DP18957 Obviously Strategyproof Mechanisms in General Environments
We consider the problem of characterizing all obviously strategy-proof (OSP) mechanisms for general preference environments. We show that any OSP mechanism is equivalent to a generalized millipede game in which agents are sequentially offered a menu of payoffs they may clinch (and thus leave the game), plus possibility the opportunity to pass (and remain in the game, hoping for better clinching options in the future). Our preference setting unifies many canonical mechanism design settings, such as single-unit auctions, public goods problems, and object allocation, and thus, many of the known OSP mechanisms are special cases of generalized millipede games. We also introduce other examples that fit our preference model that are new to the literature.
