Last week’s Public Accounts Committee report on accountability in government bodies made a series of recommendations on disproportionate reporting requirements, shared services, and functional standards. The Cabinet Office’s response cited sixteen closures or mergers emphasising plans to “re-wire Whitehall and create a modern and productive state”. Yet what has been neglected so far is the fundamental question of the role of arm’s-length bodies in UK governance.
The current programme of rationalisation largely ignores that question. The bodies merged or absorbed under the latest review include NHS England into DHSC, the Valuation Office Agency into HMRC, the UK Space Agency into DSIT, and Building Digital UK into DSIT. But the programme as a whole appear to be a narrow exercise in financial efficiency. The strategic question – what capabilities the state should discharge at arm’s length and how to build them effectively, appears lost in the frenzy to slash costs.
Chairs and chief executives running arm’s-length bodies are carriers of institutional memory about delivery and operations that the centre of government has lost. The proliferation of independent regulators, commissioners, advisers and oversight bodies over the past three decades is not an accident of poor governance. It is a response to a series of legitimate choices: that some functions require independence from ministerial direction, that some require specialist expertise the civil service cannot easily sustain in-house, that some require continuity across political cycles, and that some require a buffer between ministers and decisions that need to be insulated from short-term political pressure. Arm’s-length bodies exist because the alternative, ministers and departments performing such functions directly, was tried and found wanting.
None of this is to argue against consolidation where it makes sense. The PAC is right that the long tail of smaller public bodies is being held to standards designed for organisations twenty times their size, Meanwhile, shared services are being imposed without regard to whether they suit smaller agile organisations, while sponsor department support is uneven. The chairs and chief executives we work with would broadly agree with that diagnosis. They would also point out, as they have done consistently to successive reviews, that the proportionality problem is not a function of organisational size. It is a function of standards designed in Whitehall, then applied without modification to a sector with very different operating realities. Merging small bodies upward into bigger units does not fix that. It just buries it inside a larger organisation.
The deeper issue is that the consolidation programme is treating arm’s-length governance as an efficiency problem when it should be viewed as a strategic question. The Fifth Industrial Revolution, which we have argued is the central context for state design in the 2020s, will require more arm’s-length capability rather than less. AI governance, biosecurity, advanced therapeutics, critical infrastructure resilience, and frontier risk regulation all require the combination of independence, specialist expertise, and political insulation that arm’s-length bodies are designed to provide. The AI Security Institute, which Anthropic recently trusted to conduct the sole independent assessment of its most powerful model, is a striking example of what a well-designed arm’s-length body can deliver. It is also, on the current trajectory, an exception rather than a blueprint.
A genuinely strategic approach to arm’s-length bodies would begin from a different premise. It would ask which capabilities the British state needs to discharge at arm’s-length over the next decade, what kinds of organisations are best suited to discharge them, and what investment in capability and design would be required to build them. It would treat existing chairs and chief executives as a strategic resource, not as administrative overhead. It would distinguish between bodies whose continued existence cannot be justified, where consolidation is the right answer, and bodies whose existence reflects a real and continuing public function, where the question is how to enable them to operate effectively. And it would invest seriously in the centre-of-government capability needed to make those judgments well, which is currently thin.
The PAC recommendations are a useful starting-point. A streamlined reporting regime for smaller, low-risk bodies, clearer signposting of central support, and a more proportionate approach to functional standards would reduce wasted effort and free up time for delivery. However, the larger task, which neither the PAC report nor the Cabinet Office’s rewire vocabulary address, is the question of institutional design. The British state will need a stronger arm’s-length sector in the 2020s and 2030s. The danger of the current programme is that without a strategic frame, the state will simply be left even more hollowed out.