Consumption and Differential Mortality

Ongoing Congressional deliberations about reforming Social Security and Medicare require accurate assessments of the economic status of the retiring population. The most direct indicator of economic well-being is consumption. Yet our understanding of how consumption evolves with age is limited. The primary data source in the United States for understanding spending patterns over the life cycle has been the Consumer Expenditure Survey (CEX), which collects crosssectional data on consumption. However, for several reasons, the variation in the level of consumption with age suggested by the CEX data may not truly reflect life-cycle spending trajectories. First, because cohorts differ in the resources they have accrued over the lifetime, at least some component of the apparent variation with age is in fact the result of differing resources. Second, if mortality varies with levels of consumption, as is suggested by the well-known associations between mortality and socioeconomic status, then the cross-sectional pattern in consumption will reflect the differential mortality rates of the wealthier and less wealthy participants. That is, average cross-sectional consumption could increase with age simply because the members of a cohort who survive to advanced old age are likely to have greater economic resources than those who died. Third, a variety of evidence shows that consumption is likely to increase in the year or two preceding death due to out-of-pocket medical expenses. The purpose of this study is to document the extent and the implications of differential mortality for understanding the trajectory of consumption among the elderly.