Written evidence submitted by Byron Hyde (UKI0007)
Introduction
1. I’m an academic at the University of Bristol and Bangor University, and am visiting Fukuoka University in Japan as a visiting scholar from April until July. I also have something of a portfolio career insofar as I have previously worked in industry while an academic and currently hold a large number of public appointments, including charity trusteeship, school governorship, and membership of various committees within the Ministry of Justice and National Health Service.
2. In this submission, I argue that UKRI benefits from setting broad, flexible strategic priorities while allowing funding to be shaped from the bottom up by the research community. While the Public Accounts Committee may be inclined to sharpen UKRI’s objectives and expand the use of commissioned calls for proposals, I urge the committee to also recognise the value of responsive mode funding, in which applicants propose research topics and justify their significance.
3. I further suggest that incorporating some degree of randomization into the review process for grant applications could reduce bureaucratic burden and cost. However, this approach introduces ethical risks that must be carefully considered by science policymakers.
4. For the purpose of this inquiry, it is worth noting that I work in the humanities and social sciences as a metascientist. Most of what I say is relevant to UKRI as a whole organization, but those with a greater understanding of the sciences may argue that some of my comments only apply to AHRC or ESRC.
Balancing Top-Down and Bottom-Up Funding Priorities
5. As a humanities scholar myself, one of my criticisms of my colleagues – about which I’ve been outspoken – is that humanities research is too often self-obsessed and doesn’t sufficiently consider the end products (i.e. social impact). I emphasize that this is a real misfortune because the humanities have a lot to contribute to policy and society: in fact, some areas of public policy have failed to be as effective as they could have been because of too great an emphasis on randomized control trials and little respect for the humanities and social sciences which could have provided essential analyses about the contexts in which those policies were deployed (conditional cash transfer is a classic example of this). Nevertheless, while many humanities scholars continue to apply for research that interests them rather than benefitting society, funders such as AHRC may benefit from a more top-down approach in which humanities researchers are directly commissioned (through detailed calls for proposals) to conduct research that contributes to social problems identified by UKRI.
6. Despite the potential merits of top-down commissioning, there are clear advantages to maintaining a robust responsive mode within UKRI’s funding structure. Policymakers and funders cannot be expected to anticipate every area of emerging or neglected importance. Many breakthroughs – particularly in the humanities and social sciences – have stemmed from research trajectories that were initially considered obscure or marginal. Bottom-up proposals foster creativity and intellectual autonomy, allowing academics to identify novel topics or reframe existing problems in ways that challenge prevailing assumptions. Moreover, responsive mode proposals typically include justification of both the academic significance and potential social impact of the research, offering an organic mechanism for surfacing under-recognised areas of relevance.
7. That said, a wholly bottom-up system risks fragmentation and opportunism. Academics are skilled rhetoricians and may overstate the social utility of their research in an effort to match perceived funding trends, cherry-picking supportive literature to bolster proposals. A coherent but flexible top-level vision is therefore essential to ensure alignment with UKRI’s strategic aims. UKRI should neither micromanage research directions nor abdicate strategic responsibility entirely. In my view, UKRI’s current vision and objectives strike an appropriate balance, providing enough direction to guide funders and applicants alike, without unduly constraining academic freedom or innovation.
8. One disadvantage of the current bottom-up approach is the administrative burden it places on applicants. Academics may invest significant time and effort into crafting proposals that, regardless of methodological rigour, are ultimately rejected due to a misalignment with the implicit interests or informal priorities of the research council. This misalignment represents an inefficiency in the system: applicants are not necessarily failing due to scientific inadequacy, but due to insufficient information about strategic fit. This contributes to the growing bureaucratisation of research culture and can deter researchers from pursuing worthwhile but unconventional lines of inquiry.
9. One possible solution – already adopted by some funders such as the Templeton Trust – is a two-stage submission process. In this model, researchers initially submit a concise proposal outlining the research area and its importance. Only if this is accepted do they proceed to the full application, including methodological detail and budgetary breakdowns. UKRI already uses such staged processes in some programmes, and expanding their use could help reduce applicant workload while enabling more targeted evaluation at an earlier stage. This approach balances responsiveness with efficiency, and would also offer a valuable opportunity for feedback, reducing the perception that grant success is arbitrary.
Using Randomization to Make Funding Decisions
10. One persistent challenge in competitive grant funding is the sheer volume of applications that meet the threshold of quality. Reviewers often remark – only half-jokingly – that the final stages of assessment involve attempting to “discern between indiscernibles.” That is, many proposals are methodologically sound, innovative, and socially relevant, but the scarcity of funds necessitates rejecting a substantial proportion of them. At this stage, reviewers must spend significant time agonising over marginal distinctions. This not only slows down the funding process and increases costs but can also introduce greater subjectivity or bias in decisions that are, by the reviewers’ own admission, close calls.
11. If it’s acknowledged that a subset of proposals are of roughly equal quality and importance – yet exceed available funding – it’s arguably more honest and efficient to use randomization to select from this pool. A hybrid approach could be employed: proposals undergo several rounds of review, and only once the “equally fundable” group is identified is random selection applied. This approach reduces unnecessary reviewer labour, avoids the illusion of fine-grained discernment, and replaces opaque subjective preferences with transparent procedural fairness. This model doesn’t imply that all funding decisions should be random, but rather that a lottery could be used as a tie-breaker among meritorious proposals, where current processes already implicitly rely on chance due to reviewer inconsistency. In this context, explicit randomization may be more ethical than decisions swayed by unconscious bias, reviewer fatigue, or institutional preference. It’s my understanding that the British Academy is currently using this method to select their small grant recipients.
12. Despite its advantages, randomization carries ethical and epistemic risks. First, the assumption of indistinguishability may be premature. It’s possible that closer scrutiny or more deliberative evaluation could identify meaningful differences between proposals, and that a lottery short-circuits this important stage. A shift toward randomization might therefore discourage efforts to improve peer review quality and rigour, or to invest in better criteria for assessing societal value or research design.
13. Randomization may also be perceived as a moral abdication. Leaving impactful funding decisions to chance could be seen as removing the responsibility of human judgment, particularly when those decisions affect careers, institutions, and knowledge trajectories. While this concern may be partly psychological – since chance already plays a large implicit role – it nonetheless highlights the importance of carefully framing and justifying any randomized approach, ensuring it is used only at a late stage and only when clearly warranted by the evidence.
Conclusion
14. This submission has argued that UKRI’s funding system should preserve and strengthen its balance between top-down strategic direction and bottom-up responsiveness. While certain fields may benefit from more directly commissioned research to address under-examined social challenges, the responsive mode remains a vital mechanism for enabling creativity, surfacing overlooked topics, and fostering academic initiative. UKRI’s existing broad strategic framework, if maintained and made more transparent, can support both modes of funding in a coherent and complementary fashion.
15. In addition, this submission has proposed a carefully limited role for randomization in funding decisions. Specifically, in cases where multiple proposals are deemed equally meritorious after rigorous review, a transparent lottery system could serve as a fairer and more efficient tie-breaker than subjective or inconsistent reviewer judgments. However, this proposal must be implemented cautiously, with attention to ethical concerns around responsibility, and only in contexts where indistinguishability among applications has been credibly established.
16. To further investigate the ideas I’ve presented here, the Public Accounts Committee could commission or encourage further inquiry into three areas: first, empirical evaluation of UKRI’s responsive mode outcomes versus commissioned calls; second, piloting and independently reviewing randomization mechanisms in funding decisions; and third, gathering cross-disciplinary views on what constitutes fairness, responsibility, and strategic alignment in grant allocation. These further steps would help ensure that any reforms are grounded in evidence, ethically justified, and appropriately tailored to the diverse research landscape UKRI supports.
May 2025