Published in: PLOS ONE 13 (2). (2018). e0192038. (with Roel van Veldhuizen and Joep Sonnemans)
This paper reports the results of a lab experiment designed to study the role of observability for peer effects in the setting of a simple production task. In our experiment, participants in the role of workers engage in a team real-effort task. We vary whether they can observe, or be observed by, one of their co-workers. In contrast to earlier findings from the field, we find no evidence that low-productivity workers perform better when they are observed by highproductivity co-workers. Instead, our results imply that peer effects in our experiment are heterogeneous, with some workers reciprocating a high-productivity co-worker but others taking the opportunity to free ride.