In Sickness and In Health: Understanding the Effects of Marriage on Health

Abstract

Despite substantial research, how marriage reduces mortality remains unclear. Using a novel data set of over 150,000 elderly couples developed from the Medicare claims and examining the impact of widowhood vs. marriage on mortality, this dissertation seeks to adjudicate among the many plausible mechanisms. First, I exploit the variation between diseases in how responsive those diseases are to putative intervening processes to test which hypotheses can explain patterns of relative mortality. I find that in general the less healthy a person is, the less responsive they are to the death of a spouse – the most enduring mortality effects of spousal death are found among the healthiest elderly. Thus, support is found for models in which marriage provides immediate instrumental help; marriage does not appear to improve health by improving habits or by increasing some other stock of health capital that inheres in patients after the death of their spouse. The major mortality benefits of marriage also are not mediated via direct emotional or physiological effects, nor are they an artifact of selection or shared environment. Second, I show that the married appear to choose better inpatient hospital care, as measured in a number of different ways. They then have longer lengths of stay than the widowed, but do not appear to receive differential quality of care given the provider chosen. In a final chapter, I argue that relationships improve health by serving as a form of general capital, most useful when individuals face a wide variety of choices about their health.