by Adam Forristal Golberg, Strategic Research Development Manager (Research Growth) at the University of Nottingham. He’s also co-champion for the Association of Research Managers and Administrators (ARMA) special interest group on Research Development. This article is written in a personal capacity.
Research grant application success rates are too low. We need the pressures of competition to drive up quality, and to give funding panels options in terms of supporting projects across the range of eligible topics and disciplines. But that pressure is becoming unsustainable. The effort wasted in writing, processing, and reviewing unsuccessful proposals is now far beyond the cost of doing business. A rising tide of applications, which have almost doubled in the past seven years, means that funding rates at UK Research and Innovation (UKRI) have now dipped below 20%; at the European Research Council application success rates are even lower, at just 14%; in the US, National Institutes of Health a 20% success rate of R01 grants over the past decade plunged to 13% in the first year of the new Trump administration.
The best solution would be more funding for research, but there’s no sign of the Magic Funding Fairy. Worse still, here in the UK the universities’ omni-crisis shows no signs of abating, while Trump continues to slash US research funding and torpedo the global economic outlook. I think it’s time for the research community to look at our own practices to see what we can do.
Eliminating the Bottom 20%
What would happen if we could prevent the least competitive 20% of grant applications from being submitted? What would happen if research institutions and funders alike adopted this as a serious, collective ambition and took positive, concerted steps to make it happen? And did this in a way that was agnostic about institutional type, and put EDI considerations front and centre?
Consider a scheme with a success rate of about 20%. Of the 80% unfunded applications, how many might have been funded by another panel on another day? Or might have scored better with different reviewers? I honestly don’t know… maybe a further 20%? Based on my limited review panel experience it’s probably less than that.
For those competitive but unsuccessful bids, the main feedback should be: we ran out of cash and marginal decisions went against you. Like VAR in offside decisions in football, tiny margins can have huge ramifications. The submission of these honourable failures was entirely appropriate, and the process would have been the poorer for their absence. They gave the panel options. It’s likely the panel wasn’t just thinking about quality, but also balance across the range of disciplines (or, for themed calls, the range of topics to cover). Let’s call these projects silver medallists.
What about the applications that were never in the running for a medal? I think it’s safe to say that the writing, submission, reviewing and processing of the bottom 20% of proposals to any given call were not in anyone’s intrinsic interest because they never had a realistic chance of being funded. They were not the best use of anyone’s time. And that very much includes the applicants. This isn’t necessarily because they’re poor science – it could just be the wrong call, or that their submissions were underdeveloped or poorly articulated.
While I’m not wedded to a cull threshold of 20%, it feels like enough to make an appreciable difference to applicant and reviewer burden without risking cutting into the competitive proposals. We want to leave space for risky ‘marmite’-style projects that the panel will either love or hate.
How would such a cull work in practice? I would like to suggest that universities and funders both have roles to play.
What can universities do?
The perennial complaint from funders is that universities are submitting too many uncompetitive applications. That’s true, but it’s a challenge for any university to mark its own homework – all its experts have a glaring conflict of interest. The other problem with expecting universities to run ‘demand management’ schemes is that it just relocates effort from one part of the system to another, something the Tickell Review of research bureaucracy was anxious to avoid.
Nevertheless, expecting a university to identify and stop more of the bottom-20%-ers is not unreasonable or unachievable, given that their submission is neither in their institutional interest nor the interests of the applicants. With proper resourcing, appropriate senior support, and willingness to actively manage and monitor, this becomes doable. No doubt some uncompetitive applications will slip through, but any improvement is welcome.
Focus advice and support on scheme selection. Researchers inevitably read funding calls through the prism of their own interests. A sense check from a research development manager can save wasted effort. Some weeks, my biggest contribution has been talking a senior academic out of an application that has no realistic chance of success. An hour of my time saves weeks of hers. A lot of this already happens, but there’s scope for more. There would be more resources to focus on honing competitive proposals if we did a better job of identifying and redirecting uncompetitive proposals early.
Universities need to think carefully how they frame this support, and researchers how they perceive it. Too often it’s seen or experienced as a hoop to be jumped through (your ‘demand’ must be ‘managed’), rather than a supportive service to minimise wasted effort and maximise chances of success.
Stop colluding with uncompetitive proposals. Resources would be less stretched if we didn’t collude in supporting applications that we know to be uncompetitive. Difficult conversations are… difficult, but whether that’s because of political sensitivities or demanding characters or not wanting to hurt feelings, they need to be had. If it’s not in the applicant’s interests to waste their time, it ought to be possible to persuade them of this. But it’s tough, especially for early career fellowship applications that we know or suspect to be uncompetitive but represent the Last or Penultimate Chance Saloon for an early career researcher who wants an academic career.
Minimise perverse incentives for application spam. The submission of the bottom 20% is in absolutely no-one’s intrinsic interest. Unfortunately, the extrinsic pressure to be seen to be applying for funding – any funding – can influence behaviour. Unsuccessful? Oh well. See? I tried? It’s basically a lottery anyway.
Some institutions wrongly conflate being grant active with being research active. Some use grant activity inappropriately in promotion or redundancy processes and in doing so create perverse incentives. Often it’s because universities are over-reliant on external funding as an indicator of quality. I’d argue that virtually no-one gets funded who doesn’t deserve it, but that doesn’t help those who aren’t funded, especially those who were worthy contenders. Could the award of ‘silver medals’ (of which more below) at least partially address this for recipients?
What can funders do?
The perennial complaint from researchers and research offices is that funders aren’t giving enough feedback. Researchers can’t do any better if they don’t know the issue, and people like me can’t help them if we don’t know either. Funders always respond that they don’t have the capacity, and to be fair, it’s far from a straightforward task. But they might have at least some extra capacity if they don’t have to process and review the bottom 20%. Here are some suggestions for what they might do with that extra capacity:
Tell applicants (and their institutions) if they missed by a millimetre or a mile. Perhaps introduce a medal system. The best unfunded projects get a medal icon along with any feedback. Gold = funded. Silver means high quality and near the funding threshold at which the money runs out. Bronze proposals are high quality, but too far from the funding threshold.
What difference does this make? I don’t know how much consolation it would be to get a metaphorical silver medal but no funding. Would that just be more frustrating? But a silver medal or similar could be used as a credible quality indicator for promotion or recruitment purposes and provide useful feedback to the researcher’s university. Yes, we missed out on funding, but also, yes, we are on the right track. And yes, this application was outstanding and is a credit to everyone involved. It’s a metric that might be usefully tracked. The Wellcome Trust already shares the quartile ranking with applicants to their Discovery and thematic calls, so it can be done.
Silver medal outcomes could and should feature in academic CVs as an external measure of quality. Funders who permit resubmissions could also use medal designations to permit/deny that opportunity.
Initially I was going to propose a signifier or indicator (not a ‘wooden spoon’) for a proposal in the bottom 20% or otherwise a long way from being competitive. However, I have worries about the impact on researcher wellbeing. It’s bad enough having a funding proposal rejected; it’s worse being told that it’s one of the weakest received in the funding round. I also have worries about the potential for misuse by universities – institutional failings should not compound individual setbacks. It’s hard not to be reminded of the Professor Stefan Grimm tragedy.
Perhaps funders could report on numbers in the bottom 20% in the aggregate to institutions in an anonymised way, though this would be less useful. Previously some research councils have provided institutional level success rates, but as those didn’t distinguish between silver medals and bottom-20%-ers, they’re of limited value. Not getting funded is not a mark of failure or shame.
Tell us more about what’s competitive. It would be great to get a detailed report from funders about decisions made. What were the characteristics of funded projects? Of competitive projects? What were the major issues with uncompetitive projects? What percentage of projects were in with a realistic shot of funding? We need much greater transparency about success rates, and what it takes to be competitive. I’ve assumed that the bottom 20% have no chance and that there’s a long tail, but that’s not necessarily the case for all schemes. We simply don’t know.
Write better call guidance. There’s a lot more that many funders could do in terms of defining the call’s scope and what will or what won’t be competitive. The better written calls have ‘out of scope’ sections. These are useful for research managers like me, because we have chapter and verse to refer to when we tell researchers about our concerns about fit and competitiveness. But even these could be expanded and developed.
Towards a virtuous circle?
A concerted 20% campaign could create a virtuous circle effect. Fewer applications, greater funder capacity, better feedback/guidance/call reports. Fewer peer review requests, better peer review responses. Better informed research offices, more empowered to give better advice and filter proposals more effectively. Less wasted academic time. Better work-life balance? We can always hope!
Funders awarding quality/esteem indicators to the best unsuccessful proposals (e.g. silver medals) might at least mitigate some perverse incentives if used correctly by universities and researchers. If knowledge and acceptance of these quality indicators were to become widespread, they could start appearing on more CVs and play a role in recruitment/promotion decisions. It would also help to normalise the fact that assessing funding proposals is a difficult task that, with the best will in the world, is never executed perfectly. Extremely high-quality research doesn’t always get funded.
I’ve deliberately tried not to say too much about AI, the obvious elephant in the room, which is increasingly being adopted by applicants and funders (I understand that the UK Metascience Unit is already looking at AI and peer review). I’m no expert,but I could imagine a funder using an LLM with appropriate safeguards (e.g. active consent for training data, secure environment etc) to try to identify very uncompetitive bids. It might flag bids to a programme officer, who would then make a decision (with the panel chair or perhaps a panel subset) on whether to allow the proposal to proceed to full assessment. I’d like to see a high bar before a proposal could be rejected in this way. This approach would obviously suit more established and predictable schemes than one-off calls where there isn’t the time to build up training data. Clearly AI needs to earn applicant confidence before it has social licence to operate, so some shadow exercises would need to be piloted to start with.
I don’t have all the answers, or thought through all the challenges or unforeseen consequences. But I think a shared, collective endeavour to reduce the number of grant applications by 20%, starting with the least competitive, could reduce wasted effort and free up capacity across the whole system to do something more productive.
Adam blogs at www.socialscienceresearchfunding.co.uk and is on Bluesky @cash4questions.
In principle, a sensible idea. It relies on applicants and research organisations doing a little extra, being transparent, and acting purposefully.
I worked with UKRI for some years. A well known ECR fellowship programme attracted >1000 applications per round, leading to falling success rates, after increased administrative burden including enormous numbers of potential peer reviewers being contacted by the under-pressure UKRI team. Despite repeated requests (pleas?) from the programme Director to PVCs in the most research active, prestigious HEIs to do as you suggest here, the applications still rose in volume. The HEI in question was submitting several 10s of large scale fellowship proposals each round. Their success rate was poor. UKRI introduced formulaic demand management to address this at the sector level.
The explanation? The top universities have a ready supply of talented and ambitious people, all somewhat dependent on securing funding for career progression. These fellowships are career-changing. The host organisation accepts and understands the attrition at every level - success for them was driven by mass action.