Peer effects in the workplace

Existing evidence on peer effects in the productivity of coworkers stems from either laboratory experiments or real- world studies referring to a specific firm or occupation. In this paper, we aim at providing more generalizable results by investigating a large local labor market, with a focus on peer effects in wages rather than productivity. Our estimation strategy-which links the average permanent productivity of workers' peers to their wages-circumvents the reflection problem and accounts for endogenous sorting of workers into peer groups and firms. On average over all occupations, and in the type of highskilled occupations investigated in studies on knowledge spillover, we find only small peer effects in wages. In the type of low-skilled occupations analyzed in extant studies on social pressure, in contrast, we find larger peer effects, about one-half the size of those identified in similar studies on productivity.