What Does the Evidence Say About Devolution and UK Productivity?

By Dr. Sezer Yasar, Cardiff Business School, Cardiff University

Andy Burnham is planning a significant expansion of devolution in the UK. In economics, this is known as fiscal federalism, or decentralisation. What does the evidence tell us about whether Burnham’s plan is likely to lift UK productivity? The short answer is yes, provided the right expenditure powers are decentralised and supported by improved local institutional capacity and political accountability.

Burnham has been explicit that his priorities are growth, reindustrialisation and innovation (see his speech of 29 June 2026). The speech points to more powers for local government over transport, education, housing, water, energy, employment support, and growth and industrial policy. A new unit, “No 10 North”, would coordinate national and local government around a long-term economic strategy, making place-based collaboration “the new operating principle for UK plc.” Public and private investment would be consolidated at a place-based level. What is on offer is devolved spending and strategic decision-making, coordinated nationally, but there is no proposal to devolve taxation.

An important motivation of Andy Burnham’s plan is “Manchesterism” (Burnham, 2026), termed the “Manchester Model” at The Productivity Institute. The model is a devolved city-region growth approach whose productivity, and growth record has been attributed in part to devolved decision-making, dedicated investment funds, and transport investment (McCann and Ortega- Argilés, 2026).

However, the Manchester Model is not restricted to devolved powers, it is a larger story of institutional collaboration, coordinated leadership and coherent investment planning, built mostly before any formal devolution of powers (McCann and Ortega-Argilés, 2026). What follows concerns fiscal decentralisation, meaning formal shifts in expenditure or revenue-raising powers between levels of government. So, this post speaks to the general logic of devolving powers, not to the Manchester Model or Burnham’s full plan directly.

How centralised is the UK, really?

The UK’s public finances are unusually centralised compared with most OECD countries, and the gap is starkest on tax, the one lever Burnham is not proposing to pull. On spending, devolved government accounts for around 10% of total UK government spending, against roughly 30% for state governments in OECD federal countries (HM Treasury, 2026; OECD, 2026). Local government spending sits closer to the OECD average overall, though still below UK levels before the 2008 financial crisis, roughly when the UK’s productivity problem began. On revenue the picture is more centralised. Devolved government tax revenue is about 2.5% of the UK total, against roughly 16% for state governments in OECD federal states. Local government tax revenue, which is about 5% of the UK total, is also below the OECD average of about 10% (HM Treasury, 2026; OECD, 2026). There is little doubt that room exists to decentralise further.

Why decentralisation could help or hurt productivity

Economic theory gives decentralisation three plausible routes to higher productivity (Baskaran et al., 2016):

• Matching. Local governments know local needs, so services such as education can be tailored to them.
• Experimentation. Regions try policies in parallel, and better ones emerge and spread.
• Competition. Rivalry between regions rewards pro-business fiscal choices over overregulation and rent-seeking.

However, if spillovers across regions are ignored, or if tax competition erodes the revenue needed for productive public investment, decentralisation lowers productivity instead. And the gains only materialise where local political accountability is strong. Where it is weak, new powers are more likely to be captured for local political gain than used to serve residents.

Burnham’s plan, which has no tax dimension in his speech, leans toward the favourable channels, including matching policy to regional conditions, parallel experimentation, and competition-driven, pro-business fiscal and regulatory choices. It also carries the risk of under-investment in public goods with strong spillover effects, which No 10 North’s coordinating role is designed to manage on the spending side. What the plan does not yet address is how local political accountability will be strengthened alongside the new powers, which the theory suggests matters as much as the powers themselves.

What the empirical evidence actually shows

The empirical literature that investigates decentralisation and productivity relation is small but points more consistently in one direction. On the expenditure side, Kalyvitis and Vella (2011) find that decentralising infrastructure spending has a positive relationship with labour productivity (output per hour) in the US. On the revenue side, Bartolini et al. (2019) show that tax revenue decentralisation has a negative effect on labour productivity across the OECD countries. Burret et al. (2022) find similar patterns in Switzerland. Total expenditure decentralisation has a weak but positive relationship with labour productivity (output per worker), while total revenue decentralisation has a negative one. They find that productivity increases with fiscal competition between regions and decreases with equalisation transfers between them. Two further studies point to some channels through which this may occur. Barankay and Lockwood (2007) find that government efficiency, measured by student attainment, is higher in Swiss cantons where education spending is more decentralised, and Kappeler et al. (2013) find that investment in productive infrastructure, rather than redistribution, is higher in European countries where revenue decentralisation is greater.

Taken together, these studies suggest that some forms of expenditure decentralisation may support productivity through local governments’ informational advantage in identifying where more spending is needed, such as in infrastructure maintenance. On the other hand, tax revenue decentralisation can decrease productivity by leaving less revenue with the central government to spend on investments with strong spillover effects, like large-scale infrastructure and R&D, and because the fiscal transfer systems that often accompany it reduce competitive pressure between regions.

What this means for Burnham’s plan

UK public finances are unusually centralised, and the productivity-specific evidence, though still limited, suggests that expenditure decentralisation tends to support productivity. That evidence, however, comes from studies of particular functions, namely infrastructure and education, rather than of devolved spending in general. Read in that light, local transport and other intra-regional infrastructure, together with education, look to me like the strongest candidates for expenditure devolution in the UK. The same logic extends to labour training, where local knowledge of employer needs should matter most. They also sit broadly close to the territory Burnham’s plans cover. Revenue decentralisation looks weaker, even negative, for productivity in the evidence so far, and it is also the lever Burnham has left alone. One important caveat concerns causality. Much of this research establishes an association between decentralisation and productivity, but the direction of causation is not yet well established. It remains unclear whether decentralisation is driving the productivity gains, or whether the two are simply moving together for other reasons. The Productivity Institute raises a similar caution in its own account of the ‘Manchester Model’ (Productivity Institute, 2026).

That said, Burnham’s plan is not only about targeted expenditure devolution. It also proposes a governance and coordination model, via “No 10 North”, that has more in common with the Manchester Model’s institutional-collaboration aspect than with the fiscal-decentralisation research reviewed here. Only the expenditure-devolution element of the plan is directly addressed by the productivity evidence above. The governance dimension would need to be evaluated on its own terms. It does, though, bear on where responsibilities should sit. On theoretical rather than empirical grounds, the local-information/spillover trade-off extends the case for devolution to business support and housing policy, the latter through its labour-market effects, while functions with strong spillovers, including R&D and science funding, are better retained centrally. Inter-regional infrastructure is a different case, since its spillovers can be handled by coordination across regions rather than by central control. That is precisely the task a body like No 10 North would exist to perform.

However, productivity is not the same as growth. In the same Swiss data, Burret et al. (2022) find expenditure decentralisation has a positive association with labour productivity but a negative one with growth. More broadly, a much larger literature shows a more mixed relationship between decentralisation and growth. Greater Manchester itself shows a similar disconnect: productivity has grown faster than the UK average since 2015, but median earnings and household income have lagged behind (Peck et al., 2026). That gap is a reminder that productivity growth need not translate fully, or quickly, into broader economic gains.

Keeping decentralisation’s productivity-supporting channels strong while containing its negative ones is a sensible practical goal. But the deeper lesson from the evidence is that it is not devolution per se that determines whether productivity rises; it is which powers are devolved and how they are exercised. On the first count, Burnham’s plan already looks close to what the evidence would recommend. On the second, two conditions seem to me worth attaching to any further expenditure devolution. The first is investment in the administrative and technical capacity to spend new funds well, supported by multiyear funding settlements that let local authorities plan the long-horizon investment infrastructure and skills require. The second is stronger local political accountability, which the theory treats as the precondition for any of the gains. Routine evaluation of productivity outcomes would both hold new spending to account and reveal which practices are working across local authorities.

References

Barankay, I., and B. Lockwood (2007), “Decentralization and the Productive Efficiency of Government: Evidence from Swiss Cantons,” Journal of Public Economics, 91(5–6), 1197–1218.

Bartolini, D., E. Ninka, and R. Santolini (2019), “Tax Decentralization, Labour Productivity, and Employment in OECD Countries,” Applied Economics, 51(34), 3710–3729.

Baskaran, T., L. P. Feld, and J. Schnellenbach (2016), “Fiscal Federalism, Decentralization, and Economic Growth: A Meta-Analysis,” Economic Inquiry, 54(3), 1445–1463.

Burnham, A. (2026), Andy Burnham Speech at the People’s History Museum, Manchester – 29 June 2026 (andyburnham.org.uk/2026/07/20/andy-burnham-speech-at-the-peoples-history-museum-manchester-29-june-2026/).

Burret, H. T., L. P. Feld, and C. A. Schaltegger (2022), “Fiscal Federalism and Economic Performance: New Evidence from Switzerland,” European Journal of Political Economy, 74, Article 102159.

Kalyvitis, S., and E. Vella (2011), “Public Capital Maintenance, Decentralization, and US Productivity Growth,” Public Finance Review, 39(6), 784–809.

Kappeler, A., A. Solé-Ollé, A. Stephan, and T. Välilä (2013), “Does Fiscal Decentralization Foster Regional Investment in Productive Infrastructure?” European Journal of Political Economy, 31, 15–25.

McCann, P., and R. Ortega-Argilés (2026), “UK-Wide Growth and Development: Insights from the ‘Manchester Model’,” Productivity Insights Paper No. 095, The Productivity Institute.

Peck, J., J. P. Spencer, and A. Westwood (2026), “Is Greater Manchester Growing Significantly? If So, Why? And What Can We Learn?” Productivity Insights Paper No. 097, The Productivity Institute.

HM Treasury (2026), devolved government expenditure and tax revenue shares are the author’s estimate from HM Treasury, HMRC and devolved administrations’ data.

OECD (2026), Fiscal Decentralisation Database (Table 5, government expenditure; Table 9, tax revenue), oecd.org/en/data/datasets/oecd-fiscal-decentralisation-database.html, accessed January 2026.

Productivity Institute (2026), “The Manchester Model,” productivity.ac.uk/tpi-research/manchester-model/, accessed July 2026.