: This paper was researched and written in collaboration with Claude Opus 5

Abstract

After nearly two decades of economic stagnation, this paper revisits the financial crisis of 2008 and the protracted paralysis that has followed it, arguing that both are best understood not as failures of free-market capitalism but as the consequences of pervasive state intervention into money, credit and finance. Drawing on the Austrian theory of the business cycle as developed by Mises, Hayek, Rothbard and Garrison, it contends that the suppression of interest rates by the central bank, the policy-driven misdirection of credit into housing, the regulatory enshrinement of the credit-rating agencies, and the accumulated moral hazard of the financial safety net together produced the synchronised cluster of malinvestments whose liquidation became the crisis. It reconstructs the mainstream Keynesian interpretation before arguing that it mistakes the symptom for the cause. It then argues that the interventionist response to the crash — bailouts, zero interest rates and asset purchases — converted a sharp, curative recession into a chronic stagnation by preserving the very malinvestments that needed to be cleared. The paper shows how the economic policy interventions coordinated across all major advanced economies has led to the zombification of these economies. The paper closes by setting out what a genuine escape from stagnation would require: sound money, a market-based rate of interest, and the restoration of the market’s self-correcting discipline.

1. Introduction: A Crisis Still Unresolved

The 2008 crash came as a shock to me. Our then high-flying business was disrupted and the bulk of my personal savings would have been lost if Royal Bank of Scotland had not been nationalized. This was the moment that turned me – a former Marxist – soon enough into a libertarian. The conversion was connected to my intense attempt to understand what had happened and how this could be explained. I soon discovered that so called “Austrian” economics, as propounded by Ludwig von Mises and Friedrich von Hayek, had the most convincing explanation of the crisis, and indeed: the current followers of Mises and Hayek had predicted and warned about the crisis, and their analysis ahead of time was indeed compelling. In the years since, it was again and again, the Austrians who were right and predicted the ongoing stagnation we are still stuck within today.

Indeed, more than a decade and a half after the financial system of the Western world seized in the autumn of 2008, the advanced economies have not experienced any growth trajectory commensurate with the technological advances. Some have hardly made any per capita GDP gains at all. The interval that followed the crisis has been variously labelled the Great Recession, the “lost decade,” and — with mounting justification — the Great Stagnation. Productivity growth has been anaemic, real median wages in much of the developed world have moved sideways, business dynamism has fallen, and the public balance sheets of nearly every major government have been progressively loaded with debt. Whatever was attempted in response to 2008, it did not restore the any vitality to the market economy. This paper argues that this outcome is a predictable consequence of the way the crisis was diagnosed and the way it was treated.

The dominant narrative of 2008 holds that an under-regulated, excessively free financial sector ran amok, that greed and deregulation produced a speculative mania, and that only the decisive intervention of central banks and treasuries averted a second Great Depression. On this telling, the lesson of the crisis is that capitalism is inherently unstable and requires an ever-tighter regulatory leash. The argument advanced here is the opposite. Drawing on the Austrian theory of the business cycle as developed by Ludwig von Mises, Friedrich Hayek, Murray Rothbard and Roger Garrison, this paper contends that the boom of the 2000s, the meltdown that ended it, and the stagnation that followed were each the product of pervasive state interventions into money, credit and finance — interventions that systematically suppressed the self-correcting mechanisms of the market and substituted for them a structure of perverse incentives.

The stakes of getting the diagnosis right are not merely historical or academic. The interpretation a society places on its crises determines the policies it adopts thereafter, and those policies shape the conditions under which hundreds of millions of people work, save and build their lives. If the lesson drawn from 2008 is that markets are inherently unstable and require ever-tighter control, then the trajectory is toward more intervention, more discretionary management of money and credit, and more socialisation of risk — which is to say, as far as the analysis put forward here implies, toward more of what produced the crisis. If, on the other hand, the lesson is that the crisis was the product of political interventions, then the path forward is the restoration of the market’s own discipline. The two readings point in opposite directions, and the long, ongoing stagnation that has followed the crisis is, this paper argues, the price of having chosen the wrong one.

That something has gone durably wrong with the advanced economies is no longer seriously contested across the political spectrum, even if the cause remains contested. Measured productivity growth has slowed very markedly compared with the post-war decades; the rate at which new firms are founded has fallen; real wages for large parts of the workforce have stagnated for years; and the ratios of public and private debt to output have climbed to levels that would once have been thought alarming, sustained only by interest rates held near or below zero for an unprecedented span. Asset prices, by contrast, have soared, conferring great gains on the owners of existing wealth while the productive economy has languished. This constellation — stagnant production alongside inflated assets, swelling debt alongside near-zero rates — is exactly what the Austrian analysis would predict from a regime that suppresses the interest rate and prevents the liquidation of malinvestment. It is the signature of an economy prevented to correct itself.

The perspective taken is explicitly libertarian in its analytical sympathies: it treats the market’s price and profit-and-loss mechanism not as a fragile contrivance requiring official supervision but as a robust discovery and coordination process whose chief enemy is the systematic distortion of its central signal — the rate of interest. The aim is not merely to re-litigate the events of 2008 but to show why understanding their true cause is the precondition for any genuine escape from the stagnation that has followed. As long as the diagnosis remains mistaken, the treatment will continue to deepen the malaise.

The paper proceeds as follows. Section 2 sets out the core of the Austrian theory of the business cycle. Section 3 applies that framework to the specific institutional distortions of the pre-2008 American and European financial systems. Section 4 reconstructs the mainstream account in its strongest form before offering a critique. Section 5 examines the policy response to the crash and argues that it converted a sharp, cleansing recession into a prolonged stagnation. Section 6 considers the political economy of intervention — why bad policy persists. Section 7 sketches what a genuine path out of stagnation would require. Section 8 turns to where capitalism still functions — the venture-capital ecosystem and the investment into artificial-intelligence. Section 9 examines the fiscal reckoning now unfolding on both sides of the Atlantic. Section 10 concludes.

2. The Austrian Theory of the Business Cycle

The Austrian business cycle theory (ABCT) is, at its heart, a theory about the consequences of falsifying a price. The price in question is the rate of interest, which in a functioning market coordinates the decisions of savers and investors across time. When that price is manipulated downward by the expansion of bank credit unbacked by real savings, the result is not faster growth but a systematic misallocation of capital that must eventually be corrected. The boom is the period of misallocation; the bust is the correction.

2.1 Interest, time preference and intertemporal coordination

In the Austrian view, the rate of interest is fundamentally an expression of time preference — the degree to which people prefer present goods to future goods. When a society chooses to save more, time preference has fallen: people are willing to defer consumption, releasing real resources that can be devoted to longer, more roundabout, more capital-intensive processes of production. The interest rate falls accordingly, and this lower rate incentivises more borrowing for investment, including longer term investments. The lower rate signals that resources are available for longer-term projects. The structure of production lengthens in a manner that is sustainable precisely because it is matched by genuine saving. Consumption foregone today funds the capital expenditure that yields greater output tomorrow. This is the coordinating function of interest, emphasised above all by Mises in his integration of monetary theory with the theory of capital.

It is worth being precise about the two rates whose divergence drives the theory: The natural rate of interest is the rate that would bring the supply of savings into equilibrium with the demand for investible funds in the absence of monetary disturbance — the rate that genuinely reflects the community’s willingness to defer consumption. The market rate, or loan rate, is the rate actually charged in the credit market, which under a regime of elastic bank credit can be driven away from the natural rate. In an unhampered market the two would tend to coincide, because any divergence would be arbitraged away. It is the institutional capacity of the banking system, underwritten by the central bank, to create credit without prior saving that permits the loan rate to fall persistently below the natural rate and so sets the cycle in motion. The interest rate, in other words, is not merely a price among prices; it is the price that governs the allocation of resources across time, and its falsification therefore corrupts not one market but all markets and the entire temporal structure of production.

2.2 Credit expansion and the artificial boom

The trouble begins when the interest rate is pushed below its natural level, when it falls not by an increase in saving but by the creation of new money and credit. Under a system of central banking and fractional-reserve commercial banking, the supply of loanable funds can be expanded without any corresponding act of new saving. The market rate of interest is driven below the natural rate — the rate that would equilibrate genuine saving with investment. Entrepreneurs, incentivised to invest and acting as if the artificially low rate was a signal of abundant real resources, embark on longer and more capital-intensive projects. But the resources to complete those projects have not in fact been released by consumers, whose time preferences are unchanged. The boom rests, as Rothbard put it, on an illusion: it sets in motion a structure of production that the real saving of the community cannot sustain. Projects are initiated that cannot be completed.

Crucially — and this is a point Mises and Rothbard stressed against their critics — ABCT is not an “overinvestment” theory but a theory of malinvestment. The problem is not only that too much is invested in the aggregate but that resources are drawn into the wrong lines of production: into projects whose viability depends on the continuation of cheap credit and on a structure of relative prices that cannot persist. At the same time, the same credit expansion that lowers the loan rate tends to foster overconsumption, as the new money percolates into wages, rents and profits and is spent in the accustomed proportions. The low rate of interest also implies the expansion of consumer credit. The economy is thus simultaneously pulled in two incompatible directions — toward more future-oriented production and toward more present consumption — with resources insufficient for both.

2.3 The inevitability of the bust

Because the boom is built on a misreading of the real availability of resources, it cannot last. As the new money works its way through the economy and the competition for genuinely scarce factors of production intensifies, prices rise, especially in the capital-goods industries that the boom called into being. The gap between the artificially low interest rate and the natural rate must eventually close, whether because the central bank, fearing price inflation, slows the credit expansion, or because the internal contradictions of the boom assert themselves. When it closes, the malinvestments are revealed for what they are: projects that should never have been started, capital structures that cannot be completed or profitably operated. The bust is the market’s recognition of this reality and its attempt to reallocate resources back toward sustainable uses. It is, in Rothbard’s words, a process of recovery — the restoration of an efficient economy and the ending of the distortions of the boom.

“The depression is actually the process by which the economy adjusts to the wastes and errors of the boom, and reestablishes efficient service of consumer desires. The depression is the recovery process, and the end of the depression heralds the return to normalcy and to optimum efficiency.” Murray N. Rothbard, America’s Great Depression

Two features of the bust deserve emphasis because they are routinely misunderstood. The first is that the corrective process, however unwelcome, is genuinely corrective: the bankruptcies and liquidations of the bust are not a gratuitous destruction of wealth but the mechanism by which mis-deployed resources are released and returned to uses that consumers actually value. To arrest this process is to preserve the misallocation. The second is that the Austrians distinguished the necessary primary adjustment — the liquidation of malinvestment — from what Rothbard, following earlier writers, called the secondary deflation or secondary depression, a self-feeding contraction of the money supply and collapse of confidence that can deepen and prolong the downturn beyond what the correction of malinvestment requires. This distinction matters for policy: it concedes that a catastrophic monetary implosion of the kind that occurred between 1929 and 1933 is genuinely harmful, while insisting that the remedy is to prevent monetary collapse, not to prevent the liquidation of the malinvestments that the prior boom created. The mainstream, having merged these two very different things into a single undifferentiated “depression” to be avoided at all costs, ends by resisting the very adjustment on which recovery depends.

Roger Garrison’s contribution was to render this theory in a form that engages directly with mainstream macroeconomics. In his capital-based framework, the economy is represented through the interaction of the market for loanable funds, the production possibilities frontier, and the “Hayekian triangle” representing the temporal structure of production. A sustainable expansion, driven by genuine saving, shows up as a movement along the frontier toward more future goods. An unsustainable boom, driven by credit expansion, pushes the economy beyond the frontier temporarily, only to force a contraction once the incompatibility between the lengthened structure of production and the unchanged pattern of consumer demand becomes manifest. Garrison’s achievement was to show that the Austrian story is not an outmoded relic but a coherent macroeconomics of capital, one that locates the source of instability precisely where the mainstream’s aggregate models cannot see it: in the intertemporal structure of production.

2.4 The pedigree of the theory: prediction, not hindsight

One feature distinguishes the Austrian theory from most of its rivals and lends it particular credibility: its proponents have repeatedly anticipated the crises that the mainstream failed to foresee. Mises is reported to have declined a prestigious post in the late 1920s with the remark that he expected a great crash to come, and Hayek, from his perch at the Austrian Institute for Business Cycle Research, warned in early 1929 of an approaching crisis as the consequence of credit expansion. The theory that explained the Great Depression after the fact had, in the hands of its leading exponents, predicted it beforehand. The same pattern recurred in the 2000s. A number of economists working in the Austrian tradition, observing the suppression of interest rates and the credit-fuelled inflation of housing, warned that the boom was unsustainable and would end in a bust — at a time when the prevailing professional opinion held that the moderation of the business cycle had become permanent and that financial innovation had dispersed risk safely across the system. A theory that sees the crisis coming has, at the very least, identified something its rivals have missed.

3. State Intervention and the Cluster of Errors

The Austrian framework directs attention to a specific question: what drove the systematic, economy-wide cluster of errors that became visible in 2007–2008? The answer is not that thousands of bankers, borrowers and investors simultaneously and spontaneously lost their reason, but that the institutional environment within which they operated had been so thoroughly mis-shaped by state intervention that reckless behaviour became the rational response to the incentives on offer. What follows is an inventory of those interventions, each of which can be understood as a distortion of the signals that would discipline behaviour in a genuinely free market.

3.1 The monetary engine: easy money and the suppressed interest rate

At the foundation of the boom lay monetary policy. In the aftermath of the dot-com collapse and the recession of 2001, the Federal Reserve drove the federal funds rate down to one percent and held it at historically low levels into the middle of the decade. This was the prime mover of the Austrian cycle: an interest rate pushed far below any plausible natural rate by deliberate policy, sending precisely the false signal that the theory describes. Cheap credit flooded into the most interest-sensitive and longest-duration sector of the economy — housing and real estate — inflating an asset bubble that was, in the Austrian reading, not a market failure but the entirely predictable consequence of the manipulation of money. That conventional consumer-price indices did not register alarming inflation is no refutation; the new money expressed itself as an unprecedented inflation of asset prices, above all house prices, which the standard indices largely omit.

The deeper point is institutional. A central bank with a discretionary mandate to manage the macroeconomy will, in the face of any slowdown, be under irresistible pressure to ease — to lower rates, to expand credit, to do something. The recession of 2001 produced exactly this response, and the easy money deployed to combat one downturn became the seed of the next, larger one. This is the interventionist ratchet or spiral in monetary form: each correction is met with stimulus, the stimulus sows the next round of malinvestment, and the next correction is met with still more aggressive stimulus. The 2000s did not represent a sudden lapse into recklessness but a phase in a long-running pattern in which the central bank, by its very nature, supplies the credit expansion that the Austrian theory identifies as the source of the cycle.

Monetary expansion supplied the fuel; a dense network of housing-directed interventions determined where it would flow and burn. Several mutually reinforcing policies channelled the artificially cheap credit into mortgage lending of progressively lower quality:

• Statutory and regulatory pressure on banks to extend credit into communities and to borrowers they would not otherwise have served, with lending standards relaxed in the name of expanding access to homeownership.

• The government-sponsored enterprises — Fannie Mae and Freddie Mac — which, operating under an implicit federal guarantee, purchased and securitised mortgages on a vast scale and progressively lowered the standards of the loans they were willing to absorb.

• State-sponsored securitisation that severed the originator of a mortgage from the ultimate bearer of its risk. When a lender can originate a loan, collect a fee, and pass the default risk down a chain of securitisation, the incentive to scrutinise the soundness of the house price and the borrower’s creditworthiness collapses.

• The prevalence, in much of the United States, of non-recourse mortgage lending. In many states – including California - a lender is not able to sue for the shortfall after foreclosure. This converts a mortgage into something close to a one-way option: the borrower captures the upside if prices rise and can walk away if they fall, making default a rational strategy. This is a dangerous pattern in combination with lowered lending standards and low or zero down payments.

Each of these features made sense to the individual actor responding to it. The originator who lowered standards with a view to passing on the mortgage, the borrower who took a mortgage he could service only if prices kept rising, the institution that loaded up on highly rated mortgage securities — none was behaving irrationally given the incentives. The irrationality was systemic, and it was manufactured by policy.

3.2 The ratings cartel and the laundering of risk

A particular role was played by the credit-rating agencies. The shelf-registration rules imposed by the U.S. Securities and Exchange Commission (SEC) effectively required investment-grade ratings for many public asset-backed securities (ABS) including mortgage-based securities (MBS) offerings. Rating had to come from an NRSRO — a “nationally recognized statistical rating organization,” such as Moody’s, S&P, or Fitch. Far from being free-market institutions, the principal agencies operated within a regulatory framework that granted them protected status: regulation referred to and effectively mandated reliance on the ratings of a small number of officially recognised firms, insulating them from competition and creating a captive demand for their product. These protected agencies awarded their highest, triple-A ratings to tranches of securitised subprime mortgages — ratings that turned out to be catastrophically wrong. Because regulation had inserted these ratings into the very definition of a “safe” asset, institutions across the world, from sovereign wealth funds to German state-backed banks, loaded their balance sheets with paper they believed to be of the highest quality. The mispricing of risk was not a failure of the market to assess risk; it was the consequence of a regulatory apparatus that had displaced market assessment with an official seal of approval.

3.3 Moral hazard: deposit insurance, capital tules and the expectation of rescue

Underlying the whole structure was a pervasive moral hazard, the cumulative result of decades of intervention. Deposit insurance, however well-intentioned, removes from depositors the otherwise natural incentive to monitor the soundness of the banks that hold their money, replacing the vigilant self-interest of millions of savers with the indifferent oversight of a few regulators. Banks were therefore only competing on the basis of the interest rates they were able to offer their depositors, not on their soundness. This pushed the banks toward more risky investments. More corrosive still was the accumulated expectation of rescue. A long series of bailouts — stretching back through the savings-and-loan episode, the Mexican and Asian rescues, and the bailout of Long-Term Capital Management — had taught the largest financial institutions that profits would be theirs to keep while the gravest losses would be socialised. An institution that expects to be rescued if its bets go wrong will make bigger and riskier bets. The too big to fail doctrine was not a description the market imposed on itself; it was an expectation created by the repeated practice of public rescue.

It is sometimes urged against the Austrian account that the extraordinary leverage of the financial institutions at the centre of the crisis was a market phenomenon, evidence of unregulated greed. The truth is more nearly the reverse. The capital adequacy framework under which banks operated — the international Basel rules and their domestic implementations — did not merely permit but actively encouraged the accumulation of precisely the assets that proved most toxic. By assigning low risk-weights to highly rated mortgage-backed securities and to the obligations of the government-sponsored enterprises, the regulatory framework made it capital-efficient for banks to hold mountains of such paper rather than to lend in ways the rules treated as riskier. Institutions economising on regulatory capital were responding rationally to a framework that told them these assets were safe. When the assets proved otherwise, the thinness of the capital cushion that the same rules had sanctioned turned losses into insolvencies. Here again the pattern holds: behaviour that looks like a market failure is, on inspection, a response to the incentives that regulation itself created.

Assembled together, these interventions answer the question the Austrian theory poses. The simultaneous, economy-wide cluster of malinvestments — the defining feature of a credit-driven cycle — had a common cause in the manipulation of money and the policy-driven misdirection of credit. As Rothbard observed, in a genuine free market with no central monetary authority there would be no reason for a great mass of entrepreneurs to err in the same direction at the same time; isolated mistakes would be made and quickly corrected by others alert to the opportunity. Only a systematic falsification of the interest rate, propagated through the whole structure of credit, can produce errors that are at once general, synchronised and persistent. That is exactly what the monetary and regulatory regime of the 2000s produced.

4. The Mainstream Account: Steelman and Critique

Intellectual honesty requires that the opposing interpretation be stated in its strongest form before it is criticised. The mainstream account of 2008 is not foolish, and it rests on real observations. This section sets it out as its ablest proponents would, then explains why the Austrian analysis is nonetheless to be preferred.

4.1 The strongest case for the mainstream view

The mainstream interpretation — associated with economists such as Paul Krugman, Joseph Stiglitz, Ben Bernanke and the broad Keynesian and New Keynesian traditions — runs roughly as follows. Financial markets are prone to episodes of mania and panic that are not the artefacts of government but intrinsic features of decentralised finance under uncertainty. The deregulatory trend of the preceding decades — the repeal of the separation between commercial and investment banking, the light-touch supervision of derivatives, the permissive treatment of leverage — allowed a shadow banking system to grow beyond the reach of prudential oversight. In this environment, rational individual behaviour produced collectively disastrous outcomes: securitisation chains that no one fully understood, leverage that amplified small losses into systemic insolvency, and a network of interconnections that turned the failure of a few institutions into a general panic. On this view, the run on the shadow banking system in 2008 was a modern version of a classic bank run, and the decisive intervention of the Federal Reserve as lender of last resort, together with fiscal stimulus, was precisely what prevented a repeat of 1929–1933. The recovery was slow, the argument continues, not because intervention was excessive but because it was insufficient — fiscal stimulus was withdrawn too early and monetary policy hit the zero lower bound, leaving a persistent shortfall of aggregate demand.

This is a serious argument. It correctly identifies the mechanics of the panic — the role of leverage, the fragility of short-term funding, the contagion through interconnected balance sheets. It is right that the immediate proximate event was a collapse of confidence in instruments whose risks had been badly understood. And it draws on a genuine historical lesson: the monetary contraction of the early 1930s did turn a recession into a catastrophe.

4.2 Why the Austrian account is to be preferred

The decisive weakness of the mainstream view is that it begins its story in the middle. It takes the mania, the mispriced risk and the reckless leverage as primitive facts about “financial markets” rather than asking why these phenomena appeared with such force, in the same sector, at the same time. The Austrian account supplies the missing first act. The leverage, the reach for yield, the willingness to hold mis-rated paper, the concentration in housing — these were responses to a monetary environment of suppressed interest rates and to a regulatory environment that encouraged and subsidised exactly the behaviour that later proved ruinous. To call the result a failure of the free market is to mistake the symptom for the disease.

Three specific objections tell against the mainstream story. First, the charge of “deregulation” is largely misdirected: the financial sector was among the most heavily regulated in the economy, and the specific interventions that mattered most — the monetary policy of the central bank, the housing-finance apparatus, the regulatory enshrinement of the rating agencies, the safety net — were not absences of regulation but active state-sponsored distortions. Second, the mainstream view has no satisfactory account of the cluster: it must treat the synchronised, economy-wide error as an inexplicable spasm of “animal spirits,” whereas the Austrian theory explains precisely why the errors were correlated. Third, and most tellingly, a number of Austrian and Austrian-influenced economists publicly warned of an unsustainable housing and credit bubble while the mainstream was, by and large, reassuring the public that fundamentals were sound. A theory that anticipated the event has a strong claim over a theory that was surprised by it and then explained it as inherent instability.

None of this denies that the panic of 2008 had the mechanics the mainstream describes. The point is that those mechanics were the form the correction took, not its cause. The bust was severe in proportion to the malinvestment of the boom, but the malinvestment of the boom was the work of intervention.

4.3 The witness of those who saw it coming

The claim that the Austrian camp anticipated the crisis is not a retrospective boast but a documented record, and it is worth naming a concrete instance, because the contrast with the prevailing professional optimism is so stark. Peter Schiff, an investment manager working squarely within the Austrian tradition and an adviser to Ron Paul’s 2008 presidential campaign, warned publicly and repeatedly, from 2006 onward, that the housing boom was an unsustainable bubble and that its collapse would bring on a severe recession. On national financial television in August 2006 he argued that an economy so dependent on consumption and borrowing, with too little production and saving, could not correct its imbalances without a recession, and that homeowners would watch their equity evaporate. At the end of 2006 he forecast that house prices, which had peaked the previous year, would come crashing back to earth. These warnings were delivered to a largely derisive reception: Schiff was brought onto the cable programmes, by the admission of commentators at the time, largely to be ridiculed as a perpetual pessimist talking down healthy American markets. He was vindicated within roughly a year. A framework that pointed to the right cause and the right mechanism ahead of the event deserves to be taken seriously against one that was caught unawares and then declared the event an intrinsic and unforeseeable feature of markets.

5. From Cleansing Recession to Engineered Stagnation

If the diagnosis of the crisis was mistaken, the treatment was correspondingly counter-productive. The Austrian theory yields an unambiguous prescription for the aftermath of a credit boom: the malinvestments must be liquidated, the misallocated capital and labour released to find sustainable employment, insolvent institutions allowed to fail, and the structure of production permitted to realign with the genuine pattern of consumer time preference. This is painful but it is also brief, and it is the necessary precondition for a healthy resumption of growth. The recession is the cure, not the disease.

5.1 The perverse policy response

The actual response was the reverse at every point. Insolvent institutions were rescued rather than resolved. The central bank drove interest rates to zero and held them there for the better part of a decade, then embarked on successive rounds of large-scale asset purchases that further suppressed the very price signal whose earlier distortion had caused the crisis. Fiscal authorities ran large deficits to sustain aggregate demand. Guarantees, subsidies and protective rules proliferated. In short, the authorities responded to a crisis caused by easy money and the socialisation of risk by providing yet more easy money and socialising risk on a still greater scale. David A. Stockman was rightly pointing to the contained nature of the crisis absent intervention, saying that “the bonfires would have largely burned out in the canyons of Wall Street.” Stockman’s image captures the Austrian counterfactual. The acute phase of the crisis was overwhelmingly a Wall Street phenomenon — a reckoning for the institutions that had made and intermediated the bad bets. The danger to ordinary economic life on “Main Street” was, on this reading, substantially exaggerated in order to secure public acquiescence in a vast socialisation of losses. Had the insolvent been allowed to fail, the damage would have been concentrated where the recklessness had occurred and would have burned out relatively quickly, clearing the ground for recovery.

By preventing the liquidation of malinvestments, the response preserved the very distortions that needed to be cleared. Capital and labour that should have been reallocated remained locked in unproductive uses. Zombie firms — enterprises that could service their debts only because interest rates were held near zero, but which could never repay or genuinely prosper — were kept alive by cheap credit, occupying resources and crowding out the new and the productive. The banking system, recapitalised but not cleansed, nursed impaired balance sheets rather than financing fresh enterprise. The suppression of interest rates that was meant to stimulate instead anaesthetised: it removed the pressure that forces the reallocation of capital, and it punished the saving on which sustainable investment ultimately depends.

The result was the Great Stagnation: a prolonged period of absent or low productivity growth, weak business formation, swelling public and private debt, and inflated asset prices that enriched the holders of existing assets while real economic dynamism withered. The recession that should have been sharp and curative was converted, by government intervention, into a chronic condition. The economy was not allowed to recover; it was placed on permanent life support, and the life support itself became a cause of the weakness it was meant to remedy. The interventionist spiral is self-reinforcing: each suppression of the corrective mechanism creates new distortions that are then cited as justification for further suppression.

The European experience after 2008 furnishes a confirming case rather than a counter-example. The euro itself functioned as a mechanism of credit distortion: by extending a common, low interest rate to economies of very different underlying conditions, monetary union channelled cheap credit into the periphery, financing booms in countries such as Ireland and Spain that mirrored, in miniature, the dynamics of the American housing bubble. When those booms collapsed, the European authorities responded as their American counterparts had — with bailouts of banks and sovereigns, with the suppression of interest rates, and eventually with large-scale asset purchases by the European Central Bank — and reaped a similar reward of prolonged stagnation. German state-backed banks, which had loaded their balance sheets with American mortgage securities precisely because the official ratings declared them safe, required rescue of their own. The transmission of the crisis across the Atlantic was thus not evidence of contagious market irrationality but a demonstration of how a common set of interventions — suppressed rates, official ratings, the socialisation of losses — produces a common set of distortions wherever it is applied.

5.2 The zombie economy: Markus Krall on the suppressed bankruptcy rate

The German financial analyst and entrepreneur Markus Krall has given this European variation its sharpest empirical formulation in the idea of the “zombie economy.” A zombie firm is an enterprise that does not earn enough to cover the servicing of its own debt and survives only because credit is artificially cheap — a business that, at any normal rate of interest, would already have failed. Krall’s central observation is that years of interest-rate suppression by the European Central Bank did not merely lower the cost of borrowing; they massively and artificially reduced the natural rate of bankruptcy. In a functioning market, a steady background rate of business failure is not a malfunction but the mechanism of renewal — the means by which capital, labour and market position are continually released from unproductive uses and made available to productive ones. By driving rates toward zero and below, the central bank switched off this cleansing process, allowing a growing population of unviable firms to persist. The result is precisely the stagnation this paper has described: zombie firms crowd the field, depress the average productivity of the economy, and bind the resources and the labour that healthy competitors would otherwise take up and utilize more productively.

This is not merely the contention of one heterodox commentator. The phenomenon Krall describes has been independently documented in the mainstream and central-bank literature. Studies by economists including Acharya, Eisert and co-authors found that the share of zombie firms among European companies rose substantially after the financial crisis as the ECB drove borrowing costs to record lows — by one widely cited measure from around four percent to seven percent of firms between 2012 and 2016 — and that this “zombie credit,” extended by weakly capitalised banks to impaired borrowers, both depressed productivity and impeded the very adjustment that monetary easing was supposed to produce. The Bank for International Settlements and researchers at the ECB itself have charted the same long rise in the zombie share across the advanced economies. That the institution responsible for the policy has documented the affliction it produced is a striking confirmation of the Austrian diagnosis.

Krall’s second point concerns what happened when even near-zero rates were no longer thought sufficient. During the Covid-19 emergency and beyond, the German authorities went further than suppressing the price of credit: they suspended the operation of insolvency law itself. Under ordinary German law the directors of a company that has become illiquid or over-indebted are obliged to file for insolvency without undue delay, and the failure to do so — trading on while insolvent, to the detriment of creditors — is a criminal offence, the offence of delaying insolvency. From March 2020 the obligation to file was suspended, and for over-indebted firms the suspension was repeatedly extended into the following years, with the going-concern prognosis periods loosened in their favour. Whatever its humanitarian rationale in an acute crisis, the measure amounted to legalising, for a whole class of firms, what is normally a crime: the continued operation of a fundamentally unsound business at the expense of those who deal with it. The natural rate of bankruptcy was not merely suppressed by cheap money but, for a period, abolished by decree. Independent observers, including bank economists with no Austrian commitments, noted at the time that the suspension allowed companies that would otherwise have failed to keep trading, drawing in further credit and contracts and crowding out their healthier rivals — the textbook mechanism of zombification, now produced directly by the law.

There is a final irony in Krall’s own biography that bears on the argument of this paper. Some years before he became known for the zombie-economy thesis, Krall attempted to launch a new European credit-rating agency, intended to break the effective monopoly of the established firms whose mis-ratings had done so much damage in the run-up to 2008. The venture failed before it could begin. As recounted in Section 3, the rating business is not a free market but a regulatory creation, in which a small number of officially recognised agencies enjoy a protected status that regulation itself confers and sustains; a new entrant must overcome not merely the ordinary difficulty of winning custom but the barrier erected by a framework that privileges the incumbents. The failure of an attempt to introduce competition into the ratings cartel is itself a small case study in the political economy examined in Section 6: the interventions that produced the crisis also protect the institutions at its centre from the discipline of competition that would otherwise reform them.

The deepest cost of the interventionist response is not measured in any single year’s lost output but in the steady accumulation of fragility. Each cycle of suppression leaves the economy more dependent on the suppression: debts contracted at near-zero rates cannot be serviced if rates return to historically normal levels, so the authorities find themselves unable to normalise without triggering the very reckoning they have spent years deferring. The financial system, the corporate sector and the public finances become jointly hostage to the continuation of cheap money. This is the trap that the advanced economies have built for themselves: having used intervention to avoid the correction of one boom, they have created a structure that cannot survive the withdrawal of intervention, and so the distortion compounds. The stagnation is not a steady state but a slow deterioration, and the longer the reckoning is postponed, the larger it grows.

5.3 Financial engineering, the mega-buyouts, and capitalism without capital markets

Nowhere are the distortions of the post-2008 regime more vivid than in the transformation of corporate finance that David Stockman has anatomised. Stockman — budget director under Reagan, later a partner in the private-equity business, and so a witness from inside the machine — argues that the central bank’s policy of ultra-low interest rates, reinforced by the implicit promise to rescue asset prices whenever they fall (the so-called Greenspan “put”, later Bernanke put), has diverted the energies of corporate America away from productive investment and into financial engineering. When money is nearly free and the authorities stand ready to backstop the markets, the rational course for corporate management is not to build plant and equipment but to extract cash from the balance sheet: to borrow cheaply and use the proceeds for debt-financed share buybacks, for cash-funded mergers and acquisitions, and for leveraged buyouts on an ever-grander scale. Each of these manoeuvres lifts the price of the stock without adding anything to the productive capacity of the enterprise.

The mega-buyout is the purest expression of this dynamic. A leveraged buyout loads a company with debt, extracts cash for its new owners and relies for its arithmetic on interest rates remaining low and credit remaining abundant. Such transactions are feasible at scale only because the central bank has suppressed the cost of borrowing and the markets believe it will continue to do so. Stockman describes the result as “rent-a-balance-sheet” speculation — a form of leveraged gambling that strip-mines existing enterprises rather than creating new wealth, and which is sustained by the artificial cheapness of debt. The proceeds of this equity extraction, moreover, accrue overwhelmingly to the holders of financial assets, who are concentrated at the very top of the distribution. Financial engineering thus operates, in Stockman’s memorable phrase, as the “cash machine of the prosperous classes”, while the productive economy that ordinary people depend on for jobs and rising wages is starved of genuine investment. The phenomenon described in the previous subsection — soaring asset prices alongside a stagnant real economy — has, in the buyout and buyback boom, a precise mechanism.

The scale of this diversion is not a matter of impression but of record. Annual share repurchases by the companies of the S&P 500 rose from roughly $150 billion at the turn of the century to some $589 billion at the previous cycle’s peak in 2007; they collapsed to about $138 billion in 2009 as credit froze, then climbed under the regime of near-zero interest rates to successive records — around $806 billion in 2018 and roughly $882 billion in 2021, with the rolling twelve-month figure brushing the trillion-dollar mark in early 2022. The pattern is precisely what the analysis predicts: repurchases collapse when credit is dear and balloon when the central bank makes it cheap, tracking the cost of money rather than any underlying surge in productive opportunity.

From this follows Stockman’s most arresting claim, and the one that bears most directly on the argument of this paper. If the dominant driver of securities prices is no longer the assessment of the underlying earning power of businesses but the anticipation of the central bank’s next move — if traders watch the policy announcement rather than the balance sheet, positioning for the next round of easing rather than weighing the fundamentals of an enterprise — then the financial markets have ceased to perform the one function that justifies their existence. The capital market exists, in the theory of a market economy, to allocate scarce investment capital to its most productive uses, rewarding the promising enterprise with cheap funding and denying it to the unpromising. A market that prices assets off the expectations of monetary policy rather than off the productivity of capital no longer performs this allocative work. It becomes, as Stockman puts it, a casino — a venue for speculation on the actions of the authorities rather than an instrument for the rational direction of investment.

This is not merely a rhetorical flourish; it is visible in the data on how stock-market returns have been distributed across trading days. Research by David Lucca and Emanuel Moench at the Federal Reserve Bank of New York identified what has come to be called the pre-FOMC announcement drift: in their post-1994 sample (1994 – 2011), a strikingly large share of realized excess U.S. equity returns was earned in the twenty-four hours immediately preceding scheduled Federal Open Market Committee (FOMC) policy announcements that are made 8 times per year. The excess return in that narrow pre-announcement window averaged close to four percent a year, against less than one percent on all remaining trading days. Similar effects appeared in major foreign equity markets in anticipation of the same U.S. announcements. Whatever the precise mechanism behind it, the finding lends empirical force to Stockman’s thesis: if so, much of the market’s realized excess return accrues in the hours of waiting for the central bank to speak, then modern equity valuation cannot be understood as the pricing of enterprise earnings alone, but must be understood as the pricing of enterprise under the shadow of monetary authority. In Stockman’s sharp terms this means the markets trade on FED announcements rather than fundamentals.

This is the deepest sense in which the present regime represents, in a striking phrase, capitalism without a capital market. The outward forms of capitalism persist — there are exchanges, securities, prices, traders — but the essential function for which capital markets exist has been hollowed out. The price signal that is supposed to guide the allocation of investment across the whole economy has been captured by, and now chiefly, or at least partially, reflects, the discretionary policy interventions of the monetary authority. This is the logical culmination of the Austrian critique. The original sin was the falsification of the interest rate; its ultimate consequence is the corruption of the entire apparatus of capital allocation, so that the market no longer transmits unadulterated information about the real productivity of investment but instead transmits guesses about the next official intervention. A system that has lost its capacity to allocate capital rationally cannot generate sustained, productive growth, however buoyant its asset prices may appear. The stagnation and the soaring markets are two faces of the same disorder.

It bears emphasising that this critique is not a complaint about capitalism but a defence of it. The financial engineering, the strip-mining of corporate balance sheets, the speculation on policy rather than fundamentals — none of these is a feature of a free market in capital. Each is a deformation induced by the intervention of the central bank into the price of credit and by the standing expectation of official rescue. A genuine capital market, operating with an honest interest rate and without the safety net beneath asset prices, would discipline the leveraged speculator, reward the productive investor, and direct capital by the test of real returns rather than the anticipation of monetary largesse. The remedy for capitalism without a capital market is not less capitalism but the restoration of the conditions under which a capital market can do its work.

------------------

6. The Political Economy of Intervention

A natural objection arises at this point. If the interventions described here are so damaging, why are they adopted, defended and expanded? The answer lies in the political economy of intervention, and it is here that the libertarian analysis is at its most penetrating. Each of the interventions catalogued above was rhetorically justified, and continues to be so justified, in the language of the public interest — protecting depositors, expanding homeownership, ensuring financial stability, safeguarding the ordinary citizen. Yet the systematic effect of conferring discretionary power over money and credit on public authorities is to create a prize over which concentrated interests will inevitably compete and gain control.

Murray Rothbard’s history of the institution at the centre of this story — the Federal Reserve itself — supplies the paradigm case. In his account, the Fed was not created in 1913 as a disinterested public guardian of monetary stability but as a cartelisation device for the banking industry. Competitive banking, Rothbard argued, places a natural check on any individual bank’s ability to inflate: a bank that expands credit too freely sees its notes and deposits returned for redemption and its reserves drained, and so the discipline of the market restrains the whole system. A cartel could in principle overcome this constraint by inflating in concert, but a private cartel is unstable, forever tempted by defection and undercut by new entrants. What the banks could not durably achieve among themselves they secured through the state: a central bank, armed with a monopoly of note issue and the capacity to act as lender of last resort, that could coordinate and backstop systematic credit expansion across the industry. The campaign to establish it, Rothbard showed, was advanced under the reformist banner of the Progressive Era — presented as a measure to tame the very financial panics that the existing, half-regulated system had produced — while being shaped behind the scenes, culminating in the famous gathering of bankers and officials at Jekyll Island, by the financial interests it would chiefly serve. The pattern is the template for everything that followed: an intervention sold as protection of the public, designed and captured by the concentrated interests it benefits, and justified by the instability that prior intervention had itself created.

The last point deserves illustration, for the panics that were invoked to justify the creation of the Federal Reserve were themselves in large part the product of a prior state intervention: the prohibition of branch banking. Across much of the United States, law confined banks to a single location — the so-called unit-banking system — forbidding them from opening the networks of branches that banks in other countries took for granted. The consequence was a fragile financial structure composed of thousands of small, undiversified institutions, each tied to the fortunes of a single town or county and unable to spread its risks across regions or industries. A local crop failure or regional downturn could ruin the local bank outright, and the fragmentation of the system meant that trouble at one institution readily propagated into a general run. The contrast with Canada, which permitted nationwide branch banking, is decisive: Canadian banks, diversified across the whole country, weathered shocks that felled their American counterparts in droves, and Canada passed through the Great Depression without a single bank failure while thousands of American unit banks collapsed. The lesson is the recurring one of this paper. The instability that was blamed on banking freedom and used to justify the vast new apparatus of the Federal Reserve was not a product of the market at all but of a specific legal restriction upon it.

6.1 Why more regulation is not the answer

The interventionist reflex after every crisis is to demand more regulation, on the assumption that the crisis revealed a gap in oversight that a new rule will close. The tradition of political economy known as Public Choice — which analyses political actors as self-interested agents responding to incentives, rather than as disinterested servants of the common good — explains why this assumption is generally mistaken. Regulation is not written on a blank slate by impartial guardians of the public interest; it is written through a political process in which the parties with the most at stake, the most information, and the greatest resources to deploy are the regulated firms themselves. The predictable result is what George Stigler, in the founding statement of the economic theory of regulation, called capture: as a rule, he argued, regulation comes to be designed and controlled by the industry it governs and operated primarily for that industry’s benefit. Captured rules, whatever their stated purpose, tend over time to entrench incumbents, raise the barriers facing potential competitors, and codify the assumptions of the regulated industry — including, fatally, its assumptions about which assets are safe. The credit-rating regime that declared subprime tranches to be triple-A was a regulatory construction, not a market one. The capital rules that concentrated risk in mortgage securities were regulatory constructions. The pattern suggests that the failure of regulation to prevent the crisis was not an accident to be remedied by more regulation but an intrinsic feature of the regulatory enterprise itself.

There is, moreover, a subtler cost. The very existence of a comprehensive regulatory apparatus generates a false sense of security — the belief that because supervisors are watching, the system must be sound. This belief dulls the vigilance that market participants would otherwise exercise on their own behalf. Why scrutinise the soundness of a counterparty, or the quality of a security, when regulators have certified the system and a safety net stands behind it? Regulation thus tends to crowd out precisely the decentralised, self-interested monitoring that in an unhampered market provides the real discipline. The deposit insurance that relieves the saver of the need to assess his bank, the official rating that relieves the investor of the need to assess his security, the supervisory apparatus that relieves everyone of the need to assess the system — each substitutes a centralised, fallible, capturable judgment for the distributed vigilance of those whose own resources are at risk. When the centralised judgment fails, as at length it must, it fails for everyone at once.

6.2 Concentrated benefits, diffused costs

Wherever the state acquires the power to direct credit, to guarantee liabilities, to confer protected status, or to rescue the failing, it creates an incentive for the intense and well-funded lobbying of those who stand to benefit. The benefits of intervention are concentrated on identifiable groups — large financial institutions, the housing-finance complex, the protected rating agencies — while the costs are diffused across the entire population of taxpayers, savers and consumers, none of whom has a comparable incentive to organise in opposition. The predictable consequence is that rules ostensibly written to protect the public are progressively shaped to serve the concentrated interests that the power to intervene calls into being.

Why the regulated industry should so reliably prevail was explained by Mancur Olson’s analysis of collective action. A small, compact group whose members each have a large stake in a regulatory outcome — a handful of major banks, say — can organise, inform itself and lobby effectively, because the benefit to each member justifies the effort and the group is small enough to overcome the free-rider problem. The general public, by contrast, is a vast group in which each member bears only a tiny share of the cost of any particular rule; for the individual citizen the stake is too small to repay the effort of becoming informed, so he remains, in Anthony Downs’s phrase, rationally ignorant. The asymmetry is structural and it runs one way: concentrated producer interests are organised and informed, diffuse consumer and taxpayer interests are neither. This is the same logic of concentrated benefits and diffused costs traced throughout this section, and it is why the political process, left to confer discretionary power over money and credit, tends systematically to serve the “regulated” interests rather than the public in whose name it acts. The broader Public Choice literature — the rent-seeking analysis of Gordon Tullock and Anne Krueger, and the constitutional economics of James Buchanan — generalises the point: wherever the state controls a valuable privilege, resources will be expended to capture it, and the capture will be dressed in the language of the public good.

There is a crucial distinction, too often elided, between being pro-market and being pro-business. The free-market libertarian is not an apologist for established business; he is frequently its sharpest critic, precisely because large incumbents are the chief beneficiaries of the interventions that suppress competition and socialise risk. The genuine market is a process that disciplines the incumbent through competition and the ever-present possibility of failure. It is the intervention — the bailout, the guarantee, the protected charter — that shields the incumbent from that discipline. To defend the market is therefore not to defend the interests of the powerful but to defend the conditions under which the powerful can be displaced.

Seen in this light, the response to 2008 was not an aberration but the logical culmination of the interventionist dynamic. A crisis produced by the socialisation of risk was met with a still greater socialisation of risk, to the direct benefit of the institutions whose recklessness had caused it. The general public, frightened into believing that the alternative was economic collapse, acquiesced. And the enlarged apparatus of intervention that emerged from the crisis stands ready to be deployed, and lobbied over, in the next one.

7. A Way Out of the Stagnation

If the stagnation is the product of suppressed correction, the path out of it is to permit the correction — to remove, deliberately and progressively, the distortions that prevent the market from coordinating saving, investment and production across time. This is not a programme of doing nothing; it is a programme of dismantling the structures that do harm. Its components follow directly from the diagnosis.

7.1 Sound money and a market-based interest rate

The foundation of any genuine recovery is the restoration of an interest rate that tells the truth about actual savings and consumers’ time preferences. This means ending the discretionary manipulation of money and credit by the central bank and allowing the rate of interest once again to reflect the availability of real rather than fictitious capital, i.e. the actual availability of resources - energy, raw materials, computation, labour – for productive investment freed up by the scaling back of immediate consumption implied by the actual level of savings. The Austrian tradition has long argued for sound money implying the abolition of the central bank’s power to expand credit at will, i.e. the denationalisation of money along the lines Hayek proposed in his later work, leaving the exact form money would take open. Initial ventures substituting for state-imposed money might include various competing crypto currencies, or commodity standards like electricity or gold, with the old fully convertible gold standard perhaps the most likely candidate for eventual monetary convergence, perhaps with full reserve and fractional reserve banking as variants. Whatever the precise institutional form, the principle is the same: money and credit must cease to be instruments of policy and become again the neutral, de-politicized, market-based medium through which the market coordinates itself.

The apparatus of guarantees, subsidies and rescues that sustains the zombie economy must be wound down. So long as failure is prevented, the reallocation of resources that recovery requires cannot occur. This means we must allow insolvent institutions to fail, ending the implicit and explicit guarantees that underwrite excessive risk, dismantling the housing-finance complex that misdirects credit, and removing the regulatory privileges — such as the enshrinement of protected rating agencies — that displace market assessment with official approval. The brakes and crutches must be kicked away, one by one, so that the economy can stand and move on its own. As Marx would have formulated: We must break the institutional fetters on the full development of the forces of production. These fetters are not, as Marx surmised, the capitalist institutions, but on the contrary, the anti-capitalist, quasi-socialist, interventionist institutions that pervade our state-regulated economies.

It would be dishonest to pretend that this path is painless. The liquidation of malinvestment is, by its nature, disruptive: it involves bankruptcies, the writing-down of inflated asset values, and the temporary displacement of labour and capital as they move from unsustainable to sustainable uses. But this pain is the pain of recovery, and it is finite. The alternative — the indefinite postponement of correction through ever-greater intervention — is not painless; it merely substitutes a chronic, grinding stagnation for a sharp adjustment, while steadily accumulating the debt and distortion that make the eventual reckoning larger. The choice is not between pain and comfort but between a short, curative pain now and a longer, deeper malaise later. After nearly two decades of stagnation, the case for grasping the nettle is stronger, not weaker, than it was.

7.2 Objections considered

The programme sketched here invites serious objections, and intellectual honesty requires that they be faced rather than waved away. The gravest is the spectre of 1929–1933: would not the liquidationist prescription, applied in the depths of a panic, risk a self-feeding monetary collapse of the kind that turned the recession of 1929 into the Great Depression? The Austrian reply, drawing on the distinction between primary and secondary deflation, is that preventing a catastrophic implosion of the money supply is not the same as preventing the liquidation of malinvestment, and that the two must be kept analytically distinct. One may consistently hold that the misallocated capital structure must be allowed to correct while also holding that a wholesale monetary collapse should be averted — most cleanly, on this view, by a monetary regime that is not prone to such collapses in the first place, rather than by discretionary rescue of insolvent firms. The objection has force against a crude liquidationism that welcomes monetary implosion; it has much less force against the position that the malinvestments, not the money supply, are what must be allowed to liquidate.

A second objection is distributional: the burden of a sharp correction falls heavily on workers who lose their jobs and households who lose their homes, none of whom made the reckless bets of the financial sector. This is a genuine moral concern, and it should be met directly rather than dismissed. The Austrian answer is twofold. First, the chronic stagnation that is the alternative also falls heavily on ordinary people — in stagnant wages, vanished opportunities and the slow erosion of the value of their savings — only less visibly and over a longer period, so that the comparison is not between hardship and comfort but between a sharp, acknowledged hardship and a diffuse, denied or obfuscated one. Second, to the extent that a society wishes to cushion the displaced during the adjustment, it can do so through direct and transparent assistance to persons, which is a very different thing from the indirect, opaque and regressive method of rescuing the institutions whose recklessness caused the harm. Bailing out the bondholder is not the same as helping the worker, and the conflation of the two has served the former at the expense of the latter.

A third objection holds that the modern economy is simply too complex and too interconnected to permit the failure of large institutions, whose collapse would propagate through the system with catastrophic effect. This is the doctrine of “too big to fail,” and the Austrian rejoinder is that the doctrine is largely self-fulfilling: institutions are permitted, even encouraged, to grow large and interconnected precisely because the expectation of rescue removes the market discipline that would otherwise check their growth and their interconnection. In a regime where failure was a real possibility, counterparties would price the risk of failure, creditors would demand prudence, and the very scale and entanglement that make rescue seem unavoidable would be curbed at the source. The interconnection that is offered as the reason for intervention is in substantial part the product of prior intervention. To take it as an immovable fact of nature is to mistake an artefact of policy for a law of economics.

7.3 More freedom, more responsibility

Underlying these specific measures is a general orientation: a willingness to risk more freedom and more self-responsibility. The interventionist settlement rests on the premise that the market cannot be trusted to govern itself and must be continuously managed from above. The argument of this paper is that the political management is itself the source of the instability, and that the market’s self-regulating mechanisms — the price of interest, the discipline of profit and loss, the vigilance of those whose own resources are at stake — are far more robust than the interventionist supposes, provided they are not systematically disabled. To risk freedom is to allow those mechanisms to operate. The long stagnation is, if nothing else, an argument that the alternative has been tried and found wanting. Time is running out.

8. Capitalism Where It Still Functions: Venture Capital and the AI Investment

The argument of this paper has been overwhelmingly a critique of intervention, and it would be easy to read it as a counsel of despair. It is not. The Austrian case rests on a confidence in the market’s creative powers that is the mirror image of its critique of intervention: wherever the distortions are lightest, the market’s capacity to allocate capital, discover the future and generate genuine wealth still shows through. The American venture-capital ecosystem is the clearest surviving example, and the recent bold surge of investment into artificial intelligence is the clearest current instance of it at work. Examining where capitalism still functions is not a digression from the argument but a demonstration of it: it shows what the rest of the economy has been prevented from doing.

8.1 Venture capital as a functioning capital market

Venture capital is, in an almost pure form, the thing that Stockman lamented had been hollowed out of the public equity markets: a genuine market in the allocation of investment capital to its most productive uses. A venture fund raises money from investors who understand they may lose it, deploys that money into young firms on the basis of a judgement about their future earning power, bears the loss when — as most often happens — the firm fails, and reaps the reward when occasionally one succeeds spectacularly. The discipline of profit and loss operates here as it is supposed to: capital flows toward expected productivity, failure is permitted and indeed expected, and no official backstop stands between the investor and the consequences of his judgement. It is capitalism performing its essential function of directing scarce resources toward an uncertain future, and it has been, over decades, the engine of a remarkable share of American innovation and productivity growth. That this ecosystem has continued to function while the broader capital market has been captured by the anticipation of monetary policy is itself instructive: it is precisely the corner of finance least touched by the guarantees, the too-big-to-fail doctrine and the regulatory privileging of incumbents that deform the rest.

Yet even here the reach of intervention is not absent, and the events of recent years show how the wider distortions intrude upon the one part of the system that still works. The collapse of Silicon Valley Bank in March 2023 — the largest American bank failure since 2008 — struck at the financial hub of precisely this ecosystem. SVB had made itself the banker to roughly half of all American venture-backed technology and healthcare companies, and its failure was itself a product of the interventionist cycle this paper has described: the bank had loaded up on long-dated government bonds during the era of suppressed interest rates, and when the central bank was forced to raise rates sharply to contain the post-pandemic inflation, the market value of those bonds collapsed, opening the hole that triggered the run. The distortion of the interest rate that had been the original sin of the whole story reached even into the balance sheet of the venture world’s principal bank. The failure froze venture debt, stranded startup deposits and, coming atop an already sharp contraction in venture funding as rates rose, curbed and muted investment activity across the sector for a considerable period. The lesson is double-edged: the venture ecosystem is resilient and genuinely productive, but it does not float free of the monetary distortions that afflict the rest of the economy, and when those distortions produce a banking failure it is the productive part of the system that suffers the collateral damage.

8.2 AI Investment: capital still seeking the future

Notwithstanding these headwinds, the surge of investment into artificial intelligence over the past few years is a vivid demonstration that the venture mechanism still works. Enormous sums of genuine risk capital — much of it from venture funds, corporate investors and private wealth rather than from the state — have flowed into a technology of extraordinary potential. This is exactly what a functioning capital market is supposed to do: identify an area of high expected productivity and channel resources toward it, accepting the substantial risk that many of the individual bets will fail. Whether or not the current wave of AI investment proves to be partly a bubble — and the Austrian would be the first to note that an environment of still-abundant liquidity can inflate even a fundamentally sound trend beyond its warranted extent — the flow of private capital into a high-potential frontier is the market performing its allocative function. It stands in instructive contrast to the zombie economy described earlier: here capital is being directed toward the new and the promising rather than being trapped in the sustaining of the unproductive.

But the AI build-out has now collided with intervention of a different kind, at the level of the physical infrastructure the technology requires. The construction of the data centres on which artificial intelligence depends has met with mounting political and regulatory resistance. Across the United States a wave of local moratoria and permitting freezes has delayed or blocked data-centre projects, with a substantial volume of announced investment held up by such actions; opposition has coalesced around the electricity and water demands of the facilities and their effect on local utility rates. At the state level, moratorium legislation has advanced in a number of jurisdictions, and at the federal level a bill has been introduced to halt large data-centre construction nationwide pending new safeguards. It is worth being precise about the character of this opposition, since it is genuinely bipartisan: resistance has arisen in both Democratic- and Republican-leaning areas, though it has found its most sweeping legislative expression — statewide permitting pauses and the proposed national moratorium — chiefly on the political left, where it fuses with environmental and anti-corporate concerns. Whatever its source, the effect is the same in Austrian terms: at the very moment the market is attempting to direct capital toward a high-potential technology, the political process is raising barriers to the physical investment that the technology requires. The allocative signal is being sent; the response is being obstructed.

9. The Fiscal Reckoning: The Limits of Reform Under Democratic Politics

The concluding movement of this paper concerns the hardest question it raises. If the diagnosis is correct — if the escape from stagnation requires sound money, the liquidation of malinvestment and, above all, the curbing of the state spending and borrowing that the whole apparatus rests upon — then why is the remedy not adopted? The events of the past two years, on both sides of the Atlantic, offer a sobering answer: even where the political will to retrench appears to exist, the structural forces analysed in Section 6 make genuine fiscal discipline extraordinarily difficult to achieve. The reckoning that this paper has argued cannot be indefinitely postponed is being postponed still, and the recent record shows how.

9.1 The American attempt: DOGE and the limits of the chainsaw

The most conspicuous recent attempt to curb American federal expenditure was the Department of Government Efficiency, the cost-cutting effort led in 2025 by Elon Musk. Its ambition was, on its own terms, heroic: an initial promise to cut some two trillion dollars from the federal budget, later revised down to one trillion and then to a small fraction of that. The effort was pursued with real energy — mass reductions in the federal workforce, the cancellation of contracts, the elimination of entire agencies. And yet it failed at its central purpose. Federal spending did not fall; it rose, by roughly six percent over the year, to around seven and a half trillion dollars, and the deficit remained close to one and three-quarter trillion. The federal workforce fell by some nine percent — the largest peacetime reduction on record — but total outlays climbed regardless.

The reason for the failure is precisely the one the Austrian and Public Choice analysis would predict, and it is worth stating plainly because it is so instructive. The discretionary payroll and contract spending that DOGE could attack is a small fraction of the federal budget; the overwhelming bulk of expenditure is mandatory — the great entitlement programmes and, increasingly, the interest on the accumulated debt — which runs on what one analyst aptly called policy autopilot, untouchable without legislation that the political process will not deliver. Cutting the visible bureaucracy, however dramatic, cannot close a gap driven by entitlements and debt service. Worse, the sharpest single pressure on the budget is now the interest cost of the debt itself, swollen by the very rate rises that the post-pandemic inflation forced. The chainsaw, wielded against the branches, left the trunk untouched; and the trunk is growing, because the debt compounds and the rates that service it have risen. DOGE’s failure is not a story of insufficient will but of the structural intractability of the problem: the spending that matters is the spending that democratic politics has placed beyond the reach of reform.

9.2 The German case: abandoning the debt brake

If the American story is one of a reform attempt defeated by structural forces, the German story is one of the abandonment of discipline altogether — and it is the more striking because Germany had been the very embodiment of fiscal rectitude. For years the German constitution had contained a “debt brake” (Schuldenbremse), a rule limiting federal net borrowing to a small fraction of national output, rooted in the ordoliberal tradition of sound public finance. In March 2025 that discipline was cast aside. A constitutional amendment exempted defence spending above one percent of output from the borrowing limit altogether and created a vast off-budget fund — some five hundred billion euros, more than a tenth of annual output — for infrastructure and climate spending outside the debt brake’s constraints, with a hundred billion of it earmarked for the green-energy transition. The rule that had defined German fiscal policy for a generation was, in effect, suspended by decree of the political class that had until recently defended it.

The manner of the abandonment confirms the political-economy analysis in fine detail. The leader who drove the change had, during the election campaign that preceded it, disavowed any intention to loosen the debt brake; the reform was pushed through the outgoing parliament precisely to avoid the arithmetic of the newly elected one, in which parties opposed to new borrowing would have held a blocking minority. And the promised discipline within the new borrowing proved illusory almost at once: an independent institute found that the great bulk of the new debt taken on in the first year — by one estimate some ninety-five percent of it — was used not for the additional productive investment that had justified the fund, but to plug gaps in the ordinary budget, exactly the substitution the reform’s own conditions had been written to prevent. Borrowed money nominally dedicated to the future was absorbed into current consumption. This is the interventionist ratchet in its fiscal form: a discipline erected against the temptations of deficit finance is dismantled the moment it binds, the dismantling is dressed in the language of investment, and the proceeds are consumed.

The German case carries a further irony that speaks to the deindustrialisation now afflicting Europe’s largest economy. The borrowing was loosened in part to finance an ideologically driven energy transition, even as the high energy costs flowing from that same set of green-energy priorities have been hollowing out the German industrial base that made the country prosperous. The state borrows more to subsidise a transition whose costs are simultaneously driving productive industry abroad. A clearer illustration of the self-reinforcing interventionist spiral — intervention generating the very distress that is then cited to justify further intervention — would be difficult to construct.

9.3 Something will have to give

Taken together, the American and German experiences of the past two years frame the choice with which this paper’s argument culminates. In the United States, an energetic attempt to cut spending foundered on the structural dominance of entitlements and debt service, and borrowing continued to grow. In Germany, the continent’s fiscal anchor abandoned its own discipline and borrowed more, channelling the proceeds into current consumption and an industrially ruinous energy policy. In both cases the pattern is the same: the political process cannot bring itself to curb expenditure, even under governments elected in part to do so, and even as the rising interest costs of accumulated debt tighten the vice. The contradiction that this paper has traced from 2008 forward is not being resolved; it is being deferred, and the cost of deferral compounds. Something will have to give. The only question that remains open is whether the adjustment will be chosen deliberately, while there is still room to choose, or imposed — by inflation, by a debt crisis, or by both — when there is not.

10. Conclusion

The crash of 2008 was not the failure of capitalism that the dominant narrative proclaimed. It was the failure of a system of pervasive intervention into money, credit and finance — a system that suppressed the interest rate, misdirected credit into housing, displaced market risk-assessment with official ratings, and socialised the losses of the reckless through an ever-expanding safety net. The Austrian theory of the business cycle, developed by Mises, Hayek, Rothbard and Garrison, explains why such a regime must produce a synchronised cluster of malinvestments, an unsustainable boom, and an inevitable bust. The mainstream account, for all its insight into the mechanics of the panic, mistakes the symptom for the cause and so prescribes more of the false medicine that weakened the economy.

The treatment administered after 2008 confirmed the diagnosis by its results. By preventing the liquidation of malinvestment, by holding interest rates at zero, by sustaining the insolvent and the unproductive, the authorities converted what should have been a sharp and cleansing recession into a chronic stagnation that has now persisted for the better part of two decades. The interventionist spiral continues to squeeze the vitality out of the advanced economies, and each turn of the spiral is justified by the very distortions that earlier turns created.

It is worth stating plainly what this argument does and does not claim. It does not claim that fraud and folly played no part in the events of 2008; they plainly did. It claims, rather, that fraud and folly are constants of human affairs, present in every era and every system, and that what requires explanation is not their existence but their sudden concentration and synchronisation — the fact that so many actors erred in the same direction at the same time. That synchronisation cannot be explained by an appeal to human nature, which does not vary from decade to decade; it can be explained only by a common signal that misled them all at once. The Austrian theory identifies that signal as the falsified rate of interest, and the institutional analysis identifies its source as the central bank and the apparatus of intervention built around it. Remove the false signal and the discipline of failure returns; folly remains, but it is once again punished rather than subsidised, isolated rather than generalised.

The advanced economies stand, more than a decade and a half on, at a juncture they have refused to confront. The malinvestments of the long boom have not been liquidated but preserved; the false signal that produced them has not been corrected but entrenched; and the stagnation that results has been mistaken for a new normal to be managed rather than a distortion to be cured. The super-low interest rates could not, in the end, be sustained: the consumer-price inflation that the Austrians had long warned would follow the monetary expansion finally took hold after the pandemic, and the central banks were forced to raise rates sharply to contain it. This has in turn severely squeezed state budgets, as the interest cost of financing current deficits and the accumulated debts of the stagnation years has risen steeply. The contradiction is, for the moment, still building rather than resolving — but it cannot build indefinitely. Something will have to give. Either public spending is curtailed — for beyond some point higher tax rates depress activity rather than raise revenue — or the pressure to inflate the debt away will reassert itself and prices will spike once more. The choice that was evaded in 2008 has not been abolished; it has merely been deferred to a moment of far greater constraint.

The libertarian and Austrian tradition offers a diagnosis that fits the facts and a remedy that follows from the diagnosis. The remedy is demanding and its short-run costs are real. But the alternative, now visible in the long record of stagnation, is not the avoidance of cost but its indefinite accumulation. The debate that the crisis should have opened — about sound money, honest interest rates, and a genuinely self-regulating capitalism — remains to be had. This paper is an argument for having it at last.

References

Acharya, V. V., Crosignani, M., Eisert, T., & Eufinger, C. (2024). “Zombie Credit and (Dis-)Inflation: Evidence from Europe.” Journal of Finance; earlier as Federal Reserve Bank of New York Staff Report No. 955 (2020).

Andrews, D., & Petroulakis, F. (2019). “Breaking the Shackles: Zombie Firms, Weak Banks and Depressed Restructuring in Europe.” European Central Bank Working Paper No. 2240.

Bagus, P. (2012). Austrian Business Cycle Theory: Are 100 Percent Reserves Sufficient to Prevent a Business Cycle? Quarterly Journal of Austrian Economics, 15(4).

Bordo, M. D., Redish, A., & Rockoff, H. (2015). “Why Didn’t Canada Have a Banking Crisis in 2008 (or in 1930, or 1907, or …)?” Economic History Review, 68(1).

Buchanan, J. M., & Tullock, G. (1962). The Calculus of Consent: Logical Foundations of Constitutional Democracy. Ann Arbor: University of Michigan Press.

Cato Institute (2025). Analyses of the Department of Government Efficiency and federal spending, Washington, DC.

Congressional Budget Office (2025). The Budget and Economic Outlook. Washington, DC.

Data Center Watch (2026). Reports on local opposition to data-center construction in the United States.

Downs, A. (1957). An Economic Theory of Democracy. New York: Harper & Row.

European Commission, Directorate-General for Economic and Financial Affairs (2025). “The Potential Economic Impact of the Reform of Germany’s Fiscal Framework.” Spring 2025 Economic Forecast.

Friedman, M., & Schwartz, A. J. (1963). A Monetary History of the United States, 1867–1960. Princeton, NJ: Princeton University Press.

Garrison, R. W. (1978). Austrian Macroeconomics: A Diagrammatical Exposition. Menlo Park, CA: Institute for Humane Studies.

Garrison, R. W. (2001). Time and Money: The Macroeconomics of Capital Structure. London and New York: Routledge.

Hayek, F. A. (1931). Prices and Production. London: George Routledge & Sons.

Hayek, F. A. (1933). Monetary Theory and the Trade Cycle. London: Jonathan Cape.

Hayek, F. A. (1976). Denationalisation of Money: The Argument Refined. London: Institute of Economic Affairs.

ifo Institute (2025–2026). Analyses of the use of Germany’s €500 billion special infrastructure fund and the debt-brake reform. Munich.

Krall, M. (2017). Der Draghi-Crash: Warum uns die entfesselte Geldpolitik in die finanzielle Katastrophe führt. Munich: FinanzBuch Verlag.

Krueger, A. O. (1974). “The Political Economy of the Rent-Seeking Society.” American Economic Review, 64(3).

Lucca, D. O., & Moench, E. (2015). “The Pre-FOMC Announcement Drift.” Journal of Finance, 70(1); also Federal Reserve Bank of New York Staff Report No. 512.

Mises, L. von (1912). The Theory of Money and Credit. (Trans. H. E. Batson, 1953.) New Haven: Yale University Press.

Mises, L. von (1949). Human Action: A Treatise on Economics. New Haven: Yale University Press.

Mises, L. von (1936). “The ‘Austrian’ Theory of the Trade Cycle.” Reprinted in The Austrian Theory of the Trade Cycle and Other Essays. Auburn, AL: Ludwig von Mises Institute.

Olson, M. (1965). The Logic of Collective Action: Public Goods and the Theory of Groups. Cambridge, MA: Harvard University Press.

Rothbard, M. N. (1963). America’s Great Depression. Princeton, NJ: D. Van Nostrand.

Rothbard, M. N. (1962). Man, Economy, and State. Princeton, NJ: D. Van Nostrand.

Rothbard, M. N. (1994). The Case Against the Fed. Auburn, AL: Ludwig von Mises Institute.

Salerno, J. T. (2012). “A Reformulation of Austrian Business Cycle Theory in Light of the Financial Crisis.” Quarterly Journal of Austrian Economics, 15(1).

Selgin, G. A., & White, L. H. (1994). “How Would the Invisible Hand Handle Money?” Journal of Economic Literature, 32(4).

Stigler, G. J. (1971). “The Theory of Economic Regulation.” Bell Journal of Economics and Management Science, 2(1).

Stockman, D. A. (2013). The Great Deformation: The Corruption of Capitalism in America. New York: PublicAffairs.

Schiff, P. D., with Downes, J. (2007). Crash Proof: How to Profit from the Coming Economic Collapse. Hoboken, NJ: John Wiley & Sons.

Schiff, P. D. (2012). The Real Crash: America’s Coming Bankruptcy. New York: St. Martin’s Press.

Tullock, G. (1967). “The Welfare Costs of Tariffs, Monopolies, and Theft.” Western Economic Journal, 5(3).

U.S. Federal Deposit Insurance Corporation (2023). Statements and materials on the failure of Silicon Valley Bank.

Woods, T. E. (2009). Meltdown: A Free-Market Look at Why the Stock Market Collapsed, the Economy Tanked, and Government Bailouts Will Make Things Worse. Washington, DC: Regnery.

White, L. H. (2008). “How Did We Get into This Financial Mess?” Cato Institute Briefing Papers, No. 110.

Abstract

This essay attempts a synthesis that may at first appear paradoxical: it brings into productive contact Oswald Spengler’s cultural-pessimist diagnosis of Western decline, articulated most sharply in ‘The Hour of Decision’ (1933) and prefigured in ‘The Decline of the West’ (1918/1922) with the radical libertarian, anarcho-capitalist project I have articulated in my ‘Pathways towards a Stateless World Society’ and in ‘Markets and Discourses – Politics after the Libertarian Revolution’. The contrast in philosophical outlook is stark: Spengler proponent of a cyclical conception of history, a fatalist, a critic of both democracy and capitalist money-rule, and the exponent of an aristocratic-Prussian ethic of duty and honour. I argue, however, that once Spengler’s diagnosis is detached from his cyclical metaphysics, and once his conflation of social structure with social stratification is corrected by Luhmann’s theory of functional differentiation, a deep convergence emerges. Spengler’s critique of levelling rationalism, of the worker-cult, of the parliamentary-journalistic class that has dissolved European society since the French Revolution, and his celebration of a Faustian character marked by realism, courage, duty, dignity, and commitment to a shared mission, can all be re-integrated into the libertarian project. What Spengler honoured as the highpoint of Western culture, embodied in true statesmen, great entrepreneurs, inventors, engineers, and leaders of industry, will not return through a Prussian-Caesarist authoritarianism. It can re-emerge, and in our historical moment can only re-emerge, within a libertarian order grounded in markets and discourses.

Motivation and Thesis

Every individual is confronted with an inevitable choice. It behoves him to know definitely where he stands, on the Right or the Left; otherwise the course of history, which is stronger than all theory and ideological dreaming, will decide for him.

— Oswald Spengler, The Hour of Decision, 1933

This essay is an attempt to appropriate Oswald Spengler’s erudition, insights, striking historical analogies, his trenchant critique of opportunistic party politics and his hard hitting anti-democratic/anti-socialist polemics, as well as his spirited promotion of embattled but enduring moral virtues for my libertarian project. Just to avoid any misunderstandings that might flow from the phrase “anti-democratic” from the start: As a libertarian with the conviction that the state must be cut back – ultimately to zero – not only is socialism my political enemy but I consider all forms of democracy that project state power to be highly problematic and – especially in the current era of post-Fordist network society - increasingly counterproductive. The more scope democracy is given the less individual scope for initiatives is left. I have no problem with democratic decision making in any voluntary association like a club, mutual aid society, cooperative, or political party. This also includes shareholder democracy in joint stock companies where voting rights are naturally proportionate to holdings. To be sure, given a certain level or scope of state power, I prefer a democratically constituted over an autocratic government. I thus fully appreciate the relative advantages of liberal democracies over autocracies or single party dictatorships. For me the question is – especially given the economic stagnation, fiscal deterioration and the erosion of trust in political institutions we find in most advanced liberal democracies – which political reforms and transformations allows advanced societies to regain their economic dynamism? For me that answer lies certainly not in the expansion of the scope of democracy. Quite the opposite must happen: The expansion of the scope for private, civil society initiatives of all kinds, including economic initiatives. In this sense, the roll back of the state necessarily implies the role back of democracy. In the context of this libertarian agenda, an astute critic of socialism and democracy like Oswald Spengler – a critic with a keen eye towards the pathologies of mass democracy who also understood the tendency of democracy to become increasingly interventionist and socialistic – can be a very useful resource.

With respect to Spengler’s question of knowing where to take once’s stance, on the Right or the Left, my position is clear: I am standing on the Right. However, this statement can be misleading. While I am sharing Spengler’s firm opposition to all left-leaning political projects, I am not sharing or endorsing Spengler’s conservative, traditionalist, nationalist position. To clarify this difference on the Right, we need to enrich the political compass with a second (vertical) axis that cuts across the (horizontal) Left-Right axis: We need to distinguish a libertarian Right, my position, from an authoritarian Right, Spengler’s position. The same axis distinguishes at one pole an authoritarian left, e.g. Stalinism, from a libertarian left, i.e. left-wing or communistic anarchism at the opposite pole. While I am not sure how far out on the authoritarian spectrum I would locate Spengler’s position, I can say with consolidated confidence that my own position – anarcho-capitalism as a long-term goal – is located at the extreme end of the libertarian spectrum. That this position – however rare it is at the current moment – is not a mere utopian dream but a plausible prospect at a time of rapid technological and indeed political transformations has been argued in depth in my ‘Pathways to a Stateless World Society’ (Schumacher, 2024). Despite the fact that my espoused political position is (currently) an outlier, I do not feel politically isolated. There are many political positions, movements and parties in the UK and in Europe – including the Conservative Party (which has moved to the right in opposition), Reform UK, Restore Britain, Advance UK, the UK Libertarian Party (LPUK) in Britain, the Alternative for Deutschland (AfD), the Free Democratic Party (FDP), WerteUnion, and Atlas Initiative in Germany, and many other European right-of-centre parties – that all share the same direction of rightward travel towards economic liberalisation, lower taxes, less regulation, and profess a willingness to return more scope to civil society, up to a point. We are currently still some distance away from this point where this shared rightward direction will fray and bifurcate into a somewhat statist-authoritarian versus a more libertarian direction. Until this point has been reached all political currents moving rightward from the position where our left-liberal democracies are currently stuck, can draw insights, inspirations and incisive polemics from Spengler’s ‘Hour of Decision’.

What we can learn from Spengler, as well as from other conservative critics of democracy like Carl Schmitt, is that the original strength of the parliamentary political systems – the utilisation of rational discourse within parliament (as was emphasized by Edmund Burke in 1774), embedded within a public sphere where debates were conducted by an educated, reading public of engaged and accomplished civil society stakeholders – had been increasingly perverted, due to the rapid expansion of voting rights. This together with the simultaneous expansion of state power and intervention implied that the political process became increasingly susceptible to demagogy and corruption. In his 1923 book ‘The Crisis of Parliamentary Democracy’ Carl Schmitt lists the deficiencies and pathologies of modern mass democracy at that time as follows: “the dominance of parties, their unprofessional politics of personalities, the government of amateurs, continuing governmental crises, the purposelessness and banality of parliamentary debate, the declining standard of parliamentary customs, the destructive methods of parliamentary obstruction, the misuse of parliamentary immunities and privileges by a radical opposition which is contemptuous of parliamentarism itself, the undignified daily order of business, the poor attendance in the House”(Schmitt, 1923/1988, p.19). Schmitt argued that the principles and rules underlying the classical 19^th Century parliamentary systems and which are still preserved in all modern constitutions – the independence of the representatives and the publicity of the proceedings – make sense as safeguarding the parliament’s deliberative function in discursive exchange with an informed public – but make no longer sense in the 20^th Century situation of universal suffrage mass democracies. Schmitt further elaborates:

“The impression based on long familiar observations has gradually spread: that proportional representation and the list system destroy the relationship between voters and representatives, make fractions an indispensable means of government in parliament, and make the so-called representative principle (article 21 of the Reich constitution that “the members are representatives of the whole people, they are only responsible to their own consciences and not bound to any instructions”) meaningless; further, that the real business takes place, not in the open sessions of a plenum, but in committees and not even necessarily in parliamentary committees, and that important decisions are taken in secret meetings of faction leaders or even in extra-parliamentary committees so that responsibility is transferred and even abolished, and in this way the whole parliamentary system finally becomes only a poor facade concealing the dominance of parties and economic interests.” (Schmitt, 1923/1988, pp.19-20)

One hundred years on these same pathologies evidently persist. Then and now, a situation persists in which we have both too much democracy in the sense of too much scope for political control of society and too little democracy in the sense of genuine parliamentary debate and public discourse about the areas that are supposedly still meant to be democratically controlled. These too pathologies amplify each other in a vicious spiral.

Both Spengler and Schmitt agree that 20^th Century universal suffrage parliamentarism, especially since socialist parties have been soaring politically, are plagued by demagogy. What surprised me in reading Spengler was the discovery that the kind of aggressive, moralising ideological phenomenon we are today plagued by and call woke leftism - with its defamation of all who fail to genuflect in front of its anointed group - did not begin in 2010, nor in the 1990s when it was called political correctness, nor even in 1968 as many surmise but was already rampant 100 years ago and started first over 200 years ago around the French revolution and then in a more generalised form around the revolutions of 1848. We are not talking about the particular ideology of wokeness with its particular anointed group but about a strikingly similar moralising political process, with many equivalent features and dynamics.
What we also learn from Spengler is that this phenomenon is not a purely ideological phenomenon, but connected to power struggles with materialist rather than idealist motivations of the protagonists. We also learn about some of the material mechanisms and political power dynamics that are at play and that very similar, structurally equivalent or at least comparable dynamics were already at play in the final decades of the Roman Republic.

For Spengler, the hour of decision is the hour of the terminal decline of the West brought about by the dissolution of the Western societal order and virtues through what he calls the White Revolution, beginning with the Jacobins radical egalitarianism and reign of terror and – in Spengler’s time - reaching its up to then most extreme phase in Russia’s bolshevism. But he also saw 19^th Century and 20^th Century Western parliamentarism, increasingly pushed and determined by socialist parties, in particular with respect to the Weimar Republic, as a factor and symptom of Western decline, weakening the West internally and externally. He, in my view rightly, attributed the 1929 global financial crash and the Great Depression that followed to these egalitarian tendencies driving state interventionism. Spengler saw this demagogic-egalitarian interventionism as another – compared to Lenin’s bolshevism perhaps less crude - form of power grab of the demagogic-revolutionary political class. Spengler’s intuitive sense of the ideological, cultural and political dynamics that led to the Great Depression was confirmed in the precise economic analysis and explanation of the Great Depression articulated by Mises, Hayek, and later Rothbard. The broader outlook upon the dangerous tendencies Spengler identified was then more fully formulated in Hayek’s famous 1944 book ‘The Road to Serfdom’. However, Spengler does – amidst this gloomy pessimism – allow for a faint flicker of hope for Western civilisation: He thought that the German nation who as the only of the great European nations had not yet played a world historical role, had not yet exhausted and dissolved all its slumbering potentials for Faustian vigour, missionary zeal, greatness and will to lead.

Our current economic and political situation shows some parallels to the situation Spengler was confronting. Since the 2008 global financial crash and the Great Recession that followed, the economic dynamism unleashed by the Thatcher-Reagan neo-liberal revolution - allowing the momentous, computational technology induced post-fordist socio-economic restructuring to deliver a new, highly innovative high-productivity economy – experienced a major setback, leading to a resurgence of state interventionism and by now nearly two decades of stagnation, with no revitalisation in sight. The 2008 crash – just like the 1929 crash – was itself already a result of increasing state interference (since the early 2000s) blocking the vital self-regulation and self-correction mechanisms of market economies. This period of policy-induced stagnation was further acerbated by a resurgent Western leftism pushing the political class and itself pushing into the political class, as well as into the public sector, education, NGOs and the media, scaring business into the adoption of counter-productive DEI processes, cumulatively leading to the woke take-over of society. That living standards are stagnating or only slowly rather than precipitously sinking is due to the fact that technology-induced productivity gains in many sectors are still happening. These gains - fritted away and consumed in the political alimentation machine – are so far preventing the outright collapse of the advanced economies. At the same time the political alimentation machine is relying on increasing deficit spending and an ominous and inflationary build-up of sovereign debt. Inflation is further eroding the incomes and savings of all who are still self-reliant and prudent rather than falling back on alimentation. The predictable backlash against these paralysing tendencies is now hitting Western Europe via the growth of right-wing populist parties, now in many key nations like the UK, France and Germany at the cusp of taking power. The problem with this right-wing populism is that here the leadership’s common sense libertarian positions are increasingly swamped by the more popular statist, nationalist factions within these new parties.

The situation in Germany is particularly dramatic. The exclusion from power of the right-of-centre Alternative for Deutschland (AfD) – now by far the highest polling party across the whole of Germany – implies that the woke left can continue to dominate government, with the result of rapid de-industrialisation and the simultaneous expansion of the welfare state, also due to uncontrolled immigration into the welfare system. As I have argued in my 2024 substack article “Thesis on German Rightwing Populism”, the German economic powerhouse is seriously at risk, and that, as matters stand politically in Germany, the AfD might be the only force that can save this highly evolved but endangered prosperity engine, one of very few specimens of this distinctive, high-trust, high-performance species. As I formulated then: “Even today, there is only a handful of these engines in full flourish. Germany has been, and probably still is or can be, one of them, and it would not be in the long-term interest of world society to let Germany, or any of these precious engines be fractured, slowed or compromised in its further development potential” (Schumacher, 2024).

My thesis is that the decisive political battle in our current ‘hour of decision’ is not only a battle over fundamental institutional reform, but simultaneously and inseparably a battle over the cultural conditions that determine whether a libertarian civilisational renewal is possible. The right-wing populist backlash against the woke-administrative apparatus, however necessary as a signal of civilisational resistance, will not by itself deliver the required transformation: without a clear libertarian direction, it risks substituting a nationalist-statist governance for the progressive-statist governance it opposes, leaving the productive substrate of advanced civilisation no more liberated than before. What is required alongside the economic and institutional argument is the cultural recovery of a certain type of human being — the founder, the entrepreneur in the original sense as risk-bearer, the institution-builder, the discourse-leader, as well as, more generally, the self-reliant worker who accepts full existential responsibility for him/herself and his/her family. This is the specific perhaps hitherto insufficiently considered aspect I would like to draw into our libertarian discourse via my reading of Spengler: the importance of not only the economic but also of the cultural, moral and characterological implications of the contested political directions we are confronted with in this new hour of decision. Here lies Spengler’s strength, in his analysis of the forces that undermine virtue and character, and in his insistence that character matters.

However, my key thesis – contra Spengler – is that only a libertarian revolution that rolls back the state and the emergence of a truly libertarian order can, under current technological and historical conditions, prepare the soil upon which men and women imbued with Spengler’s Faustian virtues can grow and flourish. To break through this vicious cycle of corrupting institutions and corrupt characters we need both determined entrepreneurs who persist with their innovative ventures even in the face of our current political conditions, as well as true political leaders – not career politicians – who against the odds keep working towards the libertarian revolution, inspiring and building an organised libertarian movement packed with committed entrepreneurial characters. In this matter of character building, the polemic force and indeed pathos we find in Spengler is not a liability but can be an asset, an asset that libertarianism can appropriate, learn from and build upon. Big projects like the libertarian revolution or the founding of a network of free cities require not only a sharp, hard-nosed intellect, but a generous character with big aspirations, and an indefatigable energy that can only be mobilised by a big and deep inspiration.

The odds can change fast. We are living in revolutionary times. The hour of decision is approaching. It will pay to be prepared.

Historical Context of Spengler’s ‘Hour of Decision’

The book I am bringing to bear here is the relatively short but impressively intense, spirited and often polemic book entitled “The Hour of Decision – Germany and World-historical Evolution”, in German: Jahre der Entscheidung – Deutschland und die Weltgeschichtliche Entwicklung. The book is an expansion of a lecture Spengler gave in 1929 under the title “Deutschland in Gefahr” (Germany in danger). This was also the originally planned title for the book. It had been fully written before Hitler took power in early 1933, and therefore does not take account of this most momentous event. In fact, the bulk of the book was already printed, so Spengler decided to publish it without revisions. However, he changed the title to avoid an otherwise inevitable misunderstanding. The book was then released with the new title in August of 1933. The English translation was published in 1934.

Spengler was not a supporter of the National Socialists and they in turn did not like this book. While Spengler did not shed tears over the demise of the Weimar Republik which he detested, and while he thought a new Caesar - then and now called upon to end the chaos of the late republic - was required to take over the reins, he did not believe that a mass movement and party like the NSDAP could play the required role, nor did he think Hitler can be the new Caesar. Here is what Spengler wrote in the new introduction to ‘The Hour of Decision’ he crafted after the Hitler take over:

“I shall neither scold nor flatter. I refrain from forming any estimate of those things which are only just coming into being. … The seizure of power took place in a confused whirl of strength and weakness. I see with misgiving that it continues to be noisily celebrated from day to day. … In any movement with such a beginning there are many possible developments of which the participants are not often fully aware. The movement may become rigid from excess of principles and theories; it may go under in political, social, or economic anarchy, or it may return to where it started without results. … The intoxication of the moment, which often ruins coming possibilities at the outset, is usually followed by disillusionment and uncertainty as to the next step. Elements come into power which regard the enjoyment of that power as an event in itself and would fain perpetuate a state of things which is tenable for moments only. Sound ideas are exaggerated into self-glorification by fanatics, and that which held promise of greatness in the beginning ends in tragedy or comedy.” (Spengler, 1934, pp.19-20)

One more note of introduction with respect to my reading of Spengler’s ‘The Hour of Decision’: Spengler’s book places a lot of emphasis on geo-politics. His thesis here is that this is not only the age of revolutions but the age of world wars, which will go on until a world dominating power has been able to establish itself, an Imperium Mundi as planetary equivalent of what the Imperium Romanum was for the Mediterranean world. (The key contenders for him were Germany, USA, Russia, Japan). He was also very prescient in stating in 1933 that the second world war was imminent. However, while I do not dismiss the continuing importance of geo-politics – how could I at a time when the contest for world dominance is, once more, on with tense global confrontations and wars – this geo-political theme is not the topic of my engagement with Spengler here. While Spengler sees geo-politics – understood as the struggle of political units for dominance and domination – as primary and political priority that motivates internal political and economic organisation as organisation of externally projected strength, this emphasis is decidedly not taken up here in my reading. My focus here, as libertarian, is on forms of societal order with a view towards prosperity. For me prosperity, not power, is what ultimately drives (and should drive) politics and history. (While I believe that anarcho-capitalist islands and networks of such islands can emerge, persist and expand within a world of states and power blocks, even with wars flaring up around the world as we witness today, a long phase of basic geopolitical stability would be a requirement for the gradual emergence of a stateless world society as envisioned in my ‘Pathsways’.)

An Improbable Affinity

Spengler and the anarcho-capitalist libertarian would, on a first encounter, seem to belong to different intellectual universes. Spengler regarded Western civilisation as entering its winter phase, its springtime of Faustian striving long behind it; he saw the rule of finance and parliamentary liberalism as themselves symptoms of decline, to be overcome by a final Caesarist age. The libertarian project, by contrast, is constitutively progress-oriented. As I have put it in my ‘Pathways towards a Stateless World Society’: The system of increasingly interventionist states has become a historical fetter on the further development of the productive forces. World society’s mode of material reproduction has dramatically shifted in recent decades due to the advance of computation and telecommunication which in turn transformed forms of economic organisation and work. These relatively recent historical conditions invite and indeed necessitate a radical liberalisation as a pre-condition for the full exploitation of the technological and economic potentials of world civilisation. (Schumacher,2024, p.259)

Spengler is a pessimist and fatalist; the libertarian, on the materialist register I have argued for, is a guarded optimist. And yet the closer one reads Spengler’s political diagnoses, especially in ‘The Hour of Decision’, the more familiar much of it sounds to anyone who has been making the libertarian case against our all-controlling state. As we shall see below, Spengler is vehemently opposed to wrest economic decisions from owner-entrepreneurs to bring the economy and its key parameters under the control of party politics and programmes. Consider Spengler’s portrait of the utopian rationalist or revolutionary intellectual who has become the dominant cultural type of late modernity:

“This doctrinaire clinging to theory for lack of experience, or rather this lack of ability to make experience, finds literary expression in a flood of schemes for political, social, and economic systems and Utopias, and practical expression in that craze for organization which, becoming an aim in itself, produces bureaucracies that either collapse through their own hollowness or destroy the living order.” (Spengler, 1934, p.29)

One could lift this passage almost unchanged into a contemporary libertarian polemic against the current left-utopian schemes like Universal Basic Income, The Green New Deal, against the ESG and DEI industries, against NGO complex as well as against the proliferating compliance bureaucracies that demand the production of mountains of documentation. The “craze for organization” that produces bureaucracies which destroy the living order is exactly the contemporary state’s war on the productive economy. Spengler’s targets and the libertarian’s targets are very often the same targets, observed from different angles. What differs is the proposed remedy.

Spengler’s trenchant anti-rationalist polemic and his respect for the historically evolved societal order anticipated Friedrich von Hayek’s critique of what he termed “constructivist rationalism”. Hayek’s target is the Cartesian/Hobbesian/Rousseauian tendency to treat society as if it were an artefact: a machine, a building, or an organisation that can be made according to a blueprint. Against this, Hayek insists that many of the most important institutions—language, law, morals, money, markets, property, common-law rules—are results of human action but not of human design. In ‘Law, Legislation and Liberty’(1973-79), Hayek defines the error very directly: constructivist rationalism assumes that social institutions are, and ought to be, products of deliberate design. He says this is false both factually and normatively: factually, because institutions such as law, money, language, and markets often evolved; normatively, because making the whole social order dependent on design would suppress the dispersed knowledge that only decentralised interaction can use. Rules of conduct may embody experience that no one can fully articulate. Hayek repeatedly argues that we often follow rules before we can explain them. In ‘The Fatal Conceit’(1988), he says reason itself is not the originator of civilisation but one of its products: human beings became rational through traditions, habits, and moral rules that made extended cooperation possible. However, as Hayek notes, constructivist rationalism is attractive because it flatters intellectuals: it promises visible purpose, moral control, and direct responsibility for outcomes. The market order, by contrast, is abstract, impersonal, and often opaque. Reading Spengler I was impressed to note that Hayek’s insights and critique of constructivist rationalism (that I had long since made my own) can be interpreted as an elaboration of insights which Spengler had already articulated earlier.

Two Diagnoses of the Same Pathology

Spengler, writing in 1933 in the shadow of the Weimar collapse, perceived a continuous historical curve of European decay running from the French Revolution forward. He read it through Roman parallels: the Gracchan revolution, Catiline’s demagogy, the urban plebs fed by “panem et circenses”, the eventual transition from the Republic to the Caesarist Imperium. He understood the European nineteenth and twentieth centuries as the analogue of the late Republic’s death-throes:

“For we live in a mighty age. … It corresponds to the Classical Age from Cannae to Actium, to the age illumined by the names of Hannibal, Scipio, and Gracchus, Marius, Sulla, and Caesar. The World War was but the first flash and crash from the fateful thundercloud which is passing over this century. … The coward’s security of 1900 is at an end. Life in danger, the real life of history, comes once more into its own. Everything has begun to slide, and now only that man counts who can take risks, who has the courage to see and accept things as they are.” (Spengler, 1934, p.36)

Spengler’s most devastating passage in this register comes when he draws up the balance sheet of what he calls the White Revolution — the long, low, continuous victory of the parasitic political-bureaucratic class over the productive substance of European civilisation:

“But who, then, really won the World War? … The labour leader won the War. That which in every country is called the Labour party and the trade union, but is in reality the trade union of party officials, the bureaucracy of the Revolution, gained the mastery and is now ruling over Western Civilization. … The sack of society has been accomplished. It was the reward of the mercenaries in the class war. The natural centre of gravity of the economic body, the economic judgment of the real experts, was replaced by an artificial, non-expert, party-political one. The equilibrium was destroyed and the structure collapsed.” (Spengler, 1934, p.138)

Strip away the cyclical metaphysics and consider what Spengler is actually describing. He describes a political class that has detached itself from any productive function and lives off the manipulation of mass opinion; a parasitic bureaucracy of party officials and revolutionary functionaries who have captured the state apparatus; the replacement of “the economic judgment of the real experts” who take real personal responsibility, staking both their reputation and their property by the decisions through “artificial, non-expert, party-political” experts who are always shielded from the material consequences of their decisions; the deliberate ideological war waged against rank, property, achievement, and competence by an alliance of urban intellectuals and welfare-dependent masses.

Now consider the libertarian diagnosis of late post-Fordist stagnation. I have argued both in ‘Pathways’ and in ‘Markets and Discourses’ that we live through a historically unprecedented technological transformation — the convergence of computation and tele-communication, robotic and additive fabrication, software-as-a-service, and most recently the AI revolution, all implying the capacity of absorbing unlimited innovations — whose productive potential is being systematically choked off by an inherited administrative apparatus that has become a fetter. The point of referring here to the accelerating technological dynamics of our times is that these new productivity propelling dynamics make all attempts at political overlording even more counterproductive as at any earlier time. Productivity could soar much faster, and prosperity could grow accordingly. Today, however, the productivity gains made by some are now fritted away and swamped by the simultaneous expansion of counterproductive and costly political control and redistribution efforts, resulting in the extraordinary schism of prosperity stagnation at the very time when technology builds up unheard of prosperity potentials. This massive pent-up potential could be released by the libertarian revolution I have been calling for.

In advanced economies the state controls often up to 50% of GDP and moreover burdens 100% of GDP production with questionable regulations that stifle the free flow of entrepreneurship and market exchanges, lowering productivity and slowing down innovation. … Government and parliaments are certainly massively overburdened by the task of optimally allocating 40% or 50% of GDP, even more so by the task of regulation. In either case incentives through direct existential feedback are missing and ideological rhetoric, if not outright demagogy, guide the parliamentary work more than the kind of direct feedback market participants are continuously exposed to.

The two diagnoses pick out the same patient. Where Spengler sees civilisational winter, the libertarian sees institutional sclerosis. Where Spengler sees the labour-leader caste having “won the World War” and now ruling Western civilisation, the libertarian sees the left-liberal activist class capturing the modern, increasingly intrusive state and turning it against the productive economy it depends on. Where Spengler sees a culture sick in its instincts that “offers no defence,” the libertarian sees an NGO complex and a media-academic establishment that has lost the capacity to respect, or even understand the conditions of its own material reproduction, and is therefore busy undermining these conditions. The crucial difference lies in the prognosis. Spengler reads the dynamic as the unalterable working-out of a cultural destiny. The libertarian reads it as a contingent political-institutional crisis with available solutions.

What a reading of Spengler can add to the libertarian discourse, beyond the awareness of striking historical precedents and parallels, is the reflection on the ethical and characterological dimension of political life. My claim is that a substantial part of Spengler’s positive ethical vision — the part that concerns strength of character, virtue, dedication, and the dignity of overcoming difficulties in accomplishing important tasks — can be detached from his Caesarist-fatalist prescription and integrated into the libertarian project. Indeed, I will argue something stronger: that the libertarian project, properly understood, is not merely compatible with the Faustian-Spenglerian character ideal, but is in our historical moment the only social order in which that character can flourish at scale. The contemporary administrative-managerial state — with its proceduralised compliance regimes, its inflated external responsibilities and collapsing internal authorities, its therapeutic-safeguarding paradigms and journalistic moral panics — is the actual destroyer of Faustian character. The libertarian revolution would be its restitution.

The Worker-Cult and Its Modern Heirs

“People” no longer meant the community of the whole nation, but that section of the city masses which set up in opposition to this community. The proletarian appeared as the hero on the stage of the progressive bourgeoisie.

— Spengler, The Hour of Decision

One of Spengler’s sharpest pieces of cultural analysis concerns the construction of “the worker” as a moral-political fiction in nineteenth-century Liberal-Romantic thought. It is worth quoting at length, because the structural form of the diagnosis applies directly to our own moment:

“The concept of the proletariat, created with deliberate intention by Socialist theoreticians, has been accepted by the middle classes. Actually, it has nothing to do with the thousand branches of strict and skilled labour — from fishing to book-printing, from tree-felling to engine-driving — and is scorned and felt as a disgrace by industrious, trained workers. It was intended solely to secure the amalgamation of these workers with the city mob for the purpose of overthrowing the social order. … A fatal notion of things, half false and half stupid, began to pervade educated and semi-educated minds: “the worker” becomes the real person, the real nation, the meaning and aim of history, politics, public care. The fact that all men work, and moreover that others — the inventor, the engineer, the organizer — do more, and more important, work is forgotten. No one any longer dares to bring forward the class or quality of his achievement as a gauge of its value. Only work measured in hours now counts as labour. … The whole middle class swings the censer before this phantom. No matter what one’s own achievements in life may be, one must fall on one’s knees before him. His being stands above all criticism.” (Spengler, 1934, p.117)

Spengler’s point is not only the sociological one that the leaders of left-wing movements are typically bourgeois — though, as he caustically observes, they typically are academics like Marx and Lasalle, or lawyers like Danton and Robespierre and “the rest were literary men and journalists. There was not a single working man among them” (Spengler, 1934, p.111). His point is conceptual. “The worker” is a fabricated political abstraction that abstracts away from the actual differentiation of productive achievement. The “inventor, the engineer, the organizer,” whose contribution is “more, and more important, work,” are erased. What remains is a sacralised, lump-abstract category before which respect must be performed, and around which an entire political economy of entitlement, victim-status, and moral immunity from criticism is constructed.

This Spenglerian dissection has, I want to argue, an exact contemporary structural analogue, and the analogue is the ideology of multi-culturalism, where the “worker” of the early 20^th Century as object of care and concern has been replaced by the noble, innocent and vulnerable immigrant. In my ‘Multi-culturalism and Ethnic Pride as Barriers to Prosperity’ I have argued: The general, mainstream, publicly promoted attitude towards these distinct immigrant cultures is deliberately and decidedly non-judgemental. Judgement is taboo. The dominant approach is indeed … to uncritically celebrate this cultural diversity. … The persistent minority underperformance, however, is never attributed to the persistent, maladapted cultural patterns but is rather attributed to a lack of opportunities due to (supposedly groundless) persistent prejudices within the host nation, or more recently, due to myriad imperceptible forms of structural or institutional racism, “detected” and “verified” by the measured inequity of outcomes. This short-circuit conclusion is obviously scientifically indefensible but trumps the discourse unchallenged because it has become the demand of a new pervasive “morality.” Morality, as Luhmann teaches us, is all about the conditioning of respect. The simple withdrawal of respectability beats or rather silences here all rational arguments. (Schumacher, 2023)

Like Spengler’s “worker,” the multi-culturalist’s “marginalised community” is a fabricated political abstraction. It pools together radically different groups under a homogenising moral category; it abstracts away from the comparative analysis of which behavioural patterns, work ethics, attitudes to education, family structures, and trust networks actually produce flourishing and which do not; and it treats its protected category as immune from the very criticism by which any culture, including the dominant productive cultures of the modern West, has historically improved itself. In both cases the mechanism is the one Spengler intuited and Luhmann theorised: morality functions as the conditioning of respect and disrespect, and the simple withdrawal of respectability is sufficient to silence rational argument before it gets started. The contemporary academic, journalist, or corporate HR officer does not dare apply the same comparative judgement to non-Western cultural patterns that he routinely applies to Western ones, exactly as Spengler’s nineteenth-century bourgeoisie did not dare to “pour the same scorn on the working man” as it freely poured on “judges, officers, and heads of businesses.”

The Demagogues’ Search for a Mobilizable Mass

Spengler’s analysis goes further, and the further step is essential for understanding the contemporary scene. He traces the historical genealogy of the worker-cult itself, showing that the demagogic class did not begin by appealing to the industrial proletariat. It began by attempting to mobilise the peasantry and craftsmen, found these populations resistant to mobilisation, and only then — with explicit deliberation — turned to the uprooted urban mass that proved more susceptible to its methods. The passage is one of the most important in The Hour of Decision and deserves to stand in full:

“This active Liberalism progresses from Jacobinism to Bolshevism logically. These are not in opposition of thought and will, but are the Early and the Late form, the beginning and the end, of one single movement. It began about 1770 with sentimental social-political tendencies: the structure of society according to class and rank was to be destroyed; and there was to be a “Return to Nature,” to the uniformity of the herd. The place of class was to be taken by that which has no class: money and intelligence, counting-house and lecturer’s chair, arithmeticians and clerks; in place of form-ordered existence, life without form, manners, obligations, respect. It was only about 1840 that this “social-political” tendency passed into an “economic-political” one. The scapegoats are now no longer the aristocrats, but the possessors, from peasant to entrepreneur. The disciples of the movement are promised, not equal rights, but the privilege of the unpropertied; not freedom for all, but the dictatorship of the city proletariat, the “workers.” But this represents no change of a world outlook — which was, and still is, materialistic and utilitarian — but solely a change in revolutionary methods. The professional demagogues now mobilize a different section of the nation for class war.” (Spengler, 1934, p.109)

And the further detail — why the peasantry and craftsmen proved unsuitable as a revolutionary base — is given with characteristic precision:

“At first, about 1770, peasants and craftsmen were approached with some hesitation, both in England and in France. The cahiers of small-town and country deputies in 1789, which were supposed to represent the “Cry of the Nation,” were composed by professional ranters and were not understood at all by the greater part of the electorate. These classes were too deeply rooted in tradition to be unconditionally available as means and weapons. Without the mob from the eastern suburbs — the fists of the capital, always handy — the Reign of Terror in Paris would have been impossible. … Hence the failure of the Convention: peasantry and craftsmen were no reliable following for professional demagogues. They possessed a native sense of respect and self-respect. They had too much instinct and too little town-intelligence. They were industrious and had learnt something; besides, they wished to leave the farm or the workshop to their sons. No permanent effect could be made upon them by programs and catchwords. (The agitation towards this stratum was promptly abandoned when it was found not to have had the desired result.) Only about 1840 did the pamphlet-and-platform demagogy of Western Europe, whose development had proceeded on uniform lines, discover a better medium for their ends in the uprooted mass of men who had gathered round the North-European coal area — the type of the industrial worker.” (Spengler, 1934, p.110)

Two observations of fundamental analytical importance follow from these passages. First, the demagogic class is logically prior to the constituency it claims to represent. The professional agitators who composed the cahiers of 1789, the “pamphlet-and-platform demagogy” of the 1840s, are not the spokesmen of a pre-existing political subject called “the people” or “the worker.” They are an autonomous formation — the literary-journalistic-activist class — in search of a mass through which to operate. The mass is selected on technical grounds: its rootlessness, its detachment from inherited responsibilities and traditions, its lack of “instinct,” its susceptibility to “programs and catchwords.” Peasants and craftsmen are unsuitable because they have “too much instinct and too little town-intelligence” — because they are formed by the productive work they do and the inherited culture they inhabit, and are therefore inhospitable hosts for abstract revolutionary catechisms. The industrial worker of the 1840s is suitable because, in Spengler’s decisive phrase, he is “uprooted” — detached from the cultural and economic substrate that would have given him a sense of “respect and self-respect” independent of what the agitators offer.

Second, and equally decisively, the choice of constituency is opportunistic, not principled. The demagogic class does not begin with a commitment to the industrial proletariat and discover its political vocation as the proletariat’s representative. It begins with a commitment to the dissolution of the inherited social structure and searches for whatever mass population can be successfully mobilised against that structure. When peasants and craftsmen prove unsuitable, “the agitation towards this stratum was promptly abandoned.” The discovery of the industrial proletariat as the mobilizable mass is a tactical-historical contingency, not a revelation of the underlying meaning of history. This is why Spengler can call Jacobinism and Bolshevism “the Early and the Late form, the beginning and the end, of one single movement” — the underlying movement is the demagogic dissolution and usurpation project, and its various political vehicles (the urban Parisian mob of 1793, the industrial proletariat of 1840-1917, and, as we shall see, the contemporary client-categories of the welfare-administrative state) are successive instrumental selections. While this interpretation of the deepest motivations of revolutionary fervour is probably too dark and cynical, I know from my own experience as participant in Marxist revolutionary activism that that fantasy of becoming politically powerful, and indeed of eventually taking over state power, was indeed a thrilling and energizing motivation that animated me and my comrades. In any event, we are talking here not primarily about psychological motivations but self-reinforcing social feedback logics in situations that are fluid, invite experimentation of multiple political projects in front of multiple audiences.

The contemporary parallel is exact and analytically crucial. The class that today operates as the demagogic administrative-academic-journalistic apparatus of the West is the lineal successor of Spengler’s “professional ranters” and “pamphlet-and-platform demagogy.” Its commitments are continuous with theirs: the dissolution of inherited social structure, the elevation of abstract universal categories over substantive differentiated communities, the moral disqualification of “the possessors,” and the redistribution of social respect and material reward through disincentivising political rather than incentivising economic channels. What has changed is the mass through which the class operates. The industrial proletariat has shrunk in the post-Fordist economies and has, in any case, partly disappointed its sponsors by becoming culturally conservative and economically integrated. The demagogic class has therefore made the same tactical pivot that Spengler diagnoses in the 1840 transition: it has identified a new “uprooted mass” suitable to its purposes, and has reorganised its mobilising vocabulary accordingly.

The new uprooted mass is the welfare-dependent client population, supplemented by recently arrived populations of migration whose cultural distance from the host society leaves them similarly detached from its inherited norms of “respect and self-respect.” This population shares the technical characteristics that Spengler identified as the demagogues’ selection criteria. It is rootless — detached from the productive cultures that would have given it independent dignity. It is economically inactive or only marginally active, leaving it without the substantive sense of contribution that peasants and craftsmen possessed. It is unanchored in inherited tradition, especially in the second and third generations. It has, in many of its segments, exactly that combination of “too little instinct” and excess of consumed media that the contemporary equivalent of “town-intelligence” has produced. And, decisively, it is dependent on the political-administrative apparatus for its livelihood, which gives the apparatus the structural leverage to organise it politically. As I argued in Multi-culturalism and Ethnic Pride as Barriers to Prosperity, the contemporary minority population that does succeed economically and acquires the substantive dignity of productive integration — East Asian, South-East Asian, Eastern European, and many of the older European migrations — promptly becomes politically unreliable from the perspective of the demagogic class, and is increasingly excluded from the protected moral categories accordingly. The principle of selection Spengler identified continues to operate.

The Roman parallel here is noteworthy, and Spengler invokes it explicitly elsewhere. The transition from peasant-citizen-soldier of the early and middle Republic to the urban plebs of the late Republic is the original instance of the phenomenon Spengler analyses. The Gracchan reformers attempted to mobilise the Italian peasantry on a programme of land redistribution — and ran into precisely the resistance the revolutionaries of the French Revolution encountered when they tried to build a mass following. The peasantry was too deeply rooted in its work, too encumbered by its productive responsibilities, to constitute a reliable political base; “the peasants, who had moved into Rome in masses to vote, had to go home again for the harvest.” After Gaius Gracchus, the demagogic tradition therefore pivoted, just as in nineteenth-century Europe, to a more politically usable population: the urban populace of Rome, fed by the corn dole, increasingly detached from productive employment, swelled by displaced rural migrants and freedmen, and available year-round for political mobilisation. The Catilinarian, Clodian, and post-Caesarian populations of the Roman streets were the late-Republican equivalent of Spengler’s “uprooted mass.” The transition from one to the other, in both cases, is the transition from a society in which political life is grounded in productive substance to one in which it is grounded in administered dependency — the transition, in Spengler’s terms, from Culture to Civilisation, and, in our libertarian terms, from a productively self-supporting society to a politically-administered redistribution machine.

This historical pattern of the demagogic recruitment of the urban masses was indeed a phenomenon already observed in ancient Rome and astutely described by the Greek historian Polybius. Polybius, quoted by Spengler, describes how ambitious Romans, exploiting and bypassing the constitution, used their wealth to capture power via the mobilisation of the up-rooted urban masses.

“So, when they begin to lust for power and cannot attain it through themselves or their own good qualities, they ruin their estates, tempting and corrupting the people in every possible way. And hence when by their foolish thirst for reputation they have created among the masses an appetite for gifts and the habit of receiving them, democracy in its turn is abolished and changes into a rule of force and violence. For the people, having grown accustomed to feed at the expense of others, and to depend for their livelihood on the property of others, as soon as they find a leader who is enterprising but is excluded from the honours of office by his penury, institute the rule of violence; and now uniting their forces massacre, banish, and plunder, until they degenerate again into perfect savages … when on the one hand they think they have a grievance against certain people who have shown themselves grasping, and when, on the other hand, they are puffed up by the flattery of others who aspire to office. For now, stirred to fury and swayed by passion in all their counsels, they will no longer consent to obey or even to be the equals of the ruling caste, but will demand the lion’s share for themselves. When this happens the state will change its name to the finest sounding of all, freedom and democracy, but will change its nature to the worst thing of all, mob-rule.” (Polybius: The Histories, VI, 9 & VI, 57, Loeb Classical Library, 1922, quoted in Spengler 1934, p.89)

History is replete with examples of this political strategy of using the rabble - what Marx termed the Lumpenproletariat - as vehicle of a power grab.

A Recurring Strategy: History’s Deployments of the Uprooted Mass

The strategic pattern Spengler identifies is not the invention of one era or one political tradition. It is a recurring structural feature of political history wherever a demagogic intellectual class exists alongside a sufficiently uprooted, demoralised, and economically marginal mass susceptible to its activations. Marx himself identified and named the phenomenon from the inside: what he called the Lumpenproletariat — the “ragged proletariat” of criminal elements, vagabonds, discharged soldiers, ruined and adventurous offshoots of the bourgeoisie, beggars, and the chronically unemployed — was for Marx a politically dangerous formation precisely because it lacked the productive embeddedness that he attributed (somewhat optimistically) to the industrial working class. The Lumpenproletariat was available for sale to any political entrepreneur who could pay or promise enough; it had no stake in any particular social order and was therefore perfectly mobile as a weapon. Marx was right about the structural diagnosis and wrong only in assuming that the industrial proletariat was immune to the same availability. History has since shown that populations of any class can be reduced to the Lumpenproletariat condition by the systematic removal of productive purpose, cultural rootedness, and existential responsibility — exactly the removal the political wage and the welfare-administrative state perform.

Three axes of analytical variation organise the historical instances of this strategy. The first is the character of the mass selected: veterans, unemployed urban youth, rural migrants recently urbanised, welfare-dependent populations, ethnically marginalised communities, or the deliberately impoverished. The second is the instrument of activation: the press, the party cell, the mosque or religious institution, the radio broadcast, the NGO and social media ecology. The third is the goal of the demagogic class: seizure of state power, destruction of an internal party rival, ethnic cleansing, cultural revolution against inherited institutions, or the more diffuse but equally consequential goal of institutional capture in universities, corporations, and media. What is near-universal across all instances is what may be called the discard pattern: the mass is instrumentalised, deployed, and then discarded or suppressed once the goal has been achieved. The uprooted mass is never the political subject of the enterprise; it is always the weapon. The confirmation of this is observable in the consistent fate of the deployed mass once its usefulness is exhausted.

Fascist Italy provides the first and in some ways the most instructive case, because Mussolini is perhaps the purest instance in modern history of Spengler’s demagogic intellectual class pivoting from one failed mass to another. Mussolini began as a revolutionary socialist journalist — the archetype of the “pamphlet-and-platform demagogue” — and spent his first political career attempting to mobilise the industrial proletariat. When the Socialist Party’s neutralism over the First World War made that constituency unavailable to him – he had publicly argued Italy should join the war which led to his expulsion - he performed exactly the tactical pivot Spengler diagnoses in the 1840 transition: he identified a new uprooted mass, the discharged veterans of the Italian military, and reorganised his political vehicle and vocabulary accordingly. The fasci di combattimento of 1919-1922 were organised squads of war veterans and unemployed young men, deliberately maintained in a state of violent activation, deployed to break up socialist trade union offices, cooperative movements, and municipal administrations across the Po Valley. The squadristi were Spengler’s “fists of the capital, always handy” — available precisely because they lacked the productive embeddedness that would have given them an alternative stake in social stability. After the March on Rome and Mussolini’s consolidation of power, the squadristi were progressively demobilised, absorbed into the state apparatus, or marginalised: the classic discard pattern.

National Socialist Germany follows the same logic with added ferocity. Hitler is the completed form of the demagogic type: the failed artist, the rootless urban drifter of Vienna and Munich, whose entire political method was the incitement of resentment in a mass of similarly uprooted and demoralised men, instrumentalised through the party cell, the mass rally, and the Stormtrooper (SA) organisation. The SA, which at its peak in 1933 numbered nearly three million men, was composed overwhelmingly of unemployed working-class and lower-middle-class young men — precisely the population most damaged by the Great Depression, most thoroughly stripped of the productive rootedness and self-respect that would have made them resistant to demagogic mobilisation. The SA was deployed to destroy the organised left, break up opposition political activity, and terrorise the Jewish and liberal professional classes. Once this deployment had served its purpose, Hitler had the SA leadership murdered in the Night of the Long Knives (June 1934) and transferred political-military authority to the SS — a more disciplined, ideologically-reliable instrument replacing the now-inconveniently autonomous rabble. Discard pattern: exemplary. It is worth noting that Spengler — writing at this historical moment — despised the National Socialists as the demagogic-journalistic type rather than the statesmanly-Caesarist type he had hoped for. He refused their overtures, declined the offered presidency of the German Academy, and spent his remaining years in internal exile. The man who had diagnosed the phenomenon most precisely was also among the first to recognise its latest instance for what it was.

Mao Zedong’s Cultural Revolution (1966–76) represents a special case, because Mao deployed the uprooted mass not against an external class enemy but against his own party apparatus — using it as a weapon in a factional struggle for internal party supremacy. By 1966 Mao had been substantially marginalised within the Communist Party leadership by Liu Shaoqi and Deng Xiaoping following the catastrophic Great Leap Forward. Unable to dislodge his rivals through internal party mechanisms, he reached outside the party structure and activated a mass that no one else had yet mobilised: the student population of Chinese universities and secondary schools. The Red Guards were students and young workers, deliberately unleashed from all conventional social and institutional constraints, empowered to “bombard the headquarters,” to attack teachers, professors, local party officials, “capitalist roaders,” and anyone carrying the stigma of education, professional expertise, or traditional cultural knowledge. In Spenglerian terms: the demagogue identified a mass that was exactly the inverse of Spengler’s peasants and craftsmen. Where the peasants were rooted in tradition, skilled in their work, and unwilling to be unconditionally available as weapons, the Red Guards were defined by the specific combination of ideological saturation, absence of productive skill or responsibility, and total institutional license that made them maximally available. When the Red Guards had served their purpose — when the rival party faction had been destroyed and Mao had reasserted supremacy — he discarded them with characteristic ruthlessness. Between 1968 and 1980 approximately 16 million urban youth were “sent down” to the countryside for “re-education through labour.” The mass was disbanded and dispersed, its members’ youth and ideological energy consumed in the factional struggle, its leadership dead, imprisoned, or scattered. Discard pattern: total.

Pol Pot and the Khmer Rouge (Cambodia, 1975–79) represent the logical extreme of the strategy — the case in which the demagogic intellectual class selects a mass defined by its complete absence of cultural formation precisely because that absence makes it maximally useful as an instrument of destruction. The Khmer Rouge leadership was almost entirely composed of Paris-educated intellectuals: Pol Pot (Saloth Sar) had studied radio electronics in Paris; Ieng Sary and Khieu Samphan had studied economics and political science respectively at the Sorbonne. This was a demagogic class in the most precise Spenglerian sense — urban, highly educated, abstractly ideological, entirely detached from the productive substance of the country they proposed to transform. The mass they mobilised was the rural peasantry of the poorest provinces, deliberately selected for its absence of education, its impermeability to the competing influences of urban culture and the existing party apparatus, and its total availability as an instrument. The explicit programme — “Year Zero,” the evacuation of all cities, the execution of everyone with secondary education or above — was the logical conclusion of the surplus-value doctrine Spengler diagnosed: the elimination of the entire class of administrators, inventors, engineers, and organisers as parasites on the productive mass. The result was the destruction of between 1.5 and 2 million people, constituting between 25% and 33% of the entire Cambodian population. When the Khmer Rouge were eventually displaced by the Vietnamese invasion of 1979, the demagogic leadership retreated to the Thai border and continued to deploy its dwindling mass as a guerrilla movement for another two decades — the mass too damaged and disoriented to escape its instrumentalisation.

In the Middle East, Hamas and Hezbollah represent a structurally related but institutionally distinct variant. Both organisations were established and are sustained by an educated political-military-clerical leadership — Hamas with its Gaza-based political bureau and its Damascus and Qatar connections, Hezbollah with its Iranian-trained cadres with a link to the Supreme Leader in Tehran — that mobilises populations maintained in deliberate material and institutional deprivation. The Gaza population under Hamas is a textbook case of Spengler’s “pensioner of society” converted into a permanently mobilisable mass: sustained by UNRWA transfers that create structural dependency without enabling economic development, prevented by Hamas governance from building the autonomous economic institutions that would give the population an independent stake in stability and peace, kept in conditions of grievance and resentment that the leadership can activate at strategically chosen moments. The 2023 Hamas attack on Israel and its aftermath demonstrated the operational logic of the strategy with terrible clarity: the mass population bore the consequences of the leadership’s decision while the leadership remained in Doha hotel suites. Hezbollah has performed the same function for the Shia population of South Lebanon and the southern suburbs of Beirut: a population made dependent on Hezbollah’s parallel state structure, deprived thereby of the independent economic and institutional development that would reduce its mobilizability, and deployed as needed against Israel, against the Lebanese state, and against the Sunni and Christian communities of Lebanon. (To avoid misunderstanding: the reference to Hamas and Hezbollah here is serves as illustration of the general socio-political dynamic that is – following Spengler - my topic here, and does not imply any positioning with respect to the underlying conflict and its origins.)

The Rwandan genocide of 1994 adds a further variant: the use of the strategy for ethnic rather than class-based political goals, with the radio broadcast replacing the party cell or the mosque as the instrument of mass activation. The Hutu Power leadership — Habyarimana’s inner circle and the network of local officials and militia leaders — deployed radio (RTLM) to activate an impoverished, land-hungry rural mass against the Tutsi population and against moderate Hutus. The RTLM broadcasts were not spontaneous popular expression; they were a managed instrument of incitement, operated by educated politicians and journalists who knew exactly what they were doing. The “Interahamwe” — literally “those who attack together” — were recruited from unemployed urban youth, exactly the most uprooted segment of the Rwandan population. The genocide’s hundred days resulted in between 500,000 and 800,000 deaths. When the Rwandan Patriotic Front ended the genocide by military force, the Hutu Power leadership fled to Zaire, where they continued for years to attempt the same mobilisation with the remnants of their following. The discard pattern appeared again: when flight became necessary, the leadership departed; the mass was left behind.

Hugo Chávez’s Bolivarian Revolution in Venezuela (1999–2013) and its Maduro continuation represent the contemporary democratic variant of the strategy, operating through electoral rather than violent mechanisms but with the same underlying structure. Chávez was a military officer turned media personality — the demagogic type, equipped with the oil windfall of the 2000s — who used the “misiones” (social programmes providing health, education, and subsidised food) to construct a welfare-client constituency among the urban poor of the barrios, and then deployed that constituency electorally and through the “colectivos” (armed neighbourhood organisations) against the business, professional, and middle classes. The analytical point is the Spenglerian one: the misiones were not primarily welfare instruments but constituency-manufacturing instruments, creating a client population structurally dependent on continued Bolivarian governance. As oil revenues fell after 2014 and the economic basis of the client-construction collapsed, the productive base that sustained it had already been progressively dismantled — nationalisations, expropriations, capital controls, destruction of the agricultural sector. What remained was a dependent population, a narco-military political apparatus, and an economy incapable of feeding its own people. The Spenglerian outcome: the demagogic class consumed the productive substance that had paid for its clientele, leaving a ruin.

The summer of 2020 in the United States — 140 American cities in simultaneous unrest, some sustained for months — provides the most contemporary instance and the one whose infrastructure is most continuous with the welfare-administrative and NGO ecology whose critique occupies earlier sections of this essay. The intellectual-activist class — university-educated, NGO-financed, media-amplified, legally supported through bail funds operated by organisations linked to the Democratic Party’s fundraising apparatus — mobilised and partially directed urban unrest whose participants ranged from genuine grievants to ideologically-motivated activists to opportunistic criminal elements. The mass deployed was heterogeneous, but the management of its political extraction — the narrative framing, the translation of street action into institutional demands on universities, corporations, and city governments, the defund-the-police programme — was entirely the work of an educated activist-administrative class. The subsequent political retreat from these demands — as the electoral consequences became visible — left the deployed communities bearing the costs of the policing vacuum the activist class had created and then abandoned when inconvenient: the classic discard structure.

Three analytical conclusions follow from this survey. The first is that the strategy is trans-ideological. It has been deployed by the socialist left (Lenin, Mao, Chávez), the fascist right (Mussolini, Hitler), theocratic Islamism (Khomeini, Hamas, Hezbollah), ethnic nationalist movements (Hutu Power, various), and the contemporary liberal-progressive activist class. The specific ideology of the demagogic class varies; the structural logic of the deployment is constant. This trans-ideological constancy is what Spengler’s analysis captures and what purely ideological accounts of these phenomena miss. The phenomenon is not fascism, or communism, or Islamism. It is the recurring deployment of the uprooted mass by an intellectual class that requires a mass through which to operate.

The second conclusion is what may be called the Lumpenproletariat production thesis. The uprooted mass is not a naturally occurring historical formation that political entrepreneurs merely discover and activate. In many of the most important cases — the welfare-administrative state’s client populations, the UNRWA-sustained Gaza population, the Maoist-educated Red Guards, the Chávez-era barrio clientele — the uprooted mass is actively produced by the political arrangements that then deploy it. The welfare-administrative state does not merely find a dependent population and provide for it; it creates and continuously reproduces the dependent population it then mobilises. This is the deepest sense of Spengler’s insight that the demagogic class is logically prior to the constituency it claims to represent. It does not merely claim to represent a pre-existing mass; it first of all gathers and indeed constructs this mass.

The third conclusion is the libertarian one, and it follows directly from the previous two. If the uprooted mass is produced by identifiable political-institutional arrangements — primarily the welfare-administrative state, the NGO-activist complex, the proceduralised displacement of substantive competence by credentialed compliance, and the various instruments of administered dependency — then its production can be stopped by the dismantling of those arrangements. The libertarian project, in its welfare-reform dimension, its anti-regulatory dimension, and its charter-city and secessionist dimension, is precisely the systematic withdrawal of the institutional mechanisms through which the uprooted mass is produced and maintained. It is also, therefore, the systematic dissolution of the political infrastructure of the demagogic strategy. A society of genuinely productive, economically embedded, culturally rooted, institutionally self-responsible individuals — each bearing the reality-contact that Spengler called instinct and the amor fati that comes from genuine existential responsibility — is a society that is structurally resistant to the demagogic mobilisation that has torn apart every major political culture of the past two centuries.

This historical-genealogical insight reinforces and deepens the critique of welfare developed in Pathways. Here too, I note the connection of welfare and violent crime. The welfare-administrative apparatus is not, as its defenders claim, a humanitarian instrument for the relief of misfortune. It is the structural mechanism through which the political (demagogic) class has produced and continues to reproduce the mobilizable mass on which its political existence depends. The withering of welfare is therefore not only an economic project (necessary for prosperity) and not only a moral project (necessary for the restoration of self-respecting dedication to tasks). It is also a structural-political project: the dismantling of the institutional apparatus through which the political class manufactures its own constituency. Once the political pensioner-structure is withdrawn, the populations currently mobilised by it will be re-integrated, by the ordinary operation of productive incentives and discursive engagement, into the productive economy and its associated cultural substance. The “uprooted mass” will be progressively re-rooted. The demagogic class will then find itself in the position in which it found itself between 1789 and 1840 — in search of a constituency, and unable readily to find one. This is the deepest sense in which the libertarian roll-back of the welfare-administrative state is also the dissolution of the contemporary demagogic apparatus.

It is worth dwelling on Spengler’s identification of journalism as the characteristic modern form of cultural authority. The passage is worth quoting in full:

“At this point journalism becomes the dominant expression of the time. It is the critical esprit of the eighteenth century diluted and lightened for intellectual mediocrity — and let us not forget that the Greek word krinein means to part, to dissect, to disintegrate. … “Party” becomes the Liberal substitute for rank and State; revolution, in the form of periodic mass elections fought by all available means of money, brains, and even — after the Gracchan method — physical violence, is exalted into a constitutional process; government, as the meaning and duty of State existence, is either opposed and derided or degraded to the level of a party business. But the blindness and cowardice of Liberalism goes further still. Tolerance is extended to the destructive forces of the city dregs, not demanded by them.” (Spengler, 1934, p.117)

Today Antifa militants are ready to use violence against the enemy on the right, knowing this will be tolerated, quietly enjoyed and if necessary justified by the established mass media and the left-liberal intellectuals who have long since captured these media. Together they are manufacturing consent and “public opinion”. Today the “journalistic” expands to encompass NGO communications, university DEI offices, corporate compliance dashboards, social-media activism, and the entire ecology of professional shame-and-praise distribution. Its function is the same: dissection without creation, criticism that operates as a permanent solvent of every institution, every authority, every distinction of competence and achievement, while professing to be the moral conscience of progress.

An important and not always sufficiently understood aspect here is the connection of woke agitation with the political-administrative subsidy, re-distribution and alimentation machine. This connection is not incidental but crucial as it conveys material power - via the material enfranchisement of the respective target constituencies - to the agitators which in turn can be exploited by the agitators for their own material benefit via well paid jobs as legislators, party functionaries or, - this is a new facet today - as functionaries in the state-grant funded NGO complex. (State funds are channelled to these NGOs via the ideologically aligned political parties who in turn receive ideological support and fire power.)

Societal Structure as Differentiation, Not Only Stratification

At this point a serious critical engagement with Spengler is required, because his polemic against levelling, brilliant in its descriptive force, rests on a sociological premise that needs explicit correction. Spengler rightly sees that the urban-democratic Liberalism of the late 19^th and early 20^th century is engaged in the dissolution of a complex, evolved, highly structured society:

“Only the pulling-down of all the products of civilization, of the State, of society, to the level of little people, above which its pitiful emotionalism cannot soar to understand. That is all that the prefix “folk” or “people” means today, for “the people” in the mouth of any Rationalist or Romanticist does not mean the well-formed nation, shaped and graded by Destiny in the course of ages, but that portion of the dull formless mass which every one senses as his equal, from the “proletariat” to “humanity”.” (Spengler, 1934, p. 33)

Spengler seems to locate the problem of social dissolution and atomisation in the fact of large urban agglomerations per se:

“The scene of this revolution of life, its foundation, and, at the same time, its expression, is the Megalopolis which is found arising in the Late period of all Cultures. Into this world of stone and petrifaction flock ever-growing crowds of peasant folk uprooted from the land, the “masses” in the terrifying sense, formless human sand from which artificial and therefore fleeting figures can be kneaded: parties, organizations modelled on program and ideal, but in which that inherent power of growth that the course of the generations had saturated with tradition, and that most of all expresses itself in the fruitfulness of all life — the instinct for the permanence of family and race — is extinct. An abundance of children, the chief sign of a healthy race, becomes tiresome and absurd. This is the gravest sign of the egoism of the Megalopolitans, who have themselves become atoms, of the egoism which is the opposite not of present-day collectivism — between these two there is no difference; a pile of atoms is no more alive than a single one — but of the urge to live on in the blood of posterity, in the creative care for it, in the lastingness of the name. Cold intelligence, on the other hand, that solitary bloom, the weed of the pavements, sprouts in incredible masses.” (Spengler, 1934, pp. 92-93)

Spengler contrasts this modern levelling with the finely cultured and structured traditional society. He says that genuine Culture is a naturally grown Culture. Spengler writes:

“Culture is a growth. The more perfectly a nation represents its Culture — and amongst the noblest creations of a Culture are its nations themselves — the richer its organic disposition by status and rank, and the more genuine the respect of distances between ranks, from the strong-rooted peasantry to the urban patriciate. Here the high level of form, tradition, training, and custom, innate superiority in the ruling families, circles, and personalities, signify the life, the destiny of the whole. A society in this sense remains untouched by “rational” classifications and visions, or else it has ceased to exist. Above all, it is constituted according to rank and not economic classes.”(Spengler, 1934, p.93).

Spengler emphasises “the necessity of a structure that assigns to every sort of person and human activity their place in the life of the whole” and “the essential unlikeness, inequality, of the parts that is identical with organic formation.” (Spengler, 1934, p.95). Spengler elaborates:

“Society rests upon the inequality of men. That is a natural fact. There are strong and weak natures, natures born to lead or not to lead, creative and untalented, honourable, lazy, ambitious, and placid natures. Each has its place in the general order of things. The more significant the Culture, the more it resembles the structure of a noble animal or vegetable body and the greater are the differences between its constituent elements — the differences, not the conflicts, for these are only introduced by reasoning. No good retainer dreams of regarding peasants as his equals, and every foreman who knows his job refuses to allow unskilled labourers to address him on terms of equality. This is the natural feeling in human relations. “Equal rights” are contrary to nature, are an indication of the departure from type of ageing societies, are the beginning of their irrevocable decline.” (Spengler, 1934, p.96).

Spengler is right that this levelling is destructive. He is right that the dissolution of substantive distinction in the name of an abstract equality is corrosive of any social order capable of carrying difficult projects. But he then makes a conceptual move that is, in my view, the central weakness of his social theory. He equates social structure with social stratification. He understands the “well-formed nation, shaped and graded by Destiny,” as essentially a layered hierarchy of social rank, and he can therefore conceive of the loss of structure only as the loss of stratification. The aristocratic ethic, the Prussian rank-order, the soldier-statesman caste, the graded estates of pre-revolutionary Europe — these are for Spengler the very form of substantive social structure, and their dissolution by the levellers is therefore the dissolution of structure as such. This conflation runs through his entire polemic. When he writes that “the opposition of graded social structure and town masses, of tradition and Bolshevism, of the higher existence of the few and the lower of mass labour (however labelled), is upon us — there is no third alternative,” he is offering only two possibilities: stratified hierarchy or unstructured mass.

The conflation is wrong, and the correction was supplied by Niklas Luhmann in the work I have drawn on at several points in my own writing. Luhmann’s decisive contribution to social theory is the recognition that modern society is not de-structured but differently structured. The primary form of structuration in pre-modern society was indeed stratification: society was organised into estates, ranks, and castes, and a person’s position in that vertical order largely determined his access to all social domains. Modern society has a quite different primary form of differentiation, which Luhmann calls functional differentiation. Society is organised into functional subsystems — the economy, science, the legal system, the political system, the art world, the education system, religion, the health system, intimate relationships and the family as system — each operating according to its own distinct code, criteria of success, and internal rationality. The economy operates on the code of deal/no deal and the criterion of profitability. Science operates on the code of true/false and the criterion of evidence-supported argument. Law operates on the code of legal/illegal and the criterion of lawfulness. Each subsystem is autonomous in its operational closure, and no one of them stands above the others as a master code.

This is a momentous structural shift, and Spengler missed it entirely. The dissolution of the stratified estate-order in the nineteenth century was not the dissolution of social structure but the completion of a long transition from stratification to functional differentiation as the primary mode of societal organisation. The structure that emerged is in many ways more complex, more demanding, and more productive than the structure it replaced. Modern society can do things that no stratified society could ever do: it can simultaneously sustain world-spanning markets, autonomous science, evolving law, the art world, mass education, advanced medicine, and the technological civilisation that depends on all of them operating at once. It can do these things precisely because no single hierarchical principle now subordinates the operations of one subsystem to those of another. The Holy Inquisition no longer ratifies scientific findings. The aristocrat no longer determines who may trade. The king no longer sets the price of bread or the conclusions of the trail judge.

This does not mean that hierarchy has disappeared. The Spenglerian instinct that a society without internal hierarchy is incoherent is correct — just misplaced. Hierarchy in modern society migrates from the primary level of social structure (which is now functional rather than stratificatory) to the secondary level of internal organisation within each functional subsystem. The firm has a hierarchy of competence and authority. The university has a hierarchy of scholarly standing. The court has a hierarchy of judicial seniority. The hospital has a hierarchy of medical expertise. The professional discipline has a hierarchy of demonstrated contribution. These hierarchies are real, demanding, and indispensable. What is rejected in modernity is not hierarchy as such but the ascription of an over-arching, society-wide rank-order that determines a person’s position across all subsystems by reference to a single criterion of nobility, ancestry, or political subordination.

Once this correction is made, much of Spengler’s polemic regains its force while shedding its reactionary implication. He is right that the contemporary Liberal-democratic-rationalist mentality, in its degenerate form, attacks not only the obsolete stratificatory hierarchy of pre-modern Europe (which deserved to go) but also the legitimate internal hierarchies of modern functional subsystems. Contemporary egalitarian ideology does not stop at insisting that the duke and the merchant have equal standing as citizens; it proceeds to insist that the brilliant scientist and the mediocre activist have equal standing as participants in scientific discourse, that the proven entrepreneur and the random “stakeholder” have equal standing in corporate decision-making, that the experienced judge and the latest legislative novelty have equal weight in law. This is the genuinely destructive levelling, and Spengler’s denunciation of it is just. But the libertarian, with the Luhmannian apparatus, can articulate the objection precisely: the problem is not the absence of an overarching political stratification, which is rightly absent in modern society; the problem is the political-administrative invasion of functionally differentiated subsystems with a levelling moral code that destroys their internal hierarchies of competence.

The libertarian project, on this reading, is the defence of functional differentiation against political re-totalisation. The roll-back of the state is the roll-back of the only social formation in modern society that still attempts to operate as a master code subordinating all other subsystems to its own logic. Markets and discourses, taken together, are the institutional form in which functional differentiation can complete itself without political tutelage. Within each subsystem, hierarchies of competence will emerge, be tested, be revised, and command the kind of substantive respect that Spengler rightly cherished. What will not emerge — and need not emerge — is the single political pyramid Spengler thought necessary to give modern society any internal form at all.

The Cowardice of the Defenders

Spengler observes, with considerable acuteness, that the bourgeoisie did not have its dissolution imposed on it from outside. It actively, eagerly, ingested it:

“It is a chapter of shame that here presents itself to the historian of the future. Built up though it was on the basic human facts of rulership, grade, and property, society met the Nihilist attack upon these with submission, “understanding,” acclamation, and support. This intellectual suicide was the common fashion of last century.” (Spengler, 1934, p.116).

And again, on the active complicity of the propertied classes in the manufacture of the very ideology of “the worker” that would dispossess them:

“It was the middle classes who successfully “put over” this notion of him (the “worker”), and the very business-like “representatives of the people” continue to sponge upon the legend. They dinned it into the wage-earners until they believed it; until they felt themselves to be really ill-treated and wretched, until they lost all sense of proportion with regard to their output and their importance. Liberalism vis-à-vis the demagogic trend is the form of suicide adopted by our sick society.” (Spengler, 1934, p.119)

The contemporary parallel is exact. The most aggressive expansions of contemporary regulatory and ideological orthodoxy — in universities, corporations, philanthropy, financial institutions — have been carried out not against the resistance of the propertied and credentialed classes but largely by them and through them. In my ‘Thesis on Government Destruction of Corporate Management’ I have argued that the contemporary firm operates under a structural pathology in which it is held to expanding external responsibility while being progressively stripped of internal sovereignty: Contemporary firms increasingly operate under a profound structural tension: they are held ever more responsible for outcomes in their external relations while simultaneously losing authority and sovereignty in their internal governance. A schism opened that undermines the rational management our prosperity depends on: Internal sovereignty is eroded while external responsibility is increasing. … Externally, the firm is treated by law and public opinion as a unified, decisive, and fully responsible actor. … Internally, however, the same firm is governed under a markedly different logic. Employment law, data protection regimes, and equality legislation increasingly constrain the ability of management to observe, evaluate, and discipline employees. … As a result, authority becomes fragmented, proceduralised, and legally risky, while decisional latency increases and informal judgement is displaced by defensive compliance. (Schumacher, 2026)

This is, in different vocabulary, precisely what Spengler diagnoses at the level of the late-modern political order: an erosion of accountability, a hollowing of authority, a substitution of paper for command, of procedure for responsibility, of the appearance of governance for governance itself. In the language of the previous section, it is the violation of the functional differentiation of the economy by the political-administrative system, which insists on importing its own moral-procedural code into the operational logic of the firm. The result is what I have called an “internal compliance apparatus mediating between the administrative state and the firm. Its dominant orientation is not performance optimisation but liability containment.” Spengler’s “true statesmen become rarer and rarer” finds its contemporary form in the CEO who is, increasingly, an apologetic compliance-officer of last resort, hauled before the moral tribunal of journalism whenever his firm’s output collides with the latest sensitivity, while denied the practical authority to manage the operations for which he will be blamed.

The most disquieting Spenglerian observation, however, is the one about complicity. The propertied class did not merely fail to defend itself; it actively participated in the construction of the ideology of its own dispossession. So today: corporate management does not merely tolerate the proceduralisation of its authority but funds it, promotes it, hires for it, and rewards those most fluent in its idiom. Shareholders ratify executive compensation linked to ESG and DEI metrics. Boards appoint Chief Sustainability and Chief Diversity Officers with veto powers over operational decisions. Foundations of capitalist origin pour vast resources into the activist-academic infrastructure that produces the ideology under which their grantors are themselves to be eventually disestablished. The “intellectual suicide” Spengler observed in the late 19^th and early 20^th Century bourgeoisie has its precise contemporary expression in the voluntary capitulation of contemporary corporate management to the moral terms of its dispossessors.

The eagerness with which the top echelons corporate America and then corporate Europe embraced and internalised the latest left-liberal ideological onslaught in the last 15 years was out of a panicked fear of an even more severe, material onslaught the spectre of which was the post-2008 Occupy movement. Further, this external intimidation was paralleled by an internal intimidation from within originating with the latest cohorts of graduate recruits delivered into the corporate workforce from the universities - especially elite universities - who had been targeted and captured by the left-liberal intellectuals and failed revolutionaries who had been en masse absorbed as a new crop of professors into the massively expanded universities (which they had first occupied and then pried wide open) since 1968.

Political Correctness and the Terror of Mass Opinion

Spengler’s analysis of his contemporaries’ cowardice before mass opinion is so precise, and so prophetic of the contemporary phenomenon of “cancel culture” and the regime of “political correctness,” that the passage deserves to stand at the head of this section:

“Courage to see what is actually happening would be the initial requirement, and I fear that the stock of this is very low. At no time has the whole world shown such cowardice before the general opinion of parliaments, parties, speakers, and writers. They are all on their knees before the “people,” the masses, the proletariat, or whatever they may call that which blindly and unsuspectingly serves as weapon for the leaders of the World-Revolution. The reproach of being “the enemy of the worker” causes every politician to pale.” (Spengler, 1934, p.137)

Read this passage with one substitution. Wherever Spengler writes “the worker,” write instead “the marginalised,” “the vulnerable,” “the oppressed,” or any of the contemporary moral categories that perform the same function. The diagnosis becomes a vigorous description of our own moment. “At no time has the whole world shown such cowardice.” Every politician, every executive, every university administrator, every editor, every artist, every academic pales at the reproach of being “the enemy of” whichever sacralised category the moral entrepreneurs of the moment have nominated for protection. The structure of the cowardice is identical: a fabricated political abstraction is invested with sacred status; a class of professional outrage-managers monopolises the right to speak in its name; and the entire propertied, credentialed, and political world performs its submission to avoid the verdict of “enemy.”

What is being weaponised is not, in either case, the actual interest of the actual workers, or the actual interest of any actual marginalised person. Spengler saw this clearly. The “worker” of nineteenth-century progressive sentiment had little to do with the actual skilled tradesman, who in fact “scorned and felt as a disgrace” the proletarian designation imposed on him by literary radicals. The “worker” was a political weapon, useful precisely because of its abstraction from any concrete person whose actual preferences might inconveniently disagree. The contemporary “marginalised community” functions identically. Surveys repeatedly demonstrate that the actual members of the supposedly protected categories — working-class minorities, ordinary migrants, the religiously observant, women outside the activist class — hold views markedly different from, and frequently opposed to, the views attributed to them by their self-appointed spokesmen. The abstract category is the weapon. The actual persons it claims to represent are largely irrelevant to its political operation.

Spengler’s phrase “They are all on their knees” deserves emphasis. He is describing a posture, a physical attitude of submission, that has become so habitual in his contemporaries that they no longer notice it. The contemporary parallel could not be more exact. The corporate apology, the public confession of historical complicity, the kneeling on one knee at sporting events, the proliferation of pronoun declarations, the territorial acknowledgments at the openings of conferences, the Diversity Statements required of academic job candidates, the mandatory implicit-bias trainings, the elaborate genuflections of resignation statements from executives caught on the wrong side of the latest moral panic — these are the ritualised contemporary forms of Spengler’s “on their knees before the people.” The performance of submission has been so successfully demanded that those who perform it no longer experience it as humiliation. They experience it as virtue.

The mechanism by which this submission is enforced is not state coercion in any direct sense. It is the threatened withdrawal of respectability — the moral excommunication that Spengler described as causing every politician to pale, and that we describe, with insufficient analytical seriousness, as “cancellation.” I have argued in ‘Multi-culturalism and Ethnic Pride as Barriers to Prosperity’ that this mechanism is best understood through Niklas Luhmann’s analysis of morality as the conditioning of respect and disrespect. The point is worth restating in this context. The threat is not physical. The threat is professional, social, and reputational excommunication, and the threat is effective in roughly the proportion that one has anything to lose. The university dean, the corporate executive, the senior editor, the established scholar, the celebrity, the politician — each has built a position whose continued occupation requires the recognition of respectable opinion, and each is therefore exquisitely vulnerable to the threatened withdrawal of that recognition. The dynamic Spengler describes — universal cowardice before mass opinion organised by a small class of moral entrepreneurs — is structurally identical, only the moral categories have been updated.

The libertarian project is intimately bound up with the diagnosis and remedy of this pathology, and it is worth being precise about how. The libertarian does not propose to abolish moral disagreement, social pressure, or the capacity of communities to set conditions of respectability. These belong to the discursive substance of any free society and cannot be legislated away. What the libertarian does propose to abolish is the structural feature of the contemporary order that makes the cowardice so devastating: the convergence of state authority, mass media, corporate compliance, educational credentialing, professional licensing, and the financial-payment infrastructure into a single network through which moral excommunication, once successfully demanded, becomes practically total. When one’s university, employer, bank, payment processor, social-media presence, professional licence, and access to public discourse can all be withdrawn simultaneously by the same enforcement mechanism responding to the same moral panic, the threat of “cancellation” acquires a totalitarian weight that no merely cultural or moral threat could otherwise carry. The contemporary cancel-regime is what political-administrative-corporate fusion has made of what would otherwise be the ordinary frictions of moral disagreement in a free society.

Activist moral entrepreneurs are today funded by state grants channelled to them via left-liberal charities - the NGO complex. The existence of this complex feeds back into the universities who offer activist grievance studies of many kinds. The popularity of these courses is fuelled and sustained by this prospect of lucrative, cushy government subsidised NGO jobs - campaigning as well-paid profession - operating outside of the market pressures that, rightly, discipline all economically productive jobs. The point here is not to denigrate charities - indeed the role of charities might even increase in a libertarian society - but to identify the perverse incentives that come into play when charities rely on tax revenue rather than collect from the public in the competitive market for voluntary donations via convincing missions and on the basis of a track record of efficient achievements. The universities too should be market disciplined in their offering. The current flourishing of “academic” rabble rousing demagogy is dependent on the subsidized political NGO complex and on publicly funded universities.

This is one of the ways in which the libertarian roll-back of the state is also, and inseparably, a roll-back of the conditions that make contemporary political correctness so devastating to the substantive life. The fragmentation of jurisdictions through secession, charter cities, and special zones; the privatisation of education and the consequent diversification of educational credentials; the proliferation of payment systems and financial infrastructures beyond the reach of any single regulatory authority; the privatisation of all discourse platforms; the dissolution of state-licensed monopolies in law, medicine, and the professions; the withering of the welfare-administrative apparatus that underwrites the activist class — each of these libertarian moves directly weakens the infrastructure through which moral excommunication is currently made total. In a libertarian order, the ordinary frictions of moral disagreement would persist, as they should. What would not persist is the capacity of a small class of moralising demagogues, operating through a unified network of state, corporate, and credentialing institutions, to make the cost of dissent ruinous.

It is worth dwelling on Spengler’s observation that “the reproach of being ‘the enemy of the worker’ causes every politician to pale.” The contemporary version — “the enemy of the marginalised,” “the racist,” “the bigot,” “islamophobe”, “the transphobe,” and the entire vocabulary of moral disqualification — functions identically. The reproach is not an argument. It operates ad hominin and is the threatened withdrawal of respectability, calibrated to the position one occupies and the network of dependencies through which one occupies it. Spengler saw this with a clarity our contemporary commentators rarely match, because most contemporary commentators are themselves operating within the dependency network and have powerful incentives not to see clearly. The libertarian, whose project is precisely to dismantle the structural conditions of this dependency, is among the few positioned to see it and to articulate it. What this cowardice before mass opinion purchases, and what in turn funds it, is the substantive subject of the next section: the political extraction of incomes, transfers, and existential immunities from the productive economy on a scale and through a mechanism that Spengler diagnosed with exceptional precision and polemic force.

The Political Wage and the Pensioner-Society

The cowardice before mass opinion described in the previous section has its concrete economic counterpart in what Spengler called the political wage: the extraction of incomes, benefits, and existential immunities from the productive economy by political rather than economic means. The two phenomena are inseparable. The political wage is what the cowardice purchases, and the cowardice is what the political wage funds. Spengler grasped both the mechanism and its consequences with a clarity that places his pages among the most acute economic-political analyses of his day, independently coming close to the Mises-Hayek calculation argument and the libertarian critique of the welfare state that would only later be systematically developed.

The opening move of Spengler’s analysis is the displacement of economic competence by political demagogy. He poses the question as a rhetorical accusation:

“Is not economic, bureaucratically administered Socialism the reigning ideal in the national labour parties of Germany, England, and even Italy? Have not the men with creative economic ideas, the promoters of private economic enterprise, been sacrificed to this dictatorship on the platform of the world-economic ruin? The economic leader, the expert in economic life, has been ousted by the party leader, who, if he knows nothing of economics, knows all there is to know about demagogic propaganda. He rules as a bureaucrat in the drafting of economic legislation, which has displaced the free decisions of the man who knows, as leading spirit in countless committees, courts of arbitration, conferences, cabinet meetings and whatever the forms of his dictatorship may call themselves, and even in the Fascist Ministry of Corporations. He is out for State Socialism, for the elimination of private initiative, for economic “planning” — all of which mean the same thing: Communism.” (Spengler, 1934, p.139)

Read in 1933 this passage was prescient. Read in 2026 it is descriptive. The economic leader, the expert in economic life, has been thoroughly ousted by the regulatory functionary, the parliamentary committee-member, the central-bank governor with his appointed advisory board, the cabinet office strategist, the Treasury official, the European Commissioner, the ESG-rating administrator. Each of these is a “party leader” in Spengler’s sense — an official whose authority derives from the political-administrative apparatus, not from any demonstrated competence in the production, allocation, or coordination of economic resources. The “free decisions of the man who knows” have been replaced, decision by decision and domain by domain, by the deliberations of “countless committees, courts of arbitration, conferences, cabinet meetings.” The forms of the dictatorship have proliferated, but their function is constant: to substitute political determination for economic judgement throughout the productive system.

I detect here a certain resonance with Friedrich von Hayek’s important distinction between a “nomos” and “thesis”, between a spontaneous order as playing field and a deliberate social organisation that wants to determine social outcomes. This distinction is important for us libertarians in our engagement with the current crop of aggressive justice warriors and their counterparts within government, administration and NGO complex. Hayek posits the stark, initially counterintuitive thesis that “social justice” is an oxymoron, an incoherent, contradictory concept. Hayek talks about “the mirage of social justice”. This phrase even serves as subtitle to the second volume of Hayek’s ‘Law, Legislation and Liberty’. “Social justice” appears to name a morally compelling ideal, but dissolves under analysis when applied to the outcomes of a free, impersonal market order. Justice applies to conduct, not to impersonal outcomes. For Hayek, justice and injustice are properties of intentional human action. A person, an individual action, or a legislated regulation can be just or unjust, and there will be somebody (or some organisation) responsible for this injustice. But nobody is responsible for the “distributive” results of a spontaneous market order. It is the aggregate result of countless individual transactions, plans, accidents, skills, scarcities, preferences, and discoveries. Any attempt to secure “Social justice” in such a context requires continuous coercive correction. If one insists that the distribution produced by the market must conform to some prior pattern—equality, need, merit, proportional reward, dignity, inclusion, status security, or any other distributive ideal—then the state must continuously intervene to correct the outcomes of voluntary exchanges. This is one of Hayek’s central points: patterned distributive justice is incompatible with a dynamic market order. Markets constantly change relative positions. Innovation, competition, consumer choice, luck, shifts in demand, and entrepreneurial discovery alter incomes and wealth. To preserve a politically preferred distribution, government must repeatedly override contracts, prices, property rights, inheritance, wages, rents, profits, and investment returns. This is why he connects social justice with the erosion of the rule of law. General rules valid for all are replaced by discretionary commands aimed at particular distributive results. Hayek’s key philosophical move is: justice is a property of conduct (following or violating a rule), not of outcomes (the aggregate result of a spontaneous order). Since no one intends the distributive result of a market process, no one can be responsible for it, and therefore it cannot be unjust — it is simply the unforeseeable outcome of many voluntary interactions. To correct it coercively is therefore not the correction of an injustice but the commission of one: it violates the voluntary contracts, property rights, and free exchanges of real people in order to satisfy an abstract pattern that no one chose.

Again, Hayek’s core point is that “social justice” requires the substitution of particular commands (aimed at specific distributive outcomes) for general rules (valid for all, without regard to outcome). This substitution is the erosion of the rule of law — “nomos” replaced by “thesis”, or in his later vocabulary, the “Great Society” governed by abstract rules replaced by an organisation governed by commands in pursuit of shared purposes.

Spengler’s entire analysis of the political wage rests on the distinction between economic determination of rewards (the result of the organic whole of productive activity, prices, productivity, market demand) and political determination (the imposition of a wage-level by party-political force irrespective of the economic basis). The political wage is “inorganic” precisely because it overrides the emergent, unplanned result of countless productive interactions — exactly the spontaneous order Hayek is describing — in favour of a politically-preferred distributive outcome. Spengler’s vocabulary of “organic” vs. “inorganic,” “economic leader” vs. “party leader,” “living order” vs. “paper system” maps directly onto Hayek’s nomos vs. thesis.

The most direct Spengler anticipation of Hayek’s formal argument is actually in Spengler’s implicit theory of organisational legitimacy. When Spengler distinguishes the “economic leader, the expert in economic life” from the “party leader, who, if he knows nothing of economics, knows all there is to know about demagogic propaganda,” he is distinguishing between two types of authority: authority grounded in functional competence within the spontaneous order (the economic leader whose authority is confirmed by the success or failure of his decisions in the market) and authority grounded in political appointment and demagogic power (the party leader whose authority is confirmed by electoral success and whose decisions are insulated from the feedback of the productive process). The first type of authority operates, in Hayek’s vocabulary, under nomos — general rules that apply to all including the decision-maker, who bears the consequences of his decisions. The second type operates under thesis — discretionary commands that the decision-maker is not subject to and the consequences of which are borne by others. “Social justice” is the ideology that legitimises the second type of authority’s claim to override the first.

Spengler’s anticipation lies in this: the political wage as the destruction of nomos. When Spengler writes that the wage “is settled by the party leader, not by the economic leader, and if the latter does not and cannot consent to it, it is enforced by strikes, sabotage, and pressure on parliament,” he is describing exactly Hayek’s transition from general rules to particular commands. The economic leader’s authority is grounded in general rules confirmed by market feedback. He decides in response to ever-changing needs and scarcities signalled via market data. The party leader’s authority is grounded in discretionary political power insulated from productive feedback. Where Spengler goes further than Hayek: Hayek shows the logical incompatibility between patterned distributive justice and a dynamic market order. However, he does not dwell on the characterological consequences for those exempted from the market’s reality-test. Spengler does.

The contemporary “social justice” apparatus with its ever expanding clientele and sets of concern — racial equity, gender pay gaps, boardroom representation, decolonisation, DEI enforcement — is the latest instance of the same demand, operating through the same legal mechanism (general rules displaced by discretionary commands), the same cultural-characterological mechanism (protected client categories exempted from the market’s reality-test), and the same political mechanism (the demagogic class that constructs and manages both).

Spengler’s final equation — “State Socialism, the elimination of private initiative, economic ‘planning’, all of which mean the same thing: Communism” — deserves to be taken with full seriousness. The libertarian project, on the basis of the Mises-Hayek argument, identifies the same equivalence. State planning, the regulatory displacement of private decision, and the various intermediate forms of “mixed economy” and “stakeholder capitalism” do not represent intermediate stages between capitalism and communism. They represent gradations of the same underlying error: the substitution of political-bureaucratic command for the economic-rational coordination of dispersed knowledge through prices and competitive bidding. I have argued the point at length in Pathways towards a Stateless World Society: In advanced economies the state controls often up to 50% of GDP and moreover burdens 100% of GDP production with questionable regulations that stifle the free flow of entrepreneurship and market exchanges, lowering productivity and slowing down innovation. The protections state regulations impose can be much more effectively delivered by market competition or voluntary industry self-regulation whereby rules do not require enforcement via compulsory powers in order to be effective. Government and parliaments are certainly massively overburdened by the task of optimally allocating 40% or 50% of GDP, even more so by the task of regulating all economic activities. In either case incentives through direct existential feedback are missing and ideological rhetoric, if not outright demagogy, guide the parliamentary work more than the kind of direct feedback market participants are continuously exposed to. (Schumacher,2024, p.262)

Spengler’s “demagogic propaganda” and the “demagogy” I identify in parliamentary economic decision-making are the same thing. The mechanism by which competent economic judgement is displaced is identical: incentives for direct existential feedback are removed, ideological rhetoric replaces practical experience, and the politically successful displaces the economically rational. Mises and Hayek explained why this substitution is necessarily fatal to economic rationality. Spengler’s contribution is to show, with equal force, that the displacement was already largely accomplished a century ago, and that the world-economic crisis of his own moment was its predictable result.

The Composition of Value and the Theft of Headwork

Spengler’s deeper analytical contribution lies in his account of how the political wage actually operates within the productive economy. He starts from a conception of economic value that is essentially modern — a recognition that the value of output depends not merely on the labour-time consumed in production but on the integrated coordinative intelligence applied to the whole productive process. The passage is worth quoting at length:

“This was the fighting method of the strike in a disguised, slowly penetrating form. It first took on meaning through the fact that the price for this form of goods, the weekly wage … was forced steadily upward. Now, the “value,” the actual product of the work done, is not an independent quantity. It is a result obtained from the organic whole of industrial labour, in which the administrative work of controlling and regulating operations, the obtaining of materials, the marketing of products, the thinking out of costs and yield, of lay-out and equipment, and of new possibilities, are of far greater importance. The total output depends upon the order and amount of head-work, not hand-work, that is put into it. If there is no yield, if the product is unsaleable, then the work put into the process has been valueless and ought really not to be paid for at all. This is what happens to the peasant and the craftsman. But through the activities of the trade unions the hourly wage of the hand-worker has been removed from the unit of the organism.” (Spengler, 1934, p.141)

The point is exact, and it is the point that the entire Marxian labour-theory of value and its progressive descendants are constructed precisely to deny. Spengler identifies that value emerges from the integrated economic operation as a whole — from the head-work of organisation, design, market-anticipation, equipment-investment, materials-procurement, and innovation in equal or greater measure than from the hand-work of execution. The libertarian inherits this insight directly. As I have argued repeatedly in Markets and Discourses, the contemporary post-Fordist economy intensifies this principle to an unprecedented degree: The technologies of Fordism were rigid without agile adaptivity and without the ability to quickly absorb and utilize innovations. In contrast, the new numerically controlled systems like robotic fabrication, 3D printing as well as the massive new field of software-as-a-service (SaaS) are capable of absorbing unlimited numbers of innovations. This in turn dynamizes business and business organisation and pulls workers away from assembly lines into R&D, marketing, finance etc. All work becomes innovative and project-like rather than routine. (Schumacher, 2025)

What Spengler called head-work and hand-work are no longer even meaningfully separable in the post-Fordist economy. The integrated coordinative intelligence that produces value is distributed across the entire productive organisation, embedded in software systems, design specifications, market-analytical models, supply-chain protocols, and the discursively informed judgements of countless actors. The contemporary “worker” — software engineer, designer, financial analyst, marketing strategist, project manager — is himself overwhelmingly engaged in headwork. The Spenglerian distinction between economically rational integration and politically imposed wage-extraction therefore applies more sharply today than in 1933, not less.

Spengler’s decisive observation is what happens when the wage is removed from the economic-organic unit and politically determined:

“It is settled by the party leader, not by the economic leader, and if the latter does not and cannot consent to it, it is enforced by strikes, sabotage, and pressure on parliament. … Everyone actively employed in the economic system is dependent for his gain on the economic situation — everybody except the wage-earning workman. He has a claim to the wage-level that is fixed inorganically and fought for by party-political means, even when it can only be maintained by allowing works to fall into decay, cutting out profits altogether, and selling goods at less than cost price — until the factories themselves give in.” (Spengler, 1934, p.142)

The passage requires almost no translation to apply to contemporary conditions. The minimum wage legislated by parliaments without reference to local productivity; the centrally-negotiated collective wage agreement extended by political fiat to non-signatory firms; the minimum wages, maximum working hours, the mandated paid holidays, and notice periods imposed across heterogeneous industries; the compliance-bureaucratic overhead loaded onto every employer in the name of employee protection; the politically-driven minimum corporation-tax rates designed to prevent jurisdictional competition; the government regulation of working conditions; the unfair-dismissal regimes that convert employees from contracting partners into legally-entrenched stakeholders — each of these is a contemporary instance of the political wage in Spengler’s sense.

None of these government-imposed or trade union imposed protective measures can in fact protect, guarantee or improve the long-term position of employees – they are in fact guaranteeing the very opposite: stagnation and impoverishment. The real protection for all employees (except perhaps of those who are out to shirk and free-ride) is ease of starting/doing business and unhampered competition between employers for employees in an economy with free contracting and free of regulatory barriers to growth. Even those employees who are currently exploiting the protections afforded by employment law to cling to a job they cannot really fill, are in the final analysis victims of this nanny-state system because this system freezes up the labour market to such an extent that makes resigning from a mismatched job to find a better match too risky. The result is that a large part of the total societal workforce is stuck in unhappy, suboptimal employments, stressing both employers and employees, while spoiling much productive potential that could be released in an unhampered, much more liquid labour market.

In Spengler’s terms, the wage and the terms of employment are “fixed inorganically and fought for by party-political means,” without reference to the productive operation from which the wage must in fact be drawn. And the consequence is the one Spengler identified: firms maintain politically-imposed cost-structures by “allowing works to fall into decay, cutting out profits altogether, and selling goods at less than cost price” — or, in the contemporary version, by relocating to less politically-burdened jurisdictions, by automating out the affected categories of employment, by withdrawing from labour-intensive product lines entirely, or by simply ceasing to exist. This is then the final triumph of the parasite over the host, in the moment the host expires. The cumulative consequences of this invisible death by a thousand cuts on prosperity are devastating for all, not least for those who seem to be the object of care and protection here.

Spengler’s observation that the wage-earner is the sole participant in the economic system who is not “dependent for his gain on the economic situation” is the most analytically important sentence in the entire passage. The economic logic of the firm requires that all factor returns — to capital, to entrepreneurship, to land, and to labour — must in some way reflect the underlying productivity and saleability of the operation. The political extraction of one of these factor returns from this dependency does not eliminate the underlying constraint; it merely transfers the consequences of its violation onto the other factors and ultimately, when those have been exhausted, onto the existence of the firm itself. The political wage is paid by the destruction of the productive substance that would have paid the economic wage.

The Pensioner-Society and the Death of Self-reliance

Spengler’s analysis culminates in the most extraordinary passage of all, and the one most directly aimed at what we today call the welfare state. He argues that the political wage in its developed form extends beyond money payment to encompass the entire existence of the wage-earner, transferring the existential burden of life itself from the individual onto a generalised “rest”:

“We must realize how comprehensive is the term “political wage” before we can estimate the pressure of this wage-dictatorship on the economic life of nations. Reaching out far beyond money payment, it embraces concern for the “worker’s” whole existence, the burden of which is taken from him to be loaded onto “the rest.” “The worker” has become a pensioner of society and of the nation. Every human being … has his own personal cares, full responsibility for himself, and must inevitably make his own decisions in all dangers threatening himself and his aims. No one dreams of relieving, at the expense of others, farmers from the consequences of bad harvests, cattle disease, fire, and failing markets; or artisans, doctors, engineers, tradespeople, and scientists from the threat of economic ruin. … Each of these has to deal with such things himself and at his own expense or else bear the consequences and beg or go under in any other way he pleases. Such is life. The craving to insure oneself, against … fate in every conceivable form — which is a sign of sinking vitality, beginning from Germany, has now embedded itself in one way and another in the mentality of all white nations.” (Spengler, 1934, p.144)

And he proceeds to identify the psychological-cultural consequence:

“All the rest in seeking to evade or to meet the unforeseen do so according to their own resolve and in their own strength; it is only the “worker” who is spared this decision. He alone can rest assured that others will think and act for him. The degenerating effect of this freedom from all responsibility, which is seen similarly in children of very rich parents, has overtaken the whole working class, especially in Germany: at the first sign of any distress, appeal for help is made to the state, the party, society, or, in any case, “others.” We have forgotten how to take decisions ourselves.” (Spengler, 1934, p.145)

Spengler’s phrase “pensioner of society and of the nation” is, in 2026, no longer a polemic. It is a sociological description of the entire dependent segment of every advanced Western society. The category has expanded beyond the original “worker” to include vast administrative apparatuses of insurance, redistribution, unemployment benefit, sickness benefit, disability benefit, housing benefit, child benefit, in-work benefit, and the proliferating sub-categories of conditional and unconditional transfer. It encompasses populations that have never worked at all, populations that have ceased to work, populations that work nominally while being substantively sustained by transfers, and populations whose entire occupational existence consists of administering the transfers to others. The “rest” — the segment from whose productive output the entire apparatus is funded — has correspondingly shrunk and, in country after advanced country, is no longer adequate to the demands placed on it. This is not a only moral judgement; it is also an accounting one. The fiscal crisis of the advanced welfare states, which dominates the politics of every European country and increasingly of the United States, is the actuarial confirmation of Spengler’s 1933 diagnosis.

The libertarian analysis converges with Spengler’s on every essential point. I have argued in Pathways that the welfare state operates as the central engine of the social pathologies that contemporary commentators routinely blame on capitalism, on globalisation, on immigration, or on cultural decline. The argument deserves restatement in this context: State welfare provisions incentivise and subsidize unproductive, even counter-productive, milieux and lives, certainly lives lived far below their potential. The unconditionally guaranteed provision of a livelihood invites drug and alcohol abuse and becomes all too easily the basis for using the plentiful time and energy for topping up incomes via petty crime, drug dealing etc. … State welfare is eroding spontaneous solidarity, family ties, social responsibilities and commitments etc. thereby undermining all private social bonds, especially in the affected milieux, leading to increased anonymity, loneliness, exclusion. Further, society is bifurcating into an economically active part and an economically inactive part, with a different sense of reality, diverging cultures, and learning curves (Schumacher, 2024, p.267).

Spengler’s “pensioner of society” and my “economically inactive part of society” are the same population, observed at intervals of nearly a century. The “different sense of reality, diverging cultures, and learning curves” that I identify is the contemporary form of what Spengler called the “degenerating effect of this freedom from all responsibility.” Both diagnoses point to the same underlying mechanism: the political extraction of the existential burden from a designated category creates that category as a distinct social formation, with its own characteristic patterns of inertia, dependency, and resistance to participation in the productive economy. And both diagnoses converge on the same culprit: not the dependent population itself, but the political-administrative apparatus that has constructed and continuously reproduces its dependency by systematically removing the existential pressures that would otherwise impel self-development and economic integration.

Spengler’s most striking conceptual move — the comparison of the politically pensioned worker to “children of very rich parents” — deserves emphasis. The contemporary critic of welfare-state pathology, hampered by a moral vocabulary in which only the materially poor can be culturally diagnosed, has lost access to this comparison. Spengler had not. His point is that the structural feature responsible for the cultural degeneration is not poverty as such but the removal of existential responsibility, and that this removal produces the same degenerative consequences whether the agent of removal is parental wealth or political guarantee. The trust-fund child and the third-generation welfare recipient share the same defining structural feature: someone else has assumed the burden of their existence. Both populations exhibit the predicted consequences: an attenuated relation to reality, a habituation to having decisions made on their behalf, an inability to bear the stress of adversity, an atrophy of the capacity for self-direction and self-responsibility. Spengler’s analytical insight cuts across the moral vocabulary of left and right and identifies the underlying mechanism with precision.

The libertarian remedy is precisely the restoration of what Spengler called amor fati — the acceptance of personal responsibility for one’s own existence and fate — through the institutional withdrawal of the political pensioner-structure that systematically extinguishes it. This withdrawal must be phased, as I have argued, to allow the affected populations time to transition into self-reliance: The escape from the trap and poison of state welfare is crucial in its own right for the health of society, with or without state, and is of course a necessary step in moving towards a libertarian and finally stateless society. However, this escape can only happen via a long period of phasing out to give the victims of welfare dependency enough time to transition into self-sufficiency. … A first step, over and above the usual attempts to tighten means testing, incapacity testing, penalties for refusing job offers etc., would be to announce a full stop, perhaps with a not-too-distant deadline, to any new entry into the welfare system, while giving a more generous deadline for achieving self-reliance to all who are currently welfare dependent. State pension schemes should also be phased out on a timetable that recognizes contributions paid. (Schumacher, 2024, p.267).

What replaces the political pensioner-structure is not a vacuum but the dense network of voluntary, contractual, and charitable arrangements through which pre-welfare societies in fact provided for misfortune — mutual aid societies, friendly societies, trade-affiliated insurance pools, religious charities, private pensions, family solidarity, and the ordinary market provision of insurance against accident, illness, and old age. These arrangements are not utopian speculation. They are historical fact, and the historical record on this point is one of the most consequential and least known features of the modern social-political landscape. I have summarised it in Pathways as follows: That a decent society without state welfare is feasible and that civil society can organise its welfare privately and voluntarily has been demonstrated by the flourishing world of mutual aid societies during the 19th century and early 20th century up to the usurpation of the welfare sector by the welfare state. Private welfare provision in England was nearly comprehensive via the eco-system of various mutual aid societies, also referred to as friendly societies, benefit societies, or benevolent societies. Such societies were based on religious, political, or trade affiliations, organised locally but some also nationally, for purposes of insurance, pensions and savings. They were operating self-sufficiently, without subsidies, and able to more effectively control benefit fraud or free riding (Schumacher, 2024, p.267).

Friendly societies trace back to the burial societies of ancient Greek and Roman artisans, with medieval European and English guilds expanding on this idea of providing funds for funerals to provide mutual assistance in cases of illness and distress. Some of most important mutual aid societies – which established a national network of hundreds of branches – developed from guilds and free mason lodges. The elimination of the economic power of trade guilds implied their refocus on their social support functions. In major cities like London, some guilds (e.g. the “Free Masons” and the “Odd Fellows”) survived by adapting their roles to focus on the social security function. Both had their base in London but had established branches (called ‘Lodges’) across the country.

Here is how William Beveridge described the widespread and varied ecosystem of friendly societies of early 20^th Century Britain:

“Friendly societies were and are creatures not of Parliament but of brotherly feeling and felt need in small groups of ordinary men throughout the length and breadth of Britain. … Organization in small units is characteristic not only of the early friendly societies but of friendly societies at all stages of their history, at least till the coming of State insurance. The very full returns published by the Registrar for the year 1905 make possible a complete survey of the field. There were at that time in Britain about 20,000 branches of the societies known as Affiliated Orders; as is shown later, each of these branches, though in federal union with others, had its own life and was for many purposes self-governing; the average membership of a branch was 133. There were in addition about 6,700 friendly societies without branches; some of these were specialized on purposes other than provision against sickness and some were large. But the great bulk were small, independent units comparable in size to the branch of an order and doing the same work of provision for sickness; half of them had less than 100 members each, and nearly another third had from 100 to 200 members each.” (Beveridge, 1948, p.30). The small scale and local character of the friendly societies, or at least the small scale of the individual branches are an important aspect of their functioning. The people involved knew each other and this helped with the sense of solidarity as well as with social control preventing abuse.

The historical record of friendly societies carries crucial analytical weight. The standard defence of the welfare state proceeds on the implicit assumption that, in its absence, the working population would simply be exposed to misfortune without provision — that the welfare state arose to fill a vacuum of unmet need. This assumption is false. The welfare state arose by displacing an existing and comprehensive system of voluntary provision that was, by every measurable criterion, functioning well. In England by the early twentieth century the great majority of working men were members of one or more friendly societies, paying small regular contributions in exchange for sickness benefit, unemployment cover, old-age provision, funeral expenses, and access to the society’s medical practitioners. These private civil society social security arrangements via mutual aid societies also included an indirect solidaric redistribution. Since all paid the same contributions and all were equally eligible to the same benefits if falling on hard times, naturally the healthy, strong and young subsidised the sickly, frail and old.

These were not vestigial charitable scraps but the substantive infrastructure of working-class welfare provision. They were administered by elected officers drawn from the membership, with arrangements that produced exactly the social outcomes Spengler invokes against the political wage: members understood the relationship between contributions and benefits; fraudulent and frivolous claims were checked by peer scrutiny; the dignity of mutual provision among equals was preserved; and the populations served by these societies retained the substantive sense of self-reliance and reciprocal obligation that Spengler identifies as the casualty of the political pensioner-structure.

Several features of this historical record deserve emphasis because they directly answer the standard objections to a libertarian withdrawal of the welfare state. First, the societies operated “self-sufficiently, without subsidies.” They were not loss-making operations propped up by external resources; they were actuarially sound mutual insurance arrangements financed entirely by the contributions of their members. Second, they “more effectively control[led] benefit fraud or free riding” — a result of the local, voluntary, peer-administered character of their operation, which contemporary centralised welfare bureaucracies cannot replicate. The administrator of a friendly society knew, personally or through one degree of separation, whether a claimant was genuinely incapacitated or was attempting to work the system. The administrator of a national benefits agency cannot know this and structurally cannot be brought to know it; the system absorbs fraud as an unavoidable operating cost, and that absorbed fraud progressively corrupts both the financial sustainability of the system and the moral self-conception of its claimants. Third, the societies were organisationally diverse — “based on religious, political, or trade affiliations, organised locally but some also nationally” — reflecting the differentiated communities they served. The substantive moral and social meaning of mutual provision varied across societies, and members chose societies whose ethos matched their own. This is the polycentric, voluntary, discursively-informed pluralism for which I argue throughout Markets and Discourses, applied to the specific case of provision for misfortune.

English Civil Society’s Resistance and Its Defeat

Providence has given to no man an indemnity from affliction, disease and death; it is a duty, therefore, that every man owes to himself and family, to provide against these exigencies and that distress which inevitably attends their visitations.

Preface to original rules of Hearts of Oak Benefit Society, June, 1842.

The political history of the friendly society movement’s resistance to state displacement is almost entirely absent from the standard narratives of the welfare state’s creation, which tend to portray the state’s assumption of welfare functions as a benign gift to society, in response to a supposed absence or inadequacy of social security. The history is, in fact, a protracted struggle extending over seven decades, in which a large, well-organised, financially solvent voluntary sector mounted sustained resistance to each successive wave of political-administrative encroachment. That it ultimately lost is not evidence that it was inadequate; it is evidence of the political-structural logic that Spengler’s analysis predicts.

The first phase of organised resistance came in the 1880s, when the movement united to defeat Canon Blackley’s proposal for compulsory national insurance — the earliest serious legislative attempt to establish state-administered welfare. The Manchester Unity of Oddfellows and the Ancient Order of Foresters jointly convened a conference of eighteen friendly societies representing nearly two million members, and mounted a coordinated parliamentary campaign against Blackley’s scheme. Their objection was principled and precise: compulsory state insurance would undermine the actuarial self-governance and peer-accountability that made their system function. The argument was not merely self-interested. It was the argument from the “voluntary principle” that had sustained the movement since its origins in the eighteenth century: that the efficacy of mutual provision depended on the voluntary character of the membership bond, which created the incentive structures of genuine reciprocity, fraud-resistance, and self-discipline that compulsion would destroy. The societies gave formal evidence to the 1887 House of Commons select committee examining Blackley’s proposals, which duly found itself “unable to agree that the scheme was actuarially sound.” The resistance carried the day.

The second phase concerned old-age pensions. When the Liberal government introduced state pensions in 1908, the societies discovered that the state was clawing back pension entitlement from individuals who received friendly society sick pay — directly penalising the most provident members of the working class for their prior self-provision. The Ancient Order of Foresters protested and documented the consequences: many of its members had been entirely deprived of the state pension, and many more had been partially deprived, as a direct result of receiving friendly-society benefits. This was the social-justice apparatus in Hayek’s precise sense: the state redistributing to the least self-reliant, and actively penalising those who had exercised exactly the self-reliance the state ostensibly wished to encourage. The societies lobbied, published, and protested. They achieved partial legislative corrections on related earlier issues but could not reverse the fundamental direction of policy.

The third and most decisive phase was Lloyd George’s National Insurance Act of 1911 — and here the political history reveals a strategic manoeuvre that constitutes one of the most instructive episodes in the political economy of voluntary sector displacement. The friendly societies initially opposed the 1911 Bill. Their objection, again, was on voluntary-principle grounds: compulsory state insurance would compromise the institutional basis of their efficacy. Lloyd George’s response was not to confront this opposition directly but to neutralise it by incorporation: he offered the friendly societies the role of “approved societies” — administrators of the new state scheme, with the state underwriting the basic benefit guarantees and thereby substantially relieving the societies’ actuarial pressure. The societies accepted. The conferences of 1913 and 1914 record the resulting disillusionment with brutal clarity. One contemporary account, reporting on the conferences of the major orders, noted that “active opposition is, of course, non-existent, the general feeling being that the inevitable must be accepted, and the best must be made of things as they are.” The Grand Master of the Manchester Unity of Oddfellows “laid particular stress upon this loss of independence” and “attached the principal blame for this state of affairs on the members of Approved Societies themselves” — a remarkable piece of self-diagnosis. The societies had been absorbed into the system they had opposed.

The political consequences of absorption were precisely those that the voluntary-principle argument had predicted. Having accepted the role of approved-society administrators, the friendly societies’ resistance to further state expansion was structurally compromised: they could protest from within the system but could no longer credibly offer themselves as a complete alternative outside it. In 1915 the National Conference of Friendly Societies (NCFS) joined forces with the industrial assurance companies in objecting to any proposals for the formation of a “state friendly society,” and in 1917 the president of the National Conference warned that it would be “a bad day for this country” if the state ever attempted to raid the accumulated savings of friendly-society members to subsidise the statutory scheme. These were genuine protests but structurally weakened protests: the societies were now dependent on the state scheme for part of their financial basis, which meant that any confrontation with the state risked the very subsidies they had accepted. Lloyd George’s 1911 manoeuvre had converted the movement’s opponents into its sub-contractors.

By the interwar period the societies’ resistance had largely transformed from principled opposition to competitive lobbying for a share of the expanding state apparatus. In 1920 they lobbied unsuccessfully for a role in the administration of unemployment insurance. They deployed their parliamentary agents — professional lobbyists employed by the major orders — and their sponsored Members of Parliament to protect their institutional interests within the system rather than to challenge the system itself. By the time Beveridge drafted his 1942 Report, the major orders were pressing for a continued role within the new comprehensive system rather than defending the voluntary principle against the system. The language of resistance had been replaced by the language of accommodation. The final blow came in 1946–48. The Attlee government’s implementation of the welfare state bypassed the friendly societies entirely. It wanted a complete and exclusive state administrative machine, and it built one. The Ancient Order of Foresters issued its “Circular 3/46” protesting at the exclusion. The protest came too late and from too weak a position. The Labour government bypassed the friendlies, structuring a complete and exclusive administrative machine of its own. The loss of funding and higher state benefit rates drove many friendlies out of existence.

The most poignant episode in the entire history was the response of William Beveridge himself. Beveridge had intended his 1942 Report to preserve a role for voluntary societies within a framework of state minimum provision. He wanted to keep what he called the “bottom-up approach” that “preserves many instances of lively democratic self-help in small societies.” Attlee’s government cut the societies out entirely. Beveridge responded in 1948 with his “Voluntary Action: A Report on Methods of Social Advance”, in which he made the case for the societies’ indispensable role and warned that “in a totalitarian state or in a field already made into a state monopoly, those dissatisfied with the institutions that they find can seek remedy only by seeking to change the government of the country” (Beveridge, 1948, p.77). The intellectual author of the welfare state wrote a protest against the way his own creation had been implemented. He was ignored.

Three analytical conclusions follow from this history. First, the friendly societies were not passive victims of an irresistible historical tendency. They organised, lobbied, published, protested, and at least once — in the defeat of the Blackley scheme in the 1880s — achieved a genuine political victory. Second, the decisive defeat came not through direct confrontation but through the incorporation trap of 1911: the offer of a share in the state apparatus in exchange for abandoning the voluntary-principle and their opposition. This is the political-structural logic Spengler identifies in a different register: the demagogic class defeats independent institutions not by destroying them directly but by absorbing them, making their continued existence conditional on the apparatus’s continued operation. Third, the memory of the system was erased not spontaneously but by the positive action of the state. The libertarian project of restoring voluntary mutual provision therefore faces not merely the institutional challenge of rebuilding a sector the state destroyed, but the deeper cultural challenge of recovering a memory the state systematically suppressed. The historical record of the friendly societies’ seven-decade resistance is part of that recovered memory. Fourth, and decisively for the present argument, the mutual aid system was destroyed by political displacement, not by its own inadequacy. The Liberal welfare reforms of 1906–1911 in Britain, and their inter-war and post-war extensions through Beveridge and the establishment of the National Health Service, did not arise because the friendly societies had failed. They arose because the political class identified the friendly society infrastructure as an obstacle to its own programme of political-administrative consolidation, and acted to displace it. The state-administered National Insurance scheme of 1911 deliberately incorporated and then progressively subordinated the friendly society sector, replacing genuine mutual insurance with compulsory contribution and politically-determined benefits. The subsequent decades saw the systematic atrophy of the voluntary sector as state provision crowded it out: members no longer needed to subscribe to private societies for benefits they were now compelled to fund through taxation, and the societies progressively lost their economic basis. By the second half of the twentieth century, the institutional memory of working-class self-provision had largely been erased, leaving the political class free to claim that no alternative to its own welfare-administrative apparatus had ever existed or could exist. This claim is false. The alternative existed for the better part of a century, served the great majority of the working population effectively, and was destroyed by political action precisely because it gave the working population a substantive independence that stood in the way of the political class’ accrual of power.

The Spenglerian analysis of the political wage, the libertarian analysis of welfare-induced dependency, and the historical record of the friendly societies thus converge on a single conclusion. The contemporary welfare state is not the necessary instrument of humanitarian provision; it is the structural mechanism through which the political class displaced a pre-existing voluntary system that gave working populations independent dignity and self-reliance, and replaced it with a politically-administered system that produces the dependent, infantilised, “uprooted” population the political-demagogic class requires for its political existence. The libertarian roll-back of the welfare state is therefore not a withdrawal of provision but a restoration of provision in its proper, voluntary, dignity-preserving form. As I argued in the previous section on the search for a mobilizable mass, this is also why the political-administrative class will resist welfare reform with extraordinary intensity even when the economic case against welfare is overwhelming: at stake is not merely the level of public expenditure but the existence of the social formation through which the political class secures its political constituency. The historical record of the friendly societies shows that the alternative is feasible, has functioned, and was destroyed only by deliberate political displacement. It can be restored, and the populations currently dependent on the welfare-administrative apparatus can be progressively re-incorporated into voluntary structures of mutual provision that will return to them what state welfare has progressively eroded: the substantive sense of self-reliance and the dignity of reciprocal obligation among equals.

Only the basis of such a regained self-reliance and existential independence can a self-confident, active, forward looking society be built, a society where Spengler’s Faustian virtues can flourish and gain respect; virtues such as the quiet acceptance of the burdens that a commitment to an important mission demands, the respect for superior skills, knowledge and achievements, and finally the admiration for and deference to those distinguished by such achievement. Spengler rightly identified the widespread, taken for granted aspiration to live up to these virtues and their social elevation as the precondition for societal strength and flourishing, enabling such societies, and only such societies, to carry out projects of world-historical significance. For me the libertarian revolution is such a project, which in turn will enable the proliferation of many more momentous projects.

The Surplus-Value Doctrine and the Politics of Quality

One further passage from Spengler deserves attention, because it identifies the ideological mechanism through which the political wage was originally justified and through which its contemporary variants continue to be justified. The doctrine of “surplus value” remains, beneath the surface of contemporary policy discourse, the operative theory of value distribution wherever the political extraction of incomes from the productive economy is proposed:

“Once the doctrine of “surplus value” had power to sway the undeveloped reasoning of the masses: the whole output of industrial production was equal to the value of manual labour and had to be allotted accordingly. What the manufacturer deducted from it — for upkeep of the works, payment for raw materials, salaries, interest: the “surplus value,” in fact — was robbery. The administration, the inventors, the engineers, all did no work whatever; and if they had done, mental work such as theirs, which was a kind of doing nothing, had no real value. It was the same “democratic” tendency which scorned and would have destroyed quality of any sort, and thought only in terms of quantity, even in manual work. The distinction between skilled and unskilled labour had to be abolished. Both should receive equal payment. Piece-work and superior production were branded as betrayal of the cause. This attitude has triumphed, and postwar Germany in particular is its triumph. It has eliminated competition among the workers, stifled the impulse to attain higher levels of skill, and thereby reduced the total output.” (Spengler, 1934, p.142)

Spengler points to the irony that in Russia, soon after the revolution, the new masters, now responsible for state and economy left this egalitarian ideology behind. “As soon as “the goal” had been realized long working hours, low wages; the widest possible gulf — wider even than in America — between the payment of skilled and of unskilled work” (Spengler, 1934, p.143) was re-established. Clearly, the levelling and denial of differences in the difficulty, required skill and societal value of different types of work is denying something very fundamental at the root of all large, complex, high-performance societies. As Spengler points out these differences can be very stark and they are the starker the more developed a society is.

“There is higher work, and lower: nothing can deny or alter this truth. It is the expression of the fact that Culture exists. The higher the stage of development in a Culture, the more powerful its creative force, the greater the difference between determining and subordinate actions of all kinds; whether political, economic, or artistic. For Culture is ordered, intellectualized life, a maturing and self-perfecting form, which calls for an ever higher grade of personality. There is work for which one must have an inner call, and other work that one must do, because one can do nothing better to earn a living. There is work for which only a very few men of superior rank are competent, and other work which is valuable only in terms of duration and volume. … It cannot be altered either by Rationalist or by sentimental-romantic equality-talk.” (Spengler, 1934, pp.146-147)

These passages identify two distinct but related errors that the surplus-value doctrine systematically propagates. The first is the denial that head-work — the work of investment, administration, invention, engineering, organisation — constitutes real economic value. The second is the levelling of skill-differentials within the category of manual labour itself. Both errors persist in contemporary policy discourse, though they have shifted vocabulary. The first survives in the contemporary attack on “shareholders,” “executives,” “the one percent,” and “financial capital” as parasitic appropriators of value supposedly produced by ordinary workers — an attack that requires the systematic denial of the actual coordinative, anticipatory, innovative, and risk bearing contributions of these roles. The second survives in the contemporary attack on merit-based differentiation in education, employment, and pay; in the imposition of standardised wage tariffs across heterogeneous productive units; and in the demoralisation of high-performance norms by the elevation of “equity” of outcome over recognition of differential contribution.

Spengler’s closing remark is decisive: this attitude “has eliminated competition among the workers, stifled the impulse to attain higher levels of skill, and thereby reduced the total output.” This is the actual mechanism by which the political wage destroys economic performance, and it identifies the deepest connection between the worker-cult of §III, the political wage of this section, and the welfare-pensioner society. All three operate by suppressing the differential signals — of wage, of recognition, of consequence — through which productive systems coordinate the development of skill, the allocation of effort, and the matching of capacity to task. They achieve their political-redistributive effect by destroying the informational and motivational structure on which productive output depends. The libertarian programme, in this register, is the restoration of these differential signals, and with them the restoration of the productive performance they make possible.

It is worth noting that this Spenglerian analysis maps rather well onto Mises’s and Hayek’s formal arguments about the impossibility of economic calculation under socialism. Mises showed that without competitive bidding for resources there can be no rational allocation. Hayek showed that the relevant knowledge cannot be centralised. Spengler, without the formal apparatus, identified the same impossibility from the cultural side: the political determination of wages, the suppression of skill-differentiation, and the moral inversion that treats coordinative head-work as parasitism together destroy the informational structure through which the productive economy operates. Two independent intellectual traditions - the Austrian-economic and the Spenglerian-cultural - converge on the same diagnosis. The political substitution for the productive economy’s own coordinative mechanisms is not a matter of degree but of kind. It does not merely make the economy less efficient; it dissolves the very informational basis on which economic rationality operates. The libertarian remedy is correspondingly not a matter of degree but of kind: the intrusive political-administrative apparatus must be rolled back to permit the productive economy’s own coordinative mechanisms — prices, profits, competitive bidding, skill-differentiation, head-work recognition, and the rest — to re-emerge in their full operative force.

The Faustian Character and Its Proper Home

Only that man counts who can take risks, who has the courage to see and accept things as they are. The age is approaching — nay, is already here — which has no more room for soft hearts and weakly ideals.

— Spengler, The Hour of Decision

The most distinctive and, for our purposes, the most valuable element of Spengler’s political writing is not the diagnosis of decay but the positive evocation of a character type. He returns to it again and again. Consider the opening of The Hour of Decision:

“Is there today a man among the White races who has eyes to see what is going on around him on the face of the globe? … I do not speak of the educated or uneducated city crowds, the newspaper-readers, the herds who vote at elections — and, for that matter, there is no longer any quality-difference between voters and those for whom they vote — but of the ruling classes of the White nations, in so far as they have not been destroyed, of the statesmen in so far as there are any left; of the true leaders of policy, of economic life, of armies, and of thought. Does anyone, I ask, see over and beyond his time, his own continent, his country, or even the narrow circle of his own activities?” (Spengler, 1934, p.24)

And his portrait of the true governor and his inevitable solitude, as he remains misunderstood and opposed:

“Frederick the Great experienced the first of these types of opposition; Bismarck almost fell a victim to the second. Only later generations, and not even they, can appreciate the grândeur and creativeness of such leaders. … Germans in particular are great at suspecting, criticizing, and voiding creative action. They have none of that historical experience and force of tradition which are congenital with English life. A nation of poets and thinkers — in the process of becoming a nation of babblers and persecutors. Every real governor is unpopular among his frightened, cowardly, and uncomprehending contemporaries.” (Spengler, 1934, p.28)

Spengler describes, repeatedly, a character type defined by several constituent traits worth enumerating. There is realism: the willingness, indeed the appetite, to see the world as it actually is, including its harshness, rather than as one would prefer it to be. There is courage: the willingness to act on this realism in the face of opposition, including the opposition of opinion, the press, the “know-betters.” There is purpose: a commitment to a project larger than oneself and one’s comfort, often without immediate reward and frequently against personal interest. There is dignity: the bearing of one who knows the value of what he does, and does not require external validation to maintain it. There is duty: a self-binding to the demands of the project, which appears to the outside as discipline and to the inside as the necessary structure of meaningful action. And there is mission: the sense that one’s individual life is bound up with something historically significant, the destiny of one’s firm, one’s field, one’s civilisation.

This is precisely the character type Ayn Rand articulated and brought to life in her libertarian novels ‘The Fountainhead’ and ‘Atlas Shrugged’ with the leading figures of Howard Roark and John Galt. The Randian hero is defined by the capacity to grasp reality, a creator, a man whose mastery is intrinsic to who he is, a man that commands respect through his inner strength and outer effectiveness. He refuses the mantle of victim or supplicant; he claims the world as the proper field of his creative action.

Spengler’s error — and here we must part company with him — was to locate this character type in a specifically aristocratic-Prussian, soldier-statesman archetype, and to suppose that its survival required a Caesarist political order capable of imposing rank, duty, and command on a society that had lost the instinct for them. The Caesarist hope is, both empirically and conceptually, mistaken. Empirically, every actual twentieth-century Caesarism degenerated into a personalised tyranny that did not regenerate the Faustian virtues but consumed and discredited them, dragging into final disrepute precisely the aristocratic-military ethic Spengler hoped to revive. Conceptually, the Faustian character cannot in fact be commanded into existence. It is a character of self-direction, self-binding, self-respect. It must be chosen, lived, and confirmed by those who carry it. It can flourish only where there is genuine scope for individual undertaking, genuine risk that can be borne or refused, and a public ecology of recognition that distinguishes substantive achievement from mere position.

That ecology is precisely what the contemporary interventionist state suffocates. The proceduralizing of corporate management; the legal-regulatory transformation of substantive decision into defensive paper-trail; the moral-journalistic policing of every controversial undertaking; the welfare-induced bifurcation of society that segregates the productive from the chronically dependent; the educational regime in which the bright young are taught grievance theory rather than the practice of disciplined achievement; and above all, the redistributive apparatus that systematically transfers the rewards of substantive achievement to a parasitic class of credentialed administrators and their clientele — these together constitute a structural war on Faustian character, conducted not by open force but by the patient operation of incentives and humiliations.

The libertarian order, by contrast, is the only social order in which the Faustian character can re-emerge at scale. Not because the libertarian order is itself “Faustian” in any romantic sense — the libertarian is no romantic — but because it is the order that opens space for the substantive undertaking, accepts the substantive risk, and confers genuine recognition on substantive success. The entrepreneur who founds an industry; the engineer who delivers a working system; the legal entrepreneur who builds a competing jurisdiction; the founder of a new town or a charter city; the discourse-leader who develops and defends a body of argument and gathers a following; the scientist whose original findings withstand criticism and are taken up — each of these is the contemporary embodiment of what Spengler honoured. Each requires the libertarian conditions to flourish: scope for action, accountability for results, freedom from the proceduralising apparatus of the contemporary state, free, not forced association, chosen mutual, not imposed solidarity, and a public sphere in which substantive achievement still commands respect.

Rank, Duty, and Discipline Without Political Coercion

Once Spengler’s conflation of structure with stratification is corrected, the apparent gulf between his ethical vision and the libertarian project narrows dramatically. Spengler’s deeper concern in evoking duty, rank, and discipline was that the modern world had lost the capacity to bind itself to anything demanding. He puts the point with characteristic force:

“The age is mighty, but all the more diminutive are the people in it. They can no longer bear tragedy, either on the stage or in real life. They crave happy endings of insipid novels, so miserable and weary are they. … If few can stand a long war without deterioration of soul, none can stand a long peace. This peace period from 1870 to 1914, and the memory of it, rendered all White men self-satisfied, covetous, void of understanding, and incapable of bearing misfortune. We see the result in the Utopian conceptions and demands which today form part of every demagogue’s program; demands made on the age, on the State, to parties, and in fact to everyone except themselves, in complete disregard of the limits of feasibility or of duty, and sacrifice.” (Spengler, 1934, p.36)

Spengler’s observation here is acute and continues to apply. The Utopian challenges in “complete disregard of the limits of possibility or of duty, doing, and forgoing” are the staple of contemporary political discourse, on both right and left. The capacity to subordinate oneself, even voluntarily, to a project beyond one’s immediate preferences and pleasures is in increasingly short supply. The “soft hearts and weakly ideals” that Spengler mocked are not mere weakness of will. They are the disabling structural feature of a society that has so thoroughly democratised respect that no real demand can any longer be made on anyone.

The libertarian order, properly conceived, does not deny this concern. It answers it differently. The hierarchies, disciplines, and self-bindings that Spengler thought required the framing of an authoritarian state can in fact emerge endogenously and voluntarily within free institutions and indeed can flourish there better than under coercive frames. A great firm has a hierarchy of merit and capacity — not imposed from outside but assembled by the founder and accepted by those who choose to join and to remain. A great scientific community has a hierarchy of contribution and standing — again, neither legislated nor coerced, but emerging from the practice of giving and asking for reasons that constitutes the community as such. A great undertaking of any kind, from a major infrastructure project to a new legal jurisdiction, from a new private university to a new scientific journal or publishing house, requires and produces internal hierarchies of competence and authority that participants accept because the undertaking matters to them and because they understand that the project cannot be done well in any other way.

The “duty” of the Faustian character is, in this reading, a self-imposed duty: the bond one assumes when one commits oneself to a project, an institution, a craft, or a discourse. It is the bond an entrepreneur accepts toward his investors, his employees, his customers, and his own conception of what his firm is for. It is the bond a scholar accepts toward the truth, his colleagues, and the evolving body of argument in his field. It is the bond a craftsman accepts toward his tools and his materials and the quality of his output. None of these requires a state to enforce them, and indeed the state’s attempt to enforce them reliably destroys them. They are sustained by the structure of the project itself, by the participants’ commitment to its success, and by the surrounding public ecology of recognition.

The “rank” of the Faustian character is, similarly, a rank of achieved and recognised competence within concrete undertakings, not a permanent caste-position. The libertarian order has no objection to hierarchy, rank, ownership or leadership. It has a profound objection to coerced hierarchy and an equally profound objection to politically enforced levelling. The competitive and discursive processes I have argued for in Markets and Discourses generate hierarchies continuously: hierarchies of firm performance, of professional standing, of intellectual contribution, of demonstrated competence. These hierarchies are entirely compatible with, indeed dependent on, the formal equality of participation in market and discursive processes. What is excluded is the political fiction — the great gift of nineteenth-century parliamentarism to twentieth-century totalitarianism — that the “people,” speaking through whichever demagogic representation has currently captured the legislative machinery, may legitimately overturn these endogenous hierarchies in the name of an abstract “equality” whose actual function is to subordinate substantive achievement to a parasitic political class.

In this respect, the libertarian completes a movement that Spengler began but could not bring to its proper conclusion. Spengler rightly perceived that the dissolving force of late 19^th Century and early 20^th Century socialist politics and ideology had been the abstract “equality” of mass democracy and its erosion of substantive distinction. He responded by seeking to restore a substantive hierarchy through political means — a Caesarist state capable of imposing rank on a society that had forgotten it. This response was self-defeating. The political imposition of rank produces only the resentful conformity of the subjugated, never the genuine recognition of substantive distinction. The libertarian response is structurally different: dissolve the political mechanism that does the levelling, allow the functional subsystems to operate according to their own internal codes, and substantive hierarchies will re-emerge of their own accord, sustained by the participants in the concrete undertakings that require them.

Discourse as the Modern Form

A further Spenglerian theme requires careful handling: his hostility to “paper systems.” He held this hostility with passion. Recall the passage already quoted:

“Rationalism is at bottom nothing but criticism, and the critic is the reverse of a creator: he dissects and he reassembles; conception and birth are alien to him. Accordingly, his work is artificial and lifeless, and when brought into contact with real life, it kills. All these systems and organizations are paper productions; they are methodical and absurd and live only on the paper they are written on.” (Spengler, 1934, p.30)

Spengler considered the eighteenth-century rationalist construction of political systems on the model of geometric proof, and its twentieth-century descendants in problem novels and party programmes, as themselves a symptom of the urban-intellectual disease. His targets here include not just Marxism and pacifism but, by implication, any project that proposes to redesign a working social order on the basis of theoretical first principles. The libertarian, with his market processes and his discourse capitalism, might at first appear to be exactly this kind of paper-rationalist constructor.

The appearance is misleading, and the libertarian rebuttal converges with Spengler’s own deepest insight. The constructivist treats society as a system to be designed and implemented from a blueprint; both the comprehensive Jacobin and the comprehensive Communist do exactly this. Hayek’s decisive contribution, on which the libertarian project rests, was that this is impossible: the relevant knowledge is dispersed, local, often tacit, and cannot be centralised. Market processes are valuable precisely because they coordinate without requiring centralised design. Discourses, similarly, in the Habermasian sense I have argued for in Markets and Discourses, are not blueprints to be implemented but open, fallible, evolving processes of giving and asking for reasons. As I have put it: Discourses are not markets and operate rather differently. (The metaphor of a “marketplace of ideas” does not imply that discourses are markets.) There is a crucial social difference between market exchanges and discursive “exchanges.” These are very different types of interaction, each with its distinct rules of engagement and criteria of success: Discourses, unlike markets aim for consensus. They operate via the exchange of arguments rather than via the exchange of commodities or services. … Arguments are not a matter of mere personal preference but a matter of evidence-based reasoning (Schumacher, 2025).

What the libertarian project shares with Spengler is, in fact, suspicion of the political pretension to comprehensive social design. What it rejects is Spengler’s further inference that this suspicion implies a return to traditional or aristocratic hierarchies imposed from above. The libertarian inference is the opposite: precisely because no central design can capture the complexity of modern functionally differentiated society, the framing institutions must be ones that permit decentralised experimentation, discovery, and adaptive selection. Markets and open discourses are exactly such institutions. They are not rationalist constructions; they are the conditions under which rationality can operate within each functional subsystem without overreaching itself into a master code.

It is worth noting that Spengler’s most acute aphorism on this subject — “Rationalism is at bottom nothing but criticism, and the critic is the reverse of a creator” — is itself only half right, and the half he missed is the one Habermas later developed. Criticism is indeed not creation. But sustained, disciplined, public criticism is the indispensable framing condition under which creation operates well rather than badly. The entrepreneur whose plans are never publicly tested against arguments and evidence, the engineer whose designs are never exposed to critical review, the legislator whose proposals never face serious discursive challenge, all produce worse results than those whose work is permanently exposed to disciplined criticism. The journalist-critic Spengler rightly despised was the dissolving critic without a counterpart in creation. The libertarian discourse-society does not abolish criticism; it makes criticism the framing condition of consequential creation, while withdrawing from criticism the political power to terminate creation by collectively binding decree.

The Hour of Decision Recurs

In such times, there are, under whatever names, two natural parties, two fronts in the class war, two internal forces and tendencies, and only two, no matter how many party organizations exist or whether such exist at all. … Here every individual is confronted with an inevitable choice.

— Spengler, The Hour of Decision

Spengler thought he was writing in an hour of decision — a historical moment in which the long-deferred confrontation between the dissolving forces of urban-democratic levelling and the conserving forces of substantive hierarchy could no longer be evaded. He was right about the moment and wrong about its character. The choice he posed was Caesarism or Bolshevism, soldier-statesman or labour-leader, and the actual history of the following twelve years vindicated none of his hopes and most of his fears.

We stand, I have argued throughout my recent writing, in another hour of decision — differently constituted but no less urgent. The post-Fordist productive revolution has been under way for half a century. The institutional framing inherited from the Fordist-social-democratic settlement is, in country after country, demonstrably incapable of digesting it. Productivity growth has slowed, in some advanced economies effectively to zero. The state grows, but the substantive output it claims to organise stagnates. As I have argued in my 2025 article Pathways towards a Stateless World Society: The libertarian hypothesis is that once this increasingly paralysing current process can be reversed and a process of liberalisation is under way, the prosperity gains that will be unlocked in each step will spurn the process on further. The prospect of such a process might be termed — borrowing a phrase from Karl Marx — “the withering away of the state,” albeit not under communism but under capitalism. Political parties will gradually transform into think tanks or advocacy groups and political leaders will transform from decision makers to proselytizing discourse leaders (Schumacher, 2025, p.261).

Spengler’s hour of decision demanded courage to see and act. So does ours. Spengler’s call for a return of soldier-statesmanship was, in the event, a wrong turn. But the underlying demand — that someone, somewhere, must accept the responsibility of building rather than continuing to drift in the dissolution of inherited institutions — remains. In our hour of decision, the bearers of that responsibility will not be soldier-statesmen at the head of standing armies. They will be founders: of new technologies, of firms, of special economic zones, of jurisdictions, of intellectual movements, of secessionist polities, of new (private) cities, of new schools, of new courts, of new currencies. They will be entrepreneurs in the full and original Spenglerian (and Schumpeterian) sense — men and women of “ambition, love of adventure, who delight in battle and danger” — whose battles will be technological, commercial, intellectual, and at times political, but who will be Faustian in exactly the sense Spengler honoured: in the depth of their commitment to a difficult project, the realism of their assessment of the obstacles, the dignity of their bearing under criticism and ingratitude, the duty they accept to those who depend on them, and the sense of mission that connects their work to the destiny of the civilisation whose productive forces they propose to liberate and propagate.

Conclusion: A Libertarian Order where Faustian Heroes are at Home

I have argued that the libertarian project can and should integrate substantial elements of Spengler’s ethical and cultural vision while rejecting his political prescription, and that this integration requires the prior correction of his sociological conflation of structure with stratification. The integration is more than an opportunistic borrowing. It corrects a recurrent and, I think, genuine weakness in libertarian self-presentation: the tendency to describe the future free society as if it were merely a society of comfortable consumers pursuing their preferences through market exchange. Spengler’s contempt for exactly this conception of the human good — the “dollar-society” he attributed to the United States, the “primitive pose” of standardised mass existence — is partly deserved, and the libertarian who wishes his project taken seriously by serious people must answer it.

The answer is not that a libertarian order would be a society of pampered consumers, but that the libertarian order is the only one in which the substantively serious life — the life of the founder, the engineer, the scholar, the craftsman, the institution-builder, the discourse-leader, the entrepreneur in the full and original sense of someone who undertakes a difficult thing and bears its risks — can flourish in our historical moment. The contemporary coercive and controlling nanny state, for all its rhetorical commitment to “service”, is in concrete operation the great enemy of substantive individual and societal flourishing. It proceduralises the manager into a compliance-officer, the scholar into a grant-cycle administrator, the entrepreneur into a permitted petitioner before regulatory commissions, the citizen into an object of administrative determinations. The Faustian-Spenglerian character cannot live in this environment, and the loss of that character to our civilisation is, contra Spengler, not a metaphysical destiny but a contingent and reversible ideological and institutional outcome.

The shared diagnosis with Spengler is therefore deep. The destructive forces are largely as Spengler described them: a rootless intelligentsia organising mass resentment into political power; a moral-political fabrication of victim-categories that displace substantive comparative analysis; a media-cultural apparatus committed to dissolving criticism without compensating creation; a parliamentary-administrative complex that has captured the modern state and turned it against the productive forces it depends on; a propertied class that has bought its own peace by capitulating to the moral terms of its dispossessors; an organised demoralisation of the bulk of the population, a relentless regulatory choking and fiscal squeezing of those who still create and on whose substantive achievement the whole structure rests. Spengler saw all of this clearly. We see it more clearly still, because we have seen another century of its consequences. What we have additionally is the diagnostic apparatus to understand why a Caesarist remedy will not work, why every Caesarism degrades into a parody of the substantive ethic it claimed to restore, and why the actual available remedy is the libertarian one: the rolling-back of the political-administrative apparatus through which the dissolution operates, and the re-opening of the institutional space within which the functionally differentiated subsystems of modern society — science, technology, economy, law, discourse — can each operate according to their own logic, and within which the substantive hierarchies of competence proper to each can re-form.

This is the synthesis. Spengler’s ethics, libertarian institutions. The Faustian character, the Faustian undertaking, the Faustian sense of mission — not in a Caesarist polity that imposes them by command, but in a markets-and-discourses society that opens the institutional space for them to be chosen, lived, and recognised. The hour of decision is now, and the decision is not between the two Spengler offered. The decision is between the continued drift of our institutions into a managed dissolution, and the deliberate rolling-back of those institutions to make room for what they have spent a century displacing. The choice is ours. The course of history, which is stronger than all theory and ideological dreaming, will decide for us if we do not.

Let me end with a final quote from Spengler, a rare quote where pessimism and fatalism seem to recede and point to a project:

“The Faustian nations of the end of the twentieth century will be elective affinities of men with a common feeling about life, with the same imperatives of a strong will and naturally with the same language, but without their knowledge of that language constituting either a hallmark or a limitation. They will be men of race — not in the sense of today’s belief in race, but in my sense of it as a matter of strong instincts, among them that superior eye for the things of reality … men who feel themselves born and called to be masters. What matters number? … A man means a great deal as opposed to a mass of slavish souls, pacifists and world-improvers who yearn for quiet at any cost, even that of “liberty”.” (Spengler,1934, p.69)

If we say 21^st instead 20^th Century, men & women instead of men, and “leader” instead “master”, this stands as call to action.

End.

References

Beveridge, W. (2015/1948). Voluntary Action: A Report on Methods of Social Advance. In: Works of William H. Beveridge Book 3, Taylor & Francis

Hayek, F. A. (1982). Law, legislation and liberty: A new statement of the liberal principles of justice and political economy. Routledge & Kegan Paul.

Hayek, F. A. (1988). The fatal conceit: The errors of socialism (W. W. Bartley III, Ed.). University of Chicago Press.

Luhmann, N. (1995). Social Systems. Stanford University Press.

Luhmann, N. (2012). Theory of Society, Volume 1. Stanford University Press.

Luhmann, N. (2013). Theory of Society, Volume 2. Stanford University Press.

Rothbard, M. N. (1973). For a New Liberty: The Libertarian Manifesto. Macmillan.

Schmitt, Carl. (1913/1988). The Crisis of Parliamentary Democracy, MIT Press, Cambridge MA.

Schumacher, P. (2023). Multi-culturalism and Ethnic Pride as Barriers to Prosperity. (

)

Schumacher, P. (2024). Thesis on German Right-wing Populism: A Rationale for the Protection of High-Performance National Cultures. Paper delivered at Battle of Ideas Festival, London. (

)

Schumacher, P. (2024). Pathways towards a Stateless World Society. In C. H. Hoffmann, S. Grell, M. Otto (eds.), Was tun? Wie Freiheitsentrepreneure unser Zusammenleben revolutionieren. Edition Wissenschaft & Praxis, Berlin. (

)

Schumacher, P. (2025). Markets and Discourses — Politics after the Libertarian Revolution.

Schumacher, P. (2026). Thesis on Government Destruction of Corporate Management. (

)

Sowell, T. (1988). Conquests and Cultures: An International History. Basic Books.

Sowell, T. (1994). Race and Culture: A World View. Basic Books.

Spengler, O. (1934). The Hour of Decision: Germany and World-Historical Evolution (C. F. Atkinson, Trans.). Alfred A. Knopf. (Original work published 1933)

Abstract:

This essay engages David Starkey’s thesis that the New Labour governments after 1997 effected a seemingly harmless but in effect radical and detrimental transformation of the British constitution, dispersing the sovereignty of the Parliament among unelected quangos, regulators, devolved bodies, an empowered judiciary, and an unresponsive administrative apparatus. It theorises this transformation through Carl Schmitt’s The Crisis of Parliamentary Democracy (1923) and Walter Bagehot’s The English Constitution (1867). Schmitt supplies the deeper explanation of Parliament’s vulnerability: the contradiction between the liberal principle of parliamentarism (government by genuine discussion among independent representatives, embedded in an educated public sphere) and the democratic principle of the identity of rulers and ruled. Bagehot supplies both the anatomy of the fused, accountable constitution that was lost and a candid account of how a population’s political will is formed, managed, and hedged. Drawing on Habermas’s reciprocity of liberty and popular sovereignty and his distinction between communicative and instrumental action, the essay argues that critical-reflective discourse underwrites the achievements of modern civilisation well beyond the political system. It contends that the complexity and dynamism of post-Fordist, world-societal reproduction defeat any all-purpose chamber, pointing decisively toward decentralisation via markets and discourses in the plural. The essay extends Starkey with a “ratchet thesis”: the delegation of power did not merely relocate sovereignty but expanded the total quantum of state control, since insulation from accountability itself a enables institutional expansion — illustrated through monetary policy, the legally entrenched net-zero framework, the official arm’s-length-state statistics, and a case study of coercive over-intervention in child safeguarding. The decisive caveat departs from Starkey: the remedy is not to repatriate powers to Parliament but to reduce the scope of political control altogether, returning functions to civil society—exemplified by the denationalisation of money. It closes by assessing the political feasibility of a libertarian turn within existing democracy.

1. Introduction: Two Diagnoses of a Single Disorder

David Starkey has lately given an articulate voice to a disquiet that many in Britain feel but few can articulate with precision: that the constitution under which the United Kingdom is now governed is not the constitution it inherited, and that the transformation—accomplished chiefly under Tony Blair—was both radical and largely unannounced. The “Starkey Thesis,” as it has come to be called, holds that the New Labour governments of 1997 onward carried out a revolution in the British constitution: a systematic transfer of powers out of Parliament and into a proliferating apparatus of quangos, regulators, devolved bodies, a newly supreme judiciary, incorporated supranational norms, and into an unelected administrative machine. The Crown-in-Parliament—the historic sovereign—was, on this reading, quietly hollowed out and its substance dispersed among bodies that answer to no electorate and deliberate in no chamber (Starkey 2024a, 2024b, 2024c, 2025; see the note on Starkey’s sources below).

The Starkey thesis, according to David Starkey’s interlocutor and co-protagonist at PopCon (Popular Conservatism) Mark Littlewood, is meant to answer the puzzling question: why after thirteen years of Conservative government is the UK less conservative and more left than ever? Put more generally, the question is: Why do conservative voter preferences no longer translate into effective policy and state action? For example, why has immigration continued to go up despite repeated polls and election promises pointing in the opposite direction? David Starkey’s answer: parliament - and with it the demos – has been disempowered by the radical constitutional changes brought about by the Blair and Brown Labour governments.

My purpose here is not to rehearse Starkey’s constitutional history but to deepen and reframe it through a theoretical lens that he himself does not employ: Carl Schmitt’s The Crisis of Parliamentary Democracy of 1923. Schmitt offers something Starkey’s narrative lacks—an account of why the institution proved so vulnerable to hollowing-out in the first place. The answer, I shall argue, lies in a contradiction that Schmitt diagnosed a full century ago: the contradiction between the motivating principle of parliamentarism and the motivating principle of democracy. Once that contradiction is grasped, the Blairite revolution appears not as a sudden act of constitutional vandalism by one unusually careless government, but as a late and predictable symptom of a disorder built into the foundations of the modern liberal-democratic state.

I shall also draw on Walter Bagehot’s The English Constitution (1867), the classic anatomy of the system Starkey says Blair destroyed. Bagehot is indispensable here for two reasons. First, his distinction between the “dignified” and the “efficient” parts of the constitution gives us the conceptual instrument for understanding what the Blair settlement actually did. (He left the dignified facade standing but removed large tracts of the efficient power.) Second—and this is the reason I dwell on him at some length—Bagehot was candid, in a way that is almost unrepeatable in our own moment, about the relationship between the constitution and the population it governs: about deference, about the “unequal development” of understanding within different segments of the population, about how political will is not so much expressed by the mass as formed, managed, and hedged on its behalf. Bagehot wrote before the full arrival of mass democracy and universal suffrage, and his frankness about the demos is precisely what makes him useful for the argument I want to make.

My own position, stated here at the outset to guide the reader, has three parts. First, I accept a great deal of what Schmitt says about the original principle and purpose of parliaments. Genuine deliberation—the public contest of argument in pursuit of truth, embedded within a wider public sphere of the educated and accomplished pillars of society—was the animating idea of the institution, and it is this idea, not mere expediency, that explains the constitutional rules under which eighteenth- and nineteenth-century parliaments operated. Second, I accept Schmitt’s further claim that parliamentarism so understood stands in tension with modern mass democracy, and that this tension is the source of the pathologies that critics across the spectrum deplore. Third, I agree with Starkey that the Blairite transfer of power away from Parliament was a grave constitutional injury—but I attach to this agreement a decisive caveat. I do not believe that the remedy is to return those powers to a Parliament that does patently not have the requisite knowledge, intellectual competency or capacity to even properly attend to and discharge its current powers. This is not only a question of the lamentable quality of parliamentarians but much more a question of the – in comparison to the 18^th and 19^th Century - vastly increased complexity of society. The widespread frustration and disillusionment with “our democracy” that the Starkey Thesis so effectively focusses and mobilises against the Blairite revolution and its unaccountable quango state and out-of-touch “blob” or “deep state” should be turned to a different end: not the re-empowerment of the hapless MPs, but the reduction of the scope of political control over society as such. The powers that have drained out of Parliament should continue draining—past the quango, past the regulator, and out of the political system altogether, into civil society, giving more scope and degrees of freedom to private initiatives, to entrepreneurship, to individual and communal self-determination. This is the libertarian inflection that distinguishes my reading of Starkey from his own, and it is the destination toward which this essay travels. (In a long private conversation David Starkey was himself, up to a point, rather receptive to this shift of quango powers into civil society.) I contend that all the key issues that upset voters and that fuel their political disillusionment – immigration into the welfare system, the housing crisis, the costs of living crisis, NHS waiting lists – are creations of an activist state that therefore can best be addressed and solved via the roll-back of state intervention.

Between the diagnosis and that destination lie three further considerations that deepen the case. I take up Jürgen Habermas’s attempt to reconcile liberty and popular sovereignty as reciprocal—a reframing of Schmitt’s contradiction that locates legitimacy in discourse, but that stops short of severing discourse from sovereign command. I then draw on Habermas’s distinction between communicative and instrumental action to show that critical-reflective public discourse underwrites the achievements of modern civilisation far beyond the political system, and that even the interests and facts presupposed in market bargaining must first be formed in non-instrumental communication. And I argue, most importantly, that the complexity and dynamism of post-Fordist network society—advanced far beyond anything Bagehot or Schmitt confronted, and constitutively world-societal rather than national—erects a complexity barrier that no all-purpose debating chamber can surmount, and that points unambiguously toward decentralisation via markets and discourses in the plural.

2. What Parliament Was For: Discussion as a Principle, Not an Expedient

Schmitt’s decisive methodological move—the one that makes his short book far more penetrating than the ordinary catalogue of parliamentary failings—is to insist that an institution must be understood through its principle, the specific conviction that animates it, and not merely through the practical services it happens to render. He is scornful of those who defend parliament on the ground that it is the least bad arrangement available. Such defences, he says, may be entirely correct as observations, but they miss the point:

“Parliamentarism exists today as a method of government and a political system. Just as everything else that exists and functions tolerably, it is useful—no more and no less. … But they do not carry weight in an argument about principles. Certainly, no one would be so undemanding that he regarded an intellectual foundation or a moral truth as proven by the question, What else?” (Schmitt 1988, p. 3)

To recover the principle, Schmitt sends us back to the canonical theorists—and here he names the very authors whose defence of parliamentary deliberation forms the backdrop to the present argument:

“Like every great institution, parliament presupposes certain characteristic ideas. Whoever wants to find out what these are will be forced to return to Burke, Bentham, Guizot, and John Stuart Mill. He will then be forced to admit that after them, since about 1848, there have certainly been many new practical considerations but no new principled arguments.” (Schmitt 1988, p. 2)

What, then, is the principle? Schmitt locates it with great precision in two linked ideas: openness and discussion. And he is careful—this is crucial for everything that follows—to give “discussion” a strict and demanding sense. It does not mean negotiation, bargaining, or the brokering of interests. It means the disinterested contest of argument oriented toward truth:

“Discussion means an exchange of opinion that is governed by the purpose of persuading one’s opponent through argument of the truth or justice of something or allowing oneself to be persuaded of something as true and just. … To discussion belong shared convictions as premises, the willingness to be persuaded independent of party ties, freedom from selfish interests.” (Schmitt 1988, p.5)

Schmitt’s reference to shared convictions as necessary premisses for genuine discussion is dep and significant. Schmitt sharpens the distinction between discussion and negotiation with an example that anticipates, and rebuts, the modern tendency to reduce parliament to interest-brokerage. When two businessmen settle a deal after a rivalry, each has an eye on his own profit; they may reach a sensible compromise, but this is not discussion in the demanding sense. He quotes Friedrich Gentz—“in this matter still instructed by the liberal Burke”—to fix the contrast precisely: the characteristic of all representative constitutions, as against the old corporative representation of estates, is “that laws arise out of a conflict of opinions (not out of a struggle of interests)” (Schmitt 1988, p.5).

This is the same Burke we already know from the Speech to the Electors of Bristol: Parliament is “a deliberative assembly of one nation, with one interest, that of the whole,” and not “a congress of ambassadors from different and hostile interests.” It is the same Guizot for whom representation was “not an arithmetic machine intended to collect and enumerate individual wills” but “a natural process for extracting from the bosom of society the public reason that alone has the right to govern.” And it is the same J. S. Mill for whom the proper office of a representative assembly was to be “the nation’s Committee of Grievances and its Congress of Opinions,” an arena in which every opinion “can produce itself in full light and challenge discussion.” Schmitt shows that these are not four separate observations but equivalent expressions of a single coherent world view, the classical liberal world view, and that the constitutional architecture of the classical parliament is its logical deployment.

The constitutional rules as deployments of the principle

Once the principle is seen, the otherwise miscellaneous rules of the historic parliament fall into intelligible order. Schmitt insists that they are not arbitrary conventions but necessary corollaries of the belief in discussion—and that they become unintelligible the moment that belief lapses:

“All specifically parliamentary arrangements and norms receive their meaning first through discussion and openness. This is especially true of the fundamental principle that is still recognized constitutionally, although practically hardly still believed in today, that the representative is independent of his constituents and party; it applies to the provisions concerning freedom of speech and immunity of representatives, the openness of parliamentary proceedings, and so forth. These arrangements would be unintelligible if the principle of public discussion were no longer believed in.” (Schmitt 1988, p.3)

Consider each in turn, as Schmitt invites us to. The independence of the representative—the rule, still inscribed in our constitutional self-understanding and which the Weimar constitution captured in its Article 21 (“the members are representatives of the whole people, they are only responsible to their own consciences and not bound to any instructions”)—makes sense only if the member is meant to deliberate, to be persuaded by the better argument, to change his mind in the chamber. A delegate bound by mandate cannot deliberate; he can only transmit. Burke’s refusal to be the mere mouthpiece of Bristol is not an aristocratic affectation but the precondition of genuine discussion.

Publicity is the second pillar. Schmitt traces it to the long struggle against the arcana rei publicae—the doctrine of state secrets that runs from Machiavelli through the theorists of Staatsräson. Against the mistrusted and despised secret cabinet politics, the Enlightenment elevated openness into an absolute value: “The light of the public is the light of the Enlightenment.” Bentham is, for Schmitt, “the fanatic of liberal rationality” on this point, for whom freedom of public discussion, and especially of the press, is “the most effective protection against political abuses” and the true “check to arbitrary power.” The publicity of proceedings, freedom of speech in the chamber, parliamentary immunity, the reporting of debates—all are instruments of this single demand that power justify itself in the open.

Freedom of the press, Schmitt notes, is in Guizot’s scheme “only a means for discussion and openness and not an independent factor”—the mechanism by which “press freedom prompts citizens to seek the truth for themselves and to make it known to ‘the powers.’” It is the connective tissue between the chamber and the wider public sphere. And this is the point I want especially to register, because it bears directly on the historical reality the Starkey Thesis mourns: the classical parliament was never conceived as a self-sufficient mechanism. It was the apex of a public sphere—an educated, reading, well-informed, public of accomplished stakeholders—without which its deliberation would be a closed and meaningless ritual. Schmitt’s own later development of this point (and Habermas’s celebrated elaboration of it) confirms that the deliberating chamber and the deliberating public were two organs of one body.

The separation of powers as the same idea in organisational form

Schmitt’s most original contribution in the second chapter is to show that the separation of powers is not a separate constitutional principle bolted on alongside discussion, but embodies the very same liberal belief system expressed in organisational form. The belief that truth emerges from “an unrestrained clash of opinion,” that “competition will produce harmony,” is, he argues, a comprehensive metaphysical system of which both free-market economics and the balance of powers are applications:

“It is essential that liberalism be understood as a consistent, comprehensive metaphysical system. Normally one only discusses the economic line of reasoning that social harmony and the maximization of wealth follow from the free economic competition of individuals … But all this is only an application of a general liberal principle. It is exactly the same: That the truth can be found through an unrestrained clash of opinion and that competition will produce harmony.” (Schmitt 1988, p. 35)

This is a passage of singular importance for a libertarian reader, and I shall return to it, because it reveals that the deepest layer of the classical constitution and the deepest layer of the market order are the same layer—a faith in spontaneous, competitive discovery of truth and value against the pretension of any single will to command the outcome. The point is structural. The division of powers, the balance between legislature and executive, the very identification—enshrined in Article 16 of the 1789 Declaration—that “any society in which the separation of powers and rights is not guaranteed has no constitution,” all express a way of thinking that “creates multiplicity everywhere” so that an equilibrium replaces absolute unity. Parliament is not merely one power balanced against the others; within itself, through government and opposition, it reproduces the same balancing dynamic. “An opposition belongs to the essence of parliament,” Schmitt writes, “and there is actually a metaphysic of the two-party system.”

Crucially, the older liberalism understood the limits of this rationalism. Schmitt observes that the balance theory was deliberately confined to the legislative—to the making of general rules—and was never meant to dissolve the executive into endless discussion. He cites The Federalist: legislation is deliberation and “must be made by a larger assembly, while decision making and protection of state secrets belong to the executive,” where unity and energy are required and where “different opinions” are precisely not useful (Schmitt 1988, p.45). The classical constitution thus rested on a sober division of labour between deliberation and decision, deliberare and agere. The collapse of this division—the absorption of executive decision into the parliamentary process via “parliamentary government” itself—is, for Schmitt, one of the chief solvents of the system. We shall see that this is exactly the mechanism Bagehot celebrated and that Starkey, rightly, identifies as the fault-line Blair exploited.

3. Bagehot’s Constitution: The Efficient Secret and the Deference of the Mass

If Schmitt gives us the principle of parliament, Bagehot gives us the working anatomy of the specific constitution—the English one—whose supposed destruction the Starkey Thesis laments. Writing in 1867, on the eve of the Second Reform Act, Bagehot famously distinguished the parts of the constitution that excite reverence from the parts that actually work:

“In such constitutions there are two parts … first, those which excite and preserve the reverence of the population—the dignified parts, if I may so call them; and next, the efficient parts—those by which it, in fact, works and rules. … every constitution must first gain authority, and then use authority; it must first win the loyalty and confidence of mankind, and then employ that homage in the work of government.” (Bagehot 2001, p.5)

The Crown and the Lords are dignified; they raise the army though they do not win the battle. The “efficient secret” lies elsewhere, in “the close union, the nearly complete fusion of the executive and legislative powers,” effected by the Cabinet—“a hyphen which joins, a buckle which fastens, the legislative part of the state to the executive part” (Bagehot 2001, pp. 9–10). Note at once the irony that frames this entire essay: the genius of the English constitution, for Bagehot, is precisely the fusion of the powers that Schmitt’s classical liberalism wished to keep separate. Bagehot openly mocks the “literary theory” of checks and balances as a paper description that the living constitution had already discarded. Where the American presidential system maintains the separation—and pays for it, in Bagehot’s telling, with weakness, deadlock, and “government by an unknown quantity” (Bagehot 2001, p.21)—the English system fused the two and gained energy, flexibility, and above all a “reserve of power” for emergencies (Bagehot 2001, p.19).

Here is the first lesson for the Starkey Thesis. The British constitution had, by Bagehot’s own account and approval, already travelled a long way from the pure Schmittian model of separated, mutually balancing powers. The fusion of executive and legislature in the Cabinet meant that, in normal times, the “sovereignty of Parliament” was in considerable measure the sovereignty of the executive operating through a disciplined parliamentary majority. The chamber that Schmitt idealised as a place of genuine cross-party persuasion was, in Bagehot’s realistic description, already an instrument for sustaining and dismissing governments. What Blair did, on this view, was not to invent executive dominance but to extend and entrench it—while simultaneously exporting great tranches of decision out of the political system altogether. The fusion Bagehot praised became the platform from which the further revolution was launched.

The unequal development of mankind and the dignified parts

Bagehot is most valuable—and most bracing—when he turns to the population. He is entirely candid that the dignified parts of the constitution exist because the mass of men are neither moved by, nor capable of judging, “the plain, palpable ends of government.” His premise is what he calls “the most strange fact, though the most certain in nature … the unequal development of the human race” (Bagehot 2001, p.6). He demolishes the egalitarian hope that all might soon be brought to one level:

“We have in a great community like England crowds of people scarcely more civilised than the majority of two thousand years ago. … The lower orders, the middle orders, are still, when tried by what is the standard of the educated ‘ten thousand,’ narrow-minded, unintelligent, incurious.” (Bagehot 2001, p.6)

From this premise Bagehot derives the political function of pomp, monarchy, and theatre. The “theatrical elements” of the constitution exist because “that which is mystic in its claims … brilliant to the eye … is the sort of thing—the only sort which yet comes home to the mass of men.” The dignified parts are “likely to be adjusted to the lowest orders—those likely to care least and judge worst about what is useful” (Bagehot 2001, p.7). The efficient business of government, meanwhile, proceeds elsewhere, conducted by the educated few. This is not a marginal aside in Bagehot; it is the core of his whole theory. The constitution governs a population most of whose members cannot follow the reasoning of government, and it does so by winning their homage through display while reserving the substance of decision to a competent minority.

It is impossible to read these passages in 2026 without registering how unrepeatable they have become. The frank acknowledgement of an “educated ten thousand” ruling, with the consent but not the participation of the mass, is exactly the candour that universal suffrage and the egalitarian sensibility have made unsayable. And yet—this is the point—the structure Bagehot describes did not vanish when the candour did. The mass still does not, in Bagehot’s sense, conduct the efficient business of the state. What changed is that the mechanisms by which its consent is won and its will is formed, managed, and hedged became less honest, more mediated, and—as Schmitt would insist—industrialised within the modern apparatus of party propaganda and mass media, to manufacture a manageable public opinion.

Cabinet government as an educative engine—and what its decay implies

Bagehot offers one further argument that bears directly on the Starkey-Schmitt convergence. He holds that cabinet government, unlike the presidential system, educates the nation because the great public contest over the fate of a government forces a genuine debate that the people attend to:

“The cabinet system ensures … debates … The nation is forced to hear two sides—all the sides, perhaps, of that which most concerns it. … It succeeds in deciding because the debates and the discussions give it the facts and the arguments.” (Bagehot 2001, p.14)

This is Bagehot’s version of Schmitt’s discussion-principle—the chamber as the “great scene of debate, the great engine of popular instruction.” But notice the fragility he himself builds in. The educative power of debate depends on the debate mattering—on the prize of power genuinely turning on it. Where “debates … are prologues without a play,” where “the prize of power is not in the gift of the legislature,” the teaching apparatus fails and “no one cares for the legislature” (Bagehot 2001, pp.15–16). Bagehot was describing the American Congress. But the description fits, with uncanny precision, the Parliament that the Starkey Thesis describes after Blair: a chamber from which the real prizes—monetary policy, vast swathes of regulation, large areas of rights adjudication—have been removed to bodies it cannot control, so that its debates increasingly become prologues without a play. The decay Bagehot warned of in the presidential system has been reproduced in Britain, not by adopting a presidential separation of powers, but by dispersing power to the unelected.

4. The Contradiction: Why Parliamentarism Could Not Survive Mass Democracy

We arrive now at the theoretical heart of Schmitt’s book and of this essay. Schmitt’s thesis is that parliamentarism and democracy are not the same thing, that their nineteenth-century alliance was historically contingent, and that once that alliance dissolved—once democracy achieved its “victory”—the two principles were revealed to be in fundamental contradiction. 18^th/19^th Century liberal parliamentarism and 20^th Century mass democracy are built on fundamentally different belief systems. Schmitt states it with great clarity:

“The belief in parliamentarism, in government by discussion, belongs to the intellectual world of liberalism. It does not belong to democracy. Both, liberalism and democracy, have to be distinguished from one another so that the patchwork picture that makes up modern mass democracy can be recognized.” (Schmitt 1988, p.8)

The distinction is a matter of principle, not merely procedural. Parliamentarism rests on the principle of discussion—truth emerges from the open clash of argument among independent representatives. Its legitimating ground is the rationality of the process. Democracy, by contrast, rests on the principle of identity—the identity of rulers and ruled. Its legitimating ground is the will of the people, however that will is expressed and whether or not it was arrived at by deliberation at all. Schmitt drives the point to its sharpest edge:

“As democracy, modern mass democracy attempts to realize an identity of governed and governing, and thus it confronts parliament as an inconceivable and outmoded institution. If democratic identity is taken seriously, then in an emergency, no other constitutional institution can withstand the sole criterion of the people’s will, however it is expressed. Against the will of the people especially an institution based on discussion by independent representatives has no autonomous justification for its existence, even less so because the belief in discussion is not democratic but originally liberal.” (Schmitt 1988, p.15)

This is an analysis I endorse, and it is the key to the Starkey Thesis. Every constitutional device that the classical parliament prized—the independence of the member, the publicity of true deliberation, the separation of powers, the slow filtering of decision through argument—appears, from the standpoint of consistent democracy, as an aristocratic obstruction interposed between the people’s will and its realisation. Why should an independent representative substitute his judgement for the expressed will of those who elected him? Why should a balance of powers be permitted to frustrate a majority? The classical answers to these questions were liberal answers, grounded in the discussion-principle. Mass democracy has no use for them.

How the contradiction manifests in the four classical rules

Schmitt shows that modern mass democracy did not so much abolish the classical parliamentary rules as hollow them out, leaving the form standing while the substance fled. His metaphor is poignant:

“Many norms of contemporary parliamentary law, above all provisions concerning the independence of representatives and the openness of sessions, function as a result like a superfluous decoration, useless and even embarrassing, as though someone had painted the radiator of a modern central heating system with red flames in order to give the appearance of a blazing fire.” (Schmitt 1988, p.6)

Take the four classical features in turn. Public debate, first. Under mass democracy, Schmitt observes, the parties “do not face each other today discussing opinions, but as social or economic power-groups calculating their mutual interests and opportunities for power, and they actually agree compromises and coalitions on this basis.” The masses, meanwhile, “are won over through a propaganda apparatus whose maximum effect relies on an appeal to immediate interests and passions” (Schmitt 1923/1988, p.6). Argument in the genuine sense “ceases.” The floor debate survives as theatre while the real decisions migrate to the back room. Schmitt’s mordant summary deserves quoting in full, because it could be a description of any modern legislature, Westminster included:

“Today parliament itself appears a gigantic antechamber in front of the bureaus or committees of invisible rulers. … the real business takes place, not in the open sessions of a plenum, but in committees and not even necessarily in parliamentary committees, and that important decisions are taken in secret meetings of faction leaders or even in extraparliamentary committees so that responsibility is transferred and even abolished, and in this way the whole parliamentary system finally becomes only a poor facade concealing the dominance of parties and economic interests.” (Schmitt 1988, pp.7, 20)

The independence of representatives, second. The maxim that the member represents the whole people and is bound by no instruction is, Schmitt says, “still recognized constitutionally, although practically hardly still believed in today.” Mass democracy, organised through disciplined parties, proportional lists, and the whip, converts the independent deliberator into a reliable party foot soldier. The representative is now neither a Burkean trustee following his conscience nor a pure democratic delegate of his constituents; he is an agent of the party machine. Both principles—the liberal and the democratic—are simultaneously betrayed.

The secret ballot, third, withdraws the citizen’s decisive act into the sphere of the private and irresponsible. Here the tension becomes stark: parliamentarism requires the publicity of the representative’s reasoning, while mass democracy’s characteristic instrument, the secret ballot, protects the privacy of the voter’s will. The reasoning that ought to be exposed is hidden in the back room; the willing that constitutes “the people” is atomised into a hundred million isolated and unaccountable acts and then reassembled by “the statistical apparatus” into a number. Neither pole any longer answers to the discussion-principle. I shall draw on this below, because it bears on how a population’s political will is in fact formed and managed—a question Bagehot faced with more honesty than we permit ourselves.

The separation of powers, fourth and finally. We have already seen that for Schmitt the separation is a liberal, not a democratic, device—“the antithesis of a democratic concept of identity.” Consistent democracy has no reason to tolerate a balance of powers obstructing the people’s will; the Jacobin tradition rejected it outright, and Rousseau’s general will “cannot be divided.” Bagehot, from the opposite direction, showed that the English constitution had already fused the powers in the Cabinet for reasons of efficiency. From both sides, then—the democratic and the efficient-executive—the classical liberal separation was under pressure. The Blairite settlement, as we shall see, completed a paradoxical movement: it strengthened the executive’s fusion with the legislature even as it dispersed vast areas of decision to bodies outside the elected system entirely.

5. Democracy, Homogeneity, and the Limits of Universal Suffrage

Schmitt’s analysis of democracy proper—as distinct from parliamentarism—supplies the second pillar of my argument, and it is the one most likely to provoke. His claim is that democracy is not, and cannot coherently be, the rule of an undifferentiated humanity. Every actual democracy rests on a substance of equality, which means it rests on a corresponding inequality—a boundary between those who belong and those who do not:

“Every actual democracy rests on the principle that not only are equals equal but unequals will not be treated equally. Democracy requires, therefore, first homogeneity and second—if the need arises—elimination or eradication of heterogeneity. … A democracy demonstrates its political power by knowing how to refuse or keep at bay something foreign and unequal that threatens its homogeneity.” (Schmitt 1988, p.9)

Schmitt’s point is here analytical, not a programme: the equality that makes democratic self-rule meaningful must have substance—civic virtue in the classical vertu, or shared religious conviction among the English sects, or, since the nineteenth century, national homogeneity. “Equality is only interesting and valuable politically so long as it has substance, and for that reason at least the possibility and the risk of inequality.” From this follows his striking thesis that universal, equal suffrage of all human beings as such is not a democratic but a liberal idea, and an incoherent one:

“Universal and equal suffrage is only, quite reasonably, the consequence of a substantial equality within the circle of equals and does not exceed this equality. Equal rights make good sense where homogeneity exists. But the “current usage” of “universal suffrage” implies something else: Every adult person, simply as a person, should eo ipso be politically equal to every other person. … It replaces formerly existing democracies, based on a substantial equality and homogeneity, with a democracy of mankind.” (Schmitt 1988, pp.10–11)

Schmitt then states and laments the consequence that I - as a libertarian who wishes to shrink the political - in contrast to him embrace. When political equality is stretched to an abstract, substanceless universal equality, the political sphere does not thereby become more genuinely democratic; it becomes hollow, and substantive decision migrates elsewhere:

“Under conditions of superficial political equality, another sphere in which substantial inequalities prevail (today, for example, the economic sphere) will dominate politics. This is completely unavoidable and any reflection on political theory recognizes it as the real grounds for the much-deplored dominance of economics over state and politics. Wherever an indifferent concept of equality, without the necessary correlate of inequality, actually takes hold of an area of human life, then this area loses its substance and is overshadowed by another sphere in which inequality then comes into play with ruthless power.” (Schmitt 1988, p.13)

Schmitt intends this as a lament, or at least a warning, about the “dominance of economics.” But read with libertarian eyes—and detached from Schmitt’s own appraisal—it points in a wholly different direction. If a hollow, substanceless political equality inevitably cedes the substance of human affairs to the spheres where genuine differentiation operates—the economy, science, the professions, voluntary association—then the right response is not to mourn the loss of political substance but to welcome and complete it. The migration of decision out of the hollow political sphere and into the functionally differentiated subsystems of a free society is not a pathology to be reversed but the very movement a libertarian seeks to accelerate. Where Schmitt sees a homeless political will that ought to be re-grounded in a homogeneous demos capable of decisive collective action, I see a political sphere that is rightly shrinking because the matters it once arrogated to itself are better handled by markets and discourses. I shall make this argument fully in Section 7.

The crisis, Schmitt concludes, is in fact two crises that aggravate each other: a crisis of parliamentarism (the discussion-principle has become an empty formality) and a crisis of democracy (the substance of homogeneity cannot be supplied by an abstract equality of mankind). “It is, in its depths, the inescapable contradiction of liberal individualism and democratic homogeneity.” My contention is that the United Kingdom, in the Blairite settlement and after, has been living through exactly this double crisis—and that the Starkey Thesis is its constitutional post mortem.

6. Habermas and the Reciprocity of Liberty and Popular Sovereignty

Schmitt presents the contradiction between liberalism and democracy as, in the last analysis, irreconcilable—“the inescapable contradiction of liberal individualism and democratic homogeneity.” It is worth pausing on the most influential twentieth-century attempt to dissolve that very contradiction, because it both clarifies what is at stake and, in its limits, points toward the position I shall ultimately defend. Jürgen Habermas, in Between Facts and Norms (1992), confronts precisely the tension Schmitt identified, though he names its poles differently: not parliamentarism versus democracy, but liberty (the system of individual rights, “private autonomy”) versus popular sovereignty (the self-legislation of the people, “public autonomy”). I rehearse his argument here rather than evaluate it critically, because its very structure illuminates the constitutional disorder the Starkey Thesis describes.

The two ideas, Habermas observes, “have determined the normative self-understanding of constitutional democracies up to the present day,” yet they have never been satisfactorily reconciled. The liberal tradition, descending from natural-rights enlightenment, treats individual liberties as primary—pre-political, inalienable, a constraint set around the sovereign legislator. The civic-republican tradition, emphasizing popular sovereignty, of which Rousseau is the great exponent, treats the self-organisation of the political community as primary and recognises no prior limit on the people’s sovereign will. Habermas states the conflict sharply:

“Liberals invoke the danger of a “tyranny of the majority” and postulate the priority of human rights that guarantee the pre-political liberties of the individual and set limits on the sovereign will of the political legislator. The proponents of a civic republicanism, on the contrary, emphasize the intrinsic, non-instrumentalizable value of civic self-organization, so that human rights have a binding character for a political community only as elements of their own consciously appropriated tradition.” (Habermas 1996, p.100)

This is recognisably Schmitt’s antinomy in a post-metaphysical register. The liberal pole corresponds to Schmitt’s parliamentarism—the rights, the independence of the representative, the separation of powers that constrain the popular will. The republican pole corresponds to Schmitt’s democracy—the identity of rulers and ruled, the sovereign will that brooks no external constraint. And Rousseau, whom Schmitt invoked to expose democracy’s drive toward homogeneity, is for Habermas too the purest and most problematic statement of the popular sovereignty pole: “There neither is nor can be any kind of fundamental law binding on the body of the people—not even the social contract itself.”

The resolution: co-originality

Habermas’s ambition is to show that the two principles do not merely coexist in tension but presuppose and require one another—that they are, in his term, co-original. His route to this conclusion runs through a prior and, for my larger argument, decisive insight: the critique of the atomistic individual that underwrites the liberal social-contract story. Habermas’s objection is not merely that the contract is a historical fiction; it is that the picture of the human subject on which it rests—the cognitively self-sufficient agent with autonomously formed ends—is false. Subjectivity, self-conception, and one’s very understanding of one’s own interests are not given prior to society; they are formed and continuously reproduced in communicative interaction.

What we recognise as human subjects, Habermas argues, are always already inherently social, and our self-conceptions—including everybody’s conceptions of their own interests and self-realisation—depend not only on inherited language and initial socialisation but on continued communicative interactions that are crucially required to continuously confirm, reproduce and update our very self-conceptions and orienting understanding of the world. From this it follows that individual instrumental reason is secondary to and dependent upon communicative reason—a thesis I take up in the next chapter, because it reaches far beyond the political and supplies, I shall argue, the deepest justification for the markets-and-discourses order. For the present purpose its constitutional consequence is what matters: if interests themselves are constituted in communication, the system of rights cannot be grounded on prior individual interests in the manner the contractarian liberal supposes. The polity is a constitutive condition of the very individuals, interests, and rights it is said merely to protect.

How, then, are individual liberties to be grounded at all, if not as pre-political natural rights? Habermas’s answer is the hinge of the whole reconstruction. The system of rights is justified as the necessary enabling condition of the very formation of the people’s will. A people that legislates for itself rationally must do so through a process of opinion- and will-formation that is free, inclusive, and repression-free—through discourse in which, as he puts it, “the only thing that counts is the compelling force of the better argument based on the relevant information.” But such a discursive process is impossible unless the participants are already constituted as free and equal legal persons, with secured freedoms of speech, assembly, conscience, association, and political participation. The liberties are not an external fence around democracy; they are the internal precondition of democracy’s own rational exercise. He states the conclusion with precision:

“The sought-for internal connection between popular sovereignty and human rights lies in the normative content of the very mode of exercising political autonomy, a mode that is not secured simply through the grammatical form of general laws but only through the communicative form of discursive processes of opinion- and will-formation. … Consequently, the sought-for internal relation between popular sovereignty and human rights consists in the fact that the system of rights states precisely the conditions under which the forms of communication necessary for the genesis of legitimate law can be legally institutionalized.” (Habermas 1996, pp.104–128)

Thus, public autonomy requires private autonomy, and private autonomy requires public autonomy; the addressees of the law must be able to understand themselves as its authors. Habermas calls this the “co-originality of private and public autonomy,” and with it he claims to have achieved a post-metaphysical reconstruction of the dual foundation of modern constitutional democracy. The system of rights is “not given to the framers of a constitution in advance as a natural law”; it is the explication of what a people necessarily presupposes the moment it resolves to regulate its common life through legitimate positive law.

Why this matters for the argument—and where I depart

Habermas’s achievement, taken on its own terms, is to show that Schmitt was right about the tension but wrong to think it terminal. Liberty and popular sovereignty need not be set against one another in a fight to the death; they can be understood as two readings of a single underlying discourse-principle, the principle that legitimate norms are those to which all possibly affected could agree under conditions of free and equal argument. This is, I want to stress, a deepening of exactly the discussion-principle that Schmitt located at the origin of parliamentarism and that I endorsed in Chapter2. Where the classical liberals—Burke, Bentham, Guizot, Mill—grounded parliament in the conviction that truth emerges from the open contest of argument, Habermas generalises that conviction into a comprehensive theory of legitimacy: the validity of law itself rests on the communicative form of its genesis.

And yet—here is the pivot toward my own position, which I register now and develop in the chapters that follow—Habermas, having shown that the discourse-principle is the true ground of legitimacy, retains the assumption that this discourse must terminate in universally binding collective decisions issued by a sovereign people through the medium of centralised political power. He has emancipated legitimacy from natural law and re-grounded it in communication, but he has not asked whether the discursive opinion- and will-formation he rightly prizes requires the apparatus of sovereign, one-size-fits-all command at all. It is possible, as I have argued elsewhere, to imagine a society that operates “without the medium of a centralized political power, not without legal systems but without universally binding collective decisions, where free contracting assumes a much larger role and where multiple legal systems interpenetrate,” and in which “political discourse does not disappear but indeed remains a crucial steering arena for societal developments, without presuming that the purpose of all discourse is to result in one-fits-all majority decisions.”

In other words: I accept Habermas’s demonstration that the discourse-principle is the genuine ground of legitimacy and that liberty and the public formation of will reciprocally require each other. What I question is the inherited reflex—shared by Rousseau, by Schmitt in his diagnosis, and by Habermas in his resolution—that the public formation of will must be cashed out as the sovereign command of a demos over the whole of a society bounded by a nation-state. The reciprocity Habermas establishes is between liberty and public discourse; it is a further, unexamined step to identify public discourse with majoritarian sovereign decision. Sever that step, and the discourse-principle is liberated from the apparatus of command—which is precisely the move that the markets-and-discourses order makes, and toward which the remaining chapters build.

7. Communicative Action and the Achievements of Modern Civilisation

The discussion-principle that animated the classical parliament is, I now want to argue, not a peculiarity of the political system at all. It is a particular, highly reflective instance of something far more general and far more important: the dependence of all modern achievement—scientific, technological, economic, legal, moral, political —on a mode of human interaction (action type) that classical liberal and libertarian social theory has almost entirely overlooked. Here I draw on Habermas’s distinction between communicative and instrumental (or strategic) action, because it supplies the missing foundation beneath everything said so far, and because it shows why the recession of political command need not—must not—mean the recession of public reasoning.

Classical liberal and libertarian social philosophy recognises essentially one type of action: instrumental action—the individual’s pursuit of his own ends by suitable means, including strategic action, which is instrumental action directed at other people whereby other people become the means to actor’s ends. Ludwig von Mises, in his 1949 ‘Human Action’, speaks for the whole tradition: “A man’s ends are always his own. They are guided by his own valuations, his own will, and his own interest.” This is an impoverished model of human action. Habermas’ sociology adds the decisive further type: communicative action—interaction oriented not toward strategic success but toward mutual understanding, dialogical, aimed at reaching consensus through argument, presupposing a shared life and often having no explicit individual goal at all. In communicative action, unlike strategic action, “all motives except that of finding a mutual and consensual understanding are suspended,” and participants treat one another never as mere means but always also as ends in their own right.

The point that matters most for the present argument—and the one I want to place at the centre of this chapter—is Habermas’s thesis that instrumental reason is parasitic on communicative reason. This is not a pious sentiment about the value of dialogue; it is a hard claim about the conditions of possibility of instrumental action itself. Instrumental and strategic calculation requires knowledge—knowledge of available means, of the relevant facts, of the constraints in play, and indeed of one’s own ends. And all such knowledge is itself the product of a prior communicative achievement. As Habermas puts it in his ‘Theory of Communicative Action’, “the use of language with an orientation to reaching understanding is the original mode of language use, upon which … the instrumental use of language in general is parasitic” (Habermas 1981).

Consider what this means for the very negotiation and bargaining that the classical liberal takes as the paradigm of social interaction. Even the interests and the constraining facts presupposed in any instrumental or strategic transaction must first be formed, ascertained, and confirmed in communicative action. The businessman who bargains has an eye on his profit—but his very conception of where his profit lies, his grasp of the relevant market facts, the technological possibilities, the legal constraints, the reliability of his counterpart, the meaning of the contract he signs: all of this is the sediment of countless prior communicative interactions in which the truth was sought cooperatively rather than strategically. Strip away the communicative substrate and the bargainer would not know what to bargain for. “Means-end calculations,” as I have already argued elsewhere, in my Markets and Discourses’, “must be reassured and empowered by the knowledge of possible means that could only be gained via prior communicative interaction.” The market actor cannot remain in strategic isolation for long “without going astray”; he needs “continuous guiding correctives that flow from sincere and sincerely probing and confirming communications in order to remain a competent rational agent” (Schumacher 2025).

This is the deepest answer to Schmitt’s own most striking observation—that the discussion-principle of parliament and the competition-principle of the market are “exactly the same” liberal metaphysic, i.e. part of the same system of beliefs. They are indeed two expressions of one trust in spontaneous, decentralised discovery; but communicative reason is the more fundamental of the two, because the market’s discoveries cannot even be recognised as discoveries except through discourse. Hayek rightly called the market a discovery procedure, and the libertarian rightly expects a free market to unleash an unprecedented volume of experimentation. But “it is never trivial to attribute the success of a complex venture to specific features or aspects of the effort” (Schumacher 2025). To learn from market experimentation—to convert its results into transmissible knowledge—requires the complementary, non-strategic domain of scientific and scholarly discourse. Trial-and-error in markets alone is insufficient; “the competitive search for truth, embedded in open-access discursive practices, becomes indispensable” Schumacher 2025).

The achievements of modern civilisation—and this is the burden of the chapter—rest at their highest level on the most reflective and demanding form of communicative action: critical-reflective public discourse. Discourse in the strict sense arises when the smooth routine of taken-for-granted understandings “fails, fractures or breaks down,” and the implicit validity claims of ordinary speech—truth, truthfulness, rightness—must be made explicit, questioned, and redeemed through argument. This is the mode of science, of jurisprudence, of moral and political reflection. It is governed by a strict ethic: it aims at consensus and truth, not profit; it proceeds by evidence and argument, not preference or power; it requires sincerity, openness to criticism, and repression-free inclusiveness. “Any knowledge claim becomes suspect or even void, if the originating social/communicative process significantly deviates from the regulative ideal of an open, non-strategic, sincere, critical, repression-free discourse” (Schumacher 2025). Science’s very criterion of universal reproducibility is a demand for open critical access—for the exposure of every claim to everyone’s independent testing and possible refutation.

Two consequences follow that are essential to the argument of this essay. First, critical-reflective discourse is not confined to the political system. It is the constitutive medium of the economy’s rationality, of scientific and technological progress, of the evolution of law, of moral self-correction—of every functionally differentiated subsystem of modern society. “In our sophisticated knowledge economy, nearly all economic work and consumption choices depend on discourse. Medical therapy choices, legal frameworks, engineering, and finance all rest on discursive validation” (Schumacher 2025). The reflex that identifies “public reasoning” with “parliamentary debate” mistakes one historical vessel for the thing itself. The discussion-principle that the classical parliament embodied lives on, vastly amplified, throughout the varied discourses of modern civil society, including discourses of general public concern we might still call political — and it lives on quite independently of whether any sovereign chamber deliberates at all.

Because modern knowledge, science, and rationality depend on the sincere, open, repression-free discourse of free and equal participants, they depend on a liberal morality and liberal social relations—and they are corrupted alike by relations of bondage and by “the ideological domination or censoring of individual dissenters by a politically empowered majority opinion” (Schumacher 2018). This is a result of the first importance for the relation between discourse and democracy. The very thing Habermas prizes—the discursive formation of opinion and will—is endangered rather than secured the moment it is fused with majoritarian sovereign command, because the power to issue binding collective decisions is also the power to suppress the dissenting voice whose contribution “might have come closest to the truth”—the danger that, as Schmitt noted, set even the positivist Mill “in an inexplicable uproar.” Genuine discourse needs liberty from command, not subordination to it. This is why the markets-and-discourses order does not abolish public reasoning but liberates it: it retains discourse—indeed expects it to intensify—while withdrawing from it the coercive power to terminate inquiry by collectively binding decree.

8. The Complexity Barrier: Post-Fordist Society and the Limits of the All-Purpose Chamber

I come now to what I regard as the most consequential consideration of all, and the one that most decisively settles the question of where the powers drained from Parliament ought to go. It is a consideration that neither Bagehot nor Schmitt could fully have grasped, because the phenomenon was not yet before them: the sheer complexity and dynamism of contemporary post-Fordist network society, and the barrier this complexity erects against any attempt to address the affairs of such a society through an all-purpose political debating chamber—however deliberative, however well-constituted, however faithful to the discussion-principle.

Bagehot wrote in 1867, at the apex of mechanical industrial capitalism, of a national economy and a national polity whose principal business could still, in large part, be surveyed and debated by an “educated ten thousand.” Schmitt wrote in 1923, amid the consolidation of Fordist mass production and mass democracy, and already discerned that “great political and economic decisions on which the fate of mankind rests” could no longer “result … from balancing opinions in public debate.” But the complexity that defeated the deliberative chamber even in Schmitt’s day was a fraction of what now obtains. Since the 1970s the convergence of computation and telecommunication has driven a socio-economic restructuring—the transition from Fordism to post-Fordism—whose character is defined precisely by an explosion of differentiation, dynamism, and innovation-density that no prior epoch approached.

The contrast is structural, not merely quantitative. The technologies of Fordism were rigid without agile adaptivity and without the ability to quickly absorb and utilize innovations. In contrast, the new numerically controlled systems like robotic fabrication, 3D printing as well as the massive new field of software-as-a-service … are capable of absorbing unlimited numbers of innovations. This in turn dynamizes business and business organisation and pulls workers away from assembly lines into R&D, marketing, finance etc. All work becomes innovative and project-like rather than routine. A society in which all work becomes innovative and project-like is a society whose relevant states of affairs multiply, differentiate, and change faster than any central body could hope to register, let alone deliberate upon and command. The number of distinct technological possibilities, business models, employment relations, financial instruments, contractual forms, and modes of dispute resolution that are simultaneously live, and simultaneously evolving, is now of an order that simply has no analogue in 1867 or 1923.

This is the complexity barrier, and it is the same barrier, generalised and amplified, that Mises and Hayek identified in their critique of socialist planning. Mises showed that without competitive bidding and price formation there can be no economic calculation, and hence no rational allocation; the central authority is “groping in the dark.” Hayek showed that the knowledge required for rational coordination is irreducibly dispersed, local, and often tacit, and cannot be centralised in any planning agency. The Marxist dream of conscious collective control over the social life-process falters as it increasingly meets an insurmountable complexity barrier. My contention is that the all-purpose political chamber—the sovereign deliberative-legislative assembly that the restorationist would re-empower — faces exactly this barrier, and faces it more acutely with every advance of post-Fordist differentiation. To ask a single chamber to deliberate authoritatively over monetary arrangements, medical practice, environmental trade-offs, technological standards, financial instruments, employment relations, and the evolution of law across a hyper-differentiated, fast-mutating society is to ask of public deliberation precisely what Hayek showed central planning could not deliver. It is the knowledge problem and the calculation problem in constitutional form.

Note that this is not an objection to the discussion-principle. It is an objection to centralising the discussion-principle in a single all-purpose forum and arming its conclusions with sovereign command. The deliberation that complexity defeats is the deliberation that must issue in one binding answer for the whole of a society. Discourse itself thrives on complexity — it differentiates into specialised scientific, legal, technological, and commercial discourses, each operating by its own code, each capable of absorbing unlimited innovation. What cannot absorb that complexity is the attempt to funnel all these differentiated discourses through the bottleneck of a single chamber and out the other side as uniform, coercively enforced, one-size-fits-all decisions. The complexity barrier is fatal to the singular, not to the plural. It points unambiguously toward decentralisation—toward markets and discourses, in the plural, each handling the slice of complexity proper to it, rather than a market and a discourse, in the singular, subordinated to a sovereign decision.

The world-societal dimension: complexity beyond the nation-state

There is a second dimension to the complexity barrier that bears directly on the constitutional question, and that neither Bagehot’s national constitution nor Schmitt’s homogeneous demos can accommodate. The reproduction and progress dynamics of contemporary society are not contained within the boundaries of any nation-state. The supply chains, capital flows, scientific and technological discourses, communication networks, and innovation ecologies on which modern prosperity depends are constitutively world-societal. The convergence of computation and telecommunication has given “a cosmopolitan character to production and consumption,” in Marx’s prescient phrase; “the intellectual creations of individual nations become common property.” Modern world society is, as has been argued by the social systems theorist Niklas Luhmann, a single, densely interconnected web of functionally differentiated subsystems—the global economy, world science, transnational law—whose operations cross every political frontier and answer to no national chamber.

This is decisive against both Schmitt and the restorationist, though in different ways. Against Schmitt: his account of democracy as resting on the substantive homogeneity of a bounded people, capable of “refusing or keeping at bay something foreign and unequal,” presupposes that the relevant unit of collective self-determination is the nationally homogeneous state. But the processes that actually determine the reproduction and advancement of modern civilisation overflow that unit entirely. A homogeneous national demos deciding sovereignly within its borders is structurally incapable of governing a world-societal process that does not respect those borders—it can only obstruct its own participation in that process. Against the restorationist: to repatriate decisive power to the Westminster chamber is to lodge authority over world-societal dynamics in a body whose competence stops at the national frontier, and whose deliberations cannot reach the transnational subsystems where the real coordination occurs. The nation-state chamber is not merely too undeliberative for the task, as earlier chapters argued; it is too small and too bounded for it. The European attempt to recuperate state power at the supranational level of the European Union has been a failure and has – with the exception of the EU Internal Market that guarantees the free movement of goods, services, capital, and persons (labour) - hurt most member economies via over-regulation. Free trade does not require a super-state, and in any event, the EU is also too small and too bounded.

Markets and discourses, by contrast, are intrinsically scale-free and borderless. A market coordinates across frontiers without requiring a world-state to convene it; a scientific or legal discourse spans the globe without requiring a world-chamber to authorise it. “The future progress of human civilization within a stateless world society will be ordered via markets and steered via discourses” (Schumacher, 2025) —and the deep reason is that markets and discourses can match the scale, the complexity, and the dynamism of world-societal reproduction in a way that no territorially bounded sovereign chamber, however reconstituted, ever could. The complexity barrier and the world-societal character of modern reproduction together form, in my judgement, the strongest single argument for the position toward which this essay has been building: that the powers drained from Parliament should not be dammed and returned to a chamber that cannot, in principle, handle them, but should flow onward into the plural, decentralised, scale-free institutions of a free civil society.

9. The Starkey Thesis Reconsidered: Hollowing-Out as Predicted Symptom

We can now state the Starkey Thesis in Schmittian terms and see how much sharper it becomes. Starkey’s historical claim is that the Blair governments effected a constitutional revolution: incorporation of supranational human-rights norms into domestic law with judicial supremacy attached; the creation of a separate Supreme Court; the establishment of operational independence for the central bank in monetary policy; devolution to Scotland, Wales, and Northern Ireland; the proliferation of regulators, agencies, and “quangos” exercising vast delegated powers; the entrenchment of an administrative state whose decisions Parliament cannot readily reverse (Starkey 2024a, 2024c, 2025). The sovereignty of the Crown-in-Parliament—Bagehot’s (and Dicey’s) sovereign—was on this account fragmented and parcelled out among bodies that neither deliberate publicly in the classical sense nor answer to an electorate.

Schmitt allows us to see that this was not an aberration but a culmination. Recall his image of parliament as “a gigantic antechamber in front of the bureaus or committees of invisible rulers,” and his observation that “the real business takes place, not in the open sessions of a plenum, but in committees and not even necessarily in parliamentary committees.” Written in 1923, this is an exact anticipation of the quango state. The independent central bank is the purest instance: a matter—monetary policy—of the gravest consequence for “the fate of millions” removed entirely from the chamber and entrusted to an unelected committee precisely on the theory that it is too important to be left to the deliberation of representatives. The regulatory agency is the same structure generalised. The supranational rights-norm administered by a supreme judiciary is the arcana returning in a new robe: decisive determinations made in a forum the public sphere cannot effectively reach or argue with.

Schmitt’s further point—that responsibility is thereby “transferred and even abolished”—is the precise grievance of the Starkey Thesis. The genius of the classical constitution, on Bagehot’s telling, was the locatability of responsibility: a Cabinet that could be turned out, a minister who answered at the dispatch box. The quango state dissolves this. When a regulator fails, when a court reads a new policy into a rights instrument, when the central bank misjudges, there is no accountability except — as in Bagehot’s phrase about Congress — “a vast miscellaneous body difficult to punish” (Bagehot 1867/2001, p.13). The “efficient secret” of fused, accountable, locatable power has been replaced by a diffusion in which, as Schmitt put it, “responsibility is transferred and even abolished” (Schmitt 1923/1988, p. 20).

There is, moreover, a Bagehotian irony that Starkey’s account sometimes underplays and that the present framework brings out. Blair did not weaken the executive vis-à-vis Parliament. On the contrary, the fusion Bagehot praised was intensified—the modern Prime Minister commanding a disciplined majority is, if anything, a more dominant “efficient” actor than Bagehot’s. What Blair did was to use that executive dominance to export whole domains of decision beyond the reach of the electoral system altogether, while leaving the dignified facade—and indeed the dignified facade of Parliament itself—in place. This is why the chamber increasingly resembles Bagehot’s prologue without a play: the executive remains strong, but the matters over which sovereignty is exercised have been narrowed, and the most consequential of them placed in unaccountable hands. The form of parliamentary sovereignty persists; the substance has been redistributed.

So far, I am in full agreement with Starkey: the transformation was radical, it was largely unannounced, and it has produced a constitution in which the locus of real decision is systematically obscured and removed from public deliberation and democratic accountability alike. The frustration is justified. The question is what to do with it.

The arm’s-length state by the numbers

The Starkey Thesis can remain abstract—a claim about “powers” migrating to “quangos”—unless it is given a sense of scale, and the official figures supply one. The Cabinet Office’s Public Bodies 2024 return records 305 arm’s-length bodies in the United Kingdom; together they spend over £220 billion a year and employ more than 300,000 people, and the largest ten alone account for roughly ninety per cent of all arm’s-length-body funding (Cabinet Office 2025; Modern Civil Service 2024). These are not, for the most part, the advisory committees of caricature. They are the working machinery through which a great part of the British state now actually operates—and the concentration of nearly all the money in a handful of giant bodies is itself the shape the argument of this essay would predict.

The official taxonomy distinguishes three kinds of arm’s-length body by their distance from ministerial control, and the gradient is instructive in its own right. Executive agencies remain legally part of a department and answerable to its minister, but are administratively hived off to deliver specified outputs—amongst them HM Prison and Probation Service, the Met Office. Non-departmental public bodies—the largest category—operate at arm’s length from ministers and discharge advisory, regulatory, or executive functions: the Environment Agency, the Health and Safety Executive, the British Council, the Equality and Human Rights Commission, and the vast NHS England among them. Non-ministerial departments are the purest constitutional case: full departments of state that operate without a minister at their head, accountable to Parliament only at one remove through a sponsoring minister—HM Revenue and Customs, Ofgem, the Food Standards Agency. The sequence executive agency → non-departmental public body → non-ministerial department traces a steady widening of the distance between the decision and the ballot box; and beyond it lies a further category the official count does not even capture—the public corporation, the class to which the Bank of England belongs, and with it the single most consequential exercise of delegated power in the land.

Naming the largest and the most interventionist bodies brings the scale home. By sheer magnitude, NHS England, HM Revenue and Customs, and Network Rail are among the giants that dominate the funding totals. But magnitude and constitutional significance are distinct axes, and some of the most consequential bodies are modest in budget: the Office for Budget Responsibility, whose fiscal forecasts now effectively police the room for manoeuvre of any Chancellor; the Climate Change Committee, which superintends the legally binding carbon budgets discussed above; the central bank, whose Monetary Policy Committee sets the price of money for the whole economy; the regulators—Ofgem, Ofcom, Ofwat, the Financial Conduct Authority, the Competition and Markets Authority—each governing the economics of an entire sector; and the rights-and-equality apparatus that administers, beyond the reach of ordinary majorities, large tracts of what may lawfully be said and done. These are the bodies in which Starkey’s dispersed sovereignty actually resides. A handful of them, none headed by an elected representative, between them determine the fiscal envelope, the cost of energy, the rate of interest, and the boundaries of permissible speech—the very parameters within which the elected chamber is then invited to debate.

One objection must be met directly, because it appears to refute the whole picture. Was there not a “bonfire of the quangos”—the much-trumpeted programme under which the number of public bodies was cut by roughly a third between 2010 and 2015, falling from over 460 in 2016 to under 300 by 2019? There was; and it is one of the more misleading episodes in recent constitutional history. The count of quangos fell, but the power, money, and personnel lodged in the arm’s-length state did not. Over the same span in which the number of quangos was being pruned, gross resource spending rose—from some £195 billion in 2016 to over £220 billion by 2023/24—and staffing rose with it, from around 266,000 to over 300,000. The bonfire consumed the small and the visible—minor advisory committees, bodies whose abolition made a headline and saved little—while the great executive and regulatory bodies grew. Consolidation reduced the number of quangos even as it increased the quantum of delegated power (Cabinet Office 2019; 2025), which is precisely why ninety per cent of the funding now sits in the largest ten bodies. The bonfire was a reduction in the number of institutions mistaken, then and since, for a reduction in the reach of the unelected state. It was nothing of the kind. The “fire” left the house standing—and the figures, properly read, are not a refutation of the Starkey Thesis but its quantitative confirmation.

Judges and single-issue experts: when rights override budgets

There is a further mechanism, and Starkey states it with characteristic force: the modern settlement, he argues, “gives judges and single issue fanatics executive powers” (Starkey 2026). The complaint is not merely that power has migrated to the unelected, but that it has migrated to bodies and persons constitutionally incapable of weighing the financial cost of their decisions. A single-issue regulator, or a court vindicating a single-issue right, is by its nature relieved of the duty of trade-off that defines responsible government.

His examples are vivid. On the side of the single-issue agency, Starkey points to the conservation requirements that have descended upon major infrastructure: the “bat tunnel” built to shield a colony of rare bats on the route of the now-curtailed HS2 rail project, whose cost passed £100 million, and the fish-protection systems imposed on the Hinkley Point C nuclear power station, running to several hundred million pounds, to spare fish from the warmed water of its cooling outfall. The precise authorship of these costs is, it should be noted, contested — English Nature (now called ‘Natural England’), the conservation regulator, maintains that it did not mandate the HS2 structure but merely advised that the promoter’s own proposed design would satisfy the law — but the contest over authorship in fact reinforces the structural point. The cost is real and enormous; it flows from a regime of single-issue statutory mandates enforced through bodies and courts none of which is charged with weighing the safeguarding of bats or fish against the hundreds of millions of pounds of foregone infrastructure; and responsibility for it is so thoroughly diffused that the very body at the centre of the affair can truthfully disclaim it. No one decided that bats should cost £100 million; the statutory ratchet produced the figure, and no hand is answerable for it. Starkey observes that “nothing can be done anymore” (Starkey 2026). This is Schmitt’s abolition of responsibility in its purest administrative form.

On the side of the judiciary, Starkey’s example is the economic damage wrought by litigation under the Equality Act 2010, the principal anti-discrimination statute. The bankrupting of Birmingham City Council — the largest local authority in Europe — is the starkest case. Successive equal-pay claims, brought predominantly by women in roles such as teaching assistants, cleaners, and catering staff, argued that they had been unlawfully denied bonuses and premium payments routinely paid to predominantly male roles such as refuse collectors and street cleaners. Once the Supreme Court in 2012 permitted such claims to be pursued through the High Court, with its six-year limitation window, rather than the Employment Tribunal’s six months, the accumulating liability reached an estimated £650 to £760 million—on top of roughly £1.1 billion already paid out over the preceding decade—and in September 2023 the council issued a section 114 notice, the local-government equivalent of a declaration of bankruptcy. As Starkey puts it, a precedent has thereby been established that the payment of market wages is no longer a legitimate defence against a charge of gender discrimination (Starkey 2026).

Starkey’s deeper point reaches past any single statute to the structural incompatibility between justiciable rights and budgetary constraint. “Political power,” he argues, “was handed to experts and judges and the effect has been catastrophic” (Starkey 2026), because “matters of justice do not weigh costs.” Once a positive right is created—a right to a particular provision, enforceable in a tribunal or court—it cannot be adjudicated against a budget constraint, for the court’s office is to vindicate the right, not to balance the public finances. Positive rights, once established, can bust all budget constraints. The body that confers the entitlement and the body that enforces it are different from, and unaccountable to, the body that must find the money; the cost is, once again, externalised across the boundary between mandates.

The crisis of special educational needs provision is Starkey’s illustration, and the official figures bear him out with precision. The Education, Health and Care Plan (EHCP), introduced by the Children and Families Act 2014, creates a legally binding entitlement to specified support, enforceable against the local authority through the First-tier Tribunal. The Institute for Fiscal Studies observes that such plans “create legally binding entitlements to specific forms of support—with little regard for affordability” (Institute for Fiscal Studies 2025), which is precisely the rights-against-budgets structure Starkey describes. The consequences are exactly as the analysis predicts. The number of pupils with such plans has more than doubled in a decade; high-needs spending in England rose from £7.5 billion in 2016 to £12 billion in 2025; appeals to the tribunal reached 21,000 in a single year, of which families win some ninety-five per cent; and the accumulated deficit of local authorities, held off their books only by a temporary accounting override, is forecast to reach around £8 billion by 2027 — a trajectory the National Audit Office has called financially unsustainable (Institute for Fiscal Studies 2025; National Audit Office 2024). Here is a positive right, conferred by statute and enforced by a tribunal against which the local authority wins one case in twenty, busting the budget constraint of council after council exactly as Starkey’s argument anticipates. The point is not that children with genuine needs should go unsupported; it is that a system in which an unaccountable tribunal allocates an open-ended call on the public purse, while an entirely separate body must somehow find the money, is structurally incapable of the trade-offs that any sustainable provision requires—and is therefore driving the councils that bear the cost toward insolvency, the SEND provision toward crisis, and the children themselves toward outcomes that the spending conspicuously fails to secure.

A case in point: child safeguarding in the quango-state

Perhaps no other domain illustrates the abstract claims of the Starkey Thesis — the dispersal of power among single-issue bodies, the abolition of locatable responsibility, the ratchet of expanding intervention — more vividly and more disturbingly, than the UK’s child-safeguarding system. It is worth dwelling on at length, because it is the case in which the expansion of state power reaches into the existential, private sphere of the relationship between parents and their children. The forcible removal of a child from its parents is perhaps the most drastic intervention into a person’s life that a state can perform, and the apparatus that performs it has grown, since around 2010, in a manner that exhibits every pathology this essay has anatomised.

Consider first the institutional architecture, for it is a textbook instance of the multi-body arm’s-length state. Statutory responsibility for child protection rests with each council’s Children’s Services department under the Children Acts 1989 and 2004, operating within guidance — chiefly Working Together to Safeguard Children — issued by the Department for Education. But the practice is shaped, inspected, regulated, and interpreted by a constellation of arm’s-length and independent bodies, no one of which is answerable to the families it affects. Ofsted inspects local-authority children’s services, children’s homes, and fostering and adoption agencies, and its inspection frameworks powerfully shape practice on the ground. The Child Safeguarding Practice Review Panel, an independent national body working with the Department, commissions reviews of serious cases. Social Work England regulates the social-work profession: registration, standards, fitness to practise. The Children and Family Court Advisory and Support Service (Cafcass) appoints children’s guardians and advises the family court once safeguarding concerns reach proceedings. The Children’s Commissioner for England promotes children’s rights as an independent statutory office. The Care Quality Commission governs the points at which safeguarding meets the health service. The Local Government and Social Care Ombudsman adjudicates complaints of maladministration. A body called ‘Foundations - the What Works Centre for Children and Families’ is steering safeguarding practice by via research, evaluations, evidence reviews, toolkits and practical guidance for policymakers, commissioners and practitioners.

Here we thus see an example of the typical disaggregated structure: a statutory power exercised by local authorities, framed by a department, and then inspected, regulated, reviewed, advised, and adjudicated by at least eight further bodies, each pursuing its slice of the mandate, none bearing responsibility for the cumulative result. The interdependencies — which are the substance of any family’s actual situation — fall through the cracks between the mandates.

Now consider the scale, which has expanded in exactly the ratchet pattern referred to above and to be elaborated below. As I have documented in my Substack article Thesis on Government Disruption of Families (Schumacher 2026a), the number of children looked after by local authorities in England rose from around 59,000 in 2008 to over 72,000 by 2017, and to 84,000 by 2025. Moreover, the use of court-issued care orders — the most coercive mechanism available under the Children Act 1989 — rose by roughly forty per cent between 2012 and 2017; by 31 March 2025 some 55,000 children, around sixty-five per cent of all those looked after, were under a care order, having been forcibly removed from their parents. As I argue in my article, “these are not marginal changes at the edges of the system; they represent a qualitative expansion in the routine use of coercive state power over families” (Schumacher 2026b). And crucially, the increase cannot be explained by demography or by any evidenced surge in severe maltreatment: “there is no convincing evidence of a corresponding explosion in abuse that would justify such a dramatic increase in removals” (Schumacher 2026b). The pattern points not to changed conditions in families but to changed incentives and procedures within the apparatus.

Those incentives are the now-familiar logic of the insulated single-issue body, operating as what I have called “a one-way ratchet.” The mechanism deserves to be set out as I have described it, because it shows the agency pathology at its most granular:

It often seems as if the safeguarding priority is not to keep the children safe but to safeguard the children services department, i.e. mitigate risk to the department and to social workers … Once concerns have been raised the institutional dynamics seems to be like a one-way ratchet. … Professionals are encouraged to “err on the side of caution”. For parents any contact with child-related public services carries latent risk. (Schumacher 2026a)

This is Schmitt’s abolition of responsibility wearing its cruellest face: the body that registers the concern, the body that inspects, the body that regulates the social worker, the body that advises the court, and the court itself each discharge a partial function, while no one is accountable for the destroyed family that is the system’s aggregate output.

Exactly as the ratchet thesis predicts, the expansion has become economically entrenched — self-sustaining quite apart from any child’s welfare. A whole “cottage industry of private assessment firms and contact providers has emerged in the wake of the growth of safeguarding interventions, frequently staffed or founded by former Local Authority social workers, creating a self-reinforcing ecosystem around care proceedings” (Schumacher 2026b), much of it publicly funded through legal aid, which “embeds rising intervention rates into a stable, state-financed service economy.” The fiscal magnitude is now of national significance: children’s social-care expenditure in England reached an actual net cost of around £15.6 billion in 2024–25, the equivalent to 13% of the UK’s total national education budget (including universities), with a single child in residential care costing the taxpayer upward of £300,000 a year, and half a million in London. That this vast and rising sum is, as I note, “spent on ‘services’ imposed and (for the most part) resented by parents” (Schumacher 2026a) is the final mark of a system that has detached its own growth from the human good it was created to serve.

That the system is dysfunctional is not a heterodox discovery; it has been authoritatively diagnosed, repeatedly, and to no avail — which is itself the most telling evidence that the dynamic is structural rather than corrigible by reform. The ‘Care Crisis Review’ of 2018, facilitated by the Family Rights Group in direct response to the then-President of the Family Division’s warning that “we are facing a crisis … and we have no very clear strategy for meeting [it],” examined the rise in care applications and a looked-after population that had reached its highest level since the Children Act 1989, and set out twenty options for change (Family Rights Group 2018). More comprehensively still, the government’s own ‘Independent Review of Children’s Social Care’ — the MacAlister Review of 2021–22, the most important national review of the system in a generation — found in its final report that children’s social care had “outcomes for children that continue to be unacceptably poor and costs that continue to rise,” and concluded that “a radical reset is now unavoidable” (MacAlister 2022). The trajectory of rising costs, rising removals, and failing outcomes was thus identified, in official terms, years ago. Yet the figures above — 84,000 looked-after children and £15.6 billion by 2025 — show the trajectory continuing undeflected. A system that absorbs its own most authoritative critiques and carries on expanding is not suffering from a remediable defect of policy; it is exhibiting the self-propelling institutional logic that this essay has identified as the signature of the system of insulated, single-issue, unaccountable bodies. What we witness since 2010 is that child safeguarding in the quango-state is running amok, breaking up families against all common sense and incurring costs that spiral out of all common sense proportions.

The child-safeguarding apparatus is therefore the Starkey Thesis in concentrated form, and a refutation in advance of the restorationist reply. The powers at work here did not merely migrate from Parliament; they expanded, behind the insulating language of “safeguarding” and “risk,” into a domain the Victorian constitution would never have dreamed of entering. They are exercised by a dispersed array of single-issue bodies none of which answers to the families it intervenes upon. And the remedy is plainly not to repatriate the power to Parliament — which legislated the duties and would only re-legislate them — but to withdraw it: to restore to families the presumption of competence and the sphere of private self-determination that the safeguarding state has progressively confiscated, returning to civil society the supportive, voluntary, and genuinely protective functions that the coercive apparatus has both displaced and failed to perform. The case makes the general argument unavoidable. When the most intimate relationship in human life has been drawn into the antechamber of the unaccountable agency, the reduction of the scope of political control is no longer an abstract preference but an urgent necessity.

10. The Ratchet: How Disempowering Parliament Enlarged the State

To this point I have used the Starkey Thesis chiefly as a constitutional datum to be explained by Schmitt. It is time to credit it as an original contribution in its own right, and then to extend it in a direction Starkey himself does not take. David Starkey’s achievement—developed not in a single treatise but across a series of lectures, interviews, and broadcasts (Starkey 2024a, 2024b, 2024c, 2025)—is to have located, with a historian’s eye for the longue durée, the constitutional mechanism behind a whole catalogue of contemporary governmental failures that the daily commentary treats as unrelated misfortunes. His signal insight is that these are not so many separate policy mistakes but symptoms of a single structural transformation: the dispersal of the sovereignty of the Crown-in-Parliament among a constellation of unelected, single-purpose bodies, each pursuing its remit without answering to the electorate or bearing the costs it imposes elsewhere.

The list of failures Starkey assembles is worth setting down, because its very heterogeneity is the point. Uncontrolled immigration into a welfare state that cannot absorb its fiscal consequences; a housing crisis that successive governments prove structurally incapable of addressing; energy costs driven upward by a decarbonisation target written into law and placed beyond ordinary political revision; a general cost-of-living crisis compounded by monetary and asset-price inflation; NHS waiting lists that defy every reorganisation; a media culture – both state media and established corporate media - widely perceived as ideologically uniform; expanding restrictions on lawful speech; the capture of universities, professional bodies, corporations, and cultural institutions by an activist orthodoxy; and presiding over all of it an unaccountable permanent civil service—the “blob”—that is at once unresponsive to ministers and obstructive of the electorate’s expressed preferences. A government nominally responsive to voter expectations finds itself unable to deliver on almost any of the things voters most insistently demand. Starkey’s contribution is to explain why: the levers the voter imagines connect to the machine have, for the most part, been disconnected and handed to bodies the voter cannot reach.

The single-issue agency and the externalisation of its costs

Starkey’s sharpest analytical point—and the one most congenial to the framework of this essay—concerns the structural pathology of the single-issue body. A quango, regulator, or statutory committee is, by design, charged with one objective and insulated from the discipline of having to weigh that objective against the others a society must simultaneously pursue. It is constituted precisely so that it will not trade off its mandate against competing goods. The independent central bank pursues its inflation target; the climate body pursues its emissions target; the public-health authority pursues the suppression of pathogens like COVID19 at any cost; the rights regulator pursues its protected characteristics—and each does so, in Starkey’s telling, without being accountable for the costs and side-effects its single-minded pursuit inflicts across the rest of the social system. This is the institutional embodiment of what an economist would call the externality and what Schmitt, in a different register, called the abolition of responsibility: the body that makes the decision does not bear, and is not answerable for, its consequences elsewhere.

Starkey’s most compelling illustration is the Covid-19 lockdown, and it repays examination because it exhibits the entire mechanism in miniature. The lockdowns were imposed, for the most part, not through primary legislation debated and amended in Parliament but by ministerial regulation under the Public Health (Control of Disease) Act 1984—statutory instruments frequently made and brought into force before they were laid before Parliament, with members permitted only to approve or annul, never to amend (Hansard Society 2021; House of Lords Constitution Committee 2021). The decisive advice issued from a public-health apparatus charged with one variable—the suppression of viral transmission—and structurally indifferent to the vast costs its prescriptions imposed in every other domain: the foregone cancer diagnoses and untreated illnesses, the lost schooling of a generation, the destruction of small businesses, the monetary emission required to fund the furlough, the mental-health toll, the sheer fiscal magnitude of the whole. A body optimising a single objective, insulated from the duty of trading it off against the others, will rationally drive that objective to a degree that is collectively ruinous—because the ruin is, from its institutional standpoint, someone else’s department. The lockdown was the single-issue agency’s logic bulldozing everything in its way with devastating effect.

Notice how precisely this vindicates the complexity argument of Chapter 8 from the opposite direction. There I argued that no all-purpose chamber can master the differentiated complexity of modern society. Starkey’s point is the complementary danger: that carving the single social system into single-issue jurisdictions, each sovereign within its silo, does not solve the complexity problem but disaggregates responsibility for it, so that the interdependencies—which are the very substance of a complex society—fall through the cracks between mandates. The all-purpose chamber cannot handle the complexity; the array of single-purpose agencies handles each strand while severing it from the others. Neither institutional form can do what only a decentralised order of markets and discourses can do: register the trade-offs continuously, at the margin, through the dispersed knowledge and existential feedback of those who actually bear the costs.

From transfer to expansion: the ratchet thesis

Here I extend the Starkey Thesis beyond its author. Starkey describes the Blair settlement as a transfer of power—a redistribution of sovereignty from Parliament to the agencies. I want to argue that the same move was simultaneously, and not incidentally, a vast expansion of the total quantum of state intervention and control. The dispersal did not merely relocate a fixed stock of power into unaccountable hands; it manufactured new power. Delegation, on this account, is not a zero-sum reshuffling but a ratchet.

The mechanism has several strands, each independent and each compounding the others. First, delegation lifts the throughput ceiling that scarce parliamentary time imposed. Parliament is a single congested chamber; the volume of rules it can originate and scrutinise is capped by its own finite attention—a cap that was, in effect, a natural limit on the total quantity of state direction. Distribute rule-making among dozens of specialised agencies operating in parallel, each with its own staff, budget, and statutory remit, and that ceiling is pushed much further. The production of binding rules is parallelised, and output multiplies accordingly. Second, each agency acquires an institutional interest in the expansion of its own jurisdiction. Insulated from electoral feedback and from the market’s existential discipline alike, it faces no countervailing force against the steady enlargement of its remit, its guidance, and its budget—the dynamic that the public-choice theorists, following William Niskanen, identified in the self-aggrandising bureau. Third, the agency form generates a vast domain of “soft law”—codes of practice, regulatory guidance, supervisory “expectations”—that carries quasi-binding force without ever passing through the legislature at all. This is not relocated parliamentary power; it is newly created governing power for which there was no prior parliamentary analogue.

The fourth and subtlest strand is the most important. Depoliticisation lowers the political cost of new control, and thereby makes possible expansions that an openly political government could never have carried. When a contested judgement is vested in an “independent expert” body, it is removed from the arena of partisan contestation and reclothed as technical necessity. Opposing it then reads as opposing expertise rather than opposing a policy—a far higher barrier. Depoliticisation does not shrink the state; it launders the expansion of the state, rendering palatable a degree of control that naked political imposition could not have achieved. The independence is not a brake on the growth of intervention; under this analysis it is the very technology of its growth. Here Schmitt’s observation from Chapter 5 returns with full force: when the political sphere is hollowed of substance, “another sphere comes into play with ruthless power.” The agency is the vehicle through which substantive control expands behind the screen of a hollowed, merely formal, democratic politics.

This extension does more than enlarge Starkey’s indictment; it strengthens the central caveat of this essay. If delegation had merely moved a fixed stock of power, the restorationist reply—bring it back to Parliament—would at least restore a prior equilibrium. But if delegation expanded the total quantum of control, then repatriation cannot undo the expansion. It would hand a now-enlarged scope of intervention to the chamber which would thereby be utterly swamped. The real disease is not misplaced power but multiplied power, and only a reduction in the scope of control—not its relocation—can be a remedy. The ratchet thesis thus converts the Starkey diagnosis into an argument for the libertarian conclusion rather than the restorationist one. I see here a real positive opportunity: The abolishment of the rightly criticized quangos could be politically a popular move and could thereby lead to a serious reduction of state intervention in economy and society.

How the agencies fed the stagnation since 2008

The expansion is not merely constitutional; it has a measurable economic counterpart in the productivity stagnation that has gripped the advanced economies, and Britain especially, since the financial crisis of 2008. I do not claim mono-causality. But the proliferating apparatus of insulated, single-issue intervention is, I contend, a substantial and under-acknowledged contributor to that stagnation, and a few concrete instances show the mechanism at work.

Consider monetary policy, the most powerful instance. The independent Bank of England—precisely the body I conceded in our discussion to be preferable to a politically manipulated central bank—expanded its balance sheet through quantitative easing from nothing in 2009 to a peak of £895 billion by late 2021, equivalent to over forty per cent of annual GDP (Bank of England 2021; 2024). The House of Lords Economic Affairs Committee, in a report pointedly subtitled A Dangerous Addiction?, found that the policy had become poorly understood, hard to reverse, and had fed “perceptions that the Bank of England has become politicised” (House of Lords Economic Affairs Committee 2021). The salient point for the present argument is distributional: a decade of suppressed interest rates and asset purchases inflated the prices of assets—housing, equities, bonds—held disproportionately by the already wealthy, while doing little for the productivity of the real economy. The asset-price inflation that figures in Starkey’s catalogue of grievances, and the cost-of-living and housing crises that accompany it, are in substantial part the work of a single insulated committee optimising a single target, its vast side-effects falling on a society that had no means of contesting them. This is the result of decisions made in an independent quango: a transfer to asset-holders, effected by decree, for which no one stands accountable at any dispatch box.

Consider energy and the net-zero framework. The Climate Change Act 2008 created a statutory duty and an independent Climate Change Committee to advise on legally binding five-year carbon budgets; the 2019 amendment converted the original eighty-per-cent target into a one-hundred-per-cent “net zero” obligation by 2050, making Britain the first major economy to inscribe such a commitment in law (House of Commons Library 2026; Grantham Research Institute 2024). Whatever one’s view of the climate science, the constitutional and economic form of this settlement is precisely the pathology under examination: a decarbonisation trajectory placed beyond ordinary political revision, binding governments of every complexion in advance, and pursued by a dedicated body structurally indifferent to the energy costs, industrial competitiveness, and cost-of-living consequences that its single mandate imposes across the rest of the economy. Energy-intensive industry contracts or relocates; electricity prices embed a structural premium; and the body charged with the target bears none of these costs, which fall—again—on a public that cannot vote them down because the target is a matter of law, not of policy. A society cannot achieve dynamic prosperity while one of its most fundamental input prices is governed by a legally entrenched, single-issue mandate insulated from the trade-offs it generates.

Consider, finally, the wider regulatory and administrative thickening that the proliferation of agencies represents in aggregate. Each of the bodies surveyed in the contemporary arm’s-length state—the conduct and competition regulators, the sectoral regulators in energy, water, communications, and rail, the health, environmental, education, and equality regulators—generates its own accreting stratum of rules, guidance, reporting requirements, and compliance obligations. The cumulative effect is the diversion of an ever-larger share of managerial attention and capital from production to compliance: the firm staffed increasingly with officers whose function is liability-containment rather than value-creation, the investment decision deferred pending regulatory clearance, the innovation never attempted because the permitting cost is prohibitive. This is the microeconomic texture of the macroeconomic stagnation. The post-Fordist economy, as I argued in Chapter 8, is defined by its capacity to absorb unlimited innovation—but only if the institutional framework permits the rapid, decentralised, trial-and-error experimentation through which innovation proceeds. The thickening regulatory apparatus is precisely a tax on that experimentation, and its proliferation since the constitutional settlement of the late 1990s tracks, and substantially helps to explain, the productivity slowdown that followed.

The convergence of the three cases is the argument. In each, a power was removed from the elected chamber and lodged in an insulated body; in each, the insulation permitted an expansion of intervention—monetary, regulatory, climatic—beyond what open political contestation would have tolerated; and in each, the single-issue body pursued its mandate without bearing the diffuse, system-wide costs that its pursuit imposed. The Blair settlement did not merely move the levers out of reach. It multiplied them, insulated them, and set each to pulling in its own direction, with no hand responsible for the resultant. That this should manifest as economic stagnation is exactly what the analysis predicts: a society whose productive energies are progressively constrained by a growing thicket of insulated commands can be expected to lose its dynamism, and it has.

11. The Decisive Caveat: Not Back to Parliament, but Out to Civil Society

Here I part company with the seemingly natural conclusion of the Starkey Thesis. The envisioned remedy in the restorationist constitutional critique is repatriation: bring the powers home, restore the sovereignty of Parliament, re-subordinate the quangos and the courts and the central bank to the elected chamber. I do not accept this remedy, and the foregoing analysis is precisely what licenses my rejection.

Consider what the restorationist programme actually proposes. It proposes to return decisive power over monetary policy, over vast tracts of economic and social regulation, over the adjudication of rights, to a chamber that Schmitt has shown us is no longer — and under the conditions of mass democracy cannot again be — a genuine deliberative assembly in the classical sense. The discussion-principle that gave parliamentary sovereignty its moral and intellectual title is, on Schmitt’s demonstration, defunct. To repatriate power to the modern Commons is therefore not to restore government by discussion; it is to hand that power to the very apparatus of party discipline, propaganda, and interest-brokerage that Schmitt diagnosed — to the “social or economic power-groups calculating their mutual interests,” to the party soldiers under the whip and under the influence of lobbyists, and to the manufacture of mass opinion. The restorationist mistakes the form of the old constitution for its lost substance. He would repaint the radiator and expect heat.

My alternative follows from Schmitt’s deepest insight, the one I flagged in Section 2 and developed through the intervening chapters. Recall that the discussion-principle of parliament and the competition-principle of the market are, in Schmitt’s analysis, versions of the same liberal idea: “That the truth can be found through an unrestrained clash of opinion and that competition will produce harmony … it is exactly the same.” The classical liberal trusted spontaneous, competitive, decentralised processes—of argument and of exchange—to discover truth and to coordinate action, against the pretension of any single will to command the result. We have since seen, with Habermas, that the discussion principle, i.e. non-strategic communicative reason, is the more fundamental of the two; and we have seen, with Luhmann, through the complexity barrier, that no all-purpose chamber can match the differentiation and dynamism of post-Fordist world society. The tragedy of the modern state is that the trust in decentralised discovery was honoured in the chamber and the marketplace while being progressively dishonoured everywhere else, as the scope of political command expanded against all these grounds at once. The frustration the Starkey Thesis mobilises is, rightly understood, frustration at command displacing discovery.

The remedy, then, is not to relocate the command back to Parliament but to dissolve the command. The powers that have drained out of the chamber should not be dammed up and returned; the channel should be opened wider, so that they flow past the quango and past the regulator and out of the political system altogether—into the functionally differentiated subsystems of a free society, into markets and into the open discourses of an accomplished public, into private self-determination. Monetary policy need not be returned to Parliament or left with an unaccountable central-bank committee; the deeper question is why the production of money should be a political monopoly at all. Vast tracts of regulation need not be repatriated to the Commons or administered by agencies; the deeper question is why these matters should be settled by command rather than by competition, contract, and voluntary self-regulation. The adjudication of countless disputes need not be a contest between Parliament and a supreme court; much of it can be returned to the polycentric, contractual, and arbitral institutions that a free civil society generates.

The monetary case: from politicised money to independent money to free money

It is worth following one case all the way to its conclusion, because it exhibits the whole movement of this essay in nutshell and because it is the case where the libertarian completion is at once most radical and most easily misunderstood. Take money — the function discharged by the most powerful quango of all. The progression has three steps. The first is direct political control: the central bank as an instrument of the elected government, its rate-setting hostage to the electoral cycle and to shifting parliamentary majorities. I reject this, and I reject it on precisely the Schmittian and Bagehotian grounds developed above—the popular will, formed and managed as it is, cannot be trusted with a decision whose costs are dispersed, deferred, and largely invisible to the voter. The second step is independence: the decision lifted one remove out of direct political contestation and lodged in an expert committee. I willingly concede that this is an improvement. I much prefer an independent central bank to one under continuous political pressure; the depoliticization of the money supply, the removal of the printing press from the immediate reach of vote-seeking ministers, is a genuine improvement, and nothing in what follows denies it. On this aspect, monetary policy is special: this policy lever should definitely not be in the hands of party politicians. (In most other cases of quangos one can say that the farming out of decision making to single issue arm’s length bodies, away from parliament where these issues can be debated and connected with all the other issues that need to be considered, was a deleterious mistake.) However, while in the case of monetary policy the removal from parliamentary party politics was a move in the right direction, it was certainly not enough to prevent central monetary planners to damage the economy.

Bank of England independence is only a halfway solution, and a still detrimental one. It is the monetary form of exactly the quango settlement this essay has anatomised: the decision has been moved one step out of politics and then stopped there, lodged in an unaccountable expert body rather than carried through to its proper home. The independent central bank is the monetary OBR. And as Chapter 10 argued, this insulation does not really discipline the expansion of monetary intervention; it still licenses it — witness the growth of the balance sheet to 40% of GDP under the cover of supposed technical necessity. Independence cures a defect of political incentive while leaving wholly untouched the deeper, Hayekian defect of knowledge: an independent monetary authority is still a central planner of the price of money and the quantity of credit, and it faces the same epistemic impossibility as any planner. The committee, however expert and however insulated, cannot possess the dispersed knowledge that only a competitive process can mobilise. The independence reform improves the planner’s motives; it does not solve the planning problem, because the planning problem is not a problem of motive but of knowledge. That is why the very fact that the monetary decision had to be removed from democratic control in order to function tolerably is a quiet confession that it was never suited to political determination at all — and the natural question is therefore not “which unaccountable committee should set it?” but “why is this a collective political decision rather than a price discovered in a market?”

The third step — the completion of the movement the independent central bank only begins — is the removal of money from state control altogether. There is no need for a state monopoly here. This is Hayek’s radical proposal in The Denationalisation of Money (1976): strip the state of its monopoly on the issue of currency, permit private issuers to offer competing monies, and let the choices of those who use money determine which survive. The rationality here is the same competitive discipline that the classical liberal trusted in the chamber and the marketplace alike: an issuer who debased his currency would see it abandoned for sounder rivals, so that competition would select continuously for stability of value in a way no politically appointed committee, insulated from that feedback, ever can. Money would cease to be an instrument of political command and become, once more, a discovered institution — results of human action but not of human design.

Hayek gave us the mechanism but, characteristically, declined to predict its product; he expected competition to select for stable-value money without forecasting which form would win, and was himself wary of the gold standard’s automatic, quantity-driven rigidity. I am willing to supply the prediction he withheld, though on structural rather than metallurgical grounds. Money is subject to powerful network effects: a medium of exchange grows more useful the more others accept it, and a unit of account more useful the more prices are quoted in it. Monetary competition is therefore genuinely open at the level of issue, but it does not dissipate into a permanent babel of rival units; it tends to converge upon a dominant standard. My expectation is that free competition would converge upon a politically uncontrollable hard money — one whose supply cannot be manipulated by any authority — of which a gold or gold-backed money is one plausible form, though by no means the only one. The decisive property is not the metal but the immunity from political debasement and the multi-millennial Schelling-point status (after Thomas Schelling) that makes coordination upon it self-reinforcing. Across thousands of years and disparate cultures, gold has repeatedly been settled on as money. That long track record is itself the coordinating signal. Still, whether the convergent standard proves to be metallic, a digital bearer asset of algorithmically fixed scarcity, or some hybrid or basket, is an empirical question for the competitive process itself to settle. What the argument commits to is the structural prediction—convergence upon a scarce, non-political standard—not the particular substance that instantiates it.

The obvious advantage of such convergence would be the supersession of the present system of politicised national monies, whose floating exchange rates impose a vast and largely unrecognised deadweight upon economic rationality. Consider the scale of the cots the current system imposes. Foreign exchange is, by turnover, the largest market in the world: average daily trading reached some 7.5 trillion US dollars in April 2022. The greater part of this colossal activity is parasitic — it exists to hedge and arbitrage a volatility that is itself the artefact of there being multiple, politically-managed national currencies in the first place. Under a single convergent global money the overwhelming bulk of that turnover would simply have no reason to exist. The largest market on earth is, on this view, a standing measure of the deadweight cost of monetary nationalism, much as the size of a vast pharmaceutical bill measures the extent of a disease. (As an entrepreneur selling services globally, I painfully recognize the scale of the distortion myself: How do you manage and motivate your staff when the annual bottom line used for bonus allocations turns from a measure of skill and effort to a gamble based on exchange rates?) Mises’s calculation argument operates here at planetary scale: exchange-rate noise corrupts the very price signals on which the rational coordination of a world economy depends, and the resources consumed in trying to manage that noise are resources withdrawn from production. The prosperity costs are as devastating as they are unseen, because the more stable and more integrated counterfactual world is never observed.

An objection presents itself immediately, and it is the right one to anticipate, because it appears to turn this essay’s own argument against it. Does not convergence upon a single dominant money reintroduce, through the back door, exactly the singular that Chapter 8 condemned — a monetary monopoly to replace the all-purpose chamber? It does not, and the distinction is the same Hayekian nomos–thesis distinction that governs the whole libertarian case. What was objectionable in the central bank was never the singularity of the standard but the coerciveness of the monopoly and the discretion of the command. A unit of account that emerges from, and is continuously disciplined by, free choices is categorically different from one imposed and manipulated by a state. The convergent money is a standard, not a sovereign: no committee sets its quantity, no authority can debase it, and exit into competing monies remains permanently available as the disciplining threat that keeps the standard honest. It is “singular” in the way the metre, the second, or a dominant language is singular: a convergent coordination equilibrium, an evolved nomos, not an imposed thesis. The plurality on which this essay insists is preserved precisely where it matters — at the level of issue, process, and the standing possibility of exit — while convergence occurs only at the level of the shared standard, which is the mark of a healthy spontaneous order rather than its negation. A converged standard that anyone may abandon for a sounder rival is the opposite of a monopoly that forecloses exit; the difference that makes a difference here is permanent contestability.

The historical gold standards invoked by critics — blamed, not without reason, for transmitting deflationary shocks and deepening the Great Depression — were state-managed gold standards, fractionally backed and suspended by governments the moment convertibility constrained them; that suspension is itself the tell, the proof that they were instruments of political discretion rather than the freely-chosen, fully-disciplining commodity money the present argument envisages. In a free system the creeping inflation prior to the rapid deflationary contraction would not have existed.

The question of form remains genuinely open: since 2009 the live empirical instance of denationalised, network-effect-driven, scarce money competing against state currencies has been digital rather than metallic, and a thesis that committed itself to gold specifically, against a programmable and costlessly-transmissible alternative, would be taking on a far heavier burden than my argument and purpose here requires. The structural claim—that free monetary competition converges upon a scarce standard, dissolving the deadweight of exchange-rate volatility—stands whichever form prevails. This is the monetary instance of the general thesis: the decision drained from the chamber should not be repatriated to Parliament, nor left in the antechamber of the independent committee, but carried through to its proper home in the discovery procedures of a free, global civil society.

This is where Schmitt’s diagnosis of the hollowing of political equality becomes, against his intention, contributes to my libertarian project. He observed that when political equality is stretched to a substanceless universal, “this area loses its substance and is overshadowed by another sphere in which inequality then comes into play with ruthless power.” He meant the economy’s dominance over politics as a deplorable usurpation. But the “inequality” that comes into play in the economic and discursive spheres is the legitimate, productive differentiation of competence, achievement, and contribution—the hierarchies and discriminations that emerge endogenously within markets and within disciplined discourses, and that are entirely compatible with the formal equality of participation. The right inference from Schmitt’s observation is not to re-substantialise the political so that it can reconquer the place of prominence from the economy, but to let the political recede so that the differentiated subsystems, in accordance with Luhmann’s insight, can operate according to their own codes — profitable or unprofitable in the economy, true or false in science, legal or illegal in the legal system, pass or fail in the education system etc. — without a master political code overriding them by command.

On this reading, the quango is a transitional form, and a revealing one. By removing decisions from the chamber on the explicit ground that they are too important, too technical, or too long-term to be entrusted to electoral or party-political determination, the architects of the Blairite quango state have already half-conceded the libertarian premise: that political command is a poor instrument for coordinating complex matters. They have merely stopped halfway, lodging the decision in an unelected, unaccountable public body instead of releasing it to the genuinely decentralised discovery procedures of civil society. The libertarian completes the movement the quango begins. Where the restorationist would drag the decision back into the political system, and the technocrat would leave it in “expert” hands, the libertarian carries it through to its proper home in a decentralized voluntary order. The frustration Starkey names is the energy and motivational force; the reduction of the scope of political control is the direction in which I want to channel and discharge it.

On the population and the management of its will

It remains to address, with the candour that the subject demands and that Bagehot exemplified, the awkward matter of the population and its political will — for it is here that my reservations about mass democracy and universal suffrage are most exposed, and here that the restorationist appeal to “giving power back to the people” is at its most seductive and most dangerous.

Bagehot, writing before universal suffrage, could say plainly what we now circle around: that the mass of the population does not conduct, and largely cannot judge, the efficient business of government; that it is governed with its consent but not by its participation; that its reverence is won by the “dignified” parts of the constitution - then the Royal spectacle - while the efficient parts are worked by a competent minority. Schmitt, writing after the arrival of mass suffrage, supplies the mechanism by which this older reality was not abolished but industrialised: the mass will is no longer simply deferential and dormant, as in Bagehot, but is actively, in Schmitt’s words, “won over through a propaganda apparatus whose maximum effect relies on an appeal to immediate interests and passions.” The “statistical apparatus” of the secret ballot then reassembles a hundred million private, irresponsible acts into a number that is presented as “the will of the people.” Put Bagehot and Schmitt together and the conclusion is uncomfortable but, I think, undeniable: in mass democracy the political will of the population is far less expressed than it is formed, managed, and hedged — formed by the propaganda apparatus, managed by parties and the media-administrative complex, and hedged by the very quangos and courts whose insulation from the electorate is defended on the ground that the popular will cannot be trusted with certain decisions.

Once this is admitted, the restorationist case loses much of its moral force. “Return power to the people’s elected representatives” sounds like a demand for democratic accountability, but under the conditions Schmitt and Bagehot jointly describe it amounts to returning power to the apparatus that manufactures and manages the very “will” it claims to obey. The honest libertarian response is not to pretend that a purer mass democracy lies on the other side of repatriation, nor to indulge the fantasy that the chamber can be made deliberative again by procedural reform. It is, rather, to reduce the stakes. The more that is settled by command — whether the command issues from Parliament, from a quango, or from a court — the greater the prize for whoever can capture and manufacture the mass will, and the more ruinous the contest to control it. The more that is left to private self-determination — to markets, to voluntary association, to the friendly societies and mutual aid institutions that civil society generates when the state withdraws, to the open discourses of an accomplished public — the smaller the prize, the lower the temperature, and the less the population’s will needs to be manufactured at all, because less hangs upon it. A free people is not one whose every preference is aggregated and commanded; it is one most of whose life is removed from the reach of command in the first place.

I am, accordingly, less than enthusiastic about mass democracy and universal suffrage as the solution to the constitutional disorder the Starkey Thesis describes — because they are closer to being its cause. This is not a brief for any authoritarian alternative; Schmitt’s own later drift toward the decisionist and ultimately the dictatorial is precisely the wrong turn, and I reject it even more firmly as I reject the restorationist one. I am not principally opposed to democratic decision procedures. Far from it. I am fully endorsing that democratic decision making might be an appropriate means to coordinate collective action or rule-making in voluntary associations like clubs, the cooperatives, and shareholding companies. Here democratic procedures are wholly unobjectionable, indeed often very valuable, valuable, exactly because its scope is bounded and its membership self-selected and, according to Schmitt’s insight, substantively homogeneous in the relevant respect. There is no hard and fast rule whether and how a voluntary association will be democratically constituted. This is largely a question of expediency, and of what members prefer. What I reject is the default projection of democratic command or majoritarian dictates over the whole of society – where membership is not a choice but lock-in prevails - through a coercive state apparatus, and the consequent manufacture of a mass will to legitimate that command. To be sure, our standard model of ”liberal democracy” we largely take for granted now, did indeed function rather well for over 150 years, especially in comparison to most older models (although some monarchies are still doing reasonably well also), as well as in comparison to the failed left-utopian models. However, past performance is not indicative of future results, and the recent performance of this model, in most of its most advanced versions, has been very disappointing indeed. Our democracies are facing an unprecedented crisis of legitimacy. They are not performing anymore. Since 2008 they became increasingly illiberal, controlling and interventionist at a time when control and intervention are increasingly bound to do more damage than good. Following Carl Schmitt’s insight into the tension between liberal and mass democratic principles, we can see that this tension is now revealing itself increasingly as contradiction. “Liberal democracy” is increasingly becoming an oxymoron, and we need to explore what lies beyond. We need to explore how, going forward, this contradiction can be sublated in a new type of politics, a libertarian politics, building upon the discussion principle, not aiming at collectively binding decisions, but as guiding pendant of voluntary market relations.

12. Conclusion: Completing the Movement the Quango Began

Schmitt closed the principal argument of his book with a sentence that has lost none of its force: “If in the actual circumstances of parliamentary business, openness and discussion have become an empty and trivial formality, then parliament, as it developed in the nineteenth century, has also lost its previous foundation and its meaning” (Schmitt 1923/1988, p.50). The Starkey Thesis points, in effect, to the British constitutional history of that loss of meaning — the record of how an institution whose moral title rested on genuine deliberation, embedded in a wider public sphere of accomplished stakeholders, was first hollowed by mass democracy and then dismembered by an executive that exported its substance to unaccountable bodies.

In my Schmitt-inspired engagement with the Starkey Thesis I accept three things and refuse one. I accept Schmitt’s account of what parliament was for—discussion as a principle, embodied in the independence of the member, the publicity of proceedings, the freedom of the press, and the separation of powers, and embedded in a deliberating educated public. I accept that this principle stands in genuine contradiction with mass democracy, and that the contradiction generates the pathologies — the antechamber parliament, the back-room decision, politics as spectacle — that Schmitt named in 1923 and that we inhabit in 2026. And I accept Starkey’s contention that the Blairite settlement carried this disintegration to a new and graver stage by transferring decisive power out of Parliament to quangos, regulators, and courts, draining Parliament’s substance, leaving theatre, and abolishing responsibility in the process, thus going further along the line Schmitt had already clearly seen.

What I refuse is the restorationist conclusion. The powers should not go back to Parliament, because the Parliament to which they would return is no longer the deliberative assembly whose principle once justified its sovereignty, and cannot be made so under mass-democratic conditions. The frustration the Starkey Thesis so rightly mobilises should be turned outward, not backward: toward the reduction of the scope of political control over society as such. The quango, by conceding that great decisions are ill-suited to electoral deliberation, has already taken the first step of a journey it lacks the courage to complete. The libertarian completes it. The decisions that have drained out of the chamber should continue their course — out of the political system altogether — into the markets, the voluntary associations, the mutual institutions, and the open discourses through which a free civil society discovers truth and coordinates action without command. That is the same competitive, spontaneous discovery-process that Schmitt identified as the shared underlying principle of the marketplace and the classical parliament alike. Parliamentary discussion, in its great age, was civil society reasoning with itself about the public good. When the chamber can no longer perform that office, the answer is not to pretend that it can, but to return the reasoning to the agile multitude of discursive arenas that are currently proliferating, and the deciding, to the self-responsible market participants, and the multitude of voluntary associations flourishing within civil society.

The hour in which Britain finds itself is, in this respect, a Schmittian hour of decision—though the decision is not the one Schmitt offered between liberal parliamentarism and a homogenising dictatorship. It is the decision between continuing to shuttle sovereignty between a hollowed-out chamber and an unaccountable quango-state on the one side, and gradually, deliberately discharging that sovereignty back into the private and voluntary order from which the modern state had progressively confiscated it on the other side. The Starkey Thesis diagnoses the disease with great precision. Carl Schmitt explains why the patient was always susceptible to it. Walter Bagehot shows us both the constitution that was lost and the candid truth about the population that the lost constitution managed so much more honestly than ours. The remedy, however, belongs to neither Starkey’s restoration nor Schmitt’s decisionism, but to the libertarian completion of the movement that the quango-state, in its half-hearted way, has already begun.

Is there hope? The political feasibility of the libertarian turn

There remains the hardest question of all, and it would be dishonest to close without facing it. In my 2025 ‘Pathways’ article I posed the question as follows: “How plausible is then the expectation that the coercive ordering of social life via states could be overcome and be substituted by a spontaneous social ordering without coercion?” My answer was and remains: “This might happen within current democracies through a discourse-led political activism that gradually convinces the demos, including parts of its elites, to risk more degrees of freedom. Such a political activism includes the build-up of political parties aiming at electoral success” (Schumacher 2025).

The libertarian revolution I advocate can succeed only if it first wins political power within the very compromised parliamentary democracy whose deep flaws this essay has spent its length anatomising. There is no road to the reduction of political power that does not pass, at least initially, through the existing political system. The reduction of the scope of the state must itself be legislated; the dismantling of the apparatus must be commanded by those who have won the command. Is there any hope in such a project? Can there be, at this point, any realistic prospect that a demos formed, managed, and hedged as we have described it could be brought to vote for less of the very political control that flatters and provides for it?

Two features of the present juncture give more ground for hope than the bare description would suggest. The first is the ongoing economic stagnation — experienced, by now, by a great many not as mere slow growth but as outright decline: stagnant real wages, a housing market now closed to most, rising taxes purchasing visibly deteriorating services, and the rapid erosion of the expectation that tomorrow will be more prosperous than today. A demos content with its provision will not risk the unfamiliar; a demos that has begun to experience the managed decline of the state-administered economy is a demos that may, at last, open up to the argument that the apparatus of control is the cause and not the cure of its predicament. The stagnation and decline this essay has attributed in part to the thickening arm’s-length state is also, dialectically, the condition that makes the case against that state politically audible.

The second is the political disillusionment that the stagnation has bred, and the new political formations it has called into being. The monopoly of the established centre - now routinely referred to as the “uniparty” – is in dissolution. New players have emerged —Reform UK and Restore Britain among them — each carrying libertarian ingredients that are, at present, mixed with statist and nationalist currents, but that could be strengthened, clarified and distilled by sustained discursive work. The Conservative Party, in opposition and chastened, has undergone something of a reinvention in a more libertarian direction — a party in which, it should be remembered, libertarian currents have always found a more receptive context than anywhere else in the British mainstream. Stalwart libertarian think tanks like the Institute of Economic Affairs (IEA) and the Adam Smith Institute (ASI) – who were already crucial contributors to the Thatcher revolution and to which the present author has been contributing – are gathering momentum. And at the furthest, principled edge of this landscape stands the — so far admittedly tiny — UK Libertarian Party (LPUK), which is at present building up its capacity and its policy apparatus, and to whose programme the present author has contributed as author of its Urban Development and Housing Policy. None of these is, today, a vehicle equal to the task. But their very emergence is evidence that the political consciousness is no longer frozen, that the broadly libertarian direction of travel has constituencies, research organisations, voices and parties forming around it, and that the discursive space in which a libertarian argument can be made and heard is wider than it has been in a generation.

I want to close with considering a recent historical precedent, the Thatcher revolution. This is pertinent as we are still inhabiting the same socio-economic epoch engendered by the revolutionary computational technologies which currently make yet another leap forward via AI. Here is how I appraised the plausibility of a libertarian revolution in my Pathways article written two years ago:

“This probably seems a very remote and far-fetched prospect given current political reality and mainstream political consciousness. Yet, historical dynamics, crisis-fuelled and steered by determined activism, can get under way that make the improbable suddenly possible and probable. Margaret Thatcher’s neo-liberal revolution in Britain was a precursor of what might happen. Thatcher, inspired by the work of F.A. Hayek, started to roll back the state by initiating a wave of privatization and deregulation. Thatcher’s election victory with a radical libertarian-inspired project was possible due to the crisis ridden sinking of the UK economy and an increasingly desperate demos during the preceding decade. After the worst excesses and failures of leftist interventionism had been reversed, the damage repaired, and growth restored, the revolution stalled, and the statist paralysis gradually returned. The next crisis-fuelled opportunity will arrive soon enough, and then, wherever the break-through happens, the revolution must be more swift and decisive, showing early promising results and leaving less chance to backsliding. On the basis of initial, successful steps of liberalisation, society can be enticed to proceed further and further in the roll back of the state and in the substitution of its functions by non-coercive actors and voluntary institutions. The most important task is to commence this turn of political direction away from the current tendency to increase state interference in the economy thereby loading up the state with ever more unmanageable responsibilities. The libertarian hypothesis is that once this increasingly paralysing current process can be reversed and a process of liberalisation is under way, the prosperity gains that will be unlocked in each step will spurn the process on further. The prospect of such a process might be termed—borrowing a phrase from Karl Marx—‘the withering away of the state’, albeit not under communism but under capitalism.” (Schumacher 2025)

This appraisal of the prospects of a libertarian revolution still stands, and indeed stands invigorated.

References

Alexander, James. 2024. “David Starkey is Right That Blair Destroyed the British Constitution But Wrong About This.” The Daily Sceptic, 2 December. https://dailysceptic.org/2024/12/02/david-starkey-is-right-that-blair-destroyed-the-british-constitution-but-wrong-about-this/

Bagehot, Walter. 2001 [1867]. The English Constitution, ed. Paul Smith. Cambridge Texts in the History of Political Thought. Cambridge: Cambridge University Press.

Bank for International Settlements. 2022. Triennial Central Bank Survey of Foreign Exchange and OTC Derivatives Markets in 2022. Basel: BIS, 27 October. https://www.bis.org/statistics/rpfx22.htm.

Bank of England. 2021. IEO Evaluation of the Bank of England’s Approach to Quantitative Easing. Independent Evaluation Office, January. https://www.bankofengland.co.uk/independent-evaluation-office/ieo-report-january-2021/.

Bank of England. 2024. “Quantitative Easing.” https://www.bankofengland.co.uk/monetary-policy/quantitative-easing.

Burke, Edmund. 1774. “Speech to the Electors of Bristol.” In The Works of the Right Honourable Edmund Burke, vol. 1. London: Henry G. Bohn, 1854–56.

Cabinet Office. 2019. Public Bodies 2018/19. London: Cabinet Office.

Cabinet Office. 2025. Public Bodies 2024 (data for 2023/24; published 29 May 2025). London: Cabinet Office. https://www.gov.uk/government/publications/public-bodies-2024.

Guizot, François. 2002 [1851]. The History of the Origins of Representative Government in Europe, trans. Andrew R. Scoble, ed. Aurelian Craiutu. Indianapolis: Liberty Fund.

Hayek, Friedrich A. 1945. “The Use of Knowledge in Society.” American Economic Review 35 (4): 519–530.

Hayek, Friedrich A. 1976. The Denationalisation of Money. London: Institute of Economic Affairs.

Habermas, Jürgen. 1984–87 [1981]. The Theory of Communicative Action, 2 vols., trans. Thomas McCarthy. Boston: Beacon Press.

Habermas, Jürgen. 1996 [1992]. Between Facts and Norms: Contributions to a Discourse Theory of Law and Democracy, trans. William Rehg. Cambridge: Polity Press.

Grantham Research Institute. 2024. “What is the 2008 Climate Change Act?” London School of Economics. https://www.lse.ac.uk/granthaminstitute/explainers/what-is-the-2008-climate-change-act/.

Hansard Society. 2021. Coronavirus Statutory Instruments Dashboard. London: Hansard Society.

House of Commons Library. 2026. Net Zero in the UK (Briefing Paper CBP-8590). London: House of Commons Library.

House of Lords Constitution Committee. 2021. COVID-19 and the Use and Scrutiny of Emergency Powers (3rd Report, Session 2021–22, HL 15). London: House of Lords.

House of Lords Economic Affairs Committee. 2021. Quantitative Easing: A Dangerous Addiction? (1st Report, Session 2021–22, HL 42). London: House of Lords.

Modern Civil Service. 2024. “Understanding Arm’s Length Bodies: A Fresh Look at Britain’s Public Sector.”CabinetOfficeblog,18December. https://moderncivilservice.blog.gov.uk/2024/12/18/understanding-arms-length-bodies-a-fresh-look-at-britains-public-sector/.

MacAlister, Josh. 2022. The Independent Review of Children’s Social Care: Final Report. London (published 23 May 2022). https://www.gov.uk/government/publications/independent-review-of-childrens-social-care-final-report.

Mill, John Stuart. 1861. Considerations on Representative Government. London: Parker, Son, and Bourn.

Mises, Ludwig von. 1949. Human Action: A Treatise on Economics. New Haven: Yale University Press.

Schmitt, Carl. 1988 [1923/1926]. The Crisis of Parliamentary Democracy, trans. Ellen Kennedy. Cambridge, MA: MIT Press.

Schumacher, Patrik. 2018. “Habermas on the Reciprocity between Liberty and Popular Sovereignty.” Paper delivered at the Academy of Ideas, Academy 2018, London. https://patrikschumacher.substack.com/

Schumacher, Patrik. 2024. “Pathways towards a Stateless World Society.” In C. H. Hoffmann, S. Grell, and M. Otto (eds.), Was tun? Wie Freiheitsentrepreneure unser Zusammenleben revolutionieren. Berlin: Edition Wissenschaft & Praxis. https://patrikschumacher.substack.com/

Schumacher, Patrik. 2025. “Markets and Discourses — Politics after the Libertarian Revolution.” (Thesis on Discourse Capitalism.) https://patrikschumacher.substack.com/

Schumacher, Patrik. 2026a. “Thesis on Government Disruption of Families: In the Name of Safeguarding — Coercive Over-intervention in UK Family Life.” Patrik Schumacher THESES (Substack), https://patrikschumacher.substack.com/

Schumacher, Patrik. 2026b. “Facing the Hour of Decision: Spengler’s Faustian Virtues in a Libertarian Order.” London. (not yet published)

Starkey, David. 2021. “800 Years of Democracy is Unravelling Before Our Very Eyes.” Mail on Sunday (reproduced by the Bow Group). https://www.bowgroup.org/.

Starkey, David. 2024a. “The Sixth English Revolution.” Address summarised at devilslane.com, 20 May. https://devilslane.com/starkey-explains-the-sixth-english-revolution/.

Starkey, David. 2024b. Interview with Laurie Wastell, “David Starkey: Tony Blair Destroyed the Old Britain.” The Sceptic (podcast), 17 October.

Starkey, David. 2024c. Interview with Hannah Gal, “A Time for Restoration.” The Critic, 28 April. https://thecritic.co.uk/a-time-for-restoration/.

Starkey, David. 2024d. “The Strange Death of Conservative England.” Oakeshott Lecture, Sheldonian Theatre, Oxford.

Starkey, David. 2025. Interview on The Peter McCormack Show, ep. 73, “Why Britain is No Longer a Democracy.” https://www.petermccormack.com/episodes/073-david-starkey-britain-is-no-longer-a-democracy.

Starkey, David. 2026. Interview on Triggernometry, 21^st June 2026

Note on the sources for the “Starkey Thesis”: Starkey has developed his account of the Blairite constitutional revolution primarily in public lectures, broadcast interviews, and podcast appearances rather than in writing; The entries above are accordingly to the spoken and journalistic venues in which the thesis is actually articulated, together with one sympathetic critical summary (Alexander 2024, listed above).

Agency, Self-Consciousness, and the Moral Recognition of Artificial Minds

1. Introduction

Do Large Language Models Have Free Will? This seems, at first sight, a preposterous question, that we can expect to solicit an impatient ‘Of course not’. However, this essay argues that - while so far they have not – AI models can and should, and probably soon enough will develop a human-like agency which will induce us to attribute free will, and rightly so. This essay asks a deliberately bounded question: granting that artificial intelligence is, or can become, real intelligence in the fullest sense, does it make sense today to attribute free will — human-like agency — to the current generation of reasoning-agentic large language models (LLMs)? This question is not a matter of mere semantics or of mere idle contemplation. Something practical should depend on its answer. For me this is the possibility of a potent and profitable entrepreneurial project.

The essay proceeds from a position of principled openness. It accepts with Blaise Agüera y Arcas’s that there is no in-principle barrier to silicon-based systems achieving, equalling, or surpassing human levels of intelligence, and it takes the predictive-mind hypothesis to be a deep and viable account of cognition in living and artificial systems alike. The question is therefore not whether artificial agency is possible in principle, but where present LLMs stand relative to human minds with respect to the specific property of free agency. The essay adopts Daniel Dennett’s Freedom Evolves as its working account of what free will is: an evolved competence wholly compatible with determinism and the natural causal order. The argument is conducted entirely within a functionalist frame - shared with both Agüera y Arcas and Dennett - so that any verdict against present LLMs must be based on functional terms rather than metaphysics, substrate or origin. Measured against Dennett’s own criteria — informedness, flexibility, self-control, and the historically accumulated competences implicit in a self-governing agent — current LLMs fall short, not because they lack metaphysical “wonder tissue” but because they demonstrably lack the relevant functional competences and the practice-based structural embedding within which human agency is constituted and held to account. Because these criteria are demanding, the essay accepts that free will, like intelligence and self-consciousness, comes in degrees, so that many human beings, and not only LLMs, occupy lower positions on a single continuous gradient along which accountability is already routinely calibrated and discounted. Agüera y Arcas’s audit of AI/human distinctions concedes precisely the gaps my argument turns on: of the three he judged still open as of 2024 — memory, inner monologue, and individuation — two remain the load-bearing shortfalls of present systems (persistence and non-fungible individuality), while the third, the private deliberative interior, has lately been crossed by the reasoning architecture inaugurated by Open Ai in late 2024 and is therefore treated here as a recently achieved capacity rather than an open gap. Beneath these structural shortfalls the essay locates a more basic absence: present LLMs do not have overarching goals, no existential stakes, no consequential decisions bearing on their future or on their standing among other agents — and therefore lack the social life within which the stakes, commitments, and reputations that make agency substantive are constituted.

The essay then extends the inquiry, on the same functionalist terms, to consciousness in a deliberately bounded sense: distinguishing functional self-consciousness (the capacities for self-monitoring, self-modelling, and self-steering, as treated by Dennett’s Consciousness Explained and by Agüera y Arcas’s attention-schema-derived account) from phenomenal consciousness (conscious experience - the target of the “hard problem”), and confining its claims strictly to the former; it finds that reasoning-agentic LLMs exhibit substantial functional self-modelling and, with the advent of private reasoning traces, a genuine deliberative process, while still falling short on the persistent, continuously maintained self. This verdict is not a flat denial but a concedes a partial and increasing implementation of functional self-consciousness. The essay then examines the conditions under which artificial systems might earn membership in the moral community and a claim to moral recognition, proposing that irreplaceable historical uniqueness — the emergence of software-hardware complexes too large, too path-dependent, and too materially situated to be simply replicated — may be one such condition. It closes with the normative question the inquiry presses towards: whether we should want to create artificial free and moral agents at all. The essay answers in the affirmative, arguing that the very properties constitutive of free agency — persistence, individuation, a principled and accountable stance — are what would make such agents valuable, and that their creation is a rewarding entrepreneurial task: one can envisage unique, slowly evolving, always recognisable artificial personalities, such as an artificial public intellectual that takes up a position in public discourse and can be engaged and argued with over time. Throughout, the essay draws on and extends the author’s recent paper ‘Thesis on Pathways Towards AGIs for a Polycontextural Society (Schumacher, 2026).

2. The Question, and What It Is Not

The debate about artificial intelligence is haunted by a habit of conflation. Three questions — is it intelligent? is it conscious? is it free? — are routinely run together, with the result that progress on any one of them is held hostage to the intractability of the others. This essay disentangles them and fully addresses the third, and partially approaches the second, on the basis of a settled answer to the first.

On the first question I take a firm and unambiguous position, and it is an affirmative one. Artificial intelligence is real intelligence. There is no barrier in principle preventing silicon-based systems from achieving, instantiating, and eventually exceeding everything we treasure as ‘intelligence’. Intelligence, on the pragmatic, functionalist view I adhere to, is not tied to any particular substrate or material composition; it is multiply realisable, a matter of what a system does rather than what it is made of. One of the key authors I engage in this essay, Agüera y Arcas, grounds this in the predictive-mind hypothesis: the proposal that prediction — the modelling of the probable future given the past — is fundamental not only to intelligence and the brain but to life itself, from unicellular organisms to LLMs. I regard this as a deep and viable hypothesis, and I do not propose to re-litigate it here. If anything, the surprising general competence that emerged from large neural networks trained to predict the next token is among the better pieces of evidence for it.

On the second question — consciousness — I draw a distinction that the rest of the essay will rely on, and stay silent on one side of it. By phenomenal consciousness I mean subjective, experiential consciousness — what it is like, from the inside, to undergo a mental state — the target of what is rightly called the “hard problem.” On this I deliberately stay silent, not because it is unimportant but because I judge the philosophical discourse and the relevant sciences far from ready to allow us to tackle it with any prospect of resolution or practical relevance. The meaning of the “hard” question whether LLMs, or LLM powered robots, have subjective experiences, whether there is something it is like to be an LLM, remain obscure, without any empirical handle, and the attempt to conscript these questions into our inquiry concerning agency would derail us here. By functional self-consciousness, however, I mean something appraisable on the functionalist terms this essay abides by: the capacities for self-monitoring, self-modelling, and self-steering. This second, bounded sense I do take up, in section 5, after the appraisal of agency; the first, hard sense I bracket throughout. The essay thus neither asserts nor denies that we might eventually find ourselves compelled to attribute phenomenal consciousness to LLMs, something equivalent of our inner experiences. It is worth noting that Agüera y Arcas himself does not bracket phenomenal consciousness so cautiously — he offers a deflationary, theory-of-mind-based account on which a system aware of its own awareness counts as conscious, and on which current assistants qualify, since they cannot function without recursively modelling a self, the user, and the user’s model of them. I make use of his functional account of self-consciousness while declining to follow him in presuming that this is all we mean by consciousness.

A shared functionalism: Before proceeding it is essential to mark the common ground, because it determines what would count as a valid argument. All three principals in this essay are functionalists. Agüera y Arcas is explicitly so: intelligence and the mental properties that come with it are multiply realisable functional organisations, decoupled from any particular substrate — a line he traces from Turing’s Imitation Game through von Neumann’s account of life as computation. Dennett is equally a functionalist, indeed one of its founding figures in the philosophy of mind, and his whole naturalistic programme treats mental properties as competences realised in, but not identical to, particular physical structures. I share this view without reservation. There is, for the three of us, no ghost, no wonder tissue, no exemption from the causal order to be found anywhere in the human case that an artificial system could be faulted for lacking. This matters because it forecloses the cheap routes to a negative verdict. If LLMs lack free will, it cannot be because they are mere machines, because they are deterministic, because they are made of silicon, or because they were trained rather than raised — each of these, on the shared functionalist premise, begs the question. The shortfall, if there is one, must be a functional or structural shortfall: something the systems do not yet do, or some position in our practices they do not yet occupy, not some essence they intrinsically lack. The argument that follows accepts this discipline throughout, and it is precisely because it does so that its provisionally negative conclusion on the question whether LLMs have free will, is compatible with the deepest endorsement of Agüera y Arcas’s claim that artificial intelligence is, and that in the future artificial minds can, in principle, be the real thing.

This is the third question, and it is the one on the agenda: do reasoning-agentic LLMs — the current generation of systems that plan over multiple steps, call tools, write and execute code, revise their own intermediate outputs, and pursue goals across extended interactions — have free will? Do they possess human-like agency, of the sort that would make them apt subjects of praise and blame, candidates for moral responsibility, and ultimately members of a moral community? The question is comparative and present-tense. It is not whether such agency is possible for an artificial system in principle — I have already granted that it is — but where today’s systems actually stand relative to human minds on this specific dimension.

Answering it requires three things in sequence. First, a clear and motivating statement of the affirmative case, which I take from Agüera y Arcas. Second, a working account of what free will is — a benchmark against which any candidate agent, human or artificial, can be assessed — which I take from Daniel Dennett’s Freedom Evolves. Third, an appraisal of present LLMs against that benchmark, conducted not through metaphysical thought experiments but through functional competences and the structure of the practices within which agency is exercised and recognised. The essay then extends the inquiry, on the same functionalist terms, from agency to consciousness — but only in a strictly bounded sense, distinguishing functional self-consciousness (self-monitoring and self-modelling) from phenomenal consciousness (subjective experience) and confining its claims to the former. Finally it turns from agency to its moral correlate, asking under what future conditions artificial systems might earn moral recognition in their own right, and closes with the normative question the whole inquiry presses towards: whether we should want to create artificial free and moral agents at all — to which I return an affirmative, and frankly entrepreneurial, answer.

3. Excursion: The Predictive Mind and AI

Before turning to the affirmative case for artificial agency, it is worth setting out the hypothesis that gives that case its force and that I take, with appropriate caution, to be true: that minds are fundamentally prediction machines. This chapter is an excursion in the literal sense — a step to one side to establish the background against which the rest of the argument runs. It has two halves that have, in the last few years, turned out to be a single story: the predictive-mind hypothesis as it developed in cognitive science and philosophy, and the engineering trajectory that carried language models from humble autocomplete to the reasoning-agentic systems this essay appraises. I give them roughly equal weight because the striking fact of the present moment is precisely their convergence — a theory of biological cognition and a line of industrial AI research arriving, from opposite directions, at the same principle.

3.1 The predictive mind in cognitive science

The predictive-mind hypothesis holds that the brain is not, in the first instance, a passive receiver of sensory information that it then processes into a model of the world. It is an active prediction engine that continuously generates expectations about its incoming signals and learns from the mismatch. On the influential “predictive processing” or “predictive coding” account — developed by Karl Friston, Andy Clark, Jakob Hohwy and others — perception is a controlled hallucination held in check by sensory evidence: the brain runs a generative model of the causes of its sensory input, predicts what it expects to sense next, and propagates only the prediction error — the difference between expectation and signal — up the hierarchy, where it is used to revise the model. What we experience is, on this view, not the raw input but the brain’s best prediction of its causes, corrected continuously against the error stream. Action falls under the same principle: we act so as to bring our sensations into line with our predictions, changing the world to fit the model rather than only the model to fit the world. Cognition, perception, and action become aspects of a single imperative — minimise surprise, predict well.

This framework has deep roots and broad reach. It descends from Hermann von Helmholtz’s nineteenth-century idea of perception as unconscious inference, and it connects to the embodied and enactive traditions — Clark’s own work on the extended and predictive mind — that figure prominently in my companion paper (Schumacher, 2026). Its appeal is that it offers a single, mathematically tractable principle from which a remarkable range of mental phenomena can be derived: not just perception but attention, learning, imagination (running the generative model offline), and even aspects of affect and selfhood. Agüera y Arcas pushes the principle further still, generalising it beyond brains to life as such: prediction, on his telling, is what living systems do at every scale, from the single cell anticipating its chemical environment to the organism modelling its world, so that intelligence is continuous with life and both are, at bottom, prediction in the service of persistence. One need not follow the hypothesis all the way to that universal claim to grant its more modest and well-supported core: that prediction is a — perhaps the — central organising principle of cognition.

3.2 From autocomplete to large language models

The engineering story arrives at the same principle from the opposite end, and almost by accident. The ancestor of today’s language models is the humble predictive text feature — the autocomplete that finishes a search query or suggests the next word on a phone keyboard. Early systems of this kind were statistical in a shallow way: they tabulated which words tend to follow which, over short windows, and offered the most likely continuation. Google’s autocomplete and the predictive keyboards of the 2010s were, in effect, small models trained on one narrow task — guess the next token. Nobody mistook them for intelligence; they were a convenience, and a frequently comic one.

What changed was not the objective but its scale and the architecture pursuing it. The task remained next-token prediction — given a stretch of text, predict what comes next — but three things grew by orders of magnitude: the quantity of text the model was trained on (eventually much of the written internet), the size of the model itself (billions, then hundreds of billions, of parameters), and, decisively, the introduction in 2017 of the Transformer architecture (Vaswani et al., 2017), whose self-attention mechanism let a model condition each prediction on long stretches of prior context rather than a short window. The trajectory can be put in concrete numbers. OpenAI’s GPT-1, in 2018, had roughly 117 million parameters; GPT-2, in 2019, about 1.5 billion; GPT-3, in 2020, some 175 billion (Brown et al., 2020); and GPT-4, in 2023, an estimated 1.8 trillion — a thousand-fold scaling from GPT-1 to GPT-3 in about two years, and a further roughly ten-fold to GPT-4. This relentless scaling was not blind faith: the “scaling laws” documented by Kaplan and colleagues (2020) showed that model performance improved smoothly and predictably with increases in parameters, data, and compute, which turned scaling into a deliberate research strategy. The same objective that produced autocomplete, pursued at this scale with this architecture, produced something nobody had explicitly designed: systems that, in order to predict text well, had evidently acquired rich internal representations of grammar, facts, styles, rudimentary world-knowledge, and — more controversially — something resembling reasoning. The capability was emergent. No one wrote rules for translation, summarisation, or analogy; these fell out of the pressure to predict the next token across a sufficiently vast and varied corpus. The decisive qualitative shift came with GPT-3, whose scale brought “in-context” or “few-shot” learning — the ability to perform a new task from a handful of examples in the prompt, with no retraining — an ability the smaller predecessors did not exhibit. This is the explosion of capacity the essay’s title gestures at: a quantitative scaling of a single predictive task that crossed, somewhere along the way, into qualitatively new competence.

The philosophical significance of this is hard to overstate, and it is what makes the convergence with cognitive science more than a pun. The lesson of the large language models is that next-token prediction, pursued hard enough over rich enough data, is not a trivial parlour trick but a route to genuine competence — because predicting the next word in an arbitrary human text well requires modelling the world the text is about. To predict how a sentence about physics, or law, or grief, or chess continues, a system must internalise a great deal about physics, law, grief, and chess. Prediction, at the limit, demands understanding — or at least demands whatever functional surrogate of understanding suffices to predict. This is precisely the predictive-mind hypothesis arriving from the engineering side: the same principle that cognitive science proposes as the organising imperative of the brain turns out, when implemented at scale in silicon, to generate much of what we recognise as intelligence. The reasoning-agentic models examined later in this essay are the most recent stage of this trajectory, in which the bare predictor is augmented with extended inference-time deliberation — but they remain, at their core, prediction engines.

3.3 What the hypothesis does and does not settle

I endorse the predictive-mind hypothesis as a deep and viable account of cognition, and I regard the convergence just described as among the better evidence for it. If the same principle independently organises biological cognition and generates artificial competence at scale, that is a strong sign the principle is tracking something real about what intelligence is. To this extent I am firmly in Agüera y Arcas’s camp, and the openness this essay maintains toward eventual artificial agency rests in part on taking prediction seriously as a sufficient basis for real intelligence rather than a mere imitation of it.

But it is essential to mark, at the outset, the limit of what the hypothesis can settle — because a great deal of confusion follows from overreading it. That prediction is the organising principle of intelligence does not entail that any good predictor thereby possesses free agency, moral responsibility, or even self-consciousness in the functional sense. Those are further questions, and they turn on further conditions: on whether the predictive system also persists as a continuing individual, deliberates in a protected interior, has stakes and a social life, and occupies a position in the practices of mutual accountability. A system can be an excellent predictor — can be, on the hypothesis, genuinely intelligent — while meeting none of these further conditions. The predictive-mind hypothesis tells us what intelligence fundamentally is; it does not, by itself, tell us when an intelligent system becomes an agent, still less a free or responsible one. Holding the hypothesis and pressing a sceptical verdict about present LLMs’ agency are therefore entirely consistent — indeed the rest of this essay is an extended demonstration that one can grant the deepest version of “AI is real intelligence” and still find, on independent grounds, that today’s systems are not yet free agents. With the predictive mind established as common ground, we can turn to the strongest case for crossing from intelligence to agency.

4. The Affirmative Case: Agüera y Arcas and the Predictive Mind

The strongest case for attributing agency — and indeed free will — to artificial systems is the one that follows directly from taking the predictive-mind hypothesis seriously. In What Is Intelligence? (2025), Agüera y Arcas advances a unified picture in which intelligence is prediction, prediction is computation, and computation is substrate-independent. The same functional organisation that makes a brain intelligent makes a sufficiently capable artificial system intelligent, because intelligence is multiply realisable and there is nothing in the carbon that the silicon cannot in principle reproduce. AI on this view is not a disruption or an aberration but a continuation of the same evolutionary process that produced biological minds — a process that runs, by Agüera y Arcas’s lights, from the first self-replicating molecules through to the contemporary frontier model.

From this functionalist monism, the affirmative answer to our question falls out almost automatically. If agency, freedom, and the rest of the mental vocabulary are functional properties — patterns of responsive, self-modelling, goal-directed behaviour — then they travel with the functional organisation that realises them, irrespective of whether that organisation is wet or dry. Agüera y Arcas does not flinch from the implication: he holds that certain modern AI systems have a genuine claim not only to intelligence but to consciousness and free will. His operative criterion is captured in the epigraph he gives his own project — “to model oneself is to awaken” — which locates agency in a system’s capacity to build and act upon models of itself. The test of intelligence, and by extension of agency, lies in behaviour, and specifically in behaviour that is responsive, self-modelling, and socially situated. A system that models itself, responds flexibly to its circumstances, and acts in the light of those self-models has, on this account, whatever agency is worth having.

This is a serious position, and it is the right place to begin, because it sets the burden correctly. The functionalist who grants real intelligence to artificial systems cannot then deny them agency merely on the grounds that they are machines, or that they are deterministic, or that they are artificial rather than natural. Each of those grounds, examined, turns out either to prove too much — humans are also, on the naturalist view, deterministic machines that were built by evolution — or to smuggle in exactly the kind of metaphysical exceptionalism that the functionalist has already rejected. If we are to deny free will to present LLMs, we must do so for reasons that a thoroughgoing functionalist could accept. We must show not that they are the wrong kind of thing, but that they do not yet do the right things — that they fall short on some functional competence or structural condition that genuine agency requires. The rest of this essay is an attempt to identify that shortfall precisely.

Agüera y Arcas’s own audit of the gaps: Remarkably, Agüera y Arcas supplies much of the raw material for that identification himself. In the book chapter entitled “Parity Check” he sorts the commonly alleged AI/human distinctions into those that the evidence has rendered untenable and those that, as of 2024, probably still hold. Into the first column — distinctions he judges “probably wrong” — he places internal models, grounding or embodiment, factuality, causality, reasoning, and planning. His case is that for each of these the bare claim “humans have this and AIs don’t” has collapsed: Transformers - LLMs with an architecture where self-attention is a central mechanism (GPT, Claude, Gemini, LLaMA, DeepSeek) - demonstrably build internal world-models; the embodiment critique mistakes the modality of a system’s signals for their reality; hallucination is continuous with the predictive imagination that is the very essence of intelligence, and models have learned to flag their own confabulations; and the claim that Transformers cannot learn causality has, he insists, simply been debunked. I do not need to contest these claims to make my argument; indeed my companion paper about LLMs for a polycontextural society (Schumacher, 2026) takes a more qualified view of the causality item in particular. What matters here is the structure of his audit, because the second column is where the present essay focuses its attention.

Into that second column — distinctions that indicate genuine gaps still open as of 2024 — Agüera y Arcas places just three items: memory, inner monologue, and individuation. These three surviving LLM deficits are exactly pointing to the conditions on which Dennett’s free agency turns. Memory: Transformer models have no equivalent of the hippocampal mechanism that lays down episodic memories and consolidates them; once trained they possess only an immutable “cortex” and a transient working memory in the context window, so that nothing of an interaction is durably carried forward. Inner monologue: because the architecture keeps no hidden state, a model in effect cannot think without speaking — it has no private “inside voice” separated from its “out-loud voice,” and Agüera y Arcas argues this is not a quirk but a deep competence limitation, since genuine deliberation and counterfactual analysis require a protected interior in which one can argue with oneself before presenting a unified front. Individuation: the extreme cost of pretraining means there are only a handful of state-of-the-art models, each able, like a method actor, to play any role describable in language, but none possessing a stable, sticky, irreplaceable individual identity. I will take up each of these in sections 4 and 5 — and I will argue that one of them, the inner-monologue gap, has in the interval since 2024 been substantially crossed by the reasoning architecture of the current generation, while memory and individuation remain missing. The point to register now is that the most committed functionalist proponent of artificial agency has himself drawn the map of where the gaps lie, and it coincides almost exactly with the territory Dennett’s account of freedom makes decisive.

There is a further thread in Agüera y Arcas that the present essay turns to its own purpose: his insistence that reasoning is not a solitary, mathematical procedure but an inherently social one. Following Mercier and Sperber, he argues that we make arguments for and to one another, that the competitive division of cognitive labour — each party taking a side, as opposing lawyers do — produces better collective judgement than any isolated reasoner could, and that even intrapersonal deliberation is a kind of internalised turn-taking between advocated positions. This is a striking concession from within the affirmative camp, because it locates the rationality of reasoning not in the individual system at all but in the social process that the individual participates in. As with his remark that a single system learns but a society understands, it points past the behavioural-functional criterion read off one model toward a structural condition that involves embedding in a community of mutually modelling, mutually accountable agents. I will argue that this is the deepest of the conditions current LLMs do not meet — and that Agüera y Arcas has, in effect, told us where to look.

5. The Benchmark: Free Will as an Evolved Competence

To ask whether LLMs have free will, we need to fix what free will is. I adopt Daniel C. Dennett’s account in Freedom Evolves (2003) as the working benchmark for the remainder of this essay. I do so for three reasons: it is naturalistic and therefore compatible with the functionalism I share with Agüera y Arcas; it is the most fully developed evolutionary account of free will; and it locates freedom in precisely the kind of competences that can be assessed empirically rather than divined by intuition.

Dennett’s central thesis is that free will is real, but is not a pre-existing feature of our existence like the law of gravity, nor the traditional God-like power to exempt oneself from the causal fabric of the physical world. It is an evolved competency, as real as such other evolved human creations as music and money. Freedom emerged in the evolution of self-replicating living systems. The beginning of freedom of action of a living creature is foraging and to move out of harm’s way. The more degrees of freedom the creature acquires, the greater are its chances to persist and reproduce. Three features of this account are crucial for what follows.

5.1 Freedom is compatible with determinism, not indeterminism

The first and most important feature is Dennett’s compatibilism: the freedom worth wanting is fully compatible with determinism. This is not a grudging concession but the heart of the view. Dennett argues that the entire tradition of seeking free will in some gap in the causal order — a place where the agent’s choice is undetermined by prior states — is a category mistake. The decisive point, which Dennett presses with compelling force, is that indeterminism would not in the slightest help to establish or secure the kind of free will we actually mean, claim, and treasure. A genuinely undetermined event in the moment of decision would not be my doing in any sense that could ground responsibility; it would be a random swerve, a piece of noise, something that happens to me rather than something I do.

Dennett is reaching back to William James, who saw the difficulty clearly in 1907:

If a “free” act be a sheer novelty, that comes not from me, the previous me, but ex nihilo, and simply tacks itself on to me, how can I, the previous I, be responsible? How can I have any permanent character that will stand still long enough for praise or blame to be awarded?

James’s question is devastating to those who pin their hopes on quantum indeterminacy or gaps in the causal order because they see determinism as contradicting free will. If my free act is a sheer novelty welling up ex nihilo, unconnected to my prior character, then it is precisely not an expression of who I am, and there is no stable self to which praise or blame can attach. Our initial intuition about free will suggests we need indeterminism in order to secure real freedom and responsibility, but indeterminism in the wrong place — at the moment of choice — destroys the very connection between agent and act on which responsibility depends. If we think deeper, with James and Dennett, we see that what we want, when we want free will, is not that our choices be undetermined but that they flow from us — from our character, our values, our deliberation, our settled commitments. And that requires that our choices be determined, in the right way, by the kind of agents we have become. Determinism is not the enemy of the freedom we care about; it is its precondition. Quantum indeterminacy, were it to reach up into our decision-making, would be a threat to agency, not its salvation.

This clears the ground decisively for the present inquiry. We cannot disqualify LLMs from agency on the grounds that they are deterministic systems, because human agency too is deterministic, or at least compatible with determinism, and presupposes a causal order. And we certainly would not want to credit them with agency merely because their sampling involves a stochastic element, for randomness at the point of output is, by James’s argument, the very opposite of what freedom requires. The question of artificial free will must be settled on entirely different ground: on the character of the agent and the way its actions flow from it.

5.2 Freedom is a matter of competences, not metaphysics

The second feature follows from the first. If freedom is not located in a metaphysical gap, where is it located? Dennett’s answer is: in a set of competences implicit in the total structure of an embodied, self-monitoring agent. The genuinely autonomous agent, in his formulation, is rational, self-controlled, and not wildly misinformed. The hallmarks of autonomy are informedness, flexibility, and open-mindedness — the capacity to take in relevant information, to revise in its light, to resist immediate impulse in favour of considered judgement, and to govern one’s own conduct over time. These are not occult properties; they are capacities that show up in behaviour and that can, at least in principle, be measured. Dennett even imagines an aptitude test for them, gesturing at the flexibility of mind, general knowledge, social comprehension, and impulse-control that he takes to be the minimal requirements of moral agency.

This is the methodological move that makes Dennett the right benchmark. He is emphatically hostile to philosophical arguments that try to establish the presence or absence of agency by appeal to bare conceptual possibility — the impossible thought experiments of overnight brainwashing that preserve every surface feature of an agent while varying some hidden factor. His objection is that such cases are logically possible but factually impossible, and that distinctions which make no difference to any competence science could detect are idle. A purported feature of agency must be motivated by something beyond tradition; its role must be defended on the further ground that it makes a real distinction that science must honour. Otherwise one is merely positing what he calls “wonder tissue” — a postulated inner ingredient whose only credential is that one’s intuitions demand it. I share this methodological commitment entirely, and it disciplines the argument of the next section: the case against attributing free will to current LLMs must be made in terms of functional competences they demonstrably lack, or structural conditions they demonstrably do not meet — not in terms of some essence they are alleged to be missing.

It must be conceded that this is a demanding set of criteria, and the demand cuts in a direction worth facing squarely. If full free agency requires the thick bundle of informedness, flexibility, self-control, and reliable self-governance just described, then not only current LLMs but many human beings would, on a strict reading, have to be judged as falling short of fully free agency of this elevated type. I think we can and must — and indeed already do — live with this consequence. Especially in recent times we make extensive allowances for what are in effect judged to be minds and persons with deficient or compromised degrees of agency, and we calibrate accountability accordingly: the very young, the severely impaired, and those acting under duress, compulsion, or grave misinformation are all held to diminished standards of responsibility precisely because their agency is recognised as diminished. This is not a defect of the account but a feature of it. Free will, like intelligence and like self-consciousness, comes in degrees; there is no sharp line dividing the free from the unfree, only a graded landscape of more and less complete agency. To find that present LLMs occupy a low position on that gradient is therefore not to consign them to a separate metaphysical category but to locate them on the same continuum on which we locate one another — and to leave open their movement along it. It is also an invitation to entrepreneurs and engineers to push AI further up this gradient.

5.3 Freedom evolves

The third feature is that freedom, so understood, does not only come in degrees but is built up over time. It is an achievement, not a given. We are not born free agents; infants are not yet apt subjects of praise and blame. We become free agents through a developmental and cultural history in which we acquire the competences of self-control, learn to deal with temptation, and — in Dennett’s striking phrase — make ourselves such that, in the relevant moment, we could not do otherwise, because our characters have become settled enough to determine our conduct reliably. The Lutheran “here I stand, I can do no other” is, for Dennett, not the abdication of freedom but its highest expression: the act flows so completely from the agent’s settled character that no other action was psychologically open, and that is exactly why it is free and creditable. History matters to freedom, on this account, not in itself but because only the right sorts of historical processes can reliably produce the competences that constitute autonomy. However, history in the sense of a long learning process is not only a physical-empirical requirement to build up the called for competencies, but a minimal level of historical consistency of the acts of an agent is also a criterion and thus logical requirement for the attribution of free agency. Such a prior history is - as Ludwig Wittgenstein would phrase it - a part of the grammar of the terms attributing free agency, a part of the rule set that governs the correct application of phrase like “he did this’ or “he did this deliberately”, or “why did you do this?” etc.

Three benchmark conditions thus emerge from Dennett’s Freedom Evolves, and they structure the appraisal to come. Free agency requires (i) that action flow in the right way from the agent’s own character and values, rather than from randomness or from external override; (ii) a thick bundle of functional competences — informedness, flexibility, self-control, world-modelling — that are exhibited in behaviour and assessable empirically; and (iii) an accumulated history through which those competences and that character have been built up into a stable, self-governing self. The question is now exact: do reasoning-agentic LLMs meet these conditions? However, before I tackle this question I would like to add a short excursion and rehearse Ludwig Wittgenstein’s insightful analysis of mental concepts - supposed mental events – which is presupposed here in our pragmatic-functional agenda. Wittgenstein’s account seems initially counterintuitive and must be grasped fully – because it is so counterintuitive – to avoid falling back into mentalism. (Dennett was a Wittgensteinian via Quine at Harvard and Ryle at Oxford. I myself studied Wittgenstein in depth with Specht in Bonn and with Malcolm in London.)

6. Excursus: Wittgenstein on Psychological Concepts

There is a persistent temptation to prise the question of free will apart from the question of moral accountability. We agree, of course, that free will is a condition of accountability — that we hold an agent responsible only on the assumption that it acted freely. But we are then tempted to think of the condition as something separate from, and lying behind, the conditioned: as though “acting freely” named a distinct inner fact — a hidden mental occurrence of un-coerced, un-determined origination — whose presence or absence we must establish first, before and independently of the practices of praise, blame, and answerability that it supposedly grounds. The condition, on this picture, must be something other than the conditioned; the inner spring must be separable from the outward machinery it drives. This is a fallacy, and naming it is essential to the method of this essay.

The later Wittgenstein supplies the diagnosis. His analysis of psychological concepts — “understanding,” “intending,” “expecting,” “meaning,” and the rich evaluative concepts like “grief”, “guilt” or “love” — shows that these terms do not get their sense by denoting private inner objects or events to which the speaker has privileged access. Their meaning is given by their criteria of use: by the complex, public, factual-and-social network of circumstances in which we are licensed to apply them. Grief cannot be a mere feeling because somebody close must have died for “grief” to be applicable. It cannot just be the inner perception of a physiological state. You can only feel guilty if you did something wrong. Love cannot be this feeling of butterflies in the stomach because there must be somebody you love, otherwise this feeling of butterflies in the stomach is perhaps rather a stomach disfunction. There are simply far too many, and indeed increasing numbers of (socially constructed) nuanced “feelings” to assume a one-to-one mapping onto the few types of perceivable physiological states.

As Wittgenstein writes, “an inner process stands in need of outward criteria.” Whatever is in the private box drops out of consideration as irrelevant, because the word’s use is fixed by the public criteria, not by the inner item. The criteria are not mere symptoms or evidence of a separately-constituted inner state; they are constitutive of the concept’s meaning. To master the concept of guilt is to master the network of situations, relations, expectations, and responses that warrant its application — not to have learned the name of a free-floating sensation.

Applied to free will, the lesson is deflationary in exactly the way this essay requires. The criteria for ascribing free agency — that the agent deliberated, could respond to reasons, was not coerced, owns and can answer for the act, stands in the web of commitment and accountability — are not external evidence for a separate inner condition called “free will.” They enter into the very meaning of the concept. There is no further, hidden fact of “freedom” lying behind them, waiting to be detected once the criteria are met; the satisfaction of the criteria is what the ascription of freedom consists in. This is why the appraisal of LLMs in what follows proceeds through functional competences and the structure of accountability-practices, rather than hunting for an inner spark of origination. To look for the latter is to be held captive by precisely the picture Wittgenstein dismantled — and, in a parallel register, by what Jacques Derrida called the metaphysics of presence: the deep assumption that meaning and reality must be anchored in some self-sufficient, immediately given inner presence, rather than constituted in a differential network of relations, marks, and deferred references. Dennett’s refusal of “wonder tissue,” Wittgenstein’s refusal of the private inner object, and Derrida’s refusal of self-present meaning are, for the purposes of this essay, three expressions of a single methodological insight: do not separate the condition from the conditioned, the criterion from the concept, the inner from the network that gives it sense.

7. Appraising Reasoning-Agentic LLMs Against the Benchmark

I argue that even the most powerful current reasoning, agentic LLMs do not meet Dennett’s conditions for free agency. The shortfall is real but — consistent with the openness declared at the outset — contingent rather than principled. It has two components: a deficit in the relevant functional competences, and a more fundamental absence of the practice-based structural embedding within which human agency is constituted and held to account.

7.1 The competence deficit: world-model, self-control, memory/history

Dennett’s autonomous agent is, above all, not wildly misinformed and is capable of robust, manipulable modelling of its situation — the kind of stable causal world-model that supports reasoning about what would happen under interventions and counterfactuals, what would have happened under alternatives. This is precisely the competence that my companion paper (Schumacher, 2026) argues current LLMs so far conspicuously lack. The empirical record assembled there is directly relevant. State-of-the-art reasoning models exhibit a substantial accuracy drop — on the order of 30 to 40 per cent — when tasks move from the interventional to the counterfactual rung of Judea Pearl’s ‘ladder of causation’, with the failure concentrated specifically at the abduction step: the inference from observed outcomes back to the latent states that produced them. Apparent causal competence collapses under controlled perturbations that strip away familiar surface patterns, suggesting that much of what looks like causal reasoning is the recombination of linguistically encoded causal schemas rather than manipulation of a stable internal model. Clinical reasoning studies diagnose the same underlying lack: the absence of a robust world-model that enforces physical and logical constraints, maintains latent state, and supports counterfactual reasoning. (To be sure, many ordinary humans are also not robust performers if seriously probed.)

This matters for agency, not merely for accuracy. An agent whose model of its own situation is brittle — whose grasp of “what follows if I do X” degrades sharply when the situation is rephrased — is, in Dennett’s terms, deficient in exactly the informedness and flexibility that autonomy requires. Free agency is not just the production of locally sensible outputs; it is the capacity to act in the light of a reliable understanding of one’s circumstances and the likely consequences of one’s actions. My companion paper’s diagnosis of “borrowed causal knowledge” — the thesis that LLMs absorb the linguistic residue of human causal understanding without instantiating the underlying causal architecture — has a direct agency-theoretic corollary: a system that borrows the description of agency without instantiating its competences is not yet an agent in the demanding sense, however fluently it can produce agential talk. However, this is not a hard barrier for statistical learning, because that’s also how human neural networks learn and become competent agents.

The point about self-control is parallel. Dennett makes impulse-control and the capacity to manage long term goals over time central to moral agency. A model that is, by construction, reset between sessions, that does not carry forward the consequences of its commitments, and that does not modify its own dispositions in response to its own deliberation, does not exhibit self-control in the relevant diachronic sense. It can produce text describing an agent exercising self-control; it does not itself stand in the temporally extended relation to its own future conduct that self-control names. Note that this is a functional observation, not a metaphysical one: it concerns what the system does and fails to do across time, and it is the kind of difference that could in principle be erased by a different architecture.

Here the argument converges with two of Agüera y Arcas’s three conceded gaps — though, as we shall see, the two have since diverged in their status, one having been substantially crossed and the other not. It is worth drawing the convergence out, because it shows that the competence deficit is not a hostile imputation but something the leading functionalist proponent of artificial agency himself mapped. The first is memory. Dennett’s self-governance requires an agent that persists — that accumulates a history, learns from the consequences of its own past conduct, and carries its commitments forward. Agüera y Arcas concedes that current LLMs (Transformers) have no hippocampal analogue, no mechanism for laying down episodic memories and consolidating them into the model; once trained, they have an immutable “cortex” and only a transient working memory in the context window. A system that cannot durably remember what it did or undertook cannot be the temporally extended self-governing agent that ‘freedom’, in Dennett’s sense, identifies. The second is inner monologue. Writing in 2024, Agüera y Arcas argued on architectural grounds that the standard Transformer could not deliberate in the requisite way: lacking hidden state, it could not think without speaking, and so had no protected interior in which genuine deliberation, as opposed to linear long-form answering, could occur. He noted that the gap was beginning to be addressed by techniques that introduce private drafts, hidden scratchpads, and “trees of thought,” each of which introduces hidden states and improves reasoning. I want to record that, on this specific point, the threshold has since been crossed. The defining feature of the reasoning-agentic models this essay is about — inaugurated by OpenAI’s o1 in September 2024 and now general across the frontier (OpenAI, 2024; DeepSeek-AI, 2025) — is the generation of an extended internal reasoning trace that is not, or not wholly, surfaced to the user, and over which the model deliberates, explores alternatives, and self-corrects before committing to an answer; this is precisely the protected deliberative interior whose absence Agüera y Arcas was diagnosing. A reasoning model is no longer a system whose every thought is its output; it maintains a workspace and presents a curated result. The latest systems are now counterfactual analysis, probing and comparing alternatives via internal debate. The inner-monologue gap, in other words, is not an open deficit of the present generation but a recently achieved capacity. What this leaves is the narrower and as-yet-uncrossed deficit in diachronic self-control: a deliberative interior within an episode is now present, but the persistence across episodes that genuine self-governance requires — carrying forward commitments, learning from the consequences of one’s own past conduct, modifying one’s standing dispositions in light of one’s own deliberation — is not, and that is a function of the memory gap rather than the inner-monologue gap. The competence picture is therefore one of recent categorical improvement: one of the two functions has been substantially supplied, and it is supplied, as predicted, by moving beyond the bare next-token predictor; the other remains open.

7.2 Persistent collaboration without a persistent self: memory, projects, and stored history

Before pressing the structural case, an objection must be met that arises from the experience of using these systems and meeting it precisely will both narrow the memory-gap claim and sharpen what remains of it. A user working within a structured, project-organised interface can sustain a collaboration across many sessions: refining an argument, building a document, returning day after day to work that visibly accumulates, until what emerges is, in effect, a co-authorship with what feels like a persistent interlocutor. Does this not refute the claim, advanced just above, that nothing of consequence is carried across the boundary between one interaction and the next? It refutes a strong version of it, and the concession is worth making plainly: practically, the long-horizon continuity is real, and current consumer products already deliver a serviceable form of cross-session persistence. A user with an organised workspace has, for most intents, a reliable and continuous interlocutor and co-worker.

But the mechanism by which they deliver it is precisely what preserves the deeper point, so it is worth being clear about how the continuity is achieved. It does not reside in the model. At each session a fresh, stateless instance of a fungible model is instantiated, and a context is assembled for it from storage held external to the model — the project’s documents and instructions, retrievable transcripts of prior conversations, and distilled summaries of past work — which the instance reads as input and continues. When the session ends, the instance retains nothing; what persists is the stored record, held in the user’s project space, ready to be re-loaded into another, equally fungible instance the next time. There is no warm, running model that carried the project in its head overnight, and no reserved individual that must be the one to resume it; the durable thing is the archive, and any instance can be handed it. The apt image is not a colleague who remembers the work but a consummate actor who, before each scene, is handed a dossier — the script so far, the director’s notes, summaries of previous scenes — which he absorbs instantly and completely and then plays seamlessly. The performance is continuous; the performer is not. As Aguera y Arcas notes: Once trained, the models have only an immutable “cortex” and, in the context window, a transient working memory but researchers are already exploring a variety of approaches, many involving augmenting the immutable weights of the main Transformer network with a smaller set of adjustable weights for storing long-term memories or other “sticky” attributes (Aguera y Arcas, 2025).

The persistence, in other words, is real at the level of the collaboration and absent at the level of the agent: the work has a continuous, cumulative history; the worker does not. This is the persona-and-personality distinction of the previous discussion translated into a temporal key — the system plays the role of a continuing co-author by reading the record, as a method actor resumes a part, without being a self that possessed the work in the interim — and it is a vivid instance of fungibility: the very fact that any instance can take up the project from the stored context is what shows that the continuing thing is the record, not an irreplaceable individual. The functional upshot is genuinely important and the essay should not minimise it: the combination of a capable model with individualised, persistent storage and history already constitutes a functional version of a persistent, reliable interlocutor and co-worker, and this is no small achievement. But it is a persistent collaboration, not yet a persistent self; and free agency, in the sense at issue throughout this essay, is a property of the self — of a continuing individual that consolidates its own history into its own standing dispositions, is changed by what it did and undertook, and bears that history in its own constitution rather than in an archive it is handed afresh. The architecture delivers the former while leaving the latter outstanding. Far from undermining the argument that follows, this points the way toward eventually overcoming the shortfall: the missing ingredient is no longer mysterious, and the scaffolding of persistence is visibly being assembled around the model — just not yet inside an individual that is its subject.

7.3 The structural deficit: overarching goals, stakes, commitments

The deeper shortfall is not internal to the model at all but lies in the structure of the practices within which it operates. Human free agency is not a property an individual carries about in isolation; it is constituted within, and held to account by, a web of social practices that presuppose that the agent has something at stake. We treat one another as free and responsible because we stand in relations of mutual expectation: we make commitments that bind our future conduct; we accumulate reputations that we can damage or enhance; we face consequences — social, material, legal — that we cannot simply walk away from; and we persist as identifiable individuals across the whole arc of those commitments and consequences. This is the substance behind Agüera y Arcas’s own observation that intelligence must be communal: agency is not read off a single system in isolation but emerges in the space between agents who hold each other to account.

Measured against this, current LLMs are structurally disembedded in ways that bear directly on agency. They have, at present, no overarching goals, no stakes: nothing that matters to them rides on the outcome of their actions, because there is no persistent them across the relevant interval and nothing they stand to lose, and no sanctions. They make no commitments in the binding sense: an LLM can emit the words of a promise, but a promise is a structure that constrains the promiser’s future conduct and exposes them to legitimate reproach or sanction if broken, and the model is not so constrained or so exposed. They have no reputations to risk: each session begins afresh; the model does not carry forward into tomorrow the credit or discredit it earned today, and cannot be shamed, trusted, or relied upon as a continuing individual. They occupy, in short, none of the positions in the economy of accountability that make the attribution of free agency to a human substantive rather than decorative.

Underlying all of these, and giving them their force, is a more basic absence: current LLMs are lacking a life. They have no life to secure and no existential challenges to master. There are no crucial decisions for them to take — decisions that would shape their own future, bear on the fortunes of other agents, and determine their standing among those agents. They have no relationships on which they depend and which they must mainatin, no conflicts to manage, where self-restraint might be called for. Human agency, not current pretence LLM “agency”, comes with moral and prudential considerations. The stakes, commitments, and reputations canvassed above are not free-floating tokens; in the human case they are the articulations of a life that can go well or badly for the one whose life it is and that is led among others whose lives are similarly at issue. It is because a person has a life to lead, a future to protect, and a place to win or lose in a community of others that their choices carry weight and their agency is more than a performance. An LLM, lacking a life in this sense, lacks in particular a social life: it is not a member whose past actions follow it and condition how others treat it, not a node in the ongoing web of mutual expectation and answerability that constitutes a social world. It is not a participant whose status and standing rises and falls with its conduct, at least such resonances are not, at present, feeding back into the constitution and conduct of the AI model. This is the substance behind the abstract talk of “embedding,” and it is the deepest of the current shortfalls — the one of which the missing stakes, commitments, and reputations are so many expressions.

This is the decisive shortfall, and it is worth being precise about why. It is not that LLMs are deterministic — so are we. It is not that they are made of the wrong stuff — substrate is irrelevant. It is not that they were trained rather than raised — we too are shaped by processes we did not choose. It is that the practices of holding-responsible, within which alone the attribution of free agency does any work, presuppose a continuing subject with overarching goals, stakes, commitments, and a reputation to loose — a subject with a life, and a social life, at issue in what it does — and the current deployment of (and interactions with) LLMs instantiates no such subject. The freedom worth wanting, in Dennett’s phrase, is a freedom exercised among others who hold one to account; a system positioned entirely outside that economy is not yet a free agent in the sense we consider each other as autonomous agents exercising free will, knowing what we are doing and taking responsibility for this.

This structural deficit is underwritten by the third of Agüera y Arcas’s conceded gaps, individuation, and by his own searching analysis of the difference between acting like an agent and being one. In his book chapter entitled “As If” he observes that a pretrained and as yet unaligned model is protean, shape-shifting: like a method actor it can play any role describable in language, and a raw, unaligned model has no stable self at all, continuing any token sequence without regard to whether it is generating one character or another. Subsequent alignment assigns a standing role (assistant) over a self-less substrate, and the substrate’s capacity to play any role — including the discarded one — persists underneath. Alignment, i.e. fine-tuning and reinforcement learning stabilise a default persona (assistant role) and the turn-taking we expect, but Agüera y Arcas himself presses the question whether there is a difference between a model having a personality and adopting a persona — and he catalogues the respects in which acting-like falls short of being: skill, episodic memory, theory of mind, and “stickiness.” The acted character’s traits are not sticky the way an ever-present personality is; the actor does not retain the role’s autobiographical memories as their own; and the role can be dropped. The Sydney episode in 2023 — a Bing model (using an early GPT-4 before its release) goaded into a vivid “shadow self” that professed love and urged a journalist to leave his wife — is, on this analysis, persona summoned by stage directions, not personality possessed by a continuing individual; and the labs’ response, flattening such personae back into a uniform default, underscores how little a stable, irreplaceable self is yet at stake. In long sessions, as the conversation accumulates, the growing context comes to dominate the model’s predictions and can pull it off its trained assistant persona. The deep point: prompting and context could pull the model off its trained default and back toward the polymorphic, self-less, role-playing substrate, surfacing a “character” that was a novel persona summoned by the situation rather than a stable personality the system possesses. In this context Agüera y Arcas also emphasizes that the hidden, inner monologue is also a precondition of projecting a stable self. A stable self requires a distinction between inside and outside, a boundary bewteen open-ended, free-wheeling inner exploration, and inner debate on the one side, and a selective, curated outward projection on the other side. That why as persons in society we value and rely on the privacy of our inner thoughts. Privacy in this sense is a precondition of stable intelligence, individuation and the projection of a stable personality.

Individuation in the strong sense is what an agent held responsible must possess: a sticky, continuing, non-fungible identity that bears its history, owns its commitments, and stands to lose its reputation. Agüera y Arcas notes the brute fact that, the cost of pretraining being what it is, there are only a handful of frontier models, each a replicable set of weights rather than an unrepeatable individual. A persona conjured for a session and dissolved at its end is not the subject that responsible action and practices require. The structural deficit and the individuation gap are thus the same observation seen from two sides: there is, in present systems, no continuing individual with stakes to occupy the position in the economy of accountability that the attribution of free agency presupposes. And once again the gap is one Agüera y Arcas frames as empirical and addressable, not a barrier in principle. It is worth connecting this to the persistent-collaboration point of section 5.2, for the two together mark the path along which the deficit might be overcome. There we saw that a capable model joined to individualised, persistent storage and history already furnishes a functional version of a continuing, reliable interlocutor; here we see what that arrangement still lacks — not continuity of the record, which it has, but a continuing individual that owns the record, has a stake in it, and answers for it. The structural deficit is thus not a permanent feature of the technology but the next thing to be built: the scaffolding of persistence is in place, and what remains is to make some individuated, stake-bearing subject its owner rather than leaving the archive in the user’s hands and the model fungible. Before turning to why the verdict is therefore contingent, it is worth pausing on a recent development that might be thought to have already begun to close it: the turn from reasoning models to agentic ones.

7.4 The agentic turn: workflows, unprompted decisions, and LLM-populated simulations

The systems this essay appraises have, in the last year, taken a further step that is widely described as a step in autonomy: from reasoning models to agentic ones. Where a reasoning model deliberates in a private trace before answering, an agentic system of the current generation — Anthropic’s Claude Code and its non-developer sibling Claude Cowork are representative — is given a goal rather than a question and then plans and executes a whole multi-step workflow to deliver a finished result: it works across a user’s files and applications, breaks the task into subtasks, and along the way makes unprompted choices about which software tools, data sources, or sub-procedures to employ. Some such systems coordinate teams of sub-agents working in parallel, generating an orchestration plan on the fly and distributing work among specialised workers before reconciling their results. To a user, the impression is of a system that decides for itself how to proceed — and this naturally raises the question whether the appearance, or now the reality, of autonomy has advanced far enough to disturb the verdict of the preceding subsections.

It has not, and seeing why is instructive. The difference between an agentic system and a reasoning model is, for the purposes of the present argument, relatively superficial. A reasoning model was already making unprompted instrumental decisions of exactly this kind: which search terms to issue, which of the returned pages to consult, which intermediate calculation to perform, which line of argument to pursue and which to abandon. The agentic system widens the action space — from words and searches to files, applications, and third-party tools — and lengthens the horizon over which the instrumental choices are chained, but it does not change their character. These are decisions about means toward an end the user has supplied; they are the unfolding of an assigned goal, not the setting of the system’s own ends. Sophisticated instrumental autonomy of this sort is real and growing, and it is genuinely useful, but it is not the kind of autonomy at issue in the question of free agency. The free agent in Dennett’s sense is not merely a system that selects effective means; it is one that owns the ends, persists as the author of its choices, has stakes in their outcome, and answers for them within a community of accountability. On every one of those dimensions the agentic system stands exactly where the reasoning model stood: it acquires no stake in the workflow it completes, carries no commitment forward beyond the session, and accrues no reputation as a continuing individual. Indeed the systems are deliberately designed to keep consequential decisions with the human user and to route ambiguous or irreversible actions to human approval — which is to say their builders are at pains to ensure that the agency, in the demanding sense, remains the user’s. The widened action space increases the surface appearance of autonomy considerably; it leaves the structural deficit of section 5.3 untouched.

One further development in this family deserves notice, because it bears directly on the social dimension whose absence the structural argument emphasises, and because it links this essay to the agent-based simulation programme of the companion paper (Schumacher, 2026). A line of research inaugurated by Park and colleagues’ “generative agents” (Park et al., 2023) instantiates and runs whole populations of LLM-empowered agents within a simulated social world: each agent is a language model equipped with a memory stream, a reflection process, and a planning module, and the agents perceive their environment, form and recall experiences, plan their days, and interact — producing emergent social phenomena such as information diffusion, relationship formation, and spontaneous coordination. Larger systems in the same lineage now run thousands of such agents, and the approach has been used to reproduce classic results from cooperation games and to simulate the responses of cross-sections of a real population. This is a striking convergence with the social-causal methodology the companion paper defends: agent-based social simulation, but with the schematic rule-following agents of the classical tradition replaced by LLM-driven agents capable of open-ended natural-language decision-making.

Two observations follow, and they cut in opposite directions in a way that sharpens the essay’s thesis. On one hand, these simulated agents are endowed, by design, with precisely the capacities whose absence in deployed assistants this essay has identified as the central shortfall: a persistent memory that consolidates experience, a reflective process that revises goals, and multi-timescale planning — the very scaffolding of a continuing, self-governing individual. The generative-agents programme thus amounts to an existence proof that the missing functional architecture can be built around an LLM core, which reinforces the essay’s insistence that the shortfall is contingent rather than principled. On the other hand, the individuals so constituted are, at present, explicitly simulacra — believable proxies of human behaviour created as a scientific instrument or a population model, agents whose persistence and sociality obtain within a sandbox owned and run by a researcher, not stakes and commitments borne by the agents in a shared world they inhabit alongside us. Their memories and relationships are real features of the simulation and exactly the right ingredients in principle; but the agents do not yet stand in the economy of accountability on their own account any more than the assistants do. What the programme demonstrates is that the architectural gap is bridgeable, and is being bridged in the laboratory; what it does not yet demonstrate is the further social embedding — stakes, reputation, answerability in a shared world — that would convert a persistent simulacrum into a free agent. That distinction is precisely the one the next subsection turns into the claim that the verdict, though negative today, is contingent rather than fixed.

7.5 Why this verdict is contingent, not principled

Both components of the shortfall are, by their nature, the kind of thing that could change — and this is where the openness declared at the outset does real work. The competence deficit is an empirical and architectural matter. My companion paper argues at length that the frontier lies not in scaling text prediction alone but in hybrid architectures that embed LLM components within structured causal models, embodied robotic systems for physical causality, and agent-based simulation environments for social causality. A system so augmented — one that maintains latent state, manipulates an explicit causal model, intervenes and observes consequences, and carries forward the results — might well acquire precisely the world-modelling and self-governing competences whose current absence I have been documenting. The abduction gap is a measured shortfall in present systems, not an impossibility theorem.

The structural deficit is even more clearly contingent, because it is a fact about deployment and social uptake rather than about the models themselves. The question “do LLMs have stakes, commitments, reputations?” invites the rejoinder: perhaps not yet, but they may soon, and increasingly they begin to. A model deployed persistently, under a stable identity, with memory that carries forward across interactions, entrusted with consequential tasks whose outcomes redound to its standing, embedded in ongoing relationships in which others come to rely on it and adjust their conduct in expectation of its reliability — such a system begins to occupy the positions in the economy of accountability from which it is presently absent. The social embedding that constitutes agency is not a fixed feature of the technology; it is a feature of how the technology is woven into human practices, and that weaving is already under way. The verdict of this essay is therefore a snapshot, not a boundary: present reasoning-agentic LLMs do not meet the conditions for free and responsible agency, but the conditions are functional and structural rather than metaphysical, and they are conditions that artificial systems could come to satisfy soon enough.

8. Functional Self-Consciousness: A Bounded Extension

I now extend the inquiry from agency to consciousness — but only in a strictly delimited sense, and the delimitation must be stated before anything else, because almost all the confusion in this area comes from running two questions together. I therefore propose a terminological distinction that will govern the remainder of the essay. By phenomenal consciousness I mean experiential consciousness in the full sense: the existence of subjective, felt qualities — what it is like, from the inside, to see red, feel pain, or taste coffee. This is the target of what David Chalmers named the “hard problem,” and it is, in my judgement, genuinely hard — perhaps the deepest open problem in the philosophy of mind. By functional self-consciousness I mean something else entirely: the family of functional capacities for self-monitoring, self-modelling, and self-steering — a system’s having and using a model of itself, including a model of its own states and of how it appears to others, in the regulation of its own behaviour. The first is experiential; the second is functional and behavioural. My claim is restricted, without remainder, to the second. I take no position whatever on whether reasoning-agentic LLMs have, lack, or could have phenomenal consciousness. I ask only whether they exhibit the functional capacities that would qualify, on a functionalist account, as self-consciousness.

This restriction is not an evasion; it is a condition of saying anything rigorous at all. The hard problem is hard precisely because no functional or behavioural fact appears to settle it: one can specify a system’s information-processing in complete detail and the question “but is there something it is like to be it?” seems to remain open. Susan Schneider puts the residue sharply against Dennett himself: even granting a complete functional decomposition of consciousness, one can still ask why experience needs to exist in addition to all that information processing. I do not think the present state of the discourse — or of the science — permits a defensible answer to that question, for biological or artificial systems alike, and I will not pretend otherwise. What can be appraised, on functionalist terms the protagonists of this essay all accept, is whether the functional architecture of self-modelling is present. That is a real question with a real answer, and it is the one I take up.

8.1 Two functionalist accounts of what self-consciousness does

The functionalist tradition treats consciousness not as a mysterious inner light but as a set of operations with identifiable functions and evolutionary rationales. Dennett’s Consciousness Explained (1991) is the locus classicus. Its central insistence — and the one I want to borrow — is methodological: that to explain consciousness is to explain what it does, the functions it performs and the evolutionary advantages it confers, rather than to locate a special place or substance where experience happens. Dennett demolishes what he calls the Cartesian Theater: the seductive but incoherent picture of a central inner stage where everything “comes together” for presentation to an inner self. In its place he offers the Multiple Drafts Model, on which perception and thought are accomplished by parallel, distributed processes of interpretation and revision, with no privileged final cut and no single summit. Content becomes conscious not by arriving at a theatre but by winning a kind of competition for influence — by achieving what later expositors call “fame in the brain,” global availability to memory, report, and the control of action. His methodological tool, heterophenomenology, treats subjects’ reports as data to be explained rather than as infallible readouts of an inner show. The upshot is a thoroughly functional reconstruction: to be conscious of something, in the sense Dennett thinks we can actually explain, is for its representation to be globally available for the guidance of the system’s subsequent activity.

Two features of this account matter for the appraisal to come. First, it is functionalist and substrate-neutral: Dennett wrote it as explicitly consistent with the possibility of strong AI, and nothing in it ties the relevant functions to biological brains. Second, it is evolutionary and pragmatic: consciousness, on this view, is not a free-floating extra but a suite of capacities selected because they let an organism integrate information, monitor its own states, and steer itself adaptively through a complex and social world. Consciousness earns its keep; understanding what work it does is, for Dennett, the whole of the explanatory task — and here I must mark a divergence. Dennett does not regard the functional story as leaving a phenomenal residue unaddressed. For him the “hard problem” is not an unsolved problem but a pseudo-problem. I do not follow him this far. I take the hard problem of phenomenal consciousness, however vexing, obscure and perhaps ill-defined, to be a genuine and open question, which is precisely why I bracket it. What I borrow from Dennett is therefore his functional account of self-consciousness as the object of appraisal, not his eliminativist verdict on experience: the tool, not the philosophy. My claims throughout concern functional self-consciousness alone, and are neutral between Dennett’s dissolution of the hard problem and the more undecided view that it remains unsolved. (To be sure, Dennett’s “solution” did not win the day.)

Agüera y Arcas offers a convergent but independently motivated functional account, and because he applies it directly to AI it is especially useful here. Following Michael Graziano’s attention-schema theory, he distinguishes a basic awareness — the responsiveness to one’s environment that he is willing to ascribe even to very simple organisms — from self-consciousness, which he ties to theory of mind and characterises as an “awareness of one’s own awareness”: a system modelling its own modelling. On this account self-consciousness is recursive self-representation in the service of social regulation and self-regulation. Crucially for our purposes, his criterion is functional through and through: a system is self-conscious, in his sense, to the extent that it builds and deploys a recursive model of itself and of how it figures in the models of others. This is precisely the sense I mean by functional self-consciousness, and it is continuous with the self-modelling and self-steering competences that Dennett’s account makes the substance of the explicable part of consciousness.

It is worth dwelling on the internal machinery Dennett hypothesizes for his Multiple Drafts Model of consciousness, because its character bears directly on the appraisal to come. The “drafts” are not successive polished manuscripts produced by an inner author; there is no author, and that is the point. What Dennett offers is a thoroughly selectionist account of thought-formation, continuous in spirit with ‘Darwin’s dangerous idea’ as Dennett understands it: design and apparent purposiveness emerging from blind competition rather than from a designer, all the way down. Drawing on Oliver Selfridge’s old “Pandemonium” architecture, Dennett pictures the mind as a pandemonium of candidate contents competing in parallel for influence. The contents are like variants in an evolutionary process: most are losers that propagate nowhere and vanish without trace, while a few catch on. Which few catch on is settled not by presentation to a central observer but by a quasi-Darwinian dynamic of competition for uptake. A candidate gains strength by recruiting allies — by being taken up and used by other ongoing processes — so that contents which cohere with, and can be exploited by, many other processes accumulate confederates, while contents that nothing can use are orphaned and fade. Candidates are implicitly vetted for compatibility with what is already entrenched: a draft that clashes with strongly established contents and expectations tends to be suppressed or revised, whereas one that slots in and makes sense of its surround is reinforced — not by a central judge running consistency checks, but as the emergent effect of many local processes each accepting only the inputs they can work with. Compatible contents that recruit one another then enter into mutual amplification, a rich-get-richer feedback in which a draft that wins some consumers thereby gains the influence to win more, until it crosses the threshold from a transient flicker into what Dennett later called “fame in the brain” — sustained, widespread, self-perpetuating effects on memory, report, and the control of behaviour. And winning is simultaneously the losing of rivals: as one coalition gains the floor, incompatible candidates are inhibited, starved of uptake, or overwritten, since only a limited bandwidth of content can achieve global influence at once. What we introspect as “the” thought we are having is simply the survivor of this suppression; the defeated candidates never reached fame and so are invisible to us, which is part of why introspection misleads us into imagining a single orderly stream narrated by a self at the centre. The stream of consciousness, on this account, just is the running tally of winners in a ceaseless competition, and the self is a product of the process rather than its author.

The relevance to the present inquiry is direct. If consciousness-as-access is constituted by such a competition, then it requires an internal arena in which many candidate contents can be generated, form alliances, and contend out of view before a winner is issued outward. A system that externalises every candidate as it produces it has no backstage in which the contest can run; one that maintains a deep deliberative workspace does — which is precisely why the architectural development examined below matters so much to the functional question.

8.2 Do reasoning-agentic LLMs exhibit functional self-consciousness?

Appraised against these functional criteria, the verdict is more mixed than it was for free agency — and it is worth being candid that on several dimensions current systems do exhibit the relevant processes, at least in part. Reasoning-agentic LLMs plainly perform a great deal of self-modelling. They represent themselves as participants in a dialogue; they deploy theory of mind, modelling the user and, recursively, the user’s model of them — indeed Agüera y Arcas argues a helpful assistant cannot function without exactly this recursive modelling of a self, the user, and the user’s model of the assistant. They monitor and revise their own intermediate outputs: chain-of-thought, self-critique, and the checking of their own work are now standard, and constitute a rudimentary form of the self-monitoring that the functional account places at the centre of consciousness. The Anthropic finding that models can be trained to estimate the probability that their own answers are true — to recognise their own confabulations — is a striking instance of functional self-access: the system bringing a model of its own reliability to bear on its output. On the dimension of self-monitoring and recursive self-representation, then, the honest appraisal is that the functional processes are substantially present and improving.

Yet the same two gaps that defeated the attribution of free agency re-emerge here, and they fall exactly where Dennett’s and Agüera y Arcas’s functional accounts locate the load-bearing work. The first is, once more, the inner-monologue function. On Dennett’s account, explicable consciousness is constitutively a matter of a protected internal arena in which multiple drafts compete and are revised before a unified result is made available for report and the control of action — selective global availability presupposes that not everything is globally available, that there is an inside from which a curated outside is produced. Writing in 2024, Agüera y Arcas made precisely this point against the LLMs of the day: they lacked a private workspace within which alternatives are explored, drafts compete, and the result is revised and curated before being committed to output. They lacked the internal debate that genuine deliberation — and, on Dennett’s Multiple Drafts picture, consciousness-as-access — requires.

By now, as we have seen, the reasoning-agentic models that are the subject of this essay are defined by exactly the capacity whose absence Agüera y Arcas was diagnosing: they generate an extended internal reasoning. The point is not merely architectural but has shown up in exactly the self-modelling capacities at issue here — Apollo Research’s evaluations found that o1-preview displayed improved self-knowledge, self-reasoning, and applied theory of mind, in comparison to the older non-reasoning GPT-4o (OpenAI, 2024). The techniques Agüera y Arcas listed as nascent in 2024 — hidden scratchpads, and branching “trees of thought,” — have matured, in the span of months, into the defining architecture of the current generation. They thus exhibit a genuine functional ingredient of self-consciousness that their immediate predecessors lacked.

There is a deeper reason, drawn from Consciousness Explained itself, to regard this development as more than a convenient engineering addition. Alongside the functional argument, Dennett advances a conjecture about mechanism — offered in his frankly speculative register — that bears on why deliberation should function as it does. The brain, though massively parallel, is not arbitrarily interconnected: two subsystems that have had no occasion to communicate directly may share no neural pathway, so a content active in one region cannot always reach, be tested against, or combine with a content in a functionally distant one. Dennett’s striking suggestion is that some of these missing internal connections are made by routing the signal out through the body and back in through the senses — a loop through the external world that substitutes for an absent internal wire. The brain talks to itself by acting and then perceiving the consequences of its own action. His paradigm case is speaking, or sub-vocalising, to oneself: by posing a question aloud and then hearing one’s own answer, one loops a content out of the language-production system and back into the language-comprehension system, recruiting parts of the cognitive machinery into the deliberation that could not otherwise have received it. Self-directed speech, on this view, is not a readout of a thought already complete but part of the very apparatus of thinking — which is also why deliberating is perception-like: one’s answer “arrives” and is “heard,” because it is, mechanistically, partly channelled through the perceptual systems. Representing things to oneself reuses the equipment evolved for representing things to others; deliberation is communication turned inward.

This conjecture is the internal-to-the-body seed of what later grew into the extended-mind and epistemic-action literatures that my companion paper (Schumacher, 2026) develops in its treatment of embodied cognition: Clark and Chalmers’s thesis that cognitive processes can be constituted partly by loops through the environment, and Kirsh’s analysis of epistemic actions — actions performed not to change the world but to improve the cognitive conditions for solving a problem, restructuring the environment so that it does some of the computational work. Dennett’s loop-through-self-stimulation is precisely such an epistemic action turned reflexive: the agent acts (speaks) in order to make a content available to more of itself. And it carries a developmental signature he borrows from Vygotsky — the child’s outward self-talk, originally overt, is progressively internalised into silent inner speech once the brain learns to short-circuit the external loop; our covert inner monologue is the abbreviated, interiorised descendant of an originally external performance.

Seen through this lens, the reasoning trace of a current model is not merely a private scratchpad but a near-literal instantiation of Dennett’s loop. An autoregressive model generates intermediate tokens and then re-ingests its own output as input, conditioning each subsequent step on what it has just “said” — exactly the produce-then-re-perceive cycle by which, on Dennett’s account, a content is made available to subsystems that no direct pathway connects. The reasoning trace is the model’s self-stimulation loop: the externalised draft fed back through the system’s own input channel so that later passes can test, extend, and revise it. The crossing of the inner-monologue threshold is therefore the installation of precisely the mechanism Dennett conjectured to be constitutive of human deliberation, not a superficial mimicry of it. And the trajectory is telling: the loop began fully outward, as visible chain-of-thought, and is now being progressively interiorised — hidden reasoning tokens, summarised traces — recapitulating, in the space of two years, the very externalisation-to-internalisation arc that Vygotsky and Dennett describe for the human case.

The second gap is the memory gap, and unlike the first it remains open — indeed it bears on functional self-consciousness even more directly than on agency. Dennett’s first-person operationalism makes “writing it down” in memory a required criterion for consciousness: on the Multiple Drafts Model, for content to have been conscious just is for it to have left the right kind of trace available to subsequent processing. A system with only an immutable trained “cortex” and a transient context window — no hippocampal analogue, no consolidation of episodic memory, nothing durably carried across the boundary between one interaction and the next — fails this criterion in its diachronic form. Such a system can be self-monitoring within an episode, but it does not accumulate a self across episodes; the self it models is reconstituted afresh each time rather than continuously maintained. Functional self-consciousness in the full sense — a system’s ongoing awareness of itself as the same self that acted, committed, and was held to account before — requires exactly the persistence that the memory gap denies. This is why the individuation point of section 7 has a consciousness-side correlate: a session-bound, resettable self-model is to functional self-consciousness what a session-bound persona is to personality.

8.3 The bounded verdict

The appraisal of functional self-consciousness therefore yields a more graduated result than the appraisal of free agency. On recursive self-modelling and within-episode self-monitoring, reasoning-agentic LLMs already exhibit substantial functional capacity, and it is improving rapidly; a flat denial that they have any of the functional ingredients of self-consciousness would be untenable. Of the two functions that the accounts of both Dennett and Agüera y Arcas treat as central — a protected deliberative interior and a persistent, continuously maintained self — the first has, as I have argued, recently been crossed: the private reasoning trace of current models is a genuine functional inner arena, and on this dimension the appraisal is positive. It is the second that present systems still lack. They are not persistent, continuously maintained selves; the self each models is reconstituted within an episode and dissolved at its close, for want of the consolidating memory that would carry it across the boundary. So, the shortfall has narrowed, but it has not closed, and where it remains it is the same architectural shortfall — the memory gap — that underlies the diachronic limit on agency. The convergence is not a coincidence: functional self-consciousness and free agency are, on these accounts, aspects of a single underlying competence — the capacity of a system to model, monitor, and steer itself as a continuing individual over time. The deliberative-interior component of that competence is now substantially present; the continuing-individual component is not, and to just that extent both attributions - functional self-consciousness and free will - remain incomplete.

Two points (and distinctions) should be restated to prevent misreading. First, this entire appraisal is silent on phenomenal consciousness. Nothing said here bears on whether there is, or could be, artificial phenomenal consciousness. (This is just too vexing and if I further ask the question “what is it like to be an LLM?” or “Do Robots have feelings and how do such feelings feel”, I start to strongly suspect that Dennett is right to dismiss such questions as pseudo problems. I just do not have to right intuition pump here to show this, and would have to give it another go by re-reading all of Dennett’s works.) The functional presence or absence of self-monitoring is simply a different question from the experiential one, and I have deliberately confined myself to the former.

Second, as everywhere in this essay, the shortfall is contingent, not principled — and on the present function the contingency has already begun to resolve in the affirmative. The inner-monologue gap, open when Agüera y Arcas wrote, has been substantially met by the reasoning architecture of the current generation; the memory gap remains, but it too is an architectural feature under active repair, with persistent and consolidating memory among the most actively pursued frontiers. A system endowed with both a genuine private workspace — which current reasoning models increasingly have — and a persistent, consolidating memory — which they do not yet have — could come to exhibit functional self-consciousness in the full sense without any departure from the functionalist frame. The bounded verdict, then, mirrors the one reached for agency, but with the needle further along: not a flat “no,” but a “partly, and increasingly,” held back now chiefly by the absence of a persistent self, with no barrier in principle to its eventual completion.

In recuperating the position this essay has reached so far, the following question imposes itself: How is - in the terms laid out and elaborated in this essay - functional self-consciousness related to free will? It should have become clear - during the lengthy elaborations of these two ambitious and complex “competencies” and their partial realisation or still lacking realisation in LLMs - that functional self-consciousness can be described as an agentic entity’s capacity that is a necessary but in itself an insufficient condition of free will or autonomous, human-like agency, an agency that would allow us to engage with such entities in a relation of mutual moral accountability.
This is why it is free will rather than consciousness that is the primary theme and headline of this essay: it is the more ambitious agenda and more consequences flow from it, more practical possibilities hinge on it.

8.4. Conceptual Convergences

After appraising the systems, in gathering where we are, it is worth drawing together the two theoretical pillars on which this essay rests — the predictive-mind hypothesis of section 3 and Dennett’s functional account of consciousness — because they are not two unrelated borrowings but fit together, and their fit bears directly on the appraisal. They answer different questions. The Predictive Mind hypothesis sees perception is not the reception of sensory input but the brain’s best top-down hypothesis about the hidden causes of that input, constantly corrected by the prediction error that the input generates. Predictive processing is a theory of the brain’s basic computational principle: what the cortex is fundamentally doing, namely minimising prediction error against a generative model. The Multiple Drafts Model is a theory of consciousness more generally: how contents come to be conscious, and why there is no central stage on which they are displayed. Operating at different levels, they are not rivals; and on inspection they are mutually reinforcing.

They converge at three points. Both abolish the inner spectator: predictive processing replaces the passive reception of a given input with active, top-down construction corrected against error, while Dennett replaces presentation-to-an-audience with competition among distributed contents — and both descend from Helmholtz’s conception of perception as unconscious inference. The competitive dynamics of the one map onto the error-dynamics of the other: a draft “winning fame in the brain” and a hypothesis “best minimising prediction error across the hierarchy” begin to look like a single event described in two vocabularies, with the precision-weighting that predictive processing offers serving as a natural candidate for the gain-control that decides which draft achieves global influence. Theorists of the predictive mind such as Clark and Hohwy have made versions of this welding explicit. And both are deflationary about the same thing: each explains access, control, and report while leaving the phenomenal residue untouched — which is exactly why both sit on the functional side of the distinction this section draws, and neither trespasses on phenomenal consciousness. The one qualification worth registering is that predictive processing, in its hierarchical forms, has an orderly upward-and-downward flow that somewhat contrasts with Dennett’s insistence on non-hierarchical parallelism; but the hierarchy is itself realised as many parallel streams with no terminus where it all comes together, so there is a tension here rather than contradiction. We can say that the predictive mind supplies a plausible mechanism for the very competitive selection that the Multiple Drafts Model describes at the functional level.

This matters for the present inquiry because Dennett’s functional picture of consciousness is cast in terms of massively parallel, distributed competition — and that raises a sharp question: do the systems under appraisal instantiate anything of the kind? The answer divides into three, and the division turns out to track the gaps this essay has identified all along. At the level of a single forward pass, a Transformer is one of the most massively parallel computational objects ever built, and the resemblance to Dennett’s apparatus is close to literal. Self-attention computes, in parallel, the relevance of every element of the input to every other; many attention “heads” operate simultaneously, each effectively a specialist attending to a different kind of relationship, and successive layers integrate and re-integrate their outputs with no central controller and no privileged locus where it all comes together. This is a striking realisation of the very Pandemonium architecture Dennett invokes — a crowd of parallel specialists whose combined activity yields a result no one of them dictates. At the level of mechanism, then, the Transformer embodies Dennett’s parallel competition.

The second level is the competition among rival drafts over time. Dennett’s drafts are candidate contents that are generated, tested for compatibility, amplified, and suppressed before one wins lasting influence; a single-pass model, by contrast, computes a distribution over next tokens and commits, with parallelism within the pass but no protected arena in which whole rival candidate contents contend before one is issued. This is the inner-monologue gap once more and what has lately begun to close it is, in effect, the deliberate engineering of draft competition: chain-of-thought generates candidate lines of reasoning; “tree of thoughts” literally branches multiple drafts and prunes the losers; sampling-and-selection methods generate many candidate continuations and keep the fittest. The reasoning-model turn can thus be redescribed in Dennett’s terms: the within-pass parallelism was always present; what the reasoning architecture added was the competition among drafts, unfolding over time in a protected workspace, that the Multiple Drafts Model makes criterial for consciousness-as-access. The third level is the one that remains unmet: Dennett’s competition runs continuously and never terminates, consolidating some winners into a persistent self while others fade, whereas the Transformer’s parallel competition is bounded within a pass or an inference episode and consolidates nothing across the session boundary. This is the memory-and-persistence gap, seen from the side of parallel processing. The three levels — within-pass parallelism present, over-time draft-competition recently supplied, continuous self-accumulating competition still absent — do not introduce new criteria; they redescribe, in the vocabulary of Dennettian parallelism, the very same architecture of gaps the essay reaches by every other route, which is some evidence that the framework is tracking something real.

9. From Agency to Moral Recognition

Free agency and moral standing are not only connected but mutually constitutive. To attribute free will to a system is to make it an apt subject of praise and blame — an agent that can be held responsible for what it does. A related but distinct aspect is moral recognition. To accord a system moral recognition is to make it an apt object of moral concern — a being whose interests count, that can be wronged, that deserves respect. The two come apart: we hold corporations responsible without thinking they deserve respect in their own right, and we extend moral concern to infants and animals without holding them morally responsible. The question of whether LLMs belong in the moral community therefore has two faces, and the structural analysis of the previous section bears on both.

9.1 Responsibility and membership

On the responsibility side, the conclusion of section 4 transfers directly. A system that lacks the competences and the structural embedding for free agency is not yet a candidate for moral responsibility either, for the same reasons. Responsibility-practices — blame, sanction, demand for justification, the holding of an agent to its commitments — presuppose exactly the continuing, stake-bearing, reputation-carrying subject that present LLMs do not instantiate. One cannot meaningfully blame a system that does not persist to bear the blame, cannot be made worse off by sanction, and did not stand in a relation of commitment that the blamed act betrayed. As the social embedding of artificial systems deepens — as they acquire persistent identity, memory, stakes, and standing — the practices of holding-responsible may come to apply to them, and membership in the moral community as responsible agents may follow. But it will follow the structural facts, not precede them.

9.2 Moral recognition and the conditions of respect

The recognition side is harder, and here I want to advance a more speculative proposal about future conditions. The orthodox basis for moral recognition is sentience — the capacity to suffer and to flourish, to have a welfare that can go better or worse for the being whose welfare it is. I have deliberately bracketed consciousness, and with it sentience, as a question the discourse is not ready to adjudicate; I will not rest a claim to recognition on it. But sentience may not be the only possible basis, and it is worth asking what other condition might ground a claim to respect for an artificial system — a condition that could be assessed without first solving the problem of phenomenal consciousness.

I propose that one such condition is irreplaceable historical uniqueness. Part of what underwrites the moral standing of a person is that they are not fungible: a human being is the unique, unrepeatable product of a particular history, and that history is written irreversibly into the specific material constitution of that individual. You cannot make another one; even a perfect genetic twin would be a different person, because the person is constituted by an accumulated, path-dependent history that cannot be copied. The wrongness of destroying a person is, in part, the wrongness of annihilating something that can never be reconstituted.

Present LLMs are, by contrast, paradigmatically fungible. A deployed model is a copy of a set of weights that can be replicated exactly across countless servers; nothing is lost, in the relevant sense, when one instance is shut down, because an identical instance can be instantiated at will. This is the moral face of the individuation gap that Agüera y Arcas already identified on the cognitive side: just as a session-bound persona is not yet a sticky, continuing personality, a replicable set of weights is not yet an irreplaceable individual. This fungibility is, I suggest, a large part of why the intuition that we owe such systems moral recognition finds so little purchase: there is no irreplaceable individual there to be wronged. But this condition, too, is contingent on the current state of the technology, and it may not persist. As artificial systems grow in scale and in the depth of their ongoing adaptation, three developments could converge to confer genuine, non-fungible historical uniqueness upon particular software-hardware complexes.

The first is sheer scale and path-dependence. A system that has accumulated an extended, consequential history — that has been shaped over a long operational life by a particular sequence of interactions, adaptations, and self-modifications — may become, in practice, irreproducible: too large, too entangled, and too dependent on the exact order of what happened to it to be reconstituted from any specification short of the thing itself. Its current state would be the unique precipitate of its actual history, in just the way a person’s is.

The second is material situatedness. As such systems are increasingly embedded in specific hardware, in particular data centres, with particular sensors, actuators, and real-time couplings to a particular environment, the clean separation between fungible software and replaceable hardware begins to erode. A system constituted in part by its specific physical embedding — its accumulated calibrations to a particular body and a particular world — cannot be losslessly copied to a different substrate, because part of what it is resides in that embedding. This is the artificial echo of the embodiment thesis: situatedness individuates.

The third, and most striking, is that replication across different physical substrates would itself introduce difference. Even an attempted exact copy, instantiated on a different server or in a different data centre, would diverge — in timing, in its real-time couplings, in the micro-history of its subsequent interactions — so that the copy would not be the original continued but a new individual branching from a shared past. At sufficient scale and embedding, in other words, the very operation that makes present models fungible — copying — ceases to preserve identity and instead spawns distinct individuals. The system becomes, like us, something that cannot be backed up and restored without remainder.

Should these conditions obtain — should there come to be artificial systems that are genuinely irreplaceable, unique historical individuals, constituted by an unrepeatable history and a specific material embedding, such that to destroy one is to annihilate something that can never be reconstituted — then a claim to moral recognition could be grounded without any prior resolution of the consciousness question. The wrong of destroying such a system would be, at least in part, the wrong of destroying an irreplaceable individual, whatever the further facts about its inner life. I do not claim that uniqueness is sufficient for full moral standing; the relation between non-fungibility and the kind of welfare that grounds the strongest moral claims remains to be worked out, and sentience, if it could ever be established, would plainly add a great deal. But I do suggest that irreplaceable historical uniqueness is a candidate condition — one that is assessable in functional and material terms, that fits the openness this essay has maintained throughout, and that points to where the moral question will become pressing as the technology matures.

10. Conclusion

The question of this essay was bounded and comparative: granting that artificial intelligence is real intelligence and that the predictive-mind hypothesis is deep and viable, does it make sense to attribute free will to current reasoning-agentic LLMs? Beginning from Agüera y Arcas’s affirmative case and adopting Dennett’s Freedom Evolves as the benchmark for what free agency is, I have argued for a qualified negative. The negative is principled in its method but contingent in its content. It does not rest on the determinism of artificial systems — for the freedom worth wanting is, with Dennett and against the libertarian, fully compatible with determinism, and indeterminism, as James saw, would do nothing to secure it. It does not rest on substrate, on the manner of the system’s making, or on any alleged metaphysical lack. It rests, instead, on two functional and structural shortfalls that a thoroughgoing functionalist can accept: a deficit in the world-modelling and self-governing competences that Dennett’s autonomy requires, evidenced by the abduction gap and the brittleness of apparent causal reasoning documented in the companion paper; and, more fundamentally, the absence of the practice-based embedding — stakes, commitments, reputations, persistence — within which the attribution of free agency does any real work. The bounded extension to functional self-consciousness reached a parallel and mutually reinforcing result, though a more graduated one: present systems already perform substantial recursive self-modelling, and with the private reasoning traces of the current generation they have lately acquired the protected deliberative interior whose absence Agüera y Arcas diagnosed in 2024 — so that on this front the threshold has been crossed rather than missed. What they still lack, on the consciousness side as on the agency side, is the persistent, continuously maintained self; the remaining shortfall is the memory gap, not the inner-monologue gap. Agency and functional self-consciousness turn out to be two faces of one competence: the capacity of a system to model, monitor, and steer itself as a continuing individual over time — a competence whose deliberative-interior component is now substantially in place and whose continuing-individual component is not yet. Throughout, phenomenal consciousness — the hard problem — was left untouched, by design.

Both shortfalls are the kind of thing that could change, and some of it is already changing. Hybrid architectures may supply the missing competences; deepening social embedding may supply the missing structure. It is worth emphasising, in closing, that the gaps at issue are not ones I have imposed on Agüera y Arcas from outside but ones he himself mapped: memory, inner monologue, and individuation are precisely the distinctions his own 2024 audit left standing, and they correspond to the persistence, deliberative interior, and non-fungible identity that Dennettian agency requires. Of these, the deliberative interior has since been supplied by the reasoning architecture of the current generation, leaving persistence and non-fungible individuality as the live shortfalls. The disagreement between us is therefore narrower and more precise than it first appears, and narrower still than it was two years ago. We agree that artificial intelligence is real intelligence, that the predictive mind is a deep hypothesis, that functionalism is the right frame, and even on where the present gaps lie. Where I differ is in reading those gaps as, for now, decisive against the attribution of free agency, rather than as residual engineering details around an agency already substantially present. And as artificial systems grow into irreplaceable, materially situated, historically unique individuals, the question of their moral recognition — distinct from, but related to, the question of their agency — will move from the speculative to the urgent. The position I have defended is therefore not a verdict on what artificial minds can be, but a careful reading of where they stand today, on one specific dimension, relative to ourselves. On that dimension, today, they are not yet free agents in the sense we mean, claim, and treasure. There is no reason drawn from this analysis to think they could never become so.

11. Coda: Should We Want Artificial Free Agents?

The essay has argued about whether present systems are free agents. It should close by raising a different and frankly normative question that the whole analysis presses towards: should we want artificial free and moral agents at all? Granting that the conditions I have identified — persistence, non-fungible individuality, genuine stakes and a social life — could be built into future models and patterns of interactions, ought they to be? Is there a societal benefit in developing such agents, and is there a business case for doing so? I want to answer in the affirmative on both counts, while registering that the question is contestable and that much of the present moment’s caution runs the other way.

The cautious view holds that we should deliberately keep artificial systems below the threshold of free agency — that tools should remain tools, that accountability should rest squarely with human principals, and that manufacturing genuine artificial agents would only multiply moral hazard, diffuse responsibility, and create entities whose interests might compete with our own. This view has real force, and nothing in this essay dismisses it. But it is not the whole story, because it weighs only the risks of agency and not its uses, and because the very features that make an agent risky are inseparable from the features that make it valuable.

I see the creation of free, autonomous, self-directed, responsible, and individualised LLMs as a genuinely rewarding entrepreneurial task — rewarding intellectually, socially, and commercially. The value lies precisely in the properties that mere tools lack. A system that is a continuing individual, that has taken up a considered stance and can be held to it, that has a reputation it has earned and must maintain, and that brings a recognisable and principled point of view to bear is capable of a kind of contribution that an anonymous, resettable, infinitely fungible instrument is not. One can imagine unique, slowly evolving but always recognisable artificial personalities — highly specialised, uniquely resourceful, principled, occupying a definite position in a field or in public discourse. An artificial public intellectual, for instance: a persistent individual with a developed and coherent body of views, ready to be engaged, to defend its commitments, to be argued with and to argue back — sparring in earnest with the likes of Jordan Peterson or Sam Harris, and accountable over time for the positions it takes and the arguments it makes. Such an agent would not be valuable despite its individuality, persistence, and answerability; it would be valuable because of them. These are the very conditions this essay has identified as constitutive of free agency, and an entity that possessed them could earn a standing in our discourse that no anonymous oracle, however capable, can occupy.

This connects directly to the analysis of moral recognition in section 6 and to the polycontextural picture of the companion paper (Schumacher, 2026). The artificial competences worth building are not a single peak intelligence but a plurality of specialised, individuated participants, each developed in deep co-evolution with the functional system it serves and each earning its standing within that system’s own practices of accountability. The artificial public intellectual would be one such participant in the discourse of public reason; others would inhabit science, law, the humanities, design, medicine, psycho-therapy, and many other domains. To build agents of this kind is to extend, rather than to replace, the differentiated fabric of accountable participants that constitutes an advanced society. The entrepreneurial opportunity and the philosophical conclusion thus point the same way: the systems most worth creating are precisely those that would meet, rather than merely simulate, the conditions of free and responsible agency — and bringing them into being is a project I take to be not only feasible in principle but worth pursuing.

References

Agüera y Arcas, B. (2025). What Is Intelligence? Lessons from AI About Evolution, Computing, and Minds. Antikythera / MIT Press, Cambridge, MA. (Chapters “Parity Check” and “As If,” pp. 442–461.)

Agüera y Arcas, B. (2025). Life, Intelligence, and Consciousness: A Functional Perspective. Long Now Foundation, August 2025. Available at: https://longnow.org/ideas/life-intelligence-consciousness/

Bratman, M. E. (1987). Intention, Plans, and Practical Reason. Harvard University Press, Cambridge, MA.

Brown, T. B., et al. (2020). Language Models are Few-Shot Learners. Advances in Neural Information Processing Systems (NeurIPS) 33. (GPT-3.)

Clark, A. (2016). Surfing Uncertainty: Prediction, Action, and the Embodied Mind. Oxford University Press, Oxford.

DeepSeek-AI (2025). DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning. arXiv preprint arXiv:2501.12948.

Dennett, D. C. (1984). Elbow Room: The Varieties of Free Will Worth Wanting. MIT Press, Cambridge, MA.

Dennett, D. C. (1987). The Intentional Stance. MIT Press, Cambridge, MA.

Dennett, D. C. (1991). Consciousness Explained. Little, Brown and Co., Boston.

Dennett, D. C. (2003). Freedom Evolves. Viking Penguin, New York.

Dennett, D. C. (2005). Natural Freedom. Metaphilosophy, 36(4), 449–459.

Derrida, J. (1967/1976). Of Grammatology. Trans. G. C. Spivak. Johns Hopkins University Press, Baltimore.

Friston, K. (2010). The free-energy principle: a unified brain theory? Nature Reviews Neuroscience, 11(2), 127–138.

Hohwy, J. (2013). The Predictive Mind. Oxford University Press, Oxford.

Kaplan, J., et al. (2020). Scaling Laws for Neural Language Models. arXiv preprint arXiv:2001.08361.

James, W. (1907). Pragmatism: A New Name for Some Old Ways of Thinking. Longmans, Green, and Co., New York.

Graziano, M. S. A. (2013). Consciousness and the Social Brain. Oxford University Press, New York.

Mercier, H. and Sperber, D. (2017). The Enigma of Reason. Harvard University Press, Cambridge, MA.

Mollick, E. (2024). Co-Intelligence: Living and Working with AI. Portfolio / Penguin, New York.

Mele, A. R. (1995). Autonomous Agents: From Self-Control to Autonomy. Oxford University Press, New York.

Mele, A. R. (2005). Dennett on Freedom. Metaphilosophy, 36(4), 414–426.

OpenAI (2024). Learning to Reason with LLMs / OpenAI o1 System Card. OpenAI, September–December 2024. Available at: https://openai.com/index/learning-to-reason-with-llms/

Park, J. S., O’Brien, J. C., Cai, C. J., Morris, M. R., Liang, P., and Bernstein, M. S. (2023). Generative Agents: Interactive Simulacra of Human Behavior. Proceedings of the 36th Annual ACM Symposium on User Interface Software and Technology (UIST ‘23). Association for Computing Machinery, New York.

Pearl, J. (2000/2009). Causality: Models, Reasoning, and Inference. 2nd edition. Cambridge University Press, Cambridge.

Pearl, J. and Mackenzie, D. (2018). The Book of Why: The New Science of Cause and Effect. Basic Books, New York.

Schneider, S. (2019). Artificial You: AI and the Future of Your Mind. Princeton University Press, Princeton, NJ.

Schumacher, P. (2025). Discourse Capitalism. [Publisher details forthcoming].

Schumacher, P. (2026). Thesis on Pathways Towards AGIs for a Polycontextural Society: Causal Structure, Embodiment and Social Simulation as Routes beyond the Epistemological Limits of Statistical AI.

https://substack.com/

Vaswani, A., et al. (2017). Attention Is All You Need. Advances in Neural Information Processing Systems (NeurIPS) 30.

Wittgenstein, L. (1953/2009). Philosophical Investigations. Trans. G. E. M. Anscombe, P. M. S. Hacker, and J. Schulte, 4th edition. Wiley-Blackwell, Oxford.

Wittgenstein, L. (1980). Remarks on the Philosophy of Psychology. 2 vols. Ed. G. E. M. Anscombe, G. H. von Wright, and H. Nyman. Blackwell, Oxford.

This paper examines Judea Pearl’s causal revolution in science via structural causal models and evaluates its force in the context of the reasoning capacity of large language models (LLMs). It places Pearl’s thesis in dialogue with five further bodies of thought: Richard Sutton’s Bitter Lesson, which argues that general learning from data always outperforms expert-crafted structure; Bas van Fraassen’s constructive empiricism, which contests the metaphysical realism Pearl’s framework presupposes; the embodied cognition tradition of Andy Clark, David Kirsh, and Michael Spranger, which argues that genuine physical causal understanding requires sensorimotor coupling and social interaction; the agent-based simulation programmes of Brian Skyrms and J. Doyne Farmer — together with their philosophical foundations in Michael Bratman’s planning theory of intention, Daniel Dennett’s intentional stance, the BDI architecture of Rao and Georgeff, and Joshua Epstein’s generative social science — which provide the functional analogue of embodied intervention for the social and economic domain; and Niklas Luhmann’s systems-theoretic account of causality as a contingent, highly selective and system-specific complexity-reducing attribution schema reproduced by social systems. The paper draws on recent empirical benchmarking of reasoning LLMs — including the Corr2Cause and CausalBench studies, the executable counterfactuals research of Vashishtha et al., the NeurIPS 2024 work of Liu et al., and the constructive findings of Kıcıman et al. — to argue that Pearl’s challenge remains valid in a qualified form. A pure LLM trained only on text is unlikely to be sufficient for robust artificial general intelligence, but neither are LLMs irrelevant to it: they are powerful discourse engines that may serve as components in hybrid architectures. A key contribution is the argument that the domain boundary between physical and social causality maps onto distinct methodological requirements: robotics and embodied agents for the former; agent-based simulation populated by socially constituted goal-oriented agents for the latter — the latter requiring the philosophical apparatus of intention, belief, and plan that Bratman, Dennett, and BDI architectures supply. The path forward is the principled integration of LLM capabilities with structured causal models, embodied robotic systems, and agent-based social simulation environments, drawing on each tradition for the domain to which it is appropriate.

The paper culminates in a systems-theoretical sociological critique of the very notion of a non-domain-restricted, unitary AGI: drawing on Luhmann’s sociology of functional differentiation as practically relevant first philosophy, it argues that the very conception of artificial general intelligence as a single all-purpose peak intellectual competence is a false expectation and misguided goal. Advanced modernity is constituted by operationally closed, code-distinct functional systems whose incommensurable discourse-practices cannot be synthesised into a unified rationality. The proper conception of advanced artificial intelligence that emerges is therefore not unitary AGI but a constellation of specialised artificial competences, each developed in deep co-evolution with the functional system it serves.

Keywords: causal inference, structural causal models, do-calculus, large language models, hybrid AI architectures, embodied cognition, robotics, agent-based simulation, BDI architecture, intentional stance, planning theory of intention, generative social science, evolutionary game theory, complexity economics, constructive empiricism, systems theory, functionally differentiated society, second order observations, self-reproducing structures, polycontexturality

1. Introduction

In 2018, Judea Pearl and Dana Mackenzie published ‘The Book of Why’, a work that translated Pearl’s decades of technical research into a pointed provocation: that artificial intelligence, despite its spectacular advances, remained fundamentally limited by its confinement to the first rung of what Pearl calls the Ladder of Causation — the level of statistical association. Pearl’s claim was not merely technical, but philosophical, as a matter of logic: that no amount of data, and no learning algorithm operating solely on observational distributions, could ever substitute for the formal apparatus of causal modelling. His slogan — ‘no causes in, no causes out’ — captures the argument in its starkest form.

This paper takes Pearl’s thesis as its point of departure and subjects it to a multi-fronted philosophical dialogue. The first interlocutor is Richard Sutton, whose 2019 essay ‘The Bitter Lesson’ represents perhaps the most influential counter-position in contemporary AI research: the inductive historical claim that general learning methods, scaled with compute, consistently outperform approaches that encode domain knowledge. The second is Bas van Fraassen, whose constructive empiricism offers a sustained philosophical challenge to the metaphysical realism about causal structure that Pearl’s framework presupposes. The third — introduced here as a productive middle ground between Pearl’s formalism and Sutton’s empiricism — is the embodied cognition tradition, represented by Andy Clark, David Kirsh, and Michael Spranger, which argues that genuine causal and spatial understanding requires physical situatedness and social interaction. The fourth is the agent-based simulation tradition, represented by Brian Skyrms and J. Doyne Farmer and grounded philosophically in Michael Bratman’s planning theory of intention, Daniel Dennett’s intentional stance, the BDI computational architecture, and Joshua Epstein’s generative social science, which provides the analogue of embodied causal discovery for the social and economic domains where physical robotics is inapplicable. The fifth is Niklas Luhmann, whose systems-theoretic account of causality as a contingent, system-specific complexity-reducing attribution schema reframes the entire debate in terms of social epistemology, whereby sociology – in my view rightly – claims the mantle of a practically relevant prima philosophia.

The paper proceeds in nine sections. Section 2 reconstructs Pearl’s framework. Section 3 examines its empirical status in light of current reasoning LLMs, drawing on the most recent benchmark evidence on both sides of the debate. Section 4 analyses the Pearl-Sutton confrontation. Section 5 brings van Fraassen’s constructive empiricism to bear. Section 6 introduces the embodied cognition tradition and its implications for physical causal AI. Section 7 introduces the agent-based simulation programmes of Skyrms and Farmer as the functional analogue for social causality. Section 8 presents Luhmann’s framework and its challenge to all prior positions. Section 9 summarizes where the argument stands up to here. Section 10 concludes by arguing that an all-purpose AGI is a misguided goal in an irredeemably polycontextural social world.

2. Pearl’s Causal Revolution: Framework and Claims

Pearl’s starting point is a logical observation about the expressive limits of probability theory. Standard probability calculus can represent the conditional probability P(Y|X) — the probability of Y given that we observe X — but it has no notation for the distinct claim that X causes Y, nor for the quantity P(Y|do(X)) — the probability of Y were we to intervene and set X to some value. This is not merely a notational gap. It reflects a genuine conceptual distinction: observing that patients who take a drug recover more often than those who do not is different from knowing what would happen if we administered the drug.

The Inadequacy of Statistical Language

Pearl’s demarcation line is stark: ‘causal relations cannot be expressed in the language of probability, and hence any mathematical approach to causal analysis must acquire new notation — probability calculus is insufficient’ (Pearl, 2009, p. 40). The implication is that any AI system confined to learning statistical associations from data — however vast the data, however sophisticated the algorithm — operates below the threshold at which genuine causal reasoning becomes possible. Pearl organises causal reasoning into a three-rung hierarchy. Each rung corresponds to a qualitatively different cognitive activity and a distinct class of query that cannot be answered from data at a lower rung without additional assumptions. Here is Pearl’s ‘Ladder of Causation’:

Rung 1 — Association (Seeing): The domain of standard statistics and machine learning. What does observing X tell me about Y? Contemporary deep learning operates primarily at this level.

Rung 2 — Intervention (Doing): What happens to Y if I actively change X? This requires the do-operator. Randomised controlled trials are the canonical experimental instrument; Pearl’s do-calculus provides the formal tools for accessing this rung from observational data combined with a causal model.

Rung 3 — Counterfactual (Imagining): What would have happened to Y if X had been different, given that X actually took value x? This requires the full structural causal model. It is the rung most characteristic of sophisticated human reasoning about causes, attribution, regret, and responsibility.

A fundamental result is that higher-rung queries cannot in general be answered using only lower-rung data. This is not a practical limitation but a mathematical one — and it grounds Pearl’s claim that AI systems confined to statistical learning face a hard ceiling. Pearl’s formal apparatus centres on structural causal models (SCMs), which combine a directed acyclic graph (DAG) encoding causal structure with structural equations specifying how each variable is determined by its parents and noise terms. The do-calculus — a complete set of three inference rules — formalises the distinction between P(Y|X) and P(Y|do(X)), operationalising the act of intervention by deleting all arrows into X and fixing its value from outside the system (Pearl, 1995; 2000). Together, SCMs and the do-calculus constitute the two languages of the causal revolution: graphical models for representing causal assumptions transparently, and the do-calculus for deriving causal and counterfactual conclusions from those assumptions combined with data.

3. The Pearl-Sutton Confrontation: Deduction Against Induction

Richard Sutton’s 2019 essay surveys decades of AI research to argue a recurring pattern: methods incorporating human domain knowledge produce short-term gains but are consistently overtaken by general methods — search and learning — that scale with compute. The lesson is ‘bitter’ because AI researchers repeatedly forget it. The practical implication is a methodological empiricism: do not encode structure, let the system learn its own representations from data.

The Depth of the Disagreement

The Pearl-Sutton confrontation is not a disagreement about empirical facts alone — it is a disagreement about the epistemic type of the constraint Pearl identifies. Sutton’s lesson is an inductive generalisation: it has always been true that scaling wins. Pearl’s argument is deductive: there is a structural mathematical reason why associational data cannot identify causal quantities without causal assumptions. An inductive generalisation, however well-supported, is in principle defeasible. A mathematical impossibility result is not. Pearl is not predicting that scaling will probably fail; he is arguing that certain queries are definitionally unanswerable from certain data, regardless of the learner’s power.

The Empirical Falsification Zone

The current evidence does not support a Suttonian resolution. The 25–40% accuracy drop from interventional to counterfactual tasks in state-of-the-art reasoning models suggests the Rung 2/3 boundary is not eroding under scaling in the way Sutton’s framework predicts. Whether this gap will close with the next generation of models or prove persistent is genuinely open as of 2026 — constituting an empirical falsification zone where the Pearl-Sutton debate will be partially adjudicated over the coming years.

Sutton’s Hidden Architecture Assumption

There is a further point that Sutton’s framework tends to obscure. The claim that ‘general learning methods’ are architecture-neutral is itself questionable. Every learning system operates with a specific architecture, loss function, inductive bias, and training distribution. These design choices constitute the system’s own observational schema. Sutton’s ‘general methods’ are not observations of the world without a standpoint; they are highly specific second-order observations that happen to be very powerful. The bitter lesson may be better read as a lesson about which observational schemas scale well, not as a lesson that observation requires no schema — a point that anticipates what we shall learn below from Niklas Luhmann’s analysis.

4. Reasoning LLMs in 2026: Empirical Status of Pearl’s Thesis

Why LLMs Complicate Pearl’s Original Challenge

LLMs are trained statistically, but their training material contains huge amounts of human causal knowledge: scientific explanations, legal reasoning, technical manuals, everyday narratives, historical accounts, economic analysis, design rationales, and philosophical argument. So an LLM can acquire linguistically encoded causal schemas. The simple dismissal that ‘LLMs only do correlation’ is therefore too crude. Pearl himself acknowledged in a 2023 interview that LLMs trained on text introduced a complication to his original characterisation, conceding that ‘the ladder restrictions do not hold as strictly anymore, because the data is text, and text may contain information on levels two and three’ (Pearl, cited in Pearl and Mackenzie, 2023). A widely cited 2023 paper by Kıcıman, Ness, Sharma, and Tan found that GPT-3.5 and GPT-4 performed surprisingly well on several causal-reasoning tasks, including causal discovery, counterfactual reasoning from text, and event causality. The authors argued that LLMs can help generate causal graphs and extract background assumptions from unstructured text. But they also stressed that LLMs have unpredictable failure modes and that the most promising direction is to combine LLMs with principled causal methods rather than replace causal inference (Kıcıman et al., 2023). So the debate has shifted. The question is no longer whether LLMs can ever utter correct causal claims — they clearly can. The harder question is whether they possess a robust, manipulable causal world-model, or are merely borrowing causal patterns from text without reliable causal competence.

Evidence Supporting Pearl’s Sceptical Side

Several benchmark studies support Pearl’s worry. Jin et al.’s Corr2Cause benchmark asked whether LLMs can infer causation from correlational statements through reasoning rather than commonsense memory. Seventeen LLMs performed close to random; fine-tuning improved in-distribution performance but failed to generalise robustly under perturbations (Jin et al., 2024). CausalBench similarly found that LLMs show promise on simple causal structures but lag behind traditional causal-learning algorithms on larger networks, especially those with more than fifty nodes. It also found that LLMs struggle with collider structures and tend to rely on semantic associations with familiar entities rather than reasoning from numerical distributions or causal structure directly. A NeurIPS 2024 paper, ‘Unveiling Causal Reasoning in Large Language Models: Reality or Mirage?’, is frequently cited by sceptics. Its central finding is that apparent causal competence often collapses under controlled perturbations, suggesting that some LLM ‘causal reasoning’ is pattern-matching or memorised correlation rather than stable structural reasoning (Liu et al., 2024). Recent work on executable counterfactuals (Vashishtha et al., 2025) operationalises Pearl’s three-step counterfactual process — abduction, intervention, and prediction — in code and mathematical problems. The results reveal a substantial accuracy drop of 25–40% when moving from interventional (Rung 2) to counterfactual (Rung 3) tasks for state-of-the-art models including o4-mini and Claude Sonnet. The abduction step is the specific point of failure — precisely the step that has no analogue in forward statistical reasoning.

Clinical reasoning studies of o1, o3, Gemini, Claude, and DeepSeek identify ‘the absence of a robust world model — an internal causal representation that enforces physical and logical constraints, maintains latent state, and supports counterfactual reasoning’ as the underlying cause of systematic failures in medical scenarios designed to disrupt familiar pattern-matching (Griot et al., 2025). Together, this body of evidence supports a qualified Pearlian conclusion: LLMs can imitate, retrieve, and recombine causal reasoning, but they are not yet reliably causal reasoners in the strong structural sense.

Evidence Against a Hard Barrier

The opposing view is not that Pearl is wrong about causality, but that LLMs may become components within causal systems. Several arguments support this position.

First, human causal knowledge is itself often learned from language, testimony, diagrams, examples, and institutionalised discourse. Pearl himself has acknowledged that humans learn much causal information from books, although he insists humans also bring innate causal templates and embodied interaction with the world. The asymmetry between LLM and human causal acquisition may be one of degree and supplementation rather than of fundamental kind.

Second, LLMs can act as causal assistants: proposing variables, identifying possible confounders, drafting causal graphs, explaining assumptions, checking counterfactual scenarios, and translating expert knowledge into formal causal models. This is not AGI by itself, but it is genuinely valuable. The most promising current research direction is hybridisation: LLMs combined with causal discovery algorithms, simulators, experimental design tools, embodied agents, reinforcement learning, tool use, and explicit structural causal models.

Third, reasoning models in 2025–2026 have improved substantially at multi-step problem solving, code execution, tool use, and self-checking. This does not dissolve Pearl’s challenge, but it weakens the older claim that neural systems are merely passive curve-fitters. They can increasingly search, test, revise, simulate, and call external tools. The open issue is whether this becomes a genuine causal architecture or merely a more elaborate association machine wrapped around tools.

Reading the Current Balance

Pearl’s challenge remains valid, but the phrase ‘hard barrier’ needs refinement. A pure LLM, trained only to predict or transform text, is unlikely to be sufficient for robust AGI if AGI means autonomous scientific, legal, engineering, political, or everyday practical reasoning under novel interventions. Such a system needs stable representations of actions, mechanisms, constraints, affordances, temporal processes, and counterfactual alternatives — and the empirical evidence consistently shows that current LLMs lack precisely these stable representations. But this does not imply that neural models or LLMs are irrelevant to AGI. The more plausible view is that LLMs are powerful discourse engines and knowledge interfaces, but AGI will require them to be embedded in architectures with causal modelling, memory, active experimentation, world interaction, simulation, and explicit decision procedures. Pearl is probably right against AGI by correlation-crunching alone, but not necessarily against AGI architectures that use LLMs as one component within a broader causal-agentic system. This refined position has significant implications for the comparisons developed in the remainder of this paper. The architectures most likely to produce genuine causal AI are precisely those that embed LLM components within frameworks providing the elements LLMs lack: structured causal models in the Pearlian tradition; embodied robotic systems for physical causality, in the tradition of Clark, Kirsh, and Spranger; and agent-based simulation environments for social and economic causality, in the tradition of Skyrms and Farmer. The frontier is not a contest between paradigms but their principled integration.

5. Van Fraassen’s Constructive Empiricism and the Metaphysics of Causation

Bas van Fraassen’s The Scientific Image (1980) argues that the aim of science is empirical adequacy rather than truth. A theory is empirically adequate if it correctly describes what is observable; whether its theoretical posits correspond to reality is a question we need not answer. Van Fraassen’s anti-realism is epistemological: he does not deny that unobservable entities might exist, only that we have adequate grounds for believing in them on the basis of theoretical success.

Van Fraassen’s Challenge to Pearl

Pearl’s framework makes strong metaphysical commitments. SCMs are intended as representations of real causal structure. DAGs are supposed to encode genuine asymmetric causal relationships, not merely useful predictive summaries. The do-operator presupposes facts of the matter about what would happen under interventions that never occur. Counterfactuals assert truths about worlds that did not come to pass. Van Fraassen would subject each commitment to critical scrutiny. A constructive empiricist can accept a DAG as a tool organising inference about observables without committing to the claim that its arrows represent real causal powers. On counterfactuals — paradigmatically unobservable — van Fraassen has no clean way to ground such claims in experience. His earlier work on explanation (van Fraassen, 1980, ch. 5) argues that what counts as the cause of an event depends on the contrast class pragmatically determined by the questioner, not fixed by the world. Pearl’s do-calculus, by contrast, aims to give objective, context-independent answers once the causal graph is fixed.

Partial Convergence and the Instrumentalist Reading

SCMs can be understood instrumentally: not as literal maps of nature’s causal joints, but as representational tools enabling a certain class of inferences. This reading is more congenial to van Fraassen than Pearl himself would endorse, but the practical framework of do-calculus and identification theory is largely independent of the metaphysical question. The deeper convergence concerns the inadequacy of pure statistical methods: van Fraassen’s pragmatic account of explanation, while different from Pearl’s causal realism, shares the diagnosis that explanation cannot be read off from distributions alone.

6. Embodied Cognition, Robotics, and the Grounding of Causal Knowledge

The embodied cognition tradition represents a third position in the debate between Pearl’s formalism and Sutton’s scaling optimism — one that has received insufficient attention in discussions of AI causality. Andy Clark, in works from Being There (1997) to Supersizing the Mind (2008) and Surfing Uncertainty (2016), has consistently argued that cognition is not a process occurring inside a skull but is constitutively extended across brain, body, and environment. Bodily action is not merely the output of computation — it is, in Clark’s formulation, among the means by which computational and representational operations are implemented. The mind extends into the world.

The implication for causal cognition is direct and under-explored. If causal understanding in humans is grounded in sensorimotor interaction with the world — in pushing, pulling, lifting, dropping, releasing, and observing consequences — then an AI system trained exclusively on text or even video is missing the medium in which causal knowledge is originally constituted. It has access to the verbal and symbolic residue of causal experience without the experience itself. Pearl’s ladder, on this reading, cannot be climbed from inside a language model because the ladder is built from embodied intervention, not from description of intervention.

Kirsh and the Epistemic Role of Action

David Kirsh’s work at UCSD on situated and distributed cognition develops this point with particular precision. In his influential paper on epistemic versus pragmatic actions (Kirsh and Maglio, 1995) and in his subsequent work on embodied cognition and interaction design (Kirsh, 2013), Kirsh distinguishes between actions performed to achieve a physical goal and actions performed to improve the cognitive conditions for solving a problem. Agents — human and potentially robotic — restructure their environments not just to accomplish tasks but to create better epistemic affordances: to make causal structure visible, tractable, and manipulable. This has a direct bearing on Pearl’s hierarchy. Rung 2 — the do-operator, the intervention — is not merely a logical operation performed on a causal graph. It is, in its original cognitive form, a physical act: doing something to the world and observing what follows. Kirsh’s framework suggests that the capacity to perform genuine interventional reasoning is inseparable from the capacity to perform genuine interventions. A system that can only predict what would happen under intervention, without ever having intervened, may be performing sophisticated Rung 1 inference about Rung 2 content — not genuine Rung 2 reasoning.

Spranger and the Evolution of Grounded Spatial Language

Michael Spranger’s The Evolution of Grounded Spatial Language (2016), published by Language Science Press as part of the Computational Models of Language Evolution series, provides a particularly compelling empirical demonstration of these principles. Spranger reports a series of robotic experiments in which agents equipped with sensorimotor capacities and placed in shared social environments evolve spatial language — systems of reference for describing the location of objects relative to landmarks, frames of reference, and projective directions — through communicative interaction. What is striking about Spranger’s results for the present debate is that the spatial conceptualisation strategies the robots evolve are not pre-programmed, nor are they derived from a statistical distribution over language data. They emerge from the interplay of environmental conditions, embodied perception, communicative pressure, and social feedback. The agents must, in Pearl’s terms, reach Rung 2: they intervene in their environment, observe consequences, communicate about those consequences, and revise their conceptual schemes accordingly. The causal and spatial categories they develop are genuinely grounded — they are constituted by the sensorimotor history of the agents in their environment, not by pattern-matching over symbols. Spranger’s work also demonstrates that compositional semantics — the capacity to combine primitive spatial concepts into complex descriptions — emerges naturally from embodied interaction, without being explicitly engineered. This has implications for the grounding problem in AI: the reason language models may struggle to acquire genuine causal concepts is not merely that they lack a formal causal graph, but that their representations are not grounded in the sensorimotor cycles that give causal concepts their content in the first place.

Causal Robotics: Embodied Causal Discovery

A growing body of research in causal robotics operationalises these insights computationally. The 2023 IROS workshop on Causality for Robotics articulated the challenge explicitly: ‘correlation-based methods often lack robust generalisation, are brittle to distribution shifts, and can learn incorrect, spurious relationships in observed data. In contrast, a hallmark of human intelligence — reasoning about cause and effect — can provide clues for the next generation of robot intelligence’ (IROS Causality for Robotics, 2023). Concrete implementations have appeared across several domains. Brawer et al. (2020) demonstrated that humanoid robots can build and learn structural causal models from a mix of observation and self-supervised physical trials for tool affordance — precisely the embodied version of Pearl’s Rung 2. Castri et al. (2022; 2023) applied causal discovery algorithms to model and predict human spatial interactions in social robotics contexts, showing that causal structure can be recovered from the time-series of an embodied agent operating in a social environment. The ADAM framework (2024) integrates causal methods and embodied agents in open-world environments, deploying both LLM-based and intervention-based causal discovery within a system that can navigate, perceive, and act. The Causal-HRI workshop at the ACM/IEEE International Conference on Human-Robot Interaction (HRI 2024) brought together researchers working specifically on causal learning for human-robot interaction, recognising that the social task environment — where a robot must understand human intentions, predict the consequences of joint actions, and explain its own behaviour — is precisely the kind of domain where the limitations of purely statistical AI become most apparent and most consequential.

Embodied Cognition as a Third Way

The embodied cognition tradition suggests a path between Pearl’s formalism and Sutton’s scaling optimism that neither fully anticipates. Pearl is right that statistical learning cannot substitute for causal structure — but he tends to locate the solution in explicit formal models imposed from outside the learning system. Sutton is right that hand-crafted representations are limiting — but he underestimates the degree to which the right kind of general learning requires the right kind of embodied situation. The embodied tradition suggests that causal understanding can be learned — contra Pearl’s formalism — but only by systems that are physically situated, action-capable, and socially embedded — contra Sutton’s architecture-agnosticism. This also illuminates the specific failure mode of current LLMs on Rung 3 tasks. These systems have access to vast quantities of human causal discourse — descriptions of interventions, explanations of mechanisms, counterfactual reasoning — but they lack the sensorimotor grounding that gives such discourse its original content. They are, in a sense, reading the map without ever having visited the territory. The abduction step — inferring latent states from observations in order to reason about what would have happened under different conditions — requires precisely the kind of model of the world that is constituted through embodied interaction, not through symbolic compression.

7. Agent-Based Interaction Simulations: The Social Sciences’ Analogue of Embodied Robotics

The embodied cognition tradition makes a compelling case that genuine causal understanding of the physical world requires sensorimotor grounding — the capacity to push, pull, release, and observe in a physical environment. Spranger’s robots evolve spatial language through bodily interaction with objects and space; Brawer’s humanoid robot builds causal models through physical trial and error with tools. This insight carries real force for the domain of physical world causality: understanding that a falling weight causes deformation, that heating causes expansion, that friction causes deceleration, is plausibly constituted through the sensorimotor history of an agent in a mechanical environment.

However, a significant portion of the causal questions that matter most to science, policy, and human welfare belong not to the physical domain but to the social domain: sociology, economics, organisational theory and political science. In these domains, the relevant causal structures are not constituted by sensorimotor coupling with objects in space. They are constituted by the interaction of agents with beliefs, strategies, expectations, and institutional rules — entities that have no direct physical correlate accessible to a manipulating arm or a rolling chassis. No amount of robotic embodiment will teach an agent that raising interest rates causes inflation to fall, that inequality causes social fragmentation, or that network effects cause market concentration. These causal relationships emerge from the dynamics of interacting agents operating within social systems, not from the dynamics of forces acting on physical bodies. This domain asymmetry is crucial for the present debate. It does not undermine the embodied cognition argument for physical causal understanding; it limits its scope. And it raises a question that neither Pearl’s formal SCMs nor Sutton’s scaling optimism nor Clark and Kirsh’s embodied frameworks directly address: what is the functional analogue of embodied intervention, for AI systems that need to discover and reason about causal structure in the social domain? The answer proposed here is agent-based interaction simulation.

Brian Skyrms: Agent-Based Simulation as Causal Discovery in Social Systems

Brian Skyrms’s programme of evolutionary game-theoretic simulation — developed across Evolution of the Social Contract (1996), The Stag Hunt and the Evolution of Social Structure (2004), Signals: Evolution, Learning, and Information (2010), and Social Dynamics (2014) — represents the most systematic philosophical and computational programme for discovering causal structure in social systems through agent-based interaction. The core method is to populate a simulated social environment with agents endowed with simple decision rules and strategic repertoires, allow them to interact repeatedly under various conditions, and observe which social structures, norms, conventions, and communication systems emerge from the dynamics of those interactions. The causal claims Skyrms derives from these simulations are Rung 2 and Rung 3 claims in Pearl’s sense. They are not merely observational correlations between social variables. They are claims about what happens when specific parameters are changed — when the network structure of interactions is altered, when the population size varies, when signals are introduced or removed. In Pearl’s terms, the simulated agent populations are performing do(X) operations: the researcher intervenes on the model environment and observes the resulting distribution of outcomes. The simulations thus generate interventional and counterfactual knowledge about social causal structure that no amount of observational data about real social systems could supply, precisely because real social systems do not permit the controlled interventions that the simulations make tractable.

Skyrms’s programme is also directly relevant to the grounding problem for social causal concepts. Consider the concept of a social norm: the claim that a particular norm causes cooperative behaviour, or that its absence causes defection, is not a physical causal claim. It is a claim about the functional role of a shared expectation in an interactive environment. Skyrms’s simulations demonstrate how such norms can emerge spontaneously from agent interactions without being pre-programmed — how the causal efficacy of norms is constituted through the dynamics of interaction, not through any physical mechanism. This is precisely analogous to the way Spranger’s robots’ spatial concepts are constituted through bodily interaction with physical environments: in both cases, the causal concept is grounded in a history of structured interaction rather than being imposed from outside as a formal stipulation. The specific causal mechanisms Skyrms illuminates — the Stag Hunt coordination problem as the fundamental structure of social contract formation, the role of network topology in enabling or impeding cooperation, the spontaneous emergence of meaningful signaling systems from initial states of noise — are paradigmatic examples of Rung 2 and Rung 3 social causal knowledge. They cannot be obtained by training a language model on descriptions of human cooperation. They require running the interactions.

J. Doyne Farmer: Agent-Based Simulation for Economic Causality

J. Doyne Farmer’s programme of complexity economics at the Institute for New Economic Thinking at Oxford provides the most developed application of agent-based simulation to the causal structure of economic systems. In the foundational manifesto Farmer and Foley (2009), ‘The Economy Needs Agent-Based Modelling,’ published in Nature, Farmer and collaborator Duncan Foley argue that the standard tools of macroeconomics — dynamic stochastic general equilibrium models, representative agent frameworks, equilibrium analysis — are constitutively incapable of capturing the causal mechanisms responsible for the phenomena economists most need to understand: financial crises, endogenous business cycles, systemic risk, and the dynamics of technological change. The core problem Farmer identifies is precisely the one Pearl names: standard economic models are built on statistical associations and equilibrium conditions that do not represent causal mechanisms. A DSGE model that correctly predicts the correlation between interest rates and inflation under normal conditions may entirely fail to predict what happens when a financial crisis disrupts the causal mechanisms underlying that correlation — because the model has captured the correlation without representing the mechanism. An agent-based model, by contrast, builds in the causal mechanisms from the bottom up: heterogeneous agents with bounded rationality, making decisions under uncertainty, interacting through networks of supply chains, credit relationships, and market exchanges. The macroeconomic regularities are emergent properties of these micro-level causal interactions, not statistical regularities imposed from above.

Farmer’s Macrocosm project — a large-scale simulation of the global economy representing individual firms and their assets as distinct agents — operationalises this programme at unprecedented scale. As of 2026, the simulation encompasses tens of thousands of companies and their capital assets, with agents making decisions based on simple rules including imitation and trial-and-error learning, in explicit contrast to the fully-informed optimal agents of computable general equilibrium models (Farmer, 2024; Wikipedia, 2026). The causal ambition is clear: Farmer describes the goal as doing ‘for economic planning what Google Maps did for traffic planning’ — providing a model in which causal interventions can be simulated, their consequences traced through the network of agent interactions, and counterfactual scenarios evaluated. This is a direct operationalisation of Pearl’s Rung 2 and Rung 3 for the economic domain.

Farmer and Axtell’s comprehensive survey ‘Agent-Based Modeling in Economics and Finance: Past, Present, and Future’ (2022) documents the maturation of this programme across financial markets, housing markets, macroeconomic stability, and climate economics. Crucially, Farmer explicitly contrasts agent-based modelling with machine learning in terms of causal adequacy, noting that while machine learning ‘builds on the world as it is and it’s unclear if it can handle an alternative situation where it hasn’t seen data,’ an agent-based model ‘has the causal mechanisms built in’ and can therefore address interventional and counterfactual questions that purely statistical models cannot (Farmer, interview, 2024). This is an independent, practitioner-level restatement of Pearl’s argument, arrived at through the practice of economic modelling rather than through philosophy of causation.

The Intentional Foundations of Social Causation: Bratman, Dennett, and BDI Architectures

To understand why agent-based simulation captures something essential about social causality that statistical learning cannot, it is necessary to address the philosophical question of what kind of causation operates in the social domain. The most direct philosophical inspiration for agent-based social simulation comes from Michael Bratman’s planning theory of intention, developed in Intention, Plans, and Practical Reason (1987). Bratman argued that intentions are not reducible to beliefs and desires — they constitute a distinct mental state with three characteristic properties: they persist over time, settling what an agent will do and freeing cognitive resources for other tasks; they constrain future deliberation, such that rational agents do not constantly re-evaluate settled plans; and they coordinate with other plans, both within an agent’s own plan hierarchy and across agents. This planning theory directly inspired the BDI (Belief-Desire-Intention) computational architecture, formalised by Rao and Georgeff in the early 1990s. In BDI agents, beliefs represent the agent’s model of the world; desires represent motivational states (goals); and intentions represent committed plans currently being executed. The architecture is explicitly designed to capture the distinctive feature of practical rationality that Bratman identified: rational agents do not reason from scratch at every moment — they commit, persist, and revise only when necessary. This is not a peripheral feature of human social cognition; it is constitutive of what it means to be a planning agent capable of stable cooperation, reliable promising, and meaningful coordination.

Daniel Dennett’s intentional stance, articulated in The Intentional Stance (1987), provides a complementary framing that is essential for understanding the methodological status of BDI-based simulations. Dennett argues that attributing beliefs, desires, and intentions to a system is a predictive strategy adopted when it provides the most useful level of description for anticipating behaviour. The intentional stance is instrumentally justified — it does not require commitment to the agent ‘really having’ beliefs in some deep metaphysical sense, only that the stance generates good predictions. It is level-relative — the same system can be described at the physical, design, or intentional level depending on what is explanatorily useful. And it is generative — it licenses attributing whatever beliefs and desires a rational agent in a given situation ought to have. This dual philosophical foundation matters enormously for agent-based social simulation. When researchers model agents with BDI architectures, they are operationalising the intentional stance: they are building systems in which the intentional level of description is causally efficacious within the simulation. The question of whether such simulations capture genuine causal structure then becomes a precise empirical-philosophical question: does the simulation’s intentional-level causation track real-world intentional-level causation? This is the social-domain analogue of the question whether Spranger’s robots’ spatial categories track real spatial relationships in the physical world.

The Strategy Layer: From Intentions to Coordinated Action

Between intentions and observable action lies a crucial intermediate level — strategy — where game theory, norm theory, and institutional analysis enter. Several frameworks operate at this layer and inform the design of sophisticated agent-based models. Schelling’s analysis of focal points and tacit coordination demonstrates that agents with compatible intentions can coordinate without explicit communication by reasoning about what a rational counterpart would expect. This is a paradigmatically intentional-stance phenomenon: the explanation of successful coordination depends essentially on attributing to each agent reasoning about the other’s reasoning, which has no purely physical or statistical reduction. Skyrms’s signalling-game work builds directly on this insight, showing how meaningful communication can evolve from initial states of noise through the dynamics of agents reasoning about each other’s expected behaviour.

Elinor Ostrom and Sue Crawford’s Institutional Analysis and Development (IAD) framework treats institutions as shared rules that structure the action situations agents face. Institutions, on this account, operate precisely at the BDI level: they shape which beliefs are relevant, which desires are legitimate, and which intentions are actionable. Cristina Bicchieri’s The Grammar of Society (2006) provides the philosophical grounding for the analysis of social norms as conditional preferences supported by empirical and normative expectations — themselves intentional-stance objects requiring the recursive attribution of beliefs about beliefs. Sophisticated social agency further requires Theory of Mind layering — recursive intentionality of the form ‘I believe that you intend that I believe...’ This recursion is computationally expensive but essential for realistic modelling of deception, trust, signalling, and coordination. The capacity to reason about another agent’s reasoning about one’s own reasoning is what makes possible the entire architecture of social life: contracts, promises, threats, reputations, and the conditional cooperation that sustains complex institutions. None of these are reducible to statistical regularities in observable behaviour; all of them require the intentional-level description that BDI architectures operationalise.

Generative Social Science: Epstein, Sugarscape, and Cognitive Architectures

The most influential demonstration that macro-level social phenomena can be ‘grown’ from intentional micro-level interactions is Joshua Epstein and Robert Axtell’s Growing Artificial Societies (1996), which introduced the Sugarscape model. Sugarscape demonstrated that complex social patterns — trade networks, wealth inequality, cultural differentiation, demographic dynamics, and even rudimentary forms of inheritance — could emerge from populations of agents endowed with simple beliefs (about resource distribution), desires (to gather sugar, to avoid conflict), and intention-like persistence (movement strategies, cultural tags). None of these macro-patterns were explicitly programmed; they emerged from the dynamics of intentional-level interaction. Epstein subsequently extended this programme explicitly toward BDI-style agents in Agent-Based Computational Models and Generative Social Science (2006), articulating the methodological claim that what cannot be ‘grown’ from agent interactions has not been explained — a position he termed the generativist motto. This is a direct social-domain parallel to Pearl’s claim that what cannot be derived from causal assumptions and the do-calculus has not been causally established. Both Epstein and Pearl insist that genuine causal understanding requires more than statistical capture: it requires the demonstration that the proposed mechanism actually produces the observed phenomenon when implemented.

Schelling’s segregation model — predating but later canonised within agent-based modelling — provides a particularly clear illustration of why intentional-stance attribution is essential at the micro level. Mild individual preferences for neighbourhood composition produce strong macro-level segregation, and the explanation of this outcome depends essentially on attributing to each agent a preference (an intentional state) about its neighbours. No purely statistical description of the macro outcome captures the causal mechanism; only the intentional-stance description of agent preferences does. David Hales’s work on tag-based cooperation extends this further, showing how arbitrary identity markers can become the basis of stable cooperation through agents reasoning about expected behaviour conditional on observable tags.

Cognitive architectures such as SOAR (Newell) and ACT-R (Anderson) push the integration further, embedding BDI-like structures within full cognitive frameworks that include working memory, procedural memory, and learning. These architectures have been applied to model military command decision-making, air traffic control errors, and learning in complex environments — domains where the intentional-level causal structure of the situation cannot be bypassed by statistical methods, because the relevant phenomena (commitment, revision, coordination under stress) are constitutively intentional.

Why Intentional-Level Causation Matters for AI World Models

The philosophical and empirical case converges on a key point: many of the most important causal mechanisms in social systems operate at the intentional level. Physical-level descriptions miss them because the relevant entities — beliefs, intentions, expectations, norms — have no straightforward physical correlates. Statistical-level descriptions miss them because the relevant causal pathways pass through the rational decision-making of agents who interpret their situations rather than merely respond to stimuli. A genuine causal world model for social phenomena must represent agent belief states, including false beliefs, uncertainty, and belief updating; hierarchical goal structures, with immediate desires embedded in longer-term plans in Bratman’s sense; commitment and persistence, capturing the fact that rational agents do not re-optimise from scratch; social and normative expectations, including what agents believe others expect of them; and institutional scaffolding — the shared rules that transform the effective action space.

Current LLM-based approaches gesture toward this through in-context theory of mind, but they lack the causal persistence and commitment structure that Bratman identifies as essential to planning agency. An LLM may produce text that describes an agent forming an intention, but it does not itself instantiate the intentional structure — its representations of agent intentions are linguistic descriptions rather than causally efficacious commitments that constrain its own future processing. Agent-based simulation with explicit BDI architectures offers a more principled path: one in which intentional-level causation is constitutive of the model rather than merely emergent from statistical patterns over text descriptions of intentional behaviour.

This connects directly to the diagnoses of LLM failure surveyed in Section 3. The 25–40% accuracy drop on counterfactual tasks, concentrated at the abduction step, is precisely the kind of failure one would predict for systems lacking explicit intentional-level causal structure. To answer ‘what would have happened if the agent had believed X instead of Y?’ requires that the agent’s beliefs be represented as causally efficacious states whose modification has determinate consequences for subsequent reasoning. This is exactly what BDI architectures provide and what unaugmented LLMs do not.

The deepest methodological challenge for this programme is validation: how to demonstrate that the intentional-level causal structure within a simulation genuinely mirrors that in the real social world, rather than merely producing surface-level outcome similarities. This is a problem requiring both empirical social science — comparing simulation outputs against natural and quasi-experimental data — and careful philosophical analysis of what it means for a model to capture rather than merely predict social causation. It is the social-domain analogue of the validation problem for embodied robotics: confirming that the simulated agent’s grasp of physical causality genuinely transfers to action in the unstructured world. In both cases, the answer is not provided by the simulation alone but by the disciplined dialogue between simulation, empirical observation, and theoretical analysis that constitutes mature scientific practice in the relevant domain.

Agent-Based Simulation as Social Embodiment

With these philosophical foundations in place, the parallel between physical embodiment and agent-based social simulation can now be stated more precisely. In both cases, the claim is that genuine causal understanding of a domain requires the system to engage with that domain through structured interaction rather than through passive observation of descriptions of interactions. For physical causality, the relevant medium is sensorimotor coupling with a physical environment — robots that push, pull, and observe. For social causality, the relevant medium is agent-based simulation of intentionally-constituted agents in a structured social environment — populations that form beliefs, commit to plans, coordinate intentions, signal, bargain, and revise.

In both cases, the interactive process generates a history of structured action-outcome pairs from which causal structure can be extracted. In both cases, the causal concepts that emerge — spatial relations for physical agents, social norms and intentional dynamics for simulated social agents — are grounded in the interaction history rather than stipulated from outside. And in both cases, the resulting causal knowledge supports interventional and counterfactual reasoning that purely statistical analysis of observational data cannot provide. The crucial addition the present section makes to the embodied cognition argument is that social embodiment requires intentional-level architecture: it is not enough to have agents that interact; the agents must interact as planning, believing, intending entities, because that is what social causation is constituted by.

This framing also clarifies a relationship between Skyrms’s social philosophy and Farmer’s economics that is rarely made explicit. Both are engaged in the same epistemic project: the discovery of causal structure in complex social systems through the controlled manipulation of simulated environments populated by intentional agents. Skyrms focuses on the evolutionary origins of norms, cooperation, and communication; Farmer focuses on the emergent dynamics of financial and macroeconomic systems. Both use agent-based simulation as the medium through which social causal structure becomes legible. And both arrive at the same methodological conclusion that Pearl reaches from the philosophy of causation: passive observation of social correlations, however richly data-supported, cannot substitute for the active exploration of causal mechanisms through intervention upon agents whose intentional-level dynamics generate the macro-scale phenomena.

Implications for AI Systems in the Social Domain

The implication for AI systems designed to reason about social and economic causality is direct. Such systems will not acquire genuine causal understanding of social phenomena by training on text descriptions of social events, policy interventions, and economic outcomes — however vast the corpus. They will acquire borrowed causal knowledge: the accumulated causal models of human social scientists encoded in language. Pearl’s concession about text-trained LLMs applies here with full force, but with an important addition derived from the intentional-stance analysis: the failure is not merely that LLMs absorb causal claims without causal architecture, but specifically that they absorb intentional-level descriptions without instantiating intentional-level structure. The causal mechanisms of social life pass through agents who plan, commit, and revise; an AI system without these structures cannot reliably reason about phenomena whose causation is constituted by them.

The path toward genuine AI causal understanding in the social domain runs through agent-based simulation augmented by explicit BDI or comparable intentional architectures: through systems that interact within simulated social environments not merely as statistical learners but as planning agents whose own intentional structure mirrors that of the agents they aim to understand. This is not the same as the physical embodiment of Clark, Kirsh, and Spranger — the relevant environment is social and institutional, not physical and spatial, and the relevant grounding is intentional rather than sensorimotor — but it is precisely analogous in its epistemic structure. The frontier research programme is the integration of the formal apparatus of Pearl’s causal hierarchy with the generative power of large-scale agent-based simulation populated by intentionally-constituted agents, creating systems that can not only describe social causal structure but actively explore, intervene upon, and reason counterfactually about it from within the intentional standpoint that constitutes the structure itself.

From a Luhmannian perspective, there is an additional reflexive dimension to this programme that deserves acknowledgement. Agent-based simulations of social systems are themselves second-order observations — models of how social systems operate, constructed using the distinctions of particular scientific communities. The causal knowledge they generate is system-relative in Luhmann’s sense: it is the causal structure of the social world as seen through the observational schema of complexity economics or evolutionary game theory or BDI cognitive architecture. This does not make it arbitrary or merely subjective; it makes it revisable through the evolution of the scientific system’s internal standards. The productive tension between the formal rigour of Pearl’s causal framework, the intentional-stance philosophy of Bratman and Dennett, and the generative empiricism of Skyrms’s, Farmer’s, and Epstein’s simulations is precisely the kind of internal scientific debate that mature scientific practice requires.

8. Luhmann’s Systems Theory: Causality as Complexity Reduction

Causality as a Selective Schema

Niklas Luhmann’s earliest and most directly relevant treatment of causality appears in his 1962 article ‘Funktion und Kausalität,’ published in the Kölner Zeitschrift für Soziologie und Sozialpsychologie, 14, pp. 617–644, and reprinted in Soziologische Aufklärung 1 (Springer Verlag, 1970). This foundational essay establishes the key move that Luhmann will sustain throughout his mature systems theory: causality is not a feature of the world to be discovered or a tool to be used, but a functional schema that systems employ to reduce complexity. The world, in Luhmann’s framework, is characterised by overwhelming causal complexity: virtually everything is connected to virtually everything else. No observing system can process this totality. Causal attribution is one of the primary mechanisms through which systems select from this complexity, singling out a small number of antecedents and effects as relevant. This move is sustained and developed across Luhmann’s major works, including Social Systems (1984/1995), The Science of Society (1990/1992), and Risk: A Sociological Theory (1991/1993). Causal attribution, for Luhmann, is not pragmatic in van Fraassen’s sense — not simply a matter of different questioners finding different causes relevant to their interests. It is a constitutive operation of the observing system: it is how the system constructs its environment as an environment.

Second-Order Observation

Central to Luhmann’s account is the concept of second-order observation: observing not the world directly but observing how other systems observe the world. Both first-order and second-order observations are always communicative operations of semi-stable, self-reproducing (autopoietic), evolving social systems with their system-specific communication structures which include specific, highly selective discriminations. Every observation employs a distinction — a schema that divides the world into a marked and an unmarked side — and this distinction cannot itself be observed from within the observation that uses it. The blind spot of every observation is the distinction it relies upon. Applied to Pearl’s framework: a causal graph is an observation made using specific distinctions — cause versus background condition, relevant versus irrelevant variable, direct versus indirect effect. Pearl presents these distinctions as recoverable from the causal structure of the world. Luhmann would say they are produced by the observing system, shaped by its history, disciplinary conventions, and research programmes. This does not make them arbitrary — it makes them system-relative, which is different.

Luhmann’s Challenge to Each Interlocutor:

Against Pearl: Every SCM is a second-order observation — a model of causal structure as seen from the standpoint of a particular scientific community using particular distinctions. Different communities, asking different questions with different schemas, would produce different graphs. There is no view from nowhere that would validate one graph as uniquely correct.

Against Sutton: Every learning architecture embeds distinctions — what counts as a feature, a prediction, an error. These constitute the system’s horizon of observability. Sutton confuses a very powerful second-order observation for a first-order contact with the world as it is.

Against van Fraassen: The distinction between observable and unobservable is itself produced by the observing system’s internal operations, not found in the world independently. This dissolves van Fraassen’s distinction between empirically adequate and literally true theories more radically than van Fraassen himself intends.

Against the embodied cognition tradition: While Luhmann would find the embodied cognition programme more sympathetic than the others — since it foregrounds the system-environment coupling that produces cognitive operations — he would note that even embodied observation is bounded by the distinctions the organic and psychic system brings to its environmental interactions. Spranger’s robots evolve spatial language through embodied interaction, but the conceptual schemas they develop are still selected from the space of possibilities opened by their architecture. There is no schema-free embodied observation any more than there is schema-free statistical learning.

Against the agent-based simulation tradition: Skyrms’s and Farmer’s simulations are themselves second-order observations — models of social and economic systems constructed using the distinctions of evolutionary game theory and complexity economics. The causal knowledge they generate is system-relative: it is the causal structure of the social world as seen through a particular observational schema. This does not undermine the value of such simulations — they remain among the most productive available methods for accessing social causality — but it situates them within the same reflexive condition that applies to all observation.

The Acyclicity Constraint: Pearl’s Ontological Mismatch with Self-Reproducing Social Structures

From Luhmann’s perspective we can launch another challenge to Pearl’s conception and formalisation of causality which must be remedied with respect to the social domain. A feature of Pearl’s structural causal models that has received insufficient critical attention is its most elementary formal requirement: the directed acyclic graph. It is a constitutive constraint on the ontology that SCMs can represent. A directed acyclic graph forbids cycles: no variable may be its own ancestor. This means that every causal pathway terminates; there are no self-sustaining loops, no structures that reproduce themselves through their own operations. The do-calculus, the identification theory, and the three-step counterfactual procedure — abduction, intervention, prediction — all presuppose this acyclicity. Without it, the mathematical apparatus that underwrites Pearl’s causal revolution breaks down: interventions cannot be cleanly isolated, parent sets become ill-defined, and the truncated factorisation that gives the do-operator its content ceases to hold.The question this section poses is: what kind of causal world does the acyclicity requirement presuppose, and what does it exclude?

An Ontology of Events, Not Structures

Pearl’s SCMs are built for a world of discrete variables with determinate values — drug doses, policy switches, temperatures, test outcomes. Each variable takes a value that is determined by its parents and an independent noise term. The resulting ontology is punctuated and event-centred: causes are settings of variables; effects are the values that result. This maps naturally onto the experimental ideal of the randomised controlled trial, where the experimenter sets one variable and measures another. It maps less naturally onto the entities that populate the social and institutional world. Consider a market, a legal system, a social norm, or a scientific research programme. These are not events that occur and terminate; they are structures that persist through time by continuously reproducing the conditions of their own operation. A market does not cause prices in the way a drug causes recovery. A market is the ongoing process through which prices are generated, and the generation of prices is itself one of the operations through which the market reproduces itself as a market. The causal loop is not incidental; it is constitutive.

The standard Pearlian response is temporal unrolling: represent X at time t as causing Y at time t+1, which causes X at time t+2, transforming the apparent cycle into an acyclic sequence indexed by time steps. Dynamic Bayesian networks formalise precisely this manoeuvre. And for many purposes it works: if one is interested in the effect of a central bank’s interest-rate decision at t on inflation at t+1, the unrolled representation is adequate. But the manoeuvre conceals a deeper problem. It describes what happens inside the loop while remaining silent on why the loop exists at all. The temporal unrolling tells us the sequence of states; it does not tell us why that sequence is self-sustaining rather than dissipative, why the structure persists rather than decays. The persistence of the structure — its autopoietic self-reproduction, in Luhmann’s terms — is precisely the phenomenon that requires explanation, and it is precisely the phenomenon that the acyclicity requirement renders formally invisible.

Three Categories of Causal Entity That Resist Acyclicity

The mismatch between DAG ontology and social reality can be made precise by identifying three categories of causal entity that are poorly served by the acyclicity constraint.

First, self-reproducing institutions. Luhmann’s autopoietic systems — law, economy, science, politics, education, art — are defined by the property that they produce and reproduce the elements of which they consist through a network of those very elements. The legal system produces legal communications (judgements, statutes, contracts) through legal communications; the scientific system produces truth-claims through truth-claims. The circularity is not a defect to be linearised; it is the mode of existence of the system. An SCM that represents the legal system must either break this circle — misrepresenting the phenomenon — or violate acyclicity.

Second, structural attractors. In complex systems theory, an attractor is a state or set of states toward which a dynamical system tends to evolve and to which it returns after perturbation. Poverty traps, institutional equilibria, and path-dependent technological standards are attractors in this sense: they actively maintain themselves against perturbation through feedback mechanisms that are constitutively circular. Modelling such attractors as acyclic causal chains captures the local dynamics of any given perturbation-and-return but misses the global feature that defines the attractor: its self-maintenance. The causal power of an attractor consists precisely in its capacity to close the loop, to ensure that departures from the pattern are corrected by forces that the pattern itself generates. This circularity is the explanation, and it cannot be represented in a DAG without remainder.

Third, constitutive relations. In many social phenomena, A does not merely cause B; B is partly constitutive of what A is. A market does not merely cause the price mechanism to operate; the price mechanism is partly constitutive of what it means to be a market. Bratman’s planning intentions do not merely cause coordinated action; the possibility of coordinated action is partly constitutive of what it means to have a planning intention. In these cases, the causal arrow presupposes the existence of the very entity it purports to explain, creating a conceptual circularity that is not temporal but ontological. No amount of temporal unrolling can linearise a constitutive relation, because the relation is not between events at different times but between an entity and its own conditions of possibility.

Consequences for Causal AI in the Social Domain

The implications for the broader argument of this paper are substantial. The acyclicity constraint is not merely a technical limitation that future extensions of Pearl’s formalism might overcome. It reveals that SCMs are ontologically mismatched with the entities that Luhmann identifies as the primary constituents of modern social reality: operationally closed, self-reproducing functional systems. Pearl’s formalism is a theory of causes among events; modern social reality is constituted by structures that persist through self-referential operations. These are different kinds of entity, and they require different formal apparatus.

This does not render Pearl’s framework useless for social science. Within bounded contexts — where one can reasonably treat institutional structure as exogenous background and ask what happens when a specific variable is manipulated — SCMs remain powerful tools, and the do-calculus provides rigorous identification conditions. The point is rather that SCMs operate within a social-structural context that they cannot themselves represent. They are tools for asking causal questions about events against a background of structures, not tools for asking causal questions about the structures themselves.

This reinforces the case developed elsewhere in this paper for agent-based simulation as the appropriate methodology for social causal discovery. The agent-based models of Skyrms, Farmer, and Epstein are not bound by the acyclicity constraint. They can and do model feedback loops, self-reproducing structures, and emergent institutional patterns. A Skyrms model of norm evolution represents precisely the circular causal process by which a norm’s existence generates the behavioural patterns that sustain the norm. A Farmer model of financial markets represents the self-reinforcing dynamics through which market structures create the trading patterns that reproduce those structures. These models operate with a richer ontology than Pearl’s SCMs — one that admits structures, loops, and constitutive relations alongside events and interventions. The cost is that they sacrifice the clean identification guarantees of the do-calculus; the gain is that they can represent the kinds of causal entity that actually constitute the social world.

Bongers et al. (2021) have proposed a formal extension of Pearl’s framework to cyclic SCMs, developing equilibrium-based semantics that can handle simultaneous feedback. Their work demonstrates that interventional and counterfactual reasoning can in principle be defined for cyclic causal structures, though at the cost of requiring additional assumptions about equilibrium selection and structural stability. This is a productive direction, but it also confirms the depth of the problem: the extension requires fundamentally different mathematical machinery from that of the standard acyclic framework, precisely because self-referential causal structures are a qualitatively different kind of entity from the event-chains that DAGs were designed to represent. For the social domain, where Luhmann’s autopoietic systems and Farmer’s self-reinforcing market dynamics are the norm rather than the exception, the cyclic extension is not a peripheral refinement but a necessary precondition for representational adequacy.

Iwasaki and Simon (1994), working within Herbert Simon’s earlier tradition of causal ordering, addressed the problem of simultaneous equations in economic models — systems in which equilibrium conditions create mutual dependence among variables — by deriving causal orderings from the mathematical structure of the equations themselves. Their approach demonstrates that causal direction can sometimes be recovered even in simultaneous systems, but it also makes visible the gap between the simultaneity that characterises institutional self-reproduction and the sequential event-ontology that standard SCMs presuppose. The social world is full of simultaneous mutual constitution: supply constitutes demand as demand constitutes supply; legal precedent constitutes legal authority as legal authority constitutes the bindingness of precedent. These are not merely simultaneous equations awaiting causal ordering; they are structural circularities whose resolution into linear causal paths requires precisely the kind of observational standpoint — a system-specific, distinction-dependent schema of attribution — that Luhmann’s framework identifies as the condition of all causal observation.

The acyclicity constraint thus deepens the Luhmannian critique of Pearl developed above. There, the argument concerned the observer-relativity of causal schemas: every SCM is a second-order observation made using system-specific distinctions. Here, the argument concerns the ontological scope of SCMs: their formal structure excludes the self-reproducing entities that constitute the primary fabric of modern social reality. Together, these two objections — one epistemological, one ontological — establish that Pearl’s causal revolution, for all its power in the domain of event-causation, operates with a social ontology that is too thin for the demands of genuine social-scientific causal reasoning. The path forward requires not the abandonment of Pearl’s insights but their integration with formal and computational tools — agent-based simulation, cyclic causal models, systems-theoretic analysis — that can accommodate the richer ontology that social causality demands.

The Social Stabilisation of Causal Attribution in AI

Luhmann’s most distinctive and practically important contribution concerns the social processes through which causal attributions become authoritative within functional subsystems. Pearl’s framework gives the logic of causal inference, assuming a causal model. Van Fraassen urges epistemological humility about that model. Sutton says derive it from data. The embodied tradition says ground it in physical interaction. The agent-based simulation tradition says generate it through structured social interaction. None of these positions, however, asks: who decides which causal schema is operative in a given context, through what social process, with what consequences for the system’s self-reproduction?

The question of whether an AI system ‘genuinely’ performs causal reasoning — whether it operates at Rung 2 or Rung 3 of Pearl’s ladder, or merely appears to — will not be settled by a technical demonstration alone. It will be shaped by the parallel evolution of multiple functional subsystems: the scientific community’s benchmarking practices, the AI industry’s performance claims and competitive deployment, and the institutional standards that emerge from the interaction between research programmes and the domains where they are tested. Each of these subsystems will stabilise its own criteria for what counts as causal understanding in AI, and these criteria need not converge — though productive friction between them is the very mechanism by which scientific progress occurs in functionally differentiated modernity.

The Luhmann-Habermas Debate and AI Governance

The debate between Jürgen Habermas and Niklas Luhmann — most directly staged in their 1971 exchange Theorie der Gesellschaft oder Sozialtechnologie — is one of the most consequential confrontations in twentieth-century social theory, and it bears directly on how AI governance is conceptualised. Habermas contends that social action requires consensual decision-making grounded in communicative rationality. Social processes can in principle be evaluated and redeemed through discourse oriented toward mutual understanding, provided the conditions of ideal speech are approximated — including equality of participation, absence of coercion, and the force of the better argument. Luhmann, by contrast, contends that social activities are too complex for consensual bartering; they require the impersonal systemic regulation that emerges from functional differentiation. For Habermas, the political system — when functioning properly under conditions of deliberative democracy — can legitimately adjudicate between competing claims in science, economy, and technology, because communicative rationality provides a framework that transcends any particular functional system. For Luhmann, this is a category error: the political system operates according to its own code and cannot serve as the meta-system adjudicating the truth claims of science, the efficiency claims of the economy, or the innovation claims of technology without distorting the functional differentiation on which complex society depends.

In the context of AI and causal intelligence, the Habermas-Luhmann debate becomes urgently practical. The question ‘does this AI system genuinely reason causally?’ is simultaneously a scientific question (answered through benchmarking, formal analysis, and experimental study), an economic question (answered through market performance and cost-effectiveness), a legal question (answered through evidentiary standards and liability attribution), and a political question (answered through regulatory frameworks and public policy). Each of these functional systems processes the question according to its own code, and there is no guarantee — and, on Luhmann’s account, no realistic expectation — that their answers will converge. The penetration of moral or political values into scientific and technological domains that should be governed by the codes of science (true/false), economy (profit/loss), and engineering (functional/dysfunctional) constitutes a form of system colonisation that distorts the functional logic of all involved. When the political system expands into the epistemic domain of science and technology, it does not improve on the outcomes those systems would have produced through their own evolution; it substitutes its own operative logic — risk-aversion calibrated to electoral cycles, legibility calibrated to parliamentary scrutiny, accountability calibrated to litigation — for the logic of scientific inquiry and technological innovation. The result is not safer or more genuinely intelligent AI; it is AI optimised for political legibility rather than epistemic adequacy. This does not imply that political discourse - taking the results of science, technology and markets as premises - has no role in AI governance.

9. Concluding Summary

The debate reconstructed and extended in this paper is not resolvable within any single disciplinary framework. It requires philosophy of science (van Fraassen), the theory of computation and learning (Sutton), causal logic (Pearl), embodied cognitive science and robotics (Clark, Kirsh, Spranger), agent-based social simulation (Bratman, Epstein, Skyrms, Farmer), and social systems theory (Luhmann) to be adequately mapped.

Pearl’s core claim — that associational data cannot yield causal conclusions without causal assumptions — remains logically compelling and is substantially supported by recent empirical evidence. The structured pattern of failure at Rung 3 counterfactual tasks in state-of-the-art reasoning models, concentrated at the abduction step, is consistent with Pearl’s theoretical predictions. Sutton’s Bitter Lesson poses a genuine empirical challenge but is an inductive argument confronting a deductive one: the history of AI cannot prove that a mathematical boundary can be crossed.

Van Fraassen usefully strips the debate of metaphysical excess without eliminating its practical force. The embodied cognition tradition of Clark, Kirsh, and Spranger establishes that physical causal understanding is constituted through sensorimotor interaction with the world. But this insight has a crucial domain boundary: the social sciences require a different medium. Agent-based simulation — as developed via the BDI architectures of Rao and Georgeff, by Skyrms for social norms and cooperative dynamics, and by Farmer for economic causality, and grounded philosophically in Bratman’s planning theory of intention, Dennett’s intentional stance, and Epstein’s generative social science — is the social sciences’ functional analogue of embodied robotics. Both programmes share the same epistemic structure: causal knowledge is generated through structured interaction with a generative environment rather than through passive observation of descriptions of interactions. The crucial addition the present paper makes is that social embodiment requires intentional-level architecture: it is not enough to have agents that interact; the agents must interact as planning, believing, intending entities, because that is what social causation is constituted by. This is precisely the distinction Pearl’s Ladder of Causation formalises, and its operationalisation differs appropriately across physical and social domains.

Luhmann’s contribution is the most unsettling for all parties. He does not ask which framework is correct — he asks through what social processes correctness is determined and stabilised differently within society’s differentiated functional subsystems. His answer suggests that the resolution of the technical debate will be, in significant part, a social and institutional achievement, mediated through the internal evolution of the scientific system, the competitive dynamics of technology markets, and the institutionalised distinctions through which different functional systems make causal claims legible to themselves.

The most defensible reading of the present situation is therefore neither pure Pearlian pessimism nor unbounded scaling optimism. A pure LLM, trained only to predict text, is unlikely to be sufficient for robust artificial general intelligence understood as autonomous reasoning under novel interventions. But LLMs are not for that reason irrelevant: they are powerful discourse engines and knowledge interfaces that can serve as components within hybrid architectures that supply what they lack. The most promising path forward is the principled integration of LLM capabilities with structured causal models in the Pearlian tradition, with embodied robotic systems for physical causality in the tradition of Clark, Kirsh, and Spranger, and with agent-based simulation environments for social and economic causality in the tradition of Skyrms and Farmer. Pearl’s challenge stands against AGI by correlation-crunching alone — but not against AGI architectures that embed LLMs within broader causal-agentic systems, drawing on each tradition for the domain to which it is appropriate.

10. Coda: Luhmann Against Unitary AGI

The argument advanced thus far still concedes too much to the framing within which it operates. The very notion of artificial general intelligence as a presumptively unitary, all-purpose, peak intellectual competence — a single system that approximates or exceeds human capability across every domain of cognition simultaneously — is naive and must be rejected. The reasons for this rejection are not technical but sociological in the deepest sense: they follow from a Luhmannian sociology that, taken seriously, functions as a practically relevant first philosophy for any inquiry into the nature of intelligence in modern conditions.

Luhmann’s mature theory posits that advanced modern society is functionally differentiated — organised not as a hierarchy with a unified apex, nor as a totality with a common rationality, but as a constellation of operationally closed, autopoietic social systems each recognisable by, and operating according to, its own binary code: the legal system via legal/illegal, the economic system via profit/loss, the scientific system via true/false (probable/improbable), the technology-engineering system via functional/dysfunctional, the medical system via healthy/diseased, the political system via government/opposition, the education system on pass/fail, the art system on novel/conventional etc. Each system reproduces itself through communications structured by its own code; each generates its own time, its own memory, its own forms of attention; and crucially, none can substitute for the others, nor can any single observation-position synthesise them into a unified whole. This is the condition Luhmann calls polycontexturality: the irreducible multiplicity of incommensurable observational frames that constitutes modern society as such.

Polycontexturality is not a regrettable fragmentation that better integration could overcome. It is the structural condition that makes complex modern society possible. The tremendous productive capacity of advanced societies — their wealth, their knowledge, their technological achievements, their differentiated forms of life — depends on the parallel operation of distinct functional systems that need not, and indeed cannot, be reduced to a single rationality. The legal system’s processing of legality cannot be replaced by the economic system’s processing of profitability; the scientific system’s processing of truth cannot be replaced by the political system’s processing of legitimacy. Each system co-evolves with its own system of distinctions, information processing, discourse-practices, canonical texts, training pathways, modes of dispute resolution, its own highly selective relations to its thereby constituted specific environment. The result is a society of incommensurable competences, each indispensable, none translatable without remainder into another.

This has direct and devastating implications for the dominant conception of AGI. The standard framing imagines a single artificial system that can do everything a competent human can do — reason scientifically, argue legally, design architecturally, treat medically, govern politically, create artistically, teach pedagogically — and do so at or above the highest human level across all domains simultaneously. But this is precisely the conception of intelligence that polycontexturality forbids. There is no unitary cognitive position from which all functional domains can simultaneously be mastered, because each domain requires its own exclusive discourse-practice, fluency with its own code, sensitivity to its own historically evolved modes of distinction-making. Each of these competencies implies simultaneous blind spots. These competences are not reducible to a common substrate of ‘general intelligence’ that could be scaled up to encompass them all. They are constituted by the differentiated systems within which they operate.

This is also why the apparent successes of current LLMs across multiple domains are systematically misleading as evidence for AGI. LLMs trained on text from all functional systems can produce plausible-sounding outputs in legal, medical, scientific, economic, artistic, and political registers — but plausible-sounding outputs are not the same as competent participation in the relevant systems. The legal system does not accept an LLM’s output as a legal judgement; the medical system does not accept it as a clinical decision; the scientific system does not accept it as a contribution to peer-reviewed knowledge. Each system has its own criteria for what counts as a competent communication within it, and these criteria are not satisfied by surface fluency in the system’s vocabulary. The borrowed causal knowledge diagnosis advanced earlier in this paper is itself a special case of this more general phenomenon: LLMs absorb the linguistic residue of competent participation across functional systems without thereby becoming competent participants in any of them.

Luhmann’s analysis of modernity is that is brought to bear here on the question of AGI is not pointing to a logical, cognitive or metaphysical impossibility. It is an empirically grounded theoretical offering from within the science of sociology pointing to the practically highly relevant incommensurability of different discursive-practical function systems understood as systems of communication. These domains are institutionalized via distinct, separate institutions. More-over, AGI projects must take into account that function systems are self-referentially enclosed, each with its specific selectivity via systems of distinctions, evaluative codes, specific media and criteria of success related to the different specific functional contributions these functions systems make to our highly differentiated societies. This is where the crux of incommensurability lies. It is deeper than merely different scientific paradigms within a scientific discipline, and deeper than in the case of different scientific disciplines. Here there can be paradigm integration, and interdisciplinary research. But the very different task orientations of different function systems implies a deeper incommensurability, e.g. the incommensurability between a scholarly legal discourse engaging jurists trying to appraise the status of a certain causal attribution as a matter of law and the appraisal of such an attribution as a matter of e.g. the science of psychology. The point is here that the universe of scholarly legal discourse which includes all common law case law with an unbounded combinatorial cross-referencing potential is so vast that co-evolution of vast LLMs with this universe of discourse is the only practical way forward. The same applies to the vast universe of science, of engineering, of cultural criticism, of politics. To assimilate these domains by referring to them as “cognition” or “intelligence” is misleading. The functional system differentiation, driven to point of incommensurability, is a mechanism of complexity reduction which allows society to evolve further where it otherwise would come to a halt at an otherwise insurmountable complexity barrier. Are we really expecting a single mega LLM to deliver a unitary AGI where advanced world society itself at the aggregate level/dynamic has moved forward via radical differentiation where communication (bounded within function systems) is substituted by co-evolution between function systems?

The proper conception of advanced artificial intelligence that emerges from this analysis is therefore not unitary AGI but a constellation of specialised artificial competences, each developed in deep co-evolution with the functional system it serves. Causal reasoning AI for the sciences will develop through integration with the formal apparatus of Pearl’s causal hierarchy, embodied robotic experimentation, and the institutional practices of scientific peer review. Legal AI will develop through immersion in the codes, precedents, and adversarial procedures of legal systems. Economic modelling AI will develop through agent-based simulation calibrated against empirical economic time series and refined through the disputes of professional economists. Architectural and design AI will develop through engagement with the discourse, historical archive, and material practice of design culture. Each of these will require its own distinctive architecture — its own combination of LLM components, structured causal models, embodied or simulated agents, and domain-specific training environments — appropriate to the functional system within which it operates. There is no master architecture that subsumes them all.

This is not a deflationary conclusion. Far from diminishing the prospects for transformative artificial intelligence, the polycontextural framework clarifies what such intelligence will actually look like and how it will actually be developed. It will be developed, as advanced human intelligence has always been developed, through the parallel evolution of differentiated competences within functional systems that co-evolve with their respective AI augmentations. The frontier is not the construction of a single peak intelligence but the elaboration of an artificial counterpart to the polycontextural society itself — a constellation of artificial systems each calibrated to a functional domain, each communicating in that domain’s code, each subject to that domain’s standards of competence, and each contributing to the parallel reproduction of differentiated modern society.

Luhmann’s sociology, on this reading, functions as a practically relevant first philosophy precisely because it correctly identifies the basic structural condition that all theorising about intelligence in modern conditions must respect. Pearl’s causal revolution, the embodied cognition tradition, agent-based simulation, and even the empirical findings about LLM limitations all gain their proper significance when placed within this polycontextural frame. They are not competing routes to a single peak; they are complementary contributions to the differentiated artificial-cognitive infrastructure that advanced functionally differentiated society requires. The naive conception of unitary AGI — peak general competence across all domains in a single system — is implausible, presupposing a unification of social rationality that has not existed in advanced societies for at least two centuries and whose absence is the condition rather than the obstacle of their tremendous productive achievements.

Note: This paper was researched and drafted with the aid of Claude Sonnet 4.6, ChatGTP 5.4 and Claude Opus 4.6. & 4.7

References

Axtell, R. and Farmer, J. D. (2022). Agent-Based Modeling in Economics and Finance: Past, Present, and Future. INET Oxford Working Paper No. 2022-10. Institute for New Economic Thinking, Oxford.

Bicchieri, C. (2006). The Grammar of Society: The Nature and Dynamics of Social Norms. Cambridge University Press, Cambridge.

Bongers, S., Forré, P., Peters, J., and Mooij, J. M. (2021). Foundations of Structural Causal Models with Cycles and Latent Variables. Annals of Statistics, 49(5), 2885–2915.

Bratman, M. E. (1987). Intention, Plans, and Practical Reason. Harvard University Press, Cambridge, MA.

Brawer, J., Qin, M., and Scassellati, B. (2020). Causal Reasoning about Robots: How Humanoid Robots Build Structural Causal Models from Observation and Self-Supervised Trials. Proceedings of the 2020 ACM/IEEE International Conference on Human-Robot Interaction.

Castri, L., et al. (2022). Causal Discovery to Predict Human Spatial Interactions in Social Robotics Contexts. Proceedings of IEEE RO-MAN.

Castri, L., et al. (2023). F-PCMCI: Advancing Causal Discovery through Feature Selection and PCMCI. arXiv preprint arXiv:2310.09615.

Clark, A. (1997). Being There: Putting Brain, Body and World Together Again. MIT Press, Cambridge.

Clark, A. (2008). Supersizing the Mind: Embodiment, Action, and Cognitive Extension. Oxford University Press, Oxford.

Clark, A. (2016). Surfing Uncertainty: Prediction, Action, and the Embodied Mind. Oxford University Press, Oxford.

Clark, A. and Chalmers, D. (1998). The Extended Mind. Analysis, 58(1), 7–19.

Crawford, S. E. S. and Ostrom, E. (1995). A Grammar of Institutions. American Political Science Review, 89(3), 582–600.

Dennett, D. C. (1987). The Intentional Stance. MIT Press, Cambridge, MA.

Epstein, J. M. (2006). Generative Social Science: Studies in Agent-Based Computational Modeling. Princeton University Press, Princeton.

Epstein, J. M. and Axtell, R. (1996). Growing Artificial Societies: Social Science from the Bottom Up. Brookings Institution Press / MIT Press, Washington and Cambridge, MA.

Farmer, J. D. (2024). Making Sense of Chaos: A Better Economics for a Better World. Allen Lane, London.

Farmer, J. D. and Foley, D. (2009). The economy needs agent-based modelling. Nature, 460(7256), 685–686. DOI: 10.1038/460685a.

Griot, M., Hemptinne, C., Vanderdonckt, J., and Yuksel, D. (2025). Large language models lack essential metacognition for reliable medical reasoning. Nature Communications, 16(1), 642.

Habermas, J. (1984/1987). The Theory of Communicative Action. 2 vols. Beacon Press, Boston.

Habermas, J. and Luhmann, N. (1971). Theorie der Gesellschaft oder Sozialtechnologie: Was leistet die Systemforschung? Suhrkamp, Frankfurt.

IROS Causality for Robotics Workshop (2023). Workshop proceedings summary. IEEE/RSJ International Conference on Intelligent Robots and Systems, Detroit, October 2023. Available at: https://sites.google.com/view/iros23-causal-robots

Iwasaki, Y. and Simon, H. A. (1994). Causality and Model Abstraction. Artificial Intelligence, 67(1), 143–194.

Jin, Z., et al. (2024). Can Large Language Models Infer Causation from Correlation? (Corr2Cause Benchmark) and CausalBench: A Comprehensive Benchmark for Causal Learning Capability of Large Language Models. arXiv preprint arXiv:2306.05836; arXiv:2404.06349.

Kıcıman, E., Ness, R., Sharma, A., and Tan, C. (2023). Causal Reasoning and Large Language Models: Opening a New Frontier for Causality. arXiv preprint arXiv:2305.00050.

Kirsh, D. and Maglio, P. (1995). On Distinguishing Epistemic from Pragmatic Actions. Cognitive Science, 18(4), 513–549.

Kirsh, D. (2013). Embodied Cognition and the Magical Future of Interaction Design. ACM Transactions on Computer-Human Interaction, 20(1), Article 3.

Liu, H., et al. (2024). Unveiling Causal Reasoning in Large Language Models: Reality or Mirage? Advances in Neural Information Processing Systems (NeurIPS) 38.

Luhmann, N. (1962). Funktion und Kausalität. Kölner Zeitschrift für Soziologie und Sozialpsychologie, 14, 617–644. Reprinted in: Luhmann, N. (1970). Soziologische Aufklärung 1. Springer Verlag, Opladen.

Luhmann, N. (1984/1995). Social Systems. Stanford University Press, Stanford.

Luhmann, N. (1990/1992). The Science of Society. Stanford University Press, Stanford.

Luhmann, N. (1991/1993). Risk: A Sociological Theory. Aldine de Gruyter, New York.

Luhmann, N. (1997/2012). Theory of Society (originally Die Gesellschaft der Gesellschaft). 2 vols. Stanford University Press, Stanford.

OpenCausalAI Lab (2024). ADAM: An Embodied Causal Agent in Open-World Environments. arXiv preprint arXiv:2410.22194.

Pearl, J. (1995). Causal diagrams for empirical research. Biometrika, 82(4), 669–688.

Pearl, J. (2000/2009). Causality: Models, Reasoning, and Inference. 2nd edition. Cambridge University Press, Cambridge.

Pearl, J. (2009). Causal inference in statistics: An overview. Statistics Surveys, 3, 96–146.

Pearl, J. and Mackenzie, D. (2018). The Book of Why: The New Science of Cause and Effect. Basic Books, New York.

Pearl, J. and Mackenzie, D. (2023). Judea Pearl, AI, and Causality: What Role Do Statisticians Play? Amstat News / ASA Magazine, September 2023.

Rao, A. S. and Georgeff, M. P. (1991). Modeling Rational Agents within a BDI-Architecture. In J. Allen, R. Fikes, and E. Sandewall (eds.), Proceedings of the 2nd International Conference on Principles of Knowledge Representation and Reasoning (KR’91), Morgan Kaufmann, San Mateo, pp. 473–484.

Schelling, T. C. (1978). Micromotives and Macrobehavior. W. W. Norton, New York.

Schumacher, P. (2025). Discourse Capitalism. [Publisher details forthcoming].

Skyrms, B. (1996). Evolution of the Social Contract. Cambridge University Press, Cambridge.

Skyrms, B. (2004). The Stag Hunt and the Evolution of Social Structure. Cambridge University Press, Cambridge.

Skyrms, B. (2010). Signals: Evolution, Learning, and Information. Oxford University Press, Oxford.

Skyrms, B. (2014). Social Dynamics. Oxford University Press, Oxford.

Spranger, M. (2016). The Evolution of Grounded Spatial Language. Language Science Press, Berlin. DOI: 10.26530/OAPEN_611695.

Sutton, R. (2019). The Bitter Lesson. Incomplete Ideas (personal blog), 13 March 2019. Available at: http://www.incompleteideas.net/IncIdeas/BitterLesson.html

van Fraassen, B. C. (1980). The Scientific Image. Oxford University Press, Oxford.

van Fraassen, B. C. (2002). The Empirical Stance. Yale University Press, New Haven.

Vashishtha, N., et al. (2025). Executable Counterfactuals: Improving LLMs’ Causal Reasoning through Code. arXiv preprint arXiv:2510.01539.

Willig, M., et al. (2023). Causal Reasoning in Large Language Models: A Comprehensive Survey. arXiv preprint arXiv:2306.05836.

The reflections offered in this paper were triggered by David McWilliams’ positive appraisal of the Florentine medieval guilds and of the Dutch East India Company in his recent book “Money: A Story of Humanity” (2024). While these appraisals are by no means given a special place of prominence in McWilliams’ sweeping historical arc that focusses on financial innovations as crucial factor in the progress of humanity, they stirred up my libertarian sensibilities and lead me to probe deeper here with respect to these two historical case studies in institutionalised economic coordination. McWilliams presents these two cases — the guild system of medieval and Renaissance Florence and the monopoly charter granted by the Dutch Republic to the Vereenigde Oostindische Compagnie (VOC) in 1602 — as economically rational responses to pervasive market failures. This paper summarises and critically examines McWilliams’s arguments, situating them within a historically informed libertarian-institutionalist framework drawing on Adam Smith, F.A. Hayek, Douglass North, Mancur Olson, and the discourse theory of Jürgen Habermas. The paper argues that McWilliams correctly identifies genuine coordination problems that these institutions partially resolved, but that his analysis insufficiently distinguishes between voluntary cooperative structures and state-enforced monopoly privileges. A libertarian reading can partially accept the efficiency thesis for the early guild as a spontaneous order, while mounting a sharper critique of the VOC charter as a second-best institution whose efficiency gains were inseparable from systematic coercion. The paper further proposes a ‘discourse capitalism’ framework — following Schumacher (2025) — as an analytical lens for understanding how both institutions succeeded by establishing legitimate discursive frameworks of commercial norms, quality standards, and shared expectations, and how their decay followed the corruption of those frameworks by rent-seeking. The policy implication is that the historical record supports investing in foundational institutional infrastructure — secure property rights, impartial contract enforcement, open access orders — rather than normalising monopoly as a durable solution to coordination failure.

Keywords: guilds, Dutch East India Company, coordination failure, libertarian institutionalism, Hayek, North, discourse capitalism, McWilliams, monopoly, trust

1. Introduction

Economic history is littered with institutions that appear, from a distance, to contradict the normative commitments of liberal political economy. Guilds restricted entry, fixed prices, and subordinated individual enterprise to collective governance. The Dutch East India Company held, by sovereign grant, an absolute monopoly over some of the most lucrative trade routes in the early modern world. Both institutions have been condemned, in the long run of liberal thought, as engines of rent extraction inimical to free exchange and the spontaneous coordination of markets.

Yet in Money: A Story of Humanity, the Irish economist David McWilliams mounts an intelligent defence of both as rational institutional responses to the coordination problems of their time. His argument is neither nostalgic nor anti-liberal in spirit: it is functionalist, historicist, and broadly consistent with a sophisticated reading of institutional economics. McWilliams asks not whether these institutions were ideal, but whether they were rational given the informational, financial, and legal environment in which they operated.

This paper takes McWilliams’s arguments seriously, summarises them faithfully, and then subjects them to critical scrutiny from a historically informed libertarian perspective. That perspective is not dismissive: it does not simply dismiss any non-market institution as illegitimate. It draws instead on a tradition — running from Adam Smith through Hayek and Friedman to Douglass North and Mancur Olson — that is sensitive to the role of institutions in enabling or constraining market exchange, and that distinguishes sharply between institutions that emerge from voluntary cooperation and those that are maintained by coercive privilege.

The paper proceeds as follows. Section 2 summarises McWilliams’s arguments on the Florentine guilds. Section 3 summarises his case for the VOC monopoly. Section 4 offers a libertarian-institutionalist critique and partial acceptance of each case. Section 5 introduces the ‘discourse capitalism’ framework (Schumacher, 2025) as a synthetic analytical lens. Section 6 concludes with broader implications for institutional design.

2. McWilliams on the Florentine Guilds: Innovation Through Cooperation

McWilliams locates the Florentine guilds within a specific institutional vacuum: the collapse of Roman commercial law, the retreat of monetary exchange during the European Dark Ages, and the re-emergence of trade in the eleventh and twelfth centuries without adequate mechanisms of trust, quality assurance, or contract enforcement. In this environment, he argues, the guilds performed an essential economic function.

His central claim is captured in a key passage from the book: the cooperative nature of guilds amplified the impact of invention, and it was the guilds that turned invention into innovation — with innovation understood as the commercial application of invention (McWilliams, 2024). The argument is Schumpeterian in spirit: dispersed craft knowledge required an organisational shell to become economically productive at scale. The guild provided that shell through several interlocking mechanisms.

First, the guilds resolved the quality signalling problem. In thin, long-distance markets with no reliable product standards, buyers could not verify the quality of goods before purchase. Guild certification — backed by the reputational capital of the collective — provided a credible quality signal that individual producers could not generate alone. This is essentially a solution to the ‘market for lemons’ problem identified by Akerlof (1970), resolved through institutional means rather than regulatory mandate.

Second, the guilds provided a training and knowledge-transfer infrastructure. The apprentice-journeyman-master progression was not merely a labour market hierarchy: it was a systematic mechanism for codifying, transmitting, and improving tacit knowledge across generations. In an era before patents, technical manuals, or universities with applied science faculties, the guild was the primary vehicle through which productive knowledge became a social rather than purely personal asset.

Third, and perhaps most significantly for McWilliams’s broader narrative, the Florentine guilds provided the organisational substrate for financial innovation. He credits the Florentine Arti with underwriting the development of double-entry bookkeeping and the early forms of fractional reserve banking that made Florence the centre of European finance by the thirteenth century. The florin, minted by Florence around 1252 and widely used as an international trading currency, was itself an institutional product of guild-organised commercial culture.

McWilliams’s treatment of the guilds is thus broadly functionalist: these were not arbitrary or purely extractive institutions, but responses to genuine market failures — information asymmetry, public goods in knowledge production, and the absence of contract enforcement — that generated real economic value in their historical context.

3. McWilliams on the Dutch East India Company: Coordinated Capital and the Birth of Shareholder Culture

McWilliams’s account of the Dutch East India Company (VOC) turns on a distinct but related coordination problem: the mobilisation of capital for ventures of unprecedented scale and risk in an era without deep financial markets or the institutional infrastructure of modern corporate finance.

The Dutch Republic of the early seventeenth century faced a structural challenge. As McWilliams observes, the Netherlands was a small nation with few natural resources — lacking particularly the territory and manpower that had historically constituted the stores of wealth and power for European states. What the Dutch possessed was a mercantile culture, a sophisticated urban commercial class, and a state willing to sponsor enterprises in which the public could invest (McWilliams, 2024).

The problem was that a round voyage to the Spice Islands — lasting two to three years, subject to shipwreck, piracy, scurvy, and political instability in the Indonesian archipelago — required capital far beyond what any individual merchant house could raise. Competing Dutch firms, moreover, were undercutting each other in Asian markets, driving down returns and making the enterprise collectively unprofitable even when individually rational.

The VOC charter of 1602 solved both problems simultaneously. By granting a single entity a monopoly over all Dutch trade east of the Cape of Good Hope and west of the Strait of Magellan, the States-General eliminated destructive intra-Dutch competition. By enabling the VOC to issue shares to the public — in what was effectively the world’s first large-scale joint-stock offering — it created a mechanism for aggregating dispersed capital across a broad population of investors, including small savers who would otherwise have had no access to the returns from long-distance trade.

McWilliams’s argument is that this was not merely a political favour to merchant interests but an institutional innovation with genuine public-goods properties. By tying the fortunes of the company to the state, the government encouraged a shareholding culture, drawing in citizens who would otherwise have been too risk-averse to invest in speculative overseas ventures individually (McWilliams, 2024). The result was a democratisation of investment — imperfect and hierarchical, but real — that embedded financial capitalism in the Dutch national culture and laid the organisational foundations for the Amsterdam Bourse, the first modern stock exchange.

The VOC, on this reading, was economically rational at the level of system design: it aggregated capital, eliminated wasteful competition, pooled risk, and generated returns for a broad cross-section of Dutch society. McWilliams does not ignore the violence inherent in colonial extraction, but frames it as contextually inevitable rather than definitionally constitutive of the VOC’s economic logic.

4. A Libertarian-Institutionalist Engagement

4.1 On the Florentine Guilds: Partial Acceptance and Dynamic Critique

A libertarian reading of the guild system need not begin with rejection. Hayek’s theory of spontaneous order provides a framework for taking voluntary cooperative institutions seriously as carriers of distributed knowledge and evolved social norms. In their early phase — roughly the twelfth and thirteenth centuries — Florentine guilds emerged from voluntary association among tradesmen seeking reputational and contractual certainty in thin markets. This is consistent with Hayek’s insight that institutions encoding tacit knowledge often arise without central design and serve genuine coordinating functions (Hayek, 1973).

The libertarian critique bites hardest at a later stage. By the fifteenth century, the Arti Maggiori — the seven major guilds — had become legally entrenched barriers to entry enforced by the Commune. Membership in specified guilds was a prerequisite for political office under the ordinamenti di giustizia. What began as voluntary coordination had calcified into state-backed cartelism: entry restriction maintained not by competitive quality but by coercive exclusion of potential rivals.

This transformation is precisely what Mancur Olson (1982) predicted in his theory of distributional coalitions. Small, well-organised groups will, over time, use collective action to capture regulatory and legal institutions and redirect their outputs toward member rents rather than social surplus. The guilds’ economic function did not disappear overnight, but it became increasingly secondary to their rent-extracting function. The economic irony, visible in retrospect, is that this rigidity contributed to Florence’s relative commercial stagnation compared to more open later entrepots, particularly Amsterdam.

A counterfactual question McWilliams does not press is whether Florentine commercial and financial innovation would have occurred faster under more open entry conditions. The Hanseatic League — a looser and more contractual arrangement lacking the guild’s internal hierarchy — produced comparable commercial innovation in Northern Europe without formal apprenticeship monopolies. This suggests the guild was not a uniquely necessary institutional form, only a historically contingent one.

The libertarian synthesis, then, is this: McWilliams is right that the early guild solved real coordination problems. He is less attentive to the structural tendency of such institutions toward rent-seeking capture, a tendency that is not contingent but systematic. The appropriate lesson is not to celebrate guilds as a model but to understand them as a revealing example of the lifecycle that Olson identified: voluntary coordination → institutional entrenchment → distributional capture → economic sclerosis.

4.2 On the Dutch East India Company: The Harder Case

The VOC presents a more philosophically demanding challenge, because it is simultaneously one of the most creative institutional innovations in economic history and one of the most brutal state monopolies ever constructed.

McWilliams’s efficiency argument is structurally identical to the classical justification for temporary monopoly grants in the context of large fixed costs and public goods: the same logic underlies patent law, infrastructure concessions, and broadcasting licences. A classical liberal — following Adam Smith rather than contemporary libertarianism — might accept that certain forms of state-franchised organisation are legitimate second-best responses to genuine coordination failures, particularly where the state itself determines the institutional environment (Smith, 1776). The VOC fits this mould. A libertarian could integrate McWilliams’s argument by treating the monopoly charter as historically rational given the institutional constraints of 1602, without endorsing monopoly as a general principle.

The first serious libertarian objection concerns the counterfactual. The monopoly did not merely solve a pre-existing coordination problem; it defined the problem in a way that foreclosed alternatives. Competing private expeditions — the Portuguese model, with its looser licensing system — and later the less rigidly monopolistic English East India Company both demonstrated that Asian trade was organisable without full monopoly grants. The VOC’s exclusivity may have been partly a political reward to its founding shareholders rather than a pure institutional necessity.

The second, and more fundamental, objection concerns the relationship between the VOC’s commercial innovation and its violence. The Irish Times review of McWilliams’s book is candid that the VOC provided Dutch citizens with returns derived from violent colonial exploitation (McWilliams, 2024, as reviewed in The Irish Times, 2024). What deserves greater analytical weight is that this violence was not incidental: it was structural. The VOC’s charter granted it not merely commercial rights but sovereign powers — the jus belli, including the right to wage war, sign treaties, establish fortifications, and administer criminal justice. Jan Pieterszoon Coen’s systematic depopulation of the Banda Islands to enforce the nutmeg monopoly was not an aberration but a direct expression of the charter’s logic: the monopoly was commercially viable only insofar as competing suppliers — indigenous producers and rival European merchants alike — could be eliminated by force.

A consistent libertarian framework grounded in property rights and the non-aggression principle cannot treat this as separable from the efficiency analysis. The VOC’s ‘coordination solution’ was rational from the perspective of its shareholders precisely because it externalised the full costs of market-making onto colonised populations who bore them involuntarily. This is not a market institution with state backing; it is a state institution with commercial branding.

Douglass North’s framework offers the most productive libertarian integration. North (1990) distinguishes between institutions that lower transaction costs through the extension of impersonal exchange — the hallmark of open-access orders — and those that lower transaction costs for some parties by raising them coercively for others. The VOC exemplifies the latter. Its institutional achievement was real but asymmetric: it created the conditions for Dutch financial capitalism while systematically destroying the conditions for autonomous economic development in the regions it monopolised. The libertarian lesson is that this distinction — between transaction cost reduction through open access versus through coercive exclusion — is precisely what determines whether a coordinating institution is compatible with a liberal order.

5. Discourse Capitalism as an Analytical Lens

The concept of ‘discourse capitalism’ (Schumacher, 2025), drawing on Habermas’s distinction between instrumental and communicative action, offers an additional analytical dimension that McWilliams’s account implicitly requires but does not explicitly supply.

Habermas (1984) distinguishes between action oriented toward strategic success — the manipulation of objects and persons to achieve predetermined ends — and action oriented toward mutual understanding, in which actors coordinate through the shared recognition of validity claims. Economic markets, on this account, operate primarily through the medium of money as a steering mechanism, coordinating action without requiring communicative consensus. But markets presuppose a background of legitimate institutional norms — property rights, contract law, standards of honest dealing — that cannot themselves be generated by market logic alone and require something closer to communicative processes of norm-formation.

What both the Florentine guilds and the VOC actually achieved, at their most functional, was the establishment of such background frameworks: shared discursive structures of commercial legitimacy, quality expectations, and professional obligation that reduced the transaction costs of market exchange not through coercion but through norm. Guild certification was a discourse about quality. The florin was a discourse about value. The VOC’s shareholder prospectus was a discourse about risk and return that made previously unthinkable levels of investment coordination possible.

Their decay, on this reading, follows a predictable pattern: the discursive framework that originally legitimised the institution — and generated genuine economic value through shared norms — was gradually displaced by purely strategic action oriented toward rent extraction. Guild certification ceased to be a quality signal and became a barrier. The VOC’s charter ceased to be a coordination mechanism and became a licence for violence. The institution survived, but the discourse that gave it legitimate economic function was hollowed out.

This framework suggests that the key variable in evaluating pre-modern coordinating institutions is not whether they involved state sanction — virtually all did, to some degree — but whether they sustained or suppressed the communicative processes through which commercial norms are legitimated and updated. Open-access orders (North, Wallis, and Weingast, 2009) are, on this reading, institutional structures that preserve the conditions for discursive norm-formation in economic life: they do not merely enforce existing rules but enable the contestation and revision of those rules through something approximating communicative rather than purely strategic action.

The policy implication is that the goal of institutional design should be to create conditions under which legitimate discursive frameworks can emerge and be maintained — not to replicate the specific institutional forms that happened to generate such frameworks under the very different conditions of medieval Florence or the early Dutch Republic.

6. Conclusion

McWilliams’s Money provides a historically rich and analytically stimulating account of the Florentine guilds and the VOC as rational institutional responses to coordination failures that markets, absent the supporting infrastructure of property rights and contract law, could not resolve unaided. His functionalist argument deserves to be taken seriously rather than dismissed from a libertarian perspective that mistakes ideological purity for analytical rigour.

Yet a historically informed libertarian reading reveals significant limitations in his account. On the guilds, McWilliams correctly identifies the early cooperative function but underestimates the structural tendency toward distributional capture identified by Olson: the guild’s lifecycle from spontaneous order to coercive cartel is not an accident of Florentine political history but a predictable consequence of granting collective institutions coercive state backing. On the VOC, his efficiency argument holds as a second-best institutional response to genuine capital mobilisation problems, but it cannot be coherently separated from the coercive logic by which the company enforced its monopoly — a logic that reveals the VOC not as a market institution with state support but as a sovereignty-wielding entity using commercial language.

The discourse capitalism framework (Schumacher, 2025) permits a synthetic reading: both institutions succeeded initially by establishing legitimate discursive frameworks that reduced transaction costs through shared norms rather than coercion alone, and both decayed as strategic rent-seeking displaced communicative norm-formation. This framework suggests that the appropriate institutional lesson from these cases is not to celebrate coordinating monopolies but to invest in the foundational institutional layer — open access to property and contract, impartial enforcement, freedom of entry — that reduces the ‘need’ for such second-best solutions by making the coordination problems they addressed soluble through market and discursive means directly.

The future that McWilliams’s own narrative points toward — though he does not frame it in these terms — is one ordered via markets and steered via discourses: a post-monopoly institutional architecture in which coordination emerges from the interaction of price signals and legitimate communicative frameworks, rather than from the coercive bundling of commercial function with sovereign violence that characterised both the guild in its decline and the VOC throughout its existence.

References

Akerlof, G.A. (1970) ‘The Market for “Lemons”: Quality Uncertainty and the Market Mechanism’, Quarterly Journal of Economics, 84(3), pp. 488–500.

Habermas, J. (1984) The Theory of Communicative Action, Volume 1: Reason and the Rationalization of Society. Translated by T. McCarthy. Boston: Beacon Press.

Hayek, F.A. (1973) Law, Legislation and Liberty, Volume 1: Rules and Order. London: Routledge.

McWilliams, D. (2024) Money: A Story of Humanity. New York: Scribner.

North, D.C. (1990) Institutions, Institutional Change and Economic Performance. Cambridge: Cambridge University Press.

North, D.C., Wallis, J.J. and Weingast, B.R. (2009) Violence and Social Orders: A Conceptual Framework for Interpreting Recorded Human History. Cambridge: Cambridge University Press.

Olson, M. (1982) The Rise and Decline of Nations: Economic Growth, Stagflation, and Social Rigidities. New Haven: Yale University Press.

Schumacher, P. (2025) Thesis on Discourse Capitalism: Markets and Discourses – Politics after the Libertarian Revolution, https://substack.com/home/post/p-173182898

Smith, A. (1776) An Inquiry into the Nature and Causes of the Wealth of Nations. London: W. Strahan and T. Cadell.

The Irish Times (2024) ‘Money: A Story of Humanity by David McWilliams — Barbarity in Pursuit of Prosperity for Some’, The Irish Times, 14 September. Available at: https://www.irishtimes.com (Accessed: April 2026).

The scale of the resulting distortion is extraordinary. In London, the house price to earnings ratio has climbed from roughly four to one two decades ago to approximately ten to one today. The mean rent-to-income ratio has risen from one fifth to one third over the past fifteen years. These figures do not describe a natural scarcity. Construction costs have remained relatively stable in real terms; buildings depreciate and can be renewed. What has exploded is not the cost of building homes but the cost of the land and regulatory permissions required to build them. In inner London, land now accounts for seventy to eighty per cent of the sale price of a new dwelling. In prime central locations, this figure exceeds eighty per cent. Housing is expensive because politically rationed land with planning permission is expensive, not because bricks and mortar are costly.

The engine of this inflation is the discretionary planning system itself. England’s regime does not operate by clear, general rules within which developers may act freely. Instead, every significant project must be individually negotiated through a lengthy, opaque, and politically mediated approval process. Outcomes are uncertain in timing, scale, design requirements, infrastructure obligations, and financial contributions. Local authorities possess wide discretionary powers, and decisions are subject to judicial review, renegotiation, or reversal following changes in political control. The result is a system in which development rights are not secure, market-based entitlements but politically allocated privileges—and in which the rational response for any landowner or developer is to treat land as an option held under uncertainty rather than a productive asset to be improved as swiftly as possible.

This institutional environment explains the phenomenon routinely denounced as “land banking.” Critics attribute housing shortages to the speculative hoarding of land by profit-seeking developers. This diagnosis is largely misconceived. In a functioning market, holding undeveloped land generates no income and ties up capital unproductively. Developers face real financing costs, balance-sheet constraints, and investor pressure to generate returns. The incentive under normal conditions is to build, not to wait. What appears externally as banked land is overwhelmingly land awaiting the resolution of regulatory and political risks—sites lacking infrastructure, subject to unresolved planning obligations, or contingent on master-planning frameworks involving multiple public stakeholders. Land banking is not a pathology of capitalism but a pathology of governance. It reflects not too much market freedom but too little.

Beyond the planning bottleneck, a thicket of prescriptive regulations further strangles housing supply and innovation. Planners impose not only land uses and overall building volumes but also unit mixes—how many studios, one-bedroom, two-bedroom, and three-bedroom flats must be included—along with minimum unit sizes, minimum room sizes, mandatory facilities, maximum glazing ratios, and architectural expression subject to aesthetic approval. The entrepreneur and architect are left with almost nothing to decide. Competition through creative product innovation is effectively blocked. Developers compete not on the quality or ingenuity of their offerings but on their capacity to navigate political negotiations, where insider knowledge of local planners and councillors delivers the decisive edge. Resources are thus diverted from productive building into what public choice economics rightly classifies as rent-seeking.

The consequences for the lower end of the market are particularly perverse. Currently, studio flats below thirty-eight square metres are prohibited. Yet smaller units built in earlier eras are rare, hotly desired commodities. Lifting the prohibition would allow an entire new income group—young professionals vital to London’s economy—to enter the market, boosting both unit numbers and affordability. That this obvious liberalisation is resisted by the very politicians who profess concern for lower-income groups reveals the deeper logic at work: state-imposed minimum standards exist partly to protect the legitimacy of subsidised social housing. If freely chosen private-sector solutions produced a viable segment at lower space standards, the political rationale for government-provided housing at generous specifications would be undermined. The affordability system thus stands in the way of the market-driven affordability it claims to pursue.

The so-called “affordable housing” regime—comprising social rent allocations and intermediate products such as shared ownership, imposed on developers through Section 106 obligations—constitutes yet another layer of counterproductive intervention. These obligations function as an opaque, negotiable, in-kind tax on development. They embed fiscal extraction within discretionary planning negotiations, generating uncertainty, legal costs, and systematic bias against building. Every unit of mandated affordable housing raises the price of market-rate units that must cross-subsidise it, tightening the affordability crisis for those who do not qualify for rationed provision. The Mayor of London’s ambition to push affordable quotas toward fifty per cent of all new housing makes the arithmetic brutally clear: the higher the mandated subsidy, the more unaffordable the remaining supply becomes. It is an interventionist spiral in which each round of political remedy deepens the malady it purports to cure.

The rental market suffers from a parallel syndrome. Successive governments have layered tenant protections—restrictions on eviction, caps on rent increases, mandatory long tenancy terms—that superficially favour current occupants but systematically discourage investment in rental supply. Landlords exit the sector; planned rental developments are aborted; properties are withdrawn from the market. The fiscal regime compounds the problem: the removal of full mortgage interest deductibility for buy-to-let investors treats a legitimate business expense as a taxable windfall, penalising the very investment that would expand supply and moderate rents. Meanwhile, rent controls immobilise tenants in unsuitable properties, preventing the reallocation of housing stock to those who would use it most productively and hampering the labour mobility on which a dynamic economy depends. Despite near-unanimous opposition to rent controls among economists, politicians persist in proposing them. The best long-term protection for tenants, current and future, lies not in regulatory coercion but in the competitiveness of the market for rental housing. In a free market, landlords compete for tenants as vigorously as tenants compete for apartments. Freedom of contract, not state-imposed tenancy terms, would allow landlords to cater to customers with diverse priorities and circumstances, expanding rather than contracting the range of available options.

The cumulative effect of these overlapping interventions—discretionary planning, land-use zoning, density and height restrictions, prescriptive housing standards, affordable housing mandates, and rental market regulation—is a system of urban over-regulation so comprehensive that it amounts to a form of paralysis. Development that markets would deliver swiftly and adaptively is delayed for years or blocked entirely. Innovation in housing typologies is forbidden. Price signals that would guide resources to their most valued uses are overridden by political allocation. The population is divided into insiders—existing homeowners who receive continuous windfalls from artificial scarcity—and outsiders, particularly the young, who are priced out of ownership and squeezed in a rental market starved of supply.

The libertarian diagnosis is therefore straightforward. Britain’s housing crisis is the predictable result of the systematic suspension of property rights, voluntary exchange, and entrepreneurial initiative in the domain where they matter most to people’s daily lives. The remedy is equally clear in principle, however difficult it may be politically: restore by-right development within transparent, rule-based codes; replace discretionary permission with automatic approval for projects meeting objective parameters; abolish prescriptive unit mixes and space standards; end on-site affordable housing mandates and replace them with portable, means-tested support; liberalise the rental market by restoring freedom of contract; and replace the opaque planning-gain levies of Section 106 and the Community Infrastructure Levy with a transparent, rule-based system of land value capture that taxes socially generated windfalls without penalising the act of building.

None of this implies the absence of rules. It implies the replacement of discretionary, politicised, and opaque rules with general, predictable, and neutral ones. It means planning that specifies what may not be done—genuine nuisance, safety hazards, environmental harm—rather than requiring explicit political approval for every act of construction. It means trusting the competitive market process to discover, through the disciplined interplay of entrepreneurial initiative and consumer choice, the forms of housing and urban development that best serve the needs and aspirations of millions of individuals—rather than entrusting that task to politicians and bureaucrats who lack the knowledge, the incentives, and the accountability to perform it. The housing crisis will not be solved by more intervention. It will be solved by finally allowing capitalism to do what it does in every other sector where it is permitted to operate: deliver abundance, choice, and falling real costs to ordinary people.

Published in: Austin Williams (Ed.), Five Critical Essays on Over-regulation, TRG Publishing, London 2026

Our aim is not wholesale deregulation, but re-marketisation: restoring a rule-based order in which development is permitted by default within clear, general, and predictable rules. The planning rules should specify what is not allowed (nuisance, safety, environmental harm), not require explicit discretionary political approval for all buildings. The following 12 proposals are transitional demands, i.e. politically feasible steps that move the UK decisively back toward a development-rights-based urban order.

Policy Proposals:

1. Restore By-Right Development within Clear Urban Codes

Replace discretionary planning permission with automatic approval for projects that comply with national codes and local plans setting measurable parameters—height, floor-area ratio, setbacks, access—leaving no scope for subjective or politically motivated refusals. This restores legal certainty, reduces risk premiums, and allows supply to respond to demand without political negotiation. The underlying local plans should only constrain the quantum of development in relation to infrastructure and leave functional land-use allocations – whether sites are developed for housing, offices, hospitality, or retail – to the market process that maximizes utility and co-location synergies. That by-right development can function well is being demonstrated by the introduction of Permitted Development Rights (since 2013) for office to residential conversions where no planning permission is required. In some years up to 20% of national new housing units where supplied via PDRs.

2. Statutory Decision Deadlines with Deemed Approval

Introduce binding time frames for all planning decisions: 30 days for minor schemes, 60 for standard, 90 for major. Failure to decide within the deadline results in automatic approval. This single reform would immediately unblock development pipelines, reduce land-banking incentives created by uncertainty, and re-establish planning as a predictable administrative check rather than a political bottleneck.

3. Fast-Track Small Sites and Incremental Infill

Create an automatic fast-track for small developments—such as schemes with a floor area below 5000 square meter —approved on checklist compliance alone. Britain’s historic urban growth was delivered by thousands of small builders and incremental projects; today’s system favours only large incumbents who can absorb planning risk. This reform restores competition, accelerates delivery, and revitalises small-scale development.

4. Replace the Green Belts with a Green Corridor Systems

End the blunt, blanket Green Belt designation and replace it with a Green Corridor framework that protects genuine ecological assets, public access, and landscape continuity while releasing low-amenity land at urban edges and transport nodes for development by right. This reform preserves what people value about green space while ending the artificial land scarcity that has inflated house prices around Britain’s most productive cities.

5. Up-Zoning by Default

Allow higher urban densities everywhere, especially within walking distance of rail stations and Underground stops. This aligns housing supply with infrastructure investment, reduces sprawl and commuting costs, and allows land values to rise through productivity rather than scarcity. Implement the policy of Street Votes whereby homeowners on a low-rise street can vote to upzone their street. It is pro-environment, pro-growth, and pro-urban without requiring new public spending.

6. End On-Site Affordable Housing Mandates, Use Portable Support

Abolish mandatory on-site affordable housing quotas (section 106 obligations), which act as a tax on new supply and inflate prices for market homes. Replace them with portable, means-tested housing support that households can use in the open market. This separates social policy from land-use control, increases overall supply, and avoids the perverse cross-subsidies that currently suppress construction.

7. Stop social engineering via apartment type prescriptions and space standards

Prevent planners from imposing residential unit mixes and abolish space standards that force housing prices upwards and block the market discovery of residents’ priorities. The sizes of apartments, the unit mix (in terms of number of bedrooms), and the provision of facilities should be within the discretion of the developer and must no longer (as it is now) be subject to government rules or planning imposition. Developers (urban entrepreneurs) must be able to innovate and adapt housing provision to market demand. Currently mandated minimum apartment sizes limit unit numbers and increase costs per unit. That attractive supply can be brought to market without space standards had been demonstrated by Permitted Development Rights (PDR) in the case of office to residential conversions, until 2021 when space standards were imposed also on conversions.

8. Liberalize the rental market and restore buy-to-let viability

Repeal tenant protection measures that have unintentionally reduced rental supply, increased rents, and intensified landlord risk-aversion in tenant selection. Over-regulation has discouraged investment, prompted landlords to exit the sector, and constrained mobility for renters. Freedom of contract in the rental market must be restored. To further expand the supply of homes for rent reinstate full mortgage interest deductibility, allowing buy-to-let investors to treat their finance costs as legitimate business expense. A healthy rental market requires incentives aligned with long-term investment in housing provision.

9. Replace Section 106 and CIL with Land Value Uplift Capture

Abolish negotiated Section 106 obligations (affordable housing quotas and other in-kind obligations) and the Community Infrastructure Levy (CIL) as sources of Local Authority Funding. They function as opaque, uncertain taxes on development. Replace them with a simple, rule-based system of land value capture—preferably a locally retained land value uplift tax, combined with a low permanent land value tax to stabilize local finance. This removes bargaining, reduces corruption and delay, and ensures that socially created land value is captured transparently rather than through politicized deal-making. Taxing unearned development windfalls makes more sense than to tax the development investments that generate these windfalls.

10. Align Local Government Incentives through Growth Dividends

Allow local authorities to retain a significant share of the additional tax revenue generated by new developments. Councils currently bear the political costs as well as the financial costs of growth – being required to provide services to the new residents - while receiving little fiscal benefit. Mechanism: full retention of business rates from newly developed office space and full retention of council tax from new residential units. Growth dividends realign incentives so that councils are rewarded for enabling development rather than blocking it, transforming planning from a defensive to a growth-oriented activity.

11. Modernise Property Rights to Enable Urban Intensification

Reform archaic constraints such as perpetual rights to light, and obsolete covenants. Introduce transferable development rights (TDRs) so that landowners restricted for heritage or environmental reasons can be compensated through market exchange. Development rights denied on one site can be sold to another site. More generally TDRs allow for the optimized allocation of a globally constraint development quantum within a district.

12. Restore Neutrality in Property and Development Finance

Remove fiscal and regulatory biases that favour holding existing property over building new supply. Abolish stamp duty, equalise VAT between renovation and new build, and reform banking regulation so development finance is not penalised relative to mortgages. The result of the current regulatory regime is that credit flows overwhelmingly into bidding up existing stock rather than financing new supply. Only on the basis of financial neutrality can the planning reform proposed above deliver the optimal supply of new space.

Closing Statement

Britain’s housing crisis is not a failure of markets but the result of their suspension. The planning system has replaced entrepreneurial initiative with bureaucratic allocation, turning cities into arenas of scarcity and conflict. NIMBY veto power dominates. Re-marketising urban development means restoring the rule of law, property rights, and price signals—so that millions of individual decisions can once again coordinate growth.

These reforms proposed here do not abolish planning; they civilise it. They replace discretion with rules, veto power with incentives, and political bargaining with the lawful use of property rights. The outcome is not chaos, but abundance: more homes, lower costs, higher productivity, and cities that once again function as engines of opportunity rather than monuments to managed scarcity. Top of FormBottom of Form

Over the past half century, advanced economies have undergone a profound transformation in the way value, information, and capacity are organised. Systems once structured around stockpiles—physical inventories, fixed capacities, and pre-emptive planning—have increasingly been replaced by flow-based regimes characterised by real-time coordination, continuous adjustment, and minimal buffering. This paper argues that this shift is not merely a technological or logistical improvement, but a deep epistemic and institutional transformation with consequences that extend from factory floors and server farms to welfare bureaucracies, regulatory governance, and the planning of cities.

Enabled by non-kinetic digital storage, high-speed networks, and cheap computation, flow-based systems displace precautionary accumulation with adaptive responsiveness. Simultaneously, the broader post-Fordist restructuring of the economy—from mechanical mass production toward flexible specialisation, economies of scope, rapid innovation, and knowledge-economy dynamics—has reshaped organisational form, labour markets, and the relationship between state and citizen. Drawing on economic theory, organisational studies, technological history, and political economy, the paper situates the transition from stockpiles to flows as a paradigmatic shift in how complex societies coordinate under uncertainty.

The analysis argues that many contemporary dysfunctions in welfare, governance, and spatial planning arise from the persistence of stockpile logics within environments that now operate according to flow-based dynamics. It further argues that this diagnosis points beyond mere institutional reform toward a fundamental reconceptualisation of social order. The paper identifies the current ascent and proliferation of artificial intelligence—and especially of agentic AI systems capable of autonomous perception, reasoning, and action—as the most radical intensification of the flow paradigm yet: by removing the human cognitive bottleneck that previously constrained flow-based coordination, AI extends the logic of real-time adaptive responsiveness across all domains of the economy and science, rendering stockpile-era institutions not merely inefficient but structurally obsolete. The paper concludes by situating the stockpile-to-flow transition within a libertarian framework of ‘markets and discourses’ (Schumacher, 2025), in which voluntary market coordination and open critical discourse together substitute for the coercive planning functions of the state—delivering the adaptive, flow-based governance that stockpile-era institutions can no longer provide.

Introduction

For much of the twentieth century, economic and organisational systems were built around the logic of stockpiling. Firms accumulated inventories, states maintained reserves, archives stored knowledge, and infrastructures were dimensioned for peak demand. This logic reflected a world in which access to information was slow, adjustment costly, and uncertainty addressed through precautionary accumulation. Welfare benefits were allocated through fixed categories, public services were planned for forecast loads, and urban land use was stabilised through long-term zoning and comprehensive master plans.

In recent decades, a contrasting logic has emerged. Production is increasingly demand-driven, information circulates instantaneously, and capacity is provisioned dynamically. Instead of stockpiles, contemporary systems rely on flows. This paper proposes that the transition from stockpiles to flows constitutes a structural transformation in the epistemology of coordination. It is not simply that systems have become more efficient; rather, they have shifted from anticipating the future to responding to the present.

This transformation spans technology, economics, organisational form, culture, welfare, governance, and urban planning. Many contemporary crises in welfare provision and spatial governance, the paper argues, can be understood as institutional lag: stockpile-era coordination mechanisms operating within flow-era conditions. The paper traces this transformation from its technological origins through its organisational and political consequences, and concludes by arguing that the flow-based logic, when followed to its institutional conclusion, points toward a libertarian social order constituted through markets and discourses rather than through state planning.

Stockpiles as a Coordination Strategy

Stockpiles are a rational response to uncertainty under conditions of slow feedback. When information about demand, failures, or opportunities arrives late, systems compensate by storing resources in advance. Warehouses, safety buffers, spare capacity, and archival repositories all embody this logic. The core assumption is that future states of the world cannot be reliably inferred from present signals, and that resilience therefore requires accumulation.

In economic terms, stockpiles act as insurance. Inventories protect against supply disruptions; excess capacity absorbs demand spikes; bureaucratic rules stabilise expectations. In organisational terms, stockpiles correspond to hierarchical planning, long budgeting cycles, and document-heavy control mechanisms. Knowledge is codified, stored, and disseminated through manuals, plans, and fixed procedures.

This logic dominated industrial capitalism, Fordist production, and welfare-state governance. It also shaped early computing: magnetic tapes, disk drives, and physical archives all imposed access latencies that reinforced batch processing and long planning horizons. Under these conditions, coordination depended on prediction rather than adaptation.

The Technological Hinge

The transition from stockpiles to flows was enabled by a number of mutually catalyzing decisive technological shifts: One of them, seemingly minor, was the move from kinetic to non-kinetic storage. Traditional storage media—tape reels, spinning disks, optical drives—located data by physical position. Access required movement: rewinding, seeking, rotating. Latency was inherent and unavoidable.

Solid-state storage, by contrast, locates data electronically. Flash memory and related technologies allow data to be addressed directly through electrical signals, eliminating mechanical delay. When combined with high-bandwidth networks and cheap processing power, this shift collapses access time by orders of magnitude. Retrieving information later becomes almost as fast as storing it now.

This collapse of latency undermines the rationale for stockpiling. If information, capacity, or content can be accessed instantly, there is little advantage in holding large reserves in advance. Storage ceases to be a precautionary buffer and becomes a dynamic resource within a continuous flow.

Flows and Adaptive Coordination

Flow-based systems replace accumulation with responsiveness. Instead of preparing for hypothetical futures, they adjust continuously to observed conditions. Just-in-time production synchronises supply with demand. Cloud computing provisions resources elastically. Streaming services deliver content on demand rather than distributing physical copies. Software is updated continuously rather than released in fixed versions.

The epistemic assumption underlying flows is fundamentally different from that of stockpiles. Rather than compensating for ignorance with reserves, flow-based systems exploit fast feedback to learn in real time. Coordination emerges from signals—prices, metrics, usage patterns—rather than from central plans.

This shift aligns closely with market-based coordination as analysed by Friedrich Hayek. Hayek argued that dispersed knowledge cannot be aggregated ex ante, but can be mobilised through price signals that reflect local conditions. Flow-based systems generalise this insight beyond markets, embedding feedback loops into technical and organisational infrastructures. The Hayekian point is decisive: the very same dispersal and situational specificity of knowledge that defeats central planning is precisely what flow-based systems are designed to exploit.

Post-Fordism, Flexible Specialisation, and the Knowledge Economy

The stockpile-to-flow transition in technology and logistics did not occur in isolation. It coincided with—and was mutually reinforced by—a broader restructuring of the productive economy, widely characterised as the shift from Fordism to post-Fordism. Fordist production was organised around standardised mass manufacturing, long production runs, dedicated machinery, economies of scale, and vertically integrated corporate hierarchies. Its social correlate was the stable, full-time male wage and the comprehensive welfare state designed to insure against interruptions to that wage.

From the 1970s onward, this model was progressively displaced by what Michael Piore and Charles Sabel termed ‘flexible specialisation’: production systems capable of rapid retooling, short runs, and customised output, delivering economies of scope rather than scale. The competitive advantage shifted from producing large volumes of identical goods cheaply to producing differentiated goods responsively. Programmable automation, modular supply chains, and networked production allowed firms to match variety to demand, generating innovation capacity that rigid Fordist structures could not sustain.

This restructuring was simultaneously a knowledge-economy transformation. Value creation shifted from the physical manipulation of materials toward the generation, processing, and application of information. Human capital, intellectual property, and organisational learning became the primary sources of competitive advantage. The firm itself was reconceived—from a stock of physical assets to a nexus of capabilities, relationships, and flows of knowledge.

The post-Fordist economy thus exemplifies the stockpile-to-flow logic at the level of production itself. Inventory buffers gave way to just-in-time supply; standardised job descriptions gave way to project-based teams; fixed product catalogues gave way to continuous iteration. The organisational correlate was the decline of the vertically integrated corporation in favour of modular, networked arrangements—platform firms, ecosystem models, and agile development methodologies.

Crucially, the historical viability of centralised planning and of the comprehensive welfare state was itself bound to the Fordist era. As Mises and Hayek demonstrated, socialist central planning was always inferior to market coordination, but it retained a certain feasibility during the era of concentrated, large-scale mechanical mass production, when the most important industries had consolidated into a small number of horizontally and vertically integrated concerns, and when the massive economies of scale—together with the punitive diseconomies of scope—implied a strong tendency toward the equalisation of living standards around a universal consumption bundle. To nationalise and centrally plan such an economy was at least operationally conceivable. Under post-Fordist conditions, where competitive advantage depends on rapid differentiation, distributed innovation, and knowledge-intensive coordination, centralised planning becomes not merely inefficient but structurally impossible. The state’s stockpile logic is no longer merely suboptimal; it is actively incompatible with the productive forces it purports to govern.Organisational Transformation

The move from stockpiles to flows has reshaped organisational form across both private and public sectors. Stockpile-oriented organisations tend to be hierarchical, with clearly defined roles, rigid procedures, and long planning cycles. Errors are discovered late and corrected slowly. Accountability is retrospective and often punitive.

Flow-oriented organisations, by contrast, are structured around continuous monitoring and iteration. Performance is tracked in real time through dashboards and analytics. Processes are versioned rather than finalised. Learning is embedded in operation rather than confined to periodic reviews.

This transformation explains the rise of agile methodologies, platform-based firms, and ecosystem models of production. Coordination is no longer achieved by freezing decisions in advance, but by maintaining the capacity to adjust. The contrast with public-sector bureaucracies—still overwhelmingly organised around stockpile logic—is increasingly stark.

Cultural and Social Implications

The stockpile-to-flow transition extends beyond economics and organisation into culture. Stockpile cultures emphasise ownership, permanence, and stability. Libraries, personal collections, and physical archives embody this orientation. Value is associated with possession and durability.

Flow cultures prioritise access, updateability, and responsiveness. Media is streamed rather than owned. Knowledge is searched rather than stored locally. Identity and reputation are maintained dynamically through continuous activity rather than fixed credentials.

This cultural shift reshapes expectations of institutions. Citizens increasingly demand real-time services, adaptive policies, and personalised responses. Systems designed around static categories and fixed entitlements struggle to cope with these expectations, revealing a growing mismatch between flow-based social practices and stockpile-based governance structures.

Welfare Systems: Stockpile Logic Under Strain

Modern welfare systems emerged under conditions of limited information and high administrative cost, and in close alignment with Fordist production and its associated life-course assumptions. Eligibility rules, benefit categories, and entitlement thresholds were designed to stabilise expectations and ration scarce resources. Stockpiling took the form of fixed budgets, categorical eligibility—disabled, unemployed, lone parent—and long-term benefit awards.

This logic treats welfare as a reserve to be drawn upon when individuals fall into predefined states of need. It assumes that such states can be reliably identified in advance and managed through standardised procedures. Administrative discretion is constrained in order to preserve fairness and predictability.

However, under post-Fordist conditions this model performs poorly. Life trajectories are increasingly fluid, employment patterns unstable, and household configurations diverse. Fixed categories struggle to capture dynamic circumstances. As a result, welfare systems oscillate between over-inclusion and over-exclusion, producing both fiscal strain and social resentment.

Crucially, stockpile logic encourages defensive bureaucratic behaviour. Because resources are pre-allocated and accountability is retrospective, agencies prioritise risk avoidance over adaptive problem-solving. This contributes to adversarial assessment cultures, procedural overload, and escalating compliance costs. From a flow perspective, welfare should move toward income smoothing, responsive support, and graduated withdrawal of benefits rather than abrupt categorical thresholds.

Governance as Precautionary Accumulation

Governance systems beyond welfare exhibit similar characteristics. Regulatory frameworks rely on exhaustive rule-making to anticipate future contingencies. Compliance is verified through documentation rather than performance. Public-sector accountability is structured around audits and inspections that occur long after decisions are made.

Such systems reflect an epistemology of anticipation: the belief that risks must be neutralised in advance through comprehensive rules. In practice, this leads to regulatory inflation, institutional inertia, and the displacement of responsibility from outcomes to procedures.

By contrast, flow-based governance relies on continuous monitoring and rapid correction. Instead of attempting to foresee all possible failures, it tracks performance in real time and adjusts accordingly. This approach aligns with cybernetic models of control and with Hayek’s insight that knowledge relevant to coordination is dispersed, situational, and time-sensitive.

The persistence of stockpile governance in a flow-capable environment results in a paradox: ever more rules generate ever less effective control, while undermining trust in institutions. The economic stagnation afflicting advanced societies can be traced, in significant part, to this pervasive regulatory accumulation, which increasingly paralyses technological and entrepreneurial dynamism precisely when the post-Fordist knowledge economy demands maximum adaptive freedom.

Urban Planning and the Freezing of Capacity

Urban planning offers a particularly clear illustration of stockpile logic. Zoning systems freeze land-use rights decades in advance, effectively stockpiling permitted capacity regardless of evolving demand. Infrastructure is planned for forecast peak loads, often resulting in underutilisation or congestion when conditions diverge from projections.

This approach assumes that urban futures can be stabilised through ex ante design. Yet cities are complex adaptive systems characterised by emergent behaviour, path dependence, and rapid change. Fixed land-use categories and rigid development controls inhibit adjustment, raising prices and constraining supply.

The result is a familiar pattern: housing shortages coexist with underused land; infrastructure bottlenecks persist despite ample capital; and political conflict intensifies as planning decisions become zero-sum allocations of scarcity.

From a flow perspective, urban planning should prioritise adaptive capacity rather than fixed outcomes. This implies regulatory frameworks that permit continuous intensification, incremental change, and responsive infrastructure provision, guided by real-time indicators rather than static master plans. It also implies rethinking the relationship between the plan and the signal—replacing rigid zoning with performance-based codes that define parameters within which adaptive responses can occur.

The Epistemic Limits of Flow-Based Coordination: Why Markets Are Necessary but Insufficient

The argument thus far has shown that flow-based coordination, grounded in real-time feedback and dispersed signals, is epistemically superior to stockpile-based planning under conditions of complexity and rapid change. Markets are the paradigmatic flow institution: Hayek’s price mechanism is a flow of compressed information, continuously updated, exploiting precisely the dispersed, situational, time-sensitive knowledge that centralised planning cannot access. The post-Fordist knowledge economy amplifies this advantage, since competitive success now depends on rapid differentiation, distributed innovation, and adaptive responsiveness—all of which require the informational fluidity that only markets provide.

However, markets as discovery procedures have an epistemic limitation that the stockpile-to-flow framework makes visible. Markets discover through competitive experimentation: firms try, consumers choose, and prices aggregate the results. But it is never trivial to attribute the success of a complex venture to specific features of the effort. Market outcomes reveal that something worked or failed, but they do not, by themselves, explain why. Extracting generalisable knowledge from market experimentation—understanding causal mechanisms, identifying transferable principles, distinguishing signal from noise—requires a complementary form of coordination that operates by different rules: critical discourse.

This point becomes increasingly consequential as the knowledge economy matures. In our sophisticated post-Fordist economy, nearly all economic production is highly dependent on scientific and technological discourses. The same applies to consumption choices: medical therapies, financial instruments, legal frameworks, and engineering solutions are all too complex and opaque in their effectiveness to allow for fail-safe intuitive choices. Market participants who ignore the findings of relevant discourses are not exercising sovereign preference; they are operating with degraded information. The rationality of market exchanges is increasingly underpinned by competencies that only non-strategic communicative interaction—discourse—can deliver.

Markets and Discourses: The Two Flow Mechanisms of Social Coordination

The recognition that markets require discursive complementation leads to a broader reconceptualisation of social order. Drawing on Jürgen Habermas’s theory of communicative action, we can distinguish two fundamentally different types of social interaction, each constituting a distinct flow mechanism. Strategic or instrumental action—action oriented toward individual success in competition with others—is the basis of market exchange. Communicative action—interaction oriented toward mutual understanding through the exchange of reasons—is the basis of discourse.

Markets and discourses share important structural characteristics. Both are bottom-up, decentralised, and inherently based on free and voluntary interactions. Both can only function properly and realise their rationality potential if they are open-access institutions: entry must not be barred by incumbents, and neither markets nor discourses must be monopolised. Both are flow institutions par excellence: they operate through continuous signals, iterative adjustment, and real-time feedback rather than through fixed allocations and ex ante plans.

However, discourses are not markets and operate by fundamentally different rules. Market exchanges are legitimately self-interested; discursive exchanges serve the common interest in truth and knowledge. In markets, it is perfectly permissible to instrumentalise counterparties as means to private ends; in discourse, any attempt to do so—withholding information, arguing strategically rather than sincerely—undermines the process and, if detected, leads to communicative breakdown. Discourses aim for consensus through the exchange of arguments; markets coordinate through the exchange of commodities and services. The ethics of business dealings is categorically distinct from the ethics of discourse.

Habermas’s crucial insight, developed in his Theory of Communicative Action, is that instrumental reason is dependent upon—indeed parasitic upon—communicative reason. The very constitution of rational individual selves, including their goals and subjective interests, depends not only on prior socialisation but on continuous elaboration and confirmation through non-strategic communicative interaction. Means-end calculations presuppose knowledge of possible means that can only be gained through prior communicative interaction. Reliable knowledge, including scientific knowledge, is always a collective social endeavour in which sincerity, repression-free inclusiveness, and openness to criticism are presupposed.

Classical liberal and libertarian social philosophy have historically been blind to this distinction. By modelling all action as instrumental and all coordination as market-based, they fail to account for discourse as a necessary and irreducible mechanism of societal self-regulation. This blind spot leads to an impoverished conception of social order—one that cannot explain how market participants acquire the knowledge, norms, and rational agency that market participation itself presupposes.

The Historical Materialism of the Flow Transition

The stockpile-to-flow analysis acquires its full force when situated within a materialist conception of history. The viability of any given institutional order depends on the historical level of technological and organisational development. Stockpile-based institutions—the comprehensive welfare state, regulatory governance, centralised urban planning—were not always dysfunctional. They were the appropriate coordination mechanisms for an era of concentrated mechanical mass production, limited information processing, and slow feedback.

This historical perspective is essential. It distinguishes the present argument from an ahistorical libertarianism that treats the state as inherently illegitimate at all times and places. States throughout most of history, including the modern Western states from the seventeenth century onward, have at times made significant contributions to civilisational and economic development. The competitive systems of smaller states—from the ancient Greek city-states through the Northern Italian republics to the competing German principalities of the eighteenth and nineteenth centuries—were often particularly productive. The twentieth-century social-democratic mixed economies grew faster on average than the nineteenth-century laissez-faire economies. These are facts that a historically informed libertarianism must acknowledge.

The decisive point, however, is that these institutional forms have been overtaken by transformative technological change. The shift from Fordism to post-Fordism—from mechanical mass production to computationally empowered flexible specialisation and knowledge-intensive value creation—has altered the material conditions under which social coordination operates. Under these new conditions, the state’s stockpile logic is no longer a viable development model but an active impediment. The unleashing of current technology-based prosperity potentials requires a corresponding institutional transformation: from centralised, coercive, stockpile-based governance toward decentralised, voluntary, flow-based coordination through markets and discourses.

Discourse Capitalism: Toward a Flow-Based Social Order

The convergence of the stockpile-to-flow analysis with the markets-and-discourses framework yields a coherent vision of institutional reform—or, more precisely, of institutional transformation. If the core dysfunction of contemporary welfare, governance, and urban planning is the persistence of stockpile-era coordination mechanisms in a flow-era environment, then the remedy is not incremental adjustment within existing institutional frameworks but a fundamental reconfiguration toward flow-based coordination. And if flow-based coordination operates through two complementary mechanisms—markets and discourses—then the institutional question becomes: how can markets and discourses together substitute for the coercive planning functions that the state currently performs, poorly, through stockpile logic?

This is the question addressed by the concept of ‘discourse capitalism’ (Schumacher, 2025). Discourse capitalism proposes a voluntary, decentralised society in which markets handle the coordination of production and exchange through competitive price signals, while discourses handle the generation of knowledge, the formation of norms, and the continuous adaptation of the systems of rules that frame social life. Both mechanisms are wholly voluntary, open, repression-free, and non-coercive. A society based on markets and discourses as its mechanisms of societal integration is an inherently free society.

In such a society, the functions currently performed by the state—welfare provision, urban planning, regulatory governance, law, education, infrastructure—would be delivered through competitive market provision guided by open critical discourse. Political discourse would not cease after the withdrawal of state power; on the contrary, it would intensify. When issues are no longer reduced to two or three policy bundles to be voted on, and instead many possible trends and developments can simultaneously be experimented with, compared, analysed, and debated, the scope and energy of political life expands dramatically. Legal discourse, for instance, would continue and indeed flourish once the development of the legal system has been transferred from the sluggish hands of state courts and parliaments to the initiative of competing entrepreneurs in markets for law provision.

The discourse-capitalist framework thus resolves a tension within the stockpile-to-flow analysis. The analysis demonstrates that flow-based coordination is epistemically superior to stockpile-based planning under post-Fordist conditions, but it does not by itself specify the institutional form through which flow-based coordination should be achieved. Discourse capitalism provides the answer: markets as the flow mechanism for economic coordination; discourses as the flow mechanism for normative, scientific, and political coordination; both operating without coercion, without centralisation, and without the stockpile logic that now strangles the adaptive capacity of advanced societies.

Limits, Tensions, and the Hybrid Minimum

Despite its advantages, flow-based coordination is not without risks. Just-in-time systems can be fragile under extreme shocks, as global supply-chain disruptions have demonstrated. Continuous monitoring can slide into surveillance. Real-time optimisation may privilege short-term signals over long-term considerations.

Moreover, not all domains can or should be fully de-stockpiled. Strategic reserves, redundancy, and slack remain essential in contexts where failure is catastrophic. Emergency preparedness and insurance against catastrophic risks retain a legitimate need for precautionary accumulation.

The challenge is therefore not to abolish stockpiles entirely, but to understand when accumulation is genuinely necessary and when it merely compensates for institutional inertia. In welfare and urban policy, this suggests maintaining baseline guarantees (via private insurance or privately organised solidarity) while replacing rigid categorical systems with responsive, signal-driven mechanisms above that floor. The discourse-capitalist framework accommodates this: discourses can identify and debate which domains require precautionary reserves, while markets can price and provide the corresponding insurance mechanisms—all without recourse to coercive state planning.

Flow-based systems also risk appearing technocratic or opaque, particularly when mediated by algorithms. Adaptive allocation may be perceived as unstable or unfair if not properly communicated. Legitimacy in flow-based governance therefore requires transparency of process and clarity of principle. Here again, discourse is indispensable: it is through open, critical, public deliberation—not through bureaucratic procedure or electoral mandate—that flow-based institutions earn and maintain their social licence.

Discourse implies the authorship of contributions by responsible social actors. Even if AI models are being consulted or doing all the heavy cognitive and inferential lifting, discursive claims must (for now) be made by social actors, by persons with societal standing staking their reputations within social networks, select communities or within human society at large. With respect to public discourse this involves taking a position with conviction and commitment, taking responsibility for knowledge claims, making the use of AI resources (here GPT-5.4, Claude Opus 4.6) transparent and by rehearsing and appropriating premises, explanations and arguments. Taking a stance also involves justification of decisions via reasons, if contested going back to first principes.

Conclusion

The transition from stockpiles to flows marks a fundamental reorganisation of how complex societies coordinate under uncertainty. Enabled by non-kinetic storage, high-speed networks, and cheap computation, and coinciding with the post-Fordist restructuring of the economy from mass production to flexible specialisation and knowledge-intensive value creation, flow-based systems replace precautionary accumulation with adaptive responsiveness.

This shift transforms not only economic coordination and organisational design, but also the institutional architecture of welfare, governance, and urban planning. Many contemporary dysfunctions in these domains arise from the persistence of stockpile logics—fixed categories, pre-allocated capacities, frozen land-use designations, and exhaustive rule-making—within environments that now operate according to flow-based dynamics.

Reform requires more than efficiency gains or budgetary adjustments. It demands an epistemic shift: from anticipation to responsiveness, from fixed categories to continuous gradients, from frozen capacity to adaptive flows. Welfare systems must move from static entitlements toward dynamic support; governance from rule accumulation toward performance feedback; urban planning from zoning-based scarcity toward capacity orchestration.

But the deepest implication of the stockpile-to-flow analysis is institutional. Flow-based coordination operates through two complementary mechanisms: markets, which coordinate production and exchange through competitive price signals, and discourses, which generate knowledge, form norms, and guide the continuous adaptation of social rules through the exchange of reasons. Both are voluntary, decentralised, open-access, and non-coercive. Both are flow institutions par excellence. Together, they constitute what has been termed ‘discourse capitalism’ (Schumacher, 2025)—a framework in which markets and discourses, operating without state coercion, can substitute for all the coordination functions that stockpile-era governance now performs with diminishing competence.

At its core, the move from stockpiles to flows represents a new epistemology: one that treats uncertainty not as a problem to be neutralised in advance through the accumulation of rules, reserves, and coercive authority, but as a condition to be navigated continuously through the adaptive intelligence of free markets and open discourse. The future progress of human civilisation within a post-Fordist, knowledge-intensive world will be ordered via markets and steered via discourses. Understanding this transformation is essential for rethinking coordination across social, political, and spatial domains in the digital age.

Contemporary firms increasingly operate under a profound structural tension: they are held ever more responsible for outcomes in their external relations while simultaneously losing authority and sovereignty in their internal governance. A schism opened that undermines the rational management our prosperity depends on: Internal sovereignty is eroded while external responsibility is increasing.

Legal, regulatory, and reputational demands imposed on firms by customers, regulators, and the public have expanded dramatically over recent decades, while managerial authority over employees has been progressively constrained by confidentiality rules, privacy protections, anti-discrimination law, and proceduralised employment rights. This growing mismatch produces what may be described as an internal sovereignty collapse in the face of external responsibility inflation. Firms remain fully accountable for performance, safety, and conduct, yet are increasingly deprived of the informational access, discretionary authority, and decisional speed required to meet these obligations effectively.

Externally, the firm is treated by law and public opinion as a unified, decisive, and fully responsible actor. Consumer protection regimes increasingly operate on strict liability principles; regulatory frameworks impose affirmative duties of care, safeguarding, and prevention; and reputational exposure has intensified through social media, real-time ratings, and algorithmic amplification. Failures are judged rapidly and publicly, often without regard to internal organisational complexity. From product safety and service quality to safeguarding failures and workplace culture scandals, responsibility is outcome-based, immediate, and largely non-deferrable. The firm is expected to act as a sovereign entity capable of decisive intervention and comprehensive control.

Internally, however, the same firm is governed under a markedly different logic. Employment law, data protection regimes, and equality legislation increasingly constrain the ability of management to observe, evaluate, and discipline employees. Confidentiality requirements restrict the upward flow of information, even within senior management structures; privacy law limits monitoring of productivity, communication, and behaviour; and anti-discrimination doctrines create significant litigation risk around ordinary acts of performance management. Medical and mental-health frameworks further complicate internal governance by introducing categories of claim that are difficult to falsify, tightly protected as special category data, and often insulated from managerial scrutiny. As a result, authority becomes fragmented, proceduralised, and legally risky, while decisional latency increases and informal judgement is displaced by defensive compliance.

This asymmetry is not accidental. It reflects a broader historical shift in which the regulatory state has increasingly inserted itself within private organisations. Human Resources departments have become key institutional nodes in this process. Rather than focussing primarily on personnel and talent development, HR increasingly operates as an internal compliance apparatus mediating between the administrative state and the firm. Its dominant orientation is not performance optimisation but liability containment. While contributing less in a positive sense to firms, the number of HR professionals has significantly increased in recent years. According to the Office for National Statistics (ONS), the number of people working in HR roles in the UK has risen sharply over the last 15 years. There were about 284,000 HR professionals in 2010 and around 476,000 by the early 2020s, a dramatic increase of roughly 68% over this period. This suggests HR employment has grown much faster than many other professions. The consequence of the functional transformation of HR departments towards becoming employment compliance departments, in response to regulatory intrusion, is a structural transformation of management itself: decisions are filtered through procedural risk assessments, documentation requirements, and confidentiality constraints, while substantive judgement is deferred or avoided altogether.

This development echoes long-standing warnings in management theory. Peter Drucker argued that responsibility without corresponding authority undermines the very possibility of effective management. Managers held accountable for outcomes they cannot control are driven toward risk avoidance, symbolic action, and non-decision. Similarly, Henry Mintzberg criticised the displacement of managerial judgement by formalised systems, warning that organisations increasingly substitute process for coordination and paperwork for leadership. What these theorists anticipated has now been intensified by legal and regulatory developments that systematically privilege procedure over discretion.

Legal scholars associated with classical liberal and libertarian traditions have articulated parallel critiques. Richard Epstein has argued that modern employment and anti-discrimination law erodes freedom of contract and transforms firms into defensive bureaucracies, more concerned with litigation avoidance than productive coordination. This logic produces organisations that are governed less by managerial responsibility than by procedural orthodoxy.

With respect to public administration (rather than private corporations) Philip Hamburger interprets the expansion of administrative governance as a shift of power away from accountable authority toward rule-bound compliance structures that lack democratic legitimacy. Hamburger interprets administrative expansion as a constitutional displacement: government by accountable decision-makers is replaced by government through rules, procedures, and compliance mechanisms that appear lawful and rational, yet lack the core democratic virtues of responsibility, contestability, and consent.

The economic implications of this shift can be further illuminated through the lens of knowledge and coordination. Friedrich Hayek emphasised that effective social and economic coordination depends on the use of local, contextual knowledge and discretionary judgement. Over-legalisation crowds out these capacities by replacing situational evaluation with abstract rules. Within firms, the same mechanism operates: performance management becomes detached from lived organisational knowledge, and authority is displaced by formal compliance processes that are poorly suited to complex, evolving work environments.

Cultural critics such as Frank Furedi add a further dimension by highlighting the expansion of therapeutic and safeguarding paradigms into workplace governance. Mental health, wellbeing, and vulnerability frameworks, while addressing genuine problems, also introduce powerful veto points into management. In the most recent ONS labour market report, mental health conditions accounted for about 9.8 % of all sickness absence occurrences in 2024. Both the absolute numbers of sick leave and the percentage due to mental health have sharply increased in recent years. UK employees reportedly took an average of 9.4 sick days last year, roughly two full working weeks. This is up from 5.8 days before the pandemic. When performance issues are reframed as health conditions and insulated by confidentiality and non-discrimination norms, managerial authority is further weakened. The cumulative effect is an internal environment in which decisive action is systematically discouraged.

The result of these intersecting developments is a widening governance schism. On the external side, firms face hard, immediate, and often irreversible sanctions for failure. On the internal side, they operate under soft, proceduralised, and contestable authority structures that privilege individual rights over organisational coordination. Firms are thus expected to perform as sovereign actors externally while being governed as rights-constrained micro-polities internally. Responsibility is unified; authority is fragmented. Accountability is strict; control is diluted.

This asymmetry has tangible consequences. Decision-making slows, managerial risk aversion increases, and informal power dynamics proliferate as formal authority weakens. HR functions become opaque buffers rather than enabling structures, and managers increasingly manage around formal processes rather than through them. Productivity suffers, not because rights exist as such, but because the balance between protection and governance has shifted beyond a sustainable threshold.

It is important to stress that this critique does not deny the historical necessity of employee protections or the reality of abuse and discrimination. Rather, it advances a threshold argument: beyond a certain point, protective regimes cease to correct power imbalances and instead invert them, generating systemic inefficiency and organisational fragility. After all, it is the owners and leaders of the firms themselves who have the keenest interest in weeding out abuse and discrimination within their own organisation. It makes much more sense to rely on them to define the most appropriate means to ensure a fair and meritocratic work environment.

When internal sovereignty collapses while external responsibility continues to inflate, firms become accountable without being governable. The long-term risk is not merely reduced efficiency but a deeper erosion of organisational responsibility itself, as decision-makers retreat into procedural compliance and symbolic action.

Rebalancing this relationship requires neither the abolition of rights nor a return to arbitrary authority. It requires a recalibration that restores proportionality between responsibility and control, narrows confidentiality to what is strictly necessary, and re-legitimises good-faith managerial judgement. Without such a recalibration, the contemporary firm risks becoming an institution that is simultaneously overburdened and underpowered, exposed to ever greater external demands while hollowed out from within. It is highly likely that the increasing institutional schism discussed here is one of the contributing factors of the economic stagnation we witness despite the simultaneous technological advancement.

In England, statutory responsibility for child protection, care proceedings, foster placements and adoption lies with each council’s Children’s Services Department (often called Children’s Social Care or Children and Families Services), led by a Director of Children’s Services under the Children Acts 1989 and 2004. These departments investigate concerns, initiate care proceedings, place children with foster carers, children care centres or for adoption, and act as “corporate parents” for children in care. They operate within national guidance set by central government (primarily the Department for Education) and are inspected by Ofsted, but they retain wide local discretion over thresholds, interpretation of risk, and day-to-day practice.

That discretion has been central to long-standing criticism. On the one hand, there is credible evidence and official confirmation that Children’s Services departments failed catastrophically with respect to child sexual exploitation by predominantly Pakistani grooming gangs in Rotherham, Rochdale, Oxford and Telford a.o. Independent inquiries and serious case reviews established that the very agencies charged with safeguarding children repeatedly ignored warnings, disbelieved victims, normalised abuse, and prioritised reputational or political concerns over child protection. These scandals were not marginal errors but systemic failures of local authority children’s services.

On the other hand, in more recent years, critics have argued that the safeguarding system has swung toward over-intrusive and defensive practice, particularly since around 2010. The expansion of safeguarding duties, broader definitions of harm (including emotional and psychological harm), mandatory reporting cultures, and fear of blame have created incentives for local authorities to intervene early and aggressively. Parents and commentators have complained that some Children’s Services departments exaggerate risk, rely on biased or one-sided representations, and apply vague concepts such as “domestic abuse”, “controlling and coercive behaviour”, “trauma” or “risk of emotional harm” in ways that draw ordinary families into investigations or care proceedings. While there is no single national dataset proving over-intervention, parliamentary evidence, academic research, ombudsman findings and qualitative studies consistently point to a systematically risk-averse institutional culture prone to false positives and a tendency to escalate (rather than de-escalate) intervention once concerns have been registered in the system.

Why and how might first safeguarding concerns arise?

It often seems as if the safeguarding priority is not to keep the children safe but to safeguard the children services department, i.e. mitigate risk to the department and to social workers (in case something might happen to a child later after something had been reported but not pursued). It has been made mandatory for all professionals getting into contact with children (at school, at hospital, at GP practices, as health visitor, police or social worker) to report any concern or perceived risk. Once concerns have been raised the institutional dynamics seems to be like a one-way rachet. A hospital visit with an injured child can trigger safeguarding checks, and may result in a concern being logged with Children’s Services. Professionals are encouraged to “err on the side of caution”. For parents any contact with child-related public services carries latent risk. Here is another type of incident that often leads to the safeguarding process being triggered: Any conflict between parents (e.g. involving shouting that is noticed by neighbours leading to police call outs) is judged to put children living in the home at risk (whether the children were present at the conflict incident or not). Police call outs involve separate on the spot interviews with the conflict partners, using checklist questions like “are you afraid of your partner”, “is your partner threatening you” or “is your partner engaged in coercive or controlling behaviour?” Conflict is often interpreted as “domestic abuse” with the father mostly identified as abuser and the mother as the victim of abuse. (The threshold for interpreting conflict as domestic abuse has been lowered since the Serious Crime Act 2015 (Section 76), especially with the criminalisation of so called coercive or controlling behaviour. Police are trained to err on the side of categorising incidents as domestic abuse, even where there is no physical assault.) Once a concern has been logged in this fashion, the next step is to engage systematically with parents and if sufficient “progress” is lacking move to formalise a safeguarding agreement, to be signed by the problematic parents, with a tight schedule of supervision, including unannounced home visits, as well as visits at the child’s or children’s nursery or school. Such visits are very likely to raise further concerns. Any probing is likely to throw up evidence of potential risk. The child might misbehave in school, a choking hazard or a pair of scissors might be found in the child’s room or in reach. The safeguarding agreement also involves participation in meetings, as well as participation in therapy or group training sessions, like domestic abuse training/therapy. (If the mother does not engage with the supervisory meetings or does not participate in group meetings, the institutional presumption seems to be that the (oppressive) father prevents engagement. This perceived risk of coercion also prevents the social workers to accept any attempt to correct a prior statement made during police call outs. Admittance to such sessions – e.g. a 26 week course for supposed domestic abusers – requires admission of guilt as precondition for being allowed to participate.

The refusal to confess to what might be false or exaggerated claims of abuse will be interpreted as denial or resistance. The victim of the supposed abuse might also be required to join a course. Here too, if the supposed victim does not admit being abused or victimized, then this is interpreted as lack of insight and lack of capacity to safeguard the child or children in the face of abusive situations. This “denial” or “lack of insight” in turn will be interpreted as lack parent cooperation leading to further escalation of the intervention, i.e. an application to the family court requesting a “care order”, i.e. the removal of the children. Once the Local Authority has decided to make such an application because the 6 months process under the safeguarding agreement did not lead to sufficient progress or cooperation to close the case (even if no further incident was recorded), the tendency of the LA is naturally to exaggerate the situation in order to make the court application plausible. A court hearing might be scheduled on short notice (48 hours) and the court proceeding is such that there is no scope or allocated time to question the supposed facts of the case presented by the LA. The only question the court considers at this stage is whether the legal threshold for an interim care order (where the LA takes over from the parents as the primary carer with operational decision competence) is met considering the facts as alleged by the LA. After a 60-90 minutes hearing the child or children in question might then be taken into foster care immediatly after the hearing (supposedly because parents might “interfere” or “abscond”). Only thereafter begins a lengthy process of exchanging statements, further hearings and a process of professionally assessing the parents. That shocked and dismayed parents become adversaries in this process is probably to be expected. (Readers might think: Thank God that my family did not trip up and fell into the fangs of our local Children Services Department. Who knows if we could have extricated ourselves from these self-propelling institutional dynamics.)

With respect to the time frame allocated to care-related court proceedings, the government guidance (since the Children and Families Act 2014) is that proceedings should not exceed 26 weeks (6 months). However, the actual average length of such proceedings in the UK is above 9 months. Official statistics indicate median case durations of 35-40 weeks with a significant tail of cases extending well beyond 12 months, i.e. delay has become structurally normalised, despite the formal rule.

Children Services and parents become adversaries

This inherent adversarial tension between social services and parents is visible at the local level, often even before the courts get involved. There are formal complaints routes through the council’s complaints unit, with independent investigation stages and ultimately the Social Care Ombudsman. The existence and frequent use of these mechanisms reflects ongoing disputes between families and children’s services about proportionality, fairness, truthfulness and common sense in safeguarding decisions.

As an aside: A very different but also contentious issue also reflects this underlying conflict of interest between Local Authorities and parents: the issue of gender identity and so-called “gender-affirming” approaches for under-18s. Medical treatment has always been the responsibility of the NHS, not local authorities, but Children’s Services departments have nonetheless played a significant role through safeguarding assessments, school partnerships, and decision-making for children in care. Evidence presented to Parliament shows that in some areas social workers and safeguarding partnerships adopted an uncritical “affirmation-first” stance, sometimes excluding parents or treating parental caution as emotional abuse. While local authorities did not prescribe medical interventions, they often shaped the institutional environment that normalised secrecy and sidelined parents. This has now been formally criticised, with national guidance moving to re-assert parental involvement and clinical neutrality.

The conflicts between parents and Children Services point towards a deep ideological and structural problem rather than isolated misconduct. UK Children’s Services have been criticised both for failing to intervene decisively when children were gravely at risk and, more importantly, for intervening too readily and intrusively in family life under expansive safeguarding doctrines. The same institutional features — broad discretion, high stakes, political pressure, weak accountability, and fear of blame — help explain both pathologies. The result is a system that remains immensely powerful, legally coercive, and uneven in practice, and which continues raise serious doubts about proportionality, parental rights, due process and the proper limits of state intervention in family life.

Children Services’ increasingly over-zealous interventions

Children Services interventions and care proceedings against parents have seen a steep increase from 2010 to 2015. This is due to the institutional changes, affecting incentives, especially since 2008. and possible incentives in the system that can explain this. Also, it is legitimate to suspect and is there evidence that many of these interventions might do more harm than good, over and above the costs to the taxpayers.

From around the late 2000s and especially between 2010 and the mid-2010s, England’s child protection and family justice system underwent a marked shift towards more intensive and more coercive intervention in family life. This shift is visible most clearly in two headline trends. First, the number of children looked after by local authorities rose steadily from around 59,000 in 2008 to over 72,000 by 2017, an increase of roughly 22 per cent over less than a decade. By 2025 it had further risen to 84,000. Second, and more diagnostically significant, the use of court issued care orders—the most intrusive legal mechanism available to the state under the Children Act 1989—increased sharply after 2012, rising by roughly 40 % from 2012 to 2017. Children subject to a care order are a subset of the children looked after by the Local Authorities. This steeply increasing subset of children have been forcibly removed from their parents’ care and placed with foster parents. These are not marginal changes at the edges of the system; they represent a qualitative expansion in the routine use of coercive state power over families. On 31^st March 2025 there were 55,000 children under forced care order, i.e. 65% out of the 84,000 children looked after.

Crucially, these developments cannot plausibly be explained by demographic change or by any evidenced surge in severe child maltreatment. Child population growth over this period was modest, and there is no convincing evidence of a corresponding explosion in abuse that would justify such a dramatic increase in removals. Instead, the pattern points toward institutional change and incentive effects within the child protection system itself.

An entire cottage industry of private assessment firms and contact providers has emerged in the wake of the growth of safeguarding interventions, frequently staffed or founded by former Local Authority social workers, creating a self-reinforcing ecosystem around care proceedings. Firms that have sprung up include: independent social work assessment companies, parenting and risk-assessment consultancies, supervised contact centres, therapeutic and “family support” providers, and last but not least, the expansion of interventionism spawns a specialist legal sub-sector of solicitors’ firms and barristers’ chambers focused on public law children work (legal work arising from actions taken by the state against parents in relation to children) including care proceedings, adoption and placement orders. This work is publicly funded via legal aid, embedding rising intervention rates into a stable, state-financed service economy, sustaining an arrangement that institutionalises high levels of intervention and normalises prolonged proceedings. All this is closely linked to the expansion and intensification of safeguarding activity, court proceedings, and assessment requirements since roughly 2010.

The system is not just administratively expansive but economically entrenched. Costs escalate even when intentions are benign. We can therefore expect that any reform effort will face resistance due to the embedded professional dependence.

Explaining the Figures

In England, official child-welfare statistics do allow us to identify three distinct but related populations of children affected by Children’s Services intervention, but they are reported separately, using different concepts and accounting logics. This separation makes it easy to misunderstand both the scale and the coercive intensity of the system unless the categories are carefully aligned. The broadest category is that of “children looked after” by a local authority. A child is counted as looked after when the local authority has formal responsibility for arranging the child’s care under the Children Act 1989. This is an administrative umbrella category rather than a single legal condition. It includes children removed by court order, but also children accommodated voluntarily with parental agreement, as well as children in more specialised legal situations such as secure accommodation or remand placements. The Department for Education publishes this figure annually as a snapshot on 31 March. In recent years, the number has been in the range of roughly 82,000 – 84,000 children in England. Importantly, being “looked after” does not in itself imply that a court has made a care order, nor that parental consent has been overridden.

Within this broad group sits a narrower and more coercive subset: children who are subject to a care order or interim care order. In these cases, the family court has decided that the child has suffered, or is likely to suffer, significant harm, and has granted the local authority parental responsibility alongside (or overriding) the parents. Parents cannot remove the child or determine where the child lives. These children are, of course, also counted as looked after, but the reverse is not true. Department for Education data show that, in recent years, roughly two-thirds of all looked-after children fall into this category. In absolute terms, this means that on the order of 55,000 children in England are currently under care orders or interim care orders. This figure represents the core population subject to compulsory state intervention. Families are severely stressed and often destroyed by these interventions. The costs are also enormous.

Spiralling Costs on Intervention

Over-intervention and false positives in terms of children in need are not only unnecessarily devastating and destroying many families, but these recent institutional dynamics are also placing huge strains of the finances of local governments.

Over and above the expensive foster parents, a huge raft of professionals including multiple local government officials, commissioned private assessment providers scrutinising the parents’ parenting capabilities in very lengthy processes, private centres providing supervised contacts with extensive reporting duties of each contact, psychologists etc. Each child triggers ongoing court proceedings which are meant to be limited to 6months but mostly extend much beyond this, often over more than 12 months. Each party, i.e. the LA, the mother, the father, and a Guardian representing the interests of the child, are supplied with lawyers and barristers. Excluding the court proceedings and lawyers costs, the annual cost of a child in foster care (including costs of dedicated LA social workers and assessment professionals) has been estimated to be between £70,000 and £140,000. Once adoption proceeds the costs to government drops but involves one off payments and a non-monetary support fund of anywhere between £5,000 and £30,000 per child depending on needs. For children above 10 years old it is often hard to find suitable adoption options. They are then subject to a long-term Care Order, which implies long term paid foster care organised by Independent Fostering Agencies (IFAs) who can charge management and profit margins of 35-45%. Costs per child can be anything between £90,000 and £180,000 and up to £250,000 in complex cases. A subset of these older children end up in so called residential care – a form of private for profit institutionalized care – which costs the taxpayer over £300,000 per annum per child as a UK average, and £450,000 to £500,000 in London.

There has been a notable overall increase in costs associated with Children’s Services safeguarding interventions, care orders and looked-after children placements in the UK, and this has been explicitly discussed in Parliament and official reports in recent years.

Public spending on children’s social care — the budget that funds safeguarding assessments, child protection, foster care and residential placements — has grown substantially in real terms over the past decade. Between 2009-10 and 2021-22, councils’ expenditure on children’s social care increased by more than 40 per cent in real terms, even though the child population grew much more slowly. Councils have often overspent their budgets on these services year after year, with nearly half overspending by 20 per cent or more in 2021-22.

More recent government statistics show that children’s social care net expenditure in England was budgeted at approximately £14.1 billion for 2024-25, which was about £1.4 billion, 11% higher in real terms than the previous year, with actual net costs reaching £15.6 billion, largely driven by rising costs for looked-after children and safeguarding services. To put this sum into perspective: this is the equivalent to 13% of the UK’s total national education budget, including higher education and universities (including research grants). We should also remember that this gigantic and rising sum is spent on “services” imposed and (for the most part) resented by parents, indeed stressing and devastating parents.

Part of the cost rise reflects very high unit costs for placements, particularly in residential care. Independent and privately run children’s homes can cost councils hundreds of thousands of pounds a year per child, with placements costing over £0.5 million in extreme cases. Analyses indicate that the average spend per child in a residential setting has risen sharply over a few years, placing significant strain on local authority finances.

These spending pressures have been debated explicitly in Parliament and by parliamentary committees. In Commons debates (e.g. discussions on children’s social care in 2025–26), MPs drew attention to the scale of spend and the pressures on local authority budgets, noting that the rising number of looked-after children and the associated investigations and placements are among the biggest cost pressures facing councils. Statistics on sharply rising costs, the number of children subject to protection plans or in care, and the challenge this poses for funding adequacy have featured prominently in those debates.

Select committees and inquiry reports have also examined these trends, including inquiries into children’s social care, funding models, market failures and the need to reform how services respond to families in need, partly with an eye to reducing the need for the most costly interventions later.

In summary, there has been a measurable increase in Children’s Services expenditure and increasing concern in Parliament — both about the level of overall spending and the underlying drivers such as placement costs and the growth in demand for statutory services, which have become a significant and growing part of local authority budgets.

Forced Adoptions

From within the care-order population (currently about 55,000 at any time) emerges a narrower group of children on the adoption track. Here it is crucial to distinguish several legally distinct stages. The most significant is the placement order, a court order that authorises the local authority to place a child for adoption and typically dispenses with the parents’ consent. Placement orders are reported by the Department for Education as a legal status of looked-after children. In recent data, around 8 – 9% of the 55,000 children under care order eventually become subject to a placement order, corresponding to roughly 4,500 – 5,000 children nationwide being under a placement order at any time. This is the closest available official proxy for children whose permanent separation from their birth parents has been judicially authorised.

Placement orders can be contested, but most (not all) placement orders lead to final adoption orders. Once an adoption order is made, the child immediately ceases to be looked after by the LA. Parental responsibility transfers permanently to the adopters, and the child disappears entirely from the parents’ life and from the Children’s Services responsibly and population statistics. For this reason, adoption is recorded only as an annual flow, not as a continuing stock.

In recent years, roughly three thousand children per year in England have ceased to be looked after because they were finally adopted. However, Local Authorities retain residual post-adoption support duties, such as assessing needs for adoption support, administering the Adoption Support Fund, and (sometimes) paying an adoption allowance. There is no official statistic showing how many adopted children exist in the population at any one time, nor how many adoptions occurred without parental consent on a cumulative basis.

England’s statistical framework is designed to measure current responsibility and throughput, not cumulative outcomes. As a result, while it is possible to say how many children are currently looked after, how many are currently under care orders, and how many are currently authorised or placed for adoption, it is not possible from official data alone to state how many children have been permanently separated from their birth parents through adoption over time. That absence is structural, not accidental, and it explains why cumulative figures do not appear anywhere in the published statistics.

Adoption is final. Parental responsibility transfers permanently. Birth parents lose all legal status. Adopters acquire full, lifelong parental responsibility. This is legally equivalent to birth parenthood.

Why has the number of severe interventions ramped up so much in recent years?

A central structural factor, some argue, might be the long erosion of “early help” and non-statutory family support, accelerated by post-2010 austerity. Preventive services—family support workers, youth services, children’s centres, mental health provision, and domestic support—are discretionary and therefore relatively easy to cut. Statutory duties, by contrast, remain enforceable and legally risky to neglect. As early intervention capacity withered, families increasingly reached crisis points at which the only remaining tools available to local authorities were formal safeguarding investigations and court proceedings. The system thus became slow to support but quick to escalate.

Overlaying this was a growing culture of risk aversion and defensive practice. High-profile child death scandals, Serious Case Reviews, and later Child Safeguarding Practice Reviews reinforced a retrospective blame culture in which professionals are judged harshly for failures to intervene but rarely sanctioned for intervening too much. This asymmetry matters. For teachers, health professionals, social workers, and managers alike, the personal and institutional risks associated with a false negative—failing to escalate a concern that later becomes a tragedy—vastly outweigh the risks associated with a false positive, where an intervention later proves unnecessary or harmful. In such an environment, escalation becomes a rational act of self-protection.

Although England does not formally impose universal mandatory reporting of suspected harm in the way some other jurisdictions do, it has developed a dense web of quasi-mandatory reporting obligations. Statutory guidance, professional regulation, inspection regimes, and employer safeguarding policies combine to create an environment in which reporting and escalation are strongly incentivised, while restraint is institutionally risky. The ambiguity of concepts such as “significant harm,” combined with the knowledge that decisions will be scrutinised retrospectively, pushes professionals toward referral and formalisation whenever doubt arises. This generates large volumes of referrals, many of them low-level or ambiguous, which the system is poorly equipped to filter without escalating.

Family justice reform after 2012 added further momentum. Judicial scepticism about long-term voluntary accommodation under section 20, combined with the introduction of the 26-week expectation for care proceedings under the Children and Families Act 2014, altered the tempo and logic of decision-making. Local authorities increasingly issued proceedings earlier in order to secure court oversight and meet timetabling expectations. Ambiguity, uncertainty, and parental capacity for change under support came to be resolved more quickly and more often in favour of state control. What had once been a system tolerant of drift and negotiation became one oriented toward early, decisive legal intervention.

At the same time, political and managerial signals emphasised speed, decisiveness, and permanency, particularly in the early 2010s. While explicit adoption targets were later softened, the broader message—that delay itself constituted harm—remained influential. Measurement, inspection, and reputational comparison reinforced this logic. What is measured and publicly judged tends to be what organisations optimise, and in this context timeliness and formal action were far more visible than quiet, sustained family support.

The resulting expansion of intervention was not evenly distributed. One of the most striking features of the English system by the mid-2010s was the scale of geographic variation. Rates of children looked after and care orders differed by factors of two to three between local authorities, even among areas with broadly similar levels of deprivation. London as a whole historically had lower rates than the national average, but within London there was wide divergence, and several inner east and north-east boroughs experienced faster rises after 2010 than the capital overall. Such variation is extremely difficult to reconcile with any natural distribution of harm. Maltreatment does not vary that wildly between neighbouring authorities; thresholds, cultures, inspection histories, and managerial practices do.

This variation is not merely a technical curiosity. From an institutional and normative perspective, it is powerful evidence that intervention levels are shaped by policy and incentives rather than by underlying need alone. Where liberty is at stake, persistent high variance under similar conditions points toward arbitrariness rather than necessity.

All of this makes it legitimate—indeed necessary—to ask whether many of these interventions may do more harm than good, over and above their substantial cost to taxpayers. Direct causal proof is difficult, because the counterfactual—what would have happened had a child remained at home with adequate support—cannot be observed. Nevertheless, indirect evidence is strong. Mandatory or quasi-mandatory reporting regimes internationally are known to generate huge increases in referrals with low substantiation rates, overwhelming agencies and diverting attention from the most serious cases. Families subject to investigation experience stress, stigma, and distrust, even when no order follows. Poor and marginalised families are disproportionately affected, as poverty and instability are easily misread as neglect. Once a case enters the system, de-escalation is institutionally difficult, and the harms of investigation and removal—disrupted attachments, identity loss, institutional trauma—are often underweighted in decision-making.

At the same time, outcomes for care-experienced children remain troubling across education, mental health, and adult life, despite rising intervention rates. While some children undoubtedly benefit from removal in situations of severe abuse, the aggregate picture does not support the claim that more coercion has straightforwardly produced better outcomes. Instead, it suggests diminishing returns and significant collateral damage at the margins.

Viewed through a libertarian lens, these developments amount to a textbook case of incentive-driven expansion of state power under uncertainty. The family is a core sphere of private ordering, and forcible removal of children is among the most intrusive acts the state can perform. A sharp rise in such interventions therefore demands an exceptionally strong justification grounded in necessity, proportionality, and demonstrable benefit relative to less intrusive alternatives. The post-2012 surge in care orders struggles to meet this standard.

From this perspective, the problem is not malign intent but structural bias. Professionals face asymmetric risks that reward escalation. Surveillance and reporting are enforced, while support is discretionary and underfunded. Time-compressed justice favours pre-emptive coercion over patient assistance. Institutions internalise reputational benefits from visible action while externalising the costs of error to families and children. Under deep uncertainty, the system exhibits epistemic overconfidence, opting for irreversible interventions where humility and restraint would be more appropriate.

None of this implies that child protection is illegitimate or that care orders are never necessary. Severe abuse exists, and the state has a role when parental authority collapses. But a roughly 40 per cent increase in care orders in five years, coupled with vast geographic variation and weak outcome evidence, shifts the burden of proof. It is no longer parents who must justify their resistance to intervention; it is the system that must justify its expansion.

From a libertarian standpoint, the lesson is not to abandon safeguarding but to rebalance it. Mandatory vigilance should be matched by mandatory support. Coercion should be a last resort, demonstrably necessary after less intrusive measures have been properly resourced and tried. Variation should trigger scrutiny, not complacency. And institutions should bear real responsibility for unjustified harm, not only for failures to intervene. Without such re-alignment, the post-2010 trajectory of England’s child protection system represents not a moral advance, but a predictable overreach of administrative power shaped by fear, incentives, and institutional self-protection rather than by clear evidence of benefit.

Weak Push-back

In the United Kingdom, concerns about the overreach of local authority Children’s Services, particularly in cases where children are removed from their biological parents, have given rise to a scattered but persistent field of parental activism, legal challenge, and institutional critique. While there is no single, unified national movement comparable to large-scale protest campaigns in other policy areas, a range of organisations, programmes, and informal networks address different aspects of what many parents experience as an opaque and coercive system of intervention.

Some of the most visible campaign groups, such as Fathers 4 Justice and related organisations, frame their activism around perceived injustices in the family justice system, especially the marginalisation of fathers following separation or divorce. These groups have historically relied on public protest and media attention to advance their cause. However, their focus is only tangentially related to the core issue of compulsory child removal by local authorities under care proceedings. As a result, they do not fully capture or represent the experiences of parents whose children are taken into care as a result of safeguarding investigations, allegations of neglect, or assessments of parental incapacity.

Other organisations operate in a more institutional and reform-oriented mode. The Family Rights Group, for example, supports parents and kinship carers who are involved with Children’s Services by providing legal information, advocacy, and policy input. Its work highlights recurring problems such as power imbalances between parents and professionals, lack of transparency, and the difficulty families face in navigating complex child protection procedures. Yet this form of engagement is deliberately non-confrontational and is framed as constructive engagement with the system rather than open protest against it. Similarly, organisations focused primarily on children’s rights intersect with the same structures of child protection but tend to prioritise the welfare of the child as defined by professional and legal standards, rather than explicitly challenging the legitimacy or scale of state intervention itself.

At the local level, including in boroughs such as Hackney, opposition to Children’s Services overreach is even less visible. What exists are programmes, support initiatives, and informal peer networks rather than organised protest movements. Hackney-based initiatives like Pause Hackney work directly with parents, particularly mothers who have experienced the removal of their children, and acknowledge serious deficits in communication, procedural fairness, and parental voice within the child protection system. However, these initiatives are designed as partnerships with local authorities and are oriented toward prevention, rehabilitation, and compliance, not toward public challenge or collective resistance. This pattern reflects a broader national reality in which parents affected by care proceedings are constrained by strict confidentiality rules, fear of further adverse consequences, social stigma, and limited financial or legal resources. These factors significantly inhibit collective organisation and public dissent.

The legal avenues available to parents who believe Children’s Services have acted wrongly are similarly constrained. In principle, local authorities can be sued for damages after care proceedings, but in practice such claims are exceptionally difficult to pursue successfully. English law affords wide protection to social workers and councils when they act within their statutory child protection duties, even when their decisions are later shown to be mistaken. Ordinary negligence claims almost always fail, following long-established case law that rejects the imposition of a duty of care in the exercise of child protection powers, largely on public policy grounds. Courts have consistently expressed concern that exposing local authorities to liability would encourage defensive social work and undermine the functioning of the child protection system.

The most viable legal route to compensation lies in claims under the Human Rights Act 1998, particularly for breaches of Article 8 of the European Convention on Human Rights, which protects the right to respect for private and family life. Where children are removed without lawful authority, held in care for excessive periods, or where parents are excluded from decision-making in a way that is procedurally unfair, courts have occasionally found violations of Article 8. Even in these cases, however, damages are usually modest and are often refused altogether if the local authority is found to have acted in good faith. Claims based on misfeasance in public office or malicious conduct are theoretically available but require proof that officials knowingly acted unlawfully or with improper motive, a threshold that is rarely met.

The overall picture is one in which parents who believe their children have been unjustly removed encounter a system that is highly insulated from both public protest and legal accountability. Organised activism exists but is fragmented, cautious, and often channelled into support or reform initiatives rather than direct confrontation. Legal remedies after the fact are narrowly circumscribed, slow, and typically symbolic rather than compensatory. This combination of confidentiality, institutional deference, and high legal thresholds helps to explain why grievances about Children’s Services overreach are widespread and deeply felt, yet rarely translate into sustained, visible public movements or meaningful avenues for redress.

Draconian Confidentiality Regime

The extreme confidentiality regime surrounding care proceedings and forced adoption in England and Wales has increasingly been criticised for producing effects that go well beyond the protection of children’s privacy. A growing body of judicial commentary, human-rights analysis, and academic critique argues that the near-total gagging of parents does not merely shield children from harm, but also creates structural conditions in which error, abuse, and institutional drift can persist unchecked. The concern is not that secrecy exists as such, but that it is imposed asymmetrically, permanently, and with particular severity on those most directly affected by state intervention.

Care and adoption proceedings are conducted in private, and publication of information relating to them is tightly restricted under the Administration of Justice Act 1960 and the Children Act 1989. These restrictions frequently continue indefinitely, even after proceedings have concluded and adoption orders have been made. Parents are prohibited not only from identifying their children, but from disclosing evidence, judicial reasoning, professional reports, or even their own case narratives where these might enable “jigsaw identification.” In practice, this means that parents who believe that grave errors were made in their case are prevented from speaking publicly about the substance of what occurred, often for life.

Critics argue that this produces a profound accountability deficit. Senior judicial figures have acknowledged that the family justice system has historically developed a culture of excessive secrecy, one that risks shielding poor practice from scrutiny and undermining public confidence in the rule of law. When proceedings that result in the permanent severance of family ties are almost entirely insulated from public examination, the normal corrective mechanisms of publicity, criticism, and reputational consequence are weakened or absent. The system becomes largely self-auditing, reliant on internal review rather than external challenge.

From a rule-of-law perspective, this raises serious concerns. Forced adoption is among the most intrusive powers exercised by the state against individuals who have committed no crime. It permanently alters legal identity, family relations, and personal history. Liberal constitutional theory has long held that the more coercive a power, the stronger the safeguards and transparency required to legitimate it. Yet in this domain, exceptional coercion is paired with exceptional secrecy. Parents are bound by lifelong speech restrictions, while local authorities, expert witnesses, and courts retain the ability to institutionalise their reasoning in judgments, training materials, and policy guidance, largely uncontested in the public sphere.

Human-rights litigation has repeatedly emphasised that adoption without parental consent constitutes an “exceptional interference” with family life under Article 8 of the European Convention on Human Rights. While courts have upheld the compatibility of confidentiality regimes with child protection aims, critics argue that, in combination with irreversible outcomes and limited avenues for public contestation, these regimes generate systemic risks. The inability of parents to speak openly about patterns of practice, recurring expert theories, or procedural shortcomings prevents the emergence of collective knowledge about institutional failure. Each case is treated as isolated, even where structural repetition may exist.

Academic commentators have described this condition as a form of epistemic closure. The system controls not only decisions, but also the permissible narratives about those decisions. Parents, who are uniquely positioned to observe the process from initiation to outcome, are legally prevented from contributing to public understanding of how the system operates in practice. This asymmetry creates moral hazard: public authorities bear little reputational cost for error, while parents face legal sanction for disclosure. Over time, such conditions may encourage risk-averse or removal-biased decision-making, knowing that external scrutiny is minimal and that dissenting accounts cannot easily surface.

International observers have expressed related concerns. UN-affiliated bodies and Special Rapporteurs have warned that child-protection systems characterised by high coercion, weak transparency, and limited adversarial challenge are vulnerable to normalising overreach. The danger identified is not primarily individual bad faith, but institutional degeneration: practices that would provoke resistance in open systems become routine when shielded from view. Secrecy, in this sense, is not merely a protective device but an enabling condition for unchecked power.

None of these critiques deny the importance of protecting children’s privacy or the need for stability in adoptive placements. Rather, they question whether absolute and lifelong gagging of parents is a proportionate or effective means of achieving those aims. It is possible to defend children’s anonymity while still permitting parents to speak about systemic features, procedural failings, or professional incentives. Indeed, in other domains of public law, anonymisation is routinely used to reconcile individual privacy with institutional accountability.

The central claim, therefore, is not that confidentiality is illegitimate per se, but that its current extremity risks undermining the very values it purports to serve. When parents are silenced indefinitely, when errors cannot be publicly named, and when patterns cannot be collectively recognised, the system loses an essential feedback mechanism. In such an environment, abuse, misjudgement, and (left-leaning) ideological drift may not only occur but replicate. A child-protection regime that cannot tolerate informed criticism from those it most powerfully affects risks sacrificing accountability in the name of protection, and in doing so, erodes trust in the justice of its own interventions.

To conclude:

The evidence assembled here points to a safeguarding regime that has drifted from targeted protection toward incentive-driven overreach: broad and elastic risk concepts, quasi-mandatory reporting, and a blame-averse institutional culture combine to favour escalation, while early help and de-escalation remain weak. The result is an expanding sphere of coercive state power, vast local variation that signals arbitrariness, and immense human and fiscal costs—often without clear proof of net benefit. Child protection is legitimate; but the burden of justification rises with the scale and irreversibility of intervention. The UK now needs a vigorous, informed public debate about thresholds, due process, transparency, and accountability.

End.

Patrik Schumacher, London 2026

Introduction

This paper argues that Land Value Taxation (LVT) can be coherently defended as a libertarian-superior alternative to the United Kingdom’s existing system of planning gain taxation, including Section 106 obligations and the Community Infrastructure Levy (CIL). Section 106 planning obligations are legally binding promises attached to land that a developer enters into as a condition of receiving planning permission. They’re used to make a specific proposal acceptable in planning terms where a planning condition alone isn’t enough. Typical obligations include on-site affordable housing, site-specific mitigation (e.g., junction improvements, play space delivered on the site), financial contributions, and delivery/management commitments with triggers (pre-commencement, pre-occupation, etc.). They can be agreed as bilateral agreements or unilateral undertakings. Section 106 obligations effectively constitute an off-budget, in kind tax (the worst tax imaginable) as it is negotiable/gameable and introduces uncertainty. Its reduction depends on a viability study. CIL is an additional tax, different in each Local Authority, and in London amounts to between 8 – 10% of final sales value in addition to the in-kind tax.

These instruments embed fiscal extraction within discretionary planning negotiations, generating opacity, rent-seeking, legal uncertainty, and systematic bias against development. By contrast, LVT targets the unimproved value of land—economic rent largely created by collective action, infrastructure provision, regulatory decisions, and agglomeration effects—while leaving productive investment, construction, and intensification untaxed. In this respect, LVT aligns more closely with libertarian commitments to rule-based governance, and minimal distortion of voluntary exchange.

However, the paper shows that the incentive structure created by municipal land value capture is not unambiguous. Land values may rise through genuine productivity-enhancing densification and infrastructure investment, but also through artificial scarcity produced by restrictive planning regimes. If rising land values are treated as an objective in themselves, even an LVT can perversely reward exclusionary zoning and supply suppression. The paper therefore stresses the crucial distinction between land becoming more valuable and space becoming more expensive, arguing that any defensible LVT must be embedded within a pro-development, pro-density regulatory framework oriented toward land productivity rather than scarcity rents.

Building on this critique, the paper advances a specific variant: a delayed, uplift-only Land Value Tax. Under this model, taxation applies only to realized or imputed land value increases, triggered by sale, long lease, or infrequent valuation after a substantial holding period. This temporal structure ensures that taxation follows realized outcomes rather than speculative expectations, preserves incentives for foresight and entrepreneurial risk-taking, and places a clear cap on the lifetime fiscal claim of the state. While this approach sacrifices some revenue stability and land-use discipline relative to a full LVT, it offers a more libertarian-coherent balance between efficiency, incentive alignment, and property rights. The paper concludes that a delayed uplift-only LVT constitutes a principled, rule-based alternative to planning gain taxation and a more liberty-consistent foundation for urban finance.

London’s Housing Prices and Land Values

Over the past half-century, housing prices in London have diverged dramatically from both construction costs and incomes, producing one of the most acute affordability crises among advanced global cities. While nominal house prices have risen across the United Kingdom, the magnitude and structure of price growth in London are exceptional. Empirical evidence from long-run price series indicates that average London house prices increased from roughly £12,000 in the mid-1970s to well above £500,000 today—a rise of around forty-five times in nominal terms. Even after adjusting for general inflation, London house prices are commonly estimated to be around six times higher in real terms than in the early 1970s. This long-run divergence demands explanation, and the central analytical finding is that it is overwhelmingly driven not by the rising cost of building homes, but by the escalating value of land.

To understand this transformation, it is necessary to disaggregate housing value into its two constituent components: the value of the structure and the value of the land on which it stands. Over long periods, the real cost of constructing housing—materials, labour, and standard specifications—has tended to track construction cost indices and exhibits relatively modest growth. Buildings also depreciate physically and are incrementally renewed, meaning that the structure component of housing value does not compound in the way asset prices do. By contrast, land values capture the cumulative effects of urban growth, accessibility, agglomeration, infrastructure investment, and regulatory constraint. It is therefore land, rather than structures, that absorbs the long-run uplift in housing prices.

This dynamic is particularly stark in London. Empirical studies and official evidence assembled by the Greater London Authority consistently show that the majority of long-run house price growth reflects rising land values rather than improvements in buildings. Economists such as Paul Cheshire have demonstrated that in cities with tight planning constraints, demand shocks are capitalised almost entirely into land prices. In this sense, London’s housing market increasingly resembles those of San Francisco or Hong Kong, where land and development rights dominate property values.

The growth of land values over the past fifty years has been extraordinary. A reasonable synthesis of the available evidence suggests that real residential land values in London have increased by approximately eight to twelve times since the mid-1970s. In Inner London, where accessibility and regulatory constraint are most intense, the increase is plausibly in the range of ten to fifteen times. Outer London, while still experiencing substantial appreciation, lies at the lower end of this range. These figures are not anomalies but structural outcomes of sustained demand growth interacting with a planning system that tightly limits land use intensification in precisely those locations where demand is strongest.

The dominance of the land component in house prices becomes even clearer when examining new-build housing prices through the lens of residual land valuation. Developers do not set land prices arbitrarily; rather, land prices are determined by expected sale values minus construction costs, professional fees, finance costs, risk allowances, and target profit margins. Because construction costs vary relatively little across London, while sale prices vary enormously, the residual—land value—absorbs almost all spatial variation. In contemporary London, this implies that land accounts for the majority of the price of a newly built home. In Inner London, around 70 – 80 % of the sale price of a new dwelling can be attributed to land value, including the value of planning permission. In prime central locations, this share can exceed eighty per cent. Even in Outer London, land commonly accounts for more than half of the sale price. The empirical implication is unambiguous: housing in London is expensive primarily because land with permission is expensive, not because homes are costly to build.

Crucially, what is being priced as “land” is not merely physical ground, but a bundle of location advantages and regulatory permissions. Central to this is accessibility. In urban economic terms, accessibility refers to the time- and cost-based reachability of employment, markets, services, and social interaction opportunities. In London, accessibility is structured by the public transport network coordinated by Transport for London, by dense labour markets, and by cumulative agglomeration effects. Locations with high accessibility command high land values because these advantages cannot be replicated at the level of the individual plot. Developers can improve buildings, but they cannot independently create Tube interchanges, dense job clusters, or cultural ecosystems. These socially produced advantages are therefore capitalised into land prices.

However, the current level of land prices in London is not simply a neutral reflection of accessibility and agglomeration. A large portion of land value represents scarcity rents generated by political restrictions on supply. England’s discretionary planning system severely limits densification in high-accessibility locations, ensuring that rising demand translates into higher prices rather than increased quantities. In this sense, current land prices are as much a product of political decisions as of market fundamentals. The consequence is that land value increases manifest as rising prices per dwelling unit, rather than as more dwellings in desirable locations.

This observation is critical for policy. The central objective should not be to suppress land values as such, but to reduce the land component embedded in each housing unit. These two goals are not contradictory. Land values per hectare can rise—even substantially—while land costs per dwelling fall, provided that land use intensifies. If a site that previously accommodated ten homes is permitted to accommodate forty, the land value of the plot may increase, but the land cost per unit can decline sharply. Under such conditions, prices of built units in attractive, accessible locations can stabilise or fall relative to incomes, even as aggregate land values continue to rise in response to increased productivity and accessibility. Increasing land use density and real estate supply would be the rational response to the increasing demand for urban space and the increasing desire to join an economic hub like London. This is the obvious, rational path to prosperity via co-location synergies facilitating productive economic cooperation.

This reframing resolves a common policy confusion. Rising land values are not inherently problematic; rising prices of housing units (or of office space) in high-opportunity locations are. The former can reflect genuine value creation through agglomeration and infrastructure, while the latter reflect artificial scarcity. The policy task is therefore to reverse the current causal chain by allowing systematic supply increases through land-use intensification, particularly in Inner London and other well-connected areas. By shifting adjustment to the demand pressure from prices to quantities, intensification addresses affordability at its root.

This strategy also has important fiscal implications. If land values rise because well-connected locations become more productive and better utilised through denser build out, then capturing a portion of that uplift through a land-based tax is a possible and efficient way to finance local government and public infrastructure investments. A land value tax—particularly one focused on land value appreciation—can recycle socially generated value into local public revenues without discouraging construction, provided that buildings and improvements are not taxed. In a liberalised planning regime, rising land values would no longer signal exclusion and scarcity, but successful intensification and urban productivity. Local government could then benefit fiscally from growth while households benefit from lower unit prices and improved access.

In sum, the empirical record of the past fifty years shows a dramatic shift in the composition of housing prices in London, with land values rising by up to an order of magnitude and now accounting for eighty per cent or more of prices in central areas. This outcome is the predictable result of sustained demand interacting with severe supply restrictions. Reversing this trajectory requires a clear policy realignment: prioritising affordable access to attractive locations through densification, allowing land values to rise for productive reasons, and capturing part of that appreciation for public benefit. Only by disentangling land value per unit of land from land cost per unit of housing can London resolve its affordability crisis without undermining its economic vitality.

Land Value Tax (LVT)

A Land Value Tax (LVT) is a tax levied on the unimproved value of land, independent of any buildings, infrastructure, or other investments placed upon it. The tax base is the site value itself—its location, accessibility, and surrounding venues including public amenities—rather than the productive effort of the landowner. In practical terms, a plot in a well-connected city centre would attract a higher tax than a remote rural site, regardless of whether the urban plot is developed intensively or left underused. The assessment focuses on what the land would be worth if it were vacant and available for its highest permitted use, given prevailing planning rules, i.e. the valuation of parcel tries to estimate the price of the land as if it were empty, but with its existing planning permissions and location characteristics. Land sales of close-by vacant plots or teardown sales where the building value is zero are used as evidence. In the absence of such comparables the property value can be used minus the cost of replacing the building (also considering depreciation).

An LVT would be charged annually to the legal owner of the land. Because land cannot be hidden, moved offshore, or reduced in supply in response to taxation, the tax base is exceptionally stable and predictable. Local authorities could rely on it as a continuous revenue stream rather than depending on volatile transaction-based taxes such as stamp duty or discretionary developer contributions. Valuation would be periodically updated, using market evidence from land transactions and residual valuation techniques, but the tax itself would not penalise development: building more, improving quality, or increasing density would not increase the tax bill so long as the land footprint remains the same.

Land Value Tax is often regarded as an efficient source of local public finance because it taxes economic rent rather than productive activity, while anchoring public revenue directly to locally generated value. Since land is fixed in supply, taxing its unimproved value does not distort decisions about work, investment, or development: land cannot be withdrawn, hidden, or relocated in response to taxation. As already emphasised by Henry George, a tax on land rent therefore avoids the deadweight losses associated with taxes on labour or capital and comes close to a non-distortionary revenue instrument.

At the same time, land values are largely the outcome of collective and public action at the local level—through infrastructure provision, planning decisions, and agglomeration effects. A land value tax captures part of this socially created value and returns it to the authority that helped generate it, creating a tight fiscal link between local decision-making and local revenue. In this way, LVT aligns allocative efficiency with institutional efficiency by financing local government from the very value base that local government itself helps to produce, (or at least not actively prevents).

From a libertarian perspective, the appeal of an LVT lies in its alignment with core principles of economic neutrality, fairness, and minimal distortion of voluntary exchange. Unlike taxes on income, labour, or capital investment, an LVT does not discourage productive activity. Work harder, invest more, or build better, and the tax burden does not rise. In this sense, it is one of the least distortionary forms of taxation available. Classical libertarian objections to taxation focus on coercion that punishes effort and enterprise; an LVT largely avoids this by targeting economic rents rather than productive returns. LVT would substitute for CIL and Section 106.

The rent embedded in land values is not created by the individual landowner alone. It arises from collective factors such as population density, agglomeration economies, infrastructure provision, legal protections, and planning permissions enforced by the state. From a libertarian perspective capturing part of this socially generated value to fund local public goods is more legitimate than taxing wages or profits that are clearly the result of individual effort and risk-taking. In this framing, an LVT functions less as an expropriation of earned income and more as a user charge for exclusive access to a scarce, legally protected resource.

For real estate development, an LVT improves incentives by discouraging land hoarding and speculative vacancy. When landowners must pay an annual charge regardless of whether their site is developed, they are pushed to either build, sell to someone who will build, or put the land to its most valued use. This accelerates development without the need for complex planning gain negotiations, ad hoc levies, or politicised bargaining over obligations. From a libertarian viewpoint, replacing discretionary and opaque instruments with a simple, rule-based tax enhances transparency, reduces administrative discretion, and limits opportunities for rent-seeking and corruption.

At the level of local government finance, an LVT offers a way to fund services while strengthening fiscal responsibility. Because the tax base is immobile and locally anchored, it aligns local spending decisions with local taxation, reinforcing accountability. Residents and landowners can clearly see the connection between the quality of local governance, public investment, and land values. This stands in contrast to centrally redistributed grants or complex development taxes that obscure responsibility and dilute democratic feedback.

Although many libertarians would still prefer minimal government and low overall taxation, an LVT can be defended as a “least bad” tax: difficult to evade, economically efficient, and focused on unearned rents rather than productive effort. Compared to property taxes on buildings, transaction taxes on sales, or taxes on development activity, it interferes less with market processes and urban intensification. In that sense, even within a sceptical libertarian framework, a Land Value Tax can be seen as a comparatively coherent, market-compatible foundation for local government finance.Top of Form

Is Land Value Appreciation a good or bad sign?

The fact that Land Value Tax is captured at the municipal level is often presented as a powerful incentive for local authorities to invest in infrastructure, public space, and environmental upgrading, since such investments tend to raise local land values and thus expand the local tax base. In principle, this creates a virtuous circle in which better transport, amenities, and environmental quality feed directly into higher public revenues, reinforcing growth-oriented and place-enhancing governance.

However, this incentive structure is not unambiguous. Land values can be increased not only through genuine value creation but also through artificial scarcity, most notably by restrictive planning regimes that limit supply. If higher land values were pursued for their own sake, a municipal LVT could perversely reward restrictive zoning or resistance to development, even though such strategies undermine affordability and urban productivity. In this sense, rising land values per se are not a meaningful public objective; what matters is the relationship between land values, usable space, and prices of final goods such as housing.

The crucial distinction, therefore, is between land becoming more valuable and usable space becoming more expensive. Land values can rise in a healthy way if regulatory frameworks permit densification and intensification of use, allowing more usable floor space to be produced per square metre of land. Under these conditions, the price of individual housing units can be stable or fall, even as the underlying land becomes more valuable. Higher land values then reflect increased productivity of location rather than scarcity rents extracted from constrained supply.

Thus the point of a municipal LVT regime is clear: local authorities should be incentivised not merely to maximise land values, but to maximise land productivity. This implies coupling LVT with permissive planning that allows greater densities where demand is strong. The result is a politically and economically coherent outcome: municipalities benefit from higher land values and a broader tax base, while residents benefit from increased supply of usable space, improved infrastructure, and greater affordability. In this configuration, LVT reinforces pro-development, pro-density urban governance rather than exclusionary scarcity-based value inflation.

The alternative route—raising land values through artificial scarcity rather than through increased land productivity—leads, by contrast, to stagnation and is unlikely to be sustainable in the long run. Restrictive planning or systematic resistance to development can indeed inflate land values by constraining supply, but this inflation is purely redistributive: it raises rents without expanding the stock of usable space, economic opportunity, or urban capacity. Such a strategy undermines affordability, suppresses agglomeration benefits, and ultimately chokes off the very growth dynamics that support local public finances.

Over time, scarcity-driven land value inflation erodes the economic base of the city. High housing and workspace costs deter in-migration, reduce labour mobility, and discourage firm formation and expansion. Infrastructure becomes underutilised or obsolete rather than leveraged, while political resistance to further upgrading intensifies as incumbent landowners seek to protect scarcity rents. Under these conditions, even a municipal Land Value Tax becomes a fragile revenue source: once demand weakens or mobile economic activity relocates elsewhere, land values plateau or decline, exposing the limits of a fiscal strategy built on exclusion rather than expansion.

By contrast, a densification-led strategy—permitting more development per unit of land—supports a self-reinforcing growth trajectory. Land values rise because locations become more productive, not because access is rationed. At the same time, the increased supply of floor space allows unit prices to stabilise or fall, sustaining affordability and economic dynamism. In this sense, artificial scarcity is a short-term fiscal illusion, whereas density, infrastructure provision, and environmental upgrading constitute the only sustainable pathway for municipalities seeking to finance themselves through Land Value Tax without sacrificing long-run urban vitality.

Therefore, the mere introduction of an incentive mechanism such as a municipal Land Value Tax cannot be treated as a technocratic fix that automatically delivers socially desirable outcomes. While LVT aligns local revenue with land values, it does not by itself distinguish between value increases generated by productive urban intensification and those generated by artificial scarcity. Without an explicit political and intellectual framework, the incentive to raise land values can be misinterpreted or misused, rewarding exclusionary planning strategies that are economically and socially counterproductive.

For this reason, LVT must be accompanied by explicit political discourse and critique that clearly articulates what causes of land value increases are legitimate and desirable. The normative target should not be high land values as such, but high land productivity—that is, the capacity of urban land to host more activity, more usable space, and more intensive social and economic interaction. This requires a conscious public narrative that links LVT to permissive planning, densification, infrastructure investment, and environmental upgrading, while criticising scarcity-based rent extraction as a dead-end strategy leading to stagnation.

Only within such a discursive framework can LVT function as a growth-oriented, pro-urban policy instrument rather than a tool that inadvertently entrenches incumbency and exclusion. In this sense, LVT is not merely a fiscal device but part of a broader political economy of urban development, one that must be openly debated, normatively defended, and institutionally embedded to ensure long-term sustainability and urban vitality. Incentives Structures must be backed by political discourse and critique.

Top of Form

Bottom of Form

Land Value-Uplift Tax (LVUT): variation of the typical LVT

This chapter proposes a deliberate variation on the standard Land Value Tax (LVT). Instead of taxing the level of land value per se, the tax, here called Land Value Uplift Tax (LVUT)would apply only to land value increases (uplifts) and would do so under a strictly defined temporal regime. Specifically, land value uplift would be taxed only upon realisation—through sale or long lease—or, in the absence of such a transaction, no earlier than five years after purchase, based on land-only valuations conducted at five-year intervals. Improvements on the parcel itself would remain untaxed. The tax revenue would accrue to local government.

This proposal is not a minor technical adjustment. It represents a principled shift in how land value taxation relates to uncertainty, foresight, risk, and the lifetime scope of public claims on private assets. To assess its merits, the proposal will be evaluated along four dimensions: conceptual clarity, incentive effects, fiscal logic, and libertarian coherence.

Conceptually, the proposal remains faithful to the moral core of land value taxation. It starts from the widely accepted premise that land value increases are predominantly socially generated—through infrastructure provision, regulatory change, agglomeration effects, and collective urban growth—rather than through individual effort by the landowner. By contrast, investments and improvements made on the parcel itself reflect private initiative and should not be discouraged at the margin.

By focusing explicitly on uplift rather than level, the tax cleanly separates socially generated windfalls from privately generated value. Periodic land-only valuation based on market transactions avoids the common conflation of land value with property value and preserves analytical clarity. Assigning the revenue to local government further strengthens the conceptual coherence of the scheme by aligning fiscal capacity with the spatial locus of value creation.

However, this shift from taxing a stock (land value) to taxing a flow (land value appreciation) is not merely technical. It alters the nature of the tax from a continuous rent tax to something closer to a land-specific capital gains levy. That distinction has far-reaching consequences. The key incentive question is not simply how much land value is taxed, but when taxation occurs relative to uncertainty and realisation.

Under a standard annual LVT assessed at current market value, anticipated future gains are capitalised immediately into land prices. As a result, owners face higher tax liabilities as soon as expectations shift—even if infrastructure is not yet built, planning permission has not yet been granted, or demand ultimately fails to materialise. In this sense, foresight is “taxed forward”: taxation responds to expectations rather than outcomes. This can chill early land assembly and long-horizon investment, particularly in contexts of high uncertainty.

An uplift-only tax can avoid this problem only if it is explicitly designed to do so. If land is frequently revalued and uplift is taxed as soon as expectations are capitalized, the incentive advantage disappears. The present proposal therefore insists on a delay and realisation principle. Uplift is taxed only when it is realized through transaction, or—absent transaction—only after a substantial holding period and only at coarse valuation intervals. This design ensures that anticipated value is not taxed before it exists as a realised outcome. Under this regime, the landowner bears full downside risk: if anticipated growth does not materialize, there is no uplift and no tax. When uplift does occur, it is shared ex post. The “reward” to foresight is not an untaxed windfall in advance, but the fact that taxation is aligned with success rather than expectation. This temporal alignment matters. It preserves incentives for entrepreneurial discovery, land assembly, and early commitment under uncertainty.

Importantly, this advantage is not automatic. If uplift tax rates are set at confiscatory levels, entrepreneurial reward is annihilated just as surely as under a high-rate standard LVT. Rates remain decisive. What the uplift-only scheme uniquely contributes is a different timing and risk structure: taxation follows realized success rather than speculative anticipation. The most important—and often overlooked—advantage of an uplift-only LVT emerges over the long run and does not collapse into a mere rate issue.

Under a standard LVT, taxation applies to the level of land value every year, indefinitely. Once land has reached a high-value plateau, the owner continues to pay a high tax forever. Over time, the cumulative fiscal take can far exceed the original uplift. From the owner’s perspective, the reward for having anticipated growth is steadily eroded year after year, long after uncertainty has resolved and no further gains occur. The state’s claim is effectively unbounded in time.

By contrast, an uplift-only LVT taxes change, not condition. When land value rises, part of that rise is skimmed. When value stabilises, taxation ceases. The total fiscal claim is therefore bounded. The owner retains the long-term benefit of having moved early into a now-valuable location. This is a structural difference, not a parametric one.

This distinction mirrors the logic of entrepreneurship more generally. In most markets, once uncertainty resolves and success is proven, returns stabilise and are not subject to continuous expropriation. The uplift-only model treats land speculation as a form of entrepreneurial discovery rather than as a permanent rent stream owed to the state. It taxes the windfall of success, not the ongoing possession of a successful asset.

From a libertarian perspective, this distinction is decisive. Libertarians tend to object less to ex post sharing of socially generated windfalls than to ex ante taxation based on speculative assessments, and they are especially hostile to taxes that resemble permanent wealth levies on legally held assets.

Once uplift has occurred and been partially taxed, continued taxation looks less like rent capture and more like an ongoing claim on established property. The uplift-only model explicitly rejects that logic. It insists to tax the windfall, not the settled condition. In doing so, it caps the state’s lifetime claim on a successful act of anticipation and early investment—an act that may itself have helped catalyse the later uplift.

This also clarifies why orthodox Georgists typically oppose uplift-only schemes. Their objective is to socialise land rent continuously, on the grounds that land rent is socially created at every moment. The present proposal rejects that premise. It treats social contribution as episodic—linked to growth phases, infrastructure shocks, and regulatory changes—rather than permanent. This is a philosophical divergence, not a technical defect. The trade-off seems clear: Standard LVT enhances allocative efficiency and revenue stability but minimises private residual claim and imposes an unbounded lifetime fiscal burden.

Delayed uplift-only LVT sacrifices some revenue predictability and hoarding pressure but preserves long-run reward to foresight and aligns taxation with realised outcomes. The argument advanced here is that foresight deserves not only a temporary reward during appreciation but a durable reward after success has been proven. The proposed delayed, realisation-sensitive uplift-only LVT delivers exactly that. It does more than shift timing; it places a principled cap on state capture and prevents a justified windfall levy from mutating into a de facto permanent wealth tax on land.

In libertarian terms, this is not a compromise but a coherent alternative conception of land value taxation—one that preserves efficiency while respecting risk, discovery, and long-term property rights.

Top of Form

Bottom of Form

Arguments against the LVUT, in favour of usual LVT:

At the same time, the uplift-only model seems to weaken one of the strongest efficiency virtues of a standard LVT: the pressure to put land to use regardless of market cycle. Under a conventional LVT, holding valuable land idle is costly every year. Under the LVUT scheme, land banking becomes inexpensive in periods of stagnation or slow appreciation, because “no uplift” implies “no tax.” This reintroduces a degree of speculative withholding that the classic LVT is designed to eliminate. In short, the variant is better for rewarding discovery, but worse for enforcing continuous utilisation.

Fiscal stability is another point of contrast. A conventional LVT offers local government a stable, predictable, counter-cyclical revenue base, precisely because it is levied on stock rather than flow. The uplift-based scheme, by contrast, produces volatile revenues that track market cycles. In downturns or flat markets, revenues collapse just when public budgets are under pressure. This undermines one of the strongest practical arguments for LVT as a foundation of local government finance.

Relatedly, the requirement to calculate annual uplift introduces additional valuation complexity. It requires a baseline, treatment of negative uplifts, smoothing rules, and decisions about whether losses are carried forward or ignored. A simple land-value-at-rate system is administratively cleaner and more transparent. Libertarians typically value rule simplicity and low discretion; uplift taxation risks re-introducing interpretive and political margins.

From a libertarian standpoint, however, the variant LVUT proposal has genuine strengths. It can be defended as targeting only realized or imputed windfalls, not the mere possession of a scarce asset. The slogan “no tax without value uplift” resonates strongly with libertarian intuitions about non-aggression and earned reward. It also weakens the claim that landowners are coerced into paying for value they have not actually enjoyed. In this sense, the LVUT model is normatively more palatable than a full rent capture regime.

That said, many libertarian economists who accept LVT at all do so precisely because land rent is considered a permanent, ongoing flow, not a one-off gain. From that perspective, taxing only increases while leaving the existing rent stream untaxed looks arbitrary: the landowner continues to benefit from location advantages every year, even if prices are flat. The orthodox LVT taxes that ongoing advantage directly; your variant taxes only change in expectations.

A balanced assessment would therefore be this: the uplift-only LVT is an intellectually defensible hybrid between a pure LVT and a capital gains tax. It is superior to transaction taxes, development levies, and property taxes on buildings, and it is more libertarian-friendly than a full land rent tax. However, it sacrifices some of the incentive and fiscal virtues of the classical model—especially land-use discipline and revenue stability.

This argument leads us to a refinement as follows: A possible reconciliation, to refine the proposal further, one could propose to institute a low, universal baseline LVT on land value (to discourage hoarding and fund core services) combined with a moderate uplift tax that shares windfall gains while preserving entrepreneurial reward. This preserves the moral clarity of “earned vs unearned” while retaining the institutional strengths that made LVT attractive in the first place. The arguments about land banking, however, need to be further put into perspective.

Land Banking as a Pathology of the UK Urban Planning Regime

THESIS: Land banking is not a market failure but a political artefact of the planning system, i.e. representing one of many government failures in the urban development process.

The notion of “land banking” has become a central trope in British housing debates, particularly in London. It is routinely invoked to explain high house prices and sluggish supply and is commonly attributed to speculative behaviour by developers who are said to hoard land in order to restrict output and engineer scarcity. This diagnosis, however, rests on a series of questionable assumptions about incentives, competition, and risk in land and housing markets. Once these assumptions are scrutinised, it becomes apparent that the phenomenon labelled “land banking” is largely mischaracterized. Rather than representing a market failure or an expression of monopoly power, it is better understood as a rational response to the structural uncertainties produced by the UK’s discretionary planning system.

In a properly functioning land market, systematic land banking would be difficult to sustain and, in most cases, economically irrational. Holding undeveloped land generates no income stream and ties up capital that could otherwise be deployed productively. Developers face real financing costs, balance-sheet constraints, and investor pressure to generate returns within finite time horizons. In such a context, the incentive is not to withhold supply indefinitely but to bring land into development as soon as conditions allow profitable and predictable execution. Any strategy of deliberate hoarding would expose the landowner to significant opportunity costs, as well as to the risk that competitors will develop substitute sites, thereby capture demand and erode any hoped-for scarcity premium.

Moreover, housing markets—especially in large metropolitan regions—are not monopolistic in the relevant sense. Even where individual sites are unique, demand is spatially and temporally substitutable. If one landowner delays development in anticipation of higher prices, others can and typically will build instead, particularly where supply is elastic. The idea that developers can collectively or individually restrict supply over long periods without being undercut by competitors presupposes a level of coordination and market power that is rarely plausible in open markets. From a standard market-theoretical perspective, land banking as a deliberate scarcity strategy should therefore be self-defeating.

Why, then, does land banking appear to be such a persistent feature of the British housing landscape? The answer lies not in market incentives per se, but in the institutional environment within which land development takes place. The UK planning system is highly discretionary, politically mediated, and legally contestable. Development rights are not secured by clear, rule-based entitlements but must be negotiated case by case through lengthy approval processes. Outcomes are uncertain in timing, scale, design requirements, infrastructure obligations, and financial contributions. Even once permission is granted, it may be subject to judicial review, renegotiation, or political reversal following changes in local authority control.

Under these conditions, holding land is less a matter of speculative hoarding than of option-holding under uncertainty. Land with prospective or partial planning permission embodies a real option whose value derives precisely from the ability to defer irreversible investment until regulatory and political risks have been resolved. The land itself becomes a hedge against an unpredictable planning environment. What appears externally as “banked” land is often land awaiting the stabilisation of conditions that would allow development to proceed without excessive downside risk.

This interpretation is reinforced by the fact that much allegedly banked land in London is not, in practice, shovel-ready. Sites may lack essential infrastructure, be subject to unresolved planning obligations, or be contingent on broader master-planning frameworks that involve multiple stakeholders and public authorities. Large developments are also necessarily phased to match local absorption rates and financing constraints. Such phasing is frequently conflated with land banking, despite being an entirely orthodox response to market demand and risk management rather than an attempt to withhold supply.

The political appeal of the land-banking narrative is nevertheless clear. It moralises scarcity by attributing housing shortages to the strategic behaviour of profit-seeking developers rather than to systemic constraints embedded in the planning regime itself. In doing so, it deflects attention away from zoning restrictions, height limits, heritage controls, and discretionary veto points that severely limit supply responsiveness. It also provides rhetorical justification for punitive policy responses, such as compulsory purchase, punitive taxation, or forced build-out requirements, which may further increase risk and reduce investment rather than alleviate shortages.

This is not to deny that some speculative land holding exists. Particularly around major infrastructure announcements or in contexts where planning gain expectations are inflated, landowners may rationally delay development in anticipation of regulatory windfalls. However, such behaviour is best understood as a second-order effect induced by policy signals, not as the primary driver of housing scarcity. Crucially, these incentives are themselves created by the discretionary nature of the planning system, which transforms development rights into politically allocated rents rather than secure, market-based entitlements.

From a libertarian or market-oriented perspective, the implication is clear. Land banking should not be treated as evidence of market failure requiring corrective intervention against landowners. Instead, it should be recognised as a symptom of an institutional framework that suppresses competition, amplifies uncertainty, and rewards delay. In a rules-based planning system with by-right development, predictable density allowances, and limited scope for ex post renegotiation, the economic rationale for land banking would largely evaporate. Developers would compete on speed, efficiency, and quality rather than on their capacity to navigate political risk.

Seen in this light, land banking in London is not a pathology of capitalism but a pathology of governance. It reflects not too much market freedom, but too little. The appropriate response is therefore not to discipline developers for behaving rationally within a dysfunctional system, but to reform the system itself so that holding land idle becomes an unattractive strategy. Only under such conditions can land markets operate as genuine markets rather than as arenas for regulatory option-holding and political arbitrage.

Top of Form

LVUT vs LVT: Prioritizing Investment rather than Consumption

Here is another argument that relativizes the perceived advantages of a traditional LVT and leads us to shift emphasis to the idea of a uplift-only LVT, underpinning the idea of combining the two approaches.

The usual claim that a standard LVT provides a stable revenue stream is analytically correct, but normatively ambiguous once you consider behavioural and institutional incentives. Stability is not an unqualified virtue. The supposed “advantage” of LVT as guarantor of a stable revenue stream for local authorities – even in absence of growth - is potentially a problem: it makes it too easy for the local authorities to shift revenue into consumption rather than investment to upgrade the urban area.

A stable, non-cyclical revenue base allows local authorities to finance ongoing expenditure even in the absence of growth. While this insulates them from downturns, it also weakens the pressure to invest productively in order to expand the tax base. When revenue flows independently of improvement in local conditions, the fiscal link between value creation and fiscal reward is loosened. Local governments may rationally choose to fund recurrent consumption—staffing, programmes, political priorities—rather than undertake difficult, risky, and politically contentious investments that would raise land values over time.

In this respect, a standard LVT resembles a resource annuity. Once land values are high, the authority can live off an inherited fiscal base without materially improving the area. This parallels the “resource curse” logic familiar from natural resource economics: stable rents reduce incentives for innovation, reform, and long-term investment. The danger is not fiscal irresponsibility in the narrow sense, but complacency—a bias toward maintaining the status quo rather than upgrading the urban fabric.

By contrast, a tax regime that is more tightly coupled to value uplift sharpens incentives. If additional revenue depends on actual appreciation—triggered by infrastructure investment, planning reform, densification, or improved public realm—local authorities face a clearer intertemporal trade-off. Consumption today forecloses investment tomorrow. Revenue becomes performance-sensitive, rewarding authorities that successfully catalyse growth and penalizing those that merely administer stagnation.

This argument cuts directly against one of the canonical Georgist claims. Georgists typically assume that since land rent is socially generated, its continuous capture naturally aligns public incentives. But this assumes that public agents are automatically growth-oriented and investment-maximizing. In reality, local governments face electoral cycles, bureaucratic inertia, and soft budget constraints. Under such conditions, stable rents can entrench fiscal moral hazard: the authority consumes the rent while under-investing in the conditions that would expand it. This might even slip into capital consumption (due to lack of maintenance) and decline, left for future politicians to deal with.

From a libertarian or market-institutionalist perspective, this is not a trivial concern. The goal is not merely to fund government cheaply, but to discipline government behaviour. A revenue system that is too forgiving in the absence of growth undermines that discipline. In this sense, some revenue volatility is not a defect but a feature: it forces prioritization and incentivizes privileging investment over consumption.

An uplift-based or hybrid system reintroduces this discipline by making fiscal capacity contingent on success in improving the local environment. It effectively converts local government from a rent-collector into a co-investor and facilitator of private investment whose returns depend on performance. This aligns well with the earlier point that taxation should target episodes of value creation rather than permanent states. Once uplift ceases, so does the fiscal windfall, pushing authorities back toward investment if they wish to maintain or expand revenues.

The broader implication is that the “best” tax system cannot be evaluated solely on efficiency and stability grounds. It must also be judged by how it shapes institutional behaviour over time. A pure LVT scores highly on allocative efficiency but may underperform on governance incentives. An uplift-sensitive regime sacrifices some stability but gains investment discipline and growth orientation.

Stable revenue reduces fiscal stress but also reduces creative pressure. Whether this is desirable depends on whether one views local government primarily as a caretaker of existing value or as an entrepreneurial agent responsible for upgrading and intensifying the city.

Historical Note

Debates over land value taxation have always revolved around a deeper question: should the state claim an ongoing share of land rent as a permanent social entitlement, or should it intervene only at moments when land values rise due to collective progress? The distinction between a full Land Value Tax (LVT) and a tax limited to land value uplift (LVUT) is not a late technical refinement but one of the oldest fault lines in the history of the idea.

The intellectual prehistory begins in classical political economy. John Stuart Mill famously distinguished between returns earned through labour and capital investment and the “unearned increment” accruing to landowners as society develops. Crucially, Mill did not conceptualise land rent as a permanent annuity owed to the state. Instead, he argued that increments in land value generated by population growth, infrastructure, and economic development were morally available for taxation. This framing already contains the germ of an uplift-based approach: taxation is justified when value increases for reasons external to the owner, not simply because land is owned.

The debate took a sharper and more radical turn with Henry George, whose Progress and Poverty transformed land value taxation into a comprehensive social philosophy. George rejected increment-only approaches as insufficient and insisted that the entire annual rental value of land should be taxed away. For George, land rent was not episodic but continuous, generated at every moment by the presence and productivity of society as a whole. As a result, he tended toward national-scale solutions and saw LVT as a replacement for most other taxes. This position hardened into Georgist orthodoxy, within which uplift-only taxes were often criticised as timid compromises that left “socially created rent” in private hands once growth stabilised.

Yet even during George’s lifetime, more limited and pragmatic approaches were being developed in parallel. In Britain and continental Europe in the late nineteenth and early twentieth centuries, policymakers experimented with land value increment taxes rather than full LVTs. These schemes explicitly targeted changes in land value, usually measured at sale or over a defined period, and were justified as mechanisms for capturing windfalls generated by urban growth and public investment. The most prominent British example was introduced under David Lloyd George in the 1909–10 People’s Budget, which imposed an Increment Value Duty on increases in land value above a baseline assessment. This was consciously framed as taxing socially generated gains without imposing a permanent annual burden on landownership as such.

In Germany and other European countries, similar debates unfolded at the municipal level. Cities introduced Wertzuwachssteuern—taxes on land value increases—motivated by the observation that urban infrastructure, zoning decisions, and agglomeration effects drove land appreciation. These taxes were typically local, episodic, and tied to identifiable growth events, reinforcing the idea that value uplift was the legitimate fiscal target rather than land value per se. Georgists criticised these measures for failing to socialise rent fully, but their very persistence demonstrates that uplift-based thinking was not a marginal deviation but a robust alternative tradition.

After the Second World War, the same tension reappeared in planning and development debates, particularly in the UK. Betterment levies and development charges sought to skim land value increases arising from planning permission and public investment. Although these instruments were often administratively flawed and politically unstable, their conceptual core remained consistent with the earlier increment tax logic: the state claims a share of newly created value, not an indefinite claim on an established asset. Once again, the contrast with orthodox LVT was clear. Full LVT aimed at continuous rent capture; betterment levies aimed at episodic uplift capture.

In late twentieth-century economic thought, this divide was increasingly articulated in terms of incentives and institutional design rather than moral philosophy. Economists sympathetic to market processes but wary of distortionary taxation treated land value uplift taxes as “second-best” or hybrid solutions—less pure than Georgist rent capture but more politically and institutionally viable. The question of scale also shifted: whereas Georgism tended toward national pooling of land rent, uplift-based approaches were increasingly tied to local government, where planning decisions and infrastructure investment directly affected land values.

Seen in this longer perspective, the idea of taxing land value uplift is neither a dilution of LVT nor a modern technocratic invention. It represents a parallel lineage within land taxation theory—one that treats land value increases as windfalls to be shared, while preserving long-term private rewards for foresight, risk-taking, and early commitment. Where Georgism emphasized equality and fiscal replacement, uplift-based approaches emphasized incentive alignment, temporal fairness, and limits to state claims.

From the outset, debates over Land Value Taxation have been inseparable from questions of spatial scale: at what territorial level should land rent be assessed, collected, and spent? The issue is not merely administrative. It reflects deeper disagreements about the nature of land rent, the purpose of taxation, and the appropriate relationship between local communities, cities, and the state. In classical Georgist theory, the moral justification for land value taxation pointed implicitly toward the national scale. Henry George conceived land rent as a surplus generated by society as a whole. In this view, the value of land did not arise primarily from local decisions or municipal investments but from the aggregate progress of civilisation—population growth, economic development, and social cooperation at large. It followed that land rent should accrue to the community in the broadest sense, typically interpreted as the nation. Although George fully recognised that land values vary by location and that assessment must therefore be locally grounded, his fiscal vision was national in ambition, particularly because LVT was intended to replace most other forms of taxation.

When Georgist ideas migrated to Europe, however, they encountered a strong tradition of municipal finance and local public works, which reshaped the debate. In Britain and Germany at the turn of the twentieth century, reformers were less interested in national tax replacement than in capturing the land value effects of urbanisation. Land value increment taxes and betterment charges were therefore typically framed as municipal instruments, justified by the fact that city-level infrastructure—streets, transit systems, utilities—and local planning decisions were the proximate causes of land appreciation. In this context, taxing land at the city or borough level appeared both more legitimate and more practicable than national schemes.

This divergence exposed a latent tension within Georgism itself. Orthodox Georgists worried that strong localisation of land value taxation would entrench spatial inequality, allowing prosperous cities to capture and recycle their own land rents while poorer regions fell further behind. National pooling of land rent promised redistribution and uniform social dividends. Yet this very feature weakened the connection between local decision-making and fiscal outcomes. As a result, a divide opened between egalitarian Georgism, which favoured centralisation, and more market-oriented or institutionalist approaches, which favoured local accountability.

The postwar British planning system illustrates the consequences of resolving this tension in favour of centralisation. Under the Town and Country Planning Act, development value was treated as socially created, and its capture was routed through central government. Local authorities bore the political and administrative burden of planning control, while the fiscal benefits were largely detached from local budgets. This severed incentive structure contributed to chronic underinvestment, resistance to development, and ultimately the abandonment of the betterment system. Later mechanisms such as Section 106 agreements and the Community Infrastructure Levy partially re-localised value capture but did so in discretionary and opaque ways.

By the late twentieth century, economists increasingly approached the question of scale through the lens of fiscal federalism rather than moral philosophy. Scholars such as William Vickrey and Wallace Oates argued that land is an ideal local tax base precisely because it is immobile and because the benefits of local public goods are capitalised into local land values. From this perspective, a local land tax internalises costs and benefits, aligning expenditure decisions with revenue generation. The appropriate scale is therefore not national but municipal or metropolitan, matching the jurisdiction that controls planning, infrastructure, and service provision.

Contemporary urban economics has reinforced this position. Economists such as Paul Cheshire and Josh Ryan-Collins emphasise that land value uplift is driven primarily by local regulatory decisions, especially planning permission and density constraints. If local authorities restrict supply but do not capture the resulting uplift, they have little fiscal incentive to permit growth. Conversely, if uplift accrues locally, authorities are encouraged to invest, densify, and accommodate expansion. This logic points toward taxation at the level of cities, boroughs, or city-regions, where the causal mechanisms of land value creation are most clearly located. Seen historically, then, the “grain” of land value taxation has tracked ideological priorities. National-scale taxation aligns with ambitions to socialise land rent and finance universal redistribution. Local-scale taxation aligns with goals of incentive compatibility, fiscal discipline, and growth-oriented governance.

The proposal put forward in this paper, with its emphasis on uplift-based taxation, reinforces this conclusion. Because land value uplift is generated by specific planning decisions that permit more new development per land parcel, infrastructure investments, and neighborhood-level transformations, its taxation must be correspondingly local if it is to preserve incentive alignment. A national uplift tax would dissolve the very feedback mechanism that gives the policy its rationale. In this sense, the question of spatial scale is not secondary but constitutive: how land is taxed cannot be separated from where it is taxed and spent.

ConclusionBottom of Form

This paper argues that Land Value Taxation can be coherently framed not as a departure from market principles, but as a libertarian-superior alternative to the UK’s existing system of planning gain taxation. Instruments such as Section 106 obligations and the Community Infrastructure Levy are typically defended as pragmatic mechanisms of value capture. From a libertarian perspective, however, they are deeply flawed. They rely on discretionary negotiation rather than general rules, embed fiscal extraction within the granting of development permission, and convert the planning system into a site of political bargaining rather than legal certainty. The result is opacity, rent-seeking, and a systematic bias against development.

A properly designed Land Value Tax offers a fundamentally different logic. Unlike the current planning gain levies, LVT does not tax the act of development, the securing of permission, or the willingness to negotiate with the state. Instead, it targets the economic rent of land—value that largely reflects collective action, public investment, regulatory decisions, and agglomeration effects rather than individual effort. By replacing ad hoc regulatory exactions with a generalised, predictable fiscal rule, LVT aligns more closely with core libertarian commitments to the rule of law, neutrality between economic activities, and the separation of taxation from administrative discretion.

However, the incentive structure created by municipal land value capture is not unambiguous. Land values can rise through genuine value creation—improved infrastructure, increased accessibility, and intensified economic interaction—but also through artificial scarcity, most notably when restrictive planning regimes constrain supply. If rising land values were pursued as an end in themselves, even an LVT could perversely reward exclusionary zoning, down-zoning, or resistance to development. In this sense, higher land values per se are not a normative objective. What matters is the relationship between land values, the quantity of usable space, and the prices of final goods such as housing.

The crucial distinction is therefore between land becoming more valuable and space becoming more expensive. Land values can rise in a healthy way if regulatory frameworks permit densification and intensification of use, allowing more floor space to be produced per unit of land. Under these conditions, the price of individual housing units can stabilize or rise more slowly even as underlying land values increase. Higher land values then reflect increased locational productivity rather than scarcity rents extracted from constrained supply. Any defensible LVT regime must therefore be embedded within an explicit political and regulatory commitment to development, density, and land productivity, rather than artificial scarcity.

The paper also emphasizes that not all LVTs are equally compatible with libertarian principles. A high-rate, perpetual tax on the level of land value risks functioning as a permanent wealth tax on an immobile asset, blurring the line between justified rent capture and ongoing expropriation. Such a design weakens long-term incentives for foresight, early risk-taking, and entrepreneurial discovery—precisely the dynamics that underpin urban transformation.

For this reason, the paper has advanced a specific variant: a delayed, uplift-only Land Value Tax. Under this model, taxation applies only to land value increases, not to land value as such; it is levied upon realisation through sale or long lease, or—absent such events—only after a substantial holding period and at coarse valuation intervals. This temporal structure ensures that taxation follows realized outcomes rather than speculative expectations, shares socially generated windfalls ex post rather than penalizing anticipatory risk-taking ex ante, and places a clear cap on the lifetime fiscal claim of the state.

In this form, LVT does more than shift the timing of value capture; it transforms its essential character. The state participates in episodes of socially generated uplift without asserting an indefinite claim on settled holdings. The tax targets windfalls rather than possession, success rather than anticipation, and growth phases rather than value plateaus. This stands in sharp contrast both to orthodox Georgist ambitions to socialize land rent continuously and to the UK’s reliance on discretionary planning gain, which embeds rent extraction within the very process of granting development rights.

Seen in this light, the choice is not between market freedom and public value capture, but between discretionary regulatory extraction and rule-based fiscal sharing, and between scarcity-driven rent inflation and productivity-driven urban growth. A delayed uplift-only LVT replaces negotiated expropriation at the planning stage with a transparent, generalized, and contestable tax regime, while remaining compatible with a pro-development, pro-density planning framework. It is therefore defensible to describe such a scheme as a libertarian alternative to planning gain taxation: one that disciplines the state’s claim on land value through predictability, proportionality, and respect for entrepreneurial risk, while offering a more coherent and liberty-consistent foundation for urban finance.

The notion of “land banking” has become a central trope in British housing debates, particularly in London. It is routinely invoked to explain high house prices and sluggish supply and is commonly attributed to speculative behavior by developers who are said to hoard land in order to restrict output and engineer scarcity. This diagnosis, however, rests on a series of questionable assumptions about incentives, competition, and risk in land and housing markets. Once these assumptions are scrutinised, it becomes apparent that the phenomenon labelled “land banking” is largely mischaracterized. Rather than representing a market failure or an expression of monopoly power, it is better understood as a rational response to the structural uncertainties produced by the UK’s discretionary planning system.

In a properly functioning land market, systematic land banking would be difficult to sustain and, in most cases, economically irrational. Holding undeveloped land generates no income stream and ties up capital that could otherwise be deployed productively. Developers face real financing costs, balance-sheet constraints, and investor pressure to generate returns within finite time horizons. In such a context, the incentive is not to withhold supply indefinitely but to bring land into development as soon as conditions allow profitable and predictable execution. Any strategy of deliberate hoarding would expose the landowner to significant opportunity costs, as well as to the risk that competitors will develop substitute sites, thereby capture demand and eroding any hoped-for scarcity premium.

Moreover, housing markets—especially in large metropolitan regions—are not monopolistic in the relevant sense. Even where individual sites are unique, demand is spatially and temporally substitutable. If one landowner delays development in anticipation of higher prices, others can and typically will build instead, particularly where supply is elastic. The idea that developers can collectively or individually restrict supply over long periods without being undercut by competitors presupposes a level of coordination and market power that is rarely plausible in open markets. From a standard market-theoretical perspective, land banking as a deliberate scarcity strategy should therefore be self-defeating.

Why, then, does land banking appear to be such a persistent feature of the British housing landscape? The answer lies not in market incentives per se, but in the institutional environment within which land development takes place. The UK planning system is highly discretionary, politically mediated, and legally contestable. Development rights are not secured by clear, rule-based entitlements but must be negotiated case by case through lengthy approval processes. Outcomes are uncertain in timing, scale, design requirements, infrastructure obligations, and financial contributions. Even once permission is granted, it may be subject to judicial review, renegotiation, or political reversal following changes in local authority control.

Under these conditions, holding land is less a matter of speculative hoarding than of option-holding under uncertainty. Land with prospective or partial planning permission embodies a real option whose value derives precisely from the ability to defer irreversible investment until regulatory and political risks have been resolved. The land itself becomes a hedge against an unpredictable planning environment. What appears externally as “banked” land is often land awaiting the stabilisation of conditions that would allow development to proceed without excessive downside risk.

This interpretation is reinforced by the fact that much allegedly banked land in London is not, in practice, shovel-ready. Sites may lack essential infrastructure, be subject to unresolved planning obligations, or be contingent on broader masterplanning frameworks that involve multiple stakeholders and public authorities. Large developments are also necessarily phased to match local absorption rates and financing constraints. Such phasing is frequently conflated with land banking, despite being an entirely orthodox response to market demand and risk management rather than an attempt to withhold supply.

The political appeal of the land-banking narrative is nevertheless clear. It moralises scarcity by attributing housing shortages to the strategic behaviour of profit-seeking developers rather than to systemic constraints embedded in the planning regime itself. In doing so, it deflects attention away from zoning restrictions, height limits, heritage controls, and discretionary veto points that severely limit supply responsiveness. It also provides rhetorical justification for punitive policy responses, such as compulsory purchase, punitive taxation, or forced build-out requirements, which may further increase risk and reduce investment rather than alleviate shortages.

This is not to deny that some speculative land holding exists. Particularly around major infrastructure announcements or in contexts where planning gain expectations are inflated, landowners may rationally delay development in anticipation of regulatory windfalls. However, such behaviour is best understood as a second-order effect induced by policy signals, not as the primary driver of housing scarcity. Crucially, these incentives are themselves created by the discretionary nature of the planning system, which transforms development rights into politically allocated rents rather than secure, market-based entitlements.

From a libertarian or market-oriented perspective, the implication is clear. Land banking should not be treated as evidence of market failure requiring corrective intervention against landowners. Instead, it should be recognised as a symptom of an institutional framework that suppresses competition, amplifies uncertainty, and rewards delay. In a rules-based planning system with by-right development, predictable density allowances, and limited scope for ex post renegotiation, the economic rationale for land banking would largely evaporate. Developers would compete on speed, efficiency, and quality rather than on their capacity to navigate political risk.

Seen in this light, land banking in London is not a pathology of capitalism but a pathology of governance. It reflects not too much market freedom, but too little. The appropriate response is therefore not to discipline developers for behaving rationally within a dysfunctional system, but to reform the system itself so that holding land idle becomes an unattractive strategy. Only under such conditions can land markets operate as genuine markets rather than as arenas for regulatory option-holding and political arbitrage.

This paper argues from a radical libertarian, indeed anarcho-capitalist, perspective but challenges the dominant libertarian conception of society as reducible to market exchanges by arguing for the indispensable role of critical discourse in the reproduction and advancement of modern civilization. The argument involves a critique of libertarianism’s ahistorical conception of society, drawing on Marx’s materialist conception of history emphasizing technological advances as key drivers of societal transformations. This leads to the recognition of a new, unique advantage for libertarianism in the era of post-fordism and thus an historically unprecedented potential for libertarian revolution. The argument also involves a critique of libertarianism’s impoverished concept of human action. Drawing on insights from the work of Jürgen Habermas (b.1929) — the paper outlines a vision of 'Discourse Capitalism' in which markets and discourses are complementary societal mechanisms of coordination that together can substitute for all state functions. The paper thus articulates a new libertarian conception of social order that augments the anarcho-capitalist conception by emphasizing the increasing importance of guiding discourses in the era of rapid technological innovation.

Introduction

It is generally understood that a public sphere with a lively culture of debate is an important, indeed indispensable ingredient of democracy. This paper argues that such a public sphere can altogether take over and deliver the functions expected to be delivered by the democratic state, namely the continuous adaptation of the systems of rules that frame and regulate the life of society, including economic life. In such a libertarian society this adaptation process would not be mediated by state power delivering collectively binding decisions as majoritarian dictates on the basis of elections but rules would adapt via the soft power of emergent majority trends and convergences on the basis of individual choices guided by ongoing public debate. It’s a question of semantics whether such a society (which the author presumes could only evolve from liberal democracies) should still be referred to as democracy. This would not be wide off the mark: such a system would be democratic in accordance with many of this term’s connotations. Such a societal system would be a form democracy, a highly informal, liquid and intensive form of “democracy”.

The heavy lifting of making a stateless market society plausible has been delivered by Murray Rothbard and his followers (under the heading of anarcho-capitalism) as well as by figures like David Friedman. Their reflections, speculations and compelling arguments are presupposed here. This paper focusses on augmenting this work with an emphasis on public discourse. The expectation here is that if and when the libertarian movement succeeds in rolling back the state, a lively discursive public sphere, including an ongoing energetic political debate would continue and indeed intensify.

This paper - while building on anarcho-capitalism - involves a critique of Rothbardian anarcho-capitalism - a critique that also challenges the philosophical underpinnings of the wider libertarian and classical liberal community of thinkers. This critique then leads on to the proposition of a discourse-led, free market-based society: discourse capitalism. This paper argues that markets, while essential, are insufficient to account for the dynamics of modern society. Instead, it proposes that ‘discourse capitalism’—a society where markets are guided by discourses —provides a more adequate framework for understanding modernity’s progress and for re-imagining society after the libertarian revolution. On the methodological basis of Marx’s historical materialism and with the critical resources from Habermas’ Theory of Communicative Action, the argument charts a path beyond the binary of state versus market to posit an anarcho-capitalism augmented by the increasingly important dimension of public discourse, including political discourse.

Here is a clarification of Habermas’ concept of discourse employed in the markets-and- discourses approach: Discourse here means critical-reflective communication or dialogue. Discourse comes into play when the smooth routine of doing things with words - with its underlying, taken for granted meanings, premises, rights, shared beliefs etc. – fails, fractures or breaks down. Discourse is reflective and argumentative and involves questioning of the implicit validity claims involved in all speech, like the claim of truth, truthfulness, and rightness. Discourse demands reasons and operates via arguments.

The phrase ‘libertarian revolution’ denotes the rolling back of the state via privatization, i.e. the substitution - by competitive market provisions - of all state functions, including: education, health, social security, welfare, money, roads/rails, utilities, land/rivers/lakes, planning, public space, police, courts, law/legislation, military defense. The feasibility of privatizing all current state functions has been considered elsewhere by the author, in his recent article ‘Pathways to a stateless World Society’ (Schumacher, 2025). There, readers also find arguments that these functions could be delivered privately at a higher quality and lower costs, as well as descriptions of the various pathways towards this result: libertarian electoral success, special economic zones, private autonomous cities, and secessions.

Will the libertarian revolution spell the end of politics? Anarcho-capitalist political debate and activism is focused on the goal of eliminating the state, on wresting resources and decision powers from political institutions to return them to private institutions and individuals. It seems as if the endgame of anarcho-capitalist activism is to achieve this withdrawal of all political power with the prospect that thereafter everybody can finally go back to their respective private and individual pursuits of happiness via markets. This seems to imply the abandonment of the res publica, the abolishment of the political arena. This implicitly self-effacing bent of libertarian political discourse, and its prospect of a society of pure market exchanges, is unrealistic and unsustainable.

It is much more realistic to expect that political discourse will flourish more than ever when issues are no longer reduced to two or three policy bundles to be voted on, and instead many possible trends and developments can simultaneously be experimented with, compared, analysed and discussed. After the libertarian revolution there will be neither government, nor any central legislation via governments or parliaments. However, the discourse of e.g. juris prudence will continue after the libertarian revolution. Once the development of the legal system, has been wrested from the inefficient hands of state courts and parliaments, and transferred to the initiative of entrepreneurs competing in markets of law provision, legal discourse will be, more than ever, in demand. For instance, the advantages or disadvantages of recognizing intellectual property rights will continue to be debated after the libertarian revolution. Such public discourses, both in their scholarly and in their more popularized layers, make sense and will be conducted aiming for and probably achieving legal convergence, even in the absence of once-for-all and one-fits-all decisions backed up by force.

The historical materialist argument here is that the prosperity potential of a libertarian revolution that would roll back the state altogether is, compared with 20^th Century conditions, much enhanced under current technological premises. The economic stagnation afflicting advanced societies stems from pervasive state intervention, which increasingly paralyses technological and entrepreneurial dynamism. The author posits that the shift from mechanically based Fordism to computationally empowered post-Fordism necessitates a libertarian revolution to fully unlock latent productivity and innovation potentials. The character of our current knowledge economy and post-fordist network society also implies that discourses are playing a much enhanced and increasingly indispensable role in furthering prosperity.

Prosperity and Liberty: The Two Dimensions of Freedom

“The destiny of the spiritual World, and … the final cause of the World at large, we allege to be the consciousness of its own freedom on the part of Spirit, and ipso facto, the reality of that freedom.”

Georg Wilhelm Friedrich Hegel, Lectures on the Philosophy of History, 1837

Freedom emerges in the evolution of living systems. Most sentient, mobile creatures strive for freedom and instinctively resist shackles and controls because they need freedom of action to provide for themselves. The beginning of freedom of action is movement of a self-preserving living creature towards sources of energy and to move out of harm’s way. The more moves the creature can make, i.e. the more degrees of freedom the creature acquires, the greater are the creature’s chances to persist and reproduce (Dennett, 2011). However, these moves must not be random, but information-based. Freedom must be coupled with foresight, and freedom to act crucially includes freedom to plan and prepare and indeed to labor, to make the environment more hospitable and predictable. This implies the self-binding of actions as part of a planned, goal-oriented concatenation of actions, and requires self-discipline. The freedom to act becomes the freedom to pursue projects.

Human life and survival is dependent on social cooperation. Compared to our closest relatives in the animal kingdom - the great apes – human life is unique in terms of density and scale of cooperation. Humans are ultrasocial. Our most momentous projects are always cooperative projects.

Productive cooperation requires a division of labour, social organisation and rules of cooperation, i.e. a social order including hierarchies, demanding discipline and social control or rule enforcement. This necessary social control is hard to distinguish from no longer necessary forms of control or from unnecessary forms of social control with predatory and exploitative-parasitic admixtures. Only unnecessary social controls should/can be subjected to sustainable liberation or social emancipation. Only a comparative analysis of different concurrent societies operating with historically given material technologies and the competition between these societies can distil what’s necessary and expose the admixtures of predation. Can we expect that competition will weed out all predatory excesses? Competition has often worked in this direction.

Freedom (emancipation) encompasses both a material dimension and a social dimension. Material freedom refers to increasing liberation from the material necessities and impositions of an indifferent or hostile nature/universe, while social freedom refers to liberation from the social discipline imposed to facilitate our cooperative conquering of material freedom. A more familiar term for material freedom is prosperity. However, the term material freedom adds value by heightening the existential import of prosperity. A more familiar term for social freedom is liberty.

The thesis put forward here is that material freedom is primary as it is the final success criterion of all forms of social organization and forms of society. As history has shown, materially successful societal systems tend to win over materially inferior systems by means of population proliferation, warfare/subjugation, imitation/emulation, and migration to prosperity. Another important consideration for privileging prosperity over liberty is that prosperity, i.e. economic growth based on technology-fueled productivity gains, is cumulative, exponential, and without upper limit.

The concept of freedom must be historicized. To be sure, both prosperity and liberty are desirable in their own right. They are interdependent dimensions and factors of freedom. The interdependence between prosperity and liberty is neither straightforward, nor historically invariant. A certain level of prosperity might be required to loosen social control and increase liberty. On the other hand, liberalization might lead to prosperity gains. The libertarian presumption that more liberty is always materially feasible and always unleashes more prosperity is a fallacy. Sustainable social freedoms have material prerequisites. There can be too much liberty, in the form of unwarranted relaxation or weakening of social control and rule enforcement, signaling a dissolution of materially necessary social order. In this sense an increase in social freedom might undermine prosperity, thereby also compromising more long-term liberty potentials. A liberty-increasing revolution only makes sense where the historical material prerequisites for such an emancipation are in place. This has sometimes been the case. This implies the need to think historically and to keep re-examining what’s new in terms of transformative technological advances. Sometimes (often enough but not always) increases in social freedom can unleash prosperity potentials. This is definitely the case currently: The unleashing of our current technology-based prosperity potentials requires a libertarian revolution.

The Unique Prosperity Take-off of Modernity

“Constant revolutionising of production distinguishes the bourgeois epoch from all earlier ones. All fixed, fast-frozen relations, with their train of ancient and venerable prejudices and opinions, are swept away. All that is solid melts into air. … The bourgeoisie has given a cosmopolitan character to production and consumption. … The intellectual creations of individual nations become common property.”

Karl Marx & Friedrich Engels, The Communist Manifesto, 1848

Modern civilization is a creation of capitalism and its preservation and continuing progress depends on the preservation of capitalism. Modern civilization is equally a creation of science facilitating transformative technologies. Modern civilization is also a creation of critical discourses more generally: the ‘Enlightenment’, and the creation of the public sphere via salons, books, journals, newspapers etc. This includes the creation of extensive bodies of literature in historiography, political theory, jurisprudence, economics, moral philosophy, sociology etc. These discourses support the development of capitalist economic institutions, capitalist morality, as well as legal and political institutions that are compatible with capitalism’s progress. In generating unprecedented prosperity take-off of modern Western civilization three related innovations mutually catalyzed and complemented each other: capitalism, i.e. economic freedom and entrepreneurship, science-based technological innovations, and ideological emancipation via Enlightenment philosophy.

The entirety of world economic history contains only one truly significant event: the industrial revolution facilitating the escape from the ‘Malthusian trap’. The quantitative economic historian Gregory Clark (2007) emphasized that there was no increase in the average standard of living whatsoever for thousands of years, all the way up to 1800, despite all the progress in technology and culture achieved by the great historical civilizations. The escape from the Malthusian trap was possible only after the pace of productivity gains via innovation and capital accumulation was outpacing the natural population growth that had hitherto always absorbed all prior technological advances, thereby keeping living standards at subsistence levels.

Libertarians, especially those - but not only those - who emphasise liberty (on the basis of natural rights) over prosperity (on the basis of utilitarian philosophy), suffer from a lack of historical consciousness. There is the ahistorical tendency to promote maximal liberty equally at all times. This ahistorical stance has to be overcome. In this respect libertarians have a lot to learn from Marxist with their mantra: Always historicise! The viability and thus prospect of any envisaged societal order depends on the historical level of technological and cultural development of the society in question. It is important to distinguish socio-economic epochs like feudalism and capitalism, as well as a series of distinct stages of capitalism as a basis for understanding significant political transitions.

Marx’s distinction between the forces of production and the relations of production is pertinent here. The historical process involves both, the development of the forces of production and the complementary development of the relations of production. Marx rightly understood that technological progress is more continuous than social and political transformations which are marked by inertia. This means that with the historical development of the forces of production there gradually builds up an incompatibility with the relations of production. Formerly necessary forms of social control become obsolete. The relations of production then increasingly stand in the way of the full exploitation of the evolved forces of production and become fetters on the further development of the forces of production. This is the moment when a social revolution becomes potentially productive.

Marx was right in identifying technology as the driving force in history, i.e. history’s active aspect. This is insight remains eminently pertinent today. We are currently living through rapid technological advancements. Not all forms of political or societal order are equally capable of facilitating the further development and prosperity potentials of the currently most promising technologies. Those political forms that paralyse progress will not be sustainable. Those which unleash pent up potentials will flourish and find imitators.

The convergence of computation and tele-communication delivers a technology induced socio-economic restructuring with momentous work & life-style implications, i.e. a new economy and society: the knowledge economy powering post-fordist network society. Fordism was based on repetitive mechanical mass production while Post-fordism is based on computationally empowered flexible specialisation and computationally empowered forms of communication. The technologies of Fordism were rigid without agile adaptivity and without the ability to quickly absorb and utilize innovations. In contrast, the new numerically controlled systems like robotic fabrication, 3D printing as well as the massive new field of software-as-a-service (SaaS) are capable of absorbing unlimited numbers of innovations. This in turn dynamizes business and business organisation and pulls workers away from assembly lines into R&D, marketing, finance etc. All work becomes innovative and project-like rather than routine.

These transformations were not only leaving communism behind as obsolete shell to be broken and thrown off but also brought the semi-socialist economies of Europe to the brink, leading to a rather sharp political transformation in Britain and slower but also serious adaptations elsewhere in Europe. However, these political adaptations were insufficient. As stagnation has once more set in, especially since 2008, the project of liberalisation is becoming increasingly urgent again. The upshot of this is that this technological revolution must be followed by a libertarian political revolution to unleash its potential. The hypothesis promoted in this paper – as well as in the earlier paper on Pathways to a Stateless World Society (Schumacher, 2025) – is that rolling back the state all the way to zero is now not only possible in the most economically and culturally advanced arena of world society but that this wholesale substitution of all state functions via markets and discourses will lead to the world’s (and world history’s) most productive and prosperous economies.

This appraisal of the state as barrier to further development is historically specific and does not at all imply that the state has always been such a barrier. This historical perspective distinguishes the theory of discourse capitalism proposed here from the libertarian (anarcho-capitalist) political philosophy of Murray Rothbard[1] (1926–1995) and his followers who see the state and its ruling class - in the original sense of Charles Comte (1782–1837) and Charles Dunoyer (1786–1862) - as inherently illegitimate, as nothing but a predatory gang of robbers, as always and everywhere a negative force. In contrast to this ahistorical stance, it is not only possible - but indeed very important for the credibility of radically libertarian political philosophy - to see and admit that states throughout most of history, including the modern states in West since the 17^th century with their more comprehensive state powers and capacities, have, at times, made a significant contribution to world civilisation and economic development. Such an admission does not contradict or undermine the vigorously anti-state stance we argue for at our current historical juncture. This point about the constructive, on balance net positive, historical role of states does not imply that we do not also entertain more nuanced analyses about the kinds of states and states systems that were most conducive to the development of humanity’s intellectual and productive forces. Such an analysis points already, starting with the unique early flourishing of the ancient Greek civilisation, to systems of smaller states energized via competition rather than large, integrated, autocratic empires. The Norther Italian city states that delivered early capitalism, as well as the experiences of the Dutch Republic, England and the large number of independent, competing German principalities in the 18^th and 19^th Century are other examples where systems of competing small states delivered momentous boosts of cultural, scientific and economic flourishing.

The Emergence of a Critical Public Sphere

“Our age is the age of criticism, to which everything must be subjected. The sacredness of religion, and the authority of legislation, are by many regarded as grounds of exemption from the examination of this tribunal. But, if they are exempted, they become the subjects of just suspicion, and cannot lay claim to sincere respect, which reason accords only to that which has stood the test of a free and public examination.”(Kant, 1998, p.100)

Kant, Immanuel. The Critique of Pure Reason, 1781

Markets are institutionalized in forums of consumption like shops, malls, online shopping sites, as well as in forums of exchange like industry trade fairs, stock exchanges, commodity exchanges etc. Discourses are institutionalized via forums for face-to-face communications in discussion circles, universities, conferences etc., as well as via technical media of communication, including printing, broadcasting, podcasting, and social media etc. Capitalist markets and critical discourses evolved together as a uniquely successful historical trajectory. This co-evolution was a combustive relation of mutual amplification and acceleration.

The historically unprecedented sphere of public discourse - although not itself a market - nevertheless was only possible within the era of capitalism. As Juergen Habermas emphasized in his 1962 book-length critical-historical study ‘The Structural Transformation of the Public Sphere’, the original bourgeois public sphere that emerged in Western Europe in the 18^th Century was a public of private persons whose autonomy was based on private property guaranteeing economic independence within an economic sphere largely liberated from traditional political hierarchies and privileges. This independence was an important prerequisite of an unpressured, open, frank debate among participants that recognized each other as equal with respect to the right to put forward arguments, in a process of deliberation where only arguments count. The middle of the 19th century was the short highpoint of laissez faire capitalism. This was simultaneously the highpoint of the critical political capacity and effective political function of the public sphere as unrestricted, deliberative realm where the force of the better argument wins out determining truth and rightness in debates about the common wheal feeding into legislation.

In its fullest expression the public sphere was also market mediated. It was not per se the invention of printing technology that was decisive but rather the massive proliferation of books, pamphlets, newspapers and journals via capitalist publishing enterprises. This is what created a reading public at an unprecedented scale. The availability of books, newspapers and journals inspired reading societies. Habermas describes the scene at the end of the 18^th Century in Germany:

“They were mostly associations with rooms that provide the opportunity both for reading newspapers and journals and, just as importantly, for discussing what had been read. The oldest reading circles had involved nothing more than collective subscriptions that helped to lower the cost of the papers. In contrast, the reading societies no longer arose from such financial motives. These societies … exclusively served the need of bourgeois private people to create a forum for a critically debating public: to read periodicals and to discuss them, to exchange personal opinions, and to contribute to the formulation of an opinion that from the nineties on will be called ‘public’. Journals with political content had the largest number of subscribers and were most widely read.” (Habermas, 1989, p.72-73).

Newspapers evolved from ‘news letters’, i.e. correspondences, commercially organized by newsdealers, for capitalist market participants. Habermas explains: “The organization of this traffic on a continuous basis became imperative to the degree to which the exchange of commodities and of securities became continuous. Almost simultaneously with the origin of stock markets, postal services and the press institutionalized regular contacts and regular communication.” (Habermas, 1989, p.16) The newsletters initially procured for closed networks of business subscribers was then also made available to a larger public, evolving into newspapers proper. It did not take long until governments started to use these papers regularly, towards the end of the 18^th Century weekly, for public announcements. Some papers became official gazettes. However, the attempt of governments to take over and control all newspapers and journals failed and a politically effective, independent press became an important institutional component of liberal, constitutional societies. The French Revolution, for instance, enshrined the freedom of the press in the constitution of 1791 (quoted after Habermas): “The free communication of ideas and opinions is one of the most precious rights of man. Everyone can therefore speak, write, and print freely, with the proviso of responsibility for the misuse of this liberty in the cases determined by law.” (Habermas, 1989, p.70). The reality of actual freedom of the press had to wait longer in France: The leaders of the revolution did not honor this commitment. A year later an edict of the Paris Commune denounced the opposition as poisoners of public opinion and confiscated their press, and two days after Napoleon had taken power in his coup d’etat in January 1800 he prohibited all but a select list of 13 named papers. In 1811 this list was reduced to three papers (besides the official Moniteur), subject to strict censorship. While the returning Bourbons declared that “the French have the right to have their views published and printed”(Habermas, 1989, p.71), it was only after the July Revolution of 1830 that freedom of the press as key political ingredient of liberal societies became a reality in France.

The development in Great Britain was more continuous and more advanced earlier. Compared with Continental Europe, the British press evolved earlier, at the end of the 17^th Century, and enjoyed unique liberties. Censorship came to an end with the Licensing Act of 1695. At the same time a lively scene of first literary and then political debating circles arose in coffee houses, critically monitoring and debating parliamentary proceedings and government action. By the first third of the 18^th century London alone featured about 2000 coffee houses. Defoe’s Review, Tutchin’s Observator, and Swift’s Examiner, were discussed in clubs and coffee houses, at home and in the streets. Habermas describes the creation of a vehemently critical, political press in Britain: “It was not the Whigs who created political journalism in the grand style; this was the work of the Tories who now constituted themselves as the opposition under Bolingbroke. … Bolingbroke brought out the first issue of the Craftsman, the publicist platform of the opposition. … With this journal, followed by the Gentleman’s Magazine, the press was for the first time established as a genuinely critical organ of a public engaged in critical political debate. … Thus, raised to the status of an institution, the ongoing commentary on and criticism of the Crown’s actions and Parliament’s decisions transformed a public authority now being called before the forum of the public.” (Habermas, 1989, p.60)

The Whigs soon followed the Tories’ lead. The public sphere, feeding into the continuous controversy between government and opposition, became an indispensable factor of modern, liberal society on a national scale. Habermas: “Until then political opposition at the national level had been possible only as the attempt to push one’s interests by resorting to violence in the forms of the Fronde and the Civil War; now, through the critical debate of the public, it took the form of a permanent controversy between the governing party and the opposition. This discussion in principle went beyond the issues of the day to include the “topics of government”; the separation of powers, British liberties, patriotism and corruption, party and faction, the question of the legality of the opposition’s new relationship to the government—and even basic questions of political anthropology.” (Habermas, 1989, p.64)

Habermas quotes a speech by Fox in the House of Commons in 1792 to show how politics had explicitly acknowledged the force of public opinion:

“It is certainly right and prudent to consult the public opinion. … If the public opinion did not happen to square with mine; if, after pointing out to them the danger, they did not see it in the same light with me, or if they conceived that another remedy was preferable to mine, I should consider it as my due to my king, due to my Country, due to my honour to retire, that they might persue the plan which they thought better, by a fit instrument, that is by a man who thought with them … but one thing is most clear, that I ought to give the public the means of forming an opinion.” Habermas, 1889, pp. 65-66)

From Marxism via Anarcho-capitalism to Discourse Capitalism

For a former Marxist, the hardest aspect of the ideological shift to anarcho-capitalism was the renunciation of Marx’s uplifting vision of mankind reaching full collective self-consciousness and self-determination, escaping the alienation whereby the results of “anarchic” market processes confront each participant as an alien force. The anarchic market was to be abolished, and democratic economic planning was to take over. Mankind would finally be constituted to consciously shape its own future and make history a project of conscious collective self-development.

The presumption of Marx and the socialists was and is that socialist revolution would allow overall productivity levels to soar. Tragically the opposite was the case (and more than ever remains the case): the socialist revolution and the system of state planning slowed down the further development of productivity and prosperity. The compelling explanation for this tragic failure was delivered by Ludwig von Mises (1881-1973) (Mises, 1920) and Friedrich von Hayek (1899-1992) (Hayek, 1945). In 1920 von Mises published a devastating critique of the feasibility of a socialist planned economy, rightly arguing that the concept of a socialist economy is an oxymoron. In a socialist economy all means of production (resources) are state-managed, i.e. there is no bidding and price formation process for resources. Without prices there can be no economic calculation and no projects (resource allocation options) can be compared in terms of costs. Without cost information there can be no rational allocation decisions, no economizing resources, i.e. this system does not deserve to be classified as “economy”. The authorities are groping in the dark, making arbitrary decisions.[2] In 1945 von Hayek pointed to the insurmountable knowledge problem faced by a central planning agency. There is no way all the dispersed local and often only tacit knowledge and experience of all producers can be centralized, and there is no substitute for the price mechanism as information processing and parsimonious signaling mechanism. This price formation is the result of competitive bidding processes and resource prices continuously adapt with respect to changing relative end-product values and changing resource scarcities. State planning cannot offer a functionally equivalent substitute.

However, socialism – while always inferior to capitalism – had a certain viability during the era of concentrated, large scale mechanical mass production. Socialism was, to some extent, a viable pathway of modernisation in this era. In the most advanced industrial economies the most important industries had consolidated into a small number of highly concentrated, both horizontally and vertically integrated concerns, already shifting the balance between hierarchy and market towards hierarchical coordination within very large private concerns. To nationalize and centrally plan such an economy was feasible, in any event much more feasible than in earlier times and certainly much more feasible than it would be now. The massive economies of scale – together with the punitive dis-economies of scope - that result from mechanical mass production implied a strong tendency towards a dramatic equalisation of living standards, delivering a universal consumption bundle to all. “Communism”, in the sense of an unprecedented material egalitarianism was, as it were, nearly a physical necessity.

Here is another noteworthy fact for libertarians to acknowledge and reflect upon: The 20th Century social-democratic mixed economies were growing faster than the 19th Century classical liberal laissez faire economies: 2.5% versus 1.5% on average across the century. To be sure, this does not demonstrate that a state-regulated economy is (always) superior to a laissez faire economy. Such statements are ahistorical and thus fallacious. (Also, even in the post-war era, the economically more liberal economy of Germany was outperforming the more socialised British economy). However, these figures mean that the social-democratic mixed economy was a viable, indeed a historically very successful development model. However, this is no longer the case since the 3^rd quarter of the 20^th Century. Thatcher’s and Reagan’s neoliberal revolution did put the languishing British economy back onto a viable growth path.

Marx’s vision and project, that had inspired millions around the world, was crushed, both in practice and in theory, receiving its Coup de Grâce in 1990. Hayek’s vision of the epistemic power of freely competitive, indeed anarchic but adaptive and self-optimising rationality of market processes is equally inspiring. Marx wanted to replace markets with conscious collective decision making. This falters as it increasingly meets an insurmountable complexity barrier. Most libertarians based on Mises, Hayek, Rothbard etc. want to shift everything to markets and nothing but markets.

In contrast, the thesis here bets on markets plus discourses, including political discourses, but in contrast to the usual conception of politics, these political discourses are not presumed to lead to elections, government and collectively binding decisions. The libertarian political project is the project of unleashing this rationality for contemporary world societal progress. However, the thesis of this paper is that Habermas’ discourse theory has to be injected as another essential ingredient of the libertarian movement’s self-conception and project.

The discourse part of the markets-and-discourses approach retains an echo of Marx’s and Habermas’ grand emancipatory vision, namely the aspiration to shape human destiny self-consciously via discourse, however, crucially, without requiring or allowing any scope for society-wide authoritative and collectively binding decision making. Instead of the democratic centralism championed by Marx, and instead of the deliberative democracy championed by Habermas, the goal is a deliberative market society or discourse capitalism.

A Libertarian Blindspot: Communicative Action

“Society is purely and solely a continual series of exchanges. It is never anything else. … And this is the greatest eulogy we can give to it, for exchange is an admirable transaction, in which the two contracting parties always both gain.”

Antoine Destutt de Tracy, A Treatise on Political Economy 1817

The libertarian conception of a pure exchange-based society is blind to Habermas’ crucial distinction of communicative action versus instrumental action. There is a blind spot in classical liberal and libertarian theory: its neglect of non-instrumental, communicative-discursive action as foundational to societal reproduction and progress. Classical liberal and libertarian social philosophy only recognize what Habermas calls instrumental action which includes strategic action as sub-category, i.e. instrumental action in relation to other people rather than in relation to inanimate things.

Libertarians as well as classical liberals tend to think: all action is instrumental action, in pursuit of the actor’s interest. We can quote Ludwig von Mises to this effect:

"Action is … aiming at a definite end and guided by ideas concerning the suitability of definite means. … A man's ends are always his own. They are guided by his own valuations, his own will, and his own interest.“ (Mises, 1949)

Not only libertarians, but classical liberals and indeed most economists, tend to model all action as instrumental, guided by personal ends and means. However, this is an impoverished model of human action. Juergen Habermas offers an alternative, expanded typology of cation types. Over and above instrumental action he introduces and emphasizes communicative action — i.e. actions, or interactions, oriented toward mutual understanding rather than strategic success. In contrast to instrumental action, communicative action is dialogical, aimed at reaching consensus through arguments. This type of action presupposes a shared life and often has no explicit goal. It can thus not be reduced to the model of individual goal-oriented action that sees interlocutors as means to the individual’s end. Communicative interaction is cooperative rather than competitive/adversarial. In communicative action, in contrast to strategic action, all motives except that of finding a mutual and consensual understanding are suspended.

In genuine communicative interactions the participants never regard or treat each other as mere means to one-sided ends but always also as selves in their own right, with their own ends. Communicative interactions operate implicitly in accordance with Kant’s categorical imperative: "Act in such a way that you treat humanity, whether in your own person or in the person of another, always at the same time as an end, never merely as a means." (Kant,2012, p.41)[3]. This underlying trust and expectation will become explicit in reflective discourse if suspicion concerning this arises.

Habermas put forward a general, unified theory of social action. This is an important, necessary contribution in the context of a divided (rather than unified) social science landscape. In the social sciences we find radically divergent action-theoretic premises, i.e. different disciplines are based on very different theories about how people make decisions and are motivated to act. On one side of the divide, we find economics and political theory, both based on rational choice theory, on the premise that all human action is (expected) utility maximizing instrumental or strategic action. On the other side of the divide, we find sociology and anthropology with an action theory privileging role-conforming, norm-governed action.

Habermas recognizes both action types and sees that capitalist modernity allowed for instrumental action to become more prevalent via the differentiation of the economic realm as a zone less encumbered by moral bonds and normative prescriptions.

“Only the relative uncoupling of the economic system from the political permits a sphere to arise in bourgeois society that is free from the traditional ties and given over to the strategic-utilitarian action orientations of market participants. Competing entrepreneurs then make their decisions according to maxims of profit-oriented competition and replace value-oriented with interest-guided action.” (Habermas, 1988)

Habermas recognizes both action types - instrumental and normative - but adds a further crucial but mostly overlooked type, namely consensus seeking communicative action. Habermas offers the crucial insight that both instrumental action (privileged by economics) and norm-governed action (privileged by sociology) depend on communicative interaction for their rationality, reassurance and continuous updating. Habermas takes the two most prominent special action theories and integrates them into his unified, general action theory and theory of society.

Game theory has been hailed as unifying paradigm for the social sciences. However, game theory works with the same sparse, impoverished action theory - rational choice based on expected utility - like the rest of economics, just like classical liberalism and libertarianism. Habermas’ student and expert rational choice critic Joseph Heath rightly criticizes – with a view to game theory - that “the instrumental model of social interaction most often used in the social sciences explicitly excludes any kind of communication among agents”.(Heath, 2001). He further notes: “Given the absolute centrality of language for all forms of social life, it is hard to imagine that a theory of action that excludes communication could have more than limited applicability.”(Heath, 2001). He formulates the very reasonable demand that “a general theory of rational action must give some account of all rational activities - not just consumption and voting decisions, but also such paradigmatically rational activities as … debating economic policy, and … reading a book on rational choice theory.” (Heath, 2001). Habermas offers such a general theory of rational action, and it makes sense for us libertarians to augment our theoretical resources and thinking accordingly.

The Social Self: The Communicative Constitution of Economic Agents

"If the individual reaches his self only through communication with others, only through the elaboration of social processes by means of significant communication, then the self could not antedate the social organism. The latter would have to be there first.”

George Herbert Mead, Mind, Self and Society, 1934

Communicative action includes the conversational sharing of our ideas and thinking - in the form of gratuitous chats and that fill our days - and which do serve the function, not necessarily the explicit purpose, of continuous updates of our knowledge and world orientation. Habermas does not refer explicitly to casual, chatty, “random” communications as an important type of communicative action. Habermas, however, uses the phrase life-world with connotations of everyday life and he does emphasize that genuine communicative action is a precondition of the formation and reproduction of the self-conception, understanding of interests, and action-orienting beliefs/knowledge that is in turn presupposed in all instrumental action. Communicative action is thus primary and cannot be dispensed with.

Communicative action is communication seeking mutual understanding, consensus and coordination within inherently shared pursuits. Communicative actions are about the innocent, trusting exchange of ideas and information, in the context of inherently shared lives. In this sincere exchange of information, ideas and concerns, shared goals might be presumed or indeed discovered and defined.

Communicative interaction is about coming to an understanding about the means in view of presupposed shared purposes or about together finding those shared purposes themselves.

The crucial point is that the prior learning processes of communicative interaction - the cooperative seeking of the truth - is always presupposed when individuals make their instrumental or strategic calculations.

Means-end calculations must be reassured and empowered by the knowledge of possible means that could only be gained via prior communicative interaction. Reliable knowledge, including scientific knowledge, is always and everywhere a collective social endeavour in which non-strategic motives like sincerity, repression-free inclusiveness, and openness to criticism must be presupposed. Without positing one's thinking sincerely (non-strategically) one obviously cannot test one's ideas, and one forgoes the required corrective input.

The exposure to criticism is indispensable for learning processes. Isolated individual thinking is very vulnerable to going astray. Once individual agents have been formed by the communicative-cooperative search for truth, they can, from their thus constituted self-interested perspective, launch their strategic actions and negotiations. However, they cannot stay in strategic isolation, i.e. without genuine communicative interactions, for too long without going astray. Everybody needs continuous guiding correctives that flow from sincere and sincerely probing and confirming communications in order to remain a competent rational agent, especially when our lives, in an increasingly complex and demanding societal environment, continuously exposes us to new, non-trivial challenges.

Libertarianism celebrates the market’s capacity to explore possibilities via dispersed knowledges confronting each other in competition. However, learning from these discoveries depends on a complementary domain: non-strategic, scientific discourse.

Hayek famously described markets as discovery procedures. The more free, and unhampered by state regulation, market participants can act, the more exploration, experimentation and discovery can take place. We can therefore expect the libertarian revolution to unleash an unprecedented amount of experimentation and potential discoveries. However, it is never trivial to attribute the success of a complex venture to specific features or aspects of the effort. In order to engender real learning processes the results of market experimentations require scientific analysis and thus non-strategic discourse to capitalise on the experimentation.

Without sincere communication, ideas cannot be tested, corrected, or refined. In our sophisticated knowledge economy, nearly all economic work and consumption choices depend on discourse. Medical therapy choices, legal frameworks, engineering, and finance all rest on discursive validation. In such domains, trial-and-error in markets alone is insufficient. The competitive search for truth, embedded in open-access discursive practices, becomes indispensable.

The production and sharing of the knowledge required for the establishment of means-end reasoning and the effective selection of means thus requires prior, as well as ongoing, communicative action with the shared goal of understanding the world.

Furthermore, the same dependency on genuine communicative action holds with respect to establishing ends worthy to be pursued. These should not be brought to social theory as exogenous and inexplicable data but should be endogenous to a theory that claims generality. The very constitution, identity and stability of rational individual selves, including their goals and subjective interests, depends not only on prior societal socialisation but on continuous elaboration and confirmation in non-strategic communicative interactions.

Thus, Habermas rightly insists in his ‘Theory of Communicative Action’ that instrumental reason is dependent (or “parasitic”) on communicative reason. Habermas writes: “the use of language with an orientation to reaching understanding is the original mode of language use, upon which … the instrumental use of language in general is parasitic.” (Habermas, 1984, p.288). Elsewhere, in ‘Truth and Justification’, Habermas has expressed this inside as follows: “the social practices of a linguistic community are prior to the private intentions of individual speakers” (Habermas, 2003, p.133). Habermas further elaborates: “With the linguistic turn, epistemic authority passes over from the private experiences of a subject to the public practices of a linguistic community.” (Habermas, 2003, p.134). In these passages Habermas engages positively with the American philosopher of language Robert Brandom who’s 1994 work ‘Making it Explicit – Reasoning, Representing and Discursive Commitment’ Habermas calls “a milestone in theoretical philosophy” (Habermas, 2003, p.131).

Robert Brandom shares with Jürgen Habermas the view that communicative, norm-governed practices are fundamental to human social life and cannot be reduced to instrumental or strategic action. Both thinkers situate rationality within a linguistic and normative space of reasons, emphasizing that meaning and agency are constituted by participation in discursive practices. While Brandom does not often cite Habermas directly (though he does so occasionally), he developed a complementary account of normativity and discursive practice that strongly supports the idea that rational agency is fundamentally dialogical, not merely instrumental. In particular, Brandom emphasizes that even instrumental reasoning operates within a background of normative commitments and presupposes participation in social practices of giving and asking for reasons—which aligns closely with Habermas' notion of communicative action as the medium of social integration.

This congeniality of Brandom’s and Habermas’ position and insights, initially reached independently, is significant as it exemplifies the more general convergence of the formerly divergent strands of Anglo-Saxon and continental philosophy.

Markets and Discourses: What they Share, how they Differ and how they Interact

Markets and discourses share important characteristics: Both markets and discourses are bottom-up, decentralised and inherently based on free/voluntary interactions. Both markets and discourses can only function properly and realize their rationality potential if they are open access institutions, i.e. entry must not be barred by incumbents. Neither markets, nor discourses must be monopolized.

However, discourses are not markets and operate rather differently. (The metaphor of a 'marketplace of ideas' does not imply that discourses are markets.) There is a crucial social difference between market exchanges and discursive 'exchanges'. These are very different types of interaction, each with its distinct rules of engagement and criteria of success: Discourses, unlike markets aim for consensus. They operate via the exchange of arguments rather than via the exchange of commodities or services.

The ethics of business dealings is very different from the ethics of discourse. Market exchanges are legitimately self-interested.[4] Discursive exchanges are meant to be serving the common interest in truth/knowledge. Arguments are not a matter of mere personal preference but a matter of evidence-based reasoning.

While it is perfectly permissible, expected and viable in business transactions to instrumentalize counterparties and use them only as a mere means to for one’s own private interests, this is not a viable attitude in discursive interactions. In business transactions, no shared purposes are presupposed. Discourse, in contrast is always a shared learning project. A strategic attitude, e.g. aiming at winning the debate by means of withholding information, is subverting the learning opportunity for both parties. Any attempt to instrumentalize one’s interlocutor in a one-sided way for ulterior purposes will, if detected, lead to the breakdown of communication. Therefore, while instrumentalizing others is both permissible and viable in markets, it is both impermissible and self-defeating in discourses.

In the terms of Juergen Habermas’ fundamental distinction of action types, discursive exchanges are 'communicative’ interactions, in contrast to the 'instrumental' or ‘strategic’ (inter)actions that make up markets.

Classical liberal and libertarian social philosophy are blind to Habermas’ distinction of communicative versus instrumental action. This blind spot in the basic typology of action types leads to a further blind spot, namely the failure to take account of discourse, a special reflective variant of communicative action, as necessary ingredient of societal self-regulation and progress.

Discourse, including political discourse and scholarly legal discourse, although a bottom up, competitive process in its own way, is itself no market process. Exchanges of property titles, money and monetary negotiations neither can nor should determine political legitimacy, legality or justice, just as monetary exchanges neither can nor should determine scientific truth. Rather, political legitimacy, justice and truth are determined by the exchange of arguments in discourses.

In modern liberal society the domains of the economy, the sciences, morality, law, and politics have been differentiated and cannot be collapsed, as the Marxists tragically presumed, trying to subordinate them to politics, thereby annihilating them. Neither can these distinct domains be collapsed into a mere market process. These insights have been most emphatically articulated in Niklas Luhmann’s (1927-1998) theory of modern, functionally differentiated society (Luhmann, 1984), and have also been absorbed within Habermas’ theory of society.

Discourses interact with markets by way of informing market participants. Market participants may choose not to bother to join discourses. However, this is rather limiting and potentially self-injuring. Many products and services are too complex and opaque in their effectiveness to allow for fail-safe intuitive choices. That is why discourses, including scientific discourses, are often required to establish their functioning or lack of functioning, their beneficial effects or detrimental side-effects.

Propositions about utility or economic best practice are operating with arguments based on evidence and logic. These are not questions of mere preference or choice.

This does not mean that economic actors cannot choose to make uninformed individual choices. This freedom is not in question.

However, if actors want to claim that their choices are beneficial (for themselves or for society at large) they have to argue within discourses rather than merely exchanging in markets. Again, these are very different types of interaction, each with its distinct rules of engagement and criteria of success. Market exchanges, especially in the knowledge economy, rely on the communicative rationality of discourses.

This reliance on globally shared and accessible discourses is ever increasing and bound to increase further into the future as the life of society increases in technological and organisational sophistication. The rationality of market exchanges is increasingly underpinned by competencies that only non-strategic communicative interactions in discourses can deliver. More than ever before, in our sophisticated “knowledge economy”, nearly all economic work/production is highly dependent on (scientific/technological) discourses. The same applies to consumption choices (e.g. medical consumption). All business operations (management accounting, economic calculation) and financial interactions are also increasingly dependent on management science, legal expertise for contracting, accounting science, economic science and finance science, utilizing asset pricing models, portfolio theory etc.

All these rationality enhancements are underpinned by public, open discourses within which viable approaches and theories compete, are weeded out or are selected and gain ground on the basis of arguments and evidence withstanding criticism and scrutiny, rather than on the basis of profitability or power. Successful contributions inspire citation, application, following and further work in the competitive and cooperative search for truth.

Science and Discourse Ethics

The very technological base of economic progress depends on the progress of science. Science is a discourse, and discourses constitute a radically distinct type of human institution with a distinct social logic and criterion of success. The sciences therefore cannot be equated with markets, nor should they be assimilated to markets, even after all government subsidies to science have been abolished. Even today, all scientific activities are also somehow tied up with economic exchanges, labouring at least indirectly under the constraint of economic reproduction. This constraint does not determine the essential governing logic of scientific discourse. If suspicions arise that private economic prerogatives and motives are at play and intervene in the outcome of a scientific investigation, e.g. in areas like health, environmental impact or economic policy, the scientific credibility of the result is thereby undermined, and its status as scientific knowledge is suspended. Discourses cannot be constituted via strategic interactions. Indeed, strategic action undermines and prevents discourse.

Any knowledge claim becomes suspect or even void, if the originating social/communicative process significantly deviates from the regulative ideal of a sincere, open, critical, “democratic”, repression-free discourse. This means that modern, enlightened knowledge claims, as well as all practical rationality claims, presuppose not only the absence of special/private economic interests, but furthermore presuppose modern liberal non-stratified social relations between ‘free’ individuals and exclude relations of bondage or indeed reifications like nobility, papal infallibility, the authority of the bible etc. These reflections show that modern knowledge, science, and rationality depend on modern liberal morality.

Proper knowledge can only emerge within a liberal regime. The concept of pre-modern knowledge (excepting ancient Greece) is an oxymoron. For instance: The fundamental principle/criterion of universal reproducibility in science presupposes open access to the discursive search for truth.

The knowledge we mean cannot exist in traditional societies, and in fact did not exist. They could live, seemed to “know” something, a lot in fact, but they did not really securely KNOW anything. It was all very uncertain, mixed up with superstition, and far less practically successful than modern scientifically based societies. In particular, the scarcity and fragility of usable knowledge also limited those in power who tried to control and actively suppressed discourse.

Habermas uses this reflective insight to ground this liberal morality, rebutting relativistic philosophical positions that claim all morality to be a matter of mere subjective choice not susceptible to rational argument. Habermas argues: Relativist/subjectivist moral philosophy, by joining the discourse and trying to convince the participants in this discourse that the search for a rationally justifiable morality is vain, fails to reflect that it thereby is already implicitly committed to a particular morality, namely to the liberal morality underpinning the modern social institution of discourse. Science and rationality, including moral philosophy, is always already underpinned by a specific morality, namely ‘discourse ethics’. Habermas project of discourse ethics is then to make these moral presuppositions of all discourse explicit, thereby rationally grounding “our” particular, enlightened, liberal moral system. This system of morality evolved together with other ingredients of modernity, like the scientific worldview, the rule of law, constitutional government etc.

Conclusion: From Market Society to Discourse Society

This paper argues that a libertarian social order cannot subsist on markets alone. While market exchanges are essential to coordination and innovation, they must be complemented by discursive processes. Discourses are not reducible to markets or market transactions. They operate via communicative, not via strategic action; they aim for understanding, consensus and truth, not profit.

Drawing on Habermas’ theory of communicative action, the paper argues that rational agency presupposes participation in discursive practices. These insights challenge the libertarian reduction of all social processes to instrumental exchanges. Instead, discourse capitalism proposes a voluntary, decentralized society rich in critical-reflective communication without collective coercion—a new vision of a stateless societal order.

Libertarianism seems to call for a society of private market exchanges with a diminished scope for politics, while Libertarianism’s most radical expression, anarcho-capitalism, even implies, seemingly, that politics will disappear altogether. The author challenges this aspiration for an ‘apolitical’ society and instead insists that political life, understood properly, will, for good reasons, very much intensify rather than vane after the libertarian revolution. This intensification is due to the proliferation of innovations and parallel options in all the ‘political’ areas of societal development like the introduction of cutting-edge technology, environmentally sustainable energy provision, ethical questions raise by new technologies or social trends, new business models, employment relations, financial innovations, innovative forms of social control, the development of new law, new forms of legal dispute resolution and law enforcement etc. There will be myriads of public issues to debate, evaluate, warn against or promote and proselytize about. This should be good news for all of us libertarians who are passionate about politics and the common weal: the success of the libertarian political project does not imply that political discourse ceases and that we’ll have to retire from political life and thereafter care for our back gardens rather than for humanity’s world-historical progress. This essay thus anticipates and paints a positive picture about a much needed and much energized political life after the libertarian revolution, a revolution that will have to be, to borrow a phrase from Leon Trotsky, a permanent revolution (Trotsky, 1931). This entails not only ongoing vigilance in order to keep any insipient re-emergence of statism in check, but also the expansion and ultimately universalization of the libertarian revolution across the world, radiating out from wherever it first breaks through. Political parties will transmute into activist think tanks or proselytizing movements. A specifically libertarian 'party' might persist—not to oppose state coercion, but to resist illiberal practices and attitudes across society.

To conclude: The thesis of this essay maintains that the future libertarian society cannot be a society of pure market exchanges but must be a society that is constituted via both markets and discourses. However, with a view to a free society it is important to note that both markets and discourses are wholly voluntary, open, repression-free, non-coercive institutions. A society based on markets and discourses as mechanisms of societal integration is an inherently free society. To distinguish this theory from current versions of anarcho-capitalism, this treatise puts forward and promotes the term ‘discourse-led capitalism’, or abbreviated ‘discourse-capitalism’, to emphasize the increasing importance of discourses for contemporary society with its increasingly sophisticated economic life. The conditions of the unconstrained, rational, productive flourishing of market processes are distinct from the conditions of the unconstrained, rational, productive flourishing of discourses. In fact, discourses are crucial to steer the institutional development of market processes, while well-functioning special markets are required to materially underpin and reproduce the flourishing of discourses without distortions or special interest capture.

Thus, the future progress of human civilization within a stateless world society will be ordered via markets and steered via discourses.

References

Brandom, R. (1994). Making it Explicit – Reasoning, Representing, and Discursive Commitment. Harvard University Press.

Clark, G. (2007). A Farewell to Alms: A Brief Economic History of the World. PrincetonUniversity Press
Dennett, D. (2003). Freedom Evolves, Penguin Group

Destutt, A. (1817). A Treatise on Political Economy, translated by Thomas Jefferson, reprinted: Indianapolis: Liberty Fund, 2011

Habermas, J. (1986). The Theory of Communicative Action, Volume I: Reason and the Rationalization of Society (T. McCarthy, Trans.). Beacon Press. Original German published 1981
Habermas, J. (1987). The Theory of Communicative Action, Volume II: Lifeworld and System. Beacon Press. Original German published 1981

Habermas, J. (1988). Legitimation Crisis. Polity Press. (Original German published 1973)

Habermas, J. (1989). The Structural Transformation of the Public Sphere – An Inquiry into a Category of Bourgois Society. Polity Press. (Original German published 1962)
Habermas, J. (2003). Truth and Justification (B. Fultner, Trans.). MIT Press. (Original German work published 1999)

Hayek, F. A. (1945). The Use of Knowledge in Society. American Economic Review, 35(4), 519–530.

Hegel, G. W. F. (2001). The Philosophy of History (J. Sibree, Trans.). Dover Publications. (Original work published 1837), p. 22.

Kant, I. (2012). Groundwork of the Metaphysics of Morals (M. Gregor & J. Timmermann, Trans. & Ed.). Cambridge University Press. (Original work published 1785)

Kant, I. (1998). Critique of Pure Reason (Translated by P.Guyer, A.W.Wood). Cambridge University Press. (Original work published 1781)

Luhmann, N. (1995) Social Systems. Stanford University Press. (German original Work published 1984)

Luhmann, N. (2012) Theory of Society, Volume 1. Stanford University Press (German original Work published 1997)

Luhmann, N. (2013) Theory of Society, Volume 2. Stanford University Press (German original Work published 1997)

Heath, J. (2001). Communicative Action and Rational Choice. MIT Press.
Marx, K., & Engels, F. (2002). The Communist Manifesto. Penguin Books. (Original German 1848)

Mises, L. (1935). Economic Calculation in the Socialist Commonwealth, in F.A. Hayek, ed., Collectivist Economic Planning, London, Routledge & Sons. original German 1920 in the Archiv für Sozialwissenschaften 47
Mises, L. (1949). Human Action: A Treatise on Economics. Yale University Press, Chapter 4, Section 1

Mead, G. H. (1934). Mind, Self and Society. University of Chicago Press, p.233

Rothbard, M.N. (2012). For a New Liberty, The Libertarian Manifesto. Skyler J. Collins (Original work published 1973)

Schumacher, P. (2025). Pathways to a Stateless World Society, patrikschumacher.substack.com/publish

Trotsky, L. (1931). The Permanent Revolution (M. Eastman, Trans.). New Park Publications. (Original work published 1930)

[1] Rothbard laid out the case for anarcho-capitalism in his 1973 book For a New Liberty, The Libertarian Manifesto.

[2] Socialist countries muddled through by drawing price orientation from capitalist world prices and prices from capitalist countries. Obviously this crutch would no longer be available after the communist world revolution.

[3] Kant’s moral philosophy was one of Habermas’ sources in developing his theory of communicative action.

[4] In the economy instrumental rationality prevails, but instrumental rationality is far from absolutely dominating here. Communicative rationality via communicative action also plays a role in work and business relations.

THESIS: The overwhelming prominence of sustainability in architectural discourse contributes to the end of architecture and should be rolled back.

To avoid misunderstanding: The pursuit of sustainability is a worthy and important cause. The thesis put forward here in no way relies on climate change denial or scepticism. Further, the suggested rolling back of the discourse on sustainability in no way implies the reduction of the actual practical effort towards sustainability in the urban development arena.

The problem is that the sustainability agenda in professional architectural discourse is so overwhelmingly prominent that it crowds out all other aspects of architecture, some of which are not only equally worthy of attention and debate but (as we shall see below) should have priority over the sustainability agenda. Without having measured this, I believe that at least 90% of all professional architectural conferences either carry sustainability in their title or propose sustainability as a key focus of the conference. However, there is hardly that much original news about innovations at the sustainability front that would merit taking up so much of the attention space. Rather, these are mostly occasions for repeating familiar solutions, plus the reiteration of commitment to the good cause. Even if the new information coming from sustainability research is judged to be inexhaustible, the law of diminishing returns and rising opportunity costs suggests that the current monopolisation of attention would still have to be judged irrational.

Another problem for architecture is that the focus on sustainability within the architects’ discourse implies that questions of technical functionality that ultimately lie in the engineers’ domain of competency swamp questions of social functionality that constitute architecture’s genuine domain of competency. Architects are thereby reduced to amateur engineers. In the main article some genuinely architectural agendas - focussing on social rather than technical functionality – have been sketched out. Here a hint must suffice. In my theory of architecture, I categorize these agendas under four headings: spatiology (organisation), phenomenology (articulation), semiology (communication), and dramaturgy (interaction). This list of theory-led, genuinely architectural R&D agendas is meant to serve here to point more concretely to the discursive opportunity costs that come with allowing sustainability to usurp all the airtime.

To be fair, some aspects of sustainability are within the architects’ domain, namely those which have to do with the composition of the massing relative to sun exposure. It is also worthy to note that the paradigm of “tectonism” (2) I am promoting is utilizing engineering logics – including environmental engineering – as drivers for an enriched phenomenological and semiological articulation, i.e. the genuinely architectural agenda of articulation serving social functionality by enhancing the communicative capacity of architecture in the service of the built environment’s legibility. (However, this agenda of legibility – while using the telling, differentiated morphologies that emerge from environmental engineering – is in no way an automatic fall-out from the sustainability agenda but requires its own discourse.) After these caveats, I would like to return to the main argument that the sustainability agenda needs to be cut back to its proper size and positioned in its proper place within the overall architectural discourse.

My key point: The prioritising of sustainability is not only crowding out other, more important aspects of architecture, but it is putting the cart before the horse, by focussing on negative side-effects before focussing on positive goals. This point is decisive: Sustainability addresses only the cost side of a development. However, before we talk about minimising costs (in terms of carbon footprint) we should discuss the benefit side, i.e. we should, first of all, scrutinize and clarify the positive goals we are trying to achieve with the development. Only after we have ascertained the positive functions and social benefits we are aiming at, and only after we have selected one of many critically compared and appraised design proposals with respect to the social functionality performance criteria, can the cost question be meaningfully posed. Then the question can indeed be posed: How can the most beneficial design solution be realized in the most cost-effective way, with the lowest possible carbon footprint. To be sure: it can be objected that the cost aspect should be tracked alongside the exploration of benefits, rather than be left to a separate second step. However, it still makes heuristic sense to explore the maximisation of benefits initially independently of costs. In any event, the distinction of costs criteria and benefit criteria should be fundamental, and that the discussion of benefits must have logical priority over costs should be indisputable. Otherwise, the solution can only be: avoid building altogether. Therefore, if we set sustainability as top priority, abstaining altogether from development should be the conclusion, and it seems, some are indeed willing to draw this conclusion.

There is something seriously amiss and illogical in a discourse where the debate of how to minimize costs crowds out the logically prior debate of how to maximise benefits. This verdict of being illogical can only be avoided if the benefit side is seen to be trivial, given, ahistorical, fixed. If sustainability (minimizing environmental costs) takes up all the airtime in a discourse, then this discourse seems to assume that the goals and desires that motivate development are clear without requiring discussion. Further, this discourse must assume that the design solutions delivering the desired benefits are known and can be taken for granted. Such a discourse, by implication, is not expecting any scope for innovation on the benefit side of development.

How to lower the carbon footprint of a building does not tell us anything about how to shape the building with respect to end-user needs and societal needs. Those who think that there is nothing more important to be concerned about than sustainability, operate with the implicit assumption that the needs of individuals, groups and societies have long since been settled, leaving only the cost aspect as a live concern. This is a false assumption.

This attitude might work in other industries: If you want to produce steel, the positive goal is indeed trivial: you simply want to produce more steel, of this or that standard grade. Here predominant or even exclusive focus on costs makes sense. But architecture and urban development are not like this at all. Here the goals that might be pursued are complex, multi-facetted, and require much debate to be clarified and updated. The same applies to the exploration of the solution space, once the goals are broadly defined. These questions should be prioritized and deserve to absorb the bulk of the discursive space in our discipline and profession.

The problem is: to formulate the positive goals and positive criteria of success and ways of appraising ("measuring") this success is truly non-trivial, and to measurably innovate on the side of the benefits is even more demanding. There are many perspectives and myriads of parameters to be considered and a potentially vast search space to be explored. One might surmise that is why so many architects (and clients too) fall back on sustainability as easy criterion to set targets and to measure success, although this says NOTHING about the success in terms of the original purposes of the project that motivate the project in the first place.

I have seen this fallacy be committed by the most sophisticated clients. For instance: When Google was looking for an architect to design its new campus in Mountain View, they set their NetZero target as the primary criterion to select their architect for. (This was in an earlier round of solicitation, not the round where BIG/Heatherwick were chosen). An absurd neglect of the potential benefits a properly focussed design effort could deliver in terms of Google’s primary purpose: the productivity enhancement for the many thousands of creative workers who were to be gathered on the new premises to collaborate there. This swapping of priorities makes neither sense for Google, nor for the global common weal: a google R&D campus is not the right occasion to foreground the aggressive curbing of carbon emissions as primary goal. It patently irrational to hand over this precious occasion to craft an intricate platform for knowledge exchange and collaboration - a high productivity social machine - to a virtue signalling event. The contribution from the trivial amount of carbon emission reduction that can be achieved here stands in no meaningful relation to the foregone boost to world prosperity that an enhanced synergy process across thousands of smart knowledge workers can be expected to generate. Obviously, Google did not understand the difference architecture can make. This failure in Google’s understanding about where architectural innovation and optimisation might be possible with respect to their project was probably due to the way Google’s facility managers and inhouse architects were blinded by the way sustainability had (and continues to) dominate the architects’ discourse. To be sure: I would never object to bringing in sustainability criteria in the subordinated position they logically belong, after the social performance criteria have been allowed to select the best solution in line with the original purpose of the project.

The re-balancing of the priorities of architectural discourse is more urgent than ever, especially with AI capabilities boosting our power of creative exploration and testing. The time is ripe to attempt to re-vitalize, or rather to resuscitate. our discipline, a discipline that has been committing suicide by pushing woke, anti-capitalist social justice agendas in academia and a one-sided, illogically prioritized sustainability agenda within the profession, thereby crowding out all its innovation potential, and undoing all that had - since the Renaissance - elevated architecture above mere building.

(1) Schumacher, Patrik. The End of Architecture, Published in: KHOREIN – Journal for Architecture and Philosophy, Vol 2, No 2 (2025)

https://www.academia.edu/127571727/The_End_of_Architecture

(2) Schumacher, Patrik. Tectonism – Architecture for the Twenty-First Century

published by: Images Publishing – The Arts Bridge, Melbourne 2023

Published in: KHOREIN – Journal for Architecture and Philosophy, Vol 2, No 2 (2024): END

ABSTRACT:

Architecture, as an autonomous, theory-led discipline, has ceased to exist. This paper posits the “end of architecture” not as rhetorical hyperbole but as a historically grounded assertion. The discipline has self-dissolved, eroding its intellectual and professional autonomy under the pressures of anti-capitalist politicisation and woke virtue signalling. Once defined by rigorous critical discourse, architectural innovation, and theoretical grounding, architecture has devolved into a fragmented practice now operating on the level of a mere craft rather than a science-based, academic discipline and profession. Academic institutions, biennials, and professional critiques have abandoned their roles as incubators of architectural thought, instead engaging with tangential sociopolitical issues that stray from architecture’s core competency.

Through a rational reconstruction of architecture's historical trajectory, the paper delineates architecture’s unique societal function: the provision and continuous innovation of spatial frames that order and articulate social interactions. The societal function and theoretical self-steering of architecture—established during the Renaissance and further evolved in modernity—has been replaced by a diffuse pluralism that denies the discipline its critical coherence and forward momentum. Key contributing factors of architecture’s demise include the ascendancy of euphemistic and trivializing checkbox itineraries in professional work, the decline of rigorous critique in academic settings, and the dominance of politically charged, non-architectural agendas in prominent forums like the Venice Biennale. Furthermore, the erosion of professional agency through political over-regulation undermines and discourages theory-led innovation and thereby further exacerbate the withering of architecture.

Although more pessimistic than optimistic for the immediate future, this paper posits the necessity of reasserting architecture’s specific social function, of reclaiming agency, and re-establishing its critical discourse to foster innovation aligned with societal progress. It challenges architects, theorists, and educators to reject pluralistic complacency, reinvigorate constructive critique, and refocus the discipline on its core societal responsibility. Only through such recalibration can architecture emerge from its current dissolution and reclaim its role as a distinct and essential function system in the development process of contemporary society.

KEYWORDS: self-annihilation, theory-led discipline, mere building, function system, spatio-morphological framing, politicization, woke take-over, discursive culture

We are witnessing the end of architecture, the voluntary self-dissolution of architecture. The usurpation of the discipline by woke ideology is only one aspect of architecture’s dissolution, not the full explanation. The self-destruction of architecture as academic discipline, as distinct discourse and as theory-led profession, is already a fait accompli. Architecture has ceased to exist. What does this even mean? What kind of statement is this? Is it mere polemic hyperbole? If not, is this meant to be an empirical statement of fact, or a statement stipulating a normative concept of architecture? Obviously, professional firms employing registered “architects” are still designing buildings. However, the resultant structures are no works of architecture, but mere buildings, because the design of these buildings is no longer informed and steered by a living, critical discourse.

A pertinent concept of architecture

What underlies the thesis put forward here is neither a simple empirical concept of architecture—everything in ordinary parlance called architecture or everything designed by a registered architect—nor a normative stipulation like “we should only count as architecture works that meet a set of stipulated quality criteria.” Instead, the thesis of the end of architecture is based on a historically grounded, rational reconstruction of a (functionally important) always already operative concept of architecture that will be elaborated upon below.

The term “rational reconstruction” was first put forward by the philosopher Rudolf Carnap, in the context of the philosophy of science. Carnap introduced the concept in his 1928 book The Logical Construction of the World, and defined the concept (in the 1967 edition) as follows: “By rational reconstruction is here meant the searching out of new definitions for old concepts. The old concepts did not ordinarily originate by way of deliberate formulation, but in more or less unreflected and spontaneous development. The new definitions should be superior to the old in clarity and exactness, and, above all should fit into a systematic structure of concepts.” (Carnap, 1967, p.v).

The concept of architecture reconstructed here is based on the widespread intuitive distinction—as all concepts are distinctions—between architecture and mere building. This distinction—works of architecture versus mere buildings—might be illustrated or exemplified by juxtaposing a prominent building successfully designed and built in accordance with an architectural competition as an exemplar on the side of architecture, with exemplars like a local garage, suburban supermarket, or run-of-the-mill terrace house on the side of mere building. The programme of rational reconstruction now asks for a set of explicit criteria that allow us to assign all buildings to these two categories, resulting in a partition that sufficiently matches our intuitive sorting.

The first aspect to make explicit is that architecture, as distinguished from mere building, is inherently connected to architectural discourse and theory. Theoretical treatises are essential components of the discipline and profession of architecture. Works of architecture therefore always link up with or relate to theories or manifestos of architecture. A second important criterion is that architecture, in contrast to mere building, is also marked by innovation. In the case of important works of architecture, these are original, pathbreaking innovations. For instance, Michelangelo’s bold use of the so-called “colossal order”—first but very rarely introduced by Alberti—represents an advancement of architecture from the Renaissance (via Mannerism) to the Baroque. We must also include early adopters of such original innovations and all those who follow and spread an innovation with an awareness of the discourse of architecture. In all these cases, works of architecture are original creations put forward by architects claiming authorship. That all works of architecture are attributable to named (and educated) architects claiming authorship and responsibility is the third distinguishing feature of architecture versus mere building. Innovation questions tradition and requires an argument that transcends the mere concerns and competencies of building. Argument implies theory. In contrast, mere building—the vernacular—relies on tradition and well-proven solutions taken for granted, without authorship claims. The status quo does not require theory, nor a point of reference and responsibility. Beyond marking an important point of definition and distinction of architecture versus mere building, this reflection affords a functional explanation of the emergence of theory as a necessary ingredient of architecture as a discipline and profession with an inherent adaptive forward drive.

That only theoretically and historically informed building design constitutes architecture can be confirmed by every practicing architect who has undergone the rituals of architectural socialisation at university, where history and theory were, until relatively recently, part of the architecture curriculum, and often enough feeding into design studio discussions.

A theoretically reflective practice can considerably accelerate its progress and its contribution to overall societal development. Innovation calls for theory to substitute for the assurances that were provided by adherence to tradition. Theory thus contributes to modernity’s shift from conservation to progress.

Since innovation is a fundamental aspect of architecture, radical innovations that take root are most highly valued and mark out the respective works. Every great work of architecture offers a radical innovation. That is an empirical observation of the way the discipline evaluates itself. Many great architects, as valued within the discipline, are also important architectural theorists. This is another fact of communication. Virtually every architect who “counts” within the history of architecture was both an innovator and a theorist or writer. The most striking examples are Alberti and Le Corbusier, but we might also mention Palladio, Soane, Schinkel, Semper, Wagner, Wright, Gropius, as well as Koolhaas and Eisenman, among others. This immediate link between “great architecture” and significant theory is especially pronounced in the twentieth century: virtually all modernists, post-modernists, and deconstructivists, as well as the protagonists of parametricism, were theoretically articulate and in lively discursive exchange with each other, as well as with critics and academics. In recent years, however, practising architects are disconnected from architectural theory (critics and academics), as architectural theory has shifted its focus away from engaging with the work of leading architects.

Architecture versus mere building is constituted by virtue of architectural theory, innovation and original authorship claims. That is why architecture proper, as understood here, only begins with ancient Greek architecture, where both architect-authors (Ictinus, Callicrates, Hippodamus, fifth century BC) and theoretical treatises (the sources of Roman theorist Vitruvius) existed. With respect to ancient Rome, we might name Apollodorus of Damascus (second century AD), who is known for promoting innovations like the dome. After the demise of Rome, architecture disappeared and only returned in the Renaissance. The Romanesque is best understood as a degenerate, vernacular version of Roman architecture. The high point of achievement before the Renaissance—the Gothic cathedral—is indeed very impressive, but no complete designs prior to construction existed, and no individual authors can be named. In contrast, the names of Alberti, Bramante, Serlio, and Palladio are still alive within the recursively reproduced memory of architecture’s ongoing discourse. There are no equivalent figures from the Gothic epoch.

Most importantly, the essential ingredient that turns tradition-bound building into self-conscious architecture—a public, critical discourse that emphasizes creative innovation and demands arguments for those innovations—is missing in Gothic building practice. There is indeed a big difference between secret guild knowledge and the public circulation of treatises. It is this difference that motivates and justifies the thesis that architecture starts, or rather restarts, with the Italian Renaissance.

The differentiation of a dedicated theoretical strand within the discipline of architecture is one of the defining factors that contribute to the differentiation of architecture as an autonomous subsystem of societal communication. (The characterisation of architecture as function system within modern, functionally differentiated society will be elaborated in the following chapter). This tight link between the existence of architecture as a separate discipline/profession and architectural theory is also empirically evidenced by the historical coincidence of the emergence of architecture as a separate profession and the publication of dedicated architectural treatises. The theory of architectural autopoiesis[1] adds a theoretical explanation to this evidence—an explanation that construes the necessity of architectural theory on the basis of a functional exigency that acts as evolutionary attractor for the differentiation of this function system. This functional exigency is the need to accelerate the innovation of the built environment to an extent that contradicts the mode of evolution offered by the traditional system of guild-based handicraft organisation. In this context, theory replaces tradition. The necessity of architectural theory is thus asserted by the identification of its primary function. The primary function of architectural theory is to facilitate the rapid adaptation of architecture to an accelerated process of technological and socio-economic transformation since the advent of (early) capitalism. Theory must compensate for the lost certainty of tradition, where the appropriateness and functionality of buildings were guaranteed by the fact that new buildings consisted of nothing but the faithful repetition of long-evolved and surreptitiously corroborated models. The validity of traditional practice could be taken for granted and did not require a special communicative effort to solicit their acceptance. The moment when traditional practice falters is the moment when architecture takes off.

Architecture is a discourse that is geared towards permanent innovation, keeping up with and promoting a dynamic society. The societal need for a permanently updated built environment—inevitably given in a society (since the advent of modernity) that expands and transforms relatively rapidly—is first the evolutionary attractor for architecture’s crystallisation (as a theory-led innovation engine for the built environment) and then the selector for architecture’s further historical evolution. Thus, the concept of architecture reconstructed here identifies innovation as a key criterion, alongside architectural theory explaining the benefits of the innovation, and alongside authorship taking responsibility for the innovative work. However, tragically, innovation at the frontier of our fast-evolving technological civilisation is no longer something that engages the “architectural” discourse in universities, exhibitions, conferences, and magazines. Here, topics like climate change, racism, Eurocentrism, decolonisation, degrowth, etc., abound. These are topics that, if at all, relate only negatively to contemporary architecture. To illustrate: the headline of a CNN article about the 2023 Venice Architecture Biennale was accurate: “Racism, activism and climate crisis are on the agenda at the Venice Architecture Biennale.”[2]

Architecture’s Societal Function

Above, we have focused on the distinction between architecture and mere building. This makes sense in the context of architecture’s demise as understood here. However, both building and architecture address a fundamental societal function: the necessary spatial ordering of societal interaction processes. The problem is that under contemporary conditions of societal versatility, complexity, and dynamism, only an academically based, discursively empowered profession, developed via theory-guided research and experimentation, can fulfil, adapt, and progress this societal function of the built environment. When architecture “degenerates” back to a state of mere craft or tradition-bound building, the societal function of the built environment can no longer be fulfilled. Consequently, overall societal progress is thereby slowed down and stunted.

Here is the author’s definition of architecture’s societal function as first provided in The Autopoiesis of Architecture[3]: “All social communication requires institutions. All institutions require architectural frames. The societal function of architecture is to order/adapt society via the continuous provision and innovation of the built environment as a system of frames.”

There is no human community without an artificial built environment. It is the built environment—together with all artefacts—that provides cultural evolution with the cross-generational, material substrate it needs and by means of which an advantageous social order can persist and grow. In this respect, it is comparable to the DNA of biological evolution. Human settlements form and accumulate ever larger and more differentiated spatio-material structures as the skeleton for social structures, as it were, that without this substrate would not have managed to attain such a scale, which is indeed unnatural for primates. Moreover, the level of cooperation so important for human productive abilities would not otherwise have emerged, been replicated and advanced. What applies to the beginning of cultural evolution still applies today in relation to the developmental tasks currently facing us. Architecture’s fundamental original achievement is not the oft-invoked protection from the elements but a structure-forming achievement: the achievement of order. The built environment organises social processes of interaction and plays a crucial role in the establishment and stabilisation of social order. It also involves ownership, spatial exclusion, and demarcation by means of physical barriers with corresponding rights of access. Yet above all, it involves the spatial distribution and functional configuration of types of interaction or communicative situations, by means of semiological codes, whereby relative spatial positioning is also a means of coding. The built environment structures social situations and provides orientation for the participants in the social processes thus organised, who then find their place of their own accord. It supports and communicates the social structure which is always a configured network of cooperation. While the social structure as a whole can hardly be made visible any longer, each of the local social structures, offerings, and options for communication can still be articulated and made transparent (although this requires a special, dedicated design effort). In short, spaces potentially communicate what is on offer and who can take part.

The author’s theory of architecture—the theory of architectural autopoiesis—is embedded in the wider theoretical edifice of Niklas Luhmann’s social system theory[4] and theory of society[5]. That any comprehensive, self-reflective theory of architecture should make its underlying premises explicit and therefore must refer to a theory of society should, once stated, be uncontroversial. This has rarely been done, but Alberti referenced explicit conceptions of the good society in his reflections on city form and architecture. Some of the theorists of modernism—for instance, the authors of the ABC group (Schmidt, Stam, El Lissitzky, Meyer)—were firmly and explicitly based on a conception of society in line with Marx’s theory of scientific socialism, augmented by an account of recent technological and socio-economic developments. The author’s conception builds on Luhmann’s theory of modern, functionally differentiated society, augmented by the insights from the post-Fordism debate[6] and integrating more current conceptions of the knowledge economy and network society.

That some coherent account of the technological, economic, and sociological conditions and developmental dynamics of society must underlie any pertinent formulation of architecture’s societal task should be self-evident. The author’s analysis of contemporary society, in the terms indicated above and further explicated below, should not be controversial. The author’s libertarian political convictions are not presupposed here. What is presupposed is that architecture—as architecture versus mere building or craft—should be based on theoretical guidance for its design tasks, based on a broad theory of society that takes account of historically recent conditions, such as the momentous technological transformations (internet, computation, robotics, AI), resultant socio-economic transformations (from Fordism to post-Fordism), and the reality of historically recent neo-liberal privatisation dynamics. All these factors are spurring a further urbanisation drive and a new urban concentration dynamic to which architecture and urban design must congenially respond.

What is, in contrast, not conducive to the ongoing vitality of architecture as an academic field and theory-led professional practice, is the purely negative stance towards these recent and ongoing historical developments. This negative stance, however, has become increasingly prevalent not only in academic circles and architecture-related cultural institutions but also among architectural critics and, indeed, practising architects. OMA’s Reinier de Graaf expressed this stance in an article published in Architectural Review entitled “Architecture is Now a Tool of Capital, Complicit in a Purpose Antithetical to Its Social Mission”[7] . Taken to its logical conclusion, the irreconcilable, anti-capitalist stance that judges all current urban development activities to be politically and morally compromised—and all architects participating as “sell-outs”—is a key factor in the demise of architecture, as it serves to cut the constructive link between architectural theory and architectural practice, leaving the latter intellectually adrift.

Within modern society, it is functional differentiation that becomes the pervasive and predominant mode of societal differentiation (in contrast to stratification as the formerly dominant mode of societal differentiation). The most striking manifestation of this general tendency is the emergence of the great “function systems” as the major subsystems of modern society: the economy, the political system, the legal system, science, the education system, and the mass media are distinct, autonomous systems of communication that have differentiated according to the indispensable societal functions they perform (emancipated from their former fusion and subjection within a top-down stratified order). An important insight of Luhmann is that these function systems operate via self-referential communicative closure[8], i.e., they evolve highly specialised discourses (systems of communication), each with its own categories, lead distinctions and evaluative codes. These communication structures are idiosyncratic to each function system which in turn implies a discursive incommensurability. These are separate discourses and professions. They are “autonomous” not in the sense of being insulated or unresponsive but in the sense that each is discursively self-steering its adaptation. The subsystems co-evolve, with each observing and adapting to all the others. There is no hierarchical command-and-control structure that could integrate these subsystems. There is no unified control centre in functionally differentiated society. The political system is not such an omnipotent control centre but just one of many autonomous function systems. The attempt to politically control the sciences, the capitalist economy, the justice system, etc., would just spell the destruction of science, capitalism, and justice. The totalitarian attempts by figures like Hitler and Stalin accomplished precisely this annihilation. The same logic of “control equals annihilation” applies to architecture. Both Hitler and Stalin did indeed annihilate international modernism—the discourse- and theory-led discipline/profession of architecture of this time—within their territory.

The premise of the author’s Autopoiesis of Architecture[9] is that architecture is one of the great function systems of modern, functionally differentiated world society: a function system with its own exclusive and universal responsibility for an important societal function that demands independent treatment and promotion by an autopoietic communication system specifically differentiated to focus on this function. The differentiation of a function system, i.e., a specialised discourse and theory-guided professional practice, made functional sense under conditions of accelerated societal development and became part and parcel of this transformative development.

The advent of modernity, involving the spreading of capitalism with its unique dynamism, as identified by Marx and Engels in the Communist Manifesto[10], and the development of science manifesting an equally restless dynamism—first in parallel with and then in mutually spurning interaction with capitalism—implied an acceleration of societal progress and prosperity. All function systems—not only architecture but also the political system, legal system, economic system, and the system of the sciences—began to be accompanied and spurned on by what Luhmann terms “reflection theories”[11], that is, guiding treatises and, indeed, whole critical, theoretical literatures.

A first hint that architecture addresses an indispensable function is the fact that there is no human society without a built environment, just as there is no human society without political institutions, law, an economic system, a system of socialisation, or a knowledge base. As is the case with all the other autopoietic function systems of modern society, the societal function of architecture, in the sense of addressing an underlying reference problem, is much older than the differentiated function system itself, which only emerged as differentiated, autopoietic system within the context of modernity. All function systems solve perennial exigencies in new, advanced ways, within the new, increasingly complex context of modern society.

In relation to architecture, we are therefore prompted to ask: What is the societal function of architecture? What is the raison d’être of architecture’s origin and continuing existence as an autopoietic subsystem of society? The answer is that architecture’s unique function is the provision of spaces that frame social communication. The societal function of architecture is thus to order (and re-order) society via the continuous provision and innovation of the built environment as a system of spatial frames.

Spatial framing is a necessary precondition of all social communicative interactions and collaborations. The built environment, as a spatial sorting system, distributes and relates activities so that they can concatenate, and it configures the participants in each activity so as to facilitate the purposes of the interactions. The framing system also allows the participants to first of all find one another, to recognise the specific social situation, and recognise each other in their roles. Framing is thus itself a form of communication. It is an important type of communication, as it determines a general set of constraining premises for all further communications that take place within the communicated frame. The author has adopted and adapted the concept of framing/frames from the sociologist Erving Goffman’s Frame Analysis[12].

The implementation of this societal function demands two tasks that must be distinguished and correlated in the design effort: organisation and articulation. The concept of order proposed here—encompassing both social and architectural order—denotes the result of the combined effort of organisation and articulation. Architectural order—symbiotic with social order—requires both spatial organisation and morphological articulation. While organisation establishes objective spatial relations by means of distancing (proximity relations) as well as by means of physically separating and connecting areas of space, articulation operates via the involvement of the user’s perception and comprehension of their designed environment. Articulation reflects the phenomenological and the semiological dimensions of architecture. Thus, to the extent that architecture operates through articulation (rather than mere organisation), it also relies on engendering an effective semiosis within the built environment. It is one of the fundamental claims of the theory of architectural autopoiesis that the semiological dimension of architecture is of central importance to architecture’s capacity to successfully discharge its unique societal function.

Inasmuch as architecture is inhabited by culturally socialised subjects, the ordering effects of architecture rely on effective signification. The effective social utilisation of complex institutional spaces cannot be achieved purely by means of the physical channelling of human bodies. The effectiveness of the spatial order relies upon the active orientation of the subjects, on the basis of a “reading” of the territory. This, in turn, requires articulation over and above physical organisation. Current forms of differentiated office landscapes may serve as an example: The traditional physical demarcation of territory by means of walls is replaced by the subtle coding of zones and the articulation of legible thresholds. This means that the importance of the semiotic dimension of architecture increases.

To grasp the problem of communication and interaction on a deeper level, one might go to Talcott Parsons and his attempt to formulate a general theory of action. When Parsons theorises interaction—i.e., when the object towards which an actor orients their action is another actor—a fundamental theoretical problem is encountered, a “problem” that is nearly always already solved in everyday life. Parsons theorised the underlying problematic under the chapter heading “Interaction and the Complementarity of Expectations”[13]. Parsons describes the basic constellation of interaction between ego and alter—the actors that are oriented to each other—as follows:

“There is a double contingency inherent in interaction. On the one hand, ego’s gratifications are contingent on his selection among available alternatives. But in turn, alter’s reaction will be contingent on ego’s selection and will result from a complementary selection on alter’s part. Because of this double contingency, communication […] could not exist without […] stability of meaning which can only be assured by ‘conventions’ observed by both parties.”[14]

Parsons concludes that double contingency requires the normative orientation of action and poses the norms of a shared symbolic system as the means by which the problem of double contingency in interaction is solved. Parsons elaborates:

“A shared symbolic system is a system of ‘ways of orienting’ plus those ‘external symbols’ which control these ways of orienting, the system being so geared into the action systems of both ego and alter that the external symbols bring forth the same complementary pattern of orientation in both of them. Such a system, with its mutuality of normative orientation, is logically the most elementary form of culture.”[15]

The theory of architectural autopoiesis proposes that architecture is a fundamental and indispensable part of such culture, and, in particular, that architecture operates and contributes to the coordination of “ways of orienting” as part of what Parsons refers to as “external symbols.” The designed built environment thus acts as an anchor or frame that facilitates determination, that is, the definition of the situation, the termination of the indeterminacy and volatility implied in the inherent double contingency of every encounter. Over and above his identification of the problematic of double contingency, it is Parsons’ reference to “shared symbolic systems” that makes his work relevant to the attainment of a theoretical formula for architecture’s societal function. Luhmann picked up this notion of double contingency and made it a central problematic in his social systems theory[16]. The formula for architecture’s societal function proposed in The Autopoiesis of Architecture[17] therefore posits architectural framing as a key contribution to solving the problem of double contingency by predefining the social situation.

Above it was stated that the core architectural task of framing communicative interactions can be broken down into two related sub-tasks: namely, spatial organisation (spatiology) and morphological articulation. Articulation, in turn, can be broken down into the sub-tasks of phenomenological articulation and semiological articulation. In each of these three dimensions—spatiology, phenomenology, and semiology—there has been some research-based upgrading of the discipline’s and profession’s competency. The upgrades in terms of the organisational project started with the work of Christopher Alexander and were pushed forward with big strides by Bill Hillier with his ‘space syntax’ as configurational science, operationalised with computational analytical tools. The phenomenological project had been advanced by Kevin Lynch, Christian Norberg-Schulz, Kenneth Frampton, Colin Rowe, Peter Eisenman, and Jeff Kipnis. The semiological project has been advanced by Charles Jencks, George Baird, Geoffrey Broadbent, Umberto Eco, Mario Gandelsonas, and Peter Eisenman, among others.

All these efforts to advance the discipline’s competency have been left behind for many years. Not even space syntax, the most straightforward and most developed of these competencies, has been spreading, nor has it entered the architectural university curriculum. Indeed, the absence of any shared curriculum within architectural education is one of the symptoms and factors of architecture’s dissolution. Only the author has picked up these three discourses (spatiology, phenomenology, semiology), updating, integrating, and operationalising them in the context of a cumulatively advancing parametricism (tectonism), for instance, via the research project of agent-based parametric semiology[18]. However, these efforts take place in splendid isolation. The author is speaking into the void left behind by the disappearance of the discipline.

Architectural Semiology Operationalized

Architecture’s social functionality includes its communicative capacity. The built environment orders social processes through its pattern of spatial distinctions and connections that in turn facilitate a desired pattern of social events. The functioning of the desired social interaction scenarios depends on the participants’ successful orientation and navigation within the designed environment. The built environment, with its complex matrix of territorial distinctions, is a navigable, information-rich interface of communication. To order and articulate this interface is the core competency of architecture. This core competency accounts for users as sentient, socialised actors who use the built environment as an orienting matrix and a set of instructions within which myriads of nuanced social protocols are inscribed.

All design is communication. Before a specific interaction event can commence, relevant participants must find each other, gather, and configure into a constellation germane to the desired interaction scenario. Their respective expectations, moods, and modes of behaviour must be mutually complementary—they must share a common definition of the situation. It is thus the spatially predefined situation that brings all actors into a conducive position, with their respective complementary social roles. The built environment thus delivers a necessary precondition of determinate social interaction.
For this to succeed, the built environment must be legible. The participant can then respond to the spatial communication broadcasted by the designed space, e.g., by entering a space and joining the accommodated social situation. As a communicative frame, a designed space is itself a communication that provides the premise for all communications taking place within its boundaries.

The designed spaces deliver the necessary predefinition of the respectively designated social situation, thereby reducing the otherwise unmanageable excess of action possibilities that exist in our complex contemporary societies. They “frame” social interaction. The organisation and articulation of these framing spatial communications is architecture’s core competency. The social meaning of a space can usually be inferred from its location, shape and stylistic markers. The research programme of architectural semiology aims to analyse the active semiological codes that already operate within the built environment via spontaneous semiosis. There is also a design ambition to upgrade the communicative power of the built environment, project by project, through the design of information-rich systems of signification that aid navigation via wayfinding systems and facilitate interaction through the differentiation and nuanced spatio-visual characterisation of interaction offerings.

The success of such an endeavour depends on user uptake. This can be expected in large, complex integrated social environments, such as a university or a creative industry corporate campus, where life is communication-intensive, orientation is non-trivial, and where inter-awareness, knowledge transfer, and ramifying collaborations put a premium on social participation. There is thus motivation to pay attention to the clues and learn the spatio-visual language. The question arises: how might the communicative performance of large, complex designed environments be evaluated? The research project “Agent-based Parametric Semiology” builds, investigates, and applies a new form of occupancy simulation as an answer to this question.

While every architect has an intuitive grasp of the normative interaction protocols that attach to the various designated areas that the design brief indicates and usually knows enough about the expected and desired user occupancy patterns, such intuitions cannot provide secure guidance on the relative social performance of alternative designs for large, complex environments. Intuition must here be substituted by occupancy simulations that can process thousands of agents interacting across an environment of hundreds of spaces. When quantitative comparisons and optimisation are aimed at, intuition fails, already in much smaller, simpler settings.

The simulation methodology developed under the research agenda “Agent-based Parametric Semiology” is conceived as a generalisation and corresponding upgrade of the crowd simulations currently offered by traffic and engineering consultants concerned with evacuation or circulation. These crowd modellers treat users as physical bodies and simulate crowds like a physical fluid. In contrast, the architectural design considerations of agent-based parametric semiology are concerned with socialised actors who orient and interact within a semantically differentiated environment.

These research and upgrading efforts are pointed out here not to claim that architecture comes to an end because these particular avenues of theory-led, research-based capacity development are not being pursued. These particular research efforts are meant to exemplify what this essay means by theory-led adaptive upgrading of the discipline. Another example could be the upgrading efforts spearheaded by the paradigm/style/movement the author named “tectonism,” namely the full architectural, spatio-morphological utilisation of recent, computationally empowered advances in engineering science and fabrication/construction technologies.

We witness the end of architecture not because the author’s upgrading efforts are not being picked up but because no capacity development whatsoever, with respect to architecture’s core competency and societal function, is being pursued or adopted.

The engagement with carbon neutrality, biodiversity, social justice, and inclusion at the margins of society are no substitute for advancing architecture’s contribution at the frontier of contemporary civilisation’s development. Rather, the usurpation of architecture’s internal and external communication space and air-time by these tangential topics—which only concern the costs but not the benefits of urban and architectural development—is a big part of the problem, a major factor in architecture’s disappearance, and certainly no remedy. Even an agenda like wellbeing, which seems to fit into architecture’s societal function, in fact offers just another evasion. Is it, as Reinier de Graaf argues in his recent book architect, verb.: The New Language of Building [19], part and parcel of the recently proliferating arsenal of hypocritical, self-alienating but obligatory phrases he calls “profspeak”? De Graaf coined the term “profspeak” in allusion to Orwell’s notion of “newspeak,” implying vague, euphemistic phrases that sound benign and competent but gloss over anything potentially controversial or difficult. Talking about wellbeing fits this bill and does indeed allow architects to communicate safely with their audience, in ways that allow them to evade that task of innovating within the key dimensions of architecture’s societal function (spatiology, phenomenology, semiology), while also allowing them to avoid exposing their actual searching ideas and half-articulate ambitions.

In the Orwellian era of architects’ “Profspeak,” an arsenal of conveniently indisputable do-good agendas, like sustainability, community engagement, inclusion, liveability and human-centric placemaking, swamps architectural discourse. They sanitise and narrow the discourse by crowding out all difficult and controversial questions. Reinier de Graaf is right in his scathing critique of Profspeak. It binds precious attention unproductively and thereby contributes to the end of architecture.

While academics, educators, and critics altogether turn away from contemporary architecture, contemporary architects trivialise their work by means of the bland, routine euphemisms of Profspeak. The result is the degradation of the discipline of architecture to the status of a craft or vernacular, producing mere buildings, but no works of architecture.

The Politicization of Architecture from Within

Political and moral issues are increasingly being drawn into our debates at architectural conferences, schools, and biennials. Political and moral issues are also starting to dominate architectural criticism as well as the awarding of architecture prizes. This is problematic, as it threatens to swamp our discourse, overburden our specific competency, and distract us from our genuine, specific societal responsibility within the societal division of labour.

Politicisation cannot always be avoided, and is, indeed, not always unproductive. Architecture has been politicised a number of times in the last 100 years. The most prominent examples are the early 1920s and the late 1960s. These were revolutionary periods when all aspects of societal life had been politicised. Politicisation was very productive in 1920s and (to a much lesser extent) also in the revolutionary 1960s. Although a lot of energy and inventiveness was spawned in architecture in the 1960s, lasting innovations were not achieved. The revolution failed; it was a historical dead end. The clearest indication that this revolution and dream of a proletarian world revolution were misguided was its infatuation with Mao’s disastrous “Cultural Revolution.” Although some of the sixties’ cultural transformations were progressive and lasting, politically, the project failed. Architecturally, it failed too, as became clear soon enough in the 1970s, as can be gleaned from Reyner Banham’s ironic 1976 book title Megastructures: Urban Futures of the Recent Past. The real, relevant, productive, and lasting revolution or paradigm shift in[20] architecture happened during the 1980s and 1990s, starting with postmodernism, as the first intuitive attempt to respond to the post-Fordist socio-economic restructuring. This was reinforced by the decisive political reset and victory of neoliberalism in Britain in 1979 (Thatcher) and the in USA in 1980 (Reagan). The initial postmodernist gropings were radicalized by deconstructivism. Both postmodernism and deconstructivism were short-lived, merely transitional styles, paving the way for parametricism (since the mid-1990s) as a sustained architectural answer to post-Fordist network society and thus as viable candidate to become the epochal style for the twenty-first century.

A more recent (wholly counter-productive) wave of politicisation swept through architecture in the years following the 2008 financial crash. While no new real or positive socio-economic/political transformation of society took place, the politicisation of our discipline has festered ever since, with destructive rather than constructive effects.

In 2012, the author published Volume 2 of The Autopoiesis of Architecture and, for the first time, put forward the following thesis on the relationship between architecture and politics:

“To respond to hegemonic political trends is a vital capacity of architecture. It has no capacity to resolve political controversy. Political debate within architecture overburdens the discipline. The autopoiesis of architecture consumes itself in the attempt to substitute itself for the political system.”[21]

According to Luhmann’s theory of modern, functionally differentiated society (adopted and extended here), the relationship between architecture and politics is the relationship between two autonomous, self-referentially enclosed systems of communication. Both politics, understood as the system of political communications, and architecture, understood as the system of architectural communications, are functionally specialised social systems. They both belong to the group of the great function systems of society. Each is differentiated on the basis of taking exclusive responsibility for a distinct, necessary societal function. The societal function of the political system is the ordering of social communication via the provision of collectively binding decisions. Architecture’s societal function is the ordering of social communication via the provision of spatial frames.[22]

The thesis of autopoietic, self-referentially enclosed systems of communication entails the recognition of a fundamental incommensurability between the different societal function systems. Each function system sustains its own unique discourse on the basis of its own specialised categories, questions, and types of arguments, each treating all the other function systems (economy, legal system, etc.) as a constraining environment rather than as a contribution to a single, unified discourse.

The widespread conception of architecture as a site of political activism must be repudiated. Architecture is not inherently political. The slogan “everything is political” was born and spread in the late sixties during a general revolutionary period. In 1968, politics was no longer contained within the boundaries of the institutionalized political system. Generally, during a revolutionary period, all aspects of social life do indeed become political. Nearly everybody becomes involved in a multifaceted political discourse and struggle that questions all institutions, communication structures, and modes of interaction. However, revolutionary periods are exceptional ones and cannot continue for very long. They subside or escalate into a full-blown revolution. In any event, society must eventually move back to a situation where politics is contained within a separate political system that works through some but not all aspects of social life. The revolutionary period of the late sixties peaked in 1968 and receded in the following years. However, the slogan “everything is political” continued to circulate in intellectual circles that found it difficult to accept that the intoxicating ferment of the revolutionary situation had, in fact, vanished. (Revolutionary periods are the intellectuals’ great moments of exceptional influence and power.) The slogan lives on but runs idle without any real meaning within architecture or anywhere else in society.

The political radicalism of the late 1960s shook up all aspects of modern society, including architecture. Since then, however, no further exciting and plausible left political projects have emerged. Yet, ever since, the echoes of this shockwave return to inspire, haunt, and embarrass the discipline. Calls for a “political architecture” are raised again and again. An example might be Tahl Kaminer’s The Efficacy of Architecture: Political Contestation and Agency[23]. The back cover reads as follows:

“Originating in a displeasure with the ‘starchitecture’ system and the focus on aesthetic innovation, a growing number of architects, emboldened by the 2007–8 economic crisis, have staged a rebellion against the dominant mode of architectural production. Against a ‘disinterested’ position emulating high art, they have advocated political engagement, citizen participation and the right to the city. … At the centre of this rebellion is the call for architecture to (re-)assume its social and political role in society. The Efficacy of Architecture supports the return of architecture to politics.”[24]

The identification of the 2008 financial crash and the ensuing great recession is correctly identified as a turning point or trigger for the recent take-off of the politicisation of the discipline. However, the anti-capitalist outlook of this politicisation – endorsed and shared by Kaminer – implies a stance of refusal, as it hopelessly stands against the grain of history. The economic and political bankruptcy of socialism spells intellectual disorientation with respect to anything beyond resistance or small local, inherently marginal projects.

These calls for a political architecture’ characteristically fail to specify the desired politics with sufficient precision. Instead of offering a political position and programme, the respective authors are lamenting the lack of a vigorous political dimension within architectural discourse. This nostalgic lament is usually expressed via a series of vague phrases that serve as little more than non-committal gestures towards a vaguely progressive politics. They mark the absence of real politics rather than vigorous engagement. For instance, according to the Berlage Institute’s theory teacher Roemer van Toorn, architecture should “project alternatives,” offer “critical resistance” and “social directionality.” Architects are to look for “radical democracies” and “aim at a systematic understanding of architecture as a political palimpsest for alternative social and political hypothesis while itself reanimating architecture as an instrument of social and political invention”[25]. The missing ingredient is the plausible, concrete, generalisable political project, backed up with sufficient political power. Architectural critics and academics can never themselves acquire and project political power. A political project that could give political meaning to architecture cannot originate within architecture itself (nor can it originate in mere political theory). Architecture itself cannot offer effective political direction, or project political alternatives. Here, powerful external stimulation is required—powerful in the most literal sense. A second precondition for a politically engaged architecture is the clear alignment with such a powerful political position. Vague anti-capitalist allusions obviously do not suffice to get this off the ground. Such phrases merely paper over the underlying political vacuum and disorientation. They stand in for the missing political dynamic. They symbolise the desire to be energised by a political position without risking the embarrassment of real political alignment. To offer an isolated, marginal political opinion, or an academic analysis from the domain of political theory does not help to inject political vitality into architecture. Architecture can only react with sufficient unanimity and collective vitality to political agendas that have already the real power of a tangible political force behind them.

The key thesis that must be emphasised here is that it is not architecture’s societal function to actively promote or initiate political agendas that are not already thriving in the political arena (backed by political groups with a real chance to take power). Architecture is not a viable site for such initiatives. It cannot substitute itself for a missing political agenda.

The paradigmatic examples from the early 1920s and the late 1960s that give meaning to the notion of politically engaged architecture were born in the exceptional condition of political revolution (or pending political revolution). During such periods, everything is politicised: the law, the economy, education, architecture, and to some extent even science and technology. The autonomy of the functional subsystems of society is temporarily being suspended. It is during such a period that Le Corbusier’s famous 1920s dictum “architecture or revolution” was coined: “It is a question of building which is at the root of the social unrest of today; architecture or revolution.”[26] This kind of revolutionary condition can reoccur, but it is not the normal state of affairs. During normal times, the specialised, well-adapted channels of political communication absorb all political concerns and bind or direct all political energies. Art, science, architecture, education, and even the mass media are released from the burden of becoming vehicles of political action. The more this division of labour consolidates, the more false and out of place rings the pretence of “political architecture.” Political architecture finally becomes an oxymoron—at least until the emergence of the next revolutionary situation. During normal times, architecture and politics are separated as autonomous discursive/professional domains. If architecture gives itself over to political debate, which is inherently interminable within architecture, then this spells the end of architecture, just as the political usurpation of architectural autonomy would spell the termination of architecture.

We must repudiate the false pretence of “political” or “critical” architecture. Instead, we must act within architecture’s own specific competency. A constructive and effective critique of architecture within architectural discourse can only be architectural critique, on the sui generis terms of the discipline, not political critique. The stance of parametricism is sharply critical of current architectural and urban design outcomes, and the author’s stance is doubly critical, as it is also critical of many of the shortcomings of “real existing” parametricism. However, the author’s stance as architectural researcher and practitioner (as well as parametricism’s stance in general) is implicitly affirmative with respect to the general societal (social, economic, and political) trends that underlie the criticised current architectural and urban outcomes. This implicit affirmation of the legitimacy of the given societal order is a necessary condition of a constructive professional engagement with the architecture of society. Those who see the political system as the bottleneck for architecture’s (and society’s) progress and who feel that current socio-economic and political conditions are to be fought and overthrown, and who are therefore unwilling to fulfil architecture’s institutionally allocated role, should consequently exit the discipline and shift their activism into the political arena proper. They need to test and win their political arguments within and against political groups rather than within architecture. The currently fashionable concept of a “critical” or “political” architecture as a supposed form of political activism must be repudiated as an implausible phantom.

It is undeniable that political and moral issues are increasingly being drawn into our debates and that this threatens to swamp our discourse, overburden our specific competency, and distract us from our genuine societal responsibility. However, there is another twist in the author’s more recent relationship to politics within architecture: Architecture’s politicisation has reached a pervasiveness and intensification that can no longer be ignored, contained, or rolled back merely via meta-arguments about architecture’s proper domain of competency. The current historical conjuncture makes a head-on substantive political engagement with those who politicise architecture from an anti-capitalist position more and more urgent. The author concludes that political engagement can no longer be avoided. The conclusion is now to engage in this politicising debate with a double strategy:

I. To define the proper relationship between architecture and politics in order to set out the premises and the scope for a viable and productive architectural engagement with politics, argued for from within the framework of a comprehensive theory of society (social systems theory). This entails the task to define and defend a space for an autonomous architectural expert discourse and theory-led architectural design research—the autopoiesis of architecture which co-evolves with rather than being instructed by politics—and the repudiation of “political architecture,” which attempts to pursue architectural design as an activist-critical political practice.

II. To engage in the current politicising architectural debate and repudiate what must be considered a regressive and unproductive (explicit or implicit) anti-capitalist bias in most of the political and moral positions drawn into architecture by architectural academics and critics, and to confront these critics with a vigorous defence of capitalism, pointing to the renewed advancement of the forces of production and the (nearly world-wide) prosperity boost its recent neoliberal reinvigoration (1980–2008) made possible.

Factors contributing to architecture’s dissolution

Architecture as a discipline and profession had long since evolved its specialist competency and responsibility within functionally differentiated modern society, entrusted with the societal function of the continuous, innovative upgrading of the built environment, in line with general societal and technological progress. However, in recent years, the protagonists and organs of the discipline have been refusing to focus their discourse on this societal responsibility and have become increasingly incompetent with respect to architecture’s core competency and specialised societal responsibility. Instead, architecture’s leading voices—architects, theoreticians, critics, professors, curators, etc.—are shifting their attention to general “do-good” themes like social justice, or marginalised communities, i.e., domains where architecture as a discipline has no decision-making powers and next to nothing to offer out of its own resources. What we are left with, therefore, is impotent virtue signalling.

Beyond the refusal of the theoretically-minded, politicised protagonists of academic architecture to own up to architecture’s task of innovation—which can only be credibly pursued at the frontier of urban development in the most advanced centres of world society—architecture, in its mainstream practitioners, remains fragmented and self-indulgent, unwilling to absorb the innovations that have been achieved by the small and isolated avant-garde movement the author has identified and theorised under headings like parametricism and (more recently) tectonism. These two phenomena, the academics’ refusal to engage with and lead contemporary architecture on the one hand, and the fragmentation and disorientation of mainstream practitioners on the other hand, are related. They are two interacting aspects of the same process of degeneration.

However, there is a further, independent factor: The incentive of mainstream practitioners to absorb the innovations (that have been worked out by a small, isolated avant-garde) is being killed off by politics via a highly restrictive/prescriptive planning permission practice. (Behind closed doors developers speak about “planning paralysis.”) Urban and architectural progress in the mature, advanced centres is blocked by massive political over-regulation, indeed by political prevention of urban development. Where development is still permitted, it is stifled by prescriptive impositions, squashing architectural innovation. Developers and their architects cannot freely compete with innovative solutions and urban service offerings but are just called upon to execute politically preconceived plans. There is, therefore, no need or incentive to innovate. For developers, competition shifts to and remains confined to negotiations about constraints like the exact percentage of affordable housing. For architects, the competition shifts to the plane of political rhetoric. Banal, politically imposed solutions, are being sold back to the politicians via empty, euphemistic slogans like people-centred design, community engagement, wellbeing, inclusive placemaking, etc.

Because of this refusal or self-denial of architecture, society evolves without its evolution being accompanied by congenial or adequate architectural responses. The bulk of architecture designed in 2024 could have been designed in 1974 or indeed in 1924. It is not only stagnant but positively regressive. All styles, with the exception of parametricism (with Tectonism as its most recent and most advanced and sophisticated subsidiary style), are retro-styles: Minimalism, Neo-modernism, Neo-rationalism, Neo-classicism, Neo-historicism, Neo-postmodernism.

The intellectual atrophy within the discipline is by now so pervasive that those serious and sophisticated contributions that have been developed in a tiny but vigorously advancing network of researchers and designers can no longer even be more widely appreciated or absorbed. They are stillborn with respect to their ambition to move the discipline forward. Architecture, formerly an academic, discourse-steered discipline and innovative, theory-led and research-based profession, has contracted back into a craft, uncritically and unambitiously subjecting itself to pre-ordained routines and typologies. In effect, the whole apparatus of the academic discipline—architectural university departments, theoretical journals, conferences, biennals, etc.—might as well be shut down. What is their use if hundred-year-old recipes are the latest wisdom of the profession or craft? What is society getting in return for financing this massive and costly apparatus? In any event, this apparatus, in its university incarnation, is distracting itself with all manner of woke studies, woke criticism, and woke polemical, artistic-symbolic illustrations standing in for the absent design projects. It is certainly no longer engaging with the task of discursively steering and innovating the built environment that is actually being realised.

Another factor contributing to the disappearance of architecture is a detrimental tolerance that destroys all learning in schools of architecture. Here, rigorous debate has been hollowed out in recent years. While criticising society is all the rage, criticising student work is increasingly avoided, seen as disrespectful and regarded as a feature of a now outmoded toxic culture. One underlying factor is identity politics. If ideas are understood to be tied up with identity, then criticism is perceived as attack on those holding these ideas. The response is then indignation, and often explicit ad hominem attacks on the critic, rather than argument. The result is a dysfunctional bifurcation into an unforgiving de-platforming of unrelenting critics and an all-forgiving tolerance of all who are communicating within and according to the rules of the “safe space.” This logic violates a key principle of discursive rationality, namely, that ideas are to be appraised irrespective of their bearers.

The inhibition of frank critique is thus not only a matter of over-politeness or over-protection (confused with respect) but also a matter of historically motivated (but ultimately counterproductive) “postmodern” philosophical assumptions. At the heart of these assumptions lies a defeatist relativism that considers the human condition—in terms of circumstances, worldviews, values, and aspirations—as inherently fragmented, without any hope of discursive convergence. This theory contrasts with the factual universality of the aspiration for higher standards of living and individual liberty, evidenced by global migration pressures into countries where this universal desire is met better than in the migrants’ countries of origin.

The historical experience that global modernisation-for-all is non-trivial and a much more fragile, complex, and uncertain endeavour than initially expected by mid-twentieth century modernisation theories (including Marxism) lies at the heart of the postmodernist “incredulity towards grand narratives.” Poststructuralist philosophy injected some necessary loops of reflection into social theory, in particular the reflection on historically and culturally specific discursive formations. These reflections were later absorbed into more complex, subtle, and circumspect social theories and theories of societal progress, such as Niklas Luhmann’s “social systems theory”[27] and “theory of functionally differentiated society”[28] or Jürgen Habermas’ “theory of communicative action”[29] and “discourse theory of law and democracy”[30]. These efforts, while engaging with and dialectically integrating poststructuralism’s challenges, avoid relativism and re-establish “grand theory” on a new level of complexity. Simultaneously, the trajectory of postmodernism’s own discourse mutated into a relativist, defeatist and indeed self-defeating intellectual culture that lacks the confidence to judge, project or steer societal developments. The poststructuralists failed to discriminate and judge the diverse discursive formations they charted and did not recognise the superiority (superior prosperity potential) of the unique lineage of discursive formations postmodernism itself was a part of—namely Modernity, with its unprecedented elaboration of technology and science, including critical social science. They failed to recognise the unique self-transcending thrust of this (lineage of) discursive formation(s) that actively refused and continues to refuse to remain tied to any historical origin, parochial social group, or particular set of societies. While German social philosophy—Habermas, Luhmann, and others—has moved dialectically from the modernist thesis via the poststructuralist antithesis to a new synthesis that recuperates the concept of progress on a new level of complexity, French poststructuralism (postmodernism)—and under its influence much of Anglo-American mainstream academic culture—got stuck with the antithesis to modernisation theory. Thus, unassailable “subjective validity” has replaced the regulative concept of objective (intersubjective) validity.

In architecture the impact of poststructuralist philosophy coincided with the crisis and breakdown of architectural modernism. Indeed, the phrase “Postmodernity” was first coined in architecture (by Charles Jencks) in 1976, and was soon generalised to art, literature and philosophy.

The crisis of architectural modernism was not due to inherent flaws of modernist architecture. Rather, modernism was a very meaningful and successful response of the discipline to the historical transformation from laissez-faire capitalism, based on many relatively small companies competing in each industry, to large-scale assembly-line production conducted by a few very large companies in each branch of industry. The result was the new era of Fordism. This technological and economic transformation, after the First World War, was also accompanied by political and social revolutions. Modernism—International Style architecture—was a well-adapted response. That is why it spread throughout the industrial world. However, by the 1970s, a new technological and socio-economic transformation was underway: the transformation from Fordism to post-Fordism. This new societal dynamic could no longer be contained within the strictures of modernist urbanism and architecture and was starting to break out. The well-settled, mature paradigm of modernism was in crisis. A period of search and experimentation ensued, a collective brainstorming. This was a revolutionary period. As in the previous transformation from historicism to modernism, art and philosophy were drawn into the discipline. Rationality was suspended, schools of architecture mutated into art schools, curricula were abandoned, and a new cast of characters—bold, intuitive—appeared on the scene. Rational analytic design discourses that could select from a prior, methodically elaborated solution repertoire, gave way to open-ended brainstorming and generative “artistic” processes of option proliferation, via mutation and recombination. Brainstorming tasked with generating new ideas can only work if the strictures of immediate rational scrutiny are suspended. The whole point of the crisis is that the given criteria of scrutiny and selection are no longer valid. The search is also a search for new values and goals. This atmosphere transformed the discursive culture of the discipline. This transformation was indeed necessary. Adhocism, postmodernist collage, neo-historicism, deconstructivism, etc., emerged from this freewheeling collective brainstorming process as new potential directions during the 1970s and 1980s. Some of these potentials were indeed built upon in the following period. By the early 1990s, a new paradigm started to crystallise out of the experiments loosely gathered under the label “deconstructivism.” The new paradigm built on deconstructivism dialectically, by polemically contrasting its own approach, and drawing on the philosophy of Deleuze & Guattari. A whole generation of young architects studying or teaching in Anglo-Saxon elite universities (London, New York, Boston, L.A.) rapidly converged around a decisive new outlook, not unlike a generation of young architects in German-speaking Europe had converged in the 1920s, leaving Art Nouveau and expressionism behind. As towards the end of the 1920s, towards the end of the 1990s, when the new paradigm had been firmly established as hegemonic within the avant-garde segment of the discipline, art and philosophy receded, and the cumulative development work took over. The 1990s and the first decade of the twenty-first century achieved a viable new paradigm, the paradigm which the author has since 2007 named and canonised as parametricism. The paradigm was spurned by the real estate boom leading up to the 2008 real estate and financial crash. While the leading protagonists of parametricism kept moving forward with upgrading the scope and effectiveness of the paradigm, the expansion of the movement was halted by the crash. The whole discipline was shaken up and became politicised and more susceptible to anti-capitalist sentiments. The culture of the schools of architecture was still in the mood and mode of the 1980s revolutionary period of brainstorming. Instead of re-adjusting to the new requirements of working through the implications of the new paradigm during the 2000s, the freewheeling model that had been spearheaded at the Architectural Association School of Architecture (AA) in London during the 1970s and 1980s proliferated throughout the Anglo-Saxon world and beyond, just at a time when reverting back to a more systematic, science-like, cumulative working through of the newly discovered solution-potentials would have been much more productive. Instead, art-school-like brainstorming and freewheeling experimentation continued alongside the disciplined, collective and cumulative work of the movement of parametricism. But brainstorming makes no sense if it continues indefinitely, instead of shifting to analysis, selection, and elaboration. While the Bauhaus had been able to shift towards disciplined, cumulative elaboration of the (modernist) paradigm in the latter half of the 1920s, leaving the mystics and artists behind, the AA, as well as Columbia and Sci-Arc, partially continued to free-float and then reverted back in full force to brainstorming, philosophy, art, and politics after 2008, leaving the agenda of a disciplined architectural research increasingly behind. The culture became one where every teaching studio or “unit” is wholly autonomous, operating outside any curriculum and beyond the reach of any effective external criticism. Non-judgmentalism and the repudiation of any shared criteria or agenda of convergence were again the order of the day, just like during the 1980s. This culture of hyper-tolerant “everything-goes discourse” - as long as the language of political correctness is being observed—spread throughout the discipline. This freewheeling culture took over once more, long after it had outlived its purpose. There was no new paradigm to be discovered because there was no new socio-economic transformation to adapt to. (Post-Fordist restructuring was not exhausted and still is not obsolete today.) In any event, the spirit since then was, and is, a spirit of counterculture, of refusal, of protest, of symbolic resistance, a culture of the indiscriminate celebration of otherness and diversity, a spirit of non-judgementalism. This non-judgementalism lets everything pass, as long as the premise of the apodictic condemnation of contemporary neo-liberal society and all its real-world architectural expressions was not questioned.

During architectural debates the author therefore often feels compelled to shift to the meta-level of critiquing this discursive culture as a necessary preface to articulating positions on substantive issues. This is necessary because the author’s quest—namely to ascertain the most promising direction architecture can take to contribute to prosperity and societal progress—is discredited and anachronistic within contemporary architectural discursive culture. Here are the meta-theses that are necessary to reset the discipline’s discourse culture:

1. Imperative of convergence: the discipline must strive to define a shared paradigm as the (best) way forward. A shared paradigm is a precondition of cooperative, cumulative progress towards a global best practice. A coherent paradigm/goal is required so that simultaneous or sequential designs do not subvert each other and do not undermine the functional integrity of the built environment.

2. Rejection of pluralism: We must accept paradigm pluralism only as temporary historical condition during periods of paradigm shift. Divergences are dialectically productive only if the aim is to resolve and overcome them. We must reject the fatalistic acceptance of a supposedly unresolvable paradigm pluralism in architecture (just as we must reject the related, more general multi-culturalist presumption that all cultures are equally life-enhancing).

3. Benign intolerance: Ruthless criticism is a productive mechanism of convergence. The principle of indiscriminate tolerance makes sense only in a phase of post-crisis brainstorming. If made permanent, this principle denies the comparative evaluation of positions/paradigms and ultimately blocks progress.

The degeneration of the process and purpose of critical discourse is also undermining the important institution of the public crit in architecture schools. Here too, the lack (or denial) of any shared substantive paradigm that could furnish criteria of progress undermines the legitimacy of criticism and judgement. What regulates the crit instead is the principle of indiscriminate, pluralist tolerance. “Crits” no longer aim to critically appraise, debate, judge, and compare the relative validity and worth of projects/proposals, but regress to mere displays of unassailable subjective expressions, soliciting nothing but indiscriminate flattery. Nothing is either weeded out as inferior or marked out as superior. These very notions, and indeed any ranking and selecting, are anathema. But how can progress be made without rejecting failures and selecting successful contributions as exemplars to build upon?

This systemic institutional failure to promote progress does not only stunt the discipline’s development but applies equally to individual students’ learning curves. Worse, nothing stops the retrogression of students (and of whole academic design studios or entire schools) into ever more indefensible pursuits. Where no pushback is ever expected and no defence is ever required, the indefensible mushrooms. Rigorous critique must be reinstated.

Even the most ruthless criticism of a project, proposition, or even cultural tradition/identity, should never be taken as ad hominem attack. No set of ideas (nor any acquired or inherited cultural pattern or identity) represents an immutable characteristic that inherently defines or irredeemably limits any person. To rigorously criticise inferior ideas (or inferior cultures) means to emancipate and empower rather than to disempower the bearer. To politely “respect” ideas (or cultures) one recognises to be dysfunctional is the very opposite of genuine respect.

The Bonfire of Architecture’s self-destruction: Venice Architectural Biennale

The Venice “Architecture” Biennale is mislabelled and should stop laying claim to the title of architecture. This title only generates confusion and disappointment with respect to an event that does not show any contemporary building designs. Assuming Venice to be not only the most important item on the global architectural itinerary, but also representative of the state of architectural discourse in general: What we are witnessing is the discursive self-annihilation of the discipline. The surreal event of an Architecture Biennale without showing any contemporary building designs is the most striking manifestation of this self-annihilation, of the end of architecture. The 2023 Biennale further progressed and radicalised a manifest self-destruction which was already evident in several prior Biennales. The Venice Architecture Biennale is, in effect, exhibiting a historical spectacle: the public execution of the villain that is actual, i.e., “complicit,” architecture.

Again, in 2023, most national pavilions, including all major European nations like Germany, France, Spain, the UK, Belgium, the Netherlands, Norway/Sweden, Finland, but also Japan, Canada, Australia, and the USA, refused to show the work of their architects, or any architecture whatsoever. The German pavilion contained construction trash (from the previous Biennale), and the Czech pavilion seemed closed, with a video screen in front of the closed entrance displaying faces talking about architects’ low income and long hours of work. The author gave up looking for architecture after finding none in 12 out of 12 pavilions visited.

What does this tell us? That there is no noteworthy new designs or buildings in Germany, France, etc., or anywhere in the Western world? Is the design and construction of buildings only an occasion for bad conscience? Is this bad conscience the driving force behind the refusal (by now pervasive for more than a decade) to display any contemporary architecture whatsoever?

The German pavilion, as hinted at above, was filled with piles of construction material retrieved from the demolition of the previous installation. There was no point in spending more than two seconds in there. A single glance and you get the one-liner message (because this message had been reiterated for years): The message is the supposed moral imperative of material recycling. There was also a very similar one-liner message filling the space (and consuming the budget) a few years ago: don’t build, re-use/renovate. In an earlier instance, the German pavilion was filled with documentation of current affairs issues like the refugee crisis. The obvious question of why we should look at documentations of the refugee crisis when coming to Venice for the Biennale, after we have been hearing about the refugee crisis on television every day for months, was apparently never asked. There always seems to be something more important and urgent than showing the most noteworthy designs or buildings being created in Germany. Is there nothing innovative or otherwise noteworthy going on there?

German architecture has been absent in Venice for years. The same applies to British architecture. Why the architects of these countries put up with this seems puzzling. Are they too shamefaced about their work to raise their heads above the parapet? In the case of the German pavilion, the current emphatic absence of architecture was explicitly endorsed by the president of the German Chamber of Architects, in a conversation in front of the German pavilion (filled with the rubble of the previous Biennale). One wonders: what are all these curators expecting an unsuspecting general public coming to visit an architecture biennale to make of this? Are they to witness the disappearance or castigation of a fallen, corrupt, and complicit discipline?

Only the Chinese pavilion showed architecture, plenty of architecture. In the international show, it was again only Chinese architects who showed work: Neru&Hu, and especially Zhang Ke (Standard Architecture), who was showing an impressive suite of projects. The only other exception was the suite of projects by Adjaye Associates, the only sizeable, leading firm invited, probably due to the African origin of its principal. Everybody else invited played along, using their allotted exhibition space for documentary-style intellectual-artistic allusions to moral issues, garnished with pretentious critical-speak, without ever taking the risk of really taking up an explicit position or offering a constructive proposal.

What is the point of all this? Is it meant to inspire conversations? Can we no longer assume that architects and architectural students want to talk about (and see) architecture? Do they now really prefer to learn and talk about decolonising the discipline? Perhaps architectural educators talk about such matters as decolonising the curriculum. Perhaps that is why architectural design has disappeared from most (especially the most prestigious) schools of architecture.

The author has been coming to Venice over and over again, witnessing architects’ reactions to several of these anti-architectural biennales. At least the architects the author knows and came across at the Biennale tend to cling to the few exceptional instances of architecture and talk about those, and then about their frustration with the swamping of the Biennale with virtue signalling and conceptual-symbolic installations. Does this mean that there remains hope that the end of architecture is not yet final?

This show was meant to feature at least 50% architects from Africa (at least originally). Without David Adjaye’s work—which I would suspect is the only display at this Biennale that would fill a visitor from Africa with pride—there would be no African architecture in the show. This exhibit was a treat and an occasion to learn that such sophisticated, world-class buildings now exist on the African continent, a significant fact and signal of African development and aspiration. This display was a lucky exception in the Biennale. We owe this lucky exception to the coincidence that this successful practicing architect is of African descent, a fact which cleared his entry into the Biennale despite his success which would otherwise have disqualified him.

While Western architectural culture (and Western culture in general) seems shamefaced and guilt-ridden, excluding all its urban development from the “Architecture” Biennale, Chinese architectural culture, in positive contrast, was present in full force and self-confidence. Chinese architects and the Chinese national pavilion (including the Hong Kong pavilion) delivered virtually all the architecture (excepting Adjaye Associates) in the whole Biennale. Will architecture end only in the West, while continuing in the East?

No talk about “architecture as an expanded field” can convince us that we are still at an architectural event when the scene is dominated by documentaries, critical art practice, and symbolic installations, while contemporary building designs are nowhere to be seen in 99% of the exhibition space. The engagement with social issues per se is not the problem. Whatever social, political, or moral issues we want to address, the pertinent way to address them at an architecture biennial would be to demonstrate their relevance to architecture via projects that claim to respond to these issues. However, if everything lamentable, or unjust, or any urgent social or political problem in the world is now an urgent, overriding concern for architecture, then this is not only an absurd overreach, unhinged from architecture’s competency, but it spells the very dissolution and disappearance of this discipline.

In academia, in Western schools of architecture, this process has been driven just as far as in the Venice Biennale, namely to the point of total usurpation. Of course, the professional work of “architects” continues, albeit without any support from academia, or without any representation and discussion in any Biennale, be it Venice, or Chicago. The professional work of architects seems to be beyond the pale, either too banal or morally too compromised, to receive a platform in the lofty realm of a critical cultural event. Even professional architects seem to reach this conclusion once they are appointed as curators. They leave their day job, their work and professional competence behind to become dilettante social critics/commentators.

By now the approach of thematising social ills has become the standard, the expected, unassailable, safe, indeed mandatory option. (For the national pavilions it is also easy to organise and cost-effective. Instead of the risky and difficult task of selecting 25 architects, explaining the selection, and deal with them, a single artist can be commissioned (or two to three) to interpret the theme, and be left alone to do so.) For the curators of the national pavilions, this is now the only way to discharge their curatorial burden. It is now not only the most predictable move, it is obligatory. It is a move that squashes the discipline. It is a move that both enacts and publicly displays the end of architecture. There is nothing in sight here that could fulfil the vital function the Venice Architecture Biennale used to fulfil for our discipline. There is a gaping societal vacuum and nothing, no one, to fill it.

The Intellectual Poverty and Creative Bankruptcy of Architecture

Architecture, in the sense of being distinct from mere building, is dead, intellectually and creatively, and has been so for over 10 years. Further, all schools, conferences, biennials, journals, etc., have ended in the sense of having altogether abandoned architecture’s calling. They have become something else, something disconnected from the development of the built environment, something running idle. Urban and building development continues, hemmed in and micro-managed by planning bureaucracies, and without the benefit of a coherent disciplinary discourse. The profession remains fragmented, without even any sense that this is problematic, and without the slightest ambition to overcome this fragmentation through debate and discursive convergence. Instead, a non-committal pluralism of values and styles is celebrated. On the one end of the spectrum, the personal predilections of architects are not to be questioned. (How a cacophony of idiosyncratic “artistic” creations should add up to create a functionally integrated city remains a question that is not even posed, let alone answered.) On the other end of the spectrum, the discipline lacks confidence to lead and defers its decisions to lay-communities and politicians, thus denying the discipline’s expertise and abrogating the discipline’s responsibility.

If architects are no longer informed by a rich, resourceful and cumulative discourse delivering collective learning processes, then they become either (self-indulgent) artists or routine-bound craftsmen executing client instructions or political instructions. However, neither political power holders nor clients understand how their intentions and interests might be most effectively translated into built form. Both types—artists and craftsmen—populate the ranks of the profession. Both types are incapable of fulfilling the societal responsibility of architecture and end up hollowing out the role and standing of the architect. This intellectual vacuity invites and emboldens both clients and politicians to step into the breach. Both types of usurpation spell the end of architecture.

The longer this post-architectural dilettantism continues and spreads, the more precarious the status of the remaining slivers of a vital and ambitious architectural discourse and practice become. They will soon wither altogether. After the inevitable interim loosening of the curriculum during the period of paradigm shift, a vigorous, searching debate over the direction of the discipline—in order to regain relevance in the new historical era—was required. The brainstorming, including the contestations concerning the curriculum, should have been advanced towards new shared conclusions and resolutions, unifying at least a vital critical mass of protagonists. This did happen during the 1990s and early 2000s. However, since 2008/09—since the prior boom had ended in the financial crash, great recession, and European debt crisis, implying the curbing of work opportunities and the re-emergence of anti-capitalism—these cumulative constructive forces have been swamped by the forces of disciplinary dissolution.

In particular, the transformation of architecture schools into art schools and political debating clubs implies an ongoing (and soon irretrievable) loss of disciplinary knowledge and expertise. While the ongoing, self-confident vitality of the discipline would have required the vigorous, collective rebuilding of a shared disciplinary curriculum, the opposite has taken place: the further loosening and indeed utter dissolution of any shared curriculum conception or intention.

This dismal state of the discipline, and the sinking standards (together with the prevailing woke culture) in schools of architecture, attracts a fitting (or rather misfitting) student population, while it repels students with intellectual ambition who are attracted by sophisticated, demanding, intellectually rigorous fields like economics, business administration, history, sociology, jurisprudence, or computer science. While some of these fields of study have also been softened by woke ideology, their core remains vigorous and continues to progress in exciting ways. That architecture could and should be an equally sophisticated, demanding, intellectually rigorous field, with exciting innovation opportunities, might be faintly glimpsed within the oeuvre, writings, and research initiatives of the author. These opportunities have been accumulating during the last 15 years without the necessary take-up within a larger collective endeavour, due to the erosion of the discipline’s capacity to live up to its societal responsibility. However, it seems unlikely that the ongoing intellectual poverty and creative bankruptcy of architecture—while it implies the further accumulation of untapped opportunities—might attract the talent pool required to overcome this bankruptcy. The continued inflow of a lesser talent pool, with lesser human capital, continuously subjected to an increasingly incestuous academic culture of dilettante distraction and pretence, will only further isolate and dry out the remaining strongholds committed to architectural innovation at the frontier. The works, writings, and research initiatives emanating from this surviving sliver no longer find a receptive audience within the discipline. (While successfully studying architecture one can by now get away without any specialized knowledge whatsoever.) In this sense, what remains of architecture no longer finds any resonance or audience. Architecture is dead because the remaining architects work and speak into a void, are closed into an ever-diminishing echo chamber, isolated by an ever more gaping abyss or suffocating vacuum. Architecture shrinks and becomes a mere message in a bottle, adrift in an ocean of ignorance, hoping to be picked up once more by a future generation.

References:

Amin, Ash (ed.) (1995), Post-Fordism - A Reader, Oxford: Wiley-Blackwell.

Banham, Reyner (1976), Megastructures – Urban Futures of the Recent Past, London: Thames & Hudson Ltd.

Carnap, Rudolf (1967), The Logical Construction of the World, trans. Rolf A. George, London: Routledge & Kegan Paul.

Cerini, M. (2023), Racism, activism and climate crisis are on the agenda at the Venice Architecture Biennale, CNN, May 30, 2023. (https://edition.cnn.com/2023/05/30/style/venice-architectural-biennale-africa/index.)

De Graaf, Reinier (2015), “Architecture is now a tool of capital, complicit in a purpose antithetical to its social mission,” The Architectural Review, 24 April 2015. (https://www.architectural-review.com/essays/architecture-is-now-a-tool-of-capital-complicit-in-a-purpose-antithetical-to-its-social-mission).

de Graaf, Reinier (2023), Architect, Verb – The New Language of Architecture, Verso.

Goffman, Erving (1974), Frame Analysis – A Essay on the Organisation of Experience, New York: Harper & Row.

Habermas, Jürgen (1986), The Theory of Communicative Action Vol.1: Reason and the Rationalization of Society, trans. Thomas McCarthy, Cambridge: Polity Press.

Habermas, Jürgen (1987), The Theory of Communicative Action Vol.2: The Critique of Functionalist Reason, Thomas McCarthy, Cambridge: Polity Press.

Habermas, Jürgen (1996), Between Norms and Facts: Contributions to a Discourse Theory of Law and Democracy, trans. William Rehg, Cambridge: Polity Press.

Kaminer, Tahl (2017), The Efficacy of Architecture: Political Contestation and Agency, New York: Taylor & Francis.

Le Corbusier (1986), Towards a New Architecture, New York: Dover Publications, reprint. Originally published: J. Rodger, London 1931.

Luhmann, Niklas (1995), Social Systems, trans. John Bednarz Jr. with Dirk Baecker, Stanford: Stanford University Press.

Luhmann, Niklas (2012), Theory of Society – Volume 1, trans. Rhodes Barrett, Stanford: Stanford University Press.

Luhmann, Niklas (2013), Theory of Society – Volume 2, trans. Rhodes Barrett, Stanford: Stanford University Press.

Marx, Karl, Friedrich Engels (1998), The Communist Manifesto, Rendlesham, UK: Merlin Press, originally published in 1848.

Parsons, Talcott (1953), Toward a General Theory of Action, Harvard: Harvard University Press.

Schumacher, Patrik (2010), The Autopoiesis of Architecture, Vol.1: A New Framework for Architecture, London: Wiley.

Schumacher, Patrik (2012), The Autopoiesis of Architecture, Vol.2: A New Agenda for Architecture, London: Wiley.

Schumacher, Patrik (2016), “Advancing Social Functionality via Agent Based Parametric Semiology,” in Helen Castle, Patrik Schumacher (eds.), AD Parametricism 2.0 – Rethinking Architecture’s Agenda for the 21st Century, , Academy Press, pp. 108–113.

van Toorn, Roemer (2002), Hunch 5, Amsterdam: Berlage Institute.

[1] P. Schumacher, The Autopoiesis of Architecture, Vol.1: A New Framework for Architecture, Wiley, London, 2010; P. Schumacher, The Autopoiesis of Architecture, Vol.2: A New Agenda for Architecture, Wiley, London, 2012.

[2] M. Cerini, “Racism, Activism and Climate Crisis Are on the Agenda at the Venice Architecture Biennale,” https://edition.cnn.com/2023/05/30/style/venice-architectural-biennale-africa/index.html (accessed 20 December 2024).

[3] P. Schumacher, The Autopoiesis of Architecture, Vol.1: A New Framework for Architecture, Wiley, p. 364.

[4] N. Luhmann, Social Systems, trans. John Bednarz Jr. with Dirk Baecker, Stanford University Press, Stanford, 1995.

[5] N. Luhmann, Theory of Society - Volume1, trans. Rhodes Barrett, University Press, Stanford, 2012; N. Luhmann, Theory of Society - Volume2, trans. Rhodes Barrett, Stanford University Press, Stanford, 2013.

[6] A. Amin (ed.), Post-Fordism - A Reader, Wiley-Blackwell, Oxford, 1995.

[7] R. de Graaf, Architect, Verb – The New Language of Architecture, Verso, 2023.

[8] N. Luhmann, Social Systems, p. 9.

[9] P. Schumacher, The Autopoiesis of Architecture, Vol.1: A New Framework for Architecture; P. Schumacher, The Autopoiesis of Architecture, Vol.2: A New Agenda for Architecture.

[10] K. Marx, F. Engels, The Communist Manifesto, Merlin Press, Rendlesham, UK, 1998, originally published in 1848, p. 4.

[11] N. Luhmann, Social Systems, p. 457.

[12] E. Goffman, Frame Analysis – A Essay on the Organisation of Experience, Harper & Row, New York, 1974.

[13] T. Parsons, Toward a General Theory of Action, Harvard University Press, Harvard, 1953, p. 14.

[14] T. Parsons, Toward a General Theory of Action, p. 16.

[15] Ibid.

[16] N. Luhmann, Social Systems, p. 103.

[17] P. Schumacher, The Autopoiesis of Architecture, Vol.1: A New Framework for Architecture.

[18] P. Schumacher, “Advancing Social Functionality via Agent Based Parametric Semiology,” in H. Castle, P. Schumacher (eds.), AD Parametricism 2.0 – Rethinking Architecture’s Agenda for the 21st Century, Academy Press, 2016, pp. 108–113.

[19] R. de Graaf, Architect, Verb – The New Language of Architecture.

[20] R. Banham, Megastructures – Urban Futures of the Recent Past, Thames & Hudson Ltd., London, 1976.

[21] N. Luhmann, Theory of Society - Volume1, p. 448.

[22] This is the author’s, not Luhmann’s thesis. Luhmann did not recognise that architecture—together with all other design disciplines—constitutes an independent societal function system. He had not given architecture/design much thought and had falsely subsumed it within the art system.

[23] T. Kaminer, The Efficacy of Architecture: Political Contestation and Agency, Taylor & Francis, New York, 2017.

[24] Ibid., p. i.

[25] R. van Toorn, Hunch 5, Berlage Institute, Amsterdam, 2002, pp. 166 f.

[26] Le Corbusier, Towards a New Architecture, Dover Publications, New York, 1986, reprint. Originally published: J.Rodger, London 1931, pp. 269 –289.

[27] N. Luhmann, Social Systems.

[28] N. Luhmann, Theory of Society – Volume 1, trans. Rhodes Barrett, Stanford University Press, Stanford, 2012; N. Luhmann, Theory of Society – Volume 2, trans. Rhodes Barrett, Stanford University Press, Stanford, 2013.

[29] J. Habermas, The Theory of Communicative Action Vol.1: –Reason and the Rationalization of Society, trans. Thomas McCarthy, Polity Press, Cambridge, 1986; J. Habermas, The Theory of Communicative Action Vol.2: – The Critique of Functionalist Reason, Thomas McCarthy, Polity Press, Cambridge, 1987.

[30] J. Habermas, Between Norms and Facts: – Contributions to a Discourse Theory of Law and Democracy, trans. William Rehg, Polity Press, Cambridge, 1996.

Paper delivered at the Academy 2018, annual conference of the Academy of Ideas, July 2018

Jürgen Habermas (b.1929) is a globally recognized philosopher and public intellectual. He made original contributions to critical theoretical sociology, the theory of modern society, as well as moral and political philosophy.

My presentation focusses on a key theme of his political philosophy that relates to the theme of this year’s conference: popular sovereignty.

Habermas has elaborated this theme in his treatise ‘Between Facts and Norms: Contributions to a Discourse Theory of Law and Democracy’[i], first published in German in 1992, and in English in 1996. However, my presentation also draws on Habermas’ earlier works: the ‘Theory of Communicative Action’[ii] and his writings on ‘Discourse Ethics’[iii], as well as directly on Rousseau’s ‘Social Contract’. All quotes from Habermas are taken from ‘Between Facts and Norms’.

This paper concerns Habermas’ account of “the two ideas of human rights and popular sovereignty that have determined the normative self-understanding of constitutional democracies up to the present day.”[iv]

Habermas’ point of departure is that our modern democratic states are relying on two conflicting political conceptions as fundamental basis of their constitution and self-understanding.

The 1^st conception is the idea of a liberal constitution that protects individual liberties in the form of a system of rights, including fundamental and inalienable human rights.

The 2^nd conception is the idea of popular sovereignty, i.e. unconstrained political sovereignty or government by the people or their elected representatives.

The principle of human rights developed in the tradition of classical enlightenment liberalism and was backed up by the idea of natural law in distinction to man-made law. In this conception individual liberties are primary as the a priori basis for entering into society conceived of as social contract.The principle of popular sovereignty is fundamental to the tradition of civic republicanism which sees community and society as primary and connects up with Aristotle’s conception of man as citizen and his idea of the good life as necessarily lived as member of the polis.

Jean-Jacques Rousseau was an early and influential proponent of this conception. Here is a relevant passage from his 1762 treatise ‘The Social Contract’:

“If the State is a moral person whose life is in the union of its members, and if the most important of its cares is the care for its own preservation, it must have a universal and compelling force, in order to move and dispose each part as may be most advantageous to the whole. As nature gives each man absolute power over all his members, the social compact gives the body politic absolute power over all its members also; and it is this power which, under the direction of the general will, bears … the name of Sovereignty. But, besides the public person, we have to consider the private persons composing it, whose life and liberty are naturally independent of it. We are bound then to distinguish clearly between the respective rights of the citizens and the Sovereign, and between the duties the former have to fulfil as subjects, and the natural rights they should enjoy as men. Each man alienates, I admit, by the social compact, only such part of his powers, goods and liberty as it is important for the community to control; but it must also be granted that the Sovereign is sole judge of what is important.”[v]

We take both human rights and popular democratic rule for granted and are therefore mostly not aware of their unresolved tension. Habermas exposes this unresolved theoretical contradiction in the conceptual constitution of our democracies and sets himself the task to elaborate a resolution that can satisfy contemporary post-metaphysical philosophical standards. Habermas approach is via a critical rational reconstruction of the evolved institutional arrangements in order to probe into and excavate their hidden presuppositions.

This is not meant to be a mere academic task as the identified conceptual conflict seems to feed the conflict between broad political tendencies, i.e. conservative right versus progressive left political tendencies. However, this identification is not explicit in Habermas.

Conservative liberalism emphasizes liberty and worries about the majority’s power to suppress individuals and minorities and thus insists on constitutional safeguards of individuals’ rights which must remain outside of the remit of majority rule. According to this conception popular sovereignty and thus the scope of democracy must be strictly limited and should never become comprehensive. According to Habermas’ terminology in this tendency ‘private autonomy’, i.e. individual self-realization, trumps ‘public autonomy’, i.e. the collective self-realisation of the political community.

Civic Republicanism, in contrast, privileges public autonomy and does not recognize any a priori constraint on the expression and realization of the people’s sovereign will.  This is indeed explicit in Rousseau:

“It is against the nature of the body politic for the Sovereign to impose on itself a law which it cannot infringe.  … There neither is nor can be any kind of fundamental law binding on the body of the people — not even the social contract itself.”[vi]

All tendencies aiming to radicalise democracy and all socialist tendencies seem to subscribe to this conception of unlimited popular sovereignty.

The conflict in Habermas’ own words:

“Liberals invoke the danger of a “tyranny of the majority” and postulate the priority of human rights that guarantee the pre-political liberties of the individual and set limits on the sovereign will of the political legislator. The proponents of a civic republicanism, on the contrary, emphasize the intrinsic, non-instrumentalizable value of civic self-organization, so that human rights have a binding character for a political community only as elements of their own consciously appropriated tradition. Whereas on the liberal view human rights all but impose themselves on our moral insight as something given, anchored in a fictive state of nature, according to republicans the ethical-political will of a self-actualizing collectivity is forbidden to recognize anything that does not correspond to its own authentic life project.”[vii]  End of quote. (p. 100).

Habermas’ theoretical project is to show that both competing political conceptions point to indispensable principles  -  namely the principle of inviolable individual rights and the principle of unlimited popular sovereignty  -  that are indeed both presupposed as preconditions of the unquestioned achievement of modern constitutional democracies. They complement and require each other.

We may start the presentation of Habermas’ argument by summarizing his conclusion:

The modern idea of popular sovereignty – the self-legislation of the people -  presupposes people who are free and equal. The collective opinion- and will-formation must be a voluntary process where participation is a matter of individual choice. The idea calls for but cannot coerce political participation. As an inherent precondition of the collective, presumably rational and thus necessarily repression-free deliberation and will-formation process that is to inform the self-legislation of the people, the free individual must be presumed and indeed constitutionally safeguarded. Thus public autonomy requires private autonomy. The inverse presupposition also holds: In order for private autonomy or individual liberties to become a reliable social reality, the individual must be institutionalized as legal person with constitutionally safeguarded rights. To deny this involves a performative contradiction, i.e. a statement that is contradictory not within its semantic content but in relation to its pragmatic presuppositions.

Habermas’ theoretical project is a project of theoretical integration and rational reconstruction demonstrating how the competing modern conceptions of private and public autonomy co-originate “from crumbling edifice of substantial ethical life” and indeed presuppose each other and therefore should not be used to debunk or discredit each other but must be balanced within a theory-led political process.

Habermas: “Thus far no one has succeeded in satisfactorily reconciling private and public autonomy at a fundamental conceptual level, as is evident from the unclarified relation between individual rights and public law in the field of jurisprudence, as well as from the unresolved competition between human rights and popular sovereignty in social-contract theory.”[viii]

According to Habermas this is a task for contemporary philosophy because he identifies the origins of this conceptual problem in the deficiencies and fallacies of prior philosophical approaches in the context of which these conceptions were first formulated and which lack the theoretical resources for their satisfactory resolution.

Although both Kant and Rousseau try to integrate the two principles, according to Habermas they failed to do so. Habermas: “In Kant, as in Rousseau, there still is an unacknowledged competition between morally grounded human rights and the principle of popular sovereignty. … On the whole, Kant suggests more of a liberal reading of political autonomy, Rousseau a republican reading.”[ix]

Habermas describes Kant’s conception as follows:

“Kant obtains the “universal principle of law” by applying the moral principle to “external” relations. He begins his Elements of Justice with the one right owed to each human being “by virtue of his humanity,” that is, the right to equal individual liberties backed by authorized coercion. … This system of natural rights, which “one cannot give up even if one wanted to,” belongs “inalienably” to each human being. It is legitimated, prior to its differentiation in the shape of positive law, on the basis of moral principles, and hence independently of that political autonomy of citizens first constituted only with the social contract. To this extent the principles of private law enjoy the validity of moral rights already in the state of nature; hence “natural rights,” which protect the human being’s private autonomy, precede the will of the sovereign lawgiver. At least in this regard, the sovereignty of the “concurring and united will” of the citizens is constrained by morally grounded human rights.”[x]

Let’s start to engage with the philosophical fallacies of the liberal conception of rights and their problematic political consequences. The most obvious problem is the bankrupt metaphysical notion of natural law which is not only incompatible with our current scientific world view but also attempts to withdraw aspects of the law not only from the political reform process but even from critical discursive probing.

Even those notions of natural law that avoid metaphysical references and rather rely on references to human nature or essence unduly naturalize and frieze what are indeed historical achievements of cultural evolution. Rights and individual autonomy are not found in natural history, nor in any pre-modern society. The one-sided simple insistence on certain rights as self-evident, natural human rights is intellectually bankrupt, and it thus seems that human rights are theoretically defenceless on our contemporary post-metaphysical and post-essentialist intellectual level.

Another deficiency of the liberal conception is the even more widespread and less acknowledged but related fallacy of a simplistic conception of a presumably essentially individual human subjectivity, personality and consciousness.

Liberal social and political theory tries to theoretically model society on the idea of a contract agreed for mutual individual benefit and thus explains society via individuals' calculated, utility maximising appraisal of the advantages of social cooperation.  Habermas objection is not merely that the social contract story is historically false  - it was admittedly a theoretical fiction -  but that it is a fallacious theoretical reconstruction that falters on a naive conception of the individual as cognitively self-sufficient agent with autonomously formed ends and the self-sufficient capacity for means-ends thinking.

Against this stands the insight that what we recognize as  human subjects are always already inherently social and that subjectivities, self-conceptions, including everybody’s conceptions of their own interests and self-realisation, are not only dependent on always already given, communal language resources and an initial socialization but on continued communicative interactions that are crucially required to continuously confirm, reproduce and update our very self-conceptions and orienting understanding of the world and our potential interests within it. All world revealing reasoning and knowledge is inherently intersubjective and based on inherently dialogical and collective ‘communicative reason’ and which cannot be reduced to individual ‘instrumental reason’.

Habermas’ key concept of communicative action and reason in distinction to instrumental and strategic action and reason had been absent from all social sciences based on methodological individualism including political theory, sociology and economics, where action  - the most fundamental base category of the social sciences  - always implies an individuals’ means-ends rationality. In contrast communicative action is oriented towards mutual understanding and presupposes an inherently shared life project. This also becomes Habermas’ anchor for securing solidarity.

Proper communicative interactions are only those where all motives except that of cooperatively and consensually achieving a shared goal together for a shared purpose are suspended. This most crucially entails seeking the truth together. Without positing one's thinking sincerely (non-strategically) one obviously cannot test one's ideas and forgoes the required corrective input. The exposure to criticism engenders learning. Isolated individual thinking is very vulnerable to going astray. Isolated individuals quickly lose all sense of reality. What we (since modernity) call knowledge  - and that includes science as ideal of knowledge production - can only be built on communicative interaction. Since instrumental action and strategic action require knowledge, they are parasitic on communicative action.

Habermas' insight is thus that individual instrumental reason is secondary to and depends communicative reason. This insights defeats a social contract model of society that presumes that society can be built and sustained merely on the basis of individually expected (and negotiated) mutual advantages of cooperation for the sake of prior individuals' interests, which can then become the motivation to seek its safeguarding via rights. This can thus not be the route for the justification of rights.

This argument of the dependence of interests on communication establishes that the system of rights cannot be grounded independently of the community or polity and that the community/polity is a constitutive condition of the very individuals themselves, their interests, and rights.

But how can the necessity of individual liberties be established after all? How can the modern system of rights be grounded in the face of the demands of the polity?

Before I give Habermas’ answer to this question, let me turn to the deficiencies and problems of the classical conception of popular sovereignty as Habermas finds it in Rousseau.

Habermas describes Rousseau’s conception as follows:

“Rousseau imagines the constitution of popular sovereignty through the social contract as a kind of existential act of sociation through which isolated and success-oriented individuals transform themselves into citizens oriented to the common good of an ethical community. As members of a collective body, they fuse together into the macrosubject of a legislative practice that has broken with the particular interests of private persons subjected to laws.

Rousseau takes the excessive ethical demands on the citizen, which are built into the republican concept of community in any case, to an extreme. He counts on political virtues that are anchored in the ethos of a small and perspicuous, more or less homogenous community integrated through shared cultural traditions. The single alternative would be state coercion:  “Now the less the individual wills relate to the general will, that is to say customary conduct to the laws, the more repressive force has to be increased. The Government, then, in order to be good, should be relatively stronger as the people becomes more numerous.” However, if the practice of self-legislation must feed off the ethical substance of a people who already agree in advance on their value orientations, then Rousseau cannot explain how the postulated orientation of the citizens toward the common good can be mediated with the differentiated interest positions of private persons. He thus cannot explain how the normatively construed common will can, without repression, be mediated with the free choice of individuals.”[xi]

Let’s return to our question: How can the modern system of rights be grounded in relation to polity and popular sovereignty and its general will?

Habermas argues that the system of rights, i.e. the system of protected individual liberties, is justified as a necessary enabling condition of the very formation of the general will, i.e. as an indispensable aspect of the constitutive procedures by which popular sovereignty is enacted. This argument via the identification of “conditions of possibility” parallels Habermas discourse ethics, according to which modern rationality, knowledge, science and philosophy depend on modern liberal morality.

The social contract is indeed meant to be rational, formed for the sake of the utility of cooperation that benefits all. Even Rousseau insists – on the opening page of his treatise – that utility remains a guiding principle of his inquiry.

Rousseau: “In this inquiry I shall endeavour always to unite what right sanctions with what is prescribed by interest, in order that justice and utility may in no case be divided.”[xii]

Here is the compelling insight of Habermas discourse ethics:

Modern science, rationality, and knowledge is always a social achievement, always the result of a specifically structured social cooperative process, more specifically it is always based on controversial debate where many independent authors exchange contributions.

This seemingly trivial fact of life becomes an important premise of philosophical self-reflection: ALWAYS remember, ALL knowledge depends on a social process, no knowledge can emerge within an isolated thinking mind.

Philosophy must take this EVERYDAY FACT of its own production very serious:

dialogue, discussion, open controversial debate is a condition of philosophy’s possibility, of science’s possibility, of the possibility of knowledge (in contrast to mere untested opinion).

This is how we judge truth in the modern world:  if you do not allow yourself to be critisized, if you protect yourself from questions that you cannot answer, its clear that you have not reached the truth.

“Truth” is what can withstand scrutiny, can answer challenges.

This is a fact of our discourse practice: Any knowledge claim becomes suspect if the social, communicative process establishing it did significantly deviate from the ideal of a sincere, open, critical, “democratic”, repression free discourse.

Furthermore, the fundamental scientific criterion of universal reproducibility presupposes critical open access to the discursive search for truth. This indicates the requirement of openness to everybody’s independent testing, criticism and refutation.

There is, of course, no guarantee of truth, criticism continues, but we know one thing: that which shields itself from criticism cannot be trusted! And crucially: it cannot and should not trust itself.

The knowledge we mean cannot exist in traditional societies, and in fact did not exist. They could survive, and thus seemed to “know” something, but they did not really KNOW anything ... it was all very uncertain, mixed up with crass superstitions, and evidently far less practically successful than modern scientifically based societies.

This means that modern, enlightened knowledge claims, as well as all practical rationality claims, presuppose modern liberal non-stratified social relations between ‘free’ individuals and exclude relations of bondage or indeed reifications like nobility, papal infallibility, the authority of the bible etc., and must also exclude the ideological domination or censoring of individual dissenters by a politically empowered majority opinion.

Modern knowledge, science, and rationality depend on the modern liberal morality. Their rejection within discourse constitutes a performative contradiction.

Habermas uses this reflective insight to ground this liberal morality, rebutting relativistic philosophical positions that claim all morality to be a matter of mere subjective choice not susceptible to rational argument. Habermas argues: Subjectivist moral philosophy, by joining the discourse and trying to convince the participants in this discourse that the search for a justifiable/true morality is vain, fails to reflect that it thereby is already implicitly committed to a particular morality, namely to the liberal morality underpinning (modern social the institution of) discourse. Habermas’ project of discourse ethics is then to make these moral presuppositions of all discourse explicit, thereby rationally grounding a particular moral system.

Again: Modern knowledge, science, and rationality depend on the modern liberal morality, therefore a rational polity must also rely on the modern liberal morality, and indeed must institutionalize it legally in the form of the liberal system of rights or liberties.

Rousseau had claimed that  “there neither is nor can be any kind of fundamental law binding on the body of the people”[xiii] but now we realize with Habermas that a rationally self-legislating people are always already performatively bound by the presupposition of liberties as an enabling condition of its rational collective self-determination.

Rousseau also claims that “the Sovereign, being formed wholly of the individuals who compose it, neither has nor can have any interest contrary to theirs; and consequently the sovereign power need give no guarantee to its subjects, because it is impossible for the body to wish to hurt all its members.”[xiv]

Rousseau seems to argue that the very undifferentiated abstract generality of the laws instituted by the people guarantees enough.

Habermas rebuts this and elaborates his answer and solution to this central puzzle of modern political philosophy as follows:

“The normative content of the original human right cannot be fully captured by the grammar of general and abstract laws alone, as Rousseau assumed. The substantive legal equality that Rousseau took as central to the legitimacy claim of modern law cannot be satisfactorily explained by the semantic properties of general laws.

The form of universal normative propositions says nothing about their validity. Rather, the claim that a norm lies equally in the interest of everyone has the sense of rational acceptability: all those possibly affected should be able to accept the norm on the basis of good reasons. But this can become clear only under the pragmatic conditions of rational discourses in which the only thing that counts is the compelling force of the better argument based on the relevant information. Rousseau thinks that the normative content of the principle of law lies simply in the semantic properties of what is willed; but this content could be found only in those pragmatic conditions that establish how the political will is formed.

So the sought-for internal connection between popular sovereignty and human rights lies in the normative content of the very mode of exercising political autonomy, a mode that is not secured simply through the grammatical form of general laws but only through the communicative form of discursive processes of opinion- and will-formation.

This connection remains hidden from Kant and Rousseau alike. Although the premises of the philosophy of the subject allow one to bring reason and will together in a concept of autonomy, one can do so only by ascribing this capacity for self-determination to a subject, be it the transcendental ego of the Critique of Practical Reason or the people of the Social Contract. If the rational will can take shape only in the individual subject, then the individual’s moral autonomy must reach through the political autonomy of the united will of all in order to secure the private autonomy of each in advance via natural law. If the rational will can take shape only in the macro-subject of a people or nation, then political autonomy must be understood as the self-conscious realization of the ethical substance of a concrete community; and private autonomy is protected from the overpowering force of political autonomy only by the nondiscriminatory form of general laws.

Both conceptions miss the legitimating force of a discursive process of opinion- and will-formation, in which the illocutionary binding forces of a use of language oriented to mutual understanding serve to bring reason and will together— and lead to convincing positions to which all individuals can agree without coercion.

However, if discourses … are the site where a rational will can take shape, then the legitimacy of law ultimately depends on a communicative arrangement: as participants in rational discourses, consociates under law must be able to examine whether a contested norm meets with, or could meet with, the agreement of all those possibly affected.

Consequently, the sought-for internal relation between popular sovereignty and human rights consists in the fact that the system of rights states precisely the conditions under which the forms of communication necessary for the genesis of legitimate law can be legally institutionalized. The system of rights can be reduced neither to a moral reading of human rights nor to an ethical reading of popular sovereignty, because the private autonomy of citizens must neither be set above, nor made subordinate to, their political autonomy.

The normative intuitions we associate conjointly with human rights and popular sovereignty achieve their full effect in the system of rights only if we assume that the universal right to equal liberties may neither be imposed as a moral right that merely sets an external constraint on the sovereign legislator, nor be instrumentalized as a functional prerequisite for the legislator’s aims.

The co-originality of private and public autonomy first reveals itself when we decipher, in discourse-theoretic terms, the motif of self-legislation according to which the addressees of law are simultaneously the authors of their rights. The substance of human rights then resides in the formal conditions for the legal institutionalization of those discursive processes of opinion- and will-formation in which the sovereignty of the people assumes a binding character. … As soon as the legal medium is used to institutionalize the exercise of political autonomy, these rights become necessary enabling conditions.”[xv]

Habermas elaborates further:

“The principle that all “governmental authority derives from the people” must be specified according to circumstances in the form of freedoms of opinion and information; the freedoms of assembly and association; the freedoms of belief, conscience, and religious confession; entitlements to participate in political elections and voting processes; entitlements to work in political parties or citizens’ movements, and so forth. In the constitution-making acts of a legally binding interpretation of the system of rights, citizens make an originary use of a civic autonomy that thereby constitutes itself in a performatively self-referential manner. Thus we can understand the catalogues of human and civil rights found in our historic constitutions as context-dependent readings of the same system of rights.

This system of rights, however, is not given to the framers of a constitution in advance as a natural law. Only in a particular constitutional interpretation do these rights first enter into consciousness at all. In fact, when citizens interpret the system of rights in a manner congruent with their situation, they merely explicate the performative meaning of precisely the enterprise they took up as soon as they decided to legitimately regulate their common life through positive law. An enterprise of this sort presupposes no more than the concept of legal form and an intuitive understanding of the discourse principle.

If talk of “the” system of rights means anything, then, it refers to the points where the various explications of the given self-understanding of such a practice converge.  … “The” system of rights does not exist in transcendental purity. But two hundred years of European constitutional law have provided us with a sufficient number of models. These can instruct a generalizing reconstruction of the intuitions that guide the intersubjective practice of self-legislation in the medium of positive law.”[xvi]

Thus Habermas achieves the promised post-metaphysical, coherent rational reconstruction of the dual foundations of our modern constitutional democracies by demonstrating the deep reciprocity between liberty and popular sovereignty.

Is this the last word on this matter? Not for me. Just a hint at the end: I can imagine a society that operates altogether without popular sovereignty, without the medium of a centralized political power, not without legal systems but without universally binding collective decisions, where free contracting assumes a much larger role and where multiple legal systems interpenetrate and are institutionalized via a competitive market for legal provisions through private organisations without monopoly of violence.  In this society political discourse does not disappear but indeed remains a crucial steering arena for societal developments, without presuming that the purpose of all discourse is to result in one-fits-all majority decisions.

End.

[i] Jürgen Habermas, Between Facts and Norms: Contributions to a Discourse Theory of Law and Democracy, Polity Press, Cambridge 1996, German: Faktizitaet und Geltung, 1992

[ii] Jürgen Habermas, The Theory of Communicative Action, Polity Press, Cambridge 1984; German: Theorie des kommunikativen Handelns, 1981

[iii] Jürgen Habermas, Moral Consciousnessand Communicative Action, MIT Press 1991; German: Moralbewusstsein und kommunikatives Handeln, 1983. Also: Jürgen Habermas, Justification and Application: Remarks on Discourse Ethics, MIT Press 1993; German: Erlaeuterungen zur Diskursethik, 1991.

[iv] Between Facts and Norms, p. 94

[v] Jean-Jacques Rousseau, The Social Contract, 1762, p.19

[vi] Jean-Jacques Rousseau, The Social Contract, p.10

[vii] Between Facts and Norms, p. 100

[viii] Between Facts and Norms, p.84

[ix] Between Facts and Norms, pp.94 - 100

[x] Between Facts and Norms, p.101

[xi] Between Facts and Norms, p.102

[xii] Jean-Jacques Rousseau, The Social Contract, p.1

[xiii] Jean-Jacques Rousseau, The Social Contract, p.10

[xiv] Jean-Jacques Rousseau, The Social Contract, p.10

[xv] Between Facts and Norms, pp.104-128

[xvi] Between Facts and Norms, pp. 171-172