Under (peer) pressure: Experimental evidence on team size and task performance

Causal evidence for the effect of team size on performance is lacking despite the high relevance of this question. From an economic perspective, one would expect performance incentives to decrease with increasing team size, but the psychological phenomenon of peer effects could mitigate the free‐rider problem. To analyze the behavioral implications due to changes in team size, we exploit a rich dataset from a controlled experiment with teams of either two or three participants of a university research project performing a real‐effort task. Our study provides three main findings. First, increasing team size does not change team performance on average, which is a robust result, be it across performance dimensions and even when introducing pay inequality between two work periods. Second, positive performance spillovers from peer to peer alleviate the free‐rider problem when team size increases. Third, changes in peer pressure due to changes in the transparency of others' performance could explain the variation in peer effects across team size. In contrast to discussions in previous literature, our evidence points to a potentially negative role of peer pressure for team performance. While lower peer pressure in teams of three allows for more positive performance spillovers, a high‐skilled peer in a team of two seems to pressure the other team member to produce more mistakes instead of more work output in high quality.