Peer Effects in Cheating on Task Performance

Recent research has shown that the presence of peers can increase individual output both in the lab and the field. As a new explanation for higher individual output levels, we test whether peer settings are particularly prone to cheating even if peer settings do not provide additional monetary benefits of cheating. Participants in our real effort experiment had the opportunity to cheat when declaring their output levels. Although cheating did not have different monetary consequences when working alone than when working in the presence of a peer, we find that cheating on task performance is a more severe problem in peer settings. Our results potentially have far-reaching repercussions regarding organizational design in the context of group settings where principals are not fully able to observe agents' output levels.