Increasing institutional diversity in the social sciences
“University social scientists won. Since the early 1970s, [the ESRC] has not thought hard about the conditions within which knowledge is best produced.”
David Walker, Exaggerated Claims (2016)
What sorts of organisations receive funding to do social science research? The ESRC’s funding in practice goes to two types of bodies: (i) universities and (ii) a small number of ‘independent research organisations’ or IROs, most prominently the Institute for Fiscal Studies and the National Centre for Social Research. The first category represents over 95% of the research funding we provide. Universities are not the only game in town, but they are not far off.
This is an extreme version of a more general idiosyncrasy of UK public research funding: to quote Richard Jones, “the UK is an international outlier in the degree to which non-business R&D is concentrated in universities”. To put it another way, our publicly funded R&D system seems to have less institutional diversity, in terms of the types of organisations we fund to undertake research.
This is partly a result of our own rules and processes, which have accumulated over time.
First of all, we limit the ability of non-university organisations to apply for funding in the first place. Currently, very few organisations that are not universities are allowed to apply for funding from ESRC or any other research council. Merely to be allowed to apply for competitive funding, you need to be officially accredited as an “Independent Research Organisation” by the Government.
This is not an easy process. Having taken an organisation through the accreditation process for IRO status twice (once unsuccessfully, once successfully), I can attest it is uncertain, opaque and time-consuming. And, at least in my experience, the process seems to have a strong presumption in favour of rejection. The unstated aim seems to be to ration IRO status and to manage downside risk, rather than to increase the pool of organisations able to undertake research for the public. (And of course, becoming an IRO does not in itself lead to any funding: it merely allows the IRO to apply to the same highly competitive contests to which universities are automatically allowed to apply.)
Secondly, the UK has a second stream of research funding that only universities can benefit from: quality-related (QR) funding (from Research England, based on the Research Excellence Framework). Organisations that are not universities (“eligible higher education providers”, according to s39 of the 2017 Higher Education and Research Act, which also excludes so-called “private” universities) have no equivalent of QR funding. Since research grants typically provide 80% of the “full economic cost” of undertaking a project, and expect the remainder to be topped up from QR (or other surpluses, such as those from international student fees), this leaves many IROs in a tricky financial position. (For some IROs in fields beyond the social sciences, this difference is made up by other funding streams, including core funding from research councils or from the Government; but this is not currently the case for most social science IROs.) ESRC could in theory set up its own version of a QR-like funding stream to support non-university organisations (at least, this would be within the provisions of the 2017 Act that governs us), but currently no such scheme exists*.
It seems pretty clear to me that this state of affairs is not conducive to institutional diversity. Let’s move on to a second question: does that matter?
One line of argument would say that it doesn’t matter: the UK’s universities are big and varied organisations, and perhaps can provide all the diversity we need in a research ecosystem. By this logic, we should not fixate on the type of organisation delivering a research grant, but on the specifications for the grant. Proponents of this view might draw attention to applied research organisations in the social sciences and elsewhere that are based in universities, such as the Economics Observatory at Bristol, UK in a Changing Europe at King’s, or the Bennett Institutes for Applied Data Science at Oxford or for Public Policy at Cambridge.
I am not fully convinced.
When I look (admittedly as an outsider) at the culture of the IFS and the incentives on IFS researchers (many of whom previously worked in universities and go on to university jobs), I see a number of factors that might help explain why the organisation’s work is unusually impactful. Some of those things, it seems, are products of the Institute’s institutional form: the existence of an editorial line, the timeframes to which researchers work, and the internal systems of mentorship and advancement.
Or consider another social science research organisation that does not currently have official IRO status and is hence ineligible to apply for research council funding, the Institute for Government. The IfG’s staff (often former and future civil servants) and modes of publication and quality control would sit awkwardly in most universities. This was also my experience of running a non-university applied research organisation in my time at Nesta - our team had markedly different incentives, skills and background from the university researchers we worked with (and in some cases competed with for funding), and these differences meant our work was different in a useful way.
Other established research funders are less precious about the types of organisations they fund. The Nuffield Foundation funds both academic research organisations and credible non-academic research organisations, and I have never heard anyone complain that this leads to lower quality work or the foundation’s money being irresponsibly spent. Indeed, speaking to the Foundation about one of their recent high-profile projects, the Economy 2030 Commission, suggests that their ability to fund a collaboration between a think-tank and an academic research centre made a significant difference to the success of the project.
The case for institutional diversity also chimes with wider currents in research policy. There is a growing interest in “focused research organisations” and other novel organisational structures, sitting outside the traditional university system, from the Arc Institute to ARIA to Eric Gilliam’s investigations into the role of contract research organisations like Bolt, Beranek and Newman in the early successes of ARPA. Geoff Mulgan recently made an eloquent case for more “exploratory social science”, which he implies sits uneasily in most modern universities. And my impression of the university-based centres I mentioned earlier is that they are for the most part outliers, operating despite the systemic incentives rather than because of them. In any case, even if it is possible, with great effort, to recreate the incentives of IROs within universities, why not take the simpler route of having a system less skewed against IROs?
Indeed, I wonder if the success of the IFS obscures quite how inhospitable our social science funding system is for IROs. That the IFS was funded by the ESRC at all seems rather fortuitous. The story I am told is that the Institute, in a tight spot financially, approached David Sainsbury for funding in 1990, and were told that David would fund as long as ESRC co-funded. ESRC said yes - perhaps because governance of such things was somewhat looser three decades ago.
Various other non-university social research organisations have sprung up in the UK since then, including the Institute for Government, Centre for Cities, the Education Endowment Fund, New Economy Manchester, and the Resolution Foundation. None were set up with ESRC funding. With the exception of the EEF (which I understand was funded by Michael Gove out of DfE underspends after ESRC was unable to fund it) and New Economy Manchester (part-funded by Nesta), they were funded by philanthropists. While the UK is fortunate to have people like David Sainsbury and Clive Cowdery willing to fund policy-relevant research, it leaves me wondering what worthwhile social science institutions might exist today under a funding climate more supportive of institutional diversity.
So what might a pro-institutional diversity funding model look like? Here are three things we could consider:
First of all, we would be more supportive of applications for IRO status, working on the presumption that having more organisations able to apply to undertake research is a good thing.
Secondly, we would give more thought to the “metascience” of IROs, seeking to understand what we can learn about how they work and what lessons this holds for the wider system, and in institutional gaps there might be in the social research ecosystem. In David Walker’s words, we should “[think] hard about the conditions within which knowledge is best produced”, from an organisational point of view. Should, for example, there be an IRO that sits alongside the What Works Centres (perhaps as an incarnation of the What Works Network) allowing these organisations greater scope to apply for research funding?
Finally, we should consider about what steps ESRC should take to support a wider range of institutions, and in particular whether a second stream of funding should be created, along the lines of QR or the schemes used by other research councils to support non-university organisations that play a distinctive role in the social science ecosystem.
As we enter the ESRC’s 60th year, revisiting these questions seems especially timely.
*One concession to this is that the ESRC provides grant funding to four selected IROs (including the IFS and the National Centre for Social Research) at 86% of FEC, rather than the usual 80% - which seems to be an implicit response to this issue, but not a very satisfactory solution.