IZA DP No. 8645: All-Star or Benchwarmer? Relative Age, Cohort Size and Career Success in the NHL
published in: Frick, B. (ed.) Breaking the Ice: The Economics of Hockey, Berlin, Heidelberg, New York: Springer, 2017
We analyze the performance outcomes of National Hockey League (NHL) players over 18 seasons (1990-1991 to 2007-2008) as a function of the demographic conditions into which they were born. We have three main findings. First, larger birth cohorts substantially affect careers. A player born into a large birth cohort can expect an earnings loss of roughly 18 percent over the course of an average career as compared to a small birth cohort counterpart. The loss in earnings is driven chiefly by supply-side factors in the form of excess cohort competition and not quality differences since the performance of players (as measured by point totals for non-goalies) is actually significantly greater for players born into large birth cohorts. Performance-adjusted wage losses for those born in large birth cohorts are therefore greater than the raw estimates would suggest. Second, career effects differ by relative age. Those born in early calendar months (January to April) are more likely to make it into the NHL, but display significantly lower performance across all birth cohorts than later calendar births. In short, those in the top echelon of NHL achievement are drawn from fatter cohorts and later relative age categories, consistent with the need to be of greater relative talent in order to overcome significant early barriers (biases) in achievement. We find league expansions increase entry level salaries including the salaries of those born into larger birth cohorts, but they do not affect salaries of older players. Finally we find that the 2004-05 lockout appears to have muted the differentials in pay for large birth cohort players relative to their smaller birth cohort counterparts.