Strictly strategy-proof auctions (Q2203494)

In this paper, the authors study strict strategy-proofness in the canonical one-dimensional mechanism design setting with private values and quasi-linear and strictly single-crossing preferences. More precisely, the main result is contained in the following Proposition 1. If a feasible direct mechanism \((X, P)\) is weakly strategy-proof, then for any \(\epsilon > 0\), there is a feasible and strictly strategy-proof direct mechanism that is uniformly \(\epsilon\)-close to \((X, P)\) (where a direct mechanism is a \(map (XP):[0,1]^n\to [0,1]^n\times R^n\) specifying an outcome \(X^i(t^i,t^{-i})\) and payment \(P^i(t^i,t^{-i})\) for each agent \(i\), given the (reported) types \(t^{-i}\) of the other agents and \((X, P)\) is weakly (strictly) strategy-proof iff \(g^i(X^i(t^i,t^{-i}),t^i)-P^i(t^i,t^{-i})\ge (>)g^i(X^i(r^i,t^{-i}),t^i)-P^i(r^i,t^{-i})\) for all \(i,\,t^{-i}\) and \(r^i\neq t^i\)). A consequence is that strategy-proofness can be made strict by an arbitrarily small modification, so that strictness is ``essentially for free''.