Duke 2008 | ASHE | Equity and Efficiency in Health and Healthcare

A substantial literature documents a positive association between various measures of socioeconomic status (SES) and health. While health economists have focused primarily on education and income, research in social epidemiology suggests an independent role of social status in health. The pioneering Whitehall Studies of British civil servants revealed a steep association between employment rank and health that was not explained by absolute living standards, unhealthy behaviors or access to medical care. The principal theory that has emerged from Whitehall is the role of psychosocial factors in the determination of health – most notably, the harmful effects of stress resulting from the absence of control at lower ranks. The findings of Whitehall are supported by recent studies showing life expectancy gains for winners of prominent awards (e.g. Oscars and Nobel Prizes).

Questions remain as to whether these documented correlations represent causal effects of status on health. Social status is not randomly assigned in the population, and is highly correlated with other individual characteristics that may independently affect health. In the Whitehall context, unless promotion decisions are completely random, occupational rank will be endogenous. With a limited set of covariates, the studies of award winners also may suffer from omitted variables bias. In the ideal experiment, individuals would be randomly assigned to different levels of social status with differences in subsequent health outcomes reflecting the causal impact of status on health. Although this experimental manipulation of status has been employed in studies of non-human primates, it is not feasible in human populations. In its absence, researchers have turned to unique contexts which yield quasi-random variation in status.

This session consists of three quasi-experimental studies which investigate the existence of a causal impact of social status on health. Two of the papers (Becker, Chay and Swaminathan; Evans and Cristia) explore contexts where the assignment of status is determined by a discrete selection rule (Baseball Hall of Fame selections and U.S. gubernatorial elections). The authors use knowledge of are completely rando the selection rule and the underlying continuous voting measures in a regression discontinuity framework. This approach allows the researchers to distinguish between those "losers" (controls) who "never had a chance" and those who "narrowly lost." Anderson and Marmot exploit quasi-random variation in departmental promotion rates to investigate the impact of rank on the long-run heart disease risk of Whitehall civil servants. Each of the three papers highlights the critical importance of the reference group in status comparisons.