Optimism, overconfidence, and moral hazard

TL;DR

This paper revisits the moral-hazard model, linking it to behavioral traits like overconfidence and optimism, providing testable axioms and parameter identification within a formal framework.

Contribution

It introduces behavioral definitions of overconfidence and optimism within the moral-hazard model and demonstrates their identification and empirical testability.

Findings

Moral-hazard model parameters are identifiable.

Behavioral traits like overconfidence are formally characterized.

The model relates to the variational choice model under uncertainty.

Abstract

I revisit the standard moral-hazard model, in which an agent's preference over contracts is rooted in costly effort choice. I characterise the behavioural content of the model in terms of empirically testable axioms, and show that the model's parameters are identified. I propose general behavioural definitions of relative (over)confidence and optimism, and characterise these in terms of the parameters of the moral-hazard model. My formal results are rooted in a simple but powerful insight: that the moral-hazard model is closely related to the well-known 'variational' model of choice under uncertainty.