[Versión en castellano]

It seems a long time since John Ziman’s label ‘steady state science’ neatly captured the fact that a tipping point had been reached, where demand for research resources would evermore outstrip their supply. From this point decisions about how to allocate research resources among competing claims would become one of the central tasks of research management and policy.

Scientific quality would no longer be sufficient, in itself, to justify financial support. Scientific projects would have to be ranked on merit. The best or the most important science would be funded, the rest would not. Grant application success rates fell. Competitive markets became the analogy for research funding, coupled with new authority structures to closely oversee how these scarce public resources were spent. The advent of the financial crisis from 2008 led to slowdowns in the growth of science funding in many countries and real reductions in others, placing further pressure on allocation mechanisms.

The underlying logic is that with finite resource availability only the highest quality science should continue to be done, but it would be done properly. Yes, researchers and funders engage in a ritual seduction dance: scientists’ pad their budgets slightly in anticipation of cuts; funders trim budgets slightly suspecting they’ve been padded. But the core principle of sufficient funding for top quality science prevails.
However, there is more than one way to cut a cake. And this is precisely what has been occurring with science funding.
An emerging approach in some systems and programmes has been to maintain higher research grant success rates, through the very simple practice of systematically under-funding proposals. In Spain for example, this café para todos (coffee for everyone) system of allocation has become an expected outcome of funding in the country’s most prestigious competitive science funding programme. Applicants expect to have their budgets cut, often viciously. Budgets for PhD students or post-docs are frequently denied, systematically reducing research capacity, compromising career pathways for young researchers and the sustainability of the system as a whole.

Café para todos, at least in Spain, draws its strength from being a culturally embedded way of coping with scarcity of resources in which no one is totally deprived and almost everyone can survive. As a social philosophy it has much to recommend it. As a principle for allocating scientific research funding, however, does it stand up to scrutiny?

On the one hand, broad distribution of research funding makes it possible for more scientists to do some research and therefore to maintain their identity as scientists. This is vitally important. However, this identity can also be undermined if insufficient resources force scientists to work below their preferred quality standards.

On the other hand, there is some evidence suggesting there are reasons to be concerned. Systematic under-funding of research projects – the process of awarding each project a reduced percentage of the funding requested – does result in more research projects being done. But can they be done ‘properly’ with a fraction of their estimated budget? The answer is that we do not know, and this could be obscuring a serious problem.

If projects cannot be done properly, in one way or another, there are likely to be scientific (epistemic) consequences – enforced changes in vital elements such as research questions, methods, equipment, and personnel, which impact on the direction and the quality of the science performed.

The precise nature of these scientific consequences of underfunding are also potentially diverse, depending on the extent of the budget cut and the type of science involved. To take one actual example from the social sciences, Project X was designed as a mixed methods study involving two principle forms of primary data creation: a large scale survey; and a subsequent series of interviews. Despite receiving a relatively substantial amount of funding and excellent reviews, the total amount of financing provided nevertheless represented a 60% cut in the budget. Without any criticism of the proposed research plan, or any other guidance on execution, it was left to the researchers to decide how to proceed.

A choice had to be made between doing all tasks foreseen in the project, but in a reduced way, or to do some tasks as well as possible. The investigators chose the second approach – a large-scale, pre-tested and rigorously designed survey was undertaken and yielded a high-quality dataset that will be the basis for several scientific publications in top journals in the researchers’ field. The interview series was completely abandoned. This type of work is relatively resource intensive and could not be undertaken on 40% of the project’s original budget.

The scientific consequences of underfunding Project X are clear. Careful modelling and statistical analyses of the survey data will enable the researchers to confirm or refute their hypotheses regarding those factors which influence the social phenomenon under investigation. However, they will not then be able to go into the field to ask a selection of the survey respondents why this is the case. This means the crucial work of identifying the causal mechanisms that produced the effects observed in the survey will not be done. Project X will thus succeed in the task of identifying crucial phenomena, but fail in the task of explaining them.

Imagine this type of story repeated over and over, in project after project, for year after year. It is easy to imagine that the epistemic consequences of project underfunding could be broad, deep and go to the very fabric of the knowledge our scientists are producing. This is worrying.

One of the ironies of this situation is that research funders who underfund research receive exactly the same acknowledgement as those who fund research fully. This does not seem right.

It might be time for funding acknowledgements to include a more transparent form of support statement, detailing not just the funding received but also what proportion of the funding requested was obtained. This could be accompanied by a statement about how the project was modified to use the available funds.

Many details are unknown about the practices of project underfunding. Are project applications from all scientific disciplines subjected to cuts by funders? Are some disciplines or specialisations cut less than others? Are there guidelines for the level of systemic underfunding applied to project proposals, or is the size of a cut random or reviewer dependent?

While the epistemic consequences of systemic project underfunding remain in the shadows, it seems unwise to wait until damaging effects start to emerge to get serious about asking questions. Research into project underfunding and its’ effects is urgently needed. So is a greater degree of transparency and accountability on the part of those who make funding decisions. In an age of every-increasing concern about the quality and validity of scientific papers and even whole sub-fields within the established scientific knowledge base, it may be perverse to continue with the risky practice of project underfunding from the perspective of the long-term epistemic health of science – even if it seems to make a great deal of sense from a short-term political point of view.

Do you have a story to tell about how you coped with a reduced research budget? If you are interested in sharing your story you can write to ricwoo@ingenio.upv.es or, if you prefer, leave a comment on this blogpost.