Almost-dominant Strategy Implementation

Though some economic environments provide allocation rules that are implementable in dominant strategies (strategy-proof), a significant number of environments yield impossibility results. On the other hand, while there are quite general possibility results regarding implementation in Nash or Bayesian equilibrium, these equilibrium concepts make strong assumptions about the knowledge that players possess, or about the way they deal with uncertainty. As a compromise between these two notions, we propose a solution concept built on one premise: Players who do not have much to gain by manipulating an allocation rule will not bother to manipulate it. We search for efficient allocation rules for 2-agent exchange economies that never provide players with large gains from cheating. Though we show that such rules are very inequitable, we also show that some such rules are significantly more flexible than those that satisfy the stronger condition of strategy-proofness.