Pay and Performance in Baseball: Modeling Regulars, Reserves and Expansion

Although the relationship between pay and performance in baseball has been convincingly demonstrated by Scully, a number of unresolved questions remain. Using a large sample of player salaries from contracts on file at the American League office, new estimates of this relationship are reported. The primary findings are as follows. First, while Scully's basic results are qualitatively robust, the salary elasticities for various performance and experience variables are substantially lower for our sample and specification. Second, for most variables, recent performance, as well as career average, contributes to the explanation of salary differences. Third, expansion has a significant effect on salary structure, and, in our model, makes it statistically invalid to estimate a single salary equation from pooled time-series data that includes an expansion year.