This article is more than 4 years old

Research on research is needed now more than ever

Marking the launch of a new research on research institute, James Wilsdon reflects on the challenges of making good research and development policy
This article is more than 4 years old

James Wilsdon is professor of research policy at University College London and executive director of the Research on Research Institute (RoRI)

In its increasingly disturbing – if, perhaps, electorally rewarding – efforts to distance itself from the cheese-eating surrender monkeys who preceded it in government, and now surround it on all sides of the House of Commons, one surprising thread of continuity between Boris Johnson’s government and that of Theresa May, is its enthusiasm for research and innovation.

Second only to its faith in Brexit, there’s no denying this government’s belief in the transformative potential of R&D: from Johnson’s first speech as PM on 24 July – which even in its 1700-word brevity, managed a shout out to battery technologies, blight-resistant crops and earth observation systems – to his contribution last week at the United Nations – where the references to neural interface technology, nanobots, pink eyed terminators and limbless chickens made it seem like he’d pureed the collected works of Margaret Atwood, Ray Kurzweil and Eric Drexler, drunk and vomited the contents over his baffled audience.

As with many of this government’s bolder moves, it’s also tempting to detect the hand of Dominic Cummings at play. Long before his return to Downing Street, Cummings had captured the attention of those of us working in science policy with his occasional blog musings on the organisation of the research system, and lessons for the UK from the US Apollo programme and ARPA/PARC style initiatives.

Last Wednesday, in the midst of the crisis provoked by the Supreme Court ruling and the recall of Parliament, Cummings still found time to organise and attend a No.10 seminar on the future of UK science and innovation. Chaired by universities and science minister Chris Skidmore, the two-hour session saw an assorted cast of Cabinet ministers, chief scientists and research barons pitching their hottest and shiniest ideas for a mooted flurry of fresh R&D spending announcements, originally slated for this week’s Tory conference, and now perhaps being held back for an election manifesto.

Good and bad spending

Over and above any eye-catching initiatives, the government’s totemic policy goal (inherited from Theresa May and Philip Hammond) remains boosting the R&D intensity of the UK economy from 1.6 per cent or 1.7 per cent of GDP (where it’s long hovered) to 2.4 per cent.

This doesn’t sound much – and would only drag us up to the OECD average. But as modelling by the Campaign for Science and Engineering has shown, it still represents a whopping £20 billion of extra public and £40 billion of extra private sector investment by 2027.

Leave aside for a moment how realistic these numbers are, most of us would line up in support of more cash for research. But as we saw in the US after the 2008 financial crisis led to a massive injection of short-term stimulus spending on research and other areas, there are good and bad ways of spending large amounts of extra money.

In research systems, as a rule of thumb, steady growth is far better than unexpected splurges or lurches at the whim of politicians or funders – particularly when these are tilted towards fashionable technology fads. In the US, the R&D investment tap was turned on and then off, leading to little more than an army of unemployed PhD graduates and post-docs.

Long term and predictable funding frameworks aren’t sexy, but they are massively preferable to research and innovation policy set by Davos dreamers (does anyone still believe we’re going through a fourth industrial revolution?) or pegged to Gartner’s hype cycle . It may be unfashionable to draw European parallels, but the German government’s example is instructive – earlier this year, it pledged year-on-year increases of 3 per cent in research spending up to 2027.

The Blair-Brown government’s ten year framework for science and innovation was, of course, an attempt to provide this kind of stability. But it lost power around the halfway point of that plan, underlining the difficulty of delivering on such frameworks – even when agreed upon – in the face of political instability. This challenge is far more acute now: given all the chaos, can any government develop a long-term framework with confidence that it can see it through?

So the political draw is inexorably back to the shiny and the new. As an increasingly grizzled veteran of these debates, I view the 2.4 per cent target with a mix of admiration – because it’s the right thing to do, and governments should set bold targets – and wariness, that the new money will, to use a Johnsonian verb, be “spaffed” away on wasteful, headline-grabbing new initiatives.

To that end, university research offices should perhaps be parsing the Prime Minister’s UN speech more carefully for a hint of what lies ahead. Should we open a centre for limbless chickens? Who’s our go-to professor for pink-eyed terminators?

Time to get meta

As Richard Jones and I argued in our Biomedical Bubble report last year, a persistent challenge for research and innovation policymakers – and for funders like UK Research and Innovation – is how to improve the data and evidence that underpins the choices and priorities used in allocating new investments, and in measuring and maintaining appropriate levels of balance across a funding system.

Balance between more fundamental, discovery-driven research, and that which is applied to specific challenges, or the needs of particular businesses and industrial sectors. Balance within and between disciplines, and into more novel realms of interdisciplinary inquiry. Balance across different types of institution, and between the R&D hotspots of the golden triangle, and the other nations and regions of the UK.

These challenges aren’t going away fast, but today, I’m happy to be able to share with Wonkhe’s readers a small yet positive step in addressing them. Together with the Wellcome Trust, Leiden University and the technology firm Digital Science, the University of Sheffield is launching the Research on Research Institute (RoRI).

The RoRI consortium will undertake transformative research on research, also known as meta-research. By analysing research systems and experimenting with decision and evaluation data, tools and frameworks, we aim to advance more strategic, open, diverse and inclusive research.

RoRI will be based for an initial two-year incubation phase at Wellcome’s offices in London. To date, twelve strategic partners have expressed their interest in becoming part of the consortium, including private foundations, academic research institutions, and public funding agencies from nine countries – Austria, Canada, Denmark, Germany, India, Netherlands, Switzerland, UK and the USA. We also have one pan-African partner – the African Academy of Sciences.

Together, the consortium will co-design projects and share data to inform comparative analysis of research systems and cultures, and develop, test and apply novel approaches to decision-making, prioritisation, allocation and evaluation. RoRI’s activities will include projects, a rolling programme of research seminars, as well as regular reports and working papers.

The first of these are published today – a provocative look at “21st Century Phds”, and the shocking dearth of data within the UK system on doctoral and post-doctoral outcomes and career pathways. And another think-piece on the opportunities, uncertainties and potential biases of artificial intelligence-assisted peer review.

All of RoRI’s research methodologies and findings will be made openly available.

And within RoRI, funders and academic research institutions will also be able to access and use a secure data platform, hosted and managed by Wellcome as a neutral convenor and broker. This will provide a trusted space, where funders and academic research institutions can securely share data, policies and practices for primary research and experimentation.

Research on research isn’t new – there is a wealth of excellent work to build on – but it can be poorly joined-up and applied in policies and practices. Working with our partners, who together invest several billion pounds a year in research, RoRI has a unique opportunity to co-design new methods, and experiment with creative approaches to decision making, prioritisation and evaluation.

I’ve no idea what Dominic Cummings will make of it. But I’m excited. And I hope you’ll join us @RoRInstitute, and in other ways over the months ahead.

Leave a Reply