Robust Implementation with Costly Information

TL;DR

This paper develops mechanisms for robustly implementing social choice functions in environments where agents face costly information acquisition and the planner faces uncertainty about preferences, beliefs, and payoffs, ensuring approximate implementation under small perturbations.

Contribution

It introduces mechanisms that achieve approximate implementation of any social choice function despite agents' costly information and various uncertainties, including noisy signals and trembles.

Findings

Mechanisms are robust to strategy trembles.

Implementation is approximate but effective under small probability perturbations.

Applicable to environments with noisy information and uncertain preferences.

Abstract

We study whether a planner can robustly implement a state-contingent social choice function when (i) agents must incur a cost to learn the state and (ii) the planner faces uncertainty regarding agents' preferences over outcomes, information costs, and beliefs and higher-order beliefs about one another's payoffs. We propose mechanisms that can approximately implement any desired social choice function when the perturbations concerning agents' payoffs have small ex ante probability. The mechanism is also robust to trembles in agents' strategies and when agents receive noisy information about the state.