Academic interest in “social relevance”, “public value” and “non-academic impact” is as old as the political and social sciences. Many of these disciplines were forged precisely out of a commitment to using scientific methods and insights to drive positive social change.

But the rise of the “impact agenda” about 15 years ago, and especially the incorporation of impact into the 2014 research excellence framework, marked a sea change in such concepts’ prominence across the disciplines. Next week’s completion of the 2021 REF – in which the weighting of impact has risen from 20 per cent to 25 per cent of overall scores – is an opportune moment to take stock of how academic practice, culture and structures have evolved during the impact era.

As a political scientist who has operated at the intersection between research and policy for my entire career, I see the changes as mostly positive. My sense is that academic culture has become far less resistant to the imposition of impact as an explicit requirement when thinking about or conducting research. Early career researchers are often particularly enthusiastic about “engaging with multiple audiences in multiple ways” – to borrow a phrase from the sociologist Michael Burawoy.

The framing of discussions about impact has also changed. Instead of focusing on the role of academics, higher education or universities, conversations generally revolve around facilitating mobility across the research, development and innovation ecosystem – that is, across traditional disciplinary, organisational and professional boundaries.

̽����Ƶ

This has prompted funders to recognise that their modern role is not just to fund research but also to build research infrastructure. They are now engaged in nurturing skills, and , not merely of academics but also of research support staff and the who are often key members of research teams. Funders are also investing in synthetic research, translational capacities, and the creation of new boundary-spanning platforms or opportunities. Anyone who doubts this drift would do well to read UK Research and Innovation’s .

It is surely going too far to accuse people who produce REF-related impact case studies of having “cravings for acceptance” and being “ensnared by an infatuation with their self-image”, as Richard Watermeyer, professor of higher education at the University of Bristol, reports many scholars do in his 2019 book Competitive Accountability in Public Life. As a former national “impact champion” for the Economic and Social Research Council, I’m guessing that my status as a posing, greedy, self-serving opportunist is beyond dispute – but could it be that hidden within such hyperbole lurks a deeper issue that warrants a more balanced discussion?

̽����Ƶ

One of the key shifts that has occurred within the research funding landscape, within and beyond the UK, is that policy engagement has come to be seen as, by definition, a good thing – as has nurturing forms of co-production and co-design of research. But such fixed ideas raise issues of academic independence, criticality and control.

Noam Chomsky’s work on “”, published in 1967, distinguishes between “technocratic and policy-orientated intellectuals” and “value-orientated intellectuals”. The former are “the good guys” in the eyes of the establishment, serving the needs of the system; while the latter are “the bad guys”, who dare speak truth to power, expose lies and engage in critical analysis. My 25 years of working in Whitehall and Westminster have taught me that, in reality, “technocratic and policy-orientated intellectuals” can exert considerable critical influence on politicians; being engaged and policy-focused is not necessarily the same as being passive. The flip side is that those intellectuals who heckle from the sidelines but refuse to engage are themselves very often impotent by choice – like political parties who hold their principles so purely that they never win the power required to influence anything.

Still, the arm in the traditional “arm’s-length” relationship between ministers and research funding agencies has in recent years become significantly shorter. This is not a criticism, just a statement of fact. And research funding is increasingly linked to state-directed societal challenges that require academics to work within a specific idiom, and ideally through forms of co-creation with potential research users. This is where Chomsky’s distinction and even Watermeyer’s concerns about impact begin to gain traction. The notion of relevance risks mutating towards forms of deference, co-option and control.

Balancing engagement and criticality – or autonomy and control – has emerged as the deep story when it comes to rethinking policy impact. We need to maintain a healthy balance between “technocratic and policy-orientated intellectuals” and “value-orientated” scholars. Disruptive disengagement from policymaking is in some ways likely to be extremely healthy, in both intellectual and democratic terms. But encouraging, nurturing and supporting scholars who are both value-orientated and policy-engaged is probably even more important.

̽����Ƶ

This demands rethinking the role, limits and paradoxes of policy engagement – alongside a broader conversation about how a commitment to relevance can retain a criticality that protects it from becoming a synonym for deference.

Matthew Flinders is professor of politics at the University of Sheffield. He is also vice-president of the Political Studies Association, chair of the Universities Policy Engagement Network and a professorial civic fellow at the Institute for Community Studies at the Young Foundation.