Measuring the Effects of Coworkers on Wages
The fast-growing literature studying the impact of co-workers on individual’s wages has recently made significant progress by developing techniques that allowed it to move from small and idiosyncratic case studies to more generalizable studies based on large labor markets. However, I show that the empirical methodology underlying this shift delivers a large positive or negative bias in measured co-worker effects in realistic settings. I combine insights from the assortative matching theory with recent computer science advances in graph embedding techniques to develop a machine learning method that allows researchers to obtain efficient and unbiased estimates in those settings. The proposed method allows to non-parametrically measure the potentially heterogeneous impact of different co-workers on individuals’ wages. I am currently using the proposed method to measure co-worker effects in the matched employer-employee panel data covering the entire population of Denmark. The paper contributes to several strands of the literature. The first contribution is to the empirical studies on peer effects using matched-employer-employee dataset with parsimonious machine-learning-based approach that enables reliable and testable results. Second, the paper contributes to the literature of identifying sorting based on unobserved heterogeneity. Complementary to the existing random-effect-based approach, my method delivers precise counterfactual predictions for any individual worker if allocated to any firm conditional on the set of coworkers. Finally, the paper speaks to search frictions and assortative matching literature, providing empirical evidence of the coworker spillovers and worker-firms complementarities in wages.