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

Recent work on the complex relationship between health and socioeconomic status has emphasized the importance of very long-run dynamics, such as the long reach of early life events in determining adult outcomes. Although longitudinal data are essential for understanding these long-run effects, very few of the datasets traditionally used by health economists include a long follow–up; for example, the Medical Expenditure Panel Study follows individuals for only two years. One dataset that is ideal for analyzing long-run dynamics is the Panel Study of Income Dynamics (PSID), which has followed the same families since 1968. Children born to original sample members or their descendants are followed from infancy, so that the PSID can be used to study the relationship between very early life events and subsequent health and labor market outcomes. The PSID has, so far, been relatively underused by health economists given the tremendous potential it has to illuminate these important questions.

This session features three studies using data on health and health insurance from the Panel Study of Income Dynamics. Two papers exploit the longitudinal nature of the PSID to explore the links between health and socioeconomic status. Ziol-Guest et al. analyze the impact of poverty in early childhood on obesity in later life. They find that poverty before the age of two significantly predicts higher body mass index in adulthood, while family income later in childhood does not, emphasizing the importance of truly longitudinal data that begin in very early childhood for understanding this relationship. Zajacova et al. use the detailed data on work, income and wealth in the PSID to estimate the long-run economic impact of cancer for cancer survivors and their families. A third paper by Levy analyzes the relatively new PSID health insurance data and concludes that the data are very high-quality, yielding accurate estimates of the full-year uninsured in spite of the relatively long recall period necessitated by the biannual PSID interviews.

This session should be of interest to researchers who study the links between health and socioeconomic status and also to those who are interested in longitudinal data. We hope that the session will encourage wider use of the PSID among health economists by illustrating how this unique, rich dataset can be used to shed light on important questions.