Ambiguity in Performance Pay: An Online Experiment

Many incentive plans are inherently ambiguous, lacking an explicit mapping between performance and compensation. Using an online labor market, Amazon Mechanical Turk, we study the effect of ambiguity on willingness to accept contracts to do a real-effort task as well as completion and performance of this task. Ambiguity about the relationship between performance and compensation affects the willingness of individuals to accept contracts and the likelihood of quitting before completion, but not performance. These effects are non-monotonic in the level of ambiguity. Information about ability at the task reduces willingness to accept and increases quitting, but does not affect performance.