shows an insignificant effect.
The above results have established that the salary determination process during the 1986-1988 collusion period was different from the process during the 1989-1992 era. The 24% decline in hitter salaries and the 29% decline for pitchers roughly agree with the Bruggink and Rose ( 1990) estimate of a 28% decline in the unweighted salary to MRP ratio for players in their sample. The general decline in player salaries during the collusion era also supports the suggestion that spillover effects from the free agent market to both the arbitration eligible and arbitration ineligible players were important in salary determination (see Burgess and Marburger, 1992; and Frederick, Kaempfer, and Wobbekind, 1992).
Given the above results showing differences in the salary determination process between the collusion era and the latter years in the sample, further analysis was conducted by splitting the sample. 20 Specifically, the salary equations for both pitchers and hitters were estimated separately for the 1986-1988 and the 1989-1992 periods. For hitters, there are widespread differences in the coefficients on almost all explanatory variables in both the OLS and fixed effects models. The hypothesis that the coefficients are the same in both models is rejected by an F-test at the 1% confidence level. For pitchers, while the coefficients show apparent dissimilarities between the two periods, an F-test failed to reject the hypothesis of equal coefficients in the two models at the 5% significance level.
In contrast to the large number of studies of the economic effects of final offer salary arbitration and free agency in the market for professional baseball players, there has been relatively little analysis of the impact of owner conduct during the 1980s on player salaries. This chapter employs a direct model of salary determination that exploits panel data techniques to control for unobservable individual effects such as those due to a player's "heart," personality, or fan appeal. Further, a rich set of human capital and experience variables makes it possible to assess the differential impacts on player salaries that stem from collusion by the team owners.
The fixed effects model shows that, on average, collusion among the owners during the 1986-1988 period lowered salaries for hitters by approximately 24%. Similarly, pitcher salaries were reduced by about 29% during the collusion years. This effect was not limited to declared free agents nor was there an additional impact on salaries for this class of player.