As research managers, we’ve always known that getting funding is something of a lottery. We can do our best to make sure that applicants have the best possible chance of success, but ultimately it comes down to chance, and the proposal is at the mercy of the competition, the assessors’ comments and the views of the panel.
Last month the Swiss National Science Foundation decided to take the next logical step and introduce a lottery element to its decision-making process. Applications would be sorted into fundable and non-fundable, and then, ‘in rare cases [where] the quality of applications cannot be differentiated any further…the SNSF will implement a physical drawing of lots.’
At first this may seem shocking. After working on an application for months, the funding outcome will depend entirely on whether your number is picked from the tombola. However, there is some logic to this, as Matthias Egger, the President of SNSF’s Research Council, explained: ‘drawing lots is the fairest solution because it is blind and rules out bias.’
True, but the process does seem brutal and there must be a better way. It’s an issue that funders (and journals, in the case of publication) have grappled with for some time. Some schemes have an interview stage so that applicants can be more robustly questioned; others, such as the forthcoming pilots for Horizon Europe, have anonymised (or ‘blind’) evaluation.
Elsewhere there have been more radical solutions. In 2017 EPSRC tried a system of ‘concept auditions’ for its Human-like Computing call, whereby applicants applied for a limited number of tickets to a Dragons’ Den-style pitch. You can’t fault them for their out-of-the-box thinking.
Similarly, the journal Shakespeare Quarterly introduced a form of ‘trial by wiki’, whereby it posted four articles online and invited comments. 41 people joined in, leaving 350 comments. The revised essays were then reviewed by the editors, who made the final decision as to whether to include them in the printed journal.
For me, these all have the same in-built fault: a binary gateway late in the process. Yes, the assessment may be blind, or the pitch may be in person, or a wider pool of comments may be invited, but ultimately there is a yes/no decision at quite an advanced stage, and months of effort go up in smoke for the unfortunate majority.
A better solution would be a more iterative process. An outline is presented to the funder, and a decision made at that stage as to whether the concept has legs. If it does, the applicant works with the funder to develop the idea, so that it results in a project that suits the abilities of the investigators and the needs of the grant giver.
It’s not easy, but there is far less wasted effort on either side. Applicants don’t have to work blindly for months trying to second guess the wishes of the panel; peer reviewers are called on less to assess proposals; and the funders waste less time on processing a huge number of unsuccessful applications.
As with all radical solutions, there’s a chance of failure, but what have we got to lose? At a time when academics, desk officers and research managers are all scrabbling to balance workloads, it seems like an experiment worth trying. When we get to the point where decisions are being made by lottery, it’s worth thinking again.
This was first published in Issue 13 of Arma’s Protagonist magazine, and is republished here with kind permission.
Photo by dylan nolte on Unsplash