Rewiring the state through increased English devolution is front and centre of Andy Burnham’s political project. In this week’s budget, John Healey spoke passionately about delivering “the most radical act of devolution in a generation”. The centrepiece of this is the devolution to mayoral authorities of income generation from business rates, with connected plans to create new devolution deals across 10 areas in England. In this brave new world, the chancellor argued, money raised locally can be spent locally, empowering people who know their areas best and driving economic growth.
While the budget was accompanied by a “fiscal devolution roadmap”, a White Paper this autumn will provide more detail for what are undoubtedly ambitious and—especially in the context of a fiscally conservative and gradualist budget—fairly radical plans.
But as any good electrician will tell you, before rewiring a house you must first test the existing circuits. The existing wiring might be fit for purpose, or the glossy new lighting controls might be a expensive choice you will soon regret. The same is true of devolution. Before the prime minister attempts a massive overhaul, he needs a finely grained understanding of the devolution that already exists.
Our major study (the first ever) of the relationship between devolution and social security across, and within, the four nations of the UK sets out to build our evidence base in this vital area and inform future policymaking. “Safety Nets” was motivated by a very simple problematic: the design, delivery and experience of social security differ according to where you live in the UK in ways that remain poorly understood.
Social security is far less centralised and uniform than people think, or than political rhetoric so often implies. Scotland, for instance, runs its own social security agency and has its own social security payments, such as the Scottish Child Payment, a weekly benefit for families on a low-income with dependent children. Key aspects of the austerity-era welfare cuts, including the Benefit Cap and the so-called Bedroom Tax are mitigated in Northern Ireland. Wales has fewer formal powers but a new Plaid Cymru government is itching for more, and the already has a novel form of administering discretionary support, involving a private contractor, a local authority and a voluntary sector organisation. And across the UK, support controlled by local authorities has grown from an almost-non-existent starting point in 2010-2011. It is still small, however: in, 2024-25 this accounted for just 1.2 per cent of total UK government spending on social security.
In other words, the system has been quietly rewired already. Burnham’s focus may be on English mayors, but the devolved nations are the UK’s longest-running experiments in exactly the kinds of trade-offs he is about to make.
What can the prime minister learn from this? We draw five lessons.
First, free school meals make an excellent devolution case study. How many people know that every part of the UK does something different? Wales offers free meals to all primary pupils in state schools, and Scotland to every child up to Primary 5 (the equivalent of year four in England). In London, the mayor funds universal provision across years three to six for every state primary pupil; and across England, children in all Universal Credit households now qualify. Northern Ireland, by contrast, has no universal provision, and a means-test that excludes many families who would qualify in England. When our project brought parents from all four nations together, many were surprised to learn how much provision differed. Almost all were angry about it, and not only those who are losing out. “How that hasn’t caused an outcry, I don’t know,” said one parent in England of Northern Ireland’s approach. This shows the classic trade-off: devolution leads to differences, and what some see as support tailored to local priorities, others experience as a postcode lottery.
Second, there is clearly a lack of awareness about the extent of the devolution of social security, not just among the public, but among politicians and civil servants too. When UK ministers discuss child poverty or young people not in work, they often show little understanding of how far devolution already reaches, as if they think they can uniformly shape the policy landscape across the UK directly from Whitehall. There are understandable reasons for this: the landscape is complex; there are no institutions to encourage learning between nations; and politicians tend to highlight what differentiates their administrations rather than championing the learnings that arise from differences in approach.
Third, devolution’s initial promise as a set of “policy laboratories” hasn’t been realised with social security. Take disability benefits, one of the most contentious areas. The forthcoming Timms Review will make recommendations for the future of Personal Independence Payments, a benefit that operates in England, Wales and Northern Ireland. But there is little sign that the Department for Work and Pensions is learning from Scotland, which runs its own benefit, Adult Disability Payment, with a very different approach to assessment.
Fourth, we all need to get more into the weeds of devolution to draw lessons both from where we see good practice, but also on what aspects of policy are most and least suitable for devolution. For example, discretionary crisis support, now the job of local authorities, benefits from local knowledge and links to other local services. But Council Tax Reduction shows what happens when the wrong thing is localised. Since 2013, every English council has had to design its own means test for a benefit whose rules were already well established. The result has been legal challenges, the return of cliff-edges that Universal Credit was meant to remove, and a pattern in which the least generous schemes tend to be found in more deprived areas. So localise where local knowledge improves decisions, and keep national anything where the rules are settled and consistency matters, including the basic adequacy of incomes.
Fifth, and finally, devolved revenue must match devolved responsibilities, not just in scale but in how both move with the economic cycle. When power is devolved without the resources to use it—or, worse, when devolution is a stealthy way to make cuts to the welfare state, as happened with Council Tax Reduction and crisis support in the 2010s—local leaders get blamed for gaps they didn’t create and people lose faith in politicians at all levels. Some have called this “austerity dumping”.
Social security has been rewired, piece by piece, for well over a decade. Some circuits are working well (such as Scotland’s approach to disability benefits), some have, historically, been underfunded and overloaded (such as local crisis support), and some should never have been wired that way (such as Council Tax Reduction).
Considered separately, or together, none of these five lessons constitutes an argument against devolution. But they are an argument for deepening our understanding of the shape, nature and implications of the existing devolved settlement before making further changes. Practically, the forthcoming white paper should apply these lessons to the devolution within England that will be its principal focus. This requires stating some underpinning principles for which domains are devolved, as well as a commitment to funding devolved powers properly, rather than devolving responsibility for cuts. But it also requires the creation of mechanisms for the UK’s various governments and local leaders to work together, and to learn from each other, even when different political parties are in charge. Now that would be a rewiring of the state we might all get behind.
Safety Nets is a Nuffield Foundation-funded study, involving researchers at seven HEIs across the UK, and policy partners, Resolution Foundation and Child Poverty Action Group. The research drawn on here is a collective effort, and the authors thank the wider research team and the participants and experts-by-experience panel who gave up their time to share their expertise and experiences for this research
