Mortality and the Baseball Hall of Fame: An Investigation into the Role of Status in Life Expectancy
David J. Becker
University of Alabama at Birmingham -- School of Public Health
Kenneth Y. Chay
University of California, Berkeley - Department of Economics; National Bureau of Economic Research (NBER)
University of Alabama at Birmingham - School of Public Health
iHEA 2007 6th World Congress: Explorations in Health Economics Paper
The role of social status in life expectancy is a burgeoning topic of interest in health research. Isolating the impact of "pure status" on longevity is empirically challenging as an individual's ranking within the social hierarchy is related to other SES measures (e.g. education, income) which are correlated with health. Previous work is based either on animal studies (e.g. Sapolksy's work on baboons) or studies of humans in unique non-experimental settings (e.g. British Civil Servants: Academy Awards and Nobel Prize winners).
We use the precise rules for induction into the Baseball Hall of Fame to examine the impact of status attainment (or non-attainment) on life expectancy. The Baseball Hall of Fame provides a compelling context since the selection rule for induction by the Baseball Writers Association of America (BBWAA) is perfectly observed - a BBWAA vote share of 75 percent or more. We explore whether this discontinuous rule for assigning status leads to differential exposure to stress depending on proximity to the cut-off.
We obtain BBWAA voting data from the Baseball Hall of Fame website (www.baseballhalloffame.org). For each player born prior to 1946 (n=597) we use complete voting histories to construct a series of variables used in our analysis. From membership lists on the Hall of Fame website we construct indicators denoting whether a player was inducted by the BBWAA or the Committee on Baseball Veterans. From two other websites (www.baseball-reference.com, www.baseball-almanac.com) we obtain player characteristics including exact dates of birth/death, height/weight, ethnicity and education. Finally, we categorize cause of death using obituaries found through Lexis-Nexis, and a book on the necrology of baseball players.
We estimate the impact of Hall of Fame induction and the number of narrow losses (vote share>50%) on life expectancy, controlling for each player's maximum vote share, the number of ballot appearances, and demographic characteristics. We use several parametric (Tobit MLE, Buckley-James) and non-parametric (Kaplan-Meier survivor analysis) methods to correct for right censoring in the life durations of players who are still alive.
The effect of induction on longevity depends both on the definition of the reference group and the ease of induction. BBWAA inductees do not live longer, on average, than players who were not inducted. However, they do live 10 percent longer than players who narrowly missed induction through the BBWAA. This results from the reduced life expectancies of players who narrowly missed induction relative to non-inductees with lower BBWAA vote shares. Life expectancy falls by 3 percent for each ballot with a vote share over 50 percent but below the 75 percent threshold required for induction. Hall of Fame induction through the Veteran's Committee increases life expectancy, with the largest effect among players who never received a BBWAA ballot share above 10 percent. The effect sizes of the Hall of Fame variables are much larger in magnitude than the effects of BMI, year of birth, and educational attainment (i.e., college matriculation).
Our results suggest that the anxiety induced by non-attainment or delayed attainment of status can lead to premature death - with heart attacks being the predominant cause of death - and that there are health benefits of status attained through 'luck' rather than through performance.
Keywords: social status, population health, censoring, regression discontinuity
JEL Classification: I12
Date posted: July 7, 2007
© 2015 Social Science Electronic Publishing, Inc. All Rights Reserved.
This page was processed by apollo3 in 0.297 seconds