Blog

Experimental devolution

How Number 10 North can build a national policy lab

1 September 2026

The new UK government’s Rewiring the State agenda makes clear that much greater English devolution will happen, and attention is now turning to how to make it a success. In the first output from a Review of Growth-Capable Devolution, Mike Emmerich warns that strategic authorities currently lack the institutional capability to use their new powers effectively; Jun Du, the Productivity Institute’s new Director, highlights the need for a central coordinating mechanism; and Geoff White reminds us that the UK has been historically poor at learning from policy innovations.

If Number 10 North is to be the true engine room of regional growth, it must solve this triple challenge of capability, coordination, and learning. Its mandate and capabilities to achieve this are further strengthened with the news that, alongside responsibility for devolution, it will also take the lead on the Test & Learn agenda and potentially key areas of economic growth responsibility from HM Treasury.

A strategic challenge is to move away from decades of centralisation without creating a system that devolves into fragmented, duplicated regional actions with unclear outcomes and limited accountability. Over recent months we’ve been exploring what an experimental approach could look like to support this agenda. How could direction and accountability be provided from the centre without Whitehall micromanagement, supporting local policy experimentation in a way that radically devolves the execution while ensuring learning diffuses nationally?

Turning local innovation into national success

Regional devolution will open a wave of new policy ideas but we are lacking a mechanism to differentiate between well-intended but ineffective interventions and new ideas that really work.

The centre, which is ultimately accountable for national growth, must combine two approaches: ensuring widespread evaluation of spend for general accountability, and facilitating targeted local experimentation to explore the potential of more radical innovations and generate powerful causal evidence.

Part of the solution will be a system that combines:

What the centre does best: It can exploit economies of scale by building a shared data infrastructure and capture national learning spillovers so evidence generated in one region benefits the whole country.
What local leaders do best: Mayors and Strategic Authorities retain the operational freedom to use their knowledge of their local areas to design, adapt, and execute delivery on the ground, informed by verified evidence.

To achieve this, No 10 North could establish a new “Productivity Lab” that serves as a national learning facility that creates a shared learning engine through three roles:

Data and Comparative Intelligence: Collating fragmented programme data into a single unified database and unlocking restricted administrative micro-data (such as HMRC business and employment records). This would offer more timely and robust metrics to track policy delivery, outcomes, and comparative analysis – e.g. is one region proving more adept at engaging high-growth businesses or driving high-wage job progression than its neighbours.
Policy Innovation Funding: Providing dedicated capital (£100M) to co-fund local policy experiments and support rigorous evaluation, absorbing testing risks so local taxpayers do not foot the bill for generating national evidence. This approach would take inspiration from the Business Basics Fund, the high-conviction risk appetite of ARIA and the rigour of the Metascience Unit.
Targeted Place-Based Testbeds: Acknowledging the political reality that Ministers will have national economic priorities or simply a need to do something new, the Lab can also proactively prompt, co-fund, and accelerate high-priority experiments across regional frontline machinery. While this central funding creates a strong incentive to follow the set agenda, local leaders would still retain control over what they scale based on the robust proof generated.

Place-Based Testbeds: De-Risking Delivery on the Ground

Under this framework, No 10 North provides the funding, evaluation support, and data infrastructure, while local Mayors and Strategic Authorities voluntarily design and run the testbeds on the ground.

By covering the additional operational costs, the centre can provide incentives for experimentation, recognising the value of generating evidence where the benefits are shared nationwide. The centre could choose to prioritise specific areas for exploration but importantly, final adoption would remain voluntary: local leaders retain control over what they scale, guided by the robust proof generated.

We suggest four initial areas for proactive exploration:

How do we reach people and businesses in all areas? Despite some promising exploration, innovation and business policy has overlooked the potential for Further Education Colleges to be engines for innovation and business productivity in the “foundational economy”, with the potential to become hyper-local anchor institutions that are frequently situated precisely in areas suffering from long-term stagnation. This testbed will explore how to shift Further Education away from purely foundational skills to drive technology adoption and AI readiness directly into local SMEs and workforce.
How do we pull innovation into local economies when regional universities face severe financial strain? Policy too often judges higher education by narrow, linear metrics like patents filed, taking regional market demand entirely for granted. This testbed will trial alternative, low-cost knowledge-exchange models that reward academics for creating regional economic and public value, while exploring interventions to stimulate local business demand so they actively pull innovation out of university labs.
The future is uncertain; can we equip people with skills to adapt and find new opportunities? Enterprise education often focuses on the tactical and administrative aspects of running a business, but fails to teach how to spot opportunities and develop solutions that create value. This testbed would evaluate how to build these skills amongst current and future entrepreneurs, but could also have the potential to expand and consider how to develop “absorptive capacity” and “problem solving” across the wider workforce.
Can local areas make more effective use of national assets to drive growth? Whitehall holds powerful economic levers and assets, including administrative data and tax incentives (e.g. full expensing and R&D Tax Credits). Through local delivery there is the potential to bring these fragmented pieces together and more proactively target businesses. This testbed will experiment with that local synthesis, exploring how regions can leverage central data to target high-potential firms before they even ask for help.

For cross-cutting structural levers that remain under central government control (e.g. tax), the centre can still adopt an open, experimental approach and actively involve specific regions in sandboxing next-generation economic tools. This could, for example, include trialling ideas such as ecosystem tax credits that allow local firms to directly fund regional public goods.

Building the “Learning State“

With the Test & Learn agenda and key growth levers moving into No 10 North, the government has an opportunity to support its devolution agenda with a process of proactive and empirical learning – devolving execution while nationalising learning to navigate between the centralisation trap and a world of policy fragmentation, duplicated effort and unclear evidence of what actions drive impacts on the ground. By positioning No 10 North as a national learning facility, every regional policy experiment becomes a shared asset, building a definitive national picture of what can work.