Why not make research grants a literal lottery?

An article in Nature highlights the trend of declining acceptance rates for research grants: “Is research funding now a lottery? Elite funder ERC swamped by applications”. The latest European Research Council’s Starting Grant competition will have a success rate of under 5%. Many other granting programs across academia face similarly abysmal odds.

AI is contributing to the problem of a skyrocketing number of applications targeting a fixed number of prizes, as preparing a competent grant application is easier than ever. Still, the current system consumes a tremendous amount of time and energy from both the researchers submitting the grant applications and the organizations and peer reviewers adjudicating them. AI will only accelerate this arms race, with applicants producing ever-more elaborate proposals and reviewers having to sift through them (or get their AIs to do so). Yet failure is near-guaranteed.

Maybe we should stop pretending this system can reliably discriminate among dozens or hundreds of basically good proposals. If applying for a major grant is like a lottery now, then why not make it a literal lottery? It would save everyone a lot of time and heartbreak if applicants were allowed to submit much shorter proposals, subject to a basic quality screen, to be chosen by a random lottery. In fact, this system and variations on it have already been tried across a number of funders and countries.

At least then we can be precise about how much of the process comes down to luck!