Optimal Contest Design: Tuning the Heat

Igor Letina,Shuo Liu,Nick Netzer
DOI: https://doi.org/10.1016/j.jet.2023.105616
2022-01-01
SSRN Electronic Journal
Abstract:We consider the design of contests when the principal can choose both the prize profile and how the prizes are allocated as a function of a possibly noisy signal about the agents' efforts. We provide sufficient conditions that guarantee optimality of a contest. Optimal contests have a minimally competitive prize profile and an intermediate degree of competitiveness in the contest success function. Whenever observation is not too noisy, the optimum can be achieved by an all-pay contest with a cap. When observation is perfect, the optimum can also be achieved by a nested Tullock contest. We relate our results to a recent literature which has asked similar questions but has typically focused on the design of either the prize profile or the contest success function.
What problem does this paper attempt to address?