The effect of management changes on winning in professional sports: Analysis using a dynamic lag adjustment model

This paper examines whether coaching and general manager (GM) changes among three professional sports leagues—the National Football League (NFL), the Major League Baseball (MLB), and the National Basketball Association (NBA)—effect on‐field performance. Our empirical methodology uses team‐level data by season and adapts a lag adjustment econometric approach designed to resolve several statistical challenges that arise both in general managerial settings and in sports settings. Our main finding is that coaching changes in the NFL boost the number of wins per season by between 0.5 and 1.2 in each of the first five seasons. Coaching changes have smaller, but still positive, impacts in the MLB and NBA. For all the three leagues, we find that GM changes have no discernable impact on performance. A separate cross‐sectional analysis suggests that those small impacts stem from coaches and GMs having extremely compressed talent distributions. The data indicate that coaches and GMs, en masse, are important, but changing the people who occupy those positions rarely seems to move teams to different locations on the performance distribution.