The evidence suggests that funding is divided between too few researchers, says Kaare Aagaard
The funding system is the single most important element in defining the scope, content and direction of public research. How we allocate resources has a fundamental influence on how the science system works.
A key question here concerns the balance between concentrating funds and dispersing them. Does giving a large proportion of funding to a relatively small number of elite scientists yield the most value for money? Or is progress better supported by allocating resources in smaller portions to more teams and people?
The answer to this question is neither simple nor clear-cut. But there are compelling reasons to give it much closer attention, and to ensure priorities are made with open eyes and with trade-offs taken into account.
Recent research suggests there is a widespread trend towards concentrating funds at the levels of individuals and groups. This has, for instance, been shown for funding from the US National Institutes of Health, for a broad range of fields in Canada, and for the UK Engineering and Physical Sciences Research Council. However, the evidence is still scattered, limited to a small set of countries, disciplines and funders.
Winner takes all
To get a fuller picture of the degree of concentration within one national system, my colleagues and I analysed close to 20,000 grants allocated by 15 of the largest public, private and non-profit Danish research funding foundations over 2004-2016. In all, close to 53 billion kroner (€7 billion) was allocated to nearly 7,500 grantees during this period.
Our analysis shows 20 per cent of Denmark’s researchers received almost 90 per cent of its allocated grant funding.
Ranking recipients by the amount of grant funding received reveals a similar picture. The top 100 researchers receive grants totalling just under 100 million kroner each (see figure). The group from 101-500 averaged just under 30 million kroner per person. Beyond the first 2,000 grantees, grant income per researcher is tiny or non-existent.
The figure also shows that women constitute only 16 and 15 per cent, respectively, of the two groups receiving most grant income. The proportion of women increases as the average total grant amount decreases, showing that funding concentration has a gender bias.
To find this degree of concentration in Denmark is remarkable, as the Danish system has historically been seen as egalitarian. This suggests we might find even higher levels of concentration in other countries.
So, although we lack a complete picture of trends towards funding concentration in science worldwide, there seems a need for a thorough examination of the underlying evidence as well as of the potential consequences of this development.
Whether concentrated funding leads to better research has been discussed for decades. This discussion has, however, seldom been based on solid evidence.
Diminishing returns
In a recently published literature review, we examined the scholarly work on this question. The results are not clear-cut, and many studies suffer from conceptual, terminological and methodological shortcomings. Even so, there are convincing arguments for avoiding high degrees of concentration.
Several empirical studies, for example, have found that on average there is a declining marginal return on each euro invested in research above a certain threshold. This threshold varies across disciplinary and national boundaries, but it is not, in general, very high. For instance, data on 2,938 grants from the US National Institute of General Medical Sciences show the research output and average journal-impact factor per lab decreases with funding above around $750,000 per year, and that funding above $250,000–300,000 is associated with only modest increases in research performance.
From a more theoretical perspective, many argue that supporting diversity rather than concentration nurtures more dynamic and adaptable research ecologies, creating the best conditions for breakthroughs and originality. This resonates with arguments by historians, philosophers and sociologists of science highlighting that scientific advancement is best promoted by ensuring competition between a diverse set of ideas, paradigms, theories, methods and approaches.
However, reducing the issue to one of evidence for or against funding concentration would be simplistic. A healthy research ecology comprises large and small groups. There also seems to be a lower limit of concentration below which funding becomes too thinly spread.
The optimal balance is a matter of degree. Both too little and too much concentration appear to be inefficient in economic and epistemic terms. Nonetheless, overall the reviewed literature presents a fairly strong case against high levels of funding concentration.
Unintended consequences
These findings have important policy implications, as they question the rationale behind current funding trends and may point towards more efficient ways to allocate resources. Even without an exact idea of the optimal balance, there are signs that in most countries and most fields funding should be less concentrated.
While policymakers worry about spreading funding too thinly, and while the differences in talent and originality among researchers certainly justify some degree of selectivity, there are reasons to believe that in most systems funding is currently too concentrated, and that this may harm the overall progress of science.
To fix these shortcomings, we need to understand what has caused them. In fact, concentration is the result of a number of interacting causes reinforcing each other.
On the one hand, con-centration of funds can be seen as a result of conscious and deliberate policy choices. Larger grants, a focus on excellence, a desire to create critical mass and economies of scale, and concerns for international competitiveness all increase concentration.
But such conscious actions are reinforced by other, less obvious, factors. First, the research system is characterised by so-called Matthew effects, where already established researchers, because of their visibility and track records, enjoy advantages in competing for prestige, funds and positions.
Unintended funding con-centration is also likely to occur when different funding agencies arrive at similar definitions of excellence, and so use similar criteria in their decision-making. Ignorance of allocation decisions made elsewhere in the system will exacerbate this issue.
Hence, even if each grant decision in isolation is sound, the systemic effects may be undesirable. Lack of coordination within and across granting bodies may result in higher degrees of concen-tration than any single funder is aiming for.
This combination of over-lapping goals and lack of oversight or coordination seems widespread in most funding systems. Add fierce competition, large grants and low success rates, and you have a recipe for increased concentration.
Potential remedies
It seems, then, that patterns of concentration are shaped by interconnected factors, within and outside science, with intended and unintended effects. If correct, this diagnosis points towards several potential remedies, which probably need to be combined to bring about real changes.
First, there is a clear need for better oversight in and across research funders to ensure allocation decisions are made more from a portfolio perspective and less on isolated assessments of individual applications.
Second, there is a need to experiment with new funding mechanisms that seek to counter the current system’s concentration bias. Among the more radical proposals is to use a modified lottery model to choose among grant applicants who pass through an initial quality screening.
Others have suggested experimenting with funding instruments that promote risk taking and diversity, for instance by fully anonymising the review process. Still others highlight the need to use a broader variety of assessment criteria, and to reform assessment and reward criteria at the university level.
These are just a few of many possibilities. It is not clear what will work best, but it is clear that there is an urgent need to start the discussions.
It is also clear that any such changes will meet strong opposition from within the system. For example, in 2017 the US National Institutes of Health backed down on a plan to cap grants to big labs after protests from sections of the research community. Real change will require political will and strong university leadership, but most of all courage from both public and private funders.
This article also appeared in Research Europe