WEEKEND READING: The £265 billion illusion: part two

This blog was kindly authored by Paul Marshall, a former director of Pragmatix Advisory.

Part one of this series argued that standard university economic impact assessments measure scale rather than distribution: they tell us how much economic activity is associated with an institution, but very little about who benefits from it or whether it contributes to wider prosperity in the places universities claim to serve.

This second part asks the practical question: what would a better assessment look like, and what would it take to produce one?

What a better model would look like

Universities that are serious about their anchor institution role need a different assessment model: one that measures inclusive impact, not just aggregate output. As part of the development of Universities UK’s blueprint for change, Andy Haldane, the former Chief Economist at the Bank of England, was commissioned to explore how the assessment of universities’ impact could be improved. He recommended that Universities UK commit to developing a methodology, in conjunction with national and local government, that the sector could use to capture a broader sense of the impact created by providers. To date, that recommendation remains undelivered almost two years on from the blueprint’s publication, which perhaps speaks to the challenge of requiring economic research consultancies to not use their own models when undertaking impact assessments for a price that is palatable to the cash-strapped sector.

Such a methodology would still need to quantify the direct economic contribution of the institution, but would also identify and measure the local distributional effects that many current assessments leave invisible.

A Social Return on Investment (SROI) approach offers one practical route. Rather than beginning with what data happen to be available, an SROI model starts with a theory of change: a structured account of what the institution is trying to achieve, for whom, and through what activities. This maps institutional inputs (spending, employment, teaching, research, civic engagement) to the full range of outcomes being generated at community level. Where outcomes cannot be directly monetised, appropriate proxies can be used to assign a financial value. The result is a figure that can sit alongside the traditional GVA calculation, offering a richer and more honest picture of institutional impact.

The Centre for Local Economic Strategies (CLES) approach to community wealth building, which was developed in partnership with Preston City Council, focuses not on growth for growth’s sake, but on how growth strategies can be shaped around the everyday needs of communities. The model focuses on the power of modifying the procurement policies of anchor institutions and companies to support local supply chains, improve employment conditions, and increase the socially productive use of land and property. This produced measurable reductions in mental health issues, improvements in life satisfaction, and increases in local wages. Demos Helsinki’s wellbeing economy framework offers a complementary approach, providing a structured methodology for measuring how economic activity translates into individual and community wellbeing using indicators that go well beyond GVA. Both could be adapted for a university context.

At a minimum, a properly designed inclusive impact framework would need to cover: disaggregated employment data, local procurement ratios, community wealth indicators, housing and cost-of-living effects, and access and progression outcomes by socio-economic background. The table below sets this out in full, contrasting what current assessments measure with what a genuinely inclusive framework would contain.

Why current data makes this hard

The concept is straightforward enough in principle, but the real obstacle is data. One structural limit is that the Office for National Statistics (ONS) does not publish GVA below NUTS3 sub-regional level, which is far coarser than the ward or neighbourhood scale this kind of analysis requires. So a neighbourhood-level economic picture cannot be measured directly; it has to be built from proxies, which is precisely why data held by local authorities and NHS trusts matters so much to this agenda.

The gap between current practice and what is needed runs right across the standard metrics. Where current reports count total jobs, the fuller picture would distinguish wage level, contract type, and how many roles go to local versus non-local employees. Graduate retention figures would be disaggregated by socio-economic background, ethnicity, and subject, and then tracked through to employment outcomes, not just first destination. Supply chain and student spend would separate what stays in the local economy from what leaks to national contractors and out-of-area landlords.

The North East graduate inversion illustrates the problem. A graduate from a North East university who takes a highly-skilled role in the regional innovation economy may contribute far more to genuinely inclusive local growth than their wage premium figure suggests, precisely because that premium will be lower than the equivalent role would command in London or the South East. The metric penalises exactly the kind of graduate behaviour that place-based growth policy is trying to encourage: a model that cannot see this is not just incomplete; it is actively misleading.

Closing the gap means drawing on three kinds of data. First, institutional data that already exists but is not disclosed in disaggregated form: employment by wage band and contract type, procurement by supplier location, graduate outcomes by background. Second, administrative data held by partner organisations that can serve as neighbourhood-level proxies: ONS Business Register and Employment Survey data, Valuation Office Agency business rates and vacancy records, Council Tax and benefits data from local authorities and health inequalities data from NHS trusts. Together these can build a credible ward-level picture even without direct GVA figures. Third, longitudinal data showing what actually changes in a neighbourhood before and after a university-led intervention to give the before-and-after evidence that a single snapshot assessment cannot provide.

The real barrier is not collection but sharing. Inclusive local impact cannot be measured by a university working alone. It requires data sharing between universities, local authorities, and NHS trusts. Without this, impact reports will keep leaning on whatever proxies are convenient rather than the ones that are actually right.

Good strategy, not just good ethics

There is a compelling strategic case for this shift, not just an ethical one. Universities in post-industrial cities are operating in a policy environment that is increasingly oriented around place-based growth, devolution, and regional rebalancing. Public confidence in higher education in many of these places is, at best, fragile. An institution that can demonstrate that it is making its immediate neighbourhood more prosperous with disaggregated, community-level evidence will be better placed to defend its public funding, attract civic partners, and contribute credibly to local growth plans than one that simply reports a larger total GVA.

The Preston model is instructive here as the CLES community wealth building approach did not require Preston’s anchor institutions to do less of what they already did. It required them to think differently about how they did it: where they spent their procurement budgets, how they structured employment, how they used their land and property. The changes were operational, not transformational. The impact on the surrounding community was measurable and significant. Universities could make the same shift, but the question is whether the current model gives them any reason to.

Universities are good for the economy – but whose economy?

The answer, as this two-part series has argued, is not always the local economy, and rarely in equal measure with the national economy.

Universities are among the most significant economic actors in their host cities and regions, and their contributions to research, teaching, and civic life are real and substantial. Some are already asking the right questions and investing in local supply chains, designing access programmes that target the most disadvantaged communities, and tracking outcomes that go beyond a GVA headline. The problem is that the dominant assessment model neither requires nor rewards this behaviour. An institution can generate a large impact number while doing relatively little for its immediate community, and the current methodology will not tell you.

Changing this requires action at both institutional and sectoral levels.

At the institutional level, universities serious about their anchor institution role should commission assessments that go beyond aggregate output by mapping inputs to community-level outcomes through a theory of change approach, and drawing on the CLES and Demos Helsinki frameworks as practical starting points. The data-sharing this requires with local authorities and NHS trusts should be treated as infrastructure investment rather than administrative overhead.

At the sectoral level, Universities UK and university mission groups should develop a shared framework for inclusive impact reporting that sits alongside traditional economic impact assessments. The metrics appropriate for a research-intensive university in a prosperous city differ from those for a post-1992 university in a former industrial town, so the framework need not be prescriptive. But it should be sufficient to allow meaningful comparison on dimensions that matter for inclusive growth, and to generate the evidence base that local authorities and combined authorities need for their own strategic planning.

None of this is straightforward, and some of it will require universities to be honest about limitations they would rather not advertise. But the alternative of continuing to report ever-larger GVA figures that speak less and less to the communities universities are meant to serve is not a sustainable position in a political environment increasingly sceptical of institutions that claim public value without demonstrating it locally.