Heterogeneous Complementarity and Team Design: The Case of Real Estate Agents

Abstract

Workers often possess characteristics, such as soft skills, that are important for teamwork but unobserved by managers. In this paper, we develop a teamwork model based on the econometric teamwork framework in Bonhomme (2021) and stochastic blockmodels for binary outcomes (e.g., Bickel et al., 2013), when only team-level outputs are observed. Our model does not impose any functional form restrictions on the complementarity between workers with unobserved characteristics, which are modeled as latent types. We apply our model to a data set from a leading Chinese real estate company; the data contain the complete history of team assignments, team performances, and property details. We find that complementarities between different agent types are heterogeneous and cannot be captured by commonly used production functions. More specifically, workers with intermediate solo performance complement all other workers the most, while those with the best solo performance are not the best team players. Our results suggest that firms can boost productivity by redesigning teams, without incurring additional hiring costs. Leveraging our complementarity estimates, our counterfactual experiments demonstrate that reorganizing teams could enhance overall team output by up to 26.6%.

Keywords: teamwork, heterogeneous complementarity, unobserved heterogeneity, networks, salesforce