Then government did what government does with anything that works. It made it legible. Roles became professions, professions grew capability frameworks and career paths, teams took on recognisable shapes, and those shapes became easy to recruit, grade, fund, report and hold to account [^1]. All the necessary, governmenty things that gave digital capability a permanence in Whitehall it would never have had otherwise.

But legibility has a cost. Because whilst we have become very good at assembling teams capable of running services, improving them and making them better over time, we have become much worse at preserving the judgement, trust, permission and relationships that made those teams transformational in the first place.

This matters now. Because if we’re talking seriously about rewiring the state, rebuilding government around outcomes, missions or whole services, that work will not stay where we put it. It will cross professional boundaries, organisational boundaries and the lines between delivery and policy itself. This type of transformation needs teams with permission to follow problems, agency to act on what they learn, and enough flexibility to change shape as their understanding changes. And these are exactly the conditions the machinery of government is least well set up to provide.

So, are multidisciplinary teams still the right foundation for the kind of transformation we now say we want? Or have the structures we built to break the silos quietly become silos of their own?

Expanding how we think about working together

In the quest for legibility at scale within government, multidisciplinary working has become something of a shorthand for an entire philosophy of empowered working. Partly, that is a legacy of what the original GDS teams made look so easy. Relational, integrated, flexible, trusted, able to cross a boundary when the work asks for it. All qualities that became bound up with what good multidisciplinary working in government was supposed to look like. But it also tells us something about how government scales what works. We put enormous effort into the parts we can describe, formalise and reproduce, building line management, funding routes, career paths and professional boundaries around the separate forms of expertise that make up a team.

The consequence, however, is that the version of multidisciplinarity government has become best at reproducing is also the most legible one. It is the definition of multidisciplinary that has scaled most faithfully – specialists working alongside one another while retaining their own disciplinary approaches and perspectives [^2] – rather than all the characteristics we came to associate with those original teams. Which is a considerably smaller promise than the one we’ve been making.

This difference is more than semantic, and becomes uncomfortably obvious when you see it drawn. This typology of relationships between disciplines comes from a study on how large, multi-institutional research programmes organise themselves around complex problems [^3]. And there, multidisciplinary – the word government has loaded with a much broader philosophy of empowered working – is almost brutally simple. Separate disciplinary columns, each with its own goal, sitting under a shared thematic umbrella.

Put that beside the shape of a mature government service team and the resemblance is hard to miss. Teams of researchers that research. Designers that design. Analysts that analyse. Policy that sits somewhere else. Everyone doing their own job well, more often than not still bloody brilliantly. But I think this is the trap we’ve fallen into. Government has learned to organise the work around the professions we’ve assembled, rather than organise those professions around the outcome. And when the work has to travel between those territories, every handover creates another opportunity for tacit knowledge, learning and perspective to get lost along the way. Familiar, right?

Where these diagrams become even more useful is in giving us a vocabulary we’ve perhaps been missing. Participatory brings people from beyond the disciplines into the work and creates knowledge exchange, while the disciplinary knowledge itself remains largely separate. Interdisciplinary goes further, integrating knowledge across disciplinary boundaries. And then transdisciplinary crosses boundaries between disciplines and sectors, forms goals in common and produces knowledge together rather than passing it between established professional territories. It looks much closer to the kind of movement rewiring the state seems to demand, and we have never really had a word for it in government.

So while multidisciplinary teams can contain excellent researchers, designers, technologists, analysts and product people, multidisciplinarity itself only tells us that those forms of expertise are present. It does not tell us whether knowledge is integrated, whether authority can move, whether the team can cross organisational or policy boundaries, or whether its shape can change when the problem changes.

All of which points at an obvious answer. If multidisciplinary is too small, perhaps what government needs is transdisciplinary teams. If only it were that simple.

When the work moves

One of the clearest things I’ve learned from working on service and policy problems across government is that the work rarely stays in the place, profession or part of the system where it first appears. A citizen cannot complete something, so the obvious place to start is the service. But the service works the way it does because of an operational rule. The rule exists because a particular piece of evidence is required. That evidence requirement encodes a judgement about risk, and that judgement was made somewhere else entirely against a policy intent nobody in the room wrote. Again and again, somewhere along that chain, the thing that first presented itself as a service problem stops being one.

Put simply, the work moves. Horizontally across different kinds of expertise, from design into operations, evidence, analysis and policy, which is the movement much of the literature on cross-disciplinary working helps us understand.

But it also moves vertically, through layers of institutional authority. From an interaction to a process, from a process to a rule, from a rule to the judgement or policy that put it there in the first place. And this is the tricky part for multidisciplinary teams. They can contain every relevant profession and still have no route through those layers. Nobody is missing from the team, but the authority to follow the problem sits somewhere else.

This is where transdisciplinary working starts to act as a genuinely useful lens. It gives us a way of thinking about expertise organised around the problem rather than handed between professional territories, crossing disciplinary and sectoral boundaries where the work demands it.

What this ends up looking like in practice is much closer to the kind of working those original GDS teams made familiar. Messy, relational, crossing boundaries and organised around an outcome rather than a fixed set of professional territories. Different combinations of people form around different parts of the problem, some lasting, some temporary, some crossing disciplines and others staying firmly within them. The architecture gets denser as the work develops because understanding a problem creates new questions, and those questions demand different combinations of expertise and relationships.

What that illustrates, really, is that the team you planned for does not always end up being the team the work needs. As the work evolves and the conditions around it change, the useful mix of skills, relationships and authority changes with it. And if the work itself cannot be fully specified upfront, neither can the architecture required to do it. That puts our business as usual governance in an awkward position, because much of the machinery we use to organise, fund and assure work assumes its shape can be known at the start. The evidence has to be able to change the structure as the work develops.

This is what the authors of these diagrams call architectures of adaptive integration. The structures and processes, formal and informal, through which different groups work together on a shared problem over time. And the point of the phrase is the movement in it. Relationships form and dissolve, different expertise becomes important at different moments, and the architecture keeps adapting to what the work is teaching.

Building depth that travels

The obvious answer at this point would be to go and build transdisciplinary teams across government. But the reality of transformation work is messier than that. Disciplinary, multidisciplinary and transdisciplinary work happens at the same time, often inside the same programme, because different questions demand different relationships between expertise. Some of it needs deep specialist craft. Some of it needs several professions working alongside one another. Some of it needs people to cross those boundaries entirely.

This is part of what we’re learning through CustomerFirst. Through our NewCo partnership with DVLA we started with a deliberately loose structure, organising workstreams around some of the biggest opportunities for transformation. But the muscle memory of professions has a funny way of kicking back in. Over time, we found those workstreams drifting towards technology, data, governance and organisation design silos. We have been trying to push back against that by reorganising around the future service processes and outcomes we are trying to create, using a shared view of that future service to hold the work together while different combinations of people form around it as the work changes. What we keep coming back to is that our aim shouldn’t be to preserve the org chart we happen to have today. It should be to keep asking what team we need for the work ahead.

That way of working hasn’t required anyone to abandon their craft. People step into unfamiliar areas when the work needs it, share knowledge and ownership across disciplines, and contribute beyond the usual edge of their role. The emerging learning is that the boundaries around that craft can be permeable enough for people to contribute past them, while still knowing where their expertise ends and when somebody with deeper knowledge needs to come in. We’re still figuring out where that works and where it doesn’t, and likely making plenty of mistakes along the way. Most of this movement has happened because individuals have made it happen, which is effort rather than architecture. We’re learning that if the same people keep bridging the gaps, those bridges have to become part of the model.

And that gets closer to the proposition underneath all of this. People need to keep their professional depth while the architecture around that depth becomes more fluid. The two depend on each other. You need enough mastery to understand what your field contributes, where its limits are, what you can safely carry across a boundary and whose knowledge needs to enter next. This is what Helga Nowotny argues in her 2003 paper Democratising expertise and socially robust knowledge; that expertise increasingly involves building links between what experts know and what other people need to know and do, and that feels remarkably close to the capability we are trying to create through CustomerFirst.

Which is why the professions still matter, as platforms rather than silos [^4]. Strong places to develop craft, methods, standards and professional community, but places people can launch from as the work demands different combinations of knowledge. Somewhere solid to stand, without deciding in advance where the work is allowed to go. The profession supplies the depth. The work determines how that depth combines with everything else around it.

The real danger is what government does next with all of this, which is try to institutionalise it. Making something legible at scale usually means making enough of it repeatable, and transformation, infamously, isn’t, despite what the McKinseys of the world tell you. Make fluid working easier to manage from a distance and the temptation will be to give it firmer edges, define the roles and write more of the instructions down. We have done versions of this before, creating structures intended to preserve a different way of working that eventually make that way of working harder [^5].

Transdisciplinarity is a practice, not an institution

The alternative is structures that enable the practice without fixing it in place. Which leaves a fairly practical question. If the answer is not another prescribed team shape, how would you know whether the conditions for this kind of working actually exist?

The characteristics of adaptive integration offer a useful lens. Not a framework to implement, or a checklist every multidisciplinary team should suddenly pass, but a way of looking at complex work when it starts to strain against the structure around it. When the problem keeps crossing boundaries, when the expertise you need changes as you learn, or when the team can see where the work needs to go but does not have a route to get there, these are useful questions to ask.

These are probably most useful as questions asked during the work and revisited as it changes, not as criteria settled before it starts. A team might look perfectly healthy against them at one point and find six weeks later that the problem has moved, a new form of expertise has become important or an organisational boundary has become the thing stopping progress. The point is to notice those changes and be able to respond to them.

They are deliberately questions rather than a maturity model. Turning adaptive working into five levels of transdisciplinary maturity would rather spectacularly miss the point.

Is this what we meant by empowered teams?

Adaptive integration points us somewhere government does not usually look. People inside complex work need enough awareness of the architecture forming around them to see where knowledge sits, where it is getting stuck, which relationships are emerging and where new ones are needed. Call it sideways legibility. Making the structure visible to the people trying to move through it, rather than only to the people trying to manage it from above.

And that gets us back to empowered. Enough freedom to follow the work when it moves, bring different expertise around it and change the shape of the team as understanding changes. This is why the future isn’t multidisciplinary, at least not in the way we’ve come to use the word. The same professions may still be there, and multidisciplinary working will still be useful. What cannot stay fixed is the architecture between them. When professional boundaries determine where the work can go, the structures we built to break the silos start behaving rather a lot like silos themselves.

I recognise there are plenty of questions left hanging. What this means for the politics of professionalisation, when digital specialists spent years fighting for the standing that professions gave them. And how people get good at moving horizontally across disciplines and vertically through institutions without sliding into generalism, which is close to what I was circling in my analysis of the Public Design Gap and my practice notes on testing and learning.

Those feel like places to keep exploring rather than answer now. So instead, here is a provocation. If the future team contains exactly the same professions as today’s, and what needs to change is the architecture between them, who in government is responsible for creating the conditions that let that architecture move?

Footnotes

1. Professions, capability frameworks and career paths. Patrick Dunleavy, Helen Margetts, Simon Bastow and Jane Tinkler, Digital Era Governance: IT Corporations, the State, and e-Government (Oxford, 2006): Their account of why technical specialists were kept in professional enclaves, away from the top of the civil service, is worth reading alongside anything about digital professions. UK IT occupations only secured recognition as a civil service occupational group on a par with others in 2004–5, seven years before GDS existed. The frameworks solved a real and old problem.

2. Specialists working alongside one another. Alice Russell, Fern Wickson and Anna Carew, “Transdisciplinarity: context, contradictions and capacity”: Their formulation is that multidisciplinarity is where disciplinary specialists work together maintaining their disciplinary approaches and perspectives, interdisciplinarity investigates the areas of overlap between disciplines, and transdisciplinarity transgresses and transcends disciplinary boundaries altogether.

3. The typology. Lois Wright Morton, Sanford Eigenbrode and Timothy Martin, “Architectures of adaptive integration in large collaborative projects”, Figure 1 is redrawn in their paper from Tress, Tress and Fry, “Clarifying integrative research concepts in landscape ecology”. Figure 2 shows three USDA-funded climate and agriculture programmes, proposed in 2011 by 115 principal investigators across 27 institutions, as they were structured at conception and as they had organised themselves by 2014. The structures were observed rather than proposed, tracked over four years through surveys, focus groups, interviews, archived meeting records, external evaluation, social network analysis and ethnography.

4. Platforms rather than silos. The phrase is Russell, Wickson and Carew's. They propose that strong foundations of intellectual capacity, including disciplinary ones, could act as platforms rather than silos: places to respond to problems flexibly and to launch from, without fixing the institution into particular strengths and trajectories.

5. Structures that destroy the flexibility. Russell, Wickson and Carew again. Writing about universities, they watched cross-disciplinary work get captured by the institutions that valued it, consolidated into designated research strengths and priority areas selected largely on commercial potential. Australia’s Cooperative Research Centres programme funded 16 centres in 2004, roughly half the 2002 number and selected down from over 50 initially established. What institutionalisation produces, they argue, is mega-silos: new structures built vertically as islands of strength, replacing disciplinary boundaries with topic and industry ones, with cross-disciplinary work starved of resources in the spaces between. Their conclusion is that structures built to institutionalise transdisciplinary activity are unlikely to succeed and are in fact an anathema to the flexibility and responsiveness required. What they argue for instead is enabling structures: opportunities for interaction, researcher-driven groupings, incentives and quality criteria.

I largely agree, because we’ve done this before. Energy got its own department in 2008, was folded into BEIS in 2016 and pulled back out in 2023. DSIT itself was carved out of DCMS in 2023, and digital is now heading back. The Institute for Government estimates direct costs of around £15m to stand up a new department, with substantial productivity losses on top. It has recommended for years that a cost-benefit case be assessed once a change has bedded in, and describes the current arrangement as one where experience resets to zero every time. I mean, they’ve only just painted the walls again in our office, which is a small and slightly ridiculous detail, but one that captures something about the nature of these changes. Reorganisations interrupt work already under way, loosen relationships that took years to build, and leave good people trying to work out who now owns the decisions they were making the week before.

The more important question, though, is what this tells us about the government’s promise to “rewire the state”. Because, beneath the technocratic language of “machinery of government” and a “stronger, more strategic centre” sits a familiar theory of change – one in which government becomes more effective when responsibilities are gathered beneath the right ministers, reporting lines become clearer, and an unwieldy institution is made legible enough for the centre to direct.

These structures matter, of course. They shape authority, budgets and political attention, and in some cases give reformers the power to make departments behave differently. Yet the container itself tells us very little about whether that authority is real, which levers sit behind it, whether the centre controls any of the infrastructure the rest of government depends on, or whether its capability can survive the minister who sponsored it.

My favourite illustration of this theory of change is in James C. Scott’s “Seeing Like a State”, where he describes the state’s drive for legibility, simplifying a complicated world into categories that can be counted, planned and controlled. His example of scientific forestry in Prussia captures both the attraction and the danger. Mixed woodlands were cleared and replanted in straight rows of a single fast-growing species, producing forests that were easier to measure and initially more productive. By the second and third generations, the yields collapsed. In making the forest easier to see from above, the drive for order had erased the interdependent ecology that sustained it in the first place.

Machinery of government changes offer the same seductive promise. They make a complex state easier to see from above, giving ministers clearer lines of control and the appearance of a system that can be directed with greater precision. They also pull reform towards the things the centre can readily count: headcount, budgets, duplicated functions, reporting lines and the number of departmental boundaries a decision has to cross. And, these are legitimate concerns. Waste is not neutral in government. Every unnecessary handoff, repeated form or badly designed process consumes money, time and attention that could have been used elsewhere, and the cost rarely remains inside the organisation. It appears as longer waits, avoidable contact and less space for people whose circumstances require judgement.

The harder work – and albeit much more interesting for me – is identifying the conditions under which good work can happen, then allowing evidence from the work to shape the structures around it. That means understanding where teams have enough authority to act, which relationships help them move across formal boundaries, and where leaders create room for professional judgement. These conditions are harder to display than an org chart, and they tell us far more about whether reform will survive contact with reality.

Navigating bureaucracy: red tape and blue tape

This is what we’re exploring at CustomerFirst. By starting with the lightest governance our work can safely sustain, we the let decisions, dependencies and risks that hit us as we learn show us where something more formal is needed. Accountability stays close to the work, and additional governance gets added where the evidence shows it has a useful role. What we learn then has to travel back towards the centre. A single team can expose where a shared rule, assurance process or missing piece of infrastructure repeatedly blocks the work, but only central institutions can change those conditions across government. Direction and public purpose travel down. Evidence about what helps or obstructs delivery travels back up.

Importantly, that evidence needs to be precise, because bureaucracy is rarely one thing. From above, every delay, approval or handoff can look like friction waiting to be removed. Closer to the work, those same processes may be carrying very different kinds of value, and the time they consume tells you little about which kind you’re looking at.

David Eaves’ distinction between red tape and blue tape has long offered me a useful way of seeing the difference.

Red tape describes the formal bureaucracy of law, democratic checks, audit, safeguarding and due process. These constraints can become excessive, poorly designed or disproportionate, but they carry an identifiable public purpose. They protect fairness, safety, accountability and the principle that public power should be exercised within rules. The task is to navigate them with integrity, understanding which rules matter and why rather than treating every constraint as an obstacle to delivery.

Blue tape grows differently. It accumulates through interpretation, caution, habit and organisational memory, until a preference, workaround or local convention starts behaving like a rule. It is the approval introduced after one difficult case, the meeting inherited from a project nobody remembers, or the phrase “we can’t do that” when the underlying meaning is closer to “we haven’t done that before”. This bureaucracy may once have responded to a real concern, but its relationship to that concern has often been lost.

The distinction sharpens the relationship between efficiency and effectiveness. Efficiency measures the burden: how long a process takes, how many people touch it, what it costs. Effectiveness asks what that burden protects, whose risk it carries, and whether the same public purpose could be secured another way. From the centre, both kinds of tape look like delay waiting to be cut. Inside the work, one may be holding public power to account while the other shields the organisation from the discomfort of exercising judgement.

Learning the difference means following a rule back to its source. Is it required by law, regulation or an agreed standard? Which risk was it created to hold? Has an interpretation become stricter through repetition? What happens when the process is narrowed, changed or removed?

Answering those questions needs legal, policy, operational and service knowledge in the same room, with the people who experience the consequences close enough to challenge the account being given. It’s slower and less theatrical than setting a target for how many approvals should disappear. It also gives the centre a much firmer basis for action. Government can protect genuine guardrails, give teams permission to question inherited interpretations, and change shared rules when the same obstruction keeps appearing. That’s how evidence from one service becomes institutional change rather than another local workaround.

Some thoughts…

Perhaps scrapping DSIT was the right decision. There is something to be said for a politician moving quickly within a system most people find impossible to hold, and a 3-year-old department is hardly a constitutional principle.

What I keep returning to, though, is whether anyone will find out. The informal networks across government will likely absorb this reorganisation as they have absorbed many others. Relationships will be rebuilt, operational memory will remain in many of the same heads and the work will continue under a new letterhead. Government’s resilience makes the disruption survivable, but it can also make the costs difficult to see and the value of the change easy to leave untested.

Whitehall may be able to reorganise itself in a day. Learning which structures help good work happen, which bureaucracy protects public value and which survives through habit takes considerably longer. I think that is the work sitting underneath all of this. Without it, action risks becoming action without judgement.

And, thankfully, the approach feels like it is experiencing a real sense of renewed energy in the UK right now. A spark, even beyond the usual bureaucracy wonks, around changing how government delivers change – testing assumptions earlier, learning through delivery rather than discovery alone, starting with real service conditions, and building capability through practice rather than waiting for the perfect framework.

Test and learn is not necessarily a new idea. But something does feel like it is gathering. Where The Radical How helped sharpen the case for changing how government changes, the Cabinet Office’s Test, Learn and Grow programme has given the practice a more visible centre of gravity. Perhaps most significantly from a policy perspective, the Magenta Book — the UK government’s evaluation bible (as Nick Kimber put it) now includes a Test and Learn annex. Even if only as a signal, this matters.

For me, this is currently exactly the work we are doing through CustomerFirst. A deliberately high-velocity slice of test and learn in a NewCo space, making small, real changes to a live service that can begin to gather momentum for wider transformation, while also learning what those changes reveal about the organisation around them.

So, in the midst of that momentum, it felt like a good moment to pause and reflect on what actually happens when an organisation tries to practise test and learn seriously. Not as a method to master, but as a way of thinking out loud. What it reveals, what it makes possible, what it unsettles, and what it asks of the people trying to change a live public service from within.

As ever, I feel very fortunate to have daily conversations with brilliant colleagues, and to have the space to use writing as a way of making sense of what we are learning together. I hope they see some of their own thinking reflected here too.

1. Start by testing the problem

One of the things test and learn has made us more attentive to is how quickly a service problem can smuggle in its own solution.

It may sound obvious, but in practice it is surprisingly easy to miss, something I’m all too guilty of. A problem can arrive already shaped by the language around it. Too many people are calling. Too many cases are waiting. Too many applications are delayed. Too many people are doing the wrong thing. The temptation is to accept that framing and move straight into fixing it. Reduce the calls. Clear the queue. Speed up the process. Change the behaviour.

Sometimes that is exactly what is needed. But often the first problem statement is only the visible edge of something else.

A service might appear to have a demand problem when it actually has a confidence problem. People are not calling because they want to call. They are calling because the service has not given them enough certainty to wait. A process might appear to have a compliance problem when it actually has a comprehension problem. People are not failing to follow the rules because they are careless. They may be trying to navigate a form, a letter, a threshold or a decision they simply do not understand.

That is why problem definition matters so much in test and learn. The work is not just to ask, “What can we test?” It is to ask, “What do we think is really happening here, and how would we know if we were wrong?”

This is where the craft sits. Not accepting the first causal story too quickly. Not defining the solution into the problem. Not treating the data point as the explanation. Ten per cent of people filing late, a rise in avoidable contact, a queue that keeps growing — these things tell you something is happening. They do not, on their own, tell you why.

In practice, a better problem definition makes the work more testable. It gives you something behavioural and concrete to learn about. What are people unsure about? Where do they lose confidence? Which decision points create avoidable work? What would need to change for someone to act differently?

This is also where test and learn starts to move beyond the pipe logic of public service transformation. If the problem is defined only as demand exceeding capacity, then the answer will usually be to widen the pipe, unblock the pipe or move people through it faster. But if the problem is partly about uncertainty, trust, comprehension, risk or decision-making, then the work changes. You are no longer only improving flow. You are testing whether the service is organised around the right understanding of the problem in the first place.

For me, that is where test and learn begins. Not with the experiment, but with the discipline of refusing to pretend the first problem statement is the whole truth.

2. Hold the tension without turning it into pressure

One of the biggest pulls I’ve felt through this work is between the urgency to fix things at scale and the discipline to start small enough to actually learn something.

The ambition is deliberately big. We want to improve outcomes for citizens, make the work genuinely better for staff, and fundamentally rethink how a service operates. But when that ambition meets the reality of a live service – existing backlogs, growing operational pressure, partial data, legacy systems, local knowledge, risk and accountability – it also meets a history of attempted transformations and political pressure to demonstrate progress.

I always think about this type of work as stretching an elastic band. On one side is the scale of change people need and expect. On the other is what we can responsibly claim to know and do today. Starting small does not remove that tension. It creates space inside it, allowing us to make something real enough to learn from without pretending we already have the answer.

That may sound intense, so it is important to say that holding this tension together as a team is not the same as creating pressure. If the band is pulled too tightly, people brace against it. The work becomes hurried, performative or overly certain. If it is simply let go, things snap back to the way they were before. The challenge is to make the discomfort survivable, and even useful, so that people can stay open to what the work is revealing.

We have tried to name this directly within the CustomerFirst team. There will be moments when it feels like we are pushing too hard or testing something before we know the right answer. That feeling is not necessarily a sign that something has gone wrong. Sometimes it means we are getting closer to shipping something real rather than something safe. Holding this tension is something we have to support one another to do – noticing the discomfort, talking about it openly, and judging together whether it is productive.

This is not a tension you resolve once, either. Holding the tension can become the work. Every new idea, test or way of framing the problem stretches the elastic again. You do not get to bank the judgement from last time. You have to keep deciding how far to pull, when to hold and when to ease off. Hopefully, with practice, the team becomes a little better at recognising the difference between the discomfort that comes with learning and the pressure that stops learning altogether.

Don’t get me wrong, the point is not to make discomfort a virtue. We are all human, and what we are really working towards is a psychologically safe space where people can do brilliant work, support one another and enjoy themselves while doing it.

3. Test the organisation, not just the service

At its simplest, the high-velocity version of test and learn we’re working with keeps returning to three core questions: what is the hypothesis, what can we test to find out, and – the one that matters most – what does that unlock for the wider transformation?

The first two questions are principally about the service. Did this message reduce uncertainty? Did this small change help colleagues make better decisions? Does it move anything at all? But the third question tests something quieter and often unappreciated in this type of work. Can we work this way? Can we act before everything is known and live with what that feels like? Can the organisation tolerate the uncertainty that learning requires?

What I think this is really building toward is something less visible than the test itself. The harder transformation decisions that come later – the ones involving operating models, risk tolerance, policy constraints and organisational boundaries – will not be made by people simply reading a report. They will be made by people who have, or have not, experienced what it feels like to test something, learn something and change something as a result.

That experience changes what people believe is possible. It is not a side effect of testing. It is part of what the test is for. The external learning tells you something about the service. The internal learning – the growing confidence that evidence-led change is possible at all – is what allows you to keep doing it.

4. Start small enough to snowball

This is why a small experiment with one cohort or one part of a journey can matter, even when it feels counterintuitive in a transformation programme and inadequate against the scale of what needs to change. Early tests are not always trying to prove a business case or demonstrate statistical significance. Sometimes they are trying to create a signal – enough to know what breaks, what moves, what surprises you and whether a bigger test is worth running.

One of our first tests involved sending a hundred people a text message telling them their documents had been received. That was never going to provide definitive statistical proof. But that was never really the point. It took an idea out of a slide deck and put it into a live service within a matter of weeks. It made us confront the practical questions, bring the right people together and surface the blockers that would matter if we tried to do it at greater scale.

Dan Hill, writing about mission-oriented design, describes a related approach as snowballing, beginning with small, located prototypes that gather evidence, relationships, credibility and momentum as they move, until the work has enough gravity to attract wider institutional and political support that is needed for transformation.

Starting small, then, is not a compromise on ambition. It is a decision about sequencing. One test creates the conditions for the next, and each turn of the snowball should leave the organisation with more than it had before. More evidence, more confidence and more people who have experienced a different way of working.

This is an organisation becoming more capable as it moves. That is a different kind of rigour, but it is still rigour because the evidence that eventually convinces institutions to commit, reshape governance or unlock investment is rarely produced in advance. It is produced through the accumulation of real-world interactions in meaningful contexts.

5. Notice what travels beyond the protected space

What happens, though, when the learning reaches the edges of the protected space where it began? I have written before about innovation labs and other protected spaces for change. At their best, they create the conditions for people to work differently, crossing boundaries and testing ideas that the ordinary organisation struggles to hold. But those conditions can also be fragile. The work can remain contained within the lab, disappear when attention moves elsewhere or be absorbed back into the organisation without changing very much at all.

And at the moment, that history is sitting with me in this work. Our NewCo deliberately creates different conditions around a live service, giving a multidisciplinary team the space and permission to move quickly. But the real test is not only whether those conditions help us make progress. It is whether anything travels beyond them.

Do these protected spaces eventually replace parts of the original organisation? Do they get absorbed back into it? Or do they diffuse into nothing once the attention moves elsewhere? We are too early to know which of these is happening to us.

It has been fascinating, and slightly disorienting, to watch the language and energy of this way of working begin to travel beyond the team that built it. Both in our team and within the DVLA, people are using the phrase “NewCo” – invoking it in exactly the kind of institutional argument it was designed to sidestep.

I’m unsure whether that is a good sign or a warning. It might mean the idea has enough weight that people reach for it instinctively, which feels closer to Andrew Greenway’s distinction between a method and a movement for test and learn – where a method can be handed over, but a movement has to travel through people, practice and repeated use. Or it might mean the protected space is already being folded into the politics it was meant to remain separate from, language and all.

I genuinely don’t know which way it is going yet. I can point to the experiments, but not yet to evidence of which direction the learning is travelling. That is an uncomfortable thing to admit in a piece that has otherwise been fairly confident about what test and learn is doing. But I think the discomfort is the honest answer, rather than a gap to paper over.

Some reflections going forward

Whatever happens next, the work has to stay rooted in the people who run the service. Many have spent years watching transformation programmes arrive and leave, consultants come in, strategies launch, governance get restructured, and the service stay much the same. They know the workarounds better than anyone because they are often the people who have had to build them.

What I keep relearning, every time I sit with the people who actually run the service, is that nothing replaces what they know. They remember the change that did not make sense at the time and still does not. They know which small thing would make the greatest difference – an extra field on a form, a sentence reordered, a button moved – because they are closest to the person on the other end of it. The biggest improvements are very rarely the ones that look biggest on a roadmap.

Test and learn cannot arrive as another solution looking for buy-in. It has to be an invitation into a process, and that invitation has to lead somewhere. In practice, this requires more than listening. It means listening and responding — taking what we have heard and pushing an idea through, rather than simply noting it down, while keeping an open door for teams to share what is getting in the way before we have decided what the answer is.

The risk, as test and learn gathers momentum, is that organisations scale the visible parts too quickly. The terminology, the templates, the governance decks, the pilot labels, the show-and-tells. Those things can help, but scaling the language does not necessarily scale the learning.

The question going forward is what each test leaves behind. Does it make the service better? Does it also make the next act of learning easier because the relationships are stronger, the evidence is clearer and more people have experienced change happening in practice?

Building that capacity to keep improving feels like the harder project. And probably the more important one.

Back in 2017 I worked at Bosch Power Tools in an internal UX and product innovation lab. Shortly after, I took part in IBM’s Extreme Blue programme – multidisciplinary teams in rapid delivery sprints with organisations like the NHS, designing and prototyping new products in weeks rather than months. My time at Wilson Fletcher sat somewhere adjacent: digital transformation and organisational change. Then HMRC Policy Lab, a public sector innovation lab explicitly designed to explore policy problems differently. And now CustomerFirst within Government Digital Services (GDS), which is attempting something else entirely – a NewCo-style model focused less on generating insight and more on building, testing and learning in live environments.

Looking back now, the pattern feels clearer than it did at the time.

The first wave genuinely worked

Over the last two decades, governments across the world have built hundreds of innovation labs, policy labs, behavioural insights teams and multidisciplinary transformation units. They emerged from a growing recognition that traditional institutional structures were struggling to respond to complexity, digital transformation and increasingly interconnected public problems. Labs offered a different organisational form: protected spaces within existing systems where smaller multidisciplinary teams could experiment outside normal bureaucratic logic.

And for a while, they worked remarkably well.

The early generation of public sector innovation labs genuinely changed government. Denmark’s MindLab helped bring ethnographic and participatory approaches into policymaking long before most governments were seriously talking about user-centred design. Helsinki Design Lab explored how strategic design could help states work across interconnected systems rather than departmental silos. Policy Lab UK experimented with multidisciplinary approaches inside Whitehall, while the early GDS movement in 2011 fundamentally reshaped expectations about what public digital services could be.

These were not vanity projects or post-it-note theatres. They produced rigorous work, brought new professions into government, changed the legitimacy of user-centred design inside the state and created new ways of understanding public problems. More importantly, they created conditions most institutions struggle to sustain – protected authority, proximity to decision-making, permission to experiment and the ability to generate evidence through building rather than reporting.

Then the conditions disappeared

But over time, many of these environments lost the conditions that had made them effective in the first place.

MindLab was eventually folded into a new government taskforce following shifts in political sponsorship and administrative priorities. Helsinki Design Lab closed in 2013 after SITRA concluded that strategic design had become more valuable embedded across the organisation than concentrated inside a standalone lab. Despite Policy Lab being more than 10 years old, there is very little evidence that design-led approaches have had a significant impact on high-level policy development. And even GDS, arguably the most influential public sector transformation movement of the last twenty years, gradually became more institutionalised, governance-heavy and operationally constrained as it matured, something Mike Bracken, its founder, later reflected on after leaving government.

The common explanation is political churn. Budgets disappear, ministers move on, leadership changes. This is all true. But the deeper issue is structural.

Schuurman and Tõnurist famously described innovation labs as “islands of experimentation”. That framing still feels right. But islands can become isolated. And isolation, over time, becomes impotence. What the last few decades have revealed is that the further labs moved from live delivery and meaningful authority, the more they drifted toward insight generation detached from operational consequence – describing the future rather than testing it.

The architecture never changed

At the same time as labs proliferated, governments more broadly began adopting professions and methods designed around iterative learning – service design, agile delivery, user research, multidisciplinary product development. These approaches changed how institutions thought about services, users and delivery.

But the deeper architectures of decision-making remained largely intact. Major commitments continued to be shaped through sequential processes where assumptions harden early and learning arrives late – often only after implementation has already begun failing at scale.

The result is a strange institutional contradiction where governments have spent the last two decades building highly capable professions for iterative learning, while continuing to position those professions too far downstream from the decisions where iterative learning matters most.

Why the early GDS story still matters

The early GDS story matters not because government should nostalgically recreate 2011, but because it demonstrated what becomes possible when authority, learning and delivery remain connected long enough for evidence to accumulate.

When Tom Loosemore, Mike Bracken and the early GDS team described digital transformation, they were not really describing a design function or an innovation programme. They were describing a protected delivery environment: politically sponsored, operationally empowered and capable of generating evidence through working products rather than strategy documents. GOV.UK did not become persuasive because GDS produced a compelling vision deck. It became persuasive because working software changed the institutional argument in real time.

The working thing was the evidence. Governance, funding, resources all followed the evidence. Not the other way round.

That is the part organisations have consistently struggled to reproduce – partly because it requires political capital that is genuinely difficult to sustain, and partly because the governance systems that eventually surrounded GDS were designed for predictability and assurance, not for the kind of learning-through-delivery that made GDS persuasive in the first place.

And yet, I do not think the lesson from the last twenty years of public sector innovation is that governments should stop experimenting. If anything, the opposite may be true.

Rebuilding the conditions

This is why models like CustomerFirst are worth taking seriously. Not because they represent “the answer”, but because they are revisiting some of these conditions again: smaller multidisciplinary teams, protected delivery environments, evidence generated through testing and building, and governance that follows learning rather than attempting to pre-approve every pathway in advance.

Yes, CustomerFirst is early and has not yet demonstrated sustained influence over major institutional decisions – the kind of high-stakes, high-commitment choices where the gap between assumption and evidence most damagingly widens. And yes, it is also operating within GDS, an organisation that has itself become significantly more institutionalised since 2011. But the more interesting question is whether a NewCo-style model can maintain protected authority as it scales, or whether it will gradually be absorbed into the same governance logic that constrained its predecessors.

Because the risk is not that CustomerFirst fails outright. The risk is that it succeeds at the level of products and services while the deeper architecture of decision-making – where assumptions harden, where commitments calcify, where learning consistently arrives too late – remains essentially unchanged. That is precisely what happened to the first generation of labs. They produced excellent work and lost the argument about how decisions are made.

The conditions are the point

After spending nearly a decade moving through different organisational experiments, the pattern underneath all of them feels surprisingly simple. The environments capable of creating meaningful change are almost always the ones that shorten the distance between assumption and evidence – ensuring that learning happens before institutions become too committed to decisions that are difficult to reverse.

This is the logic sitting underneath approaches like the early Government Digital Service or Public Digital’s “test and learn” approach, which feels like a contemporary evolution of that earlier ethos. The point is not to move recklessly fast or abandon rigour. It is to test earlier. Build sooner. Learn before institutions become too committed to decisions that are difficult to reverse.

It is not a rejection of research so much as a shift from generating insight about doing, to learning through doing. From discovery work that describes hypothetical futures, to prototyping that creates partial versions of those futures and tests them against reality before institutions fully commit to them. Essentially, the work becomes about speeding up the feedback loop around the things that matter most.

But a faster feedback loop only changes things if it is attached to something consequential. Which brings the argument back to conditions rather than form.

The honest lesson from twenty years of public sector innovation labs is not that the model was wrong. It is that the lab itself was never really the innovation. The innovation was the institutional conditions that made the first wave effective in the first place.

What the next generation of transformation environments needs to get right is not what they are called – lab, unit, NewCo, programme – but whether those conditions can be created, sustained and kept genuinely connected to the decisions that matter most. A protected experimentation environment without political literacy, legitimacy or stewardship of the wider system does not solve the problem. It risks being wrong faster.

The question has never really been whether governments need experimentation. It is whether they are willing to create and protect the conditions that allow experimentation to shape decisions before systems fully commit themselves to the wrong trajectory.

That was always the harder problem. It still is.

This piece draws on CustomerFirst, alongside earlier CIVICWORKS writing on institutional learning, decision-making under uncertainty and the architecture of public transformation — particularly Unfit for Uncertainty and Beyond the Waterfall State. It builds on research into public sector innovation labs, strategic design and adaptive governance, particularly the work of Andrew Greenway, Ben Terrett, Mike Bracken and Tom Loosemore, Schuurman & Tõnurist, Jenny Lewis, Anna Whicher, Francesco Leoni, Public Digital and others exploring how institutions learn under conditions of uncertainty.

There’s a question that doesn’t get asked enough in conversations about government reform: are the people we’re hiring actually set up to do the thing we say we want them to do? Not in terms of talent or commitment – the civil service has plenty of both. But in terms of how we’ve defined the work, what we’ve decided to call success, and whether the systems we’ve built to develop capability actually point people toward the problems that matter most.

In the words of Jen Pahlka, in order for government to get better at doing the things it says it does, we need the right people, working on the right things. And in the UK, I don’t think we have.

Service design as a public panacea

For over a decade, design has been championed as the discipline that would help government think differently — bringing user-centred practice, experimentation, and systems thinking to the knotty, contested work of policymaking in the face of the many wicked and interconnected challenges we’re facing. The UK’s Government Digital Service was built on this promise. Public sector innovation labs like Policy Lab UK sprung up on the same foundations. And yet, despite measurably improved digital services, better citizen journeys, genuine, documented savings, and a professional community that simply didn’t exist twenty years ago, there is very little evidence that design has meaningfully scratched the surface of high-level policy development. Why?

The problem isn’t that government lacks talented people, or even the right instincts. It’s that it has spent the last decade hiring, training, and organising itself around a skills framework focused on delivery – and then acting surprised when those same institutions struggle with transformation. Nowhere is this clearer than in the rise, and limitations, of service design in government.

As a profession, service design was adopted through the creation and professionalisation of the digital, data, and technology (DDaT) career paths in the civil service by GDS in 2011. A huge moment. One that was genuinely transformative for digital public services. But – and this is integral to the argument – it inherently (perhaps inadvertently) embedded a commercial logic into the DNA of the practice that has never fully been interrogated. GDS mixed ideas and ways of working from commercial tech and agile software development to reshape how government procured, developed, and delivered digital services. The key phrase there is commercial practices – because service design is, at its roots, a discipline based on the role of design in organising production and consumption in industry. That origin matters more than we’ve admitted.

Since then, service design has been used as a “panacea” within the public sector – for both improving public services and, more ambitiously, as an approach to policymaking itself. The latter of which I’ve been practicing for the past three years. There are now hundreds of service designers across UK central and local government. Which makes the gap between what the profession is thought to do and what it’s actually doing all the more important to examine.

The public design gap

Despite a relatively large degree of success in improving digital public services, there is very little evidence that service design has had a significant impact on high-level policy development. Instead, a public design gap has emerged – where service design is deployed downstream, focused on improving citizen experiences of services without necessarily involving citizens in the significant policy decisions that shape those services in the first place.

What this means in practice is that service design – a capability explicitly built to reduce risk through iterative learning – is not being used to test assumptions before decisions are made. It arrives after the policy is set, after the commitments are made, after the assumptions have hardened. As Public Digital document in The Radical How, waterfall-style programmes start with many risky assumptions. By the time design enters, the window for real influence has often already closed, sometimes at significant public cost.

This downstream focus reflects broader institutional constraints – ones I’ve explored in previous articles. But here I want to focus on something more specific: service design as a practice. Because I think the discipline itself, as it has been institutionalised in UK government, lacks the political literacy and systemic capabilities needed to work upstream – and that this isn’t an accident. It’s a consequence of how we have professionalised a discipline without interrogating it’s commercial bias.

The key difference between commercial and public design is often framed as a question of outcomes. Businesses direct design toward shareholder value, market share, brand. Public design, by contrast, is oriented toward the creation of public value – “as both an outcome and a process”. This involves designing in unavoidably political contexts, across systems, with citizens, within longer-term socio-technical goals that no single organisation controls. And, fundamentally, these different purposes mean the practices themselves – the inputs, outputs, skills, and mindsets required – are fundamentally different. Yet, the framework that governs how service designers in government are hired, trained, and evaluated doesn’t reflect that difference.

Wrong framework, wrong place

This commercial orientation is visible in black and white in the Government Digital and Data Profession (DDaT) Capability Framework, which defines the service design role as “a method of designing end-to-end journeys of a service to help users complete their goals and deliver policy intent.”

Deliver. Not shape. Not challenge. Not co-produce. Deliver.

To understand why this matters, it helps to set it against what the UK’s own public design review – a landmark piece of evidence synthesis published by the Cabinet Office last year – identifies as the actual practices of public design. The review defines seven core practices that characterise how design creates public value across the full policy cycle: from understanding people’s lived experiences and generating ideas, to visualising possibilities, synthesising perspectives, enabling citizen co-creation, facilitating cross-organisational collaboration, and practically experimenting with potential options. Taken together, these aren’t a list of tools – they are a vision of what design is in a public context: participatory, systemic, experimental, and oriented toward outcomes that no single team or institution can define alone.

When you map the DDaT’s service designer skills against these practices, the gap is stark. What follows is that analysis – and it makes for uncomfortable reading.

What this reveals is a lack of publicness in how service design has been institutionalised – and consequently how it is practiced, funded, and staffed. The framework aligns closely with specific phases of delivery rather than the full scope of practices needed to achieve genuine policy outcomes. In particular, three critical gaps emerge that the framework either under develops or ignores entirely.

1. Sustained citizen participation. The framework asks for user research to generate insights. It does not ask for structured, long-term approaches to genuine co-production – the kind that gives communities real power over decisions that shape their lives. The result is a reinforced top-down model of policymaking, with citizens consulted rather than involved.

2. Place-based and ethnographic research. The framework prioritises usability analytics and user insight. It doesn’t ask for the kind of deep contextual research that connects individual lived experience with the wider social, ecological, and systemic structures those lives are embedded in. Policy made without this is policy made partially blind.

3. Policy prototyping beyond the digital. The framework has a strong focus on digital iteration – prototyping products, adapting interfaces. It doesn’t ask for prototyping in physical spaces, policy sandboxes, or regulatory experiments that test assumptions before they calcify into legislation. This is precisely where design could reduce risk earliest in the policymaking cycle. It’s also precisely where the framework goes silent.

The consequences of these critical gaps (and this analysis in general) should be a much bigger conversation than they currently are – and I don’t think we’re having it honestly enough.

Because this isn’t just a skills gap. When you hire, train, and grade designers against the wrong framework, the effects compound in ways that are easy to miss but hard to reverse. The quality of the work narrows toward what the framework asks for. The perception of what design is – among policymakers, among senior leaders, across the civil service – gets shaped by what designers are seen to actually do. The culture around the profession contracts. Work that might have been ambitious becomes cautious, because the incentive structures reward delivery over inquiry. And perhaps most troubling of all, the people who are genuinely motivated by transformation, by the harder and messier work of shaping policy rather than implementing it, look at what the role actually asks of them and decide it isn’t for them.

We end up with service designers doing a great job – at the wrong jobs. This is the current reality in the civil service.

Towards public design

To close the public design gap, there is a clear need to distinguish public design from other approaches – moving beyond improving the delivery of public services to a full suite of practices that generate public value through desirable policy outcomes. After all, design is about making things better. Where its strength in opening up possibilities has typically been perceived as a weakness within the constraints of policymaking, realising that potential requires a different mindset: a belief that government should be humble about its ability to know how policy will work in practice when working in complex systems.

This starts with differentiating public design from “just good policymaking” – where design practices often appear in isolation, disconnected from each other and from the policy cycle – toward understanding how a full suite of public design practices can be applied end-to-end, driving equitable, participatory, and genuinely creative policymaking.

Some argue that service design itself is ready for this shift – that the profession is being primed to embrace the full ambition of public design, that its existing foundations are sufficient to evolve toward it. I’m not convinced. Not because the people aren’t capable, but because the structures that shape the profession – how it’s defined, graded, hired for, and resourced – are not. You can’t iterate your way to transformation using a framework optimised for delivery. And continuing to ask service designers to do the work of public design without giving them the scope, the mandate, or the capabilities to do it isn’t evolution. It’s wishful thinking.

The more honest starting point is that we likely need to expand the range of design professions in government altogether – and to do so in a way that is explicit about scale and purpose. This is where Dan Hill and Anja Melander’s conceptualisation of strategic design becomes genuinely useful – as a practice that takes the core principles of contemporary design (user research, ethnography, iterative prototyping, participation, co-design) and directs that toolkit toward ethical concerns, systemic change, and broader societal outcomes. It is, in their framing, concerned with “design doing” rather than design thinking: working across networks, scales, and timeframes to help redefine how problems are approached before solutions are reached.

What makes this compelling for government isn’t the label, it’s the logic. Strategic design, in this conceptualisation, operates at precisely the scales where the public design gap lives: across systems, upstream in the policy cycle, in the “dark matter” of policy, regulation, and organisational culture that shapes what services are even trying to deliver. It connects the practices the DDaT framework currently ignores – not as additions to a delivery role, but as the foundation of a distinct professional practice with a different remit entirely. It also helps you imagine what the other professions that work in between these gaps could be.

None of this is straightforward. Defining new design professions within government means changes to capability frameworks, grading structures, and how roles are scoped and resourced across departments. It requires political will to give designers a legitimate seat upstream, not just downstream. And it requires senior leaders willing to accept that design’s value in policymaking isn’t despite its openness to uncertainty, it’s because of it.

Fundamentally, the right people, working on the right things. That’s it. The question is whether we’re willing to be honest about the gap between that aspiration and the framework we’ve actually built to pursue it – and whether we have the appetite to do something about it.

CIVICWORKS explores the intersection of design, policy, and public practice. If this piece resonated, reach out! Orr share it with someone working at that intersection.

References and further reading…

Bason, C. (2017). Leading public design: How managers engage with design to transform public governance (PhD thesis). Copenhagen Business School.

Bason, C. (2018). Leading public sector innovation: Co-creating for a better society (2nd ed.). Policy Press.

Bason, C., & Schneider, A. (2014). Public design in global perspective: Empirical trends. In C. Bason (Ed.), Design for policy (pp. 23–40). Gower Publishing.

Bjögvinsson, E., Ehn, P., & Hillgren, P.-A. (2012). Design things and design thinking: Contemporary participatory design challenges. Design Issues, 28(3), 101–116.

Chen, D.-S., Lu-Lin, C., Hummels, C., & Koskinen, I. (2016). Social design: An introduction. International Journal of Design, 10(1), 1–5.

Clarke, A., & Craft, J. (2018). The twin faces of public sector design. Governance, 32(1), 5–21.

Conway, R., Masters, J., & Thorold, J. (2018, July 12). Conclusions—From design thinking to systems change: How to invest in innovation for social impact. Medium. https://medium.com/from-design-thinking-to-system-change/conclusions-from-design-thinking-to-systems-change-how-to-invest-in-innovation-for-social-impact-909fce4c4d7c

Conway, R., Thoung, C., Kattel, R., & Maldonado, M. (2024, March 20). The return on investment of public design: Working paper describing the modelling process and proposed design for calculating the ROI on public design. Transformation by Design Ltd. https://www.transformationbydesign.org.uk/wp-content/uploads/2024/03/ROI_Public_Design_Working_Paper_March_2024.pdf

Crilly, N. (2024). Design thinking and other approaches: How different disciplines see, think and act. Cambridge University Press.

Dorst, K. (2011). The core of ‘design thinking’ and its application. Design Studies, 32(6), 521–532. https://doi.org/10.1016/j.destud.2011.07.006

Design Council. (2015). The Design Economy 2015: The value of design to the UK. Design Council. https://www.designcouncil.org.uk/resources/report/design-economy-report

Design Council. (n.d.). The Double Diamond. Design Council. Retrieved April 7, 2025, from https://www.designcouncil.org.uk/our-resources/the-double-diamond/

Hill, D. (2012). Dark matter and trojan horses: A strategic design vocabulary. Strelka Press.

Hill, D. (2023). Designing missions: A playbook for getting it done. Vinnova.

https://www.vinnova.se/p/designing-missions-a-playbook-for-getting-it-done/

Hill, D. (2014, July 8). The city is my homescreen. Medium. https://medium.com/dark-matter-and-trojan-horses/the-city-is-my-homescreen-317673e0f57a

Kimbell, L. (2009). The turn to service design. In G. Julier & L. Moor (Eds.), Design and creativity: Policy, management and practice (pp. 157–173). Berg.

Kimbell, L., Durose, C., Kattel, R., & Richardson, L. (2024). Public design in the UK government: A review of the landscape and its future development. London: Fieldstudio Design Ltd.

Knight, A. (2023, September 22). Introducing public design. Public Policy Design. https://publicpolicydesign.blog.gov.uk/2023/09/22/introducing-public-design/

Lee, C., & Ma, L. (2020). The role of policy labs in policy experiment and knowledge transfer: A comparison across the UK, Denmark, and Singapore. Journal of Comparative Policy Analysis: Research and Practice, 22(4), 281–297. https://doi.org/10.1080/13876988.2019.1668657

Lewis, J. M., McGann, M., & Blomkamp, E. (2020). When design meets power: Design thinking, public sector innovation and the politics of policymaking. Policy & Politics, 48(1), 111–130. https://doi.org/10.1332/030557319X15579230420081

Lewis, J. M. (2025). Policy design labs and uncertainty: Can they innovate, and retain and circulate learning? Policy Design and Practice, 1–13. https://doi.org/10.1080/25741292.2025.2460821

Lorusso, S. (2024). What design can’t do: Essays on design and disillusion (Set Margins’ #26).

Manzini, E. (2015). Design, when everybody designs: An introduction to design for social innovation. MIT Press.

McKinsey & Company. (2018). The business value of design. McKinsey & Company.

Mazzucato, M., & Ryan-Collins, J. (2022). Putting value creation back into “public value”: From market-fixing to market-shaping. Journal of Economic Policy Reform, 25(4), 345–360. https://doi.org/10.1080/17487870.2022.2053537

Mazzucato, M., et al. (2024). Mission Critical: Statecraft for the 21st century (Policy Report 2024/04). UCL Institute for Innovation and Public Purpose. https://www.ucl.ac.uk/bartlett/public-purpose/publications/2024/may/mission-critical-statecraft-21st-century

Papanek, V. (1984). Design for the real world (3rd ed.). Thames & Hudson.

Potter, N. (1969). What is a designer: Things, places, messages. Hyphen Press.

Schuurman, D., & Tõnurist, P. (2017). Innovation in the public sector: Exploring the characteristics and potential of living labs and innovation labs. Technology Innovation Management Review, 7(1), 7–14. https://doi.org/10.22215/timreview/1045

Tõnurist, P., Kattel, R., & Lember, V. (2017). Innovation labs in the public sector: What they are and what they do? Public Management Review, 19(10), 1455–1479.

Whicher, A. (2021). Evolution of policy labs and use of design for policy in UK government. Policy Design and Practice, 4(2), 252–270. https://doi.org/10.1080/25741292.2021.1883834

Whicher, A., & Swiatek, P. (2022). Rise of the policy designer—lessons from the UK and Latvia. Policy Design and Practice, 5(4), 466–482.https://doi.org/10.1080/25741292.2022.2141488

Over the last few years, I’ve been exploring the shape of mission-driven governments and the need for a fundamentally different approach to decision-making—one capable of navigating uncertainty, not simply delivering outputs.

In Unfit for Uncertainty, I argued that the UK state is structurally ill-equipped to operate in conditions of complexity, relying on linear models of analysis, static evidence bases, and decision processes designed for stability rather than change. I pushed this further in Beyond the Waterfall State, exploring how these dynamics play out in practice, and what a possible alternative decision-making architecture might look like – and what it would take to deliver it.

Both pieces converge on the same deeper question: what does a “healthy” government system actually look like? Not in abstract terms, but in terms of capability. Such a system is not defined by any single method or reform. It is defined by its ability to generate, absorb, and act on different types of knowledge – to hold multiple perspectives at once, to interrogate its own assumptions, and to learn in motion. But if this is what capability looks like, where does it actually sit within the state? The institutional locations of these capabilities remain unresolved.

One route into this question comes through my experience working within HMRC’s Policy Lab, alongside futures and foresight practice more broadly, and what Jon Kolko describes as “studio culture.” I’ve long admired Lucy Kimbell’s 2019 paper “What if there were more Policy Futures Studios?”, where she proposes a model for embedding futures thinking more directly within policymaking – moving beyond episodic foresight exercises toward more situated, practice-based forms of inquiry.

Picking up this thread, this article looks to (re)imagine the “lab model” as Policy Futures Studios – not as niche functions or experimental units within Government, but as a missing piece of institutional architecture capable of supporting more anticipatory, adaptive forms of decision-making within the state.

Futures as insight not infrastructure

Futures practices matter – particularly if we take seriously the idea that governments must operate not only in response to present conditions, but in relation to multiple possible futures. This requires a capacity to both keep the lights on and shape the future – something that cannot be achieved through delivery-focused models of policymaking alone.

And to some extent, these capabilities already exist. Across government, futures and foresight functions are distributed through horizon scanning teams, strategy units, and innovation labs. The Government Office for Science has codified approaches through horizon scanning, scenario planning, and trend analysis. Policy Lab has developed experimental futures methods to support policy development. Even the Home Office, as part of their Research, Development and Innovation strategy in 2025, identified futures as one of 14 internal RDI capabilities.

Futures, in other words, is not absent from government. It is just not where decisions are made.

And yet, despite all this activity, most policy decisions continue to be made as if the future were broadly knowable, stable, and reducible to a set of assumptions.

The issue is not the limits of futures methodologies, nor the complexity of the systems they seek to shape. It is the way these practices are positioned within the state. Futures work remains structurally disconnected from decision-making. Insight is generated, reports are written, signals are identified – but when it comes to allocating resources, defining policy direction, and making trade-offs under uncertainty, these capabilities sit at the margins.

They are, more often than not:

• episodic – conducted as projects, rather than sustained as capabilities

• adjacent – informing decisions from the outside, rather than shaping them from within

• non-binding – generating insight without consequence

This helps explain why futures work so often struggles to demonstrate impact. Even where capability exists, it often arrives too late – or remains too detached – to influence the decisions that matter most.

We are attempting to build a mission-driven government without an anticipatory core.

From foresight to statecraft

If futures practice is to play a meaningful role in mission-driven government, the question is not simply how to improve its methods. It is how to reposition it within the architecture of the state itself. This is where the distinction between foresight and statecraft becomes important – and where I want to spend a moment, because I think it’s the crux of what this piece is really about.

Foresight, as currently practised, is largely concerned with producing insight about the future. Statecraft is concerned with how decisions are made under conditions of uncertainty – how institutions interpret signals, structure choices, and act in the present in ways that shape long-term outcomes. Moving from one to the other is not a methodological shift. It is an institutional one (Read more on Mission Critical Statecraft).

Foresight asks: what might happen? Statecraft asks: how do we decide, given that we cannot know?

Government has invested heavily in the first question. It has done far less to build the institutional conditions for the second. The result is what Jen Pahlka calls “policy vomit“ – where the words of policy are pushed, essentially undigested, into the administration of a policy, without the benefit of thinking, learning, or iteration. Futures insight suffers the same fate: generated upstream, absorbed downstream, and rarely allowed to alter direction in between. What accumulates instead is the institutional spaghetti of policy, delivery, and technology that so many of us recognise.

This is the gap that Policy Futures Studios are designed to occupy – not as a rebranding of existing labs or foresight units, but as a form of organisational infrastructure capable of operating across, and at times cutting through, the accumulated weight of legacy processes and fragmented mandates. Not another layer of activity. A different way of organising how policy problems are engaged with.

What a Policy Futures Studio actually is

At their core, Policy Futures Studios would operate as sites of structured inquiry within the state, bringing together futures practice, policy design, and systems thinking to support decision-making and delivery under conditions of uncertainty.

Crucially, their role is not to produce standalone artefacts – reports, scenarios, or speculative outputs. It is to work directly with policy and strategy teams to reconfigure live problems and deliver credible policy and service outcomes. This means temporarily stepping outside the constraints of existing policy and delivery environments in order to reimagine the intent, experience, and operational logic of a given problem – before translating that back into the organisation in forms that can be tested, implemented, and scaled.

In this sense, they are less concerned with prediction, and more concerned with making uncertainty actionable. These interventions might take the form of new policy propositions, redesigned services, or alternative organisational arrangements – developed as prototypes or pilots that allow ideas to be tested in practice, rather than remaining at the level of abstraction.

This is what begins to move futures work out of the realm of insight generation and into the realm of applied institutional change.

The language of the “studio” is not incidental, either. As Jon Kolko describes, studios are not simply places where ideas are generated, but environments in which problems are actively worked through – iteratively, collaboratively, with a close coupling between thinking and making. When applied to government, this suggests a shift from abstract strategy toward situated experimentation – where policy is developed through cycles of exploration, prototyping, and testing, rather than defined upfront and delivered downstream.

To make this institutional form more tangible, I find it useful to think through five interrelated design components – drawing on existing work on lab frameworks and relational infrastructure (See Anna Whicher & Sam Rye), but extending into a more explicitly decision-oriented context. A Policy Futures Studio can be understood not simply as a team, but as a structured environment for inquiry, design, and implementation, organised through: (1) purpose and proposition; (2) product and prototypes; (3) people and capacity; (4) process; and (5) place.

(1) Purpose and proposition — intervening in how decisions are made

The purpose of a Policy Futures Studio is not to produce insight about the future, but to intervene in how decisions are made in the present. Most futures work assumes that better knowledge leads to better decisions. In practice, insight accumulates, but decisions remain shaped by fixed problem definitions, institutional constraints, and delivery pressures.

Policy Futures Studios operate within this gap — reconfiguring how policy problems are understood, structured, and acted upon under conditions of uncertainty. This means working both at the point where problems, options, and trade-offs are still fluid, and where existing policies and services have become constrained by legacy assumptions and accumulated complexity.

What makes this different from conventional policy processes is not the tools, but the mode of engagement. Studios are designed to hold problems open for longer, working across multiple framings and deliberately avoiding premature convergence. This is not about slowing decision-making, but about changing when and how decisions are stabilised — so that choices are made with a fuller understanding of the systems they are reshaping.

This work is done with policy and delivery teams, not for them. The aim is not to reduce uncertainty, but to increase the state’s capacity to navigate and act within it — building shared understanding, exposing trade-offs, and enabling decisions that are better grounded in the realities of the system.

(2) Product and prototypes – making decisions tangible

The “product” of a Policy Futures Studio is not insight, but interventions that make policy decisions tangible, testable, and open to challenge. Rather than producing reports or recommendations, studios work through a set of outputs that sit closer to action: propositions, prototypes, and live interventions that allow ideas to be explored in practice – through delivery – not just described in theory. This is essential to bridging the gap between policy and delivery.

Propositions articulate alternative ways of framing a problem or organising a response–making explicit the assumptions, trade-offs, and implications that are often left implicit in policy development. They are not final answers, but structured positions that can be interrogated, compared, and evolved.

Prototypes take this further. Drawing on Sam Rye’s thinking on relational infrastructure, prototypes are not simply early versions of solutions, but devices for reorganising relationships within a system. They make abstract ideas concrete, allowing different actors to engage with them, contest them, and understand their consequences. In doing so, they shift conversations from speculation to experience – changing how decisions are made by grounding them in something that can be seen, tested, and responded to. This might take the form of a reconfigured service journey that exposes operational constraints, a policy mechanism modelled through a pilot, or alternative delivery arrangements enacted at small scale.

Live interventions extend this logic into the system itself – testing propositions and prototypes in context, generating feedback, revealing unintended consequences, and creating the conditions for iteration before decisions are locked in.

Taken together, these outputs create a different relationship between thinking and doing. They enable policy to be developed through cycles of exploration, testing, and refinement, rather than through linear stages of analysis followed by delivery. The result is not just better evidence, but evidence that is produced through action – grounded in the realities of implementation, and capable of shaping decisions as they unfold.

(3) People and capacity – unlocking capability from within

The effectiveness of a Policy Futures Studio does not depend on assembling a new team of specialists. It depends on identifying and working with the right people already inside the system.

Across government, there are genuinely talented individuals with deep domain expertise – policy professionals, analysts, service designers, operational leaders – who understand where the system is not working, where assumptions no longer hold, and what would need to change. They are often closest to the problem, and yet furthest from the conditions required to reshape it.

Policy Futures Studios are designed to work with and elevate this existing capability, not replace it. Rather than acting as an external intervention, the studio creates a space where these individuals can operate differently: with greater agency, access to different tools and perspectives, and the permission to question, reframe, and test ideas in ways that are often difficult within standard policy environments.

Capacity, in this sense, is not a question of headcount. It is about how capability is activated, connected, and sustained over time.

(4) Process – working above and across the system

The process of a Policy Futures Studio cannot simply replicate existing policy processes. If it did, it would reproduce the same constraints it is intended to challenge.

Studios operate alongside, across, and at times above bureaucratic processes – creating conditions to reframe problems, test ideas, and explore alternatives without being immediately constrained by existing procedural requirements. This does not mean bypassing governance or avoiding accountability. It means sequencing engagement differently – intervening earlier, before decisions are fixed, and creating space for exploration before formal processes begin to close options down.

Studios focus on problems where uncertainty is high, the stakes of getting it wrong are significant, and existing approaches are struggling to make progress. The process is not linear. It moves iteratively between exploration and intervention – working through propositions and prototypes, testing them in context, and using what is learned to reshape direction.

Crucially, this also means reimagining the processes themselves. If policies and services are being rethought, then the bureaucratic routines that produce them – approval pathways, evidence standards, sequencing of decisions – must also be open to redesign.

Studios therefore operate not only on policy problems, but on the processes through which those problems are addressed. Over time, this creates a dual effect: decisions improve, and so do the conditions under which future decisions are made.

(5) Place – from space to sociomaterial infrastructure

The idea of the “studio” is not incidental. It speaks to a different kind of place within the state.

As Jon Kolko suggests, place is not simply space. Space is possibility; place is what emerges when that space is shaped by shared practices, norms, and ways of working. Policy Futures Studios are designed as this kind of place – not just a physical environment, but a sociomaterial infrastructure in which new ways of working can be enacted.

Within a Policy Futures Studio, place becomes a medium for working through problems. Prototypes, conversations, visualisations, and artefacts accumulate over time, creating a shared reference point that allows ideas to be interrogated, contested, and evolved. Through repeated participation, people don’t just contribute to the work – they begin to internalise a different way of working. Drawing on Sam Rye, this can be understood as the creation of relational infrastructure: the environments, artefacts, and interactions that shape how people come together to act. In this sense, place is not background. It is an active component of how capability is formed.

This is what allows the studio to function as more than a project environment. It becomes a site of practice, where new forms of decision-making are not only explored, but rehearsed and embedded. As people move back into their respective domains, they carry these practices with them – shifting how work is done elsewhere in the system.

Place is not just where the work happens. It is how the work becomes durable.

Bringing it together – and being honest about what I don’t know

Taken together, this positions Policy Futures Studios not as discrete teams, but as part of a broader capability infrastructure within the state – one that supports how knowledge is generated, interpreted, and acted upon across domains.

But I want to be honest about the tension this piece hasn’t fully resolved.

I’ve argued that studios must work within decision-making rather than alongside it. And yet, the five components I’ve described still risk looking like something adjacent – a temporary architecture, a space to step into. The harder institutional question – how a studio gains the authority to actually reshape live decisions, rather than inform them from a comfortable distance – is one I’m still working through.

What I’m increasingly convinced of is this: legitimacy in this space is built through action, not insight. Studios that start by producing reports will be treated as report-producers. Studios that start by taking on live problems – messy, high-stakes, politically loaded ones – and demonstrably improving them, will be treated differently. That’s not a design principle. It’s a bet on how institutional trust actually accumulates.

In practice, this begins small. Not with a fully-formed team, but with a small number of live problems where uncertainty is high and existing approaches are struggling. The work starts by assembling the right people around these problems and working with them to reframe, prototype, and intervene in ways that shape decisions in real time.

From there, the work expands not through scale, but through demand – creating pull within the system to embed this way of working more deeply over time.

That feels like the right starting point. Whether it’s sufficient is, honestly, still an open question.

Despite the prevailing assumption that economic policy exists to fix markets, correct failures, patch inefficiencies, and generally take a step back, every so often a reminder appears that market-shaping is not only possible, but already underway in different parts of the world.

Last night I was invited to attend a FinTech UK & Korea event hosted by the British Embassy in Seoul and Co-Labs Global. South Korea – one of the world’s fastest-growing fintech markets with 39 million open-banking users, a rapidly maturing digital infrastructure in the form of MyData, and a clear ambition to become a world-leading fintech economy – offers a compelling example of institutional intent in practice: setting direction, building shared infrastructure, and structuring the terms on which private actors participate.

For South Korea, their fintech story is framed less as one of sectoral encouragement and more as the intentional shaping of conditions for interoperability, trust, and experimentation. Security expectations are being treated as a foundation of legitimacy rather than a compliance afterthought. Shared rails are then standardised, so innovation builds on common infrastructure rather than parallel systems. This, in turn, enables regulatory sandboxes to institutionalise experimentation within the perimeter of the state itself. In combination, these moves signal that growth is not being left to chance; it is being actively shaped and played with.

This closely resembles what development institutions increasingly describe as digital public infrastructure – shared, interoperable rails upon which private innovation can build which interestingly, as David Eaves argues, raises all sorts of real thick strategic questions of not simply whether a country adopts digital public infrastructure, but what kind of tech stack it chooses to commodify and govern – and what that implies for sovereignty, competition, and long-term value capture.

This stands in contrast to the UK’s more familiar posture, where technology policy often defaults to “push incentives”: funding early innovation, lowering costs, and waiting for markets to consolidate around the value created. The recurring outcome is that public infrastructure absorbs systemic risk while private actors capture the majority of scale-driven returns. The issue is not that these investments are misguided or wasted., it is that the terms of partnership are rarely designed with sufficient intentionality or conditionality to ensure that public risk is matched by public reward. And, over time, this produces a pattern in which the state carries early uncertainty while long-term value accrues elsewhere. The architecture of partnership, rather than the volume of investment, becomes the central design flaw.

I think these comparisons are meaningful for many reasons. Market failure theory – the intellectual foundation of most Western industrial policy – asks governments to intervene only where markets fall short, then return to the sidelines. It is a static technocratic framework, oriented toward optimising what already exists. Market-shaping, by contrast, begins with a different question: what kind of market do we want to build, who are the willing actors, and what institutional capacities are required to steer it there?

These questions imply that public support need not be blind to outcomes. Conditional mechanisms – whether linked to domestic investment, knowledge sharing or interoperability standards – can ensure that public risk is matched by public return. Without such design, partnership defaults to asymmetry (Read more from Marianna Mazucato and Dani Rodik here)

Which is why the more interesting question, watching Korea’s pitch, wasn’t how many British firms might enter the market. It was whether the partnership terms being offered would actually align private activity with the infrastructural outcomes Korea is trying to build. Market access and market co-creation are not the same thing. Who carries the cost of building the field? Who benefits when it scales? These questions don’t answer themselves – their solutions (the conditionality of these relationships) have to be designed.

Interestingly enough, the UK has shown it can do this quite well. The antimicrobial subscription model introduced by NICE and NHS England is a quiet example of genuine policy creativity. Rather than paying pharmaceutical companies per pill sold – which would have created the wrong incentives entirely – the NHS pays a fixed annual fee based on a drug’s assessed value to the health system. Revenue is delinked from volume, and private incentives are realigned with long-term public benefit. The whole mechanism was redesigned, not just the funding level.

While Fintech isn’t antibiotics, the underlying question is the same: can you design the terms of partnership so that private capability serves public infrastructure outcomes, rather than just extracting value from them? Could partnership models reward improvements in interoperability or system security, not just transaction volume? Could data-sharing frameworks be built to reduce lock-in rather than entrench it?

These are the design questions that separate a high-growth sector from something more durable – infrastructure whose governance shapes the economy over the long term.

Korea is making a credible case that it knows what kind of fintech market it wants to build. The invitation to British firms is real. But the quieter provocation is for the UK to ask the same question of itself – not just how to grow the sector, but how to govern it in a way that actually captures public value from the infrastructure we’re already building.

More broadly, the strategic question for governments is not whether markets will form, but which layers of the stack they choose to shape, commodify, and govern — and fintech is not exempt from that logic. The capacity exists. The question is whether it is exercised with the same degree of intentionality in the digital financial infrastructure now emerging.

Involving citizens in public design is much more than running an “open” process. It is about redistributing agency – enabling people to shape decisions, define problems, and participate in making change happen. It is as much about co-authoring a story of change as it is about delivering outcomes.

Over the past eight years, I’ve been fortunate to work with a broad range of people — from children on dialysis wards or someone’s friend’s uncle in Germany looking to rent power tools to everyday citizens wanting to inquire about their tax bill and the people I work with. Across all of these experiences, I’ve come to trust something quite simple. That work gets better when people (or citizens) are involved in defining needs, shaping responses and proposing designs – not just talking about change, but helping to make it happen, together.

Given that I’m currently working with a brilliant teaching team on the MPA in Innovation, Public Policy and Public Value at UCL’s Institute for Innovation and Public Purpose – with Rowan Conway, Gabriella Gomez-Mont, Agostina Argiro Fuertes, Cecilia Santucho and Nickolas Laport Aldunate – supporting students to start tuning into how we better engage, enable and empower citizens through participatory scaffolding, it felt timely to pause and reflect on what it really means to include citizens in the design process.

Drawn from fieldwork, facilitation, practice, research, and a somewhat substantial volume of reading across public innovation and social infrastructure in recent years, here are some notes from the field that I’ve been marinating on.

1. Spend more time outside than in the office.

Being and doing things together on site is where trust and respect are built across differences. Thinking doesn’t happen in spreadsheets alone; it happens in kitchens, on pavements, in school corridors, in workshops and vans. We can’t expect colleagues — or citizens — to always come to us, or to log into a Team’s call, when they are balancing lives of their own.

Instead, jump in a tradesman’s van for a ride-along, chat to your taxi driver, go on a dog walk, have spontaneous conversations where people naturally gather: the barber shop, the football pitch, the school gate. These are all so much more than romantic gestures, they are cultural and essential ethnography into how people actually live their lives.

This echoes much of Hilary Cottam’s work on relational welfare – the idea that change does not emerge from service transactions but from relationships built in the everyday spaces of life. In Revolution 5.0 A Social Manifesto (which inspired this post) Hilary argues that our public systems are over-optimised for institutional efficiency and under-designed for human connection. Spending time outside the office is not a gimmick; it is a much needed human correction.

A varied working process also opens more ways of contributing. Different spaces unlock different competencies. Some people speak best in formal rooms; others in motion, or over tea. Inclusion is not only about who is invited, but about the environments we design for participation. Basically, complex systems can’t be redesigned from conference rooms.

2. Hold back and wait.

Spend time. Percolate. Let plans evolve and be open to change. So much pressure comes from what we assume are procedural rules of the system — deadlines, reporting cycles, funding windows. But what happens when we loosen our grip? What could happen if communities genuinely shaped the direction, rather than reacting to pre-set options?

This isn’t easy. It often means moving at the “pace of trust” – and trust, that pesky, vague word, takes years to build and only moments to lose. It really means that this work cannot be rushed to meet artificial six-month timelines or funding deadlines. This theme has surfaced repeatedly in my conversations with practitioners and articulated much more fluently than I could dream of, particularly those working in place-based change like John Hitchin or Jo Blundell who have that special skill at crystallising thoughts. The Place & Evidence substack is a shining example of this.

In many of my own experiences, breakthroughs came only after long stretches of what looked, from the outside, like slow work: listening, showing up, returning again and again. Staying in the deep end long enough for patterns and signals to manifest. People (and by people I mean anyone from the policy teams I work with to the students I support) often ask, “How much research do we need?” But that’s a category error. You don’t know until you start learning – and learning takes time.

In complex systems, premature solutions often risk strengthening the status quo because they fail to shift underlying conditions. In this context, holding back does not mean being passive. It means resisting premature convergence of ideas. It means allowing relationships to form before solutions harden. It means accepting that the “messy middle” is not a phase to rush through, but that relationships are the core of transformation. Our role as practitioners is to continually pry open the space between problems and solutions, creating room for deeper exploration before closure.

3. Lead with deep generosity and an abundance mindset.

Reject the conventional economic logic of the “bare minimum” and instead bring intentional generosity and care into every interaction — throughout life and your work.

This can mean many things and, in public systems, it’s not easy. Generosity can feel naïve. We are trained to optimise, to reduce, to streamline. Time is scarce. Budgets are tight. Engagement becomes extractive almost by default: What do we need from people? How quickly can we get it? How cheaply can we deliver it?

Leading with deep generosity flips that frame.

It can mean taking the time to listen deeply to citizens to understand what they value and what makes a place work, rather than leading with data-heavy, deficit-focused profiles. It can mean hosting rather than merely facilitating — setting the boundaries of a space physically, socially and conceptually. Considering how much natural light is in the room. Whether the snacks are nourishing. Whether the space smells right. Whether the premise of gathering feels intentional and human. Even ensuring the font sizes in your document are consistent — because coherence communicates care.

I think that Dark Matter Labs are fantastic at illustrating this principle by arguing that transformation requires shifting underlying logics – not just introducing new tools (See: Dm Threads). In this framing, generosity is not softness; it is a structural move. It changes incentives. It signals safety. It invites reciprocity rather than compliance. And when people feel respected and safe, experimentation becomes possible. Risk becomes shared rather than imposed. Generosity creates the conditions for long-term relationships (in perpetuity) rather than one-off transactions. It moves us from episodic engagement to ongoing stewardship.

Fundamentally, working from abundance recognises the immense skills, networks and energy already present in communities or teams, rather than viewing them through a lens of scarcity, conflict or decline. In complex work, what you choose to amplify matters. An abundance mindset amplifies capability – and capability compounds over time.

4. Provide the “milk carton” not the “flavour.”

We have the amazing Lauren Dark to thank for this framing. As practitioners, our role is to provide the “scaffolding” and guardrails for participation (the “milk carton”) – such as safeguarding, policies, and frameworks – while allowing the local community to provide their specific identities, responses and lived experiences (the “flavour”).

This means designing participatory scaffolding around psychological safety and sensory comfort. As with my note on leading with deep generosity, the physical environment is a super easy way to think about shaping how people participate. A bland community hall with poor coffee can put people on edge. Designing lounge-room-like environments – creating front rooms rather than back offices – is part of this shift. Physical space communicates whether power is shared or retained. The container should feel safe enough for people to experiment, disagree and create.

But the “milk carton” is not only physical. It extends to the social and governance scaffolding we build around engagement. Frameworks such as Free, Prior and Informed Consent (FPIC) ensure that decisions are made freely, in advance and with full information – acting as safeguards for self-determination and sustainable development. This is not bureaucratic box-ticking; it is participatory infrastructure. Practitioners such as Julian Thompson (through Rooted) are doing super important work in this space, grounding engagement in Design Justice principles. Their work reminds us that inclusion is not neutral. It requires increasing our own capacity to build empathy, recognise positionality and actively avoid harm.

My own mental model is to remember to avoid the trap of being the “hero” designer and being eternally optimistic about solving everything and helping everyone; instead, we should be striving to act as a translator or craftsperson who helps amplify the brilliant things the community is already doing.

5. Building “breadcrumbs” for policy change

Relational work is powerful – but fragile. Without translation, it evaporates at the policy interface. To ensure local insights aren’t lost, codify community conversations into thematic “breadcrumbs” that future policy decisions can remain accountable to. Make things tangible.

Dan Hill (in his book: Dark Matter and Trojan Horses) refers to these, from a strategic design perspective, as “MacGuffins” – artefacts or interventions that carry narrative weight and create movement within complex systems. Extended through ideas like Trojan Horses, they become strategic devices: visible, workable objects or experiments that bridge the gap between radical local practice and the institutional mindsets of national policymakers. They make change experiential rather than rhetorical – ensuring that inclusive growth is rooted in lived, everyday experience rather than abstract ambition.

In practice, this means getting your hands dirty. Improvising. Being creative. Perhaps it’s my background in industrial design, but I believe in working through prototypes and experiments that allow participation to become a collective act of change. It means using probes, pilots and live tests to understand how ideas behave in real space – not just on paper. When something is built, placed, trialled or enacted physically, it shifts the narrative. It becomes evidence that listening has occurred in a way people can experience.

There is something powerful about interventions you can see, touch and use. Think about Santander Cycles — or “Boris Bikes.” Whatever your view on their politics (of both the service and the bureaucrat), they were tangible. Rows of red bicycles appearing across London were visible proof that a conversation about congestion, health and mobility had translated into action. You didn’t need to read a strategy document or look through a spreadsheet to understand that something had changed. The city felt different.

A shared reference point was created – something that allowed disagreement to be productive rather than abstract. Something for communities to respond to, debate, adapt and reinterpret – not just hypotheticals, but lived experience.

Without these breadcrumbs, participation risks becoming episodic — a series of conversations that dissolve once the workshop ends. With them, it becomes cumulative. It becomes part of how a place understands itself, and how institutions learn to act differently over time.

To close…

None of these thoughts are particularly new. Most of it is slow. All of it is relational. Fundamentally, this note is acting as a reminder. Including citizens in design is not a technique to master, but a way of working to practise – one that redistributes agency not just in theory, but in the everyday spaces where change actually happens.

This piece starts where that argument ends, asking what a different architecture for decision-making might look like when uncertainty is high – and stability still matters. To answer it, this article explores why existing innovation efforts repeatedly collapse back into business-as-usual routines, before setting out an alternative, overlaying architecture for decision-making under uncertainty: one built on the premise of strength without weight and dynamic learning at its core.

Beyond the waterfall state

To be clear, the answer is not to burn everything down and start again. It points in a very different direction from the chainsaw, scorched-earth approach of the Elon Musks of the world. As Andy Knight reminded me, most of what government does is stewardship – and that function is indispensable. It depends on stable, convergent, waterfall prediction policymaking. It is what keeps the lights on. Literally.

But the architecture that underpins stability is not the architecture that enables transformative change. Keeping the lights on today matters – but so does ensuring they stay on in the future. Missions, systems change, and early-stage sense-making require something different: space for divergent creativity, structured exploration, and collective judgement. This is the foundation for a state that is able to work with uncertainty rather than suppress it – what Kattel, Drechsler, and Karo describe as agile stability: the capacity to generate new directions while maintaining long-term institutional coherence.

The problem is that we continue to run all forms of decision-making – certain and uncertain – through the same architecture inherited from a New Public Management era, an architecture built on the assumption that if you press a button, you know something will happen – as if governing were a computational task with bounded rationality rather than a socio-technical one. While this logic may work tolerably well for stewarding stable systems, organisations – and the world around them – are not computers. When uncertainty is high and judgement, learning, and creativity matter most, this type of computational “bounded rationality” repeatedly fails (See: Statecraft for the 21st Century) leaving us with a hollowed-out version of what Charles Lindblom once described as “muddling through”: not pragmatic adaptation, but institutional drift in the face of uncertainty.

James Plunkett has suggested that this traps government in a kind of “local maximum” able to improve what it can already see, but unable to perceive or reach a wider landscape of possibility. Sophia Parker extends this argument, noting that when institutions are caught “between worlds,” the familiar tools of optimisation cannot open up a larger reality – they only make us better at navigating the status quo.

This is precisely why the re-emergence of mission-oriented and responsible approaches to policymaking – used here as an umbrella for systems thinking, futures thinking, user-centred design, agile methods and other exploratory practices – has put the limits of the administrative system in the spotlight. Instead of becoming a routinised way of organising systemic change, missions – and the transformative methods attached to them – repeatedly collapse back into incrementalism and business-as-usual routines when they are forced through an architecture built for stability.

See Kattel & Mazzucato on why missions struggle to make an epistemic turn within existing bureaucratic logics, Demos Helsinki’s “Missions for Governance” on how missions constrained by current mechanisms tend to produce only incremental improvements.

So, if the architecture is the limit, the question becomes not how to replace the existing system (we rely on its stability more than we admit), but how to create an overlaying architecture that can be activated when uncertainty is high and when public value depends on exploration. Geoff Mulgan refers to this type of bureaucracy as having “strength without weight,” the ability to use institutional power without the bureaucratic drag that stifles learning – enabling decision-making supported by intelligence, curiosity and adaptability, rather than the machinery of implementation and control.

The capabilities the state needs to work under uncertainty

Any discussion of a second, more agile organisational architecture for navigating uncertainty and producing public value must begin with capability. Not whether an organisation has the right structures on paper, but whether it can learn, coordinate, and adjust as conditions change. Working under uncertainty – and delivering any form of transformative, systemic change – requires a different way of thinking about both public value and the organisational capabilities that sustain it.

First, public value cannot be treated only as an outcome delivered at the end of a policy process, or as a means-to-an-end justified through appraisal. As Mark Moore argues, it is both an outcome and something produced through the policy process itself. Under conditions of uncertainty, the quality of these processes is not incidental; it is constitutive of public value, shaped across an organisation’s mission, sources of legitimacy and operational capacity. Treating value as both an outcome and a process provides a practical way to judge whether particular actions are improving a situation and are worth their costs.

This shifts attention to the dynamic capabilities embedded within public organisations: the capacities that allow institutions to sense change, coordinate across boundaries and deliberately reshape resources and priorities as the environment shifts. These are not specialist innovation skills or isolated tools, but foundational capabilities – central to public value management – that determine whether an alternative decision-making architecture can function at all.

This aligns closely with work on Human Learning Systems, which frames outcomes as emergent properties of complex systems and treats learning, adaptation and relationship-building not as phases of reform, but as the core work of public management under complexity.

This is where the Institute of Innovation and Public Purpose’s work on dynamic capabilities is particularly useful. For this discussion, it helps clarify both what decision-making under uncertainty requires and why so many innovation efforts struggle to influence practice. Three dynamic capabilities are especially relevant here:

• Sense-making: the ability to scan an environment, surface weak signals, and develop shared interpretations of what is happening and why. This is not about better forecasting or more data, but about collective orientation when evidence is partial, contested, or evolving.

• Connecting: the capacity to coordinate across organisational, disciplinary, and sectoral boundaries. Under uncertainty, no single team or function holds the full picture. Effective decision-making depends on integrating different forms of knowledge and negotiating trade-offs across silos that were never designed to meet.

• Shaping: the ability to reconfigure priorities, resources, and routines in response to what is being learned. Without this capability, insight accumulates but decisions remain locked in – learning exists, but it has no leverage.

Taken together, these capabilities frame public value as something produced through decision-making itself (or “strategic management” if you would rather). How problems are framed, whose knowledge is included, when commitments are made, and whether learning is allowed to alter direction all shape whether public value is created – or quietly eroded.

Why innovation efforts collapse back into business as usual

The difficulty is not a lack of innovation activity, it’s where that activity sits relative to how decisions are actually made. Uncertainty demands dynamic capabilities such as sense-making, connecting, and shaping. Yet, most public-sector decision-making architectures are designed to steward stable systems, not to navigate contested, evolving or genuinely new ones. This creates a structural mismatch where even when teams attempt to work differently – through experimentation, design or mission-led approaches – the surrounding architecture pulls decision-making back towards certainty.

Public Digital’s The Radical How illustrates this issue. In the dominant waterfall model of policymaking, the most consequential decisions are made upfront – precisely when uncertainty is highest – while learning arrives later, once direction has already been fixed. The process follows a familiar pattern: Write policy → guess requirements → procure IT systems → Inflict on users at scale → operate indefinitely.

Seen through this lens, learning is systematically pushed downstream, when its capacity to shape outcomes is weakest. Innovation efforts are not failing because they lack ambition or quality, but because they are layered onto an architecture that is structurally hostile to uncertainty.

This pattern repeats across governments. New methods are introduced, capabilities are developed, and insights are generated – but only after the moments when direction is set, resources committed, and alternatives foreclosed. The result is activity without leverage: innovation that is visible and energetic, but ultimately powerless.

Three examples help bring this structural mismatch to life.

1. Policy Labs

Policy labs are among the most significant attempts to embed learning into policymaking through user-centred and systems perspectives. They are designed to intercept policy ideas early and improve decision quality (see Anna Whicher on the evolution of Policy Labs in the UK). In capability terms, labs are particularly strong at sensemaking in many ways. A large part of my own work is surfacing lived experiences, identifying risks and challenging the assumptions of policy teams while at the same time bringing together multidisciplinary actors who would not normally work together and creating relational infrastructure across policy and delivery siloes.

The problem, however, is not what labs do, but when they are able to do it. Within the existing decision-making architecture, the moments when assumptions can genuinely be challenged are tightly constrained. Policy labs rarely control the decision gates where commitments are made, budgets locked in or political capital spent. Their insights enter policy processes already moving toward closure. Learning exists, but it arrives downstream of power.

As a result, insight is absorbed rather than acted upon – often filed away in the proverbial “policy drawer” while assumptions remain untested. Labs generate understanding, but without authority to reopen decisions, that understanding struggles to shape outcomes.

2. Technology-centric or challenge-driven funding

Technology programmes and challenge-driven funds apply shaping pressure differently. They create momentum around emerging technologies and capabilities, directing attention, investment, and experimentation towards what might be possible. In doing so, they generate valuable insight into technical feasibility and future options — often through pilots, prototypes, and partnerships that test new approaches in practice.

This activity frequently supports sense-making as well – imagining new uses of AI, automation, or data at scale, and expanding the perceived solution space.

The problem is not the quality of this learning, but its position in the decision-making system and who it involves. Insights generated through challenge funds rarely travel upstream into everyday policy work. Policy teams struggle to access or apply what has been learned – not through resistance or lack of interest, but because there are few institutional pathways for the diffusion of knowledge.

As a result, innovation happens around the policy system rather than within it – shaping the environment without reshaping the decisions that govern it. Insight accumulates, but without leverage where it matters most (see my note on the frustration of AI pilots here).

3. Generalist training, hackathons and experimentation initiatives

Perhaps the most compressed illustration of the problem comes from generalist training programmes, hackathons, and short-term experimentation initiatives. These are designed to introduce brief moments of learning and collaboration – short bursts of sense-making and connecting – often framed as ways of normalising experimental decision-making.

Yet, in practice, they frequently do the opposite. Because these initiatives are time-bound, performative and disconnected from approval, commissioning and resource-allocation processes, their outputs struggle to survive contact with existing governance routines. What is learned rarely reshapes decisions; instead, participants return to an architecture that rewards certainty, compliance, and delivery over adaptation. It is often theatre over substance.

Viewed in isolation, these initiatives can appear valuable. Participants report insight, energy, and motivation. But because they sit outside the moments where commitments are made, they offer no sustained pathway for learning to translate into public value.

This is not a failure of individuals or intent. It reflects a deeper structural issue. When learning is experienced as an exception rather than a condition of everyday decision-making, it is easily dismissed as “interesting but unrealistic.” Without repeated, consequential exposure to uncertainty – where learning genuinely alters direction – counter-intuitive ways of working struggle to stick.

Across each of these examples, the same pattern repeats.

Learning happens, but too late. Insight accumulates, but without leverage. Innovation is active, but powerless.

Seen through the waterfall architecture, this is not surprising. These mechanisms are attempting to introduce uncertainty, exploration, and adaptation into a system designed to eliminate them as early as possible. This is why mission-oriented approaches, design methods, and experimental practices so often collapse back into incrementalism: they are forced to operate after direction is set, resources committed, and alternatives foreclosed.

This is not a capability problem. It is an architectural one. Until the conditions exist for sense-making, connecting, and shaping to influence decisions when they matter most, innovation efforts will remain episodic and low-impact. Which leads to the next question: what would an alternative – complementary – decision-making architecture need to look like?

An alternative (mission-driven) decision-making architecture

Rather than replacing existing policymaking systems, this section sets out an overlaying decision-making architecture designed for moments of uncertainty. Drawing on Vinnova’s mission-design framework – and learning from the UK Government Digital Service’s agile delivery model for digital services – the model below translates these insights into a participatory decision-making architecture for moments of uncertainty. Its purpose is not to generate better analysis, but to change how direction, commitment, and resourcing decisions are formed by deliberately activating the dynamic capabilities discussed earlier: sense-making, connecting, and shaping.

Put simply, this architecture embeds learning throughout the policy design and delivery process, rather than treating it as something that arrives after key decisions have already been made.

This architecture rests on four principles.

1. Conditional, not universal: It is activated only where uncertainty is high and stakes are systemic – not as a default mode of governance.

2. Triggered by uncertainty, not default: It complements, rather than competes with, existing stewardship systems.

3. Relational and procedural, not structural: it works through roles, practices and decision rights, not new hierarchies. This means a focus on making stakeholders active participants of the design process.

4. Time-bound and reversible: Its purpose is to resolve uncertainty, not institutionalise exploration indefinitely. Results matter.

Critically, it also depends on what Dan Hill describes as “soft eyes”: the capacity to observe and work with the dark matter of bureaucracy – the informal relationships, tacit norms and interdependencies that sit between teams, directorates and policy domains. Without this, attempts to engage uncertainty are either blocked by premature certainty or absorbed back into business-as-usual routines before they can influence decisions.

This is not an abstract sensibility either. Variants of this capability have been articulated over the past decade through relational practices such as Relational Service Design, Human Learning Systems, Dark Matter Lab’s Governing Together, and the TEAL movement, which foreground trust, continuity, and institutional relationships as core operating conditions rather than delivery-side effects. The challenge is not their absence, but their marginal position relative to where decisions are actually made.

What follows sets out how this overlaying architecture works in practice: the stages it moves through, the capabilities it embeds, and how it interfaces with the state’s existing delivery and stewardship machinery.

At a high level, the architecture is organised around four overlapping learning loops. These are not phases or stage gates. They are modes of work that structure how learning, judgement, and commitment evolve under uncertainty. What distinguishes this model from familiar innovation or design frameworks is not the loops themselves, but where decision authority and resource commitment sit within them.

The flow follows Vinnova’s mission-design framework, moving from Angles → Missions → Prototypes → Demonstrators, progressively reducing uncertainty while keeping direction open for as long as possible.

A. Angles: orienting the system

The Angles loop focuses on collectively developing an initial view of the system: identifying where uncertainty lies, what matters, and where intervention might be possible.

Rather than jumping to solutions, this loop brings a core group together to surface different ways of seeing the problem space. It deliberately draws together a trans-disciplinary team of policy, operational, user, technical, and political perspectives to build a shared orientation to the system as it is experienced, not just described.

The emphasis here is on sense-making and connecting. The output is not a strategy or recommendation, but a set of plausible directions – angles of intervention worth exploring further.

B. Missions: committing to direction

The Missions loop translates the most promising angles into statements of intent and design principles. These articulate what kind of change is being pursued and why, without prematurely specifying how it will be delivered. Design principles matter because they act as decision constraints under uncertainty – a function GDS has long used to align teams without over-specifying solutions.

At this point, connecting remains central – aligning actors around a shared purpose – while shaping begins to matter more. Missions create directionality by setting boundaries, clarifying intent, and establishing criteria for learning in the next stages.

Crucially, missions are treated as provisional commitments. They are strong enough to mobilise action, but flexible enough to adapt as learning accumulates. Their role is not to close down debate, but to focus it – similar to futurestate strategy or strategic foresight practices.

C. Prototypes: learning through tangible commitments

The Prototypes loop explores missions through a portfolio of concrete experiments – policy, service, regulatory, or organisational – designed to test assumptions and surface consequences.

Prototypes are decision-shaping commitments: deliberately designed to inform judgement about whether, how, and where to scale. Multiple prototypes ask different questions, reducing uncertainty through comparison rather than consensus. Vinnova’s Street Moves programme is a good example: place-based experiments used not as pilots, but as learning devices.

Here, connecting and shaping are tightly coupled. Prototypes act as boundary objects, aligning actors, mobilising resources and enabling learning through doing. The emphasis is not speed for its own sake, but efficiency: learning as much as possible for the least irreversible cost. Or, in Public Digital’s “The Radical How” terms: doing whatever it takes to speed up the loop of testing the things that matter the most and learning from the results.

D. Demonstrators: integrating learning at scale

The Demonstrators loop brings together insights from multiple prototypes into large-scale system demonstrations. These are not end states; they are decision-shaping devices. Their purpose is to determine whether uncertainty has been reduced enough to re-enter the state’s core delivery and stewardship architecture.

Here, sense-making resurfaces, alongside continued connecting and shaping, as priorities and resources are adjusted in response to what has been learned. Demonstrators help decision-makers see how interventions interact at scale before irreversible commitments are made. At this point, work can deliberately transition from exploratory modes into more stable delivery pathways – moving from adaptive learning to reliable execution.

Switching modes

At this point, I want to stress that this proposed second architecture is conditional, not universal. It is activated only when uncertainty is high – when outcomes cannot be reliably predicted, causal pathways are contested, and early lock-in would be costly or irreversible. It’s about introducing a dynamic learning process so that the state can return to its stable delivery architecture once uncertainty has been reduced.

When activated, several conditions matter:

• decision authority must sit close to the work

• learning must have formal routes into commitment and resourcing

• participation must include those affected, not only those accountable

• exit conditions must be explicit

The intent is not to weaken the state’s capacity for delivery, but to protect it by ensuring that stability is not achieved at the cost of learning, and that exploration does not become detached from power. In other words, this is not about replacing the machinery that keeps the lights on. It is about creating the conditions under which the state can decide differently when it needs to – and then return, deliberately, to what it already does well.

Preparing the ground: agile team models when working under uncertainty

Transformative change starts by preparing the ground: creating the conditions for a small, multidisciplinary team to work differently, with political cover, protected space, and the authority to bypass routine constraints long enough to prove what is possible. This was true of the early Government Digital Service in 2011, and it remains true of any attempt to change how the state decides under uncertainty.

Fundamentally, if we expect teams to sense, connect, and shape under uncertainty, we cannot treat them as interchangeable delivery units. Different phases of decision-making require different, malleable team purposes and archetypes that evolve as uncertainty reduces.

Kate Tarling’s work on service organisations is useful here. Rather than assuming static “agile teams,” she shows how teams shift over time: from exploratory, functional groupings with hand-offs, towards whole teams responsible for delivering and improving joined-up services. She stresses that hand-offs are not inherently problematic but they are often necessary early on – with problems arising when teams are locked into a structure that no longer matches the nature of the work and are forced into incompatible handovers.

It is therefore more helpful to think in terms of complementary team types, rather than a single ideal form. Drawing loosely on team typologies developed by Skelton and Pais (and Tarling), a mission-driven architecture typically relies on a portfolio of teams with distinct roles:

• Service teams are responsible for delivering and improving outcomes for users and operations.

• Depth teams, which explore complex issues, generate insight and reduce uncertainty on behalf of others.

• Common-capability teams, which build shared tools, platforms, and processes that make learning and delivery easier.

• Enabling teams, which unblock, coach, and support other teams to adopt new ways of working.

• Coordinating teams, which align activity across services or missions when interdependencies are high.

• Operational teams, which steward stable systems and deliver continuity over time.

Crucially, these teams do not all operate in the same way, nor at the same time. Early in the lifecycle of a mission – when uncertainty is highest – depth, enabling, and coordinating teams play a larger role in sense-making and connecting. As direction stabilises and uncertainty reduces, service and operational teams take on greater responsibility for shaping and delivery.

Seen this way, team models are not an implementation detail, but a core part of the decision-making architecture. Without the right mix of teams – and the ability for them to change as conditions change – even well-designed missions will struggle to move from exploration to impact.

Conclusion

The world is not a machine where, if you pull a policy lever, something will happen. Reconnecting policymaking with this reality means moving beyond top-down, disconnected approaches and towards a decision-making architecture that treats uncertainty as something to be worked with, not engineered away. One that embeds learning, judgement, and adaptation into how direction is set – without sacrificing the stability the state depends on. What’s required is strength without weight. The ability to do the hard things well, while remaining flexible enough to change when conditions demand it.

I want to be clear about what this piece is – and is not. It’s not a story of pessimism, nor a claim that the government lacks innovation or intent. Quite the opposite. There is a great deal of thoughtful, committed work happening across UK government and beyond (See: The state of British policymaking: How can UK government become more effective). This is an attempt to frame those efforts – through theory and lived experience – in a way that surfaces a deeper constraint: not ambition or capability, but the architectures through which decisions are made.

This article is not the answer either. The critique, and the model that follows it, could easily be read as two-dimensional: one person diagnosing a system and proposing a cleaner alternative. That reading misses the point. The issue is not a shortage of models for change. It’s that institutions – Whitehall included – often convince themselves they are uniquely resistant to new ways of doing, while continuing to route uncertainty through systems designed to suppress it.

This is as much a cultural challenge as a structural one. Work on systems change reminds us that transformation rarely follows linear cause-and-effect pathways. It emerges through overlapping activity, partial learning and shifting coalitions over time. For those of us working inside the government, that means holding multiple perspectives at once: staying curious, working across boundaries, and treating thinking itself as a form of action – something that shapes what becomes possible next.

Much of what I’ve drawn on here builds on existing ground: the early work of Government Digital Service, mission-oriented innovation frameworks, action-based approaches to social change, and emerging practices around Human Learning Systems and other relational practices. The open questions, for me, are less about invention than application. How do these principles actually reshape decision-making at scale? What role does digital play – not as infrastructure alone, but as an enabler of learning, coordination and judgement? Have we laid the horizontal foundations for a third era of digital governance, or are we still struggling to move beyond enterprise-era assumptions?

Those are questions for another time.

Next steps…

As with most of my writing on substack, this is thinking-as-writing rather than a finished argument. It’s an attempt to synthesise lived experience wrestle with the ideas and practices I’ve been immersed in over the past few years, to see whether they hold together. It’s written as self-provocations, loose ends, tensions and partial conclusions – that’s part of the point for me.

For now though, this thinking is feeding directly into my own work at HM Revenue & Customs where I’m currently leading a review of our internal Budget Policy Starter technology-impacting process by working with policymakers, solution architects, cost engineers and user-centred designers to better bridge stable and agile ways of working – and, in doing so, improve how policy decisions are shaped upstream so they serve citizens more effectively downstream.

This post reflects on an article I recently published for Recurve on neighbourhood missions and social infrastructure.

Since the announcement of the UK’s Pride in Place programme – a £5 billion, decade-long investment intended to regenerate high streets, tackle deprivation and renew entire areas – there has been a steady stream of thoughtful work on neighbourhood renewal. From the Local Government Association’s Growth and Reform Network and the Independent Commission on Neighbourhoods (ICON) to Local Trust’s Big Local and Place Matters, there is no shortage of strong analysis of the structural conditions shaping neighbourhood decline.

What remains far less developed, however, is how these diagnoses are translated into meaningful policy action outside the dominant logic of growth agendas.

This is no criticism of the quality of these reports. The problem lies elsewhere. While neighbourhoods are increasingly treated as complex social, institutional and civic systems, policy responses are still filtered back through an economic paradigm designed for sectoral growth, programme rollout and aggregate outcomes. The result is a familiar slippage: relational diagnoses, followed by transactional, often technocratic, recommendations (See: A City Is Not a Computer).

In this Recurve piece, I explore why neighbourhood renewal policies struggle not because of a lack of evidence or ambition, but because they remain spatially blind in practice. Not because they ignore context, but because the mechanisms for turning diagnosis into action remain rooted in bureaucratic hierarchies that struggle to work with local complexity and relational depth.

This is not an argument against economics. Neighbourhood policies that focus on jobs, skills, transport and investment all matter. But economics alone is an insufficient organising frame for neighbourhood renewal. As Polly Mackenzie argues, if decline were only about economics, many of the places that continue to exist against the odds simply wouldn’t.

What this points to is an institutional problem, not a market one.

Drawing on work from Hilary Cottam, Sam Rye at Monash University, and research on mission-oriented innovation and neighbourhood renewal, this article argues that despite acknowledgement that relationships are central to sustained, transformative change, public systems designed around managing needs and delivering transactions systematically underinvest in relationships, and that we need greater attention to social infrastructure institutions: intermediary organisations that steward relationships, integrate fragmented systems, and hold learning and coordination over time.

Highlighting Citizen Hub in St Neots, alongside the Monash University living labs model, the article shows how neighbourhood renewal depends not just on programmes, but on institutions capable of sustaining change.

You can read the full Recurve article, The missing middle: why Labour’s neighbourhood missions need social infrastructure, here: https://www.recurve.solutions/insights/rebuilding-missing-middle

As ever, I’d be very interested in reflections – especially from those working within this messy interface between policy, place and practice.

Given Labour won the last general election on the promise of national renewal and a mission-driven state with a “relentless focus on long-term ends” after years of crisis management – frustration is understandable. After the Budget this week, it feels as if political ambition has met the practical reality of sustaining long-term political direction inside institutions designed around stability and risk management. Add a “black hole” in public finances and sporadic and often contradictory regulation, and it has rarely been harder for public servants to know what to decide – or, critically, how to make decisions in the first place.

Drawing on my work in early-stage policymaking at HMRC and research at UCL’s Institute for Innovation and Public Purpose, this short piece of “writing as thinking” argues that the UK’s public institutions are structurally incapable of navigating uncertainty because they are built on a decision-making logic designed for predictability, control and linear planning – not for complexity, learning or adaptation. Because, if the UK is serious about missions, or radically overhauling how the civil service thinks and works, the everyday practicalities of how decisions are framed, contested and made will need to change.

The issue of managing uncertainty

Across government and beyond, a lot of my work has been at the earliest stages of decision-making – where ideas are half-baked, assumptions untested, and uncertainty is at its highest. In these spaces, one of the things I see teams struggle with the most is the ability to be agile enough to experiment, learn and iterate while being stable enough to maintain direction and meet deadlines. In fact, as Rainer Kattel and James Plunkett have noted, institutions are often both too stable and too agile — highly proficient at producing procedures and frameworks that promise control, yet poor at learning, experimenting, or re-orienting in the midst of uncertainty. The result is an all too familiar blend of paralysis, short-termism, and “everythingism”: needing to respond to everything while changing little.

This dynamic is not accidental. It is a lasting hangover from New Public Management – a model that has compartmentalised and hollowed out the civil service in order to make it measurable, auditable and controllable. Designed for an era of fewer citizens, slower information flows and limited data capacity, our civil service is oriented around prediction and control despite vast improvements in data, research and digital infrastructure. Put simply, we have left ourselves in a situation where we are taking 21st century challenges, evaluating them with 20th century tools, and responding within 19th century institutions built in a different time and for a different time.

In “Seeing Like a State”, James C Scott describes this as the state’s drive for “legibility” – making societies administratively visible by simplifying their complexity into categories that can be planned for. To illustrate this he uses the example of modern scientific forestry in Prussia. In an effort to maximise timber output, foresters cleared mixed woodlands and replanted them in perfectly straight rows of a single fast-growing species. The first generation produced higher yields, fantastic. But, by the second and third generations, yields collapsed. The drive for order – monoculture, straight lines, predictability – had unintentionally erased the complex, interdependent ecology that sustained the forest and made it viable in the first place.

New Public Management simply extends this logic into government, treating institutions as machines to be ordered rather than ecosystems to be stewarded. It rewards prediction over understanding and linear planning over iterative learning in what Funtowicz and Ravetz call a “doctrine reinforced by guesswork,” reliant on cost–benefit analysis and other appraisal tools built for marginal change in predictable environments. These methods make little room for uncertainty or contested values because they flatten complexity, obscure causality and routinely miss the non-linear, cascading effects that dominate policymaking. The result is a higher risk of expensive failures, often harming those most in need.

More fundamentally, this mode of public problem-solving is simply too slow and too distant from the people it is meant to help, offering little opportunity for learning or course correction in the first place. And, for policymaking – a forum directly concerned with people and their behaviour – it is simply not tenable to expect civil servants and ministers to predict, design and justify interventions up-front without addressing these structural constraints. Put as many policy design cycles in front as me as you would like, the “waterfall” model of policy delivery, illustrated in The Radical How, makes this explicit. Planning happens early, implementation happens late, and evaluation tends to arrive long after the window for improvement has closed.

Recent literature on Human Learning Systems pushes this argument further, describing how this creates a “pathological culture of blame and defensiveness,” within the civil service, where ambiguity is treated as error and risk is transferred onto individuals. Even when uncertainty is real and unavoidable, public officials are expected to present polished cases that imply certainty, rather than surface complexity or challenge assumptions – something I see time and time again within HMRC. What this ultimately means is that talented civil servants (of which I work with many) become trapped in an environment where they are punished for being wrong, rewarded for nothing going wrong, and rarely recognised when things go right.

This paralysis and culture of negativity isn’t individualised. I work with and have met so many talented, driven and imaginative civil servants. However, because the institutional grammar, design and hierarchy of decision-making treats uncertainty as a threat to be contained rather than a condition to be worked with, they are not given the time or space to do their best work. We have designed our organisations to reduce risk and cost under the premise of control but failed to appreciate how work actually gets done in complex systems – ultimately treating learning as political performance.

Recent reflections from municipal leaders in China – long associated with rigid target-driven governance – note that obligatory numerical targets distort behaviour, narrow attention and create perverse incentives within institutions.

Ultimately, this manifests as both a bias towards inaction and an inclination to make decisions when the least is known. Jen Pahlka memorably phrases this as “policy vomit” – in which the words of the policy are vomited, essentially undigested, into the administration of a policy, without the benefit of thoughtful, learning and iteration. In a department like HMRC, where impact is traditionally inseparable from cost, this often means using cost-benefit analysis (as suggested by the UK government’s guide to policy appraisal – the ‘Green Book’) as a means to centre on the risk of doing something rather than the opportunity for it to go well – effectively excluding learning from the process altogether and overlooking or under-considering real-world dynamics.

Towards managing missions or muddling through?

Over the last decade, governments and policymakers have been increasingly turning to mission-oriented innovation as a way to break out of incrementalism and tackle the “wicked” problems causing so much trouble for our ways of working today. It signals a shift away from the narrow technocratic logic of correcting market failures towards a more ambitious role for the state in shaping and co-creating markets, mobilising collective action and directing investment towards clearly defined public purposes.

Labour’s commitment to a mission-driven government, not dissimilar to those across Europe as part of a broader “third generation” of innovation and policy frameworks, was meant to embody this shift – a break from short-term crisis management and a renewed focus on long-term ends. Essentially, a recognition that neoliberal and technocratic policy tools are fundamentally unsuited to steering complex socio-technical systems as they privilege efficiency, prediction and control over relevance, learning and adaptation.

There is plenty of noise on mission theory. The problem, however, is that while there are many practical examples of a shift towards achieving mission policy – and more pragmatic, learning centric ways of policymaking – any meaningful practices generated by this new approach to policy have struggled to take hold.

Demos Helsinki’s “Missions for governance” paper reports that a 2022 survey of more than 40 OECD countries found that only one in four mission initiatives had a clearly defined target; only 15% had a dedicated governance structure; and only one in ten had a plan for monitoring and evaluation. In other words, while governments are embracing the language of missions, they are implementing them through business-as-usual structures, inherited from New Public Management, rather than transforming the way we make decisions.

This is perhaps indicative of a larger problem of institutional change. While missions and many other forms of experimental governance look to give us a blueprint for reigniting innovation policy, they do not help shift the entrenched challenges that government faces internally. It makes me wonder – we’ve been practicing systems thinking, futures thinking and experimental forms of governance for decades (with missions as the new synthesis of some of these tools) yet regardless of how good they sound, they haven’t scaled or had the impact any of them promise because they have all failed to address these entrenched challenges (See an interesting expansion of this point here).

To me, this all suggests that regardless of the approach, the need for a reimagined architecture of our institutions has never been more important if we want missions to be actually transformative rather than symbolic. Missions cannot be delivered through a policymaking architecture optimised for certainty, marginal change and after-the-fact evaluation. They require a state capable of working with uncertainty rather than suppressing it — one that builds learning into the fabric of everyday decision-making, rather than treating it as an optional add-on.

Reimagining public value must begin in the texture of everyday decisions.

If missions are to mean anything, the civil service needs to rebuild its capacity to create public value as a practice. That requires a new grammar of decision-making: one that treats uncertainty as navigable, frames choices in terms of risk and opportunity rather than cost and control (See Simon Sharpe), and builds learning into the routine choreography of how problems are defined and acted upon.

This shift depends on everyday practices, not high-level strategies: developing shared strategic alignment; negotiating problem definitions with those affected; testing ideas early; iterating in public; and learning at portfolio scale rather than programme-by-programme. These are the capabilities that make missions possible – and without them, missions will remain symbolic rather than transformative. I will write about those practices elsewhere and be sharing my MPA thesis “(Re)imagining Mission Neighbourhoods: Emerging Practices of Urban Transformation” soon.

For now, the point is simple. If the UK is serious about missions, it must first change how everyday decisions are made.

Following a strong period of economic recovery, Ireland has positioned itself as a noteworthy participant in digital governance. Surrounded by world-class digital firms, the country is now (finally) starting to leverage technology for effective public service.

This essay examines Ireland’s current push to modernise its public sector through digital transformation, with a spotlight on healthcare. It outlines how strong ambition and new frameworks are moving Ireland forward, while also addressing the persistent challenges from fragmented legacy systems and slow progress in e-health. Drawing on international best practices and a user-centred approach, the piece explores whether Ireland can bridge the gap between policy aspirations and the delivery of genuinely effective digital services.

The Irish government are quite happy with their progress toward their Harnessing Digital policy framework, with the “ambition to continue to be a European and global digital leader, and our strong commitment to progressing the digital transformation of our economy and society.”

An eighth-place ranking in the 2023 OECD Digital Government Index would have you think that Ireland has its act together. And it does, to a degree. An intertwined series of policies, largely driven by a whole-of-government approach to implementing its Better Public Services initiative shows a very nice pathway to progress, full of good intentions.

Many would (rightly) argue that graphs, charts, and frameworks often don’t mean all that much by themselves. Governments boasting about alleged digital superiority, while ignoring the vast shadow cast by private sector development is like “celebrating being the best laggards in the world”, as one of our professors at UCL’s Institute for Innovation and Public Purpose recently put it.

But those in Government who champion these rankings, must live by the sword and die by the sword. Across the European Union, Ireland ranks dead-last in e-health performance per its EU 2024 eHealth Indicator Survey. By quite a distance, too: coming in at a participation-medal rate of 11%, compared to second-last Czechia, at 51%.

The HSE are acutely aware of the lagging nature of e-health in government, expressing the multiplicity of challenges across fragmented data, legacy systems, and historical underinvestment in their Digital Health Strategic Implementation Plan:

Our health service is held back with inconsistent, paper based and siloed documentation of patient interactions. Paper based record-keeping such as these are time-consuming, inefficient, and often susceptible to errors. Consequently, essential patient information remains trapped in isolated repositories, preventing an integrated approach to patient care. The result is a lowered efficiency of healthcare delivery and potential compromise on the quality of patient care.

Recognising the urgency to do something, the government has launched Digital for Care: A Digital Health Framework for Ireland 2024–2030 in the hopes of improving the situation, with the digital transformation of the Health Service Executive (HSE) at the forefront of its efforts.

A recent survey by Ireland’s Health Information and Quality Authority (HIQA) showed that 91% of the public and 88% of healthcare professionals think that people will be better informed if they have access to their health records. 74% of the public said that they would access their online health record when it is made available online. Achieving the robust digital transformation envisioned for Ireland’s e-health sector necessitates a shift beyond mere technological upgrades. The Government needs to fully embrace user-centred design.

The imperative of user-centred design

Enter Matt Edgar, a service design expert who argues that we need to view health transformation holistically. His work on service design provides invaluable insights to guide health transformation, emphasising the imperative to begin from the patient’s perspective. Thinking about their whole digital journey, from the initial referral process to actually getting to the appointment, while remembering their expectations are shaped by that slick banking app they use every day.

Edgar stresses the importance of holistic design, advocating for research into patient experiences across the entire service, rather than within isolated organisational siloes. He argues that people experience health as a continuous journey, not as a series of disconnected parts and if Ireland’s e-health efforts are going to be successful, they will need to provide a joined-up set of healthcare services, placing control in the hands of patients as experts of their own lives.

Edgar is adamant that it’s not just about the patients; staff need the tools and operational excellence to work across these digital services, too. They need to understand the patient’s journey, not just their part in the process. Flexible systems that show useful information about each patient (like their needs and support networks) are crucial. Edgar rightly points out that if the system is a pain for staff to use, they won’t recommend it to patients. Simple as that.

So, the plan? Ireland’s e-health needs to focus on creating systems that don’t make you want to rage-quit and throw your phone out the window. Intuitive interfaces are essential.

If Ireland manages to weave these user-centred principles into its e-health efforts, it might actually create services that are both secure, efficient, and, dare we hope, useful. It could even start to rebuild some of the public trust lost during the 2021 HSE cyberattack — when everyone’s data was at risk. As Ireland keeps investing in its e-health infrastructure, it needs to put some of that investment towards proper design and a genuine commitment to services that put people first.

HSE Health App: A step towards digital healthcare

Ireland’s Health Service Executive (HSE) has taken an important step in its e-health transformation journey with the launch of its Health App in February 2025. Developed under the Digital for Care framework, this app is designed to serve as a “digital front door” to health services, marking a significant milestone in digitising patient health records. Initial features include viewing maternity service appointments, creating digital medication lists, storing health insurance information, and accessing vaccination records. While these functionalities are promising, the app’s true potential lies in its development approach.

The HSE Health App reflects a strong commitment to user-centred design principles. Co-designed with patients, staff, and advocacy groups, the development process prioritised inclusivity and accessibility. Consulting with various patient organisations and forming a Patient Advisory Group ensured that diverse perspectives were incorporated into the app’s design. This collaborative approach was about creating an app that genuinely meets the needs of its users, including individuals with additional needs.

But this isn’t just about inclusivity; it’s about adaptability. The app’s development roadmap includes planned expansions and enhancements based on ongoing research and user engagement, and is currently inviting public servants to register for testing of the app. This iterative approach ensures that the app evolves alongside user needs and technological advancements. It’s clear that the HSE is aiming for more than just a static product, but rather it wants a service that grows and improves over time.

Agile Delivery: Borrowing Lessons from Across the Water

If some elements of the HSE Health App’s rollout feel familiar, it might be because Ireland has borrowed heavily from the UK government’s Service Manual for agile project delivery. This framework breaks down development into distinct phases: Discovery (understanding problems and user needs), Alpha (prototyping and testing assumptions), Beta (refining based on feedback), Live (launching with ongoing monitoring), and Retirement (decommissioning when necessary). It’s a structured approach that has proven effective in delivering public sector digital services, and Ireland seems keen to follow suit.

The phased rollout of the HSE Health App aligns closely with these principles. Its initial launch — offering limited features like maternity appointment viewing — functions effectively as a Beta phase. This allows the HSE to gather real-world usage data, identify areas for improvement, and refine functionality before expanding its scope. Plans to introduce additional features based on user feedback reinforce this agile approach, mirroring the iterative nature of both Beta and Live phases.

The co-design process further reflects agile principles from Discovery and Alpha phases by involving stakeholders early on to identify key priorities and challenges. By engaging patients, staff, and advocacy groups from the outset, the HSE ensured that pressing issues were addressed while designing an app that is intuitive and accessible for all users.

While some might argue that Ireland is simply copying best practices from its neighbour across the Irish Sea, there’s no denying the practicality of this approach. After all, if something works well elsewhere, why reinvent the wheel? What matters most is whether Ireland can adapt these principles effectively to meet its unique healthcare challenges.

Integration with the Government Digital Wallet: A litmus test

The next big test for Ireland’s e-health ambitions is already on the horizon: integrating the HSE Health App with the Government Digital Wallet slated for launch later this year.

This initiative is part of Connecting Government 2030, Ireland’s broader strategy to create seamless digital services across public sectors. The Government Digital Wallet promises to revolutionise how citizens interact with public services by securely storing digital versions of essential documents like driving licenses, birth certificates, and European Health Insurance cards.

This integration will serve as the litmus test for Ireland’s ability to transcend traditional departmental siloes and foster collaboration across government agencies. Success could set a precedent for future cross-departmental initiatives while paving the way for more efficient and citizen-centric public services.

However, challenges abound. Ensuring robust data privacy across interconnected systems will be critical in building public trust. Creating seamless user experiences across diverse services will require careful planning and execution. And maintaining flexibility to adapt to future technological advancements will be essential for keeping pace with evolving citizen needs.

Despite these hurdles, there’s reason for cautious optimism. If Ireland can navigate these challenges effectively, it could establish itself as a leader in integrated digital governance — not just within healthcare but across all public services. The integration of health services into this broader ecosystem will be a key indicator of progress toward creating a cohesive digital infrastructure that enhances citizen experiences at every level.

Share

Subscribe now

Leave a comment

I’ve felt a very welcome shift in tone across the World Design Congress, the Global Public Design Conference and SD in Gov these past few weeks – one i’ve taken some time to reflect on.

Pulling together threads from James Plunkett, Christian Bason, Kara Kane, Omar Idris, Andy Marsden (and many more), the emphasis has been less about technical slickness or defending methodological standpoints and more about cultivating a collective designerly way of doing things rooted in the humanity and humility of creating better public experiences.

A fantastic example is the Irish Government’s Department of Public Expenditure NDP Delivery and Reform Life-Events programme, which moves beyond department-focused service delivery to align cross-department services around singular life-event experiences – like welcoming a new child or finding a place to live. This kind of institutional reform, and the alignment of skills behind outcomes rather than structures, captures this shift in practice beautifully.

More broadly, this change reflects what Indy Johar framed at WDC as an invitation to design differently in order to design within a fundamentally different reality — ecological, social and economic — in a 3-degree warmer world. Hillary Cottam, building on Carlota Perez, anchors us in this moment in her book The Work We Need, reminding us that although periods of socio-technological transition are turbulent, they also carry the energy and promise of a new kind of future. That’s the energy at the edges James Plunkett has spoken and written so persuasively about.

This positioning also gives weight to what so many speakers have pointed towards, and was perhaps captured best in Mariana Mazzucato and Kate Raworth’s WDC conversation: that our systems have been designed – which means they can be redesigned.

Whether it’s alternative theories of economics that move us away from technocratic, neoliberal approaches and towards growth with direction and social benefit (such as Regen Melbourne’s *Melbourne Doughnut* or Leuven 2030’s mission for climate neutrality), new forms of civic democracy like Choose For Cheetham’s youth participatory budgeting, or the civic capabilities required to deliver them (which I think design sits at the heart of, and the UK’s Public Design Review is the start of) — the real question is not whether redesign is possible, but whether we’re prepared to take responsibility for it.

Of course, this shift in tone doesn’t come without tension. We’re speaking about challenging the entrenched paradigm of designing within an ultimately unsustainable set of systems – this means juggling expectations, perspectives, power and the complexities of actual human beings who are doing the work. This is perhaps (ironically) what is missing from design-led conversations. The voices of citizens, policymakers, developers and planners that are all involved in making change happen.

So where do we start? One of my highlights from this past few weeks was a session run by Julian Thompson, Ashley Evans and Courtney Mclean from Rooted by Design on reparative design. They introduced a set of principles that demanded a more participatory approach to systemic change, poking and prodding participants to question their own positionally and capacity for transformation. It was an important reminder: if our work wants to achieve systemic change, we must find ways to bring in voices beyond our profession, departments, institutions and social perspectives – not as an afterthought, but as the very condition of the work.

I can’t claim to have left the session prepared to design equitable places, policies or services but I do believe it helped open up my capacity to hold these questions in my own work – and to design ways that supports others in doing the same.

I don’t believe for a second design holds all the answers to the systemic issues we face. What I do believe is that it’s a malleable practice – one that allows us to hold systems lightly, to wade through the uncomfortableness of complex systems and differing perspectives, and to build the collective capacity to confront what we might otherwise avoid. This is not a solution in itself, but it is a start.

I’ve enjoyed reading reflections and thinking from across my network on these conferences and i’d love to keep the conversation going – it has also been a pleasure to speak more widely about my recent MPA thesis on urban mission practices culminating in the opportunity to do a lightning talk at SD in Gov.

References and resources (in order of appearance)

• James Plunkett – Reflections on design practice (LinkedIn post): https://www.linkedin.com/posts/james-plunkett-a1472827_yesterday-i-had-the-pleasure-of-giving-a-activity-7374370000444506112-3LmA

• Department of Public Expenditure, NDP Delivery and Reform, Government of Ireland – Life-Events Programme Case Study: https://publicpolicydesign.blog.gov.uk/wp-content/uploads/sites/255/2025/03/CASESTUDY.IRELAND.LIFEEVENTS_M.CASSIDY_MAR25.pdf

• Hillary Cottam – The Work We Need: https://www.hilarycottam.com/the-work-we-need/

• James Plunkett – The Energy at the Edges (Medium article): https://medium.com/@jamestplunkett/the-energy-at-the-edges-8732ac49ecc9

• Regen Melbourne – Melbourne Doughnut: https://www.regen.melbourne/news/melbdoughnut

• Leuven 2030 – EU 100 Climate-Neutral Cities Mission: https://en.leuven2030.be/

• Shared Future CIC – Hope for Cheetham: Youth Participatory Budgeting: https://sharedfuturecic.org.uk/hope-for-cheetham-youth-participatory-budgeting/

• UK Government – The Public Design Review (Evidence Review): https://www.gov.uk/government/publications/the-public-design-evidence-review

• Rooted by Design – Rooted Centre: https://rootedcentre.org/

Hi, I’m Jack Strachan. I’m a designer, researcher and strategist. I work with policymakers, digital teams and a whole host of civil servants – blending design, strategy, and policy tools to rethink public value, shape early policy ideas, design digital services, and wrestle with technology.

I’d like to introduce CIVICWORKS: a space for sharing thinking on civic bureaucracy, institutional reform, dynamic capabilities, policymaking and technology. My intention for this space is to be intentionally transdisciplinary – bridging the richness of civic design at the street level with the complexity of policymaking at the state level and beyond, drawing on as many perspectives as possible to create interesting angles, takes and provocations.

I don’t envision this as a space confined to my own thoughts either; I’m hoping to leverage a growing network of incredible practitioners and academics to share their insights and thinking – to form what I hope will become a platform for a diverse collection of thought pieces and insights into the future of public policy and the built environment.

I won’t linger on any introductions; rather, I would prefer the writing to speak for itself. Fundamentally, CIVICWORKS is about surfacing interesting people’s thinking on important civic questions – creating a space that invites thought, conflict and plurality.

Some of the first few pieces of writing will be on the OECD’s recommendations for human-centred public administration services, the economics of decision-making in the civil service and how the UK’s public design review can influence the future of civic capabilities.