Why are Americans addicted to baseball? An empirical analysis of fandom in Korea and the United States

Contemporary Economic Policy, Jan, 2008 by Young H. Lee, Trenton G. Smith

According to the GMM estimation results shown in Table 4, the chi-square tests strongly reject the hypothesis that the lagged attendance growth rate is exogenous with respect to model error terms. The Hansen tests do not reject the legitimacy of our instruments and model specification. Moreover, our test of the exogeneity of the income growth rate is not rejected (p = 0.383).

Table 5 provides validation for our choice of instruments: the lagged attendance growth rate is highly correlated with the instruments and other regressors, and the coefficient on lagged income is not statistically significant. These results indicate that our GMM estimates are unlikely to suffer from bias due to weak instruments.

Contrary to our MLB findings, GMM estimation (and in particular, the small and statistically insignificant coefficient on [DELTA]ln([ATT.sub.t])) does not support the hypothesis that KPBL attendance is habit forming: the GMM estimate of the coefficient is only 0.064 and is insignificant even at 10% confidence (Table 4, Column GMM (ii) of Panel II). The estimated coefficient on price is negative, has magnitude greater than unity, and is statistically significant even at 1% confidence. This also stands in contradiction to our MLB findings and implies that the IES for KPBL games is located in the elastic region. (11) The stadium size turns out to have a substantial effect on attendance. This result is not surprising, given the large variation in stadium capacity in our sample: Table 6 shows the four cases in which KPBL teams moved or expanded their stadiums. All four moves to larger stadiums drew more attendance than before movement even when (in two of the four cases) their team win/loss record worsened. Our finding of statistically significant (negative) effects of MLB airtime on KPBL attendance is also consistent with Lee (2006), who found that the dispersion of MLB games has negative effects on KPBL attendance.

In comparing our MLB results (Table 2) with those for the KPBL (Table 4), the differences are striking: we find that MLB attendance is strongly habit forming, whereas KPBL attendance is not; and that while MLB fan behavior exhibits a small IES, for the KPBL, it is large. These differences appear to reflect differing pricing strategies on the part of profit-maximizing team owners, which are presumably driven by fundamental differences in fan behavior. (12) In the next section, we offer a review of the literature on the psychology, anthropology, and endocrinology of the sports fan and sketch a formal model of fan behavior consistent with these literatures before discussing differences in the cultural environments or other factors that might explain our divergent findings for the United States and Korea.

IV. NATURAL ADDICTION TO ... BASEBALL?

A. Endogenizing Addiction

Theories of rational addiction define a good as habit forming if it exhibits adjacent complementarity--that is to say, if the marginal utility of consumption increases with experience. Although a number of authors have subsequently criticized the original theory of Becker and Murphy (1988) for lacking psychological realism, (13) the origins of or reasons for intertemporal complementarities have received little attention. This is unfortunate, because a deeper understanding of the reasons for such complementarities is prerequisite to any attempt to make ex ante predictions about which goods or services in the economy are likely to be habit forming. Our approach, therefore, will be to take a step back and look more broadly at what the scientific literature has to say about the human phenomenon of spectator sports. In this enterprise, we follow Smith and Tasnadi (2007), who studied the problem of habit formation in dietary preferences. In addition to showing that the process of Bayesian learning can generate adjacent complementarity in an optimal foraging framework, the authors provide a review and synthesis of the biomedical literature as it relates to the neuroscience of dietary habits and drugs of addiction.