Association for Public Policy Analysis & Management

Science is increasingly a team activity, and the size of the teams has been growing. At the same time, there has been an increase in the rate of retractions of published findings. This increase suggests the need to look beyond individual-level explanations and focus on organizational structures and institutional contexts. Drawing on the literature on organizational misconduct, we argue that increasing in division of labor may be a key factor contributing to pathologies in science and the associated retractions. Furthermore, we examine the effects of high-stakes incentives and of institutional corruption as additional predictors of scientific pathologies. Based on a matched sample of 262 retracted and 494 paired papers that were not retracted, we develop indicators of the division of labor in the team that produced a paper and find that the rate of retractions is higher as the division of labor increases (net of team size). Additionally, we find that high-stakes incentives and a culture of corruption are also associated with increased retractions. We conclude with a discussion of the implications of these findings for science policy, in particular for organizing team science projects.