Peer Effects and Observability: Theory and Evidence from a Field Experiment

Observability is crucial for the effectiveness of “peer effects”, particularly when workers engage in multiple tasks. This paper presents a theoretical and empirical investigation into the peer effects with multiple dimensions of outputs (quality and quantity), which have differential observability. The model predicts that workers will pivot their efforts towards the dimension with more observability due to peer pressure, and shows that under reasonable conditions, an increase in the obersevability of quality will lead workers to produce higher quality, while reducing the quantity of output. We test this theoretical implication with the data from a field experiment in China. Our empirical analysis shows that in team production, a worker’s effort devoted to quantity and quality production does respond to the changing observability of these two dimensions of a job. Such a finding demonstrates not only the existence of peer effects but also the strategic responses to differential peer pressures from co-workers. Our further experiment of switching workers from team-based to individual-based incentive pay yields two other findings. First, the “social norms” generated in the period of team work persisted into the period of piece rates, particularly its early period. Second, workers exhibited higher productivity under team work than piece rates, which implies that peer pressure is an effective tool to mitigate the free-rider problem of team production.