Entrepreneurs are often hailed as being a different breed from the rest of us. They are considered restless, risk-tolerant, and allergic to hierarchy. And we often hear that Steve Jobs was always going to start a company, whether he dropped out of Reed College or not.

It’s a nice story, and one that is easy to tell and get people to buy into. But it is also only part of the explanation. What is missing is context. Where you work, who you work for, how long you’ve been there, and what kind of organization you are situated in can shape whether the entrepreneurial spark ever gets struck.

Sorting or Context?

There are two competing explanations for why we observe a group of people — engineers, academics, immigrants, public employees, for example — starting companies at an unusually high or low rate. Unfortunately, it is customary to conflate them.

The first explanation is sorting. Certain kinds of people, with certain dispositions, end up in certain jobs. Risk-averse people gravitate toward stable employers. Risk-tolerant people gravitate toward more volatile ones. If a group has a low rate of entrepreneurship, perhaps it is simply because less entrepreneurially inclined types selected into that group in the first place. Nothing about the job itself is necessarily doing anything. The sorting happened before the person entered that position.

The second explanation is context. The environment itself changes people, changes their opportunities, or both. Bureaucratic rules may narrow the scope for initiative. Fast-moving, chaotic organizations may do the opposite. Long tenure somewhere can gradually erode — or perhaps build — the skills and mindset needed to strike out on your own. From this perspective, the job is an active ingredient in the process leading to entrepreneurship.

Popular explanations for why a given group of individuals becomes entrepreneurial often quietly assume one of these two stories and ignore the other. That’s a problem because they point to very different policy prescriptions. If it’s sorting, you might want to change who you recruit. If it’s context, you might want to change the environment people work in. Most of the time, of course, both are tangled together in ways that are genuinely hard to pull apart empirically.

A Case Study: The Public Sector

A few years ago, Serden Özcan and I decided to pull on this thread by studying transitions from public sector employment into entrepreneurship in the United States, using more than three decades of household panel data.

The public sector is a particularly useful test case because most people already have a strong prior about it. Ask why relatively few government employees become entrepreneurs, and you’ll likely be told that people who work for the government aren’t exactly wired for risk-taking. That’s the sorting explanation, and for many people it more or less closes the debate.

Our aim was to identify whether that was in fact true, or simply a convenient story bureaucracy-sceptics tell themselves.

So we built a model that lets both mechanisms speak for themselves. One equation captures the decision to become an entrepreneur. Another captures the decision to work in the public sector in the first place. Importantly, we allowed the two equations to be correlated in ways that capture unobserved characteristics that might nudge people toward or away from both choices.

On top of that, we separated out a third mechanism that tends to get lumped in with “context” but is worth considering in its own right: tenure. Simply staying somewhere for a long time may shape the likelihood that someone eventually chooses to establish their own firm.

What We Found

First, the sorting story is real, but hardly the whole story. People who end up in public sector jobs really do look, on average, less entrepreneurially inclined even before they take the job. Less entrepreneurial types are systematically matched with government employment. So there is some support for the popular perception that many public sector employees were less likely to start a company in the first place.

The more surprising result appears once we account for that sorting. The public sector context itself increases the hazard of quitting to start a business. Once we strip out the selection effect, being in a public sector job in a given year makes entrepreneurial exit more likely.

This is almost the opposite of what the raw pattern might lead us to expect.

Why might that happen? There are several possible explanations. One is frustration. Public bureaucracies may attract people who want to make a difference while simultaneously constraining their ability to do so through impersonal rules, slow rewards for good performance, and limited scope for initiative. For some employees, entrepreneurship may consequently become an attractive route out. Re-entry into the public sector may also be relatively low-risk if the venture fails, while the psychological payoff from having greater influence over one’s work may be particularly valuable after spending years in a bureaucratic environment.

These are possible interpretations of the result rather than something our empirical analysis can establish directly. But the important empirical point is that once selection is taken seriously, the apparent negative effect of public employment changes.

Third, tenure matters, and it works differently in the public sector. The longer someone stays in a public sector job, the less likely they are to leave for entrepreneurship. And this tenure effect is stronger in the public sector than in the private sector.

This isn’t simply about people becoming more risk-averse as they get older. One possibility is the accumulation of specialized, agency-specific human capital. Another is adaptation to a particular mode of working — perhaps more reactive than proactive and more rule-following than rule-questioning — that makes moving into entrepreneurship progressively harder. Whatever the precise mechanism, staying longer appears to raise the barrier to entrepreneurial exit.

The low rate of entrepreneurship among public employees therefore reflects a layering of three mechanisms: who gets sorted into public employment, what happens while they are there, and what happens as tenure accumulates.

And this distinction matters empirically. In our earlier models, when we ignored the sorting effect, we actually got the sign of the context effect wrong. Looking only at the observed relationship would therefore have led us toward almost exactly the wrong conclusion about what public sector employment itself was doing.

Beyond the Public Sector

While our study concerns the public sector, the distinction between these mechanisms may be relevant in many other organizational settings as well.

Consider large, rule-bound and seniority-driven organizations: banks, hospital systems, university administrations, or parts of Fortune 500 companies. It is at least plausible that similar processes operate there. Such organizations may disproportionately recruit people attracted to stability while also frustrating some of the more entrepreneurially inclined people they do hire. At the same time, long tenure can make leaving increasingly costly.

That possibility has an interesting implication for leaders worried about losing their most enterprising employees. Recruiting practices and day-to-day management practices may work against each other. An organization can disproportionately attract people seeking stability while simultaneously making life difficult for the relatively entrepreneurial people who nevertheless join it.

There’s also a policy angle worth considering. Governments spend considerable energy trying to make public organizations more entrepreneurial through innovation labs, hackathons, and chief innovation officers. Our results raise the question of how much such initiatives can accomplish if the underlying bureaucratic structure — the rules, reward systems, and seniority ladders — remains unchanged. Adding visible elements of an entrepreneurial culture may have limited effects if the organizational structures governing everyday behaviour point in another direction.

None of this means bureaucracy is simply bad, or that government should try to become Silicon Valley. Rules, consistency, and impersonality exist for good reasons, including reasons that entrepreneurs generally don’t need to care about. But if a society genuinely wants more of the entrepreneurial talent currently sitting inside its public institutions to eventually put that talent to work, we need to look beyond the characteristics of the people who apply for those jobs.

We also need to understand what happens to them once they are there — and what happens when they stay.

Based on Özcan, S. & Reichstein, T. (2009). “Transition to Entrepreneurship from the Public Sector: Predispositional and Contextual Effects.” Management Science, 55(4), 604–618.

Today social impact tends to be a strategic priority of university leadership teams. But their answer becomes weary when asked them what it concretely means and how they intend to measure it. Indeed, their language tend to become charismatic of hedging. They will recognize that it is difficult to capture. And they will probably indicate the need to set up a task force to determine that more precisely. That is both sensible since it shows awareness and action-oriented behaviour. But ultimately, you are left with the strong impression that nobody quite knows what they’ve committed to but even so that it is the only right choice regardless.

It is easy, and even too easy, to suggest this to be a clear indicator of universities being poorly managed. I personally refuse to believe that to be the case. I also do not think it to be fair to make that conclusion. It might actually be that social impact is exactly the right strategic move for universities. It is only right and becoming to start from that premise that the choice is a rational choice and try to unfold that perspective and see whether it makes sense.

Why might social impact not be a silly strategy at all?

Universities depend overwhelmingly on resources controlled by external principals including government block grants, research councils, tuition policies set by ministries, capital funding for buildings, and increasingly, public legitimacy itself as a precondition for all of the above. It is important to keep this in mind and recognize that this is a relatively small number of powerful actors. In such a scenario, resource-dependence logic suggests that an organization who wish to prosper and grow should shape its stated priorities according to what those actors want. And let us be clear about this. Some might call this cynicism. But it comes close to how any economist would expect a rational agent to behave when a counterparty controls the resources it needs and is thus not entirely cynicism at least.

Politicians and funding bodies have been explicit about wanting social impact and see signs of social impact. The UK’s Research Excellence Framework built “impact case studies” directly into the formula that allocates funding to universities. In their system impact is not merely rhetorical but directly priced by adding to the score. Research councils across Europe increasingly require impact or knowledge-exchange sections in grant applications. And it is sometimes integrated with dedicated scoring criteria. National innovation strategies routinely link funding to universities’ ability to demonstrate contribution to growth, health outcomes, or societal challenges. Adopting impact as a strategic pillar is in this way not only a choice of what is fashionable. Instead, it is contractual engagement where the university exhibit a shift towards the focus areas the people who write the cheques are asking to see. Ignoring this signal would be a strategic error and suggest that the university do not exercise a rational purity in strategic action.

It is also important to remember that universities are facing a genuine crisis of public legitimacy. Society is increasingly questioning whether the universities in fact are delivering value for money. There are debates surfacing that question if there is a return on an increasingly expensive degree. And public debates often forward the view that academic research only serves academics and not anyone else. Being able to point to concrete societal contribution is not just about chasing funders’ preferences. It is also about improving the public trust more broadly. The pursuit of social impact may thus not be an arbitrary buzzword but rather universities’ rational response to a shift in what the key stakeholders are willing to fund and tolerate. And this is what makes it interesting. It is interesting to consider it a strategic sensible and sound move to align the organization with social impact. Below we will seek to provide some reasoning that surrounds this choice considering it a rational behaviour.

The case for scepticism

Before moving on, we might gain from ask what a strategy is supposed to do. Strategy is supposed to specify an objective and afterwards let you tell whether you achieved it. That’s a low bar that it needs to clear like every research grant application has to clears, every corporate strategy deck must clear, and every PhD thesis must clear when it defines a research question. Strategizing thus involves measurement, attribution of credit, and clear ownership of objective. Let us consider these in turn.

First, universities have means of measuring research including publications, citations, grant income, and rankings. Whatever you think of these metrics we acknowledge that they are comparable across institutions and that they are updated on a timescale that matches decision-making. They are indeed measurable in a few years rather than a generation. Social impact has none of this. Effects of a shift in how government thinks about a given policy (e.g. industrial policy) will almost certainly not be identifiable until a decade later. It will probably need to diffuse across dozens of contributing bodies of work, filtered through journalists, think tanks, and civil servants who never cite the original paper, and feed into a policy document whose authorship is institutional rather than individual before it will have even the initial minor impacts. And there is no clean, agreed-upon instrument that turns this into a number. In fact, it is not obvious such an instrument could be built even with unlimited resources seeing that the thing being measured doesn’t decompose cleanly into attributable units in the way a citation does.

Second, even granting that impact happened and we did measure it, we must also be able to provide clear indication of who it was the secured that impact and which effect that agent had on the impact. This is a similar the same problem economists confront in program evaluation were we need a counterfactual for claiming causal credit. What would have happened without that agent? For a single lecture or a single student, that’s hard enough, though at least tractable in principle with the right data. For a slow-moving shift in policy or public opinion shaped by hundreds of researchers across many institutions over many years, constructing a defensible counterfactual is not just difficult. You can observe correlation with some effort (e.g. a policy paper cited in a white paper, an alumnus who ends up in a ministry). But “our university caused this outcome” is a much stronger claim than that kind of evidence can carry. It’s the identification problem, transplanted from the seminar room into the strategy document, and about as under-acknowledged there as it would be embarrassing to see in a submitted paper.

Finally, a firm has a residual claimant in that the owner captures what’s left over after costs are paid. Profit thus gives everyone inside the firm a common, if imperfect, yardstick to optimize against. Universities have no equivalent. They answer simultaneously to government funders, students, industry partners, alumni donors, and academic peers. And each of these stakeholders holds a different idea of what “impact” should mean (even if they may not openly share these views). For a funding ministry it might mean GDP growth or productivity. For a student it might mean employability. For a peer academic it might mean influence on the field, which is just research quality wearing a different hat. For a civil society partner it might mean something closer to public value that resists monetization altogether. “Impact,” left undefined, is precisely vague enough that everyone can nod along to it without anyone having to say whose definition wins. The moment someone does specify whose definition wins, some of those stakeholders stop nodding. Social impact may in fact be sufficiently poorly defined and specified in terms of measurement so that all stakeholders are able to nod approvingly at least in the short run. In the longer run it may trigger some level of disagreement and conflict. In this situation, there is no clear ownership of objective.

Social impact thus struggles to qualify as a sensible strategic focus considering even a low bar for what is strategizing. It would seem almost irrational to even consider social impact as a key item for strategizing. But let’s turn attention to why it might be sensible even so.

Why might social impact be the rational choice even so?

In economics, vagueness does not qualify as something that is temporary and functional for a brief period where the agent is search for a more optimal choice or solution. As a matter of fact, economics would instead be likely to consider vagueness as a deliberate choice in a stable equilibrium that nobody inside the system has much individual incentive to disturb. By interpreting it in this way, the phenomenon may instead be reframed from failure to a quiet institutional rationality.

To shed further light on this, it might be prudent to draw on two economist that separately received the Nobel Price. Bengt Holmström and Paul Milgrom showed decades ago that when an agent faces several tasks, and only some of them are easy to measure, the agent will rationally shift effort toward whatever gets measured. Their work even indicated that the agent would do this regardless of whether the organization have other stated priorities and that it even will happen when the agent strongly believes in those other stated priorities. It seems clear that this exactly is what is happening at universities who have declared their strategy to be social impact based (or any of the many other vaguely defined strategies in the past). Promotion, tenure, and rankings all run on things that are easily measured including papers, citations, and grants. Universities may have great aspirations for being impactful. But it is likely to remain an aspiration without changing the incentive structure of the university. A researcher who spends a year on a well-measured paper instead of a diffuse, unattributable “impact” activity is behaving completely rationally given what gets counted in career relate decisions. In fact, in this perspective, one might claim that it is the clarity and operationality of the metrics that makes the vagueness of the strategy-level tolerable among faculty of universities.

So why do universities not invest heavily in generating precise measurements of (social) impact even if it is difficult. To answer that question, we might consider Goodhart’s Law (when a measure becomes a target, it ceases to be a good measure). Let’s imagine a university did commit to a hard metric like media mentions, policy citations, spinout companies founded, industry funding raised for capturing social impact. What will be the outcome of that shift? Well, people will then optimize based on those metrics instead of the underlying thing it was meant to proxy for. It is easy to predict this behaviour. After all, look at the current system where the research metrics distort research behaviour (e.g. strategic self-citation, salami-sliced papers, journal-chasing over substance). There are good reasons to suspect that the same thing will happen to an impact metric. And the process through which this will happen is likely to be faster, and the distortion will likely to be greater given that impact is easier to simulate convincingly than a citation count. A press release is cheap to produce. A policy citation can be lobbied for. A spinout can be incorporated without doing much of anything. In this perspective, staying vague isn’t necessarily failure. It is a rational hedge against the well-documented cost of picking the wrong precise number and then living for a decade with the perverse incentives it quietly creates.

There’s also a simpler signalling story that contribute to the universities’ behaviour in pursuing these buzz words in their strategies. Game theory would argue it no longer to be cheap talk for universities to do so whenever their funders have priced impact directly. In this case, real money is genuinely at stake and gaming it carries real audit risk. Having said this, it is important to note what the universities do when subjected to these harder constraints. They tend to satisfy the specific, narrow, externally-audited requirement while they internally remain relatively unclear and undefined in their language about “impact”. While it speaks to the external stakeholders and their agendas as a function of compliance, it seldomly develops into a clear, stable, internally-owned definition that the rest of the university’s strategy and incentive structure can be organized around. Indeed, this adopting of “impact” as part of the university’s strategy remains a low-cost way of telling funders, governments, and the public that the institution is listening to what they currently want to hear. It costs almost nothing to say, and all the stakeholders and funders willingly accept the “cheap” signals as credible commitments. And indeed, every peer institution is sending the same signals of commitment in the same vague language which means no single university pays a competitive price for failing to operationalize it. Peers continue to enjoy the reputational benefit of the word without paying that cost. Considering it this way puts the behaviour into a coordination equilibrium. It becomes rational for every institution to stay vague and collectively produce the fog that makes the whole exercise be sufficiently hollow.

Short-lived or persistent?

Let us finally consider the possibility but plausible scenario that the current push for social impact as the new black for university strategizing is relatively short-lived rather than a persistent unwavering choice. After all, these buzzwords come and go only never to return. And perhaps nobody inside the system expects “social impact” to be the organizing language for very long. Indeed, universities have been through similar cycle before. Go back fifteen or twenty years and the dominant vocabulary was “excellence.” Before that, waves of “internationalization.” At various points, “third mission,” “knowledge exchange,” “engagement,” “entrepreneurial university.”. Every time these buzzwords tricked the same reactions starting with strategy development which entailed task forces. And every time the strategic focus was replaced by a new one without having resolved or achieved anything noteworthy on the previous fad.

Given the same pattern will persist for “social impact”, it becomes clear that it is not a destination for universities but merely a temporary focus that will be changed within a foreseeable future. It will be substituted with a new buzzword which, given the current international affairs, might be sovereignty or security. Alternatively, it may become AI and technological transformation seeing how AI has already become a major concern for all organizations and agents including universities.

From an economics point of view, the pattern of replacing buzzwords periodically can also be considered functional rather than accidental. A pattern and accompanying arguments that strengthen the equilibrium perspective. Everyone signalling the same commitment in the same words becomes a pooling equilibrium (different types of senders in a signaling game choose the exact same action or signal) which has a shelf life. Its whole value lies in signalling alignment with what funders and the public currently want to hear. As soon as every institution has adopted the term, it stops differentiating anyone. At this stage it starts to look like exactly the cheap talk it is. At that point the signal decays and have little to no value in terms of the relationship with sufficiently attentive funder or ministers. The rational response for any institution, or any funder trying to look forward, is to shift toward whatever term hasn’t yet been fully commoditized by universal adoption. In that sense the buzzword’s turnover is not a symptom of the system’s fickleness so much as a predictable consequence of treating the term as a signal in the first place.

We may also consider this in terms of two other Nobel Laureates, Kydland and Prescott, and their credibility problem. A university that genuinely tried to hold itself accountable to a five- or ten-year impact metric would be committing to a yardstick that outlives the strategic fashion cycle that produced it. When the cycle of that buzzword is over, that university would risk being judged by a standard nobody else is being held to anymore and using language that since has become slightly embarrassing to say out loud in a strategy meeting. It is unlikely that social impact will be the buzzword in 2033 and still focusing on it and calling out that focus may indeed come with reputational cost. That gives institutions a further reason not to over-invest in precision in terms of definition and measurement. Not only because precision is measurement-hard and Goodhart-risky, as above, but because the term itself is unlikely to be the one anyone is still using by the time precision would be possible. The investment in itself will to a great extent resemble a sunk cost.

What does this imply?

None of this means impact is a fiction that universities should give up talking about it, or that the people sitting on these task forces are doing so in bad faith. Not at all. It simply means the vagueness is not primarily a knowledge problem. It suggests that it is a fairly predictable output of a genuinely hard measurement problem which is situated inside an institution with no single principal to define the objective. Furthermore, it means the vagueness is subject to an incentive system that already rewards something else entirely and that there is a reasonable fear of what happens once you name a new proxy and watch people optimize for that instead. And at the same time, it is unlikely that universities will incur the fixed costs of engaging in transforming the vagueness into precision given that the current vocabulary only has a five-to-ten-year lifespan as a central part of university strategy. It is likely that the choice is defined by relatively well-defined equilibrium where it is only rational to engage with the new buzz and not spend too many resources in seeking to turn it into a fully functional long-term strategy and what that entails.

Recognizing that I might be completely wrong, it is nonetheless worth discussion the testable predictions that this reasoning may entail. I see two possible hypothesis that can be tested. First, wherever universities do end up adopting hard social impact metrics, those metrics will not be whatever is conceptually closest to the universal definition of social impact, but whatever is cheapest to count and hardest to game in the short run (industry funding secured, spinout companies founded, patents filed). Not because these are the best proxies for social value, but because they are the ones that survive contact with an organization that needs a number by next year’s strategy review and because they might also function well for the next buzzword that is coming along. Second, given that the buzzwords are short-lived in university strategizing, then terms will start disappearing from university strategy documents within the decade without any form of resolution or achievement. Both these predictions are falsifiable and could thus be subject to empirical scrutiny.

Every summer, I run into the same annoying problem. My workplace—like a lot of others—tells me I have to take three weeks off in a row. It’s not an option or a perk. It’s mandatory. And every year, I find myself thinking, “This isn’t really when I’d choose to take my holiday if I had a say.” Maybe I’d grab two weeks in July, then save another for October, when Denmark gets that crisp, golden look and the office feels quieter. Maybe I’d split it up around my kids’ school breaks so things made a little more sense for us. But nope—the company calendar makes the call.

As an economist, I can’t help but pause when something like this happens. Instead of just quietly grumbling, I ask myself: is there a solid economic reason behind forcing people to take their holidays like this? Or is it just a well-intentioned rule that ends up being more hassle than help for someone like me? Honestly, there isn’t a simple answer. It depends, and when you really dig in, the pros and cons get a lot more interesting—and complicated—than you’d think at first glance.

The case against forced holidays

Let’s start with something pretty basic: most people know what works for them better than their boss does. That’s not just common sense—it’s straight from standard economic theory. If you give people the freedom to make their own choices, they’ll usually line things up in a way that works out best for them. Let’s say I can weigh the tradeoffs—family time, travel costs, my workload—better than HR ever could. So when a company tries to force everyone into the same vacation schedule, it actually makes things worse. Think about it: a parent juggling school breaks and a single 26-year-old on the hunt for cheap flights need totally different things. A universal rule never fits anyone quite right, and that “deadweight loss” economists like to talk about? That’s real. It’s the difference between when I’d actually want a break and when I’m told to take it.

Money matters, too. For people paid by the hour, missing work isn’t just an inconvenience—it’s lost income, and often at a time that’s hard for their budget. Telling someone, “Stop working for three weeks,” feels a lot different if you’re salaried and secure versus if every paycheck really counts. For someone scraping by, every extra day’s wage matters.

There’s also the simple reality that some people just love to work. Revealed preference theory doesn’t look kindly on policies that ignore that. Sure, a lot of articles paint workaholics as irrational, but honestly, liking to work hard isn’t any less reasonable than liking anything else. Some of my most productive, creative stretches came when I wasn’t interrupted by a mandatory holiday; I had momentum going, and I was glad to ride it. A blanket rule about time off just can’t tell the difference between someone about to burn out and someone hitting their stride—because it’s not measuring how people actually feel; it’s just slapping a calendar over everyone’s life.

The case for forced holidays

And yet. When I try to convince myself to fight the mandate, I invariably keep thinking up reasons for why it might be clever. Reasons that explains why forced hodays is a good practice.

The Rat Race Problem: The most persuasive, I feel, is the old coordination game we saw in Akerlof’s work on efficiency wages. If taking holidays is a signal to other employees and managers alike that you’re not fully committed then obviously no one can ever risk being first to take a holiday – regardless of the fact that a holiday-averse world would be worse for everyone if everyone accepted that they had to keep working? The mandate saves everyone from the classic no-win-situation and it takes you from your personally optimal but universally damaging Nash equilibrium. It’s not so much restricting my freedoms that the mandate is actually liberating me by letting me escape a self-damaging equilibrium. It is basically freeing me form a social trap.

Present bias and self-control: There is empirical evidence suggesting that individuals value current reward far more over future reward in a similar way that we undervalue retirement savings compared to immediate needs. Many of us would love to carry on never taking a holiday because the short term disruption of doing so appears too great and the benefit of a weeks long rest appears less concrete and distant in future years. All of those factors compound however. A firm wide level the mandatory holiday is like mandatory inclusion into a retirement plan. It will do a lot of our future-self a favour by having this little nudge to take time out and enjoy.

Marginal product timing. An argument which is probably close to the one driving the policy of my workplace pertains to seasonality. The organisation may seek high labour supply when the marginal product of the labour is high and low labour supply when it is low. And marginal product of the labour may be highly seasonal. At the the university sector at which I work, only summer school is running during July and August and student numbers are thus thin. This may entail that the marginal value of an additional lecturer is very low or even zero in the lecture theatres. Forcing holiday into that season may be the institution’s attempt to minimizing the output it forgoes by my holiday. This logic only cleanly applies where output really is single-dimensional and seasonal. My output at the university is a joint product of teaching and research. While teaching capacity is idle in summer, research capacity is often higher, since the very teaching load that competes for my time during term time has disappeared. Indeed, there is a more general problem where organizations frequently run a single uniform HR policy across employees whose actual production functions are quite different from one another. The forced holiday policy may in fact ends up optimizing for the wrong margin for a subset of the people it governs. The marginal product argument applies far better in sectors where output really is seasonal and largely one-dimensional (tourism, retail, agriculture, hospitality).

Fixed costs of partial operation. There are a significant number of workplaces at which running at reduced capacity is barely cheaper than running at full capacity. Indeed, a large share of overhead, heating and cooling, security, IT support, management attention, building maintenance may be fixed rather than scale down with occupancy. However, at some organizations it might in fact be the case that the firm pays close to full overhead twice if half the staff take holiday in July and the other half in October. Synchronizing the employees holiday may allow the organisation to collapse those fixed costs into a single period rather than paying them twice over. This is a pure firm-side efficiency argument.

Production complementarities. There is a simpler coordination problem story which does not include status competition like tha raised in the rat race issue. Assuming that my output depends on colleagues being present (which more often than not is highly likely), then individually chosen holiday timing creates real coordination costs even in a culture with zero stigma attached to taking leave. Synchronizing everyone’s absence into the same window solves a genuine production complementarity problem.

Labor hoarding in seasonal industries. In sectors with strong seasonal demand cycles, tourism and agriculture being the clearest examples, firms sometimes prefer to keep workers formally employed through the slow season via paid leave rather than laying them off, because rehiring and retraining ahead of the next peak season is expensive and uncertain. Here the forced holiday functions as a disguised retention strategy: the firm uses the low season as the natural moment to discharge its leave obligation without losing workers it will urgently need again in a few months.

Accrued leave as a balance sheet liability. A duller point, but a real one. Under many accounting regimes, unused vacation entitlement sits on a firm’s books as a financial liability, since the firm may eventually have to pay it out, particularly if an employee departs. Use-it-or-lose-it rules and mandatory leave windows are sometimes driven less by any theory of productivity or wellbeing and more by finance departments wanting to prevent that liability from accumulating on the balance sheet.

Externalities and fraud detection. Two further arguments round out the case, from opposite ends of the spectrum. On one side, burnout and presenteeism impose costs the individual worker does not fully internalize: health costs partly absorbed by employer insurance or public health systems, error costs absorbed by colleagues and clients, turnover costs absorbed by the firm. If the private return to skipping vacation exceeds the social return once these spillovers are counted, a mandate functions much like a Pigouvian correction. On the other side, in finance and trading in particular, mandatory consecutive leave has long been used explicitly as a fraud detection device, since many kinds of financial misconduct require the ongoing presence of the person managing the scheme to keep the books balanced. Force them out for two consecutive weeks and someone else has to sit at their desk. This is not a wellbeing policy dressed up in economic language, it is a control mechanism that happens to look like one.

So where does that leave me

Several of the forwarded arguments (marginal product timing, fixed costs, labor hoarding) apply most cleanly to firms with genuinely seasonal and fairly one-dimensional output. My own sector is a partial exception (an argument most researchers at University seem to think). Research does not go quiet in summer even if teaching does. A policy justified by one output variable is thus faulty since it fails to recognise that at university, the time of a professor is split between administration, teaching and research by contract. The rat race and complementarity arguments works pretty well for contexts in which visibility is quietly rewarded or deadlines genuinely require a full team. In this context there is more sense in such a forced holiday policy. Yet, the policy would make less sense in a context where there is a culture that already treats time off as unremarkable, or in roles that are fairly independent of colleagues’ schedules.

The reasonability of the practice may also depend on who you are within that context. Whether you are somebody who have a strong self-control and a flexible household or you are somebody who often are caught in a status competition or are working in a highly interdependent team may dictate whether the policy make sense or not.

I would certainly place my three weeks differently if I could. And I suspect my own university’s version of the policy is really about thin summer school enrollment rather than any of the other arguments above. But I am less sure than I used to be that a world in which everyone could freely choose their own holiday timing would actually deliver more rest than the one we currently have.

We have gotten used to talking about entrepreneurs the way we talk about Lionel Messi or Ed Sheeran. Mark Zuckerberg, Jensen Huang, Brian Chesky — names that function less like CEOs and more like franchises. Business media runs profiles with the same structure as a music magazine cover story: the early struggle, the breakthrough, the singular vision, the now-untouchable status. It is a flattering comparison, and a seductive one. It is also, mostly, wrong.

The economist Sherwin Rosen gave us the cleanest account of why a tiny number of performers earn obscene multiples of the next-best alternative. Underneath his model are three conditions that have to hold together:

1. Differences in quality or talent matter to consumers. People don’t treat a great performer and a mediocre one as interchangeable.

2. Consumers can choose among many providers, so they’re free to chase the best one rather than settle for whoever’s nearby.

3. The best providers can serve very large markets at low marginal cost. A recording or a broadcast reaches a million people about as cheaply as it reaches one.

Put these three together and you get a reward structure where small differences in talent translate into enormous differences in income. Not because the market is irrational, but because it’s ruthlessly efficient at finding the very best and then letting that one performer reach almost everyone, almost for free.

This is a good description of why Adele earns what she earns, why Messi earns what he earns, why a top box-office actor commands a fee no one else in the cast comes close to. It is a much worse description of why Jensen Huang or Brian Chesky are billionaires.

Where the analogy breaks

Three of Rosen’s conditions don’t transfer cleanly to entrepreneurship.

Joint consumption applies to the product, not the person. When a billion people use Instagram, they are not consuming Zuckerberg’s talent the way an arena consumes Ed Sheeran’s performance. They’re consuming a product that a team built, that network effects made valuable, and that venture capital and acquisition strategy helped scale. The entrepreneur’s payoff comes through equity — a financial claim mediated by capital markets, board governance, and exit timing — not through an audience paying to directly experience their skill. Strip away the platform and there’s no “Zuckerberg performance” left to consume, the way there’d still be an Ed Sheeran concert.

Talent isn’t observable the way a batting average is observable. Sports and music have something close to a public scoreboard. We can rank a hundred sprinters by stopwatch before any of them becomes famous. We have no equivalent instrument for entrepreneurial talent. Most of what gets called “founder genius” is read backward from outcomes that were, at the time, deeply uncertain.

The reward structure isn’t a smooth gradient — it’s a cliff. Rosen’s model produces a continuous relationship: slightly better talent, noticeably better pay, scaling up smoothly toward the top. Startup equity doesn’t behave that way, as the next section gets into.

A useful way to hold the distinction in your head: superstar theory is fundamentally about the economics of talent and market reach. Entrepreneurship theory is fundamentally about the economics of ownership and uncertainty. A labour economist might put it like this — superstar theory explains why a small number of individuals earn extraordinary returns from their personal talent in scalable markets, while entrepreneurship theory explains why returns to business ownership are highly skewed because of risk, uncertainty, innovation, and equity ownership. Two different engines, even when the dollar amounts at the top look similarly absurd.

The better-fitting model: a lottery with loaded dice, not a talent ladder

If Rosen’s model is the wrong scaffolding, what’s the right one? Entrepreneurship looks much less like a superstar market and much more like a lottery where skill buys you extra tickets, not a guaranteed prize.

Here’s the core difference. In Rosen’s world, moving from very good to truly exceptional pays off smoothly — a slightly better singer or athlete earns a meaningfully bigger reward, and the climb up that ladder is fairly continuous. Entrepreneurial outcomes don’t climb a ladder like that. The vast majority of founders end up in roughly the same place — modest or no return — and then, out of nowhere, a tiny handful land an outcome that’s not just bigger but categorically different, an order of magnitude or more beyond everyone else. There’s no smooth slope connecting the two. It’s a cliff.

This is exactly why venture capital is built the way it is. Investors don’t try to minimize failures the way you might expect a careful investor to. They expect most of their bets to lose money, because the entire model depends on one enormous winner covering all the losers many times over. That’s a strange way to invest if you think you’re identifying the most talented founders in advance. It makes complete sense if you think you’re buying lottery tickets that have already been pre-screened for basic competence, and counting on a few of them to hit.

And the unevenness isn’t something the market adds on top of entrepreneurship — it’s there from the start. Research looking at thousands of young and high-growth firms found that almost everything we’d normally expect to be fairly evenly distributed among founders — their resources, their decisions, their environment, even basic measures of how they think about problems — turns out to be lopsided in the same lottery-like way. Most founders cluster near the bottom, a few sit wildly far out on their own. So the extreme outcomes we see later aren’t created by some amplifying market mechanism, the way Rosen’s joint-consumption mechanism amplifies a singer’s talent into stadium tours. The unevenness was baked in before any market touched it.

Put differently: the right question for entrepreneurship usually isn’t “how much more talented is this founder than the next one” — a question nobody can really answer in advance — but something closer to “did this person, at the right moment, get matched with the right co-founder, the right investors, and a market that happened to be ready, and did they survive enough rounds of being tested for that to add up to a real outcome.” Talent improves your odds. It doesn’t write the ending.

So why do VCs spend so much time on the team, then?

A natural objection: if entrepreneurial talent is so hard to observe, why does every VC pitch meeting open with deep diligence on the founders themselves? Doesn’t that prove they’re scouting talent the Rosen way?

Not quite. VCs aren’t ranking founders on a continuous scale the way a scout times a sprinter — there’s no instrument that does that. What they’re screening for is narrower: does this person have the traits that correlate with surviving the next eighteen months of chaos — founder-market fit, ability to recruit, resilience, capital discipline, coachability. That’s a pass/fail filter for ticket eligibility, not a fine-grained talent ranking. In a world with no stopwatch for entrepreneurial talent, the team is simply the least-bad observable signal available before any outcome exists — idea quality and market timing are often even harder to assess at the seed stage. Team-first diligence is the underwriting process for raising your odds in the lottery, not a Rosen-style talent audit. If outcomes really were driven by a legible, continuous talent signal, you wouldn’t need this much slow, qualitative, often-unreliable interviewing to get at it — you’d just look at the equivalent of a stopwatch.

A short list of partial exceptions

This is not to say the superstar mechanism is entirely absent from entrepreneurship. A few cases plausibly come closer to Rosen’s logic than most — and tellingly, they’re almost all American, which may say as much about US business culture as about economics:

• Elon Musk arguably trades on something like a personal-brand premium that functions independently of any one company — Tesla’s valuation has at points seemed to price in “Musk” as a factor distinct from the underlying fundamentals, closer to how an audience prices a star actor regardless of the film.

• Sam Altman, post-ChatGPT, has something resembling celebrity-grade public name recognition that attaches to him personally and travels with him across ventures and advocacy roles, somewhat like a performer’s reputation following them between projects.

• Founders who become public commentators or media figures in their own right, using a platform built on one company to sustain personal visibility independent of that company’s later performance, come closer to “joint consumption” of the person rather than the product.

Even these cases are partial. None rests on a legible, ex ante talent ranking the way an athlete’s stats do, and none scales reputation at zero marginal cost the way a song does. They’re superstar-adjacent, not superstar cases proper.

Why the mislabeling matters

This isn’t just a definitional quibble. The superstar narrative does real work, and most of it is unhelpful.

For aspiring founders, and for the PhD students and early-career researchers I spend a lot of time around, the superstar framing implies a comforting but false promise: that if you’re simply talented enough, the market will eventually recognize and proportionally reward you, the way it eventually recognizes a great tenor. The lottery-with-loaded-dice framing implies something harder to hear: Being talented is necessary but radically insufficient, that timing and variance dominate the outcome, and that the rational response is not “work harder to climb the gradient” but “make many attempts, expect most to fail, and design your career so failure is survivable.” Conflating the two leads people to over-invest in single bets they’ve been told are meritocratically guaranteed to pay off if they’re just good enough.

For innovation policy, the stakes are similar. A government or university that internalizes the superstar framing designs programs to find and back the most talented entrepreneur, much as a talent scout finds the most talented athlete. But if talent isn’t legible in advance and outcomes are dominated by lottery-like variance, the better policy instrument looks more like a venture portfolio than a scouting competition: fund many attempts, expect most to fail, and judge the system by its tail outcomes rather than by whether any single grant committee successfully identified “the next Zuckerberg” in advance.

And for the public discourse, the superstar comparison flatters entrepreneurs in ways that obscure how contingent their position really is. It also lets a tiny number of survivors retroactively narrate their success as the product of foresight the data can’t actually distinguish from luck. Rosen’s superstars really can be measured against a stopwatch or a chart position before the world crowns them. Entrepreneurial “stars” mostly cannot. We just stop noticing the ones who didn’t make it.

In a recent piece on this Substack, I tried to make the case that Nigel Farage’s rise, paradoxical as it appears on the surface, is largely what economics would predict. Stagnant growth, concentrated costs, decades of felt abandonment, and a mainstream politics that offered no credible outlet for the resulting frustration all played their part. The article generated a good deal of discussion. I thus thought it only to be fitting to post a follow up on the topic asking: Does the structure of the political system itself make these kinds of explosions more likely? Is Britain’s predicament partly a product of how its democracy is designed, rather than just what its politicians have done or failed to do?

It is a good question, and it deserves its own treatment. So consider this a companion piece, using the same economic lens, but trained on the rules of the game rather than the players.

Start with a thought experiment from economics

Imagine two ice cream vendors setting up on a long beach. Where do they end up? If each is trying to maximise customers, they both gravitate toward the middle of the beach, because that is where most people are. A vendor who sets up at one end loses all the customers on the other side to their competitor.

The economist Harold Hotelling formalised this intuition in 1929, and political scientists quickly noticed its application to party competition. Two parties competing for votes will tend, over time, to converge toward the centre, toward the median voter, because that is where elections are won. Moderation, in this model, is not a moral choice. It is a competitive strategy.

This sounds reassuring. But it has a structural problem that is easy to miss. Voters far from the centre, including the deeply dissatisfied, the economically left behind, and those with strong views that neither mainstream party wants to adopt, are systematically abandoned. Neither party needs their vote badly enough to really compete for it. So their preferences are occasionally acknowledged, rarely acted upon, and ultimately ignored.

In a two party system, these voters have nowhere credible to go. And that, over time, creates pressure.

What happens to that pressure?

In a political system with many parties and proportional representation, where seats in parliament reflect vote shares roughly accurately, dissatisfied voters have genuine alternatives. A voter who feels the mainstream has failed them can shift their support to a smaller party that actually speaks to their concerns, wins seats, and exercises real influence, perhaps even entering government as a coalition partner. The grievance finds an outlet. The pressure releases gradually. The system adjusts.

In a two party, winner takes all system, that outlet is largely blocked. Voting for a third party feels like wasting your vote, because under first past the post a party can win millions of votes and almost no seats. So dissatisfaction accumulates beneath the surface, sometimes for decades, without meaningful political expression.

When a charismatic insurgent eventually provides a spark, the release is not gradual. It is explosive. What looks from the outside like a sudden political earthquake is, in reality, the release of long accumulated pressure that the system had no mechanism for releasing slowly.

Brexit was arguably exactly this. It was not a sudden eruption of new grievances, but decades of accumulated frustration finally finding an outlet, crucially, through a referendum that bypassed the usual electoral system entirely.

But is Britain really a two party system?

Here it is worth pausing, because the picture is more complicated than the simple two party label suggests.

Britain has, in practice, several significant political forces beyond Labour and the Conservatives. The Liberal Democrats hold meaningful support, particularly among remain leaning, economically comfortable voters in southern England. The Scottish National Party dominates Scottish politics. Plaid Cymru competes seriously in Wales. The Greens have a growing presence. And now Reform itself is winning hundreds of council seats and polling consistently at the top of national surveys.

So in terms of voter preferences and party diversity, Britain looks increasingly like a multi party system. The problem is that its electoral system, first past the post, was designed for two parties, and it translates that diversity of preferences into parliamentary seats very poorly. A party can win fifteen percent of the national vote and end up with a handful of MPs, while another wins the same share in more geographically concentrated form and wins dozens of seats. The votes are pluralist. The parliament is not.

This is the tension at the heart of British democracy right now, and it matters enormously for understanding what Farage is doing.

Is Farage actually accelerating Britain toward a multi party system?

This is perhaps the most interesting question of all, and it cuts against the instinctive reaction of many who see Reform simply as a threat to democratic norms.

One could argue, and I think there is something to this, that Farage’s insurgency is not the destruction of British party politics but its forced modernisation. The traditional two party duopoly of Labour and Conservative, which served reasonably well for much of the twentieth century, has been visibly failing to represent the full range of British political preferences for at least two decades. Reform’s rise, alongside the Liberal Democrats’ recovery and the Greens’ growth, is forcing a reckoning with that failure.

In the May 2026 local elections, Reform gained massively, but so did the Liberal Democrats in their areas of strength. Labour and the Conservatives both lost heavily. The result was not a simple transfer of power from one dominant party to another, but a genuine fragmentation of the political landscape. Britain’s party system is, in real time, becoming more plural.

The question is whether the electoral system will catch up. First past the post is increasingly struggling to contain a political reality it was not designed for. Calls for proportional representation, long the hobby horse of smaller parties with nothing to gain from the existing system, are now being heard in more mainstream quarters. Whether that conversation leads anywhere is uncertain. But the pressure for it has never been stronger.

Two different ways to contain a populist surge

Multi party systems are not without their own vulnerabilities to populist insurgency. Marine Le Pen in France, the Sweden Democrats, Alternative für Deutschland, all emerged within proportional systems. The difference is not immunity but containment. And containment, it turns out, can take more than one form.

Germany illustrates one route: coalition dilution. In a proportional system, a radical party winning fifteen or twenty percent of seats enters the legislature and faces the reality of coalition politics, including compromise, accountability, and the grinding exposure to the complexity of actually governing. The AfD has grown to around twenty percent of the national vote over roughly a decade, which is alarming to many, but it has so far been kept out of government and diluted in influence by the arithmetic of coalition building. It is inside the tent, but the tent has other occupants who constrain what it can do.

Denmark illustrates a different route entirely, and one that may be more instructive for Britain, because it does not rely on proportional representation forcing a radical party into government with moderating partners.

The Danish People’s Party, Dansk Folkeparti, emerged in the 1990s with a platform built heavily around restricting immigration and protecting cultural homogeneity, not unlike Reform’s core appeal in Britain. For roughly two decades it pulled working class and rural voters away from the Social Democrats, much as Farage has been pulling them away from Labour.

But the Danish Social Democrats responded differently than Labour has. Rather than treating immigration scepticism as something to be managed, contained, or quietly disagreed with, they absorbed it. Under Mette Frederiksen, the party adopted a markedly restrictive stance on asylum and immigration, in some respects to the right of where European social democratic parties have traditionally stood, while keeping its traditional left leaning positions on welfare, labour markets, and redistribution largely intact. The result is a party that, on immigration, sits closer to Dansk Folkeparti than to its sister parties in Germany or Sweden, while remaining recognisably social democratic on almost everything else.

The effect, at least in the short term, has been to defuse much of the pressure that powered the Danish populist right. Dansk Folkeparti’s vote share has declined substantially since its peak, not because it was excluded from power and faded into irrelevance, but because its signature issue was taken from it by the party best positioned to make the absorption credible to working class voters.

This is a genuinely different mechanism from the German case. Germany contains the AfD by keeping it out of government and diluting its influence through coalition arithmetic. Denmark contained Dansk Folkeparti by letting the mainstream itself shift, co opting the grievance rather than out competing it through institutional exclusion.

Whether this is a sustainable solution or simply a deferral is a fair question. It requires that voters trust the mainstream party’s new position is sincere rather than tactical, and it carries its own risks, chiefly that a left leaning party adopting a right leaning position on a salient issue can alienate its existing base, or be seen as legitimising the underlying premise it has adopted. But as a short run containment strategy, it has clearly worked better than ignoring the issue or hoping it would fade.

Britain’s Labour Party faces a version of this choice right now, whether it acknowledges it or not. So far it has largely chosen neither full absorption nor confident rebuttal, but something closer to uncomfortable silence. The Danish case suggests that strategy may be the least stable of the available options.

The containment problem in a winner takes all system

In a winner takes all system, a party that breaks through faces none of the constraints available to Germany or Denmark. If Reform were to win a general election outright, which remains uncertain but is no longer unthinkable, it would do so with unconstrained executive power. There would be no coalition partners to dilute the programme, and no proportional ceiling on ambition, and unlike the Danish case, no mainstream party has yet found a credible way to absorb its core appeal.

The system that kept Farage out for so long offers no soft landing if he gets in.

What this suggests

The deeper lesson here is not really about Farage specifically. It is about the relationship between political system design and political stability.

Systems that provide genuine outlets for diverse preferences, that let dissatisfied voters be heard without requiring them to burn everything down, tend to be more resilient over time. They let pressure out gradually rather than storing it until it explodes. They expose radical parties to the moderating discipline of coalition governance, or give the mainstream room to credibly absorb their grievances, rather than sealing them out until they are strong enough to govern alone.

Britain is conducting, somewhat involuntarily, a live experiment in what happens when a first past the post system meets a genuinely multi party electorate. The outcome is not yet determined. But the experiment is well underway.

And Nigel Farage, whatever else one thinks of him, may turn out to be less the destroyer of British democratic norms than the pressure gauge that finally broke, revealing that the system underneath had been running too hot for a very long time.

I should be clear that everything above is filtered through an economist’s lens, one that emphasises incentives, system design, and the structural conditions that make certain outcomes more or less likely. Political scientists would bring a rather different toolkit to the same questions, including theories of party organisation, electoral behaviour, democratic legitimacy, and institutional change that I have at best a passing familiarity with. My instinct is that the economic and political science perspectives would be broadly complementary here rather than contradictory, but I hold that view with appropriate humility.

Every week, I encounter some version of the same reaction from educated, cosmopolitan Europeans. A raised eyebrow. A slow shake of the head. “How,” they ask, with genuine bewilderment, “can the man who gave Britain Brexit, which has been, by almost any serious economic reckoning, a self-inflicted wound, now be on the verge of becoming Prime Minister?”

The flabbergasted expression that follows suggests they are witnessing something outside the bounds of rational explanation. A glitch in the democratic operating system. Evidence, perhaps, that voters have simply lost their minds.

I understand the sentiment. But as an economist, I find myself offering a different reaction: mild surprise that anyone is surprised.

Because when you strip away the moral indignation, understandable though it may be, and look at what economics actually predicts will happen when growth stagnates, regional inequality deepens, and mainstream politics repeatedly fails its constituencies, the Farage ascendancy is not a mystery. It is almost textbook.

Let me walk you through the main reasons why.

Voters don’t do counterfactuals

The question most commentators are asking is essentially a moral one: how can voters reward the architect of their own misfortune? It is the right question ethically, but it is the wrong one analytically. The more useful question is: what do we actually know about how voters assign blame?

The answer from political economy is uncomfortable but consistent. Voters, by and large, do not engage in careful counterfactual reasoning. They do not sit down and ask: “What would my living standards have looked like in an alternative timeline where Brexit never happened?” That kind of thinking requires holding two parallel worlds in your head simultaneously. The world as it is, and the world as it might have been. This is cognitively demanding and emotionally unrewarding.

Instead, voters experience the world as it actually is. Stagnant wages. Crumbling public services. Expensive food. Long waits in A&E. And they attribute these conditions, quite naturally, to whoever currently holds the keys to 10 Downing Street.

Right now, that is Labour. And so Labour owns the misery. That included the misery that predates them, and the misery that flows directly from a political project that Farage himself championed. Sitting comfortably in opposition, he gets to run against the consequences of his own policy. Brilliant some might say. Political scientists call this “retrospective voting.” It sounds neutral but certainly is not.

When costs land on some people and benefits float away from everyone

There is a second mechanism at work, and it is equally important.

Brexit’s aggregate economic damage is real. Most serious estimates put UK GDP somewhere between four and six percent below where it would otherwise have been. A significant and sustained drag on national prosperity. But here is the thing about aggregate numbers: they are politically inert. Nobody experiences “GDP.” People experience their pay cheque, their local hospital, the price of a weekly shop.

What matters, then, is not the average impact but the distribution of costs and benefits.

The costs of Brexit fell with considerable precision on specific groups and places. Traded-goods industries that depended on frictionless access to European markets. Border-adjacent businesses disrupted by new customs bureaucracy. Communities that had relied on EU labour to staff their care homes and harvest their crops. For these people, the consequences were concrete, visible, and daily.

The putative benefits, on the other hand (regulatory autonomy, reduced budget contributions, the theoretical freedom to strike new trade deals around the world) turned out to be diffuse, abstract, and for the most part yet to materialise in any tangible way.

This is a pattern economists recognise immediately. When the costs of a policy are concentrated on identifiable groups and the benefits are spread thinly across the population, the aggrieved have a vivid, specific grievance while the putative winners have little to celebrate. The result is almost always political instability. This is not a Brexit quirk. It is a structurally predictable outcome of any large distributional shock.

The left-behind were left behind long before 2016

Here is the part that polite commentary often glosses over, because it complicates the clean narrative of Brexit-as-catastrophe.

The regions that voted most heavily to Leave where post-industrial towns across the Midlands and North of England, coastal communities, places that once had steel mills, coal mines, or thriving fishing fleets. These regions were already economically marginalised well before the 2016 referendum. Brexit was not only the cause of their grievance. It was also a loud, blunt expression of it.

The deindustrialisation of the 1980s hollowed out entire communities. The subsequent decades saw economic dynamism become ever more concentrated in London and a handful of other cities, while towns like Hartlepool, Stoke-on-Trent, and Grimsby were largely left to manage their own decline. For many people in these places, the feeling of being ignored, talked down to, and economically discarded by the political mainstream was not a new experience in 2016. It was a chronic one.

The economist Albert Hirschman gave us a useful way to think about this. When people feel that a deteriorating institution (a party, a system, a government) is not responding to their concerns, they have two basic options. They can use their voice: stay engaged, advocate, try to change things from within. Or they can exit: walk away, withdraw their loyalty, look elsewhere.

After forty years of feeling that neither Labour nor the Conservatives were genuinely speaking to their situation, many of these communities concluded that voice wasn’t working. Exit began to look attractive. And Farage was standing at the exit door, offering a simple, emotionally resonant explanation for why things had gone wrong: immigration had driven down wages, metropolitan elites had captured the political system, and unaccountable institutions in Brussels had stripped Britain of its sovereignty.

The explanation was factually questionable in important ways. But it was coherent, it was simple, and it named a villain. That is a very powerful combination when you have been economically struggling for a generation.

The competitor that was supposed to stop this collapsed

Under normal electoral conditions, a right-wing protest vote of this scale would be absorbed by the mainstream centre-right party. The Conservatives exist, in large part, precisely to catch this kind of disaffection and channel it into something governable. A voter angry about immigration or cultural change or economic stagnation would historically end up, eventually, back in the Conservative tent.

But the Conservative Party is not operating under normal conditions. After the remarkable chaos of the post-Boris Johnson years (multiple prime ministers within months of each other, a minibudget in 2022 that caused the pound to collapse and forced the Bank of England to intervene in the bond market, a procession of scandals that made the party a byword for incompetence and bad faith) it has simply lost the authority to serve as a credible home for disaffected right-leaning voters.

The result is that there is no functioning right flank to hold the line. Reform’s gains in the May 2026 local elections came at the expense of both major parties simultaneously. Farage is not just eating the Conservative vote. He is attracting former Labour voters in the old industrial heartlands too. That is a genuinely unusual political coalition, and it is what makes him so hard for the existing parties to counter.

Why people don’t change their minds even when the evidence arrives

There is one final piece of the puzzle, and it may be the most psychologically interesting.

The economist Timur Kuran developed a concept he called preference falsification. The idea is that people often hold views privately that they would not express publicly, because the social cost of expressing them feels too high. For years, large segments of the British public may have harboured strong feelings about immigration, national identity, and distrust of European institutions. Yet, the dominant public discourse made these views feel unsayable in polite company. You kept them to yourself at the dinner table.

What Brexit did was normalise the public expression of these views. Once preferences that have been suppressed finally come into the open and accumulate political force, they do not easily go back. And people who voted for Brexit in 2016 did not quietly become Remainers when the economic data started to come in. That would have required admitting, to themselves and to others, that they had been deceived by Farage prominently and that the promised sunlit uplands had not appeared. That is a psychologically very expensive admission to make. So instead, the grievances deepened, the attachment to the narrative hardened, and the search for new villains to explain why the promised benefits had not arrived intensified.

This is what makes the Farage phenomenon feel so paradoxical from the outside, while feeling entirely logical from within. He is not winning despite Brexit’s consequences. In a strange but analytically coherent way, he is winning because the alienation, the resentment, and the appetite for a simple story remain entirely intact.

So is there anything to be done?

The honest answer from economics is: not quickly.

The structural forces driving this (entrenched regional inequality, the erosion of working-class economic security, the decades-long geographic concentration of growth) took a long time to build and will take a long time to undo. Short-term political counter-moves, however clever, are unlikely to shift the underlying dynamic much.

What would actually make a difference? Sustained, targeted investment in the places that feel abandoned. Genuine reform of the housing, planning, and transport systems that have entrenched Britain’s geographic inequality. A political class willing to speak honestly about trade-offs rather than promising frictionless solutions to complex problems.

None of that is easy. None of it is fast. And none of it fits neatly on the side of a bus. Which is, of course, a large part of why Farage keeps winning.

There is a particular kind of letter that arrives in an academic’s inbox with reliable frequency. It is polite, collegial, and faintly flattering. A journal writes to ask whether you might find time to review a submission. The editors value your expertise. The paper falls squarely in your area. They would be most grateful.

What the letter does not mention, and what the editors know perfectly well, is that this same journal has rejected your own work. Repeatedly. The contribution was insufficient, you were told. The fit was not quite right.

I am a full professor. I have published in the field’s top outlets. I have been doing this long enough that the rejections no longer sting the way they once did. And yet there are a handful of journals — not the absolute elite, mind you, but solid second-tier outlets where I would genuinely like to place work — where I struggle to get in. I have submitted. I have revised and resubmitted. I have received the polite, terminal no. And in the meantime, I have written somewhere north of thirty reviews for these same journals. Carefully, conscientiously, on time.

So the question I have started asking myself is a simple one: Does that make me a fool?

The honest answer, I think, is that it makes me a rational actor caught inside a system whose incentives most of us have never fully examined or understood. And the economics of why that system works the way it does are worth spelling out — not to complain, but because understanding them points toward something more useful: How each of us should actually behave within it.

Two Markets, One Person

The key insight is that you are not the same economic actor in both situations. When you submit a paper, you are a supplier in a highly competitive tournament. Top journals in most fields accept somewhere between five and fifteen percent of submissions. The rejection rate is not incidental to the journal’s prestige. It is the prestige. Scarcity is the product. In this market, the journal holds all the power, can afford to be ruthless, and has no particular reason to treat any individual author as anything other than one entry in a long queue.

When you are asked to review, you enter an entirely different market. A market that is structurally starved of supply. Peer review is a textbook public good. The benefits are diffuse (science advances, the literature improves), while the costs fall entirely on the individual reviewer (time, effort, opportunity cost). The private incentive to review is weak. The result is chronic undersupply, and editors know it. Getting three qualified reviewers for a single paper can take weeks of emails, rejections, and escalating desperation. The power relationship inverts completely.

In the submission market, you need the journal. In the review market, the journal needs you.

Why You, Specifically

But this still leaves a question. Surely a journal that considers my work below its standard should find reviewers who have published in their pages? Why come back to the serially rejected. Especially when the person asking holds a chair and has the publication record to prove it?

Because the skills required to review are not the same as the skills required to crack a particular journal’s code. To evaluate a paper competently, you need to be fluent in the literature, familiar with the methodology, and capable of identifying weaknesses and strengths. You do not need to have satisfied this specific editorial board. And repeated submissions, however unsuccessful, are actually informative to an editor. They signal that you are active in the area, that you read the journal, and that you take it seriously enough to keep trying. You are pre-screened. You have already raised your hand, several times, in full view.

There is also a more personal reason I say yes, and I may as well be candid about it. Many of the editors who contact me are part of my network. A colleague from a conference, a co-author’s collaborator, someone I have known for years through the circuit of workshops and special issues. Saying yes to their request is not purely altruistic. It is a deposit into a relationship. That kind of relational capital is hard to build and easy to neglect, and in a career that runs on trust and reputation as much as citations, it can be cashed in in ways that matter: a recommendation, a co-editorship, a conversation that opens a door. The economics here are real, even if they are rarely spoken aloud.

So when I review for a journal that has rejected me, I am not simply being exploited. I am also, in part, investing. That is worth being honest about.

The Extraction Problem

And yet, taken in aggregate, something worth naming is still happening. Journals systematically extract free labour from the very pool of scholars their editorial decisions exclude. This is not a conspiracy. Nobody designed it this way. It is an emergent property of a market structure in which publication and reviewing are governed by entirely different incentive logics and in which academics, almost universally, feel a strong normative pull to say yes to review requests.

The asymmetry is especially sharp when the rejected reviewer holds seniority. A full professor who has published at the top of the field is, by any reasonable measure, more than qualified to review for a journal that has found his work wanting. The journal knows this. It is, in fact, exactly why they keep asking. What they are extracting is not marginal competence. It is expert judgment, applied for free, by someone they have declined to recognise as an author.

But here is where I want to resist the obvious conclusion. Because I think it is the wrong one.

The Ledger Looks Different From the Other Side

When I am honest with myself, the picture is more complicated than a simple story of exploitation. Yes, I have given thirty-odd reviews to journals that have not published me. But I have also published in journals where, in retrospect, I have not reviewed nearly enough. The ledger does not balance and not always in my favour.

The peer review system is the only mechanism the academic community has for maintaining the quality of its own literature. It does not work without a sufficient supply of careful, expert reviewers. And the supply is only sufficient if scholars treat reviewing as a genuine professional obligation rather than a favour they dispense when the mood strikes. The system asks something of all of us. The question is whether we are paying our share. Not just to the journals that have been good to us, but to the broader enterprise.

My own view, arrived at slowly and not without some discomfort, is this: for every paper you submit, you should be willing to review three. That is roughly what the arithmetic requires. Each submission draws on the reviewing labour of others; reciprocity demands you return it. And crucially, that obligation does not disappear because the journal in question has rejected you before. The paper you are being asked to evaluate is not responsible for the editorial decisions that stung you. Its author deserves the same careful reading you would want for your own work.

A Note for Younger Scholars

If you are early in your career, the dynamics described above can feel bewildering and occasionally demoralising. You work hard, you submit to the right places, you get rejected, and then the same places ask you to review. It can feel like the system has you at a permanent disadvantage.

It does not. Or at least, it need not.

First, rejection from a journal that keeps asking for your reviews is not a verdict on your competence. It is a signal about fit, timing, and the peculiar sociology of editorial decisions — none of which is fully within your control or fully rational. Senior scholars with strong records face the same wall at certain outlets. That is not consolation exactly, but it is context.

Second, reviewing is one of the most underrated ways to develop as a scholar. Reading work at the frontier of your field, being forced to render a structured judgment on its contribution and its flaws. It makes you a better writer, a sharper thinker, and a more realistic assessor of your own work. The journals that keep rejecting you are, in a backhanded way, offering you something valuable every time they send a manuscript your way. Take it seriously.

Third, say yes when an editor in your network asks. Not strategically, not with a running tally in your head, but with the understanding that academic careers are built as much on relationships as on publications. The editor who remembers you as a reliable, generous reviewer is a different kind of asset than a line on a CV and one that compounds quietly over time.

Fourth, and most importantly, build the habit of reviewing generously and early across the board. Not just for the journals and editors you know, but because the literature you are contributing to is a commons, and commons require stewardship. The scholars whose work you will cite, challenge, and build on deserve careful readers. Be one.

Final Thoughts

So am I a fool for having reviewed thirty times for journals that will not publish me? Perhaps slightly. But the larger accounting error, I have come to think, runs in a different direction. The journals where I have published, and where I have reviewed too little. That is the debt I am more concerned about settling.

The system is imperfect. Its incentive structure is, as I have described, genuinely skewed. But it is the system we have, and its quality depends entirely on whether people inside it choose to behave better than the incentives strictly require.

I am pleased, at least, that Substack has no review process. This piece would never have made it through.

It was 2008. had just been awarded the Nobel Prize in Economics — officially, the Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel — and he came to Copenhagen Business School shortly after. The room was charged with the particular energy that surrounds a newly minted laureate: part intellectual excitement, part celebrity adjacency.

When the floor opened for questions, someone from the political left in the audience pressed him on what many considered the defining challenge of our time: the environment. Climate change. The urgent need to invest in green technology — wind energy was the example — and to subordinate economic logic to ecological necessity.

Krugman’s answer was measured, but it landed like a chess move. Paraphrasing: the dot-com bubble bursting has had severe social consequences — especially for the most economically vulnerable groups in society. Those are the people who should be at the centre of our attention right now.

There was a moment of silence. It was, in retrospect, a checkmate. The political left, which had asked the question, defines itself precisely by its commitment to those vulnerable groups — the unemployed, the precarious, the people for whom an economic downturn is not an abstract statistic but a lost home or a skipped meal. Krugman had taken their own value system and used it to argue for a different set of priorities. Green technology could wait. People could not.

This was not the answer the questioner wanted. It felt, to some in the room, like a sleight of hand — a Nobel economist using the language of social concern to sidestep ecological urgency. But I want to argue that Krugman was doing something more interesting than being dismissive of environmental concerns. He was, perhaps without fully naming it, engaging in one of the deepest and most unresolved problems in economics: the question of how we prioritise among grand challenges.

It is a question that has only grown more urgent since.

What Do We Mean by Grand Challenges?

The term “grand challenges” came into the policy conversation in the 1980s and 1990s, borrowed from mathematics and computer science, where it referred to problems that were well-defined but stubbornly unsolvable. In the social and economic context the meaning is looser — but you know one when you see one.

Grand challenges are large-scale, persistent problems that:

• Affect many people, often across national borders

• Cannot be solved by individual actors or market forces alone

• Require coordinated effort across institutions, sectors, and disciplines

• Play out over long time horizons under deep uncertainty

• Carry high stakes — in welfare, justice, or survival

The canonical list today includes:

Climate change and environmental sustainability — decarbonising the global economy while managing the consequences of warming already locked in.

Global health and pandemic preparedness — as COVID-19 demonstrated with devastating clarity, the world remains structurally unprepared for infectious disease at scale.

Poverty and economic inequality — both within and between nations, despite decades of development economics and billions in aid.

Food security — feeding nine to ten billion people by mid-century without destroying the agricultural systems that make feeding anyone possible.

Education and human capital — the persistent failure to provide quality education to hundreds of millions of children, with compounding intergenerational consequences.

Digital transformation and AI governance — ensuring that the most powerful general-purpose technologies in human history are developed in ways that distribute benefits rather than concentrate them.

Demographic change — ageing populations in rich countries, youth bulges in poor ones, and the fiscal and care systems struggling to absorb both.

Conflict, fragility, and governance failures — the roughly 1.8 billion people living in fragile or conflict-affected states, for whom all other challenges are compounded by institutional breakdown.

This is already an overwhelming list. And it is not exhaustive. Which brings us to the hard question: when you cannot do everything, what do you do?

What Economics Actually Says About Prioritisation

Economics does not produce a ranked list. What it offers instead is a set of lenses — frameworks that each illuminate a different dimension of why prioritisation is hard and what principles might guide it. Let me walk through the most important ones.

1. Opportunity Cost: The Inescapable Logic of Scarcity

The foundational insight is brutal in its simplicity. Resources — money, attention, political will, scientific talent — are finite. Every choice to address one grand challenge is simultaneously a choice not to fully address another. This is what economists call opportunity cost, and it is not a conservative talking point or a counsel of despair. It is simply the structure of the problem.

This matters because a great deal of public discourse about grand challenges implicitly denies scarcity — as if we can and should address everything, fully and simultaneously, if only we have sufficient political will. Economics says: no. Prioritisation is not a failure of ambition. It is a precondition for coherence.

2. Marginal Analysis: The Question Is Never All-or-Nothing

Related to opportunity cost, but subtler: economics frames choices at the margin, not in aggregate. The question is never “should we address climate change or poverty?” — it is “given current allocations, does the next dollar, the next unit of policy attention, generate more value if directed at this challenge or that one?”

This marginal logic matters because the returns to investment in any grand challenge are typically diminishing. The first billion dollars of pandemic preparedness investment buys far more security than the four-hundredth billion. Which means that even if climate change is the most important challenge in the abstract, there may be enormous value to be captured elsewhere at the margin — depending on where current investment levels sit.

3. Market Failure as a Prioritisation Heuristic

A different economic tradition asks a simpler question: where do markets fail most badly? Grand challenges are, almost by definition, problems where market logic breaks down. But the type and severity of the failure differs — and that difference is a basis for prioritisation.

Consider three familiar failures:

Externalities occur when the costs (or benefits) of an action fall on parties not involved in the transaction. Climate change is the paradigm case — carbon emitters do not pay the full social cost of their emissions, so they emit too much. The market systematically underweights the future.

Public goods are non-excludable and non-rival — once produced, no one can be prevented from benefiting, and one person’s use doesn’t diminish another’s. Pandemic surveillance, basic scientific research, and biodiversity all fit this description. Because no one can be charged for benefiting from them, markets chronically underprovide them.

Coordination failures occur when individually rational actors get stuck in a bad equilibrium because they cannot credibly commit to changing their behaviour simultaneously. The clean energy transition is partly a coordination problem: no single firm or country can afford to move first if others free-ride.

The economic logic suggests prioritising where the wedge between private and social returns is largest — where, in other words, the market is most severely misdirecting resources.

4. The Discount Rate: The Deep Ethical Fault Line

This brings us back to Krugman in that lecture hall. His instinct — that immediate social distress deserves priority — is not economically naive. It reflects a high implicit discount rate on future harms: a preference for present welfare over future welfare, and an acknowledgement that the people suffering now are real and visible in ways that future people — who do not yet exist, or whose suffering is still probabilistic — are not.

The discount rate is the rate at which future costs and benefits get shrunk when translated into present-day terms. It sounds technical, but the stakes are enormous. A discount rate of 5% makes a catastrophe 50 years from now look relatively cheap to prevent today. A rate close to zero — as the economist Nicholas Stern argued in his influential 2006 review of climate economics — makes future catastrophe vastly more expensive, and dramatically raises the urgency of acting now.

What looks like a technical squabble over a number is really a philosophical argument about intergenerational justice: how much do we owe to people who haven’t been born yet? Economics frames the question precisely. It doesn’t answer it.

5. The Capability Approach: Whose Welfare Counts?

The standard welfare-economics framework — aggregate costs and benefits, discounted and summed — has a well-known blind spot. A policy that produces large aggregate gains while piling losses onto the already-vulnerable can look excellent on paper.

Amartya Sen’s capability approach offers a corrective. Rather than maximising aggregate welfare, it asks a more human question: are people actually able to live lives they have reason to value? This shifts prioritisation toward the distribution of capabilities, not just their total. A grand challenge that severely limits what a small, marginalised population can do with their lives may deserve more attention than its aggregate welfare weight would suggest.

It also asks who bears the cost of inaction — something aggregate analysis tends to paper over. On climate change, the answer is uncomfortable: the populations least responsible for emissions are generally the most exposed to their consequences.

6. Option Value and Platform Challenges

Finally, there is an insight from the economics of innovation that gets too little attention in grand challenge debates. Not all investments have equal systemic value. Some challenges, if cracked, unlock progress on several others at once.

The development of mRNA vaccine technology — turbocharged by COVID-19 investment — is now a platform for attacking cancer, HIV, and future pandemics. Breakthroughs in energy storage have implications across the whole clean energy transition. Computing infrastructure has generated spillovers into virtually every corner of the economy.

This points toward a prioritisation logic built around platform challenges — problems whose solution generates outsized option value elsewhere. It is a logic that sits awkwardly inside standard cost-benefit analysis, but it may be one of the most important considerations for anyone thinking seriously about long-run resource allocation.

Back to the Lecture Hall

So was Krugman right?

His instinct — social challenges now, environment later — captured something real. The dot-com bust was acute, visible, and hitting people with immediate force. Unemployment concentrates suffering in ways that diffuse environmental degradation, at that moment, did not. And crucially, he was speaking the left’s own language back at them: if your core commitment is to the vulnerable, then here, right now, are the vulnerable. The economic logic of present-biased prioritisation is not indefensible.

But the economics also pushes back. Climate change is precisely the kind of problem where delay is not neutral — where the costs of inaction compound over time, where we are dealing with irreversibilities, and where the gap between private and social returns is among the largest of any grand challenge. The market was not going to fix it while we attended to unemployment. And unemployment, terrible as it was, would eventually respond to conventional macroeconomic tools. Climate change would not.

What Krugman’s answer illuminated — perhaps more than he intended — is something that goes beyond who was right in that lecture hall in 2008. It is that most disagreements about grand challenges are not, at their root, disagreements about facts. They are disagreements about which economic lens to apply: how heavily to discount the future, whose welfare to weight most, where the most severe market failures lie, what counts as a platform investment versus a narrow one.

This matters enormously for how we conduct these debates. When someone argues that climate action should be subordinated to addressing poverty, they are not necessarily being cynical or captured by fossil fuel interests. They may be applying a high discount rate, or a capability-based distributional weighting that foregrounds present suffering. When someone argues the opposite, they are not necessarily being naïve about economic tradeoffs. They may be taking irreversibility seriously, or recognising that the populations most harmed by climate change are precisely the vulnerable groups the left claims to represent.

Understanding which framework your interlocutor is using — implicitly or explicitly — is the first step toward a more productive conversation. It transforms what often feels like a values clash into something more tractable: a structured disagreement about assumptions, weights, and time horizons that can, in principle, be examined and even partially reconciled.

Economics does not resolve the prioritisation problem. But it gives us a shared vocabulary for understanding why we disagree — and that, in debates as fractious as those surrounding grand challenges, may be the most valuable thing it offers. Genuine agreement is more likely to emerge from understanding the structure of a disagreement than from simply shouting louder about its conclusion.

The questioner in that Copenhagen lecture hall and Krugman were not, in the end, as far apart as the moment suggested. They shared the same underlying concern: the welfare of people who are vulnerable and marginalised. They differed on time horizon, on which vulnerabilities were most pressing, and on how much weight to give futures that had not yet arrived.

That is a conversation worth having. Economics shows us how.

You helped organize the conference. You mentored the first-years. You served on the committee nobody wanted. And then, when it mattered, none of it counted. Here is why — and why the system is designed that way.

There is a rite of passage in academia that nobody warns you about.

You arrive as a PhD student, eager, socially attuned, perhaps mildly anxious to belong. You notice that the department runs on invisible labour — someone has to book the seminar rooms, coordinate the reading groups, show up early to set out the coffee. You do these things willingly, even enthusiastically. It feels right. It feels like investment. Somewhere in the back of your mind, a quiet calculation is running: I am building credit here. This will matter later.

It will not matter later.

This is not a complaint about ingratitude, nor a cynical verdict on human nature. It is something more interesting and more structural: a set of interlocking economic mechanisms that make the academic context almost uniquely hostile to the conversion of social capital into career capital. Understanding these mechanisms does not make the experience less frustrating. But it does transform it from a personal wound into a legible institutional pathology — and that shift in framing is, at minimum, more useful.

The Credit That Cannot Be Collected

The first mechanism is what economists call the externality problem, and it operates quietly in almost every department.

When a PhD student invests in the social life of a department — organising events, being the glue that holds a cohort together, volunteering on committees — the benefits of that investment diffuse outward to everyone. Colleagues enjoy a more collegial environment. Junior students feel more supported. The department as a whole functions better. These are real benefits, and they are genuinely produced by the student’s effort.

But here is the problem: the people who absorb those benefits are largely not the people who will later decide whether the student gets a postdoc recommendation, a fellowship endorsement, or a position on the shortlist. Those decision-makers are often elsewhere, operating on different time scales, and they face no particular cost for failing to reward the contributions they never directly witnessed. In the language of economics, your social investment is non-excludable (you cannot restrict who benefits) and non-rival (one person benefiting does not reduce another’s share). These are the hallmarks of a public good — and public goods, left to individual incentives, are chronically undersupplied and systematically under-rewarded.

You produced a public good for your department. The market for public goods does not pay well.

The Problem With Being Nice

There is a second mechanism, more uncomfortable to name but perhaps more important: adverse selection in signalling.

In economics, a signal is only useful if it separates types — if it is costly enough that only the people who genuinely have the underlying quality will invest in it. Publications, in this sense, are (imperfect but real) signals of research ability: they are hard to fake, externally validated, and costly to produce. Service and sociability, by contrast, are cheap signals. They are available to anyone regardless of their intellectual contribution. A brilliant researcher who is also warm and collegial looks, from a distance, exactly like a weak researcher who compensates with charm.

Worse: heavy investment in service can shade into a negative signal in the eyes of some senior evaluators. The implicit reading is not flattering — this person evidently has time for this — which raises questions about the research pipeline. This is not a fair inference, and it is often wrong. But it is a predictable one, given how senior academics allocate their own attention and how they have learned to read the revealed preferences of junior colleagues.

The cruelty here is specific to the early career stage. A tenured professor who serves on university committees is demonstrating institutional citizenship from a position of security. A PhD student who does the same is, structurally, doing something that reads differently — and the reading is not always charitable.

The Hold-Up Problem, or: Why You Had No Leverage

Transaction cost economists have a concept called the hold-up problem, first formalised by Oliver Williamson and developed further by Klein, Crawford, and Alchian in the 1970s. It describes a situation where one party makes a relationship-specific investment before the terms of exchange are secured — and is then exposed to exploitation because the investment cannot be redeployed.

Classic examples involve physical capital: a supplier who builds a factory tailored to a single buyer’s specifications, only to find that the buyer can now renegotiate the price at will, because the factory is worthless to anyone else. But the logic applies with painful accuracy to the PhD student who invests in social integration.

Once you have built a reputation as the helpful, engaged, collegial member of the department, that investment is sunk. It cannot be taken elsewhere. It has no value on the external market. And crucially, senior colleagues — not through any particular malice, but simply through the structure of the situation — have no credible incentive to ensure reciprocation. You have already delivered. There is nothing left for them to secure. The terms of the implicit exchange, which were never written down precisely because they felt too obvious to require writing, are entirely in their favour by the time decisions are made.

This is why the feeling of betrayal that accompanies this realisation is so acute. It is not paranoia. The asymmetry is real. It was baked into the situation from the beginning.

Everyone Benefited, No One Paid

The third mechanism is the most structurally fundamental: a collective action problem of the classic kind that Mancur Olson described in The Logic of Collective Action (1965).

The department, as a collective, benefits from having socially engaged, service-oriented PhD students. The seminars run better. The atmosphere is more generative. The junior students are better mentored. These are genuine, diffuse gains. But no individual senior faculty member has a specific incentive to personally ensure that the student who produced those gains is rewarded for them. Each can reason, quite rationally, that someone else will take care of it, or that the system will recognise it, or simply that it is not really my job.

The result is a multi-player prisoner’s dilemma in which the contributing student is left in what game theorists call the sucker’s payoff — the cell in the payoff matrix occupied by the one player who cooperated while everyone else defected. The tragedy is compounded by the fact that none of the defectors need feel like defectors. They are simply doing what the institutional incentive structure asks of them: advancing their own research agendas, writing their own papers, serving on the committees that advance their own careers.

Institutions, as Olson observed, do not solve collective action problems automatically. They require selective incentives — mechanisms that make individual contribution worth the individual cost. Academic departments, for all their sophistication, rarely have those mechanisms in place for PhD-student service. The culture of recognition is thin, the memory is short, and the feedback loop between contribution and reward is so attenuated as to be functionally broken.

We Forget, Systematically

There is a behavioural dimension too, less celebrated in the economics literature but no less real.

Senior academics who make career decisions — recommendation letters, shortlist opinions, grant endorsements — are not drawing on a comprehensive ledger of everyone’s contributions. They are drawing on a much thinner dataset: recent interactions, salient publications, and whatever narrative about the candidate is most readily available in memory. Daniel Kahneman’s work on the availability heuristic is relevant here: we judge frequency and importance by how easily examples come to mind, not by how accurately we have tracked the underlying base rate.

A paper published last year is available. A committee you served on three years ago is not. The conference you helped organise is not. The first-year student you mentored informally is not. These contributions are not suppressed by bad faith; they are suppressed by the ordinary cognitive economy of busy people making judgements under time pressure.

The PhD student, meanwhile, is operating under what behavioural economists call projection bias — the tendency to assume that others share your current values, time horizon, and sense of what matters. You remember every contribution you made because you made it. You assume it is similarly salient to others. It is not, and the asymmetry in memory is not correctable by goodwill alone.

What To Actually Do About It

None of this is an argument for becoming a bad colleague or withdrawing from the social life of your department. Departments that function well are better environments to develop in, and your contribution to that — even if it goes unrewarded — has genuine value. The people you mentor informally will remember you. The relationships you build may matter in ways that are non-linear and hard to predict.

But there is a distinction worth drawing sharply: between contributions you make because you value them intrinsically, and contributions you make because you expect them to compound into career capital. The former is fine and even admirable. The latter is a systematic miscalculation, and the economics are clear about why.

So here is the practical translation — not cynical, but clear-eyed.

Treat your time budget as a portfolio. Research output is your equity: it appreciates, it is portable, and it is legible to every evaluator you will ever encounter. Service is a current expense — valuable in the moment, but it does not carry forward on any balance sheet that decision-makers consult. A rough rule of thumb: if a service commitment cannot be finished in a few hours, ask yourself whether the same time invested in a paper section would serve your five-year self better.

Make your intellectual work visible early and often. Present at every internal seminar you can. Circulate working papers. Comment publicly on others’ work in workshops — not to perform, but to establish a track record of intellectual engagement that senior colleagues actually witness firsthand. Visibility in research contexts is the one form of impression management that survives the availability heuristic.

Choose service that leaves a trace. Not all service is equal in its legibility. Organising a coffee rota leaves no trace. Co-organising an external conference, editing a special issue, or serving as a discussant for a high-profile seminar creates an artefact — a programme, a publication, a record — that can appear in a CV and be pointed to at the moment of decision. If you are going to invest in service, invest in the kind that converts into something attributable.

Build relationships deliberately, not diffusely. The hold-up problem is most severe when your social investment is spread across the entire department with no particular depth anywhere. A small number of senior scholars who know your work well — who have read your drafts, attended your talks, engaged with your ideas — are worth far more than a wide network of people who simply like you. Depth beats breadth when letters of recommendation are being written.

Name your contributions when the moment arrives. Academic culture discourages self-promotion, which means that the people who most need their contributions remembered are also the least likely to mention them. This is a costly norm to internalise. When updating your CV, when meeting with your supervisor, when being introduced to an external evaluator, make your service contributions part of how you describe yourself — briefly, factually, without apology. Do not assume others are keeping the ledger. They are not.

What compounds in academia is output that is attributable, durable, and legible at the moment of decision. Publications survive institutional forgetting. A reputation for intellectual rigour, built through visible work, survives it too. Pure service, warmth, and social glue largely do not — not because they are less valuable, but because the institutions that would need to remember them are not structured to do so.

The tragedy is that this is a genuine market failure. The academic department, as a community, wants the thing that it systematically fails to reward. The fix is institutional — formal recognition of service in promotion criteria, mentorship logs that enter evaluation processes, department cultures that make collective goods visible at the moments when they count. Most departments do not do this. Until they do, the most honest advice for the idealistic early-career scholar is both blunt and kind: contribute because you want to, not because you expect it to be remembered. Navigate accordingly. The system has a short memory. It is not personal. It is structural.

The diagnosis is by now widely accepted. The Oxford Economics/ICC report published last month put a number on it. Uncertainty alone (not the tariffs themselves) erased the equivalent of $202 billion in global business investment in 2025. If volatility intensifies further in 2026, that figure could reach $380 billion. The IMF, UNCTAD, and the World Economic Forum are unanimous. We are living through a historically unprecedented episode of economic policy uncertainty, and it is costing us dearly.

The prescription that follows seems equally obvious. What the world needs is stable, predictable trade policy. A clear tariff schedule. Rules that businesses can plan around. In other words, certainty.

But here is where economics has something uncomfortable to say. Certainty, as a remedy for uncertainty, only works if it is credible. And credibility, in the technical sense that economists mean, is not something that can be announced. It has to be earned. And it can be destroyed.

The time-consistency problem

In the early 1980s, economists Finn Kydland and Edward Prescott, who would go on to win the Nobel Prize, described a problem they called time-inconsistency. Their insight was simple but devastating. A policymaker who is free to change course at any moment will be rationally disbelieved, even if they are currently telling the truth. Promises about future policy are only worth something if there is a institutional, legal, or reputational mechanism that makes them costly to break.

This is why central bank independence matters so much. It is not that central bankers are smarter than politicians. It is that an independent central bank with a legal mandate and a public inflation target has its hands partially tied. When the Federal Reserve says it will keep rates at a given level, markets believe it. Not out of trust in any individual, but because the institutional structure makes deviation painful and visible.

Now apply this framework to the current situation. The Trump administration has, over the past 18 months, announced, retracted, threatened, paused, escalated, and partially reversed tariff measures on a timeframe that has sometimes been measured in hours. The “Liberation Day” package of April 2025, the subsequent 90-day pause, the country-by-country negotiations, the reimposition of levies. Each of these moves was presented, at the time, as a signal. A firm stance. A new baseline. Each reversal was also presented as a win.

What track record does to credibility

Credibility is a stock, not a flow. It accumulates slowly and depletes quickly. A policymaker who has changed course frequently, who has used policy announcements as negotiating tactics, and who has explicitly framed unpredictability as a strategic asset has, by definition, spent down their credibility stock.

This is not a political judgment. It is simply a description of the information environment that firms face when making investment decisions. When a CEO considers building a factory, hiring workers, or committing capital to a supply chain that crosses a border, they are not just evaluating current tariff rates. They are forming expectations about future tariff rates. And those expectations are shaped, overwhelmingly, by the track record of the policymaker in question.

A tariff announcement from an administration with a demonstrated willingness to reverse course is, from a decision-theory standpoint, not a stable signal. It is a current state that carries a non-trivial probability of changing at any moment. Rational firms respond to this not by investing on the current terms, but by waiting. By deferring. By keeping their corporate wallets, as one economist recently put it, firmly shut.

This is the deeper trap of the current moment. The uncertainty is not just about what the tariffs are. It is about whether any stated policy will still be in place by the time the investment pays off. A five-year capital project cannot be insured against a policy environment that has shown it can reverse in a week.

What credibility restoration actually looks like

History offers a handful of cases where economic credibility was genuinely rebuilt after being badly damaged. They are instructive not because they map neatly onto the current trade situation, but because they share a common structure. And that structure is the opposite of a press conference.

The most studied example is Paul Volcker’s Federal Reserve. By the late 1970s, the Fed’s credibility on inflation was essentially zero. A decade of stop-start monetary policy. Tightening when inflation rose, then easing the moment unemployment ticked up. It taught markets and wage-setters a simple lesson: the Fed will always blink. Inflation expectations had become unanchored because the institution had demonstrated, repeatedly, that it would not pay the cost of holding the line.

Volcker’s response, beginning in 1979, was not to announce a new policy. It was to demonstrate one at enormous short-term cost. The Fed raised rates to nearly 20%, triggering the deepest recession since the Great Depression, and held them there as unemployment climbed past 10%. The point was not the rate itself. The point was the consistency. Month after month, as political pressure to reverse course intensified, the Fed did not move. It was precisely the willingness to absorb punishment that rebuilt the expectation that the policy was real.

The second parallel is more recent and more geographically apt. The ECB’s “whatever it takes” moment in July 2012. At the height of the eurozone sovereign debt crisis, Mario Draghi delivered a three-word commitment that stopped a market panic in its tracks. But it is worth examining why it worked when so many prior statements had not.

It was not the forcefulness of the language. European leaders had made forceful statements repeatedly. What made Draghi’s commitment credible was its institutional architecture. The ECB had a legal mandate, an independent governance structure, and Draghi had secured the backing of the ECB’s governing council before speaking. The commitment was costly to make and costly to reverse. Markets understood, correctly, that the institutional machinery behind the words was real.

What do these cases tell us about trade policy in 2026? They suggest something specific and somewhat uncomfortable. There is no rhetorical shortcut. A statement from an executive branch that has demonstrably treated policy announcements as tactical instruments cannot, by itself, produce the credibility effect that investment decisions require.

The path forward, if there is one, runs through one of two routes. The first is institutional constraint: Congress legislating a tariff framework, bilateral agreements with legally binding dispute mechanisms, or a return to rules-based multilateral structures that impose external costs on reversal. These remove discretion structurally, as Volcker’s mandate removed the Fed’s ability to simply choose a different number. The second is the slower route that Volcker himself ultimately had to travel: demonstrated consistency over time, at political cost, until the expectation of stability becomes earned rather than asserted.

Neither is quick. Neither is painless. But the historical record is clear on one point. The investment that is currently frozen will not thaw in response to an announcement. It will thaw when firms update their expectations and expectations, having been burned, only update on evidence.

The cost of confusing noise for signal

There is a real danger in the current discourse of misdiagnosing the problem. If policymakers and the analysts advising them conclude that what is needed is simply a clearer or more emphatic announcement of stable tariffs, they will find that the investment paralysis persists. Firms are not failing to invest because they have not heard the announcement loudly enough. They are failing to invest because they have learned, empirically, that the announcements do not bind.

The $380 billion in investment that Oxford Economics estimates could be unlocked by a “favourable scenario” of policy clarity is real money. It represents factories, jobs, supply chains, and growth that are sitting frozen because the environment is too uncertain to justify committing to them. Unlocking it requires more than a press conference. It requires the kind of structural credibility that takes time to rebuild and that, for now, the current policymaking architecture is not positioned to provide.

Certainty, in other words, is necessary but not sufficient. What the world is actually short of is something harder to manufacture, Trust that the signal will still mean what it said tomorrow.

The food arrived. Strictly vegetarian. No other option. Not “we have a great vegetarian spread plus some other things”. Just vegetarian, full stop.

Now, I want to be careful here. I am not a fragile person. I can eat a vegetarian meal. The world will not end. But something about that moment felt off in a way I couldn’t quite shake, and on the way home I started to think about why.

The Principle That Wasn’t Shared

For years (decades, really) vegetarians at events, meetings, and institutional lunches have faced exactly this situation in reverse. A spread of meat-heavy dishes, perhaps a token pasta salad, and the quiet implication that their dietary choices were an afterthought at best, a nuisance at worst. The complaint was legitimate. The frustration was fair. Nobody should have to go hungry or feel invisible at a professional lunch because their diet diverges from the majority default.

The argument vegetarians rightly made was fundamentally a fairness argument. Don’t impose your majority preference on everyone. Acknowledge that people are different. Offer a real choice.

That is a good principle. A sound principle. One I agree with entirely (especially from an economics point of view). Which is precisely why I found myself staring at my plate of lentils wondering why that principle had apparently been abandoned the moment the same group found themselves in the majority. Because what happened to me today is exactly what vegetarians spent years objecting to. The default was set. No alternative was offered. The implicit message was: this is what we serve, and if it doesn’t suit you, that is your problem.The logic hasn’t changed. Only the direction of the inconvenience has.

This Is Actually Quite Common Now

What happened at my table is not an isolated quirk. Across universities, business schools, and corporate institutions (particularly in Northern Europe) vegetarian-only or vegetarian-default catering has become increasingly common, often framed as a sustainability measure or a signal of institutional values.

The trend is real and accelerating. Plant-based menus have moved from niche accommodation to mainstream default in many professional settings. Some institutions have gone further, eliminating meat options entirely from officially catered events. And there are genuine arguments for this direction. Meat production carries significant environmental costs (greenhouse gas emissions, land use, water consumption) that are not reflected in the price of a chicken sandwich. If institutions want to reduce their carbon footprint, shifting catering is a legitimate lever. The environmental case is real.

However, there is a difference between a principled environmental policy and simply flipping who gets to feel excluded. And not everyone making these decisions is being honest about which one they are actually doing.

What Economics Has to Say

Standard economics offers a useful framework for thinking through the morning situation I experienced and why “strictly vegetarian, no options” is actually a poor policy even on its own terms.

On externalities: Yes, meat has environmental costs that the market doesn’t fully price in. This creates a legitimate case for intervention — a corrective price signal, a carbon tax, making meat available but more expensive to reflect its true cost. These are economically coherent responses. Eliminating choice entirely, however, is a blunt instrument that corrects the externality by ignoring the welfare of the people sitting at the table.

On preference sovereignty: Economics is built on a foundational respect for individual preferences. People’s choices are generally the best signal we have of what makes them better off. A well-designed policy would price in the externality and let people choose. Not simply erase the option and declare the problem solved.

On default effects: Behavioral economics tells us that defaults are enormously powerful. Nudge theory, developed by Thaler and Sunstein, shows this clearly. The honest version of what institutions are doing when they set a vegetarian default is to use the power of inertia to change your behavior. That may sometimes be justified. But it should be acknowledged for what it is, not dressed up as mere inclusivity.

On the implicit social contract: The lunch was offered after I had already agreed to participate and deliver. It wasn’t compensation. It was a social offer extended in good faith, which I accepted in good faith. The reasonable expectation, in any normal professional context, is that “lunch” means something edible for the person being invited. Discovering after the fact that it came with an undisclosed dietary condition is a small but real breach of the implicit terms.

On efficiency versus equity: A strictly vegetarian lunch may score well on some efficiency metrics. But efficiency that consistently produces the same losers is not a policy. It is a transfer. When one group’s inconvenience is treated as a legitimate grievance and another’s is not, that asymmetry is not economics. It is politics dressed as principle – bordering activism.

The Smartest Version of This Policy

The best version of vegetarian-forward catering doesn’t look like what I experienced today. It is a menu that is predominantly plant-based, thoughtfully designed, genuinely delicious but with a real meat or fish option available for those who want it. This approach captures most of the environmental benefit, signals the institution’s values, and respects the people in the room enough to give them a choice.

Some caterers are already moving toward a “plant-forward” approach where vegetables and grains are the center of the plate, not a concession on the side, but where nobody is locked out. That is both more inclusive and, frankly, better food (of course my subjective opinion)

What doesn’t work (logically, ethically, or economically) is strict exclusion justified by the memory of having once been excluded. That’s not a principle. That’s just revenge with a sustainability label on it.

A Final Thought

The vegetarian argument for inclusion was always at its strongest when it appealed to universal fairness — everyone deserves a real option. That argument had genuine moral force. It moved institutions. It changed menus. It was right.But it on ly retains that force if it actually means everyone. The moment it becomes everyone like us, it has abandoned its own foundation.

I left lunch slightly hungry and more than slightly thoughtful. The lentils were fine. The principle had gone missing somewhere between the kitchen and the table.

That’s the thing about fairness. It doesn’t get to take the afternoon off just because your side is now setting the menu.

There was a time when entering politics required at least a passing acquaintance with the art of making an argument. You needed to be able to construct a sentence, follow a chain of reasoning to its conclusion, and — crucially — notice when your own logic had gone off the rails. That time, if it ever truly existed, appears to be behind us.

What we are witnessing across Western democracies is not merely a decline in the quality of political debate. It is something more fundamental: a growing tolerance for statements that are not just wrong, but structurally incoherent — claims that fail not at the level of evidence, but at the level of basic logic. And what is perhaps most alarming is not that politicians make such statements. It is that almost no one calls them out for it.

A Gallery of the Incoherent

Politicians have always stretched the truth. But there is a meaningful difference between a politician who exaggerates a statistic and one who makes a claim that is self-contradictory on its face.

Consider the genre of statement that has become depressingly familiar: the politician who argues simultaneously that a problem is so severe it demands emergency action and that their policies have already solved it. Or the official who insists that a spending programme will “pay for itself” through growth — while also insisting that the very same growth proves the programme is unnecessary. Or the leader who warns of the dangers of foreign interference in elections while actively soliciting it.

These are not nuanced positions that require unpacking. They are logical contradictions. Yet they pass through press conferences, parliamentary debates, and television studios largely uncontested. The journalists move on. The opponents seize on the optics rather than the reasoning. The public, exhausted, scrolls past.

The standard has dropped so far that we have almost ceased to notice.

The Rubin Case: A Masterclass in Muddled Thinking

The most favoured and obviuos case for illustrating the issue is to be found in the current American presidency (You-Know-Who) and his administration. But lets use a different example for the sake of novelty. Indeed, lets consider a statement recently made by the Danish politician Monika Rubin for a particularly vivid illustration of how badly political logic can go wrong and how little it tends to matter.

Rubin was commenting on a genuinely troubling phenomenon: a pattern of Danish police shelving rape cases, with the problem apparently more pronounced in certain regions — specifically, on the island of Funen — than in Copenhagen. This is a real and serious issue that deserves serious engagement.

Her conclusion, however, was remarkable. She logically deduced from the forwarded regularities that she would rather be raped in Copenhagen than on Funen, given the differential likelihood that her case would actually be investigated.

Let us be precise about why this is not merely an unfortunate turn of phrase, but a logical failure.

Rubin was presumably trying to make a point about institutional accountability — that the Copenhagen police are more likely to pursue rape cases than their counterparts on Funen. That point, if true, is worth making. But the conclusion she drew from it is inverted. What follows from the observation that cases are more likely to be shelved on Funen is not that one would prefer to be victimised there — a preference no rational person could hold — but rather that a perpetrator would find Funen more operationally convenient. The calculation of “where would I rather this crime happen to me” yields a nonsensical result; the calculation of “where would a criminal find less institutional resistance” is the coherent version of the argument.

Put plainly: if you are trying to highlight institutional failure in a region, the logical statement is “I would rather commit the crime on Funen, if I were a criminal seeking impunity.” Not “I would rather be the victim there.” These are opposite positions. One is a critique of the system; the other is a bizarre expression of victim preference that makes no sense at all.

The statement should have been immediately challenged. It was not a matter of political interpretation — it was simply wrong in its internal logic. That it apparently passed without significant scrutiny is itself the story.

So Why Does This Keep Happening?

The charitable explanation is simple incompetence. Some politicians are not very bright, or speak before they have finished thinking, or have communications teams that care more about soundbites than coherence. This is doubtless part of the answer.

But it is not the whole answer. When a phenomenon is as widespread and persistent as this one, it rewards a more systematic explanation. Economics, perhaps surprisingly, has several to offer.

Public Choice Theory: The Incentives Are All Wrong

Public Choice theory, developed by economists such as James Buchanan and Gordon Tullock, applies the tools of economic analysis to political behaviour. Its central insight is that politicians are rational actors responding to incentives — and the incentives in democratic politics are not well aligned with the production of careful, logical argument.

Voters do not, on average, reward logical rigour. They reward emotional resonance, tribal affiliation, and the appearance of conviction. A politician who says something confused but passionate will often outperform one who says something precise but dry. The market for political ideas does not clear on the basis of logical quality; it clears on the basis of emotional impact.

Moreover, the costs of making an illogical statement are low. Most voters will not notice. Journalists are under time pressure and may lack the specialist knowledge to identify the error. Opponents may prefer to score political points in other ways. In this environment, the rational politician has little incentive to invest in careful reasoning — and considerable incentive to invest in memorable, emotionally charged rhetoric, even if it makes no strict logical sense.

The Attention Economy and the Clickbait Effect

Modern political communication does not take place in the measured environs of parliamentary debate. It takes place on social media platforms optimised for engagement — which is to say, optimised for outrage, surprise, and emotional intensity.

In this environment, a provocative or unusual statement — even a deeply confused one — generates attention. Monika Rubin’s statement about preferring to be raped in Copenhagen almost certainly attracted more coverage than a carefully worded, logically coherent analysis of regional variation in Danish police case closure rates would have done. From a pure attention-maximisation standpoint, the illogical statement worked.

This creates a perverse dynamic. The very features that make a statement logically problematic — its exaggeration, its inversion of normal reasoning, its emotional register — are the features that make it travel. The attention economy does not penalise incoherence. In many cases, it rewards it.

Adverse Selection: The Wrong People in the Room

Adverse selection, a concept from information economics, describes the process by which, under conditions of imperfect information, lower-quality actors tend to crowd out higher-quality ones. The canonical example is the used car market: because buyers cannot distinguish good cars from lemons, they are only willing to pay average prices, which drives owners of good cars out of the market, leaving a higher proportion of lemons.

Something similar may be occurring in democratic politics. The conditions of modern political life — the media intrusion, the social media abuse, the salary compression relative to comparable private sector roles, the brutal hours, the permanent public exposure — are increasingly unattractive to people with strong outside options. The highly capable lawyer, economist, or policy expert can earn more, suffer less, and exercise more genuine intellectual influence outside of politics than within it.

The result, over time, may be a political class that is increasingly selected against cognitive quality. This is not universal — there are still brilliant people in politics — but the structural pressures are real. And a legislature or cabinet that has been filtered, even partially, for willingness to tolerate degrading conditions rather than raw intellectual capacity will, on average, produce lower-quality reasoning.

Moral Hazard: No One Bears the Cost

Moral hazard arises when an actor does not bear the full consequences of their decisions and therefore takes more risk than they otherwise would. In financial markets, it explains why banks that expect to be bailed out take excessive risks. In politics, it helps explain why politicians make claims they could not sustain under scrutiny.

If a politician makes a logically incoherent statement and faces no meaningful consequence — no lost votes, no media censure, no damage to their career — they have no strong incentive to be more careful next time. The cost of logical failure has been externalised. It is borne by public discourse generally, by the quality of democratic deliberation, by the citizens who make decisions based on confused information — but not by the politician who introduced the confusion.

The absence of robust fact-checking institutions, the fragmentation of media, the collapse of a shared epistemic baseline — all of these reduce the likelihood that logical failures will be identified and penalised. Moral hazard thrives.

Game Theory: A Race to the Bottom

Finally, game theory illuminates how individually rational behaviour can produce collectively irrational outcomes. Consider a simple model: if Politician A maintains high standards of logical rigour while Politician B deploys emotionally resonant but incoherent messaging, and if voters respond more strongly to emotional resonance than to logical quality, then Politician B wins. Politician A, observing this, has an incentive to lower their own standards. The result is a Nash equilibrium in which all politicians converge on low logical quality, even though all of them — and certainly the public — would be better off if high standards were maintained across the board.

This is a coordination problem. Individual actors, responding rationally to the incentive structure, produce an outcome that no one would have collectively chosen. And like most coordination problems, it is difficult to escape without some external mechanism — robust institutions, a demanding press, an informed and demanding public — that changes the payoff structure.

What Is To Be Done?

This article has been more diagnostic than prescriptive, and deliberately so. The causes of political illogic are structural, not merely personal — they cannot be fixed by electing better individuals into a system that selects against rigour and rewards incoherence.

But a few observations seem worth making.

First, the press has a role here that it is not consistently filling. Pointing out that a politician’s statement is factually contested is now standard practice. Pointing out that it is logically invalid — that the conclusion does not follow from the premises, that the argument is internally contradictory — is rarer, and arguably more important. A claim can be factually accurate and still be a non sequitur. Journalists need the tools and the confidence to say so.

Second, citizens bear some responsibility too. The appetite for emotionally satisfying nonsense is not something that happens to electorates — it is, to some degree, something they participate in. A public that rewards clarity and penalises confusion at the ballot box would change the incentive structure for politicians overnight.

Third — and perhaps most importantly — we should stop treating logical incoherence in political speech as merely aesthetic, a matter of style or communication skill. When a politician makes a claim that is structurally invalid, they are not just speaking badly. They are contributing to an environment in which citizens cannot reason clearly about the choices in front of them. That is a democratic harm, not a rhetorical one.

Monika Rubin did not mean to say something absurd. She was trying to make a real point about a real problem. But the fact that her illogic went largely unchallenged is not a small thing. It is a symptom of a broader failure — of scrutiny, of standards, of the basic expectation that people who hold public power should be able to think clearly about the things they are speaking about.

That expectation seems worth recovering.

For years, Denmark has occupied an almost mythical place in international political debate. Especially in the United States, the Danish welfare model has often been presented as proof that a highly taxed and highly redistributive society can also be prosperous, innovative, and socially cohesive. Politicians such as Bernie Sanders repeatedly pointed to Denmark and the broader Nordic model as evidence that universal healthcare, free education, strong labor protections, and extensive social security are not only morally desirable, but economically sustainable.

And to a remarkable extent, Denmark has justified that admiration. The country consistently ranks among the happiest societies in the world. Corruption remains low, institutions are generally effective, and social trust has historically been extraordinarily high. Danes tolerate tax levels that would be politically impossible in most countries because the system has long delivered tangible results in return: functioning infrastructure, universal healthcare, childcare, education, pensions, and a relatively stable social fabric.

But beneath the success of the Danish welfare state lies a reality that international admirers often overlook. The system does not merely depend on taxation. It depends on trust. Not only political trust, but administrative trust.

Something Rotten in Danish Property Taxation

A welfare state on the scale of Denmark’s cannot function through enforcement alone. No democratic government can sustainably maintain Nordic levels of taxation if large parts of the population begin to believe that the system itself is arbitrary, incompetent, or fundamentally unfair. The Danish model works because citizens broadly trust that taxes are calculated correctly, that institutions behave predictably, and that similar citizens are treated equally under the law. In many ways, the true foundation of the Danish welfare state is not taxation itself, but legitimacy. That legitimacy is now beginning to fracture. And nowhere is this more visible than in Denmark’s property tax system.

For more than a decade, Denmark has struggled with one of the largest administrative failures in modern Danish governance: the collapse and attempted reconstruction of the public property valuation system. Property taxes in Denmark are heavily dependent on state-calculated assessments of both the total value of a property and the value of the land beneath it. These assessments determine major taxes such as ejendomsværdiskat and grundskyld, making the entire system dependent on the state’s ability to produce accurate and credible valuations across hundreds of thousands of homes.

In theory, the system appears rational. The state estimates market values and taxes property owners accordingly. In practice, however, the system gradually evolved into a bureaucratic disaster. The original valuation model was eventually found to contain widespread inaccuracies. Some homes were dramatically overvalued while others were substantially undervalued. Similar neighboring properties could receive radically different assessments without any obvious explanation. The political response was extraordinarily ambitious: Denmark would replace the old system with an entirely new national valuation model built on modern data analysis, automated calculations, and large-scale digital infrastructure.

The reform, however, exposed a deeper problem inside the Danish administrative state. The new valuation system relied on enormous quantities of underlying data drawn from housing registries, zoning records, transaction histories, land classifications, renovation data, and geographical analysis. Yet much of this data turned out to be inconsistent, outdated, incomplete, or simply incorrect. The Danish state attempted to construct a highly sophisticated tax system on top of administrative foundations that were far less reliable than policymakers had assumed.

The consequences have been severe. Years of delays, provisional tax bills, retroactive recalculations, disputed assessments, legal uncertainty, and widespread confusion have undermined confidence in the system. Many homeowners no longer understand how their own taxes are calculated. Others have watched valuations fluctuate dramatically while the authorities themselves repeatedly revise the methodology behind them. This is not merely a technical issue. It is a political and institutional problem of far greater significance.

The Real Danger

The danger is not that Denmark will suddenly run out of tax revenue. The real danger is the gradual erosion of institutional trust inside a society whose entire welfare model depends on precisely that trust.

High-trust societies possess a hidden vulnerability. They are remarkably resilient as long as institutions function competently, but when failures persist over long periods of time, the psychological damage can become disproportionately large. Citizens in low-trust societies already expect dysfunction from public institutions. Citizens in high-trust societies do not. When repeated failures occur inside systems that were previously regarded as competent and fair, confidence can deteriorate far more rapidly than policymakers expect. That deterioration is now increasingly visible in Denmark.

For decades, SKAT represented one of the central pillars of the Danish state’s administrative legitimacy. Today, however, many homeowners increasingly view the tax authorities as opaque, inconsistent, and difficult to challenge. The property tax system has become so complex that ordinary citizens often struggle to understand not only the calculations themselves, but whether the authorities genuinely understand them either. That distinction matters enormously.

The growing frustration is not merely caused by complexity or delays. It is increasingly driven by a perception of institutional overreach. In some cases, homeowners who have been unable to pay the correct taxes because the authorities themselves failed to provide accurate valuations may ultimately not only face retroactive tax bills, but also interest payments on amounts that could not properly be calculated in the first place.

From a purely administrative perspective, such interest charges may appear technically defensible. But politically and psychologically, the consequences are far more serious. Many citizens do not experience this as a neutral bureaucratic procedure. They experience it as the state asking ordinary homeowners to absorb the financial consequences of failures created by the state itself. That perception is extraordinarily dangerous in a high-trust society.

The legitimacy of the Danish tax system has historically rested on a widespread belief that the authorities act competently, proportionally, and fairly. Once citizens begin to feel that institutions are no longer merely ineffective, but increasingly unreasonable or insulated from the consequences of their own mistakes, frustration can evolve into something much deeper: resentment.

Because the Danish welfare state ultimately relies less on coercion than many outsiders imagine. Compliance at Nordic tax levels requires something far more fragile than legal enforcement. It requires a broad social belief that the system is fundamentally fair, predictable, and competently administered. Once that belief begins to weaken, the long-term risks become far more serious than a failed IT project or a delayed tax reform.

International admirers of the Danish model often focus on the visible outcomes: free healthcare, strong welfare protections, social stability, and low poverty. But they frequently overlook the fragile institutional culture beneath those achievements. The Danish system is not self-sustaining. It depends on competent administration, transparency, credibility, and a population willing to trust the institutions governing them. Unless convincingly addressed, the property tax crisis may ultimately reveal what happens when those foundations begin to crack.

This does not mean that the Danish welfare state is on the verge of collapse tomorrow. Denmark remains one of the best governed societies in the world by most international standards. But the crisis should nevertheless serve as a warning. The greatest long-term threat to the Danish welfare model may not come from ideological attacks, globalization, or even fiscal pressure.It may come from something far more dangerous: the slow erosion of trust between citizens and the institutions that ask them to pay for the entire system.

Notes from a Strategy Scholar

Why Management Research Is Especially Vulnerable to Academic Activism

Why has management and organization research become such fertile ground for academic activism, particularly of the left-leaning, social justice-oriented, “critical theory” variety…

Read more

4 months ago · 2 likes · Nicolai J Foss

The field’s fragmentation, its normative ambitions, and its weak evaluative hierarchies all create conditions in which activist scholarship can thrive. But there is a mechanism missing from the account — one that helps explain not just why the structural conditions exist, but why individual scholars, even those who privately value rigor and neutrality, so often fail to resist the drift. That mechanism is the collective action problem.

The Logic of the Collective Action Problem

Collective action problems arise whenever a group of individuals would collectively benefit from a certain outcome — in this case, a field that maintains rigorous, politically neutral standards of scholarship — but where each individual has a rational incentive to free-ride rather than bear the personal costs of contributing to that outcome. The benefit, if achieved, accrues to everyone regardless of who did the work to produce it. The cost, by contrast, falls squarely on the individual who acts.

In the context of management research, this dynamic plays out with considerable force. Suppose that a majority of scholars in a given subfield privately believe that the balance has tilted too far toward advocacy, that activist commitments are distorting the selection and interpretation of evidence, and that the field’s credibility is at stake. Each of those scholars faces a choice: speak up, or stay quiet.

Speaking up is costly. The costs are not hypothetical. Peer review is anonymous, but editorial networks are not. A scholar who publicly questions activist scholarship risks being seen as politically retrograde, unsympathetic to important social concerns, or simply difficult. In a field where careers depend on publication, and publication depends on the goodwill of editors and reviewers drawn from overlapping networks, the social and professional costs of dissent are real and foreseeable. Staying quiet, by contrast, is safe. And if enough others speak up — if the problem gets corrected without one’s own intervention — so much the better. The rational individual waits.

The problem, of course, is that everyone reasons this way. The result is a field in which a silent majority may privately share concerns but none of them act on those concerns publicly. The activist minority, who bear personal costs willingly because of strong ideological commitment, face no organized resistance. They shape journals, hiring committees, and conference agendas not because they have persuaded the majority, but because the majority has opted out of the contest.

When the Collective Action Problem Is Especially Severe

The collective action problem does not arise with equal force in all settings. Several features of management research make it particularly acute, and they map closely onto the structural vulnerabilities Foss identifies.

First, the costs of dissent are high when reputational networks are dense and long-lived. Management research is a relatively small world. The same scholars serve on editorial boards, review grant applications, sit on hiring committees, and see each other at the same annual conferences year after year. This repeated interaction raises the stakes of any single act of public disagreement. Unlike a discipline where one might publish a critical piece and retreat into the anonymity of a large, impersonal field, management scholars know that they will encounter the targets of their criticism again — and that those individuals may soon be reviewing their next manuscript or evaluating their next grant application. When the shadow of the future is long, the cost of speaking out is amplified.

Second, the problem intensifies when evaluation criteria are subjective and contested. In a discipline with strong shared standards — where it is relatively clear what counts as a rigorous contribution — a dissident scholar can appeal to those standards as a neutral basis for critique. The argument “this paper does not meet the field’s methodological standards” is harder to dismiss as politically motivated than the argument “this paper reflects ideological bias.” Management research, with its coexisting methodological traditions and its chronic ambiguity about what constitutes theoretical contribution, lacks this safe ground. Critics of activist scholarship cannot easily appeal to shared standards that command wide allegiance, because those standards are themselves contested. The cost of critique is therefore higher, and the risk of being dismissed as a mere ideological opponent is greater.

Third, the collective action problem is compounded by the field’s fragmentation. In a more integrated discipline, a critical mass of like-minded scholars can more easily find each other, coordinate, and act collectively. Fragmentation makes this harder. Scholars who share concerns about rigor may be spread across strategy, organizational behavior, entrepreneurship, and international business, each operating within its own journals and conference circuits, with limited occasion to discover that their private reservations are widely shared. Without that discovery, the coordination necessary to overcome the free-rider problem never materializes.

Fourth, the “grand challenges” framing that Foss rightly identifies as a driver of activist drift also raises the cost of dissent in a specific way. When research is framed around addressing climate change, inequality, or injustice, to question the political orientation of that research is easily recast as indifference to those problems. The moral urgency of the agenda provides activists with a rhetorical tool: skeptics can be characterized not as defenders of scholarly rigor but as opponents of socially important goals. This asymmetry in how critique can be framed further suppresses the willingness of individual scholars to voice their concerns.

When Individuals Escape the Collective Action Problem

The collective action problem is not inescapable. Certain conditions reduce its force for individual scholars, making dissent more likely and less costly.

Tenure is the most obvious. The formal protection of tenure was designed precisely to reduce the cost of intellectual dissent. A tenured scholar cannot be dismissed for publishing unpopular views, which removes the most severe professional sanction that conformist pressure can impose. To the extent that collective action problems suppress dissent in academia, they should be most severe among junior scholars — graduate students, postdocs, and assistant professors — for whom career vulnerability is greatest. Senior scholars with tenure face lower personal costs and should, in principle, be more willing to resist.

Elite reputations provide a related form of insulation. A scholar whose work is widely recognized as foundational to the field is harder to dismiss or marginalize than one who is still building a record. The social costs of attacking an eminent scholar publicly are higher for the attacker than for the target. This asymmetry means that scholars with strong reputations can bear the costs of dissent more easily — not because those costs disappear, but because the countervailing weight of their standing provides a buffer.

Cross-disciplinary identity offers a third escape route. Scholars whose primary intellectual home is in an adjacent discipline — economics, psychology, or sociology in its more positivist traditions — are less dependent on the approval of the management research community. They publish in different journals, attend different conferences, and build reputations in networks where different standards apply. This independence reduces the cost of dissent within management research specifically, because the most important professional sanctions are administered by a community to which the cross-disciplinary scholar is only loosely attached.

Finally, publishing platforms outside the traditional peer-reviewed system — blogs, newsletters, Substack — lower the cost of dissent in a structural way. In the peer-reviewed system, the same community whose norms one is challenging controls access to publication. Outside that system, the gatekeepers are absent. A scholar can reach a substantial audience without submitting to the social filtering mechanisms that ordinarily suppress heterodox views. The reputational costs do not vanish, but they are no longer backed by the power to block publication.

The Structural and the Individual: Completing the Account

The collective action lens complements rather than replaces Foss’s structural account. Fragmentation, normative ambiguity, and weak evaluative hierarchies create the conditions in which activist scholarship can capture subfields. But the collective action problem explains why those conditions persist even when many individuals are privately dissatisfied with the outcome. The two arguments are mutually reinforcing: structural fragmentation raises the cost of coordinating resistance, while the collective action problem ensures that resistance remains atomized and ineffective. Together, they describe a field in which the equilibrium can remain far from what most participants would choose if they could coordinate.

This also has implications for what kinds of interventions might help. Structural remedies — clearer evaluative standards, more integrated subfield communities, greater methodological consensus — address the root conditions that Foss identifies. But remedies targeted at the collective action problem itself are also possible. Platforms that allow scholars to express private concerns publicly without individual attribution could lower the cost of dissent. Associations or working groups that allow like-minded scholars to discover each other and coordinate could help overcome fragmentation. And the development of external reputational anchors — cross-disciplinary connections that reduce dependence on the approval of any single community — could make individual resistance more feasible.

Foss as a Case in Point

It is worth noting, finally, that Nicolai Foss himself exemplifies the conditions under which the collective action problem loses its grip. He is a senior scholar with tenure, whose foundational contributions to strategy and organization theory have given him a reputation that is not easily dismissed or marginalized. His intellectual identity spans disciplines — rooted in economics and organizational economics as much as in management — which means that his professional standing does not depend solely on the goodwill of the management research community whose norms he is challenging. And he is writing on Substack, outside the peer-reviewed system, where the editorial gatekeepers who might otherwise filter his views have no role.

These are not incidental features of his situation. They are precisely the conditions that make the costs of dissent bearable. Foss is not free of all professional risk — no one who challenges entrenched norms is. But the structural features of his career and platform mean that he faces those costs from a position of relative strength. That he uses that position to say what many others are privately thinking, but feel unable to say, is not a small thing. It is, in fact, a vivid illustration of the problem he is diagnosing: the silence of the majority is not evidence of their agreement, but of the rational calculation that speaking has costs and waiting is free. Foss has chosen not to wait. That choice, and the conditions that make it possible, deserve to be recognized as part of the argument he is making.

This points to a broader and sobering implication. The value of internal critics like Foss is not only that they are knowledgeable and constructive — it is that they have a stake in the field’s health. An insider who challenges activist drift does so because he cares about the quality and credibility of management research as a discipline. His critique is aimed at improvement, not demolition. But if insiders remain collectively silent, the task of challenging politicised voices does not simply go undone. It falls, eventually, to external agents: policymakers questioning the value of academic funding, journalists skeptical of university research, politicians targeting institutions they perceive as ideologically captured, or simply a broader public that stops taking management scholarship seriously. These external critics have little to lose from being unfair, reductive, or indiscriminate in their attacks. They are not trying to save the field — they are indicting it. The irony is that the collective silence of insiders, individually rational as it may be, creates the very conditions that make external, far less sympathetic, scrutiny more likely and more damaging. Scholars like Foss, by speaking out from within, are not just defending intellectual standards. They are protecting the field from a reckoning it would be far less equipped to survive if it came from outside.