Pull yourself up by your bootstraps: Identifying procedural preferences against helping others in the presence of moral hazard

Abstract

Governments and organizations often implement policies designed to help people affected by undesirable events. Such policies can make the society better off, but they may also create moral hazard. We use a laboratory experiment to examine two questions. First, can discretionary decisions to provide assistance overcome the problem of moral hazard and lead to higher efficiency? Second, if so, will people prefer this discretionary procedure to the strict liability policy in which no assistance is provided? We find that assistance is more efficient than a strict liability procedure. However, people still prefer the strict liability regime over assistance. We conduct additional treatments that show that this effect is not driven by the presence of human discretion, nor by aversion to risk, ambiguity, loss or inequality. This suggests that when moral hazard is a concern people have procedural preferences in favor of strict liability.