Why Starmer can’t govern

Labour’s lack of preparation means it has failed to reform Britain’s amateur state

Image: Alexandros Michailidis/Shutterstock.com

Today on Arguably we have a guest column from Geoff Mulgan, the author, thinker and former No 10 director of policy under Tony Blair. This post is free to read but become a paid subscriber for just £6 a month to receive all of our pieces.

The civil service finds itself in an unenviable position: denounced by the Prime Minister and blamed for failings of policy and politics, its morale is at a low ebb and it faces the threat of mass redundancies from Nigel Farage’s Reform after the next election. So can it be fixed?

The answer, cautiously, is yes. A leaner, more capable, more agile state is achievable and the present crisis, if anything, provides the political opening that incremental reformers have lacked for years. But anyone expecting easy wins will be disappointed. Austerity hollowed out institutional memory. Brexit consumed years of administrative energy. And a decade of Conservative political instability left the UK sliding steadily down global rankings of government effectiveness. The damage is real and it runs deep.

That is not to let Whitehall entirely off the hook. The familiar pathologies endure: over-centralisation, a culture that prizes elegant prose over practical delivery, ingrained elitism, and a Treasury whose grip on departmental spending has too often substituted financial control for genuine strategic thinking.

Yet the picture is much less uniformly bleak than the current political mood suggests. Britain’s record on issues such as carbon reduction, counter-terrorism and school standards represents genuine achievement. And despite everything, 74 per cent of UK citizens report satisfaction with the services they actually use, a figure that comfortably exceeds the OECD average of 66 per cent. The public, it turns out, is rather more forgiving of the state than the politicians nominally in charge of it.

The strategic vacuum at the centre

The most pressing failure is a lack of direction. Keir Starmer’s No 10 and Cabinet Office are operating with less strategic capacity than at any point in living memory. Without a capable centre able to think strategically and work through the hard trade-offs – on trade policy, AI governance and defence procurement – governments drift, reacting to events rather than shaping them. This is where reform must begin.

One of the early priorities should be rationalisation. With the state now consuming almost 45 per cent of GDP and a civil service headcount that has risen by a third in a decade to 550,000, duplication and bureaucratic inertia are not right-wing talking points but observable realities. That the Conservatives and Reform so thoroughly captured the language of efficiency is no excuse for Labour’s failure to do anything systematic about waste and inefficiency. That needs to change.

Technology and the failure of trial and error

Technology can help, but on AI and digital transformation the government faces a genuine dilemma. The prize is considerable: smarter service delivery and better policy design. But realising it requires something Whitehall has consistently struggled to build: in-house technical capability that does not simply outsource dependency to Silicon Valley, and that is adept enough to avoid future AI variants of the Post Office/Horizon scandal.

More fundamentally, the British government remains trapped in a culture of “error and trial” – mistake, inquiry, sacking, repeat – when what is actually required is its inversion: systematic experimentation, honest evaluation, and the institutional confidence to learn from failure rather than bury it. Franklin D Roosevelt’s New Deal operated on precisely this logic. The current “Test and Learn” initiatives gesture in the right direction, but the scale of ambition remains embarrassingly modest given what is at stake. It’s odd that the UK spends tens of billions of public money on scientific innovation but so little on social and policy innovation to fix the problems people worry about most, such as adult social care or teenage mental health.

Structural anachronisms

The machinery of government itself is long due an overhaul. Nineteenth-century departmental silos were not designed for cross-cutting priorities such as economic growth or child poverty, and no amount of ministerial goodwill or committees compensates for structures that actively obstruct joined-up working. Formal partnerships with devolved administrations and local government are not optional extras but prerequisites for governing a country as complex as contemporary Britain and getting results on priorities such as housebuilding.

The Treasury’s approach to public spending deserves particular scrutiny. The distinction it draws between capital investment – subject to rigorous investment appraisal – and current spending on people, treated as a mere annual expense, is not just intellectually incoherent but actively harmful. It systematically undervalues prevention and stores up costs for the future. In an era when people outlast physical infrastructure, this is a category error with serious consequences.

The amateur state

Binding all of this together is a more fundamental problem of capability. The belated announcement of a new National School of Government in early 2026 was welcome, if thin on detail. Done properly, it could genuinely equip civil servants for a world defined by AI, data and complex systems. But the even bigger gap is political. Ministers arrive in office with no formal preparation whatsoever, and it shows: in the quality of decisions, in the management of departments, and in the recurring pattern of avoidable catastrophe.

Labour came to power without adequate preparation and then lurched towards angry briefings against the civil service, burning goodwill. The 2025 ban on officials speaking publicly was a particular misjudgement, closing off precisely the kind of engagement with civil society that a more confident, human and outward-facing state ought to be cultivating.

The inheritance, in the end, is characteristically British: a state composed of archaeological layers – Anglo-Saxon, Tudor, Victorian meritocratic, Fabian, Thatcherite, Blairite – that has never quite been rebuilt from first principles. In the age of AI, pandemics, trade wars and climate disruption, that is no longer a charming eccentricity. It is a liability. There’s now a rare opportunity to address it. But whether the political will exists is unclear.

Of course, if this chance is missed, things could get much worse. The US in 2026 gives us a hint of how a Farage government could bring an entirely new level of toxic over-politicisation and amateurism, a souped-up version of the chaos of the Johnson/Cummings era. That’s a future the UK can ill afford.