Association for Public Policy Analysis & Management

This study estimates how macroeconomic downturns such as the 2007-09 Great Recession affect the use of various types of long term care, including nursing homes and both paid and unpaid home care.  While there is a considerable amount of research on the effects of factors such as health, disability and family structure on long-term care use, the causal impact of economic factors on care outcomes have not been as well studied.  However, it is likely that changes in income, asset values and wages each have important effects on both demand for and supply of various types of care.  For that reason, from a theoretical point of view, it is difficult to predict how exactly an economic downturn that decreases both employment and income and asset values might affect utilization of long-term care. We estimate the effect of macroeconomic conditions on long-term care utilization using data from the Health and Retirement Study (HRS), pooling data from the original, AHEAD and Children of the Depression Age (CODA) cohorts for the years 1998 through 2012.  We use two different empirical identification strategies to estimate the effects of economic conditions.  The first is to estimate the relationship between utilization of care and a monthly measure of national economic conditions, exploiting variation in conditions over time.  These regressions control for each HRS respondent’s demographic and health characteristics and the characteristics of any adult children who could potentially provide informal care.  The models also include individual fixed effects to control for time-invariant unobservable personal characteristics could otherwise bias the results.  Based upon our estimates of the relationship between real per capita GDP and utilization, we estimate that during the Great Recession overall utilization of long-term care decreased by approximately 12 percent.  This was made up of significant decreases in nursing home care and informal care, but not paid home care.  In a second specification, we make use of variation in the impact of recessions across different parts of the country, as well as over time, by estimating models in which the primary independent variable is the state-level unemployment rate and state and year fixed effects are included.  Although the decline in retirement income from stock market losses is probably fairly similar across the country, some states experience larger and more persistent increases in unemployment.  For example, in the recent Great Recession states with larger construction industries and housing market bubbles, such as Nevada and California, were disproportionately affected.  We therefore expect any long-term care effects, particularly the increase in supply of informal care, to be stronger in those areas.  In these models, we find that higher unemployment is associated with significantly lower utilization of care overall, and in both paid and unpaid home care, but only for women. Finally, we consider the set of factors that might have caused a drop in long-term care utilization during economic downturns.  Our results suggest that two important factors were lower asset levels and higher self-reported health status (a finding which is consistent with Ruhm (2000)).