The Dynamic Effects of Poverty and Health Insurance Status On Health Service Utilization Among Older Adults

Background and Purpose: The Affordable Care Act of 2010 addresses one of the major concerns of our public health insurance system by expanding Medicaid eligibility to include poor, working age adults up to 133% of the federal poverty line (FPL).  Although access to health insurance, whether public or private, does not necessarily guarantee that one will access health services, the research on health service utilization among the elderly is unambiguous in finding that those who are the most heavily insured use the most health services.  Thus, while poor, older adults may eventually find it easier to access the health services that they need due to the recent Medicaid changes, those exceeding the new eligibility threshold may still find it difficult to access the care that they need.  As such, this study extends the current literature by examining the patterns of health service utilization among both Medicaid eligible and near-eligible older adults using policy-relevant cutoffs of 133% and 400% of FPL.  In addition, the study addresses the need for a dynamic model of health service utilization that explicitly takes into account the changing circumstances of older adults.  The relationship between income, health service utilization, and public program take-up decisions is a dynamic process that can best be examined longitudinally.  Accordingly, this research examines, in particular, the dynamic effects of poverty and health insurance status on health service utilization among older adults as they approach the traditional age of Medicaid eligibility, i.e., age 65. 

Methods: This investigation uses data from the Rand version of the Health and Retirement Study spanning the years 2002-2008 to estimate a discrete-time hazard model of health service utilization among older adults.  The study draws 60- and 61-year olds from Waves 6 (2002), 7 (2004), 8 (2006), and 9 (2008) and follows their trajectory to explain the dynamic effects of poverty and health insurance status on the utilization decision.  The study incorporates both contemporaneous poverty and lagged poverty variables to distinguish the effects of persistent poverty from those of recent falls into poverty.  Similarly, the study also utilizes both contemporaneous health insurance status and lagged health insurance status variables to distinguish the effects of persistent health insurance coverage from those of recent coverage.     

Results: Life table analysis reveals that poor older adults are much more likely to utilize overnight hospital services and much less likely to utilize doctor services relative to non-poor adults.  Moreover, a discrete-time hazard model reveals that poverty status interacts with Medicaid receipt to significantly influence both hospital and doctor service utilization.

Conclusions and Implications: While the recent changes in the Affordable Care Act of 2010 may render public health insurance more readily available for low income individuals, it is not at all clear whether this expansion of the public system will translate into greater health service use, let alone, greater health among affected persons.  Nevertheless, the results indicate that advocacy efforts should be directed not necessarily towards expansion of public coverage to near-eligible individuals but towards increasing Medicaid take-up among already eligible individuals.