Answers are grounded in this paper and cite their sources.
We study all-pay contests with incomplete information in which, besides the winner's prize, agents are exposed to exogenous identity-dependent externalities. An agent receives a non-positive payoff whenever a specified event occurs, where the loss function is type-dependent and does not need to be synchronized with the winning value. We look at three key specifications: general externalities (every loser suffers), status externalities (one designated opponent outperforms the agent), and second-place externalities (the agent finishes second). Externalities are considered moderate when each type's expected winning payoff exceeds its expected loss, allowing all agents to participate. Under moderate general or status externalities, every type exerts strictly more effort than in the standard all-pay contest. Under moderate second-place externalities, this is no longer true type by type, and we present a necessary and sufficient condition for aggregate effort to increase. These moderate externalities could theoretically be generated by the designer, who could then increase the agents' expected efforts, but unlike other well-known methods, no agent withdraws from the competition.
No takes yet. Share an insight, caveat, or question.
Minchuk et al. (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: