We are a group of early career academics and postgraduate students in the humanities and social sciences writing to express our profound concern with the proposed use of impact as part of the Research Excellence Framework (REF) assessment procedures. Together, we represent a range of institutions, disciplines and approaches. But we are all united in our belief that the attempt to quantify impact as HEFCE proposes to do will have a corrosive effect on the higher education sector that we are expected to lead into the twenty first century.

We reject the proposal to measure the impact of research on the following grounds.

First, we contest the assumption underpinning the proposal that current academic work has no impact to speak of, and that social impact thus needs to be encouraged by redirecting funding. We would not have entered academe unless we believed our work would have impact impact on our peers in our respective fields of expertise and equally importantly, impact on our students, whose education is enhanced through our research efforts. These are the primary ways in which academic work has an impact on society. To suggest that the direct social effect of academic work now needs to be measured in order to incentivise greater impact suggests that research hitherto produced by UK universities has been irrelevant and useless, and that educating hundreds of thousands students counts for little.

Second, we hold that attempts to measure the impact of our work will inevitably fail to capture the breadth and depth of academic effort and achievement. All sorts of research will slip through the net in attempts to establish demonstrable effects of a particular piece of research on a specified group. This will often be for reasons that have nothing to do with the intrinsic worth of the research itself. Research may and indeed should question established conventions and consensus: this may make it unpopular. New discoveries of the greatest intellectual significance may be difficult to interpret and understand within a given time frame: they may be ignored when first published. The significance of an intellectual breakthrough may only be realised at a later point: it will be ignored in the present. Research may highlight emerging trends whose direction is difficult to understand: such research would be left unfunded. Research may simply survey the current state of knowledge in a particular field without breaking new ground: the impact of this could not be measured by impact criteria. Consider the crisis in contemporary economic theory associated with the recession. Restricting research to its measurable effect in the present will cramp precisely the kind of theoretical development, intellectual ambition and long-term focus needed to break through the impasse in todays economic thinking. Oftentimes it will only be other researchers and experts in the field who are best placed to assess the significance of research. Yet now HEFCE suggests the inclusion of non-academic research users on REF panels for the first time. The way in which research has advanced and affected society in the past would escape all of HEFCEs suggestions.

Third, we hold that attempts to measure the impact of academic work will warp the impact our work already has. Attempts to assess the impact of research will distort research priorities. As academics will inevitably come under pressure to tailor research to meet the requirements of impact, our research will end up reflecting the frameworks used to measure its impact rather than intellectually engaging with the world. Academics cannot fairly be judged on the basis of our impact on public policy or the economy, because we have little to no control over how our research gets taken up. The only way that we could be sure of making an impact on public policy would be slavishly to follow the political priorities of the government of the day. The deadening effect on open-ended research does not need to be spelled out: the academy would become an echo chamber for government. As research adapts to the needs of the impact agenda, we will not have research but only the rediscovery of the needs of research users. All attempts objectively to measure the impact of research will become futile as researchers end up simply mirroring impact guidelines.

For these reasons, we roundly reject any proposal to measure the impact of research in our disciplines. Impact will undercut the ambition and drive of young researchers. If research is to be assessed more by its impact than its intrinsic worth, then impact criteria will necessarily privilege established institutions of learning, whose high-profile reputation will automatically boost their impact. This will add to the advantage such institutions already have over less well-known institutions, thereby encouraging stagnation rather than development within the UKs academic sector. Criteria for measuring impact that include such vacuous indicators as culture and quality of life have clearly been included to pre-empt protest. But we will not be patronised. We are fully confident that our combined efforts already have extensive and far-reaching impact, not least in advancing human knowledge and understanding in the widest possible sense. HEFCEs pledge to develop clear criteria to measure impact is not convincing because if it were possible to devise clear criteria HEFCE would already have offered them. No one working in social science has yet been able to generate proof that particular pieces of research lead directly to specific social effects. Even HEFCE concedes this when the REF document speaks of the need to assess rather than measure impact and calls for narrative evidence.

Everything previously achieved by social science and the humanities in the UK has been done without measuring its impact. Instead of HEFCEs embarrassed terminology and unconvincing suggestions we propose that the best way to advance UK higher education is to give academics as a profession and universities as institutions the widest possible autonomy from narrowly instrumental criteria and utilitarian means of assessment.