University:
Boston University
Course:
MET HI 215 | Special Topics in History
Academic year:

2024
Views:

305

Pages:

28
Author:

Diamond Bolton

SUMMARY By 1970, ten of the original sixteen major league baseball teams had relocated from their 1945 locations. The racial composition of the neighborhood was the most im- portant statistical predictor of relocation, even after controlling for average annual atten- dance, place in the standings, and age of the stadium. INTRODUCTION Many baseball fans have been affected by teams' relocations from one city to another. Some are glad to welcome a franchise. Others are distressed to lose teams that matter. Change has been a constant in American life, so we should not be surprised to see it in baseball. Still, the years from 1950 to 1970 saw unexpectedly frequent relocations of baseball teams. In this paper, we report the results of a study. of the predictors of baseball teams' relocations between and within cities. Throughout, we focus not on the attractions of other cities or other sites. These attractions-- new markets anxious for major league baseball and new (and often low-cost) stadiums-- were potentially equally available to teams that did not choose to relocate. Rather, we concentrate on the characteristics of a team, its stadium, and its surrounding residential neighborhoods. Did these characteris- tics impel a team to move elsewhere, or did they combine to render a team comfortable where it was? A NATIONAL OVERVIEW In many ways, the decades before 1945 were a time of remarkable stability for major league baseball in the United States. The most recent relocations from one city to another had taken place over four decades carlier, as the new American League was sort- ing itself out. The Milwaukee Brewers became the St. Louis Browns in 1902, and the Bal- The 1980 racial comparisons, though interesting, are crude because they reflect changes that could have taken place either before or after a team relocated. If the latter, teams could have been anticipating continued demographic change or could have been blind to it. We therefore examined relocations decade-by-decade. Relocations during two decades. Five teams relocated during the 1950s: the Boston Braves (1952), the St. Louis Browns (1953), the Philadelphia Athletics (1954), and the Brooklyn Dodgers and New York Giants (1957). The average black composition in 1950 of the neighborhoods surrounding these teams' stadiums was 30.8 percent. The average for the remaining eleven teams was 22,4 percent. Five of the eleven original teams still at their 1945 locations in 1961 moved during the 1960s: the Washington Senators (1961), the St. Louis Cardinals (1966), and Philadel- phia Phillies, the Pittsburgh Pirates, and the Cincinnati Reds (all in 1970). The average black composition in 1960 of the neighborhoods surrounding these five teams' stadiums was 62.4 percent. The average for the remaining six teams was 13.3 percent.- (Sce Table 2.) Note that all of the teams relocating during the 1950s moved to another city while four of the five teams relocating during the 1960s moved within their city. Put another way, every team relocating between 1952 and 1961 moved to another city, while every team relocating between 1966 and 1970 moved within the city, usually from a neigh- borhood stadium to a downtown site. This change in the type of relocation probably reflects a combination of the two forces mentioned carlier: the drop in the number of large cities actively demanding teams, owing to the relocation of teams during the 1950s, and the increased supply of franchises owing to league expansions. Each variable's relation to chance of relocation. Preparing for multivariate analyses, we performed a series of t-tests on each independent variable's association with relocation between 1951 and 1970. As shown in Table 3, teams in black neighborhoods were very likely to relocate; teams playing in older stadiums were almost as likely to relo- cate; teams with higher attendance were somewhat likelier to remain; and teams lower in the standings were very slightly likelier to relocate. Correlations among the variables. We examined the relations among the variables through the correlation matrix presented in Table 4. Many of these relationships are depicted in the series of figures following the end notes. It is interesting that area percent black is correlated positively with age of stadium and very negatively with average attendance. Avcrage age of stadium is very negatively correlated with average attendance. As would be expected, attendance is very negatively associated with games behind in the standings. Multi-varlate analyses. To attempt to gauge the best predictors of relocation, we performed ordinary-least-squares regression and logistic regression (logit). Ordinary- least-squares regression can be performed more casily on the small number of cases avail. able in this study, but is not theoretically preferable here because the real.value.of the dependent variable can be only one or zero and regression can produce, calculated (or pre- dicted) probabilities of relocation below zero and greater than one. We therefore checked the principal ordinary-Icast-squares regression result in two ways: by performing a logit on the same variables"' and by comparing predicted probabilities of relocation with teams' actual behaviors. The dichotomous dependent variable was relocation versus non-relocation between 1950 and 1960 or between 1960 and 1970. By making a second measurement for those eleven teams still in place in 1960, as described earlier, we achieved 27 cases for analysis. This is not a robust number of cases on which to perform multivariate analysis, especially for logistic regression, so results should be interpreted with some caution. In the ordinary-least-squares regression, arca percent black was the first predictive variable stepping into the equation; age of stadium entered next. Neither attendance or games bchind met the statistical criterion for entering.'2 Area percent black was sig- nificant at p = 0.095 and age of stadium was significant at p = 0.169. Together, the two predictive variables yielded an adjusted R? of 17.4 percent, meaning that they explained that share of the variance in the dependent variable, probability of relocation. The entire equation was significant at p = 0.039. In the logistic regression or logit model, the same two predictive variables were en- tered into the equation. Area percent black was again the more significant statistically at p = 0.151, and age of stadium was significant at 0.244. The entire equation accounted for 25.2 percent of the mathematical variation in actual relocation. 13 Equations' predictions versus actual behavior. How accurately did these regression and logit equations predict the 27 chances of relocation? We set the break point equal to 0.50, meaning that we counted a prediction as accurate if an individual team's predicted probability of relocation exceeded 0.50 and it actually relocated or if a team's predicted probability fell below 0.50 and it actually stayed where it was. By this measure, the regression equation and the logit equation correctly classified. 21 teams and missed six. As shown in Table 5, the equations correctly predicted half of the relocations-- the St. Louis Cardinals, New York Giants, Philadelphia Phillies, Cincin- nati Reds, and Washington Senators) but missed the Philadelphia Athletics, Boston Braves, St. Louis Browns, Brooklyn Dodgers, and Pittsburgh Pirates. Why did the equations predict half of the relocations accurately but miss the other half? One way to begin to address this question is by contrasting the teams on the two predictive variables, since teams were wrongly predicted not to relocate in light of their values on these very variables. The stadiums of teams correctly predicted to relocate averaged 47.6 years old, and their neighborhoods averaged 73.1 percent black, The stadiums of teams wrongly predicted not to relocate averaged 41 years old, and their neighborhoods averaged 14.9 percent black. What other factors might help to explain the relocations of the mixed group of teams that were wrongly predicted not to relocate? The Athletics' attendance was dropping in the 1950s; their stadium was in a neigh- borhood whose black population share was rising from 34 percent in 1950 to 62 percent in 1960. The Browns and and Braves were the second teams in relatively small citics that were beginning to experience population losses. Both suffered low attendance. The Browns averaged just over 325,000 per season during their last seven years in St. Louis, a time when they continued to play poorly. Their local competitor was sold to the Busch family with its considerable financial resources, making it harder for them to compete. Their stadium was located in an area that was 30 percent black in 1950, but which be- came almost two-thirds black in 1960. The Braves' attendance had dropped steadily from 1,455,000 in 1948 to 281,000 in 1952, their last year in Boston. Like the Browns, they competed against a local team owned by a person of wealth. They were located in a heavily commercial area with a very low black population share. (Ironically, the Braves were developing an integrated group of players that might have allowed the team to displace the Red Sox as Boston's .. most popular team had been able to struggle on in the city for a few more years.) 14. The Dodgers' attendance dropped from 1.6 million per year in the late-1940s to an average of 1.1 million a year during their last half-decade in Brooklyn even though they continued to field superb teams.IS They were located in an area with a nine percent black population share in 1950, rising to 22 percent in 1960 and to 56 percent in 1970. The Pirates' move was anomalous. Located in a neighborhood with a very low black population share, the stadium site was apparently coveted by a nearby university. The regression and logit equations were wrong only once in predicting that a team would relocate when it did not actually do so. This was the Washington Senators, which were predicted to relocate between 1950 and 1960; they actually moved in 1961, so the models could be said to have been one year carly.

Why Baseball Teams Moves

Related Documents

Recommended Documents

Description

Related Documents

Free up your schedule!

Take 5 seconds to unlock