The relationship between income, health status, and health expenditures in the United States

Abstract

The relationship between income and health has important implications for policy makers and businesses, and will continue to receive attention as healthcare reform takes hold in the U.S. Most existing literature looks at the relationship between income and either health status or health expenditures in isolation. However, in this research, we take advantage of the wealth of data available in the U.S. Department of Health and Human Services' Medical Expenditures Panel Survey (MEPS) to answer two important, related questions regarding the income-health relationship for U.S. adults. First, we seek to determine how much sicker are poorer people than richer people (if at all), both in their perception and in actual terms. Second, we seek to determine if a poorer person is likely to consume more or less care than a richer person for given level of health or condition. To answer the first question, we start by examining the relationship between family income and health status using multiple regression techniques. For both perceived health and actual health, we find a curvilinear relationship between income and health, with diminishing returns associated with membership in successively higher-income groups. Depending on the status metric, the associated health benefits of membership in highincome cohorts tend to flatten once income reaches approximately 500-600% of the federal poverty level (FPL). We also find that marginal income at low income levels tends to be more strongly associated with reduced probability of poor health than increased probability of strong health. Regardless of the dependent variable chosen, we find that the shape of the relationship between income and health status is the same once we normalize the coefficients. Perceived and actual health are strongly related, although some of our results indicate that poorer people may be more pessimistic about their health than richer people. We find similar trends when we examine the relationship between income and health expenditures using the MEPS data. In this case, however, the diminishing returns associated with membership in higher-income cohorts are more accelerated, and the associated reductions in spending for membership in successive cohorts above 200-300% FPL are not significantly different from zero. When we add controls for health status, however, we find that the wealthiest members of the population are most likely to have the highest spending on healthcare, although not drastically so. In addition, we find the poorest members of the population do not have a tendency to overconsume care relative to their level of health.