Devolved innovation funding needs investment discipline beyond cluster branding

Devolution raises the standard for local investment judgement

The case for giving places a stronger role in innovation policy is persuasive: local leaders can often see practical barriers that a national programme cannot. They may know whether firms lack access to specialist equipment, whether a major employer could become a first customer, which skills are scarce, or why smaller businesses do not use support already available.

But more local discretion does not automatically produce better investment. It transfers a difficult allocation problem closer to the ground. Government has signalled a greater role for regional leaders in shaping innovation funding and linking national innovation ambitions to local growth ↗ ↗. The immediate test is whether local institutions can make disciplined choices under uncertainty—not merely assemble coalitions around a preferred project.

That matters because innovation funding is routinely asked to achieve several things at once: commercialise research, improve productivity, create skilled jobs, strengthen supply chains, support entrepreneurship and widen opportunity. These objectives may be complementary, but they are not interchangeable. A university-business partnership might improve research translation without materially changing management capability or technology adoption among smaller firms. A specialist facility might meet a technical need while generating limited local value if local firms cannot afford to use it, if demand is weak, or if resulting contracts and supply chains sit elsewhere.

The danger is not that places tell ambitious stories. Ambition is necessary. The danger is allowing a persuasive story about a cluster, campus or flagship building to stand in for an investment case. Every area can identify an anchor institution, a promising sector, a development site or a successful start-up. Those are assets; they do not establish that public support is additional, that the proposed tool fits the problem, or that benefits will reach firms and workers beyond a small circle of incumbents.

**Sanctuary assessment.** Devolution should be judged by the quality of the investment thesis it enables. For each material commitment, local leaders should be able to state: the binding constraint; why it persists without public action; whose behaviour the intervention is intended to change; the route from that change to economic value; and the evidence that would justify expansion, redesign or withdrawal. This is a more demanding test than a cluster prospectus. It is also the appropriate one when funding, institutional capacity and political credibility are limited.

A cluster is an asset base, not a causal case for support

UK Research and Innovation’s guidance for the Local Innovation Partnerships Fund places weight on strategic vision, ecosystem capability, governance, accountability and routes from research to commercial application (UK Research and Innovation, 2025) ↗. Read properly, that is not an invitation to rebrand every local strength as a cluster. It directs attention to whether partners can convert capabilities into products, services, adoption and sustainable commercial activity.

The distinction is important because cluster policy can confuse visible activity with impact. More firms in an area, fuller workspace, new networks, events and favourable coverage may indicate momentum. Yet none of these outcomes alone shows that an intervention improved firm performance, raised productivity, created additional jobs or broadened access to opportunity. A cluster may have grown without policy; it may also have attracted activity from elsewhere rather than increasing national capacity.

Nathan’s evaluation of the Tech City programme is a useful corrective. Using workplace-level microdata, synthetic controls and placebo tests, the study found evidence that the cluster became larger and denser, while evidence on firm performance was weaker and less stable. Some apparent changes also pre-dated the programme’s rollout (Nathan, 2022) ↗. The finding does not show that cluster policy cannot work, and London cannot be treated as a template for every city-region. It does show why growth in a cluster’s size is an inadequate proxy for policy additionality.

There is a legitimate counterargument. Requiring immediate productivity or employment gains from every programme would favour short-term, easily measured activity and discourage investment in shared capabilities whose effects take time to emerge. UKRI’s guidance recognises that longer-term outcomes and impacts may develop over an extended period (UK Research and Innovation, 2025) ↗. Intermediate outcomes can therefore be valid: a credible pipeline of demonstrators, functioning partnerships, investable propositions or specialist infrastructure that firms can actually use.

The qualification is that intermediate outcomes need a credible pathway to value. A shared laboratory should be assessed not mainly by occupancy, but by whether it resolves a demonstrated technical or regulatory bottleneck; whether SMEs can access it at a viable cost; and whether use leads to prototypes, paid contracts, investment, regulatory progress or stronger in-house capability. The same applies to property-led proposals. Specialist premises may be necessary in advanced manufacturing, health technologies or clean energy. Premises alone, however, do not create customer demand, commercial management capacity, validation capability or a sustainable operating model.

This changes the question local boards should ask. Rather than asking whether a place has enough assets to justify a cluster investment, they should ask which constraint prevents existing assets from producing more commercial and local economic value—and whether the proposed intervention is the most proportionate way to address it.

Make early commitments testable rather than irreversible

Innovation policy cannot wait for perfect evidence. Local institutions will sometimes need to take risks that individual firms, investors or universities will not take alone, particularly where benefits depend on coordination, shared facilities or difficult early adoption. The practical alternative to either optimism or paralysis is to make public risk explicit, bounded and revisable.

**Sanctuary recommendation: use four decision gates for material commitments.**

**1. Diagnose the constraint.** Start with business journeys, not sector labels. Evidence should include firms that have grown, stalled, disengaged or left, as well as customers, investors, researchers, skills providers and trusted intermediaries. The constraint may concern prototype finance, specialist equipment, clinical or regulatory validation, technical skills, first-customer access, supply-chain capability, management capacity or adoption support. “We need a stronger cluster” is an ambition, not a diagnosis.

**2. Establish the public role.** A proposal should explain what is likely to happen without intervention; why private finance, routine institutional activity or an existing programme will not resolve the issue; and what public risk is justified. This helps distinguish between subsidising a proposition that is already commercially viable and addressing a coordination failure or shared capability gap. The latter is generally the stronger rationale for place-based support.

**3. Fund bounded experiments.** Initial commitments should include a hypothesis, accountable owner, decision rights, baseline and review date. Indicators should follow the mechanism. For a testbed, relevant signals may include qualified firm demand, completed trials and pilots converting into paid contracts. For supplier development, they may include new local supplier relationships, repeat orders or capability accreditation. For management and technology-adoption support, sustained use of a new process is more meaningful than workshop attendance. Such measures are not substitutes for impact evaluation; they indicate whether the proposed mechanism is operating.

**4. Scale, redesign or stop.** Exit conditions should be agreed before institutional attachment takes hold. High activity without signs of commercial progress, adoption or capability gain should trigger redesign, consolidation or closure. This does not mean demanding certainty from early-stage work. It means refusing to treat continued activity as evidence that a programme deserves permanent support.

UKRI’s early review of the Place Based Impact Acceleration Account pilot supports the operational case for embedding monitoring and evaluation in delivery, alongside partnership and cluster-building activity (UK Research and Innovation, 2025) ↗. As an early-stage review, it cannot settle the question of long-run economic impact. Its more practical lesson is valuable nonetheless: baselines, data access and review arrangements are much easier to establish before a programme becomes institutionally entrenched.

Use local knowledge, but build in challenge and inclusion

The advantage of local decision-making is not proximity for its own sake. It is the potential to improve programme fit. Local institutions may be better placed to test whether demand for a facility is real, whether an employer can act as an early adopter, which organisations smaller businesses trust, and whether a claimed skills shortage is genuinely constraining growth.

Yet proximity creates its own risks. Universities, large employers, property interests and well-networked intermediaries commonly have more time and capacity to shape agendas than early-stage firms, social enterprises, small manufacturers or businesses outside a place’s most visible sector. Capital projects can gain political momentum even when a less tangible intervention—brokerage, investment readiness, management support, procurement access or adoption assistance—would deal more directly with the diagnosed problem.

Research on the UK’s regional policy levers identifies collaboration and coordination as important assets, while also highlighting funding uncertainty and institutional capacity as constraints on sustained action (Bentley, 2023) ↗. The implication is not that local partnerships are inherently unreliable. It is that capable partnerships require governance able to challenge their own preferred solution, especially where delivery institutions also stand to benefit from the funding decision.

**Sanctuary recommendation.** For substantial investments, innovation boards should publish a concise decision record: the problem definition; options considered; conflicts and their management; baseline; pathway to value; leading indicators; review date; and conditions for materially changing or withdrawing support. This is not transparency as a procedural exercise. It makes assumptions contestable, gives smaller firms a clearer basis on which to challenge decisions, and reduces the likelihood that a regional narrative substitutes for a causal case.

Inclusion is also part of investment quality, rather than a communications add-on. Research on UK second-tier city-regions cautions that a narrow focus on established high-technology clusters can overlook networks and foundational-economy sectors that employ many more people (Pugh, 2024) ↗. This is not an argument for dispersing innovation funding indiscriminately. It is an argument for testing the route to local value: SME access, local procurement, supplier development, workforce progression, new enterprise formation, or technology adoption among firms serving neighbourhoods and high streets.

A place may reasonably back a specialised cluster with concentrated benefits. But if it also claims wider gains in productivity, employability or inclusion, it should specify how those gains will diffuse and which complementary measures are needed to make that plausible.

Build portfolios around constraints, not brands

The strongest local innovation strategy is unlikely to be a single flagship. It is more likely to be a portfolio whose tools fit different barriers: feasibility support where technical and commercial viability are uncertain; testbeds where validation is costly; challenge-led procurement where a credible first market is missing; supplier development where value is leaking from the local economy; and management or peer support where SMEs need help to adopt technology that already exists. Specialist facilities can have a place where equipment is genuinely scarce, shared and commercially consequential.

A portfolio makes trade-offs harder to hide. Local leaders should not imply that every intervention will simultaneously maximise frontier research, broad-based productivity, high-street vitality, social inclusion and near-term job creation. Each investment needs a primary objective, realistic secondary benefits and measures proportionate to both. A programme designed to accelerate research commercialisation should not be judged only by immediate employment; equally, it should not claim broad local productivity gains without a credible route to adoption beyond a narrow group of firms.

**Sanctuary conclusion.** The success of devolved innovation funding will not be measured by the number of cluster brands, new boards or flagship sites it produces. It will depend on whether local institutions can identify real constraints, take justified public risk, resist institutional capture, learn from early evidence and end support when its route to local value does not hold. That discipline is not a brake on ambition. It is what gives place-based innovation policy a credible chance of becoming more than a transfer of responsibility.