Research impact and the denial of political responsibility

By John Heathershaw

“It’s not a matter of policy, it’s a matter of fact”

This quote is attributed to UK Health Secretary Matt Hancock on the Today programme on 5 May, 2020, in response to the questioning of social distancing rules for the over-70s. But it may just as easily have been attributed to any one of dozens of government ministers and their most senior scientific advisors in recent weeks with respect to testing, PPE, lockdown rules, or any other matters where the government has sought scientific advice. ‘We are led by the science’ is repeated ad nauseam.

While there is a consensus in government on this mantra, in academia the picture is more complicated. A fearsome debate is raging. This is evident in the furore surrounding the UK government’s Scientific Advisory Group for Emergencies (SAGE). A former scientific advisor has set up a shadow group to meet in public, the minutes of official SAGE meetings have been published in heavily-redacted form, and an increasing number of SAGE members have disclosed their identity and taken exception to government policy in public.

The focus of this debate is that of transparency. The critics in science, the media and the political opposition seem to agree that more disclosure will restore trust in the independence of scientific advice. However, what if the problem is neither one of transparency nor even of independence but one of accountability?

This question requires some historical perspective. In the 1980s government funding of research was transformed after the Rothschild Review with an important new demand: demonstrable ‘public benefit’. This move, while commonsensical, established a particular and problematic relationship of accountability — from the academy to the government — in the form of research impact.

Research Impact (hereafter Impact) was formally inaugurated into immediate controversy with the first Research Evaluation Framework (REF) of 2014 which sought to assess, for the first time, “an effect on, change or benefit to the economy, society, culture, public policy or services, health, the environment or quality of life, beyond academia“.

The relationship between science and government in the Covid-19 response may be understood through the REF Impact debate. Impact has always been mired in controversy. Radical critics see Impact as creating an army of unwitting academic servants sacrificing basic research to the demands of short-term problem solving. More constructive voices point to the “heroic” narratives of REF Impact Case Studies (ICSs) in favour of the solo, senior and typically male academic identifying the problem and providing the solution.

The reality of the relationship between policy and evidence is more complex. Two fundamental problems arise from the notion of ‘being led by the science’.

The first problem with ‘being led by the science’ is that the idea of a fixed set of facts is a profound misrepresentation science, particularly with regard to a new natural phenomenon. There are few if any incontrovertible facts in science. Science is driven by controversy, by thesis and counter-thesis. New and complex problems can be cut in many different ways, according to many different models. Scientific consensus on the best approaches takes time to develop. No amount of evidence or transparency avoids the problem of having to decide which combination of models and evidence to place most weight upon.

A second problem is that scientific facts don’t emerge in vacuums but in social environments where they are subject to prevailing wisdoms and political interests. The Observer Effect means that it is not possible to observe something without changing it. Whether or not Dominic Cummings sits in on SAGE meetings, the scientific facts which are collected and the advice which is produced are, to some extent, standpoint-dependent. This much is clear from the “facts” of behavioural scientists and the advice of SAGE in March that a strong lockdown would not be respected by the British public — facts and advice which may have led to 10,000s of excess deaths.

These two epistemological problems are both widely acknowledged by observers of the Covid-19 controversy. However, fundamental though they are, they are in some ways dwarfed by two more basic political issues.

First, which facts and advice triumph is a matter of who is heard. Politicians are masters at selecting their kind of intellectual. There is no evidence to suggest this problem is entirely overcome if civil servants oversee the selection. Bearers of radical ideas are frequently dismissed as unconstructive. More moderate voices which offer win-win solutions are favoured. Daniel Drezner frames this distinction as one between public intellectuals, who often demand structural changes, and Thought-leaders, who present win-win solutions.

Second, as Anand Ghiradaradas argues, thought-leaders are chosen to provide an important function for their political masters: the evasion of responsibility. Thought-leaders within government, such as the Chief Scientific Officer, will repeat government mantras about being led by the science. They will repeat that their advice is independent and even (sometimes) transparent and open to scrutiny. But if they claim, even implicitly, that science speaks with one voice they are effectively conspiring in the denial of political responsibility.

Unfortunately, natural and physical scientists are not necessarily trained to spot the ways in which they are being co-opted for political purposes. Some may be aware of this and break ranks, others not. But ultimate responsibility for government policy, including the current debacle in relations between ministers and scientists, does not belong to them but to our democratically-elected politicians.

This is not a new problem but it is one that appears to have worsened in recent decades under the ‘regulatory state’ whose ministers behave like management consultants. Examples from foreign and domestic policy, Labour and Conservative governments, are abundant. Tony Blair insisted that it was independent evidence that required the Iraq War. David Cameron claimed there was no economic alternative to austerity, evoking Thatcher’s claims about the efficiency of the market.

This is not a problem unique to the UK but it appears to be particularly acute here. The combination of a highly centralised political system and a regulatory state creates both immense responsibility and meagre capacity. No other state appears to have quite this heady mix. But, as an emerging debate on the ‘end of democracy’ suggests, the UK may merely be a particular case the entwining of aspects of technocracy and autocracy with the institutions of democracy across the world.

From the point of view of research impact, this analysis suggests a rethink is required. The purpose of social scientific research is not to provide evidence behind which government can obscure its political choices. It is to lay out the array of possible frames and evidence bases on every possible question – natural, physical, social, cultural. We cannot expect government’s ministers-come-management-consultants to begin accepting responsibility for how they use our research, but we can call this behaviour out and lay down some alternatives. The impact of old-school public intellectuals may be greater than that of the co-opted Thought Leaders. Then, the question for the delayed REF 2021 will be how to measure our real impact on government.

John Heathershaw is Professor of International Relations and Director of Impact, Department of Politics, University of Exeter. The views expressed here are his own.