An anatomy for interpreting devolution?

Each government action is interpreted through a ‘devolved or centralised’ lens. This binary is unhelpful.

I think we are witnessing changes in the devolution policy landscape that we do not have a language to articulate.

One of the lenses through which policy-makers and observers judge UK Government action is through a devolution-centralisation binary: whether or not the Government’s decision to, for example, reduce the number of police forces is, by design, an example of either of those forces. We do this, for the most part, reflexively.

However, there is a more developed anatomy for interpreting devolution, in which policy may be centrally or locally coded, but is more complex than is given credit. I set out four distinct, but overlapping, processes of this new anatomy.

But first, why does this matter? Understanding — and articulating — what is taking place matters because how government action is interpreted shapes the direction of devolution. When reform is judged through an adversarial ‘devolved versus centralised’ binary, important distinctions are lost. New measures that may be better understood as recalibration or standardisation are too easily framed as centralisation, creating unnecessary conflict and mistrust between Whitehall and places. Over time, this weakens the culture of devolution by encouraging defensive responses rather than constructive engagement.

A more nuanced understanding is also essential for protecting devolution politically. English devolution remains fragile and contingent on continued political support. Without precise language — (alongside much else, including as I argue here, more evaluation and analytical tools) — skeptical future governments will find it easier to make the case against devolution.

Crucially, the anatomy also exposes how devolution is chipped away at not through dramatic reversals, but through incremental and technical means, such as consent regimes. These mechanisms can undermine autonomy while leaving formal structures intact. By making these dynamics visible, the anatomy functions as an early-warning system against gradual recentralisation (which as I highlight below, has happened elsewhere).

More broadly, this approach reframes devolution as a question of system design rather than ideology. Instead of reducing the debate to “local versus central” it focuses attention on scale, coordination, capacity, accountability and evidence. It encourages policymakers to be more intentional in asking where autonomy adds value, where coordination is essential, and where common standards are justified. In doing so, it makes devolution more resilient to political change.

Ultimately, the anatomy also reflects the reality that England is beginning, tentatively, to move toward: beyond a ‘campaign model’ of devolution focused on securing powers — though crucially important — and to a governance model concerned with how power is exercised and sustained, and how it evolves. In seeking to identify a shared language, I think this — very, very early iteration of an anatomy — increases the durability of devolution.

Devolution is not linear

Over the past decade, we have witnessed significant devolution. That is now a familiar and positive story, as productivity data from the North of England — and Manchester specifically — has begun to show (even taking into account Paul Swinney’s thoughtful critique of the HQ effect and data inaccuracy).

But there is a prevailing assumption that devolution is linear and, as a result, the focus has only even been on the next ‘wave’ of devolution, rather than how we preserve it.

Yet there are multiple reasons why devolution might not progress as we might envisage. Is there appetite from those in power to create more political opponents? Does the evidence tell us a clear and convincing picture of the merits of devolution? Given the critical appraisal of the Civil Service as an embodiment of the ‘stakeholder state’, will Whitehall come to view devolution as a vehicle which gums up the machine further? It is notable that the Government has decided to encourage more Foundation Strategic Authorities.

Moreover, I think weak monitoring and evaluation of devolution is an achilles heel that may be weaponised by those skeptical of the merits of devolution to pursue an alternative.

Rather than a radical departure from the current status quo on devolution though, it is more likely — indeed, we are witnessing it — that it is a little more difficult to determine what is taking place. Re-evaluating what competencies and functions sit at which spatial level — local or strategic authority, Investment Promotion Agency, Growth Company, regional, pan-regional — can often be interpreted as devolution or centralisation, depending on your starting point. Standardisation can also be viewed as the antithesis of devolution, which by design provides some tolerance of place-specific actions, when it is more benign and there is something plainly wrong with our governance arrangements.

I set out the four processes which, I think, better reflect the anatomy of devolution.

1. Devolution

Epitomised in the creation of new strategic authorities — which are overseen by an elected Mayor; in ‘double-devolution’ such as the creation of Neighbourhood Governance Board’s under the Pride in Place programme or; primary legislation that the Government is introducing.

While it is reasonably clear what devolution is, within each of these there are elements that might be described otherwise. As I explore further in this column, there is at least an element of the Pride in Place programme that is centralist in approach.

2. Recalibration

In the context of multi-level governance, functions operating within their current institutions or at their current spatial geographies may be recalibrated. The OECD describes this phenomenon as ‘upscaling governance’, with strategic authorities taking a more central role in economic development, as opposed to local authorities, for example.

The rationale is often mistaken for centralisation, but is more commonly because the scale at which a function operates is suboptimal. There are multiple reasons for that. For example the scale was poor to begin with. The decision may have been decades prior and for political reasons which still may be salient (it will divide opinion, but many would point to ‘lower-tier’ district authorities as an example of recalibration, rather than centralisation, given that the competencies are being streamlined rather than returning to the UK Government). In one respect, recalibration corrects a historic weakness in our governance.

That should also be separated from the process of re-organisation, which was led by Whitehall. Few authorities signalled an appetite for it before the General Election.

Another reason is that we may have not had the institutional architecture in place when a decision was made over who should be responsible for a competency. Demographic change, economic geography or public opinion also may be explanations.

To take one example, the Mayor of Greater Manchester has been active in supporting his local authorities to tackle homelessness. Andy Burnham made it a priority of his mayoralty (the Centre for Homelessness Impact has done some fantastic work with Greater Manchester). Many strategic authorities have avoided public service reform until now in part because of concerns that local authorities will be critical of them for stepping into their sphere of responsibility. But strategic authorities can operate at a scale that local authorities cannot. In London local authorities are currently outbidding one another for accommodation to house their homeless; Capital Letters was set up to tackle this issue but has been unsuccessful; the Greater London Authority has not stepped into that void.

Meanwhile, given their purchasing power, local authorities wish to negotiate better rates with hoteliers who are charging them market rate to place homeless families in them. Leaving aside the quality of the accommodation for a moment, international hoteliers such as Travelodge and Holiday Inn have not engaged with those authorities. Could they simply ignore Sadiq Khan in the same way they can a local authority? I suspect not. In both cases, each example provides a strong rationale where a strategic authority can provide additional value to support a local authority service.

3. Standardisation

Standardisation too is usually critiqued as being centralising, but often in stronger terms than recalibration. Commentators might suggest standardisation is anathema to the ‘managed difference’ (to steal the phrase from The Lyons Inquiry into Local Government) of the place-specific nature of interventions. Others would add that this may breach well-established principles of subsidiarity.

Yet the impetus for standardisation is usually a result of the fact that our sub-national architecture is dysfunctional. That concern is often expressed by Whitehall, but it exists beyond it too.

And some of this standardisation takes place — I argue — because the sector has not diagnosed the issue, or has done so but either hasn’t or is unable to take unilateral action to address it. The Government’s decision to require local authorities to publish data on how many potholes they’ve filled in is one example. I would rather the Government didn’t mandate authorities to do it, but it is both absurd that we do not have a national picture of pothole repairs because local authorities either don’t publish their progress or calculate the number of potholes they fill in using different methodologies and that (b) local authorities have not been able to address that proactively.

A second example might be the Ministry of Housing, Communities and Local Government’s decision to impose new strategic authorities, where necessary. The English Devolution White Paper says that:

The government will work collaboratively with local government to deliver on the ambition of universal coverage of Strategic Authorities in England. However, in order to ensure a complete national layer of Strategic Authorities is in place to devolve further powers to in future, we will legislate for a ministerial directive, which will enable the government to create Strategic Authorities in any remaining places where local leaders in that region have not been able to agree how to access devolved powers. Our commitment to working in partnership holds firm, and so the government will limit its use of this power to instances when other routes have been exhausted. We will ensure that the ministerial directive is used to conclude the process where there is majority support, or the formation is essential in completing the roll out of Strategic Authorities in England.

Though it may be undesirable for the Government to establish new strategic authorities without universal consent, it would be even less acceptable if one or two authorities simply did not create a strategy authority (which is what is happening in a handful of places because they are incapable of agreement, for example, governance arrangements).

A third example of standardisation may include the Government’s decision to consolidate police forces. The Policy Paper notes:

“We will create a more consistent service by holding the police to account for delivering a set of Local Policing Guarantees, backed up by new powers to intervene where forces fall short.”

4. Centralisation

I worry, as I set out earlier, that we are labouring under a misguided belief that devolution is linear. It is at best complacent and at worst undermines devolution.

There is precedent for this ‘recentralisation’ trend internationally. Evidence from the OECD notes that:

“Several OECD countries have experienced a reduction in the share of subnational government expenditure in GDP and (or) in general government over the past two decades. In Hungary, there was the recentralisation reform, started in 2011-12 with the constitutional reform and the Local Government Law which led the central government to reorganise the Hungarian local government sector and take over many functions exercised previously by municipalities and counties. As a result, Hungary went from being quite decentralised compared to other OECD countries to being among the most centralised. In 2010, subnational expenditure amounted to 25% of public expenditure and 12% of GDP to 13% of public expenditure and 6% of GDP in 2017, a reduction by half in both cases.”

Closer to home, Ireland witnessed a similar fate:

“The case of Ireland is … a strong decrease in subnational spending in 2005, 2010 and in 2014 resulted both from recentralisation trends (e.g. water services), from Better Local Government reforms, including the 2014 Local Government Reform Act which merged 114 local councils into 31 local governments and abolished the previous 8 regional authorities, and from the 2008 recession. The crisis significantly reduced local income and necessitated spending and staffing cuts imposed by central government.”

And while there is overwhelming evidence of devolution over little more than a decade in England, some actions taken by Whitehall can be viewed, legitimately, as centralising.

The commonly used tool in Whitehall’s toolkit is requiring approval from the Secretary of State prior to introducing a policy change. One egregious example is that the Secretary of State for Transport must approve the replacements of cattle grids where there is an objection to a decision by a highway authority (under the Highways Act of 1980). Despite the Devolution White Paper committing to remove these “outdated” and “nonsensical” consent, it is not currently included in the draft legislation. Under the Transport Charges Act of 1954, toll charges (even of a penny) cannot be approved without the Secretary of State’s approval either.

Examples of centralisation may also be embedded within legislation or policies otherwise giving places more devolution overall. Neighbourhood Boards will oversee the Pride in Place Programme, for example. The Chair is appointed and approved by the local authority and Member of Parliament. But where a Chair cannot be agreed by those parties, the Government will intervene. Likewise, local authorities (or the Neighbourhood Board itself) cannot take a view on the boundaries that the funding must cover. Instead, they will have to apply to the Government to change it. Does the Government really need to concern itself with adjudicating over who should be Chair of a Neighbourhood Board?

Finally, and most egregious, is overtly political centralisation. For example, when Grant Shapps was Secretary of State for Transport, he wrote an astonishing letter in 2022 to the Mayor of London over their impasse to secure a long-term settlement for TfL, in which he accused Sadiq Khan of “artificially creating” service reductions at TfL “as a political weapon in your campaign for a long-term capital funding deal”. He went on to add:

As you know, we do want to give TfL a longer-term capital deal. But your tactics are the wrong way to achieve one. They are harming London’s interests. If we are to make a longer-term funding commitment, you must keep your promises and we must be able to deal with you and TfL on a basis of honesty and seriousness, not campaigns of scaremongering and threats. Your latest stunt today – where you have prematurely announced details of an extension before it was finalised with the department – is just the latest example of your desire to play politics, rather than working constructively with us. TfL’s services are at no risk, unless you want them to be, and should not be undermined for political ends.

As we have stated before, we will continue revenue support, in further temporary deals if necessary. We remain open to giving you a longer-term capital settlement. But it will require a reset of the relationship.

At risk of this becoming an academic article, I will conclude. If English devolution is to mature, it must be understood not as a simple tug-of-war between Whitehall and places, but as a complex and evolving system of governance that is constantly being adjusted, constrained, and reimagined. The anatomy set out here is an attempt to provide a more precise language for that reality: one that distinguishes between genuine empowerment, necessary recalibration, functional standardisation, and corrosive centralisation. Without this shared vocabulary, we risk mistaking technical change for political intent, reacting defensively to reform, and overlooking the slow accumulation of constraints that hollow out autonomy over time. With it, policymakers, practitioners, and observers can engage more honestly with where power should sit, how it should be exercised, and how it can be safeguarded.