How Does Improvement in Commuting Affect Employees? Evidence from a Natural Experiment

Abstract We collect worker month-level panel data from two companies in Beijing for a two-year period before and after the opening of a nearby subway station, which significantly improved public transportation commutes for some workers. We find a significant difference-in-differences increase (12.6% of the standard deviation) in bonus pay, which is strongly correlated to worker-level performance measures, for affected workers relative to unaffected coworkers. We find no evidence that the improved performance is a result of affected workers spending extra time at the workplace. We find suggestive evidence for a relative decline in turnover, consistent with a gain in utility for affected workers.