Issues in the Empirical Evaluation of the Black-White Mortality Gap

A large empirical literature seeks to understand the forces that shape racial disparity in mortality. Given that factors early in one's life can be important for subsequent mortality outcomes, research in this literature often relies on panel data that follows individuals over many decades. An important example is empirical work that uses the National Longitudinal Survey of Older Men (NLS-OM), which collected socioeconomic data for men aged 45-59 in 1966 and several subsequent years, and then also reported deaths as indicated by death certificate data collected in 1990. Here we examine an important methodological issue that arises in studies that use panel data: deaths are likely to be under-reported, most likely in systematic ways. In the NLS-OM, for example, the matching procedure appears to have missed a substantial number of deaths. We work out a simple model that illustrates the effect of this particular form of measurement error, and then show that inappropriate handling of the measurement error in survival analysis causes serious inference problems in the NLS-OM data