How agents assess the (in-)tangible externalities that others might impose on them can strongly influence strategic interaction. This study explores mechanism design for agents whose externality assessments and private payoffs, exclusive of externalities, are all subject to asymmetric information; utility is quasi-linear and transferable. An allocation rule will be called strongly Bayesian implementable if it is Bayesian implementable for arbitrary type distributions. Under reasonable assumptions, the following result is established: A Paretian allocation rule is strongly Bayesian implementable through budget-balanced transfers if and only if it maximizes the sum of private payoffs exclusive of externalities. The corresponding mechanism is necessarily externality-robust in that it leaves agents' externality assessments strategically inoperative. The result emphasizes the critical incentive-theoretical role of the welfare judgment inherent to social choice. Strong Bayesian implementation of a welfare judgment inconsistent with externality-ignoring utilitarianism violates budget balance and thus entails incentive costs.
(behavioral) mechanism design externalities robust implementation social welfare bargaining
This working paper/preprint is based on chapter I of my Dissertation entitled "The Impact of Other-Regarding Preferences on Moral Hazard and Adverse Selection", which has been published by the Library of the Technical University of Munich under https://mediatum.ub.tum.de/?id=1356815 .