Mechanism Design with Endogenous Perception

TL;DR

This paper models how agents' perception of their private information in screening problems is influenced by their cognitive state, which is affected by mechanism incentives, impacting the design and accuracy of mechanisms.

Contribution

It introduces a framework for endogenous perception in mechanism design, linking cognitive investment to incentive structures and perception accuracy.

Findings

Derived a general representation of attention incentives

Showed how incentives vary with allocation rules

Defined a measure of perception accuracy

Abstract

We model endogenous perception of private information in single-agent screening problems, with potential evaluation errors. The agent's evaluation of their type depends on their cognitive state: either attentive (i.e., they correctly perceive their type) or inattentive (i.e., they might misperceive their type). The mechanism's incentives structure determines the agent's cognitive state via costly investment in cognition. We derive a general representation of attention incentives, show how they vary with the mechanism's allocation rule, and define a notion of accuracy of perception. In applications we showcase how perception both shapes and is shaped by the design of mechanisms.