The Short and Long Term Effects of In-Person Performance Feedback: Evidence from a Large Bus Driver Coaching Program

Does in-person coaching increase worker productivity? We answer this question using detailed performance data from a large-scale bus driver coaching program, exploiting natural variation in when drivers receive coaching. Our results show significant improvements in fuel economy and outcomes pertaining to passenger comfort. Drivers with lower pre-coaching productivity experience the largest treatment effect, initially closing about 40% of the productivity gap. These effects last four to nine weeks and replicate in a control region. The data suggest that the main mechanism behind the coaching effect is information transmission from higher- to lower-productivity workers, rather than peer pressure or signaling. JEL classification: D23, J24, M53, Q55.