CEFISES Seminar: Thomas Feliciani, “Funding lotteries and their impact on conservatism and the wider research landscape”
November 22@14:00-16:00 CET
Series: Digital Approaches
Speaker: Thomas Feliciani (School of Management, Politecnico di Milano)
Title: “Funding lotteries and their impact on conservatism and the wider research landscape”
Abstract
In recent years, concerns have grown over biases in science funding, particularly a potential aversion to funding innovative, interdisciplinary, and ‘high-risk, high-reward’ research projects (a tendency referred to as conservatism). Various interventions have been proposed to reduce such biases, aiming to encourage the submissions of disadvantaged proposals and leverage the wisdom of crowds within review panels, while minimizing the influence of bias on panel deliberations.
Funding lotteries are among the proposed interventions. In a funding lottery, peer reviewers identify the most promising project proposals, and grants are randomly allocated among them. Despite the growing adoption of funding lotteries by funders, their effectiveness in reducing conservatism and other biases remains contested – particularly as lottery implementations vary widely, complicating direct comparisons.
In this seminar I present an extended taxonomy of funding lotteries informed by empirical examples, and I evaluate their efficacy using empirical data from a large funder (Novo Nordisk Foundation) and counterfactual simulations. Specifically, I assess: (1) the ability of different lottery types to reduce biases such as conservatism; (2) a comparison among lottery types and between lotteries and other bias-reduction interventions; (3) some potential downsides of lotteries, relative to other interventions. Additionally, I also touch on key issues such as standards of correctness for evaluating funding decisions; aggregation and deliberation as competing models for social choice; the definition and measurement of the concept of novelty in scientific research, and the limitations of counterfactual simulations along with possible remedies.