In permanence crisis

An argument for state capacity and civil service reform

"A state without the means of some change is without the means of its conservation."

Edmund Burke

State capacity is the ability of a government to turn political aims into public progress. The UK has a deficit. I think we can do better.

I’ve written about state capacity before, most recently with my brilliant collaborator Jen Pahlka for the Niskanen Centre in DC. Many democracies are grappling with this challenge. But I will focus on home here.

In case you want to get off this train now - and fair enough - here’s the destination. A government that is serious about state capacity must do 2 things, urgently and in parallel:

One, fundamentally reboot the character of the ‘permanent’ civil service, beginning with an independent Royal Commission like the Victorian-era Northcote-Trevelyan report that founded the modern bureaucracy.

And two, create new institutions that adopt the working styles, technologies and affordances given to public officials who have shown how the state can do great things in special circumstances. A corollary of creating new institutions is that some existing, legacy organisations must be encouraged to wither and die.

Both of these are necessary; just one is not sufficient.

With a new Cabinet Secretary about to arrive, an administration that has repeatedly said it plans to ‘rewire the state’, and growing consensus across the political spectrum that something must be done, we have a rare opportunity to do both.

---

In the spirit of Dominic Cummings, who has had plenty to say about this topic (of which I agree with much, but not all), this post is of grossly self-indulgent length. His 2014 post ‘The Hollow Men’, which I remember reading when it was published, had a sense of catharsis running through it; I suppose I’m searching for similar.¹

To prolong the agony, I’m going to split this into 3 chapters. The first will go over why state capacity is important, why it is urgent something is done about it in the UK now, and what the arguments are for prioritising something which is both fiendishly difficult to land and - this said almost as an article of faith by many - ‘of no interest to ordinary people’ because it is complex and, frankly, quite boring.

The second part will argue that the UK government needs to set up a formal, independent Royal Commission on the future of the Civil Service in the style of Northcote-Trevelyan. I’ll write about what happened the first time around, how such a Commission could work now, what it might explore, and why I think setting up something that appears to be another bloody review and a Very Establishment sounding thing is a necessary act of radicalism.

The third and final part will talk about the need to create new teams in government that work in different ways. Ways that have characterised many of the most successful things governments (of different parties) have achieved to public acclaim in the last 30 years. These have almost exclusively happened because of exceptional circumstances and exceptional individuals. They have succeeded despite the orthodox systems, skills and processes used within government.

Some final bits of throat clearing. First, this isn’t written with a party political slant. I’ve had the chance to work with both Labour and Conservative ministers in the last ten years. I don’t think it’s fence sitting to say that addressing state capacity is anything other than a ‘national interest’ issue. If this is for anyone then it is for those in power, as the sitting government are the only people in a position to do anything about it. I’m pro-things working better and pro-democracy, and less good at tribalism.

Secondly, some of what follows that will be critical of the civil service. So it’s important to say: I like the civil service. I was a civil servant for 7 years. I have friends and family who work for it. I fully believe in its power to do great things for the country. That ability to do good is what animates me to want it to be better, more often. I’ve no interest in bashing bureaucrats, don’t subscribe to ‘blob’ characterisations, and think that plenty of other actors share the blame for the UK’s current level of state capacity, including politicians of all parties. I will avoid naming names in what follows, not least because I see most people working in Whitehall as the product of a system and its incentives. However, it would be dishonest not to say that the complacency of some senior officials, past and present, makes me incandescent. That has provided some fuel for what follows.

Further grist to that mill comes from the constant drip of entirely predictable and repeatable patterns of state capacity deficit. 2026 is barely 7 weeks old and two more have come to the surface - the Civil Service Pensions debacle (Cabinet Office + Capita), the ‘full-spectrum disaster’ that is the National Savings and Investments ‘Rainbow’ programme (HMT + Atos). It is not just the dismal waste of money both of these blunders represent, or the deleterious (and sometimes tragic) impact they have on thousands of humans. It’s the fact the same bloody thing keeps happening, in exactly the same ways, again and again and again. I don’t know about you, but it makes me want to push quite large things over.

Finally, I want to acknowledge the insight and influence of the many people I have read, listened to, or worked alongside as I’ve picked away at this topic over many years. I won’t list them: this isn’t the Oscars, and I would unforgivably forget many who deserve mention. There’s also no reason they should feel guilty by association with what follows.

Whenever you write about civil service reform, there’s a good chance of falling foul of the apocryphal Samuel Johnson burn: “Your manuscript is both good and original; the part that is good is not original, and the part that is original is not good.” Many others can lay claim to any good bits that follow. Anything original is entirely my fault.

---

Part 1: The case for caring

"No profit grows where is no pleasure ta'en: In brief, sir, study what you most affect."

Shakespeare, The Taming of the Shrew (Act 1, Scene 1)

Most British voters no longer think that their elected politicians can turn promises into things they can see or feel. The country is at the point where many voters don’t think that their government is capable of changing things, even when they actually have. Net migration to the UK fell by more than two-thirds in the year ending June 2025, but 67% of people polled thought it had increased over that period. NHS waiting lists are at their lowest levels in 3 years. Productivity seems to be improving. There are hints that despite relatively anaemic GDP growth, the economic picture is improving.

But state capacity is not quite the same as a government’s ability to deliver. It is a quality where the substance and the style matter. Change has to happen, and be seen and heard to have happened.

The current popularity of Reform and the Greens rests on the fact they are untainted by association with a polity the majority of voters do not believe can change their lives for the better. I am sceptical that either commands a base of support that is fully signed up to their policy platform or personalities. But both tap into the expanding group of people who simply feel a deep contempt towards what is familiar. Labour’s election victory in 2024 was won largely on the strength of just that message: ‘we’re not the other lot’. Now they too are finding themselves beyond the pale. They’re all the same, and “they’re all just looking after one another.”

As well as making voters ever more fed up, I would argue that poor state capacity is leading to many countries, the UK included, failing to make the most of their potential - in terms of economic growth, high-quality public services, infrastructure and defence.

It is also eroding public faith in democracy. For millions of voters, British democracy has become akin to following a mediocre football team. The loyalty is deep; there remains a scintilla of belief that greatness is possible. But hope is overwhelmed by the reality of regular defeats, drab draws, increasing ticket prices and not feeling listened to by the owners. Football is one of the most highly scrutinised and accountable management professions there is. Outcomes are all that counts. Bosses who don’t deliver them get fired. Blood letting as a regular release for frustration is one thing that keeps fans coming back to their clubs. Politics is following an identical path. Britain has had six Prime Ministers in the last ten years. Making it seven looks a distinct possibility. The Premier League’s average is five.

Many countries that the UK finds itself unfavourably compared to in anecdotal perceptions of state capacity are either qualified democracies or authoritarian states. China arguably leads the world in state capacity. Its success also shows why prizing ever higher state capacity may not be a great idea. China’s incredible expansion of sophisticated public infrastructure: housing, transport, energy and so on - is progress that many democratic nations legitimately envy. I’m far less sure everyone would wish their government to have the social coercion of Zero Covid lockdowns, say, as an option at its disposal. The Goldilocks point of what ‘just enough’ state capacity amounts to is a good debate to have. What doesn’t seem to be in dispute is that most democracies currently fall some way short, rather than overshoot.

Democracy becoming associated with sludge, delay and not getting things done is dangerous. Especially if authoritarian states become perceived as models of muscular progress, regardless of whether or not they actually deliver on their hype. History rarely repeats itself, but it sometimes rhymes. Even today Mussolini remains famous for the idea that ‘he made the trains run on time’,² rather than his grisly end - hanged upside-down outside a Milan petrol station.

Among those who want democracies to be more competitive, state capacity is having a moment. It has become ‘the issue of the age’ according to The Economist. Having spent more than 10 years going on about this, the recent attention feels to me like the world finally embracing the niche Belgian turbo-folk band I’ve been enthusing about for years.

Part of the trouble with going a crusade for state capacity to restore public trust is the fact it feels like such an insidery topic. Most of the arguments made for change in government are built on technocratic foundations. But the ones that really worked found a way of tapping into the id, of making people feel belief, urgency and desire.

Looking back to the ‘Good Government’ US progressive movement of 1920s and beyond, ‘efficiency’ has been the watch word for wannabe reformers of government. But efficiency in those days was a rallying cry for two linked ideas. One was stamping out endemic corruption. The other was ushering in the then new principles of scientific management enlivened by Henry Ford, Taylorism, and the nascent world of management consultancy that was beginning to emerge from the newly minted Ivy League business schools.

An efficient state that delivers value for money is both a desirable outcome and good indicator of state capacity. But it is also bloodless³. It matters, but who cares, deeply? Efficiency alone is not much a banner to rally around, for voters or ministers. Politicians don’t win elections because they saved a lot of money. Saving money so that those funds and energies are redirected into things we do care about, sure. But nobody wants to talk about those vegetables you have to eat before the dessert. Beating corruption, on the other hand - that’s enlivening on its own. Kicking out the bastards waxing fat on your tax money, hey, everyone can get exercised about that.

As the spectre of overt corruption was slain (or at least diminished as a political rallying cry), all that remained of ‘transformation’ was the logic of rigour, data, and cashable savings. These became the touchstone of technocratic change efforts from about the 1980s onwards, with management consultancies often ministering to the figures and charts. It is interesting that in the past few years, the British right, taking its cues from Trump, went hard on the story of corrupt (or at least, extremely wasteful) officials as the rousing core of their arguments for slashing the state. Far more than the centre or left, they locked into a story about reforming the state that taps into the animal brain in all of us.

But what if you think, as I do, that the corruption story isn’t the real and substantive problem? What are the alternative narratives that elevate discussion about state capacity into something that has political currency and verve?

To somebody not bothered about politics, or management theory, or bureaucracy, the reasons to sort out state capacity are much more direct. When public services are poor, people get frustrated. You swear under your breath when you can’t book a driving test or you’re stuck on the phone to HMRC. When public infrastructure is lumpen or degraded, we get depressed by the ugliness, the potholes, the peeling paint, the infantilising signs. When the extremely comfortable get away without paying their fair share of tax, it pisses people off. And as a dozen or more drips of disappointment gather, they coalesce into one of the most deep-rooted of British emotions - embarrassment. The fact nothing seems to work is embarrassing. It offends our sense of national pride, even though it gives us the opportunity to indulge in Britain’s favourite pastime: moaning. And genuinely terrible scandals of state capacity and competence, like Infected Blood, Hillsborough and the Post Office inflict further hammer blows.

One strong political argument for improving state capacity is one that is taps into national pride. Pride is a powerful emotion. Sometimes it curdles into arrogance and sin, and sometimes it is misplaced. But a case for improving state capacity that fails to authentically tap into it is doomed to be either dry or dangerous.

A state with good capacity should be capable of running public services that we appreciate when we need them (like receiving a new passport in under a week), of orchestrating innovation when the scenario demands some (like a Covid vaccine), of building public infrastructure that unlocks opportunity and lifts the heart (like the Kings Cross redevelopment), and putting on a show that impresses the world now and again (like the London Olympics). These are things to be proud of, that ever-so-slightly quicken the blood, that transcend the urge to moan. We should have a state that can do them more often than not. We instead have one that can only deliver such moments occasionally, in a crisis, despite itself.

The other argument for state capacity is to maintain and enhance sovereignty; or more plainly, a feeling of control. As geopolitics becomes increasingly febrile, more careful thought is going into how national governments maintain their ability to exercise power within their domain. The Brexit campaign got ahead of that nagging sense of loss, playing effectively on voters’ sense of decisions being made elsewhere. It crisply tied the loss of control and agency people many felt in an increasingly globalised world directly to a supranational entity who really did have some power to bind Britain’s hands in ways she may not have independently chosen. If you voted for Brexit because you were persuaded by the repatriation of sovereign power - the ability of government to ‘take back control’ - it must be frustrating to find yourself with a state that still does not appear to be in control.

In a roundabout way, this brings us back to pride. Poorer nations often find themselves substantially dependent on money from institutions like development banks. Because loans come with conditions - Country X will only get the $$$ if it makes reforms x, y and z - the parameters in which the beneficiary government can make policy choices is constrained. The sovereign act of policy making is effectively outsourced to donors. Richer nations like Britain would like to think of themselves as being above such overt forms of sovereignty loss (hence why Denis Healey ‘going cap in hand to the IMF’ 50 years ago still looms large in Westminster’s consciousness). But poor state capacity trammels UK sovereignty in a similar way. In recent years, this has become particularly visible when it comes to technology. For example, such is the tangle of IT spaghetti at the heart of the UK tax system that the government can’t make some changes to corporation tax rules because nobody knows exactly how to change the systems that would enable it.

It is not a coincidence that the fracture point tends to be at the intersection of policy and technology. Within government, these two worlds are still often oil and water. Policy has long been the discipline that is first among equals in the civil service. Technology has long been considered part of the plumbing, below the salt. Yet over the last ten years, and accelerating ever faster, the relative power of these two disciplines around the world has flipped. Our lives and our nations revolve around technology now. So does much of global capital - physical and human. Much of the policy profession has still barely absorbed this, or is in complete denial. After all, it’s just boring IT, right? Right - except the ever-changing way in which human beings interact with tech more profoundly shapes the impact a government can have on a society than any policy paper. I’m reasonably confident that the most influential ‘thinkers’ on governments won’t be think tankers writing papers; it’ll be people mucking about with AI agents who show how easy it may soon be to irrevocably break the logic of existing public services. By spending too long looking the wrong way, many of the most senior people in government have not even spotted their strength and control slipping elsewhere.

All to say that the technocratic arguments for higher state capacity - better value for money, better services, delivering faster, and so on, are of course still important. But they are not enough. The emotional and political arguments - pride, strength - must be made more powerfully, and now.

Part 2: Northcote Trevelyan II

"He who desires or attempts to reform the government of a state... must at least retain the semblance of the old forms; so that it may seem to the people that there has been no change in the institutions, even though in fact they are entirely different."

Niccolò Machiavelli, Discourses on Livy

A few weeks back, Chief Secretary to the Prime Minister Darren Jones gave a speech entitled ‘Move Fast. Fix Things.’ This set out the government’s views on ‘rewiring Whitehall and building the new digital state’, building on a similar speech made by Pat McFadden (Jones’ predecessor) in December 2024.

When you talk to normal people about the speech - people who have neither heard of and couldn’t care less about the Green Book, say - there are two quotes that stick:

“…only 7 in nearly 7,000 senior civil servants were reported to be on a development plan for underperformance last year. And only 2 of them were dismissed due to poor performance.”

“The original plans had 40 different approval processes, each one needing a time to write a paper, submit, review, feedback on it and amend. This new scheme cut that from 40 to two.”

These are the points where people who’ve had a job before, in almost any organisation, say: ‘but, wait…what? Really? Seven hundred, you say? SEVEN? But that can’t be right.’ They join a club of similar facts from the past.

“A recent review of the Department for Business...recommended that its layers of management be cut to seven. I am told that Tesco has five layers from CEO to checkout operator.”

“Civil servants are ten times more likely to die in post than to be dismissed for poor performance.”

The “Rayner Scrutinies” identified £170 million in annual savings (roughly £750 million today) simply by eliminating paper-shuffling tasks that “served no identifiable policy goal.”

“(In 2004) only 25% of Departmental Finance Directors are professionally qualified accountants.”

Before the minister set out some of the short-term actions the government is taking - on incentives for senior officials, establishing ‘taskforce’ teams, a new training school for government - there was one line that piqued my interest.

‘It may require more fundamental restructuring of Whitehall in the future.’

I think this is true. And by restructuring, I don’t mean rearranging departmental deckchairs (though new institutional shapes and accountability frameworks will play a role). I don’t mean getting rid of a few senior officials (though changing people will play a role, and I personally would favour a Thanos finger snap⁴ for the very top). I mean a fundamental refresh of the character and organisation of the civil service. And to do that kind of system change, I think a government - this one or a future administration - will need to set up something of similar scope, influence and credibility to the Northcote-Trevelyan report of 1854, the foundation stone of modern bureaucracy.

Northcote-Trevelyan was piqued by embarrassment over gross administrative failure. It turned British bureaucracy from a world of placemen and patronage to one of meritocracy. It created connective tissue to unify a fragmented set of institutions, inspired the creation of brand new institutions, and distinguished between ‘intellectual’ and ‘mechanical’ labour.

The content of Northcote-Trevelyan is one thing. The context is also important. The two principals, Stafford Northcote and Charles Trevelyan, were not minor functionaries. Northcote held many of the biggest political offices in Victorian Britain over three decades in politics⁵. Trevelyan was a tireless official who was the Treasury’s Permanent Secretary (or the equivalent title for the time) for 19 years⁶. Their report was a combined insider-outsider effort, official and political. They were serious, senior figures.

It is worth noting that for all it is lauded now, the immediate reaction to their sweeping proposals for reform was almost universal horror. What followed publication was literally years of vociferous moaning and active non-compliance. William Hayter, ‘Patronage Secretary’ in the Treasury - Trevelyan’s own department - was alleged to keep two halfwits on his payroll (known as ‘Hayter’s idiots’) whose only function was to compete for promotion with Hayter’s desired candidate, thus ensuring that the favoured man always secured the post. Even Queen Victoria expressed her concerns. It took the better part of two decades before the report’s reforms became reasonably bedded in.

I expect any attempt at a ‘fundamental restructuring’ of the civil service today would find these things to still be true. To get off the ground, it has to be done as a joint endeavour between elected politicians and permanent officials (though it can be achieved by a motivated subset of each). And if and when it happens, almost everyone will hate it.

However, I don’t think everything has remained as it was in 1854. Northcote-Trevelyan was a very Victorian endeavour. It aimed to rebuild the bureaucracy so that government was fit and able to discharge its responsibilities in an industrial, mechanical age. Those responsibilities were several and complicated. But they were, by and large, known challenges - a foreseeable trajectory. That is not true anymore.

Today’s world is an almost infinitely complex and uncertain one. Many of the occasions in which government flounders today - setting aside the comparatively rare instances of bad luck, gross incompetence or poor behaviour - are those where it acts as if the world is still predictable, mechanical, controllable. That is certainly the mindset that tends to be applied to delivery of major projects like HS2, or the Civil Service Pensions/NS&I disasters referenced earlier. But that has not been our world for a while. It is becoming ever less so. The fabric and the character of bureaucracy has not kept up. Northcote-Trevelyan moved the civil service from patronage to meritocracy. Today, we need a shift from permanence to responsiveness.

---

Why, though, do we need a Northcote-Trevelyan style intervention now?

First, nobody from any point on the political spectrum seems to be arguing that what we have is working well enough. Moreover, as I’ve said, it is falling short in consistent, well-observed patterns. Credible critiques of the British civil service go back more than a century. They invariably go over the same ground. Plenty of administrations have wrestled with these problems by increments, through which some have enjoyed moderate tactical success. But none can really lay claim to a quantum leap, a paradigm shift, a real sense of something different and better. If the government’s best answer to ‘why will it be different this time?’ is ‘because this time we’re leading it’, don’t believe them.

Second, over the last 15 years I have spoken to hundreds of officials, who are desperate for things to change⁷. At all levels, in all departments. Not all of them would agree on the nature, tone and speed of that change, obviously. But I can think of increasingly few who would say, in private, that ‘everything is fine’. The tiny handful I’ve spoken to that would say this fall into one of two camps: they are extremely senior born survivors, or they work for the Treasury⁸.

Third, Britain has a chance take to lead and gain a global advantage, and tell a compelling story to boot. I don’t subscribe to civil service malaise forming part of a ‘Britain is Broken’ story. Rather, I think improving state capacity is an enormous opportunity in which the country can excel, building health and growth and pride and belief. Other democracies around the world are waking up to declining state capacity as a bottleneck on both their national sovereignty and economic vitality. They will make changes eventually too. Northcote-Trevelyan predated the US progressive movement (which pushed for similarly meritocratic aims in official positions) by several decades. Let’s get out ahead of everyone again, and take first mover advantage.

All this is perhaps just a case for reform in the service of state capacity. What about the case specifically for a Northcote-Trevelyan review: which to clarify, I am using as a shorthand for a cross-party Royal Commission, set up on statutory basis, that is focused principally on the character and selection of people who populate government and how to arrange them.

First, state capacity is a national interest, not a partisan one. Civil service reform is one of the few topics where the Labour, Conservative and Reform parties can claim to share some common cause⁹. A government who seizes the opportunity can claim the high ground credit for their statesmanship; offering a grand bargain to those who are up for constructive radicalism, while isolating those who have nothing to offer beyond conspiratorial complaint. There is plenty of room between the parties for disagreement on emphasis and political posture. But by definition, a Commission’s proposals should work for the government of the day, irrespective of political stripe. Conversely, making this a partisan issue almost guarantees no progress will be made. Rational status quo officials are well practised in running down the clock while waiting for the next set of political masters to arrive.

Second, and linked, we clearly can’t count on Victorian levels of (relative) political and official stability. To have any chance of working, improving state capacity has to be an endeavour that can survive political and personnel change; in leadership and/or of a government. A Commission would need to start with self-dissolution powers, put into effect only when it is confident that its reforms have been carried out satisfactorily or rejected by the government. A vehicle for reform that is not placed on a firm footing is exceptionally vulnerable to shifting political sands.

Third, this is an Establishment challenge. My hunch - and it is no more than that - is that you can only tackle Establishment challenges with Establishment tools. Trying to do this in a new or novel way, something more in keeping with what passes for sensible and appropriate for 2026 is clearly the right thing to do. But Britain is Britain. I would be more confident in cloaking radical change in reassuringly crusty clothes, than do something too shiny that hands opponents a straightforward opportunity to patronise and ignore it. It has to be serious and it has to be proper. This carries obvious risks. ‘Royal Commission’ can be a synonym for long grass. To work, it needs political impetus to push on boundaries and pace - public hearings, working in sprints, perhaps testing interim recommendations for real in small areas, and so on.

---

So much for the case in favour. Why would a government not do this? There must be plenty of good reasons, because it is now almost 60 years since the last attempt at something close to Northcote-Trevelyan¹⁰. I’ll try and make a fair case against doing it, and respond to each point.

The natural first objection, particularly for this government, is to say ‘why must we have another bloody review?’. Reviews take time and they are the antithesis of action. Worse, they distract from delivery. The government is already coming under fire for commissioning lots of reviews. Can’t we just get on with it? I have plenty of sympathy with this, and it’s why setting up new institutions now is an essential complement to this (see Chapter 3). But if you are genuinely serious about ‘rewiring’ or restructuring the state, there isn’t a shortcut. We’ve had a century of shortcuts, and here we are. To build a coalition and the credibility to make a leap takes time. By all means make incremental changes in parallel; in fact, doing so is essential. But with no set piece, you have incrementalism. If you believe that’s insufficient, I think you need the set piece. It’s also worth saying that the civil service has been distracted from delivery by incremental visions of change and reform for almost every single moment of the last few decades. If distraction is already the default, why not do it properly?

The second objection is to argue that today’s world moves too fast for this type of exercise. Given the speed of technological and geopolitical developments, one could argue any set piece review is doomed to produce an answer that is immediately obsolescent. How can any review offer a direction for bureaucratic shape and character a century hence? The answer to this is to flip the argument. It’s not trying to predict what 22nd Century bureaucracy will do or be. The whole point of a sweeping review today should be to reform the civil service so that it is far more effective at responding to an uncertain, rapidly changing world, whatever it throws at us. Incrementalism is what continually leads us into trying to build a modern state on top of 19th Century frameworks.

The third objection is that this will cost a lot of money. Making people redundant costs money, so do new systems, new training, new hires, and so on. Surely transformation is about doing more with less? There are a couple of counters to this. One is to point out the questionable record various flavours of previous ‘transformation’ have actually had in delivering better public services and infrastructure for less money. The other is the point that Britain can’t afford not to do this. Low state capacity creates financial failures in all sorts of places: demand failure (extra demand on the state created by poorly designed/run services), market failure (poor value for money resulting from inefficient public procurement and/or oligopolistic supplier behaviour) and management failure (expensive major programmes that comprehensively fail due to magical thinking and upfront false certainty). Reduce that, and you save money. Not to mention the equally material but qualitative improvements to public services and public life.

The fourth objection is that effecting any substantive constitutional changes recommended by such a Commission would most likely end up requiring legislation. That would mean spending precious parliamentary time and political capital on something that won’t touch the sides with voters. This is a reasonable argument, and perhaps one of the reasons why so many governments have failed to make real headway on the issue. It also points to how important having a better political case for state capacity is (see Chapter 1). But the lack of a Civil Service Act to provide for a clear, statutory definition of relationships between ministers and senior officials is arguably a significant contributor to the breakdown in relations between those two worlds. Ambiguity has bred accusations of ‘politicisation’, briefings, frictions and finger-pointing. Voters aren’t much impressed by that either.

The last, and perhaps most telling objection, is that there does not appear to be a visible champion for such a move from within the civil service. Let’s return briefly to where I started: Darren Jones’ speech. Something was conspicuous by its absence in a major ministerial announcement about reforming the civil service - civil servants. In terms of style and substance, it looked strange that there was no senior official to present and publicly bolster the government’s case. There’s little point me speculating as to why that was. There are many possible reasons. But one implication is that however much Darren Jones may want to be the Stafford Northcote of his generation, a Charles Trevelyan figure is not yet apparent. Does one exist somewhere in the civil service today? I hope they do. Are they in a position equivalent to that of Charles Trevelyan? I don’t know. But the personnel surely matter.

Establishing a Royal Commission on the future of the Civil Service is an unambiguous signal that a government believes that deep systemic change is needed to unlock greater state capacity, and that it is willing to spend political capital to prioritise it. We haven’t had an administration do that since the invention of the computer mouse and office cubicle. It’s overdue.

A review won’t in itself solve all the problems. And just doing a review on its own will have relatively little to show for some time - so setting up new teams in parallel to demonstrate tangible progress is politically essential. But if such a review can serve as a marker, a signal and a signpost that now is the time to make a leap, not take a step, then I think both the civil service and the country could emerge from it - after a lot of good, old British moaning and pain - in a prouder and more optimistic place. It’s worth a go, no?

Part 3: The Department of Now

“There is a tide in the affairs of men
Which, taken at the flood, leads on to fortune;
Omitted, all the voyage of their life
Is bound in shallows and in miseries.
On such a full sea are we now afloat;
And we must take the current when it serves,
Or lose our ventures.”

William Shakespeare, Julius Caesar (Act 4, Scene 3)

One of the other messages in Darren Jones’ civil service reform speech was the announcement that the UK government will apply a ‘taskforce’ teaming model to other challenges, inspired in part by the Vaccines Taskforce led by Kate Bingham.

“I’m announcing that we will apply the Vaccine Taskforce model in “peace time”, not just in a crisis. These will be focussed on Prime Ministerial priorities.

And they will be tasked with bulldozing delivery obstacles, not spending weeks trying to answer policy questions.”

The speech set out four freedoms these teams would be given: freedom to hire, freedom to procure, freedom to spend and freedom to take more risks - each created by being given a direct line to ministers conferring the political cover to do so.

There are sniffs and whispers of other new team shapes sprouting up in Whitehall. HMRC’s Path team, which is “accelerating how we design, build and run the technology, platforms, and digital services the UK relies on.” DSIT’s CustomerFirst team is tasked with “transform(ing) customer services across government.”

All of these teams appear to be pretty much brand new; certainly insofar as they have broken cover publicly. Will they make any difference? Let’s come back to that. But the fact of their existence hints at something that has been strongly suspected for ages.

Our current institutions - departments, ALBs - very often can’t do things.

Again, we have been here before. During the mid-Victorian period, the high-point of radical bureaucratic reform in Britain, lots of new government institutions and shapes were created. It was recognised the existing set of public bodies lacked the capacity and competence to address the day’s pressing issues: sanitation, education and housing (some things never change). The Metropolitan Board of Works (1855), School Boards (1848 onwards), public library authorities (1850) and various local government and housing boards (1858 onwards) came into being.

Spin forward to today, and the ‘newco’ or ‘taskforce’ idea is predicated on logic that several ministers and officials appear to have landed upon:

1. My department isn’t delivering well or fast enough, and hasn’t for some time.

2. It cannot deliver its commitments without fundamentally changing how it works.

3. Despite decades of ‘change programmes’ that have attempted to redesign the organisation on paper, it has not meaningfully changed itself.

4. Despite several different leaders, it has not meaningfully changed itself.

5. When government sets up something new that’s semi-detached from the usual institutions, blending outsiders and insiders, it seems to get more done.

6. Let’s try that, and leave the old department to rumble on.

This is an important shift, because it is a step forward from a tabloid analysis which holds the view that: ‘What can you expect from this shower of useless ministers/this blob of recalcitrant officials? What we need is new people (who coincidentally have broadly the same strengths and attitudes as me).’ I would suggest that even if the country were lucky enough to be blessed with a golden generation of political figures and peerless bureaucrats, I’m 99% certain the Home Office would still be a mess.¹¹

Like most governments, this administration has taken some cues from the private sector. Less commonly, it appears to have taken inspiration not from the corporate giants that government has historically tended to view as its peers, but from relatively successful scale-ups who are making some headway in challenging well-set incumbency. Think Octopus and Kraken in energy, or Monzo in banking¹².

Personally, I’m a fan of this. When I was a civil servant just over a decade ago, I could never understand the logic of senior civil service secondments to the private sector. They happened very rarely for one thing, but when they did, almost always resulted in a high-flyer going to work at a bank for a bit¹³. Big banks, by and large, are among the few organisations even more gummed-up and byzantine than government. Voyages of innovative discovery they are not. So the government paying attention to some firms who grew up on the internet and are in the process of trying to disrupt regulated industries with complexities analogous to those of the state - that seems like a good thing to me. Does that mean government departments should be run just like Octopus? No. Will there be plenty to learn? Yes, probably.

However, one interesting aspect of the ‘newco’ approach generally taken by these parts of the private sector is not just the fact of creating a new team with more tech savvy and more responsive ways of working. It’s that newcos as conceived of by Kraken or similar are a) created outside the legacy ‘host’ organisation, so they can’t be smothered at birth by its sludgy stakeholder management and processes, and b) are usually intended to render the legacy host increasingly redundant, before ultimately killing it off altogether¹⁴. They aren’t a souped-up tech unit, or a snazzy digital innovation team, or a nice-to-have. They are the new organisation. Build a new thing, get it running, steadily transfer the good bits over from the old thing into the new thing, turn off the old thing. The King is dead, long live the King, etc.

Can a government do this to, say, DWP? There’s no reason why not, in theory. Will the government do this? That’s a different, harder question.

Governments generally set up taskforce teams when they have run out of hope. Existing institutions have shown themselves as inadequate for a task in hand; there is no choice but roll the dice.

That inadequacy tends to show up in a couple of ways. The first is as a hot, sharp pain, in the midst of crisis. Somebody - usually but not always in No.10 - has realised ‘Oh my God, it’s all completely fucked, and the responsible Department of X is miles off the pace. We need to scramble.’ The Vaccine Taskforce is a fine example of that. These teams come together to address point-in-time crises. They came, they saw, they turned a thing around, and then they largely disappeared. In these cases, the team’s nucleus exists for maybe a year, 18 months, before inexorably seeping back into their host organisation or leaving government entirely.

The second form of taskforce is a more meta construction, sitting in the centre, dealing with the dull, achey, constant pain of departments not getting their shit together. In this group you might include the Government Digital Service, the Prime Minister’s Delivery Unit, the Social Exclusion Unit, the Central Policy Review Staff, and so on. Keir Starmer’s Mission Delivery Unit was presumably intended to be something like this. It has not managed to achieve escape velocity, in large part because it was built from existing civil servants following existing civil service logic. More successful versions of this trope lasted a while, enjoying an influential half-life of maybe 5 years or so, before eventually being diluted. After this point, their artefacts and livery may endure, but only in Alice-in-Wonderland caricature. Perversely, their signature moves often become performative to the point of being counter-productive (hello service standard assessments, hello PMDU slide-deck format).

Both kinds of taskforce teams have done exceptional things in the last few decades; truly the best of the state at times. But what unites all of these past attempts at taskforce theory in government is that they ultimately lost. Again, some of the best delivered world-class outcomes when they were at their peak. They got copied around the world. They raised the game of government; sometimes at a specific national pinch-point, sometimes across the waterfront. But they did not endure. The closest example I can think of a taskforce team that almost made it from insurgency to establishment was the Office of Climate Change becoming the Department of Climate Change in 2008. They clung on for nearly 8 years before slipping beneath the waves.

What endured? The Home Office, the Cabinet Office, and many other familiar acronyms. Each bound in shallows and in miseries, but still definitively there. This is why I think the only way to sustain an increase in state capacity is to new up new institutions AND fundamentally reset the character of the bureaucracy via a set-piece review, as argued in the previous chapter¹⁵.

---

Let’s imagine that structural reform via a Northcote Trevelyan style set-piece of Chapter 2 is underway. Where might you need some taskforces to show what the new could look like? Based on where the bureaucracy consistently trips itself up, there are 3 clues that a new institution might be needed:

1) The desired political outcome falls between the cracks of the existing departmental structure. It may be several departments’ third or fourth priority, but nobody’s top priority. So it inevitably falls down the priority list, or becomes trapped in the realm of endless cross-departmental governance, Cabinet writerounds and nonsense. The Office of Climate Change¹⁶ was created in 2006 for this reason; that then led to DECC in 2008.

2) The policy and delivery of a desired political outcome is split between different organisations. I won’t revisit the issue of why breaking the feedback loop between policy and implementation is problematic; the Radical How paper gives full chapter and verse on that. Nonetheless, this split appears all over the place, and it’s makes progress on everything from farming subsidies to road pricing either impossible or much harder than it needs to be.

3) More nebulously; if a single institution is responsible for managing a very wide risk portfolio, then it is likely to become entirely captured by patterns designed to look after the high risk issues, and to neglect the low risk issues - even if the rewards are disproportionately higher. Most government departments suffer from some version of this very wide risk portfolio; the Home Office is a relatively extreme example.

Running a taskforce in government is a role with unique difficulties and freedoms. More often than not, their leaders (and some team members) are outsiders to government. Quite apart from the different skills and perspectives that someone with time working outside of Whitehall may bring, it possibly helps them to not know what odds are stacked against their success. Complementing such people with what my colleague Tom Loosemore would call ‘trained killers’, and what James C. Scott called those with ‘metis’¹⁷ - cunning, experienced and exceptional in-house bureaucratic hackers - is where the magic tends to happen.

To succeed, the whole team needs both the attitude and political backing to not wait for the rest of government to invite them to make the changes needed - that invite is not the post. You can’t wait for everyone to understand the practicalities and trade-offs, you can’t wait for everyone to read and comment on multiple drafts of papers outlining your plans, you can’t wait for everyone to agree. Some would argue this would lead to chaos and incoherence; I would invite these people to defend the coherence generated by the status quo. (I would also prefer chaos that produces some positive outcomes over coherence delivering soggy nothings.) Not always agreeing doesn’t necessarily mean being disagreeable by the way. Working collaboratively, honestly, openly with your peers, carrying a nuanced approach to risk and certainty - that all still counts for a lot. But crucially, the incentives for those successfully leading or participating in taskforces cannot be shaped by thinking about their career prospects in a post-taskforce world. They are there to do jobs that those focused on following the internal rules of the game currently cannot do, for want of storing up trouble or resentment that they believe will hurt them later.

The word permanent comes up so often in the Civil Service that it is easy to forget it. Permanence underpins so much of the behaviour; the conflict avoidance, the lack of urgency, the rule-boundedness. The UK civil service has forever presumed that its role is to offer stability and a long-term perspective, often as intentional contrast to the flighty, ever-changing faces in ministerial positions. Whether it actually provides these qualities is debatable¹⁸, but the cultural norm is deeply ingrained and staunchly defended. The house view is that politicians can’t be trusted, as individuals or as a tribe, not to be distracted by the short-term and surface-level. This received wisdom is not always inaccurate. And lots of what the state does is too big and too important to spectacularly fail. Managing risk does matter. Projecting permanence is important.

Many senior officials believe in their hearts that radical transformation is a non-starter because the edifice they are accountable for - a delicate, rickety thing which cannot, under any circumstances, be allowed to collapse - rests upon a pile of computer systems, forms and arcane processes that are decades old, and broadly work (or at least, are broken to what is deemed a politically acceptable level). Precious few understand the systems. The view is if it ain’t too broke, don’t fix it. We can squeeze out a few % efficiency gains here and there. Maybe we can write cleverer policy, somehow.

But there are two problems with this. The first is that bureaucracies bound their present and future selves in an ever-more twisted bundle of ties - both legislative and organisational. Governments accrete and calcify, with the best of intentions throughout¹⁹. This occurs not just in the laws they pass, but in the bodies and organs of state themselves. The internal processes and procedures bureaucrats impose on themselves. The iron that enters the soul of lifelong senior officials who’ve never worked elsewhere. Even the offices and buildings they occupy. They end up standing tall, mighty but dead, like petrified trees.

The second problem is the assumption that bureaucracies have to pick only one mode: stability or renewal. Only in particularly exacting circumstances - wars, crises, pandemics, high-stakes-absolutely-no-choice-but-to-nail-it events - do government bodies open themselves to a Jeffersonian streak of renewal - and only when it is obvious that the status quo can deliver nothing but doom. Enter the taskforce to do battle for a fraught few years. But after that, what happens is the desire for stability sees everything drifts back to the way things once were. Processes slip back into the same well-cut grooves of rules and norms. The external hires and ‘specials’ who were drawn into government by the call of ‘your country needs you’ are given little reason not to drift off (this happened after WWII, and it happened after Covid too).

The idea that the bureaucracy can do both stability and renewal, simultaneously and all the time, in different parts of the forest which have different risk/reward calculations, seems not to stick. After a radically different recipe delivers in the midst of crisis, the cry goes up: ‘why can’t we do this all the time?’ Or another, more recent, variant: ‘why can’t the state behave like a start-up?’ And the answer is: ‘Government doesn’t need to do this all the time, and it isn’t a start-up. It does need to provide stability and solidity, to manage money carefully, to not wilfully rattle markets or move goalposts’. The need to combine renewal and stability in government is eternal, but bureaucratic culture always cleaves towards stability and certainty regardless of whether that stability is good. Those who are good at renewal do not stick around in such an environment. Why would they?

It is inadequate for progressives or incumbents (political and official) to say ‘it would all be fine if we had more money’. The UK government currently has £200 billion earmarked for service transformation programmes in the major projects portfolio. There is money. But attempting to transform the impact of the government within the same organisational boundaries, using the same people, culture and processes, isn’t working. At best, it achieves incremental improvements against an old set of outcomes. At worst, it will be ever more expensive, complexifying, pointless and doomed.

The UK is not alone in this. Where we had the NHS’ £11 billion National Programme for IT, Canada had Phoenix, Australia Robodebt, the US Healthcare.gov. Transformation, as it is currently scoped, procured, project managed and staffed by democratic governments around the world usually does not work. A lot of money rests on that conclusion not being drawn, however, as well as an eye-wateringly high bill of sunk cost fallacy on money that has already been spent.

AI will inevitably be folding in the next wave of trying to perpetuate this magical thinking, and to some extent already is. This is a shame, because it obscures the benefits that most likely will be there. Matt Clifford has said that ‘no public organisation has an AI shaped hole, but most have an AI shaped hole in their vision’. All too many consultancies and in-house strategy teams will be queuing up to ignore this, squashing AI into holes that aren’t really there. No-one will be commissioned or allowed to think deeply about the vision, because the conclusion will most likely mean calling into question the shape of the institution itself. Left to its own devices, the civil service is not going to ask ‘how can AI unlock higher employability, improved mental health outcomes and a lower benefits bill?’, it’s going to ask ‘how can AI be applied to DWP?’ This is the wrong question.

This doesn’t mean AI won’t have some success within the current organisational construct; it surely will, particularly in extracting significant efficiency gains from processes that basically function OK in delivering existing outcomes (e.g. implementing automated note taking for probation officers). But all that gets you is, at best, a slightly better Ministry of Justice.²⁰ Not a better country.

Without systemic change, new institutions are rarely more than sticking plasters. New technologies too. Sometimes briefly effective sticking plasters, but sticking plasters nonetheless.

With systemic change, they could become the first building blocks of a quite different institutional landscape, one that eventually ends up supplanting and consuming organisations that are no longer fit for purpose.

---

There’s so much more to say. I haven’t talked enough about accountability or AI, for a start. But this is already far too long.

Amidst all this windy prose, all I’m really proposing here is a strategy that combines three different pace layers for reform in the service of state capacity. A story that works in months, a delivery mechanism for generating political and public results that works in years, and structural changes that works in decades to make better delivery mechanisms the default. Each ultimately depends on the other for lasting success. Taken together, they could change the country, visibly and for the better.

Of course, this is just a recipe. What will make it happen and stick (or not) is people. People with bravery, humanity, charisma, cunning, a certain ruthlessness, self-belief, a tolerance for unpopularity, and perhaps most importantly, a desire to deliver this kind of radical change for its own ends rather than as a convenient vehicle for their personal or political advancement. Such people are rare, but ever naive, I’m confident they exist.

When it comes to dramatically improving the UK’s state capacity in this moment, I don’t rate the chances of success particularly highly. I’m a cynical soul, and this is very difficult to do. There’s a reason we haven’t managed it in a very long while. We know that low state capacity is not a problem that will solve itself. Nor is it a cause that will excite lobby journalists or parliamentary tea rooms. It does not fit nicely with a 24-hour news cycle, the otiose business of spin, the flummery of political knockabout, the playing of games, the bluffing and the gurning. It is fundamentally less fun and less easy to play than the usual Westminster and Whitehall games.

If we remain attached to a bureaucratic method and culture that is unable to respond, and in which power and faith continues to ebb away to unelected actors, I do not foresee an especially happy future for democratic public life in the UK. Pennies are still not dropping where they should. The Public Accounts Committee’s recent report on the NS&I debacle referred to earlier in this post certainly doesn’t think so, saying “despite the lack of progress, NS&I has not demonstrated to us that it understands and accepted what went wrong.”²¹

Despite all this, I remain a qualified optimist. There are many fine people still drawn to public life, and fire residing in the bellies of some to radically improve it. If - and it is a big if - we have enough people in politics and public service who both believe in and think hard about how to run the country more effectively, there is always hope. I am hopeful that it is becoming obvious to more people that the UK should at least try to improve state capacity by adopting a different posture to lacklustre incrementalism.

Our alternative is to row on, boats against the current, borne back ceaselessly into the past.

1. For those fans of exploring how university backgrounds shape people; Dom was an Oxford History graduate, a distinction he shared with the previous Cabinet Secretary, Sir Chris Wormald. Dame Antonia Romeo, apropos of nothing, did PPE at Oxford. So did I. A majority of Perm Secs have Oxbridge History or PPE degrees. I’m no fan of this domination.

2. Except he didn’t make the trains run on time, other than the lines used by credulous tourists.

3. I’m generally unconvinced that you find savings when you chase them as a primary goal. GDS saved a lot of money in its first four years; over £5 billion in today’s money. To his credit, Francis Maude accepted advice not to set a savings target for the digital team, and instead trusted the logic that if you make government work more effectively, the savings will follow. If you instead chase after savings, government will not work more effectively (and therefore often end up costing more that it did before you tried to make it ‘more efficient’).

4. This is a reference to Avengers: Infinity War, where a character snaps his fingers and half the population immediately disappears. One of my less acceptable opinions is that if you did similar to the current top 200 civil servants, even at random, and paid the remaining half 1/3 more, I strongly suspect you’d get better outcomes.

5. He also holds the unhappy distinction of being one of the few political figures to die in 10 Downing Street, having passed away while awaiting his demotion in a Cabinet reshuffle.

6. Trevelyan is better remembered - not fondly - in Ireland rather than the UK, thanks to his role in the Irish famine of 1845-52. His name features in the song ‘The Fields of Athenry’.

7. Others, like Amy Gandon and Anna Garrod at Demos, or Joe Hill at Re:State, have done much more rigorous and interesting temperature checking with officials than my anecdata.

8. This isn’t because those two groups are inherently enemies of change, though they are the most powerful actors within the status quo. But they are also the people (rightly) concerned about accountability, and the role of the Accounting Officer in particular. You can’t have reform without reforming the model of accountability and officials’ relationship to Parliament. This post by Gordon Guthrie touches on some aspects of this. The only true horizontal accountability in today's HMG is where/how to allocate money; everything else flows through vertical line departments. I can’t see how HMG cannot shift to more horizontal accountabilities, and that they'll be made of data, not paper. A big topic for another day.

9. Witness Reform UK’s new ‘Preparing for Government’ website, or recent remarks from Danny Kruger expressing his view that “the notion of the impartial civil service is an admirable one, but there's something essentially broken about the Northcote-Trevelyan model of a professional civil service operating independently of party politics.” Whether you agree or not - this is at least evidence of some active thinking about the issue.

10. This was the Fulton report of 1968. Fulton made some headway, but was seen as a cultural failure in terms of change, and stymied by not being allowed to look at ministerial-official relationships or machinery of government changes. It was also not a Royal Commission.

11. I’m not picking on the Home Office specifically - you could insert almost every department here. Though it must be said the Home Office does seem to receive a ‘not fit for purpose’ review once every 4-5 years.

12. Both Monzo and Octopus have their fingerprints on CustomerFirst.

13. Former Cabinet Secretary Sir Jeremy Heywood went to work for Morgan Stanley, for example.

14. A move close to the ‘strangler fig’ pattern.

15. And nicely articulated by Martha Dacombe.

16. The ‘Office of…’ construction was the mid 2000s term of art for ‘newcos’. There were several.

17. James Plunkett has written a good rundown on metis and why it matters.

18. The Civil Service’s lamentable institutional memory has been a semi-regular complaint of the National Audit Office and Public Accounts Committee down the years.

19. Jen Pahlka’s Cascade of Rigidity is the last word on this.

20. There’s also the plausible possibility that AI hyperscales the Jevons Paradox in public services; loading more demand on the service faster than any efficiency gains in delivery.

21. You can read their report here. It essentially portrays the NS&I’s evidence as the ‘This is Fine.’ meme in prose form, to the point where some of it is bleakly comedic.