Care Work in the Woodwork: Medicaid Home Care and Family Caregivers’ Health

Abstract

Family caregivers provide the majority of eldercare in the United States, yet little is known about how eldercare policy affects their health. For several decades, research has shown that family caregivers experience emotional, physical, and financial strain that is associated with worse mental and physical health. These harms are particularly severe for women, who are overrepresented among family caregivers, provide more hours of more intensive care, and experience worse associated declines in health. Simultaneously, U.S. eldercare policy has undergone enormous change since the 1980s, as states have shifted Medicaid long-term care spending from nursing homes to home care services. These policy changes have restructured the social, political, and economic systems that shape the realities of family caregiving. However, this dissertation is the first work to investigate how Medicaid home care affects family caregivers’ health.

In Chapter 1, I used data from the Medical Expenditures Panel Survey, a nationally-representative longitudinal survey of households’ healthcare utilization, to assess how family members’ health changes when an adult in the household with limitations in activities of daily living (ADLs) begins receiving Medicaid home care services. I applied person-level difference-in-difference models, with person and time fixed effects, to assess the within-person change in self-rated physical and mental health associated with new-onset Medicaid home care among “likely caregivers” – non-disabled adults who lived with at least one adult with ADL limitations. I found that likely caregivers’ self-rated mental health improved by 0.075 standard deviations (95% CI: 0.010, 0.140; p=0.02), or 3.39% over their baseline mental health, approximately equivalent to an 8% decrease in the odds of screening positive for depression or a 9% decrease in the odds of screening positive for severe psychological distress. This improvement in mental health was larger among caregivers who were women, Black non-Hispanic or Hispanic/Latinx, living in or near poverty, unemployed, under 65, or caring for people with dementia or other cognitive limitations. There was no association with physical health.

In Chapter 2, I expanded on the work in Chapter 1, using the same dataset and similar methods to assess whether there was a dose-response relationship between the volume of Medicaid home care and the degree of improvement in likely caregivers’ mental health. In addition, I used latent class analysis to explore how heterogeneity in care recipients’ disability status, health events, and service providers modified this dose-response relationship. I found that overall, an additional four days per month of Medicaid home care was associated with an improvement in family caregivers’ health of 0.017 standard deviations (95% CI: 0.002, 0.032; p=0.026). However, there was heterogeneity by latent class. Among caregivers for individuals receiving post-acute rehabilitation services, four additional days of home care per month was associated with an 0.057 standard deviation improvement in self-rated mental health (95% CI: 0.025, 0.089; p.001) – equivalent to a 6% decrease in the odds of screening positive for depression or a 7% decrease in the odds of screening positive for severe psychological distress. By contrast, no significant dose-response relationship was observed among caregivers for people with long- or short-term disabilities receiving primarily personal care services.

In Chapter 3, I used data on older couples interviewed in the Health and Retirement Study and annual data on state-level Medicaid home- and community-based services (HCBS) spending to assess how spousal caregivers’ health changes when states shift long-term care funds from nursing homes to home care. I used state-level difference-in-difference models with state and year fixed effects, as well as a non-caregiver control group, to isolate the effects of Medicaid HCBS spending on caregivers from other state-level factors. I found that when states increased the portion of Medicaid long-term care spending devoted to HCBS by ten percentage points, spousal caregivers’ self-rated health improved by 0.057 points on a scale from one to five (95% CI: 0.013, 0.100; p=0.011) – a 16% reduction in the caregiver-non caregiver health gap.

Thus, across a variety of datasets and statistical methods, access to Medicaid home care is associated with substantial improvements in family caregivers’ health. This analysis, guided by feminist theories of care and ecosocial theory, demonstrates that family caregivers’ health is inextricably intertwined with the policies that structure caregiving. As a result, the population health impact of long-term care policy is far greater than previously understood. When assessing the social costs and benefits of long-term care policy, policymakers must take family caregivers’ health into account.