Join us for an ERIM research seminar
Abstract
We examine whether and how workers who participate in a contest reallocate their effort between cooperative and competitive behaviors once the contest organizer introduces individual cooperative incentives. Do workers substitute cooperative for competitive behaviors once cooperation is incentivized? Or do they cooperate less and compete more, for fear that an increase in the aggregate level of cooperation might enable others to move ahead? We study the effects of individual cooperative incentives using longitudinal data on workers active on the data science platform Kaggle. We apply a regression discontinuity in time (RDiT) design to an exogenous incentive change on the platform, through which workers could suddenly earn recognition not just for competitive outputs but also for cooperative outputs. Following the introduction of individual cooperative incentives, workers, on average, increased cooperative effort and decreased competitive effort. However, high-ability workers responded differently, by increasing both their cooperative and their competitive effort relative to other workers. Thus, contest organizers face a trade-off between maximizing aggregate levels of cooperation or competition, but individual cooperative incentives can help them elicit both cooperative and competitive effort from the highest-ability workers.
