Relative value of high-ability shirkers versus low-ability consistent workers in repeated jobs

Determine, in a repeated-game version of Part 3 of the experiment—where Workers repeatedly choose how many of up to 10 addition problems to answer and may shirk by skipping problems that pay the Worker but not the Manager—whether Workers with high addition-task ability who regularly shirk yield higher expected value to Managers than Workers with low addition-task ability who consistently work.

Background

The study measures Worker ability as the number of correct answers on a timed addition task and prosocial behavior via a Dictator Game. In Part 3, Workers perform jobs where they choose how many addition problems to attempt. In the Conflict job, shirking (skipping problems) pays the Worker but not the Manager, while correct answers pay both. In the No Conflict job, shirking yields no payoffs, aligning incentives.

The paper finds Managers generally prioritize ability over prosocial behavior in one-shot settings. The author proposes extending the framework to repeated interactions, where reputation effects likely become more salient, and explicitly notes uncertainty about the comparative value of high-ability but regularly shirking Workers versus low-ability but consistently working Workers.