“We are a vital part of a nutritious breakfast”: Stian Westlake on what UKRI is for
The ESRC’s executive chair on buckets, missions, devolution, tight budgets and AI – and why research funding alone won’t fix growth
Stian Westlake has been executive chair of the Economic and Social Research Council for three and a half years, after spells at Nesta, in government as a ministerial adviser and running the Royal Statistical Society. At Research Agenda’s first live session, UKRI, transformed, James Coe asked him how far the argument for investing in ideas has come, what UKRI is actually for, and what devolution, tighter budgets and AI mean for the UK’s biggest research funder. The interview has been edited for length and clarity.
From the margins to the mainstream
You’ve gone from Nesta to the Royal Statistical Society to the ESRC. What gets you up in the morning?
The thing I’ve always been interested in – in my research, my writing and my policy work – is the knowledge economy: how we invest in ideas, and how those ideas improve the economy and people’s lives. The two books I co-authored were about that, and so was much of my work at Nesta. The problem with writing about something is that occasionally someone asks you to do something about it. Working as a ministerial adviser, running a learned society and now running a research council have been great opportunities to put some of those ideas into practice – and you learn a lot trying to do this stuff for real.
In 2013 you wrote a report for Nesta called Plan I. It covered regional inequality, the undervalued intangible economy, the gap between financial architecture and spending, low productivity and a failure to value R&D. Thirteen years on, did they listen?
Let’s start with the optimistic side. When I was saying things like that in 2013 – and a few other people were – we really felt like voices in the desert. We called it Plan I because the whole conversation was about plan A and plan B: austerity versus a sort of neo-Keynesian stimulus. The idea that innovation and technology mattered was something people in universities cared about, but in government it was pretty marginal. To the extent government was interested in research and innovation, it was a few Tory special advisers hanging around Silicon Roundabout, if anyone remembers that. Very niche.
Since then it has become much more mainstream – in the world of ideas, among policymakers and in think tanks, driven partly by tech money but also by ideas coming from researchers in universities. It has risen up governments’ spending priorities. Theresa May’s commitment to spend 2.4 per cent of GDP on R&D was a little bit iconic in putting quantity at the forefront. Then, as the 2010s went on, there was growing interest not just in how much government spends on R&D but in how it does it: ARIA, research on research, the metascience work I’ve led. Quality as well as quantity. And that translated into money, with a series of very good spending reviews for public R&D. That’s a big win for all of us who care about this area.
But we have to be realistic. The problems haven’t gone away. The economy isn’t growing the way it could, and in some ways the gap between the innovation economy in the UK – arguably Europe as a whole – and the US has grown. Some of the work still to do sits in classic research policy. Some of it is outside it: planning, housing, land use. But there is so much more appetite to work on this than there used to be.
R&D spending has gone up over the past decade and productivity has flatlined. Either the two aren’t linked, or things would have been even worse without the spending, or the time horizons are too long and we aren’t deploying R&D in the right way. Which is it?
I don’t have a strong research-based answer, but my intuition is a couple of things. First, things would be worse had we not spent the money. Successes like Google DeepMind haven’t transformed the UK economy, but they are material. Second, there are complementary things we need to get right, many of them outside the R&D system – planning and land use, which affect how easily businesses can set up to take advantage of R&D, and access to finance for businesses to scale, which is a big priority for the government and something I spend time on at the ESRC.
Making sure we fund R&D in the best possible way won’t solve the problem on its own, but it matters – which is why metascience and the reforms we’ve been making at UKRI help at the margin. UKRI, or any funding agency, is not going to deliver economic growth on its own. I hope I’m not selling out our comms team when I say that. But we are a vital part of a nutritious breakfast, as they say.
What UKRI is for
UKRI’s strategy lists the ways it contributes to growth: productivity gains, quality jobs, regional and national growth, innovative companies, regional inequality, new markets, living standards, competitiveness, investment, social impact – on and on. That strikes me as very hard. What is your theory of how the work of UKRI grows the economy?
The list makes it sound as if we’re trying to do more things than we are. Those are all benefits you get from successful R&D. What we actually do is a smaller number of things.
The first is that we invest in curiosity-driven research. If you’re a chemical engineer it’s the feedstock of future growth; if you’re more naturally minded it’s the wellspring. One of the great things about Patrick Vallance’s time as science minister is that he put a stake in the ground for that, time and again, and it’s now enshrined in the buckets approach. It has always been part of what the research councils and UKRI did, but it’s really good to have it set down as one element.
The second is applied R&D: working with businesses and other organisations in ways that crowd in investment, further downstream, with more proximate impacts, tied to things like the industrial strategy, which has a coordinating effect. You mentioned creating markets – the advance market commitments we’re doing on quantum computing are an example. There’s a range of tools there, from advance market commitments and challenge prizes to collaborative R&D.
The third is helping businesses start, scale and stay in the UK. That’s the core activity of Innovate UK, but it feeds right through into the research councils’ work, which is why the new UKRI strategy takes a much more integrated approach. As well as being cute because it all begins with S, it represents a real change in how Innovate UK operates. People at the university end of the audience might be less interested, but it matters: across UKRI, we’re trying to be more choiceful.
My colleague Tom Adeyoola, executive chair of Innovate UK, has an anecdote that brought this home to me. Innovate UK employs lots of skilled people to make choices about how it spends money. The R&D tax credit scheme is kind of algorithmic: you apply, and if the application passes, you get the credit. Historically, the ratio of applications to funding was the same in the two. So we had a very high-touch way of funding through Innovate UK that was funding at the same rate as a high-volume scheme. In Innovate UK’s new Velocity programme we want to be more choiceful – to apply intelligence to how we support businesses, and which ones, so that they don’t just start here, but scale and stay here. That’s a really valuable step forward.
Buckets and missions
Are the buckets still the main way UKRI thinks about the world? They appear in some strategies and not in others.
Buckets are very much part of how we operate. I’ve just been writing my papers for the next ESRC Council meeting, and I use the buckets framework to describe what we do. It’s hardwired: bucket one, curiosity-driven research, is mostly run and delegated in councils, whereas bucket two is collaborative across UKRI. The lived experience of working at UKRI is that buckets really determine how we operate.
It’s worth pausing briefly to talk about the infamous word “bucket”. A few months into the discussion, before any formal government document had been published, someone in a meeting with Patrick Vallance sheepishly suggested that buckets might not be the most inspiring term – perhaps they should be pillars, or foundation stones. Patrick basically said: we could spend all year naming these things, I’m making an arbitrary decision that they’re called buckets, and we’re not discussing it any more. So buckets it is. Not a great term, but it really does shape what we do. People who know innovation policy can always problematise it – is it too linear, don’t things slip between them? – but although it’s a simplification, it is really helpful.
UKRI’s missions programme covers everything from policing to environmental change. The language feels very Starmer-era. Did you have the internal debate about whether to keep calling them missions?
This is the R&D Missions Accelerator Programme, a half-billion-pound programme for which I’m the responsible officer. You’re right that the missions we fund were originally the missions in the Starmer government’s plan for change. But they are really enduring priorities. I’ve talked to Chris McDonald, the science minister, about them, and he is very bought in. And mission-oriented innovation certainly didn’t come from Keir Starmer’s brain, big though it is.
What’s interesting about the programme is that, for a limited number of themes with a clear government priority – reducing violence against women and girls, for example, as part of the safer streets mission – we work with government to identify quite specific challenges and targets, and then very skilled challenge directors, often from outside the research system, use a range of innovation funding mechanisms to try to achieve them. It will never be the majority of how UKRI funds things, but it’s a useful tool. And there is other mission-led work: the quantum missions programme, a big area of industrial strategy investment, is mission-oriented too and involves advance market commitments. Mission-oriented investment is one tool in our toolkit, and it isn’t going away despite the brief association of missions with the previous administration.
Devolution
The Burnham government’s big plan is devolution. UKRI isn’t set up to do devolution, mayoral combined authorities are new to research, and you have ambitions to transform the economy and society. What if mayors fund things that have nothing to do with those ambitions?
I think there’s quite a lot of compatibility, and we’ve been working closely with mayors and with No 10 North on how devolution should play out for R&D. Some programmes are already running. The Local Innovation Partnerships Fund devolves funding to mayoral strategic authorities to fund specific projects in their innovation plans. It’s up and running, it’s going to be continued at the next fiscal event, and it has been a really effective partnership. I’ve been doing a lot of work on it with West Yorkshire Combined Authority – I was in Leeds yesterday with the team working on more projects – and they’ve developed a really exciting innovation plan.
Look at what the government has been saying about its theory of growth under devolution. Dan Tomlinson, the growth minister at the Treasury, gave a really interesting speech at the West Midlands Growth Summit a few weeks ago. It’s predicated on the idea that there is valuable local knowledge about what’s happening in local economies, which helps you coordinate investment in all the complementary things you need for growth: R&D, but also skills, transport and so forth. That goes back to what we were saying earlier. R&D on its own probably won’t get you growth. If you can coordinate R&D, planning, transport and skills, you have a better stab at it.
Money, bureaucracy and AI
The Budget set out how the Department for Business, Innovation, Science and Technology will lose another one per cent of its capital funding, which will flow down to UKRI. Grant success rates are around 20 per cent, so a lot of activity goes nowhere, there’s a lot of bureaucracy and there’s less cash. How do you navigate that nightmarish triangle?
A few things. First, we’re working hard to give potential applicants and research organisations as much clear guidance as we can. In some places that means tiering the process, with shorter expressions of interest so people waste less time on applications. Partly it means working more closely with research organisations on demand management. Understandably, demand management creates work for them, but if we do it the right way and give advance notice, it’s less painful. And more generally, being more choiceful about what we do, and clearer about our future pipeline, gives people more clarity. We can’t make the pie bigger, but we can make the process of getting to it as legible and painless as possible.
Will AI transform all of this and make the conversations we’re having now look antiquated?
No one really knows, but when we think about future risks and opportunities, you have to be open to the possibility of a dramatic increase in AI capability over the next year or two that transforms how research happens in a lot of areas. We’re working hard on how we respond. We have a very big programme investing in AI research itself – a huge area of focus in bucket two. We’re looking at how we can use AI in our own operations, which has to be pursued with great care: how, without taking human judgement out of the picture, we can use it to make our application processes more efficient. That’s something we’re approaching experimentally.
More broadly, you can imagine scenarios where AI totally transforms how research is done, and then you get some profound questions about what a funding agency’s portfolio should look like. At the ESRC, about half our money goes on research grants, a quarter on data infrastructure and a quarter on PhDs and talent. Maybe that balance would change in a world of genuinely superintelligent AI. I don’t know the answer, and we’re not changing anything immediately, but we’re keeping our eyes open because things could change very rapidly.
Your questions
From John Warren in the audience: pricing policy supported by UKRI is increasing unit rates for university TRAC facilities. Has UKRI looked at whether those price changes are driving specialist analysis out of universities and into commercial suppliers?
That’s very much on my mind from an ESRC point of view, because we fund a lot of survey-based work – population studies and so on – where there is a really interesting and delicate partnership between universities and suppliers of polling services, some of them commercial. All I can say is that it’s something we’re keeping an eye on, both on TRAC and on full economic costing rates. It’s a great example of a system where we need to watch for unintended consequences. A very good question.
UKRI funds excellence wherever it is, but in a devolved landscape its governance doesn’t necessarily reflect the regions and nations – there’s no guaranteed place for them on many of your governing committees. Does that need updating?
There are still questions about how devolution will work precisely. But on what has been set out so far, in the Rewiring the State Cabinet statement, the core excellence-oriented work – bucket one, curiosity-driven research – is still expected to be a UKRI and research council function rather than devolved to mayoral strategic authorities. I haven’t asked them, but I don’t get the sense that many mayors want to become funders of core curiosity-driven research in universities either, at least for the time being. So to some extent that question isn’t in play at the moment.
Where it does raise a really interesting question for me as an executive chair is capability building. In some areas of social science, and elsewhere, applied work will disproportionately be funded in the south east of England or the golden triangle if you just go on where the quality is – because of path dependence, because that’s where it has been funded for a very long time and where the excellence has built up. As we think about the future of our applicant-led mode at the ESRC, there’s a question about whether, in strategic investments, we should selectively grow capability in other places. You have to be very cautious: you can’t magic capability out of the air, and we don’t want to be spreading the jam, to use the metaphor. But where expertise is highly concentrated, it’s a relevant question whether it’s possible and desirable to grow it gradually in other selected places.
Some incentives across UKRI seem to pull against your growth story. Infrastructure funding is struggling because universities can’t find the match funding. And when we talk about economic impact, it doesn’t matter where that impact happens. How do you align the incentives?
On infrastructure and the bigger question of university finances, we have to acknowledge it’s a challenge. University finances are in a challenging position, and the economic rents universities used to rely on – from international students, and from home fees that were more sustainable – historically subsidised research more than they do now. Through Research England in particular we work really closely with universities on that. You’ll have seen yesterday that my colleague Christopher Smith is the preferred candidate to run Research England, so there’s a strong sense of continuity there.
The other, wonkish thing is that our ability to work closely with the Office for Students, and through them with the Department for Education, has really improved in the last year or so. It’s great to be working with Edward Peck, the OfS chair, and with Ruth Hannant and Polly Payne, the newish joint chief executives. When the Higher Education and Research Act was being drafted, that link was meant to work well. It’s been challenged in the past, but it’s working better now. UKRI can’t fix the university finance issue, but we need to be involved and having those conversations to make sure that we can play our role.
On growth across the country: if you look at where UKRI R&D funding goes, we are the most regionally dispersed part of what government used to call the R&D stack. Health research through NIHR, defence research and R&D tax credits are all more concentrated. We have the great good fortune that research-intensive universities exist in many parts of the country, including places with relatively little business R&D. They are the bridgeheads for that kind of investment, and it’s through UKRI that we fund research there. So I’m optimistic from that point of view.
What’s next at ESRC
What should we be looking out for from ESRC over the next few months?
Social scientists will know we’re in the middle of refreshing our applicant-led mode – what used to be called responsive mode. We have some very long-standing staff members, and as far as we can tell this is the first proper refresh in several decades, which is really exciting. The aim is to properly support great curiosity-driven research in the social sciences. That includes a scheme called Connect, which is meant to give researchers the space to come up with the big ideas of the future; challenge awards, which are like big centre-style awards and could fund centres or other large projects; and a range of awards in between, including a very rapid-response, policy-relevant award.
We’re also running some really interesting experiments with our metascience unit capability – which is itself transformative – on how to make the application process easier for reviewers, and how to make sure it works well. Peer review is a real black box, and we want to make sure we get the best possible decisions out of the system, including asking what appropriate use of AI looks like. And there’ll be further progress on the R&D missions: how we can harness R&D to tackle violence against women and girls, help the creative industries work effectively, and improve early years opportunities – all super important to the government, to citizens and to a flourishing society. Lots to look forward to.
Stian Westlake is executive chair of the Economic and Social Research Council. James Coe is Associate Editor at Research Agenda. This interview took place at UKRI, transformed, Research Agenda’s first live session, on 11 November 2026. Watch the full session.