Erratum

ENDNOTES

1. In truth, a more appropriate title for this paper could have been “An Alternative to the Scully Approach to Estimating a Player’s Marginal Revenue Product.” Apologies to Gerald Scully (RIP).

2. The first-order condition of expected profit maximization specifies that the competitive firm equates expected MRP to the price of the input, hence the motivation for using the wage to approximate MRP. Unless there exist exists systematic bias in the models used to predict the player’s marginal value, the forecast errors should be white noise.

3. While the actual method used by Forbes to estimate team revenues is proprietary, vague references to their methodology did appear many years ago when Financial World was publishing this data. Here it was implied that the estimates were derived from simple marketing models, with ad hoc adjustments made in recognition of legal and public commitments (e.g., lucrative stadium contract clauses, locally mandated financial stipulations).

4. The regression model run is TR = α + β1WINS + β2 WINS2 + β3 POP, where WINS is total number of wins, and POP is the MSA population. The sample used corresponds to the 1999 and 2000 seasons. Other specifications tried included a dummy variable for teams making the playoffs. Empirical results are available from the author upon request.

5. Sabermetricians have gone to great length to devise metrics attempting to measure the value of a player relative to the alternative player who would have been hired instead (e.g., VORP, WARP, Batter-Fielder Win.). Many of these metrics assume that the replacement player is the average player at the same position.

6. Some studies control for the starter vs. utility player effect by including dummy variables for utility players; other times, this is achieved by segmenting the dataset according to starters vs. utility players. In Krautmann (1999), the wage equation includes AB per year to capture whether the player is a starter or utility player.

7. The surplus is defined as (MRP – WAGE), the difference between the player’s (estimated) MRP and his wage.

8. These wage factors include a measure of the player’s expected marginal product (above his replacement player), his age, the market size of his signing team, and controls for both primary position and whether he is a starter or utility player. The empirical results are available on request from the author.