Abstract

Objective: The current debate about how best to meet the growing demand for person-centered, high-quality long-term care in the least restrictive setting possible has centered almost solely on the direct and indirect costs of each type of care. However, to date, there has been very little, if any, attention paid to the relative benefits of different types of care. Our study measures the causal impact of using only Medicaid-financed home health care provided by agencies, compared to other home care arrangements including services provided by family members, on health care utilization outcomes for the care recipient. Data and

Methods: We use longitudinal individual-level Medicaid claims data from three states: Arkansas, New Jersey, and Florida. Adults in these states were eligible for the Cash and Counselling demonstration project, which randomly assigned individuals to treatment that unintentionally encouraged receipt of Medicaid-financed home care provided by home health care agencies. This random assignment to the treatment provides a unique instrumental variable for our study. Using a valid instrumental variable addresses many potential biases that plague observational studies, including omitted variables, measurement error of home care, or reverse effect of institutional care use. We control for baseline disease history and health status to further reduce omittedvariables bias and to enhance the overall precision of the results. Because treatment is a binary variable, we use two-stage residual inclusion methods to estimate the causal impact.

Results: We find that being eligible for the Cash and Counselling program is a significant predictor of relying solely on Medicaid-funded home health care provided by agencies, compared to other home care arrangements. Further, we find that addressing the endogeneity is important; individuals relying only on formal home health care have significantly higher inpatient costs and inpatient days. However, once the endogeneity is addressed, we find that relying on formal home health care significantly decreases Medicaid inpatient costs.

Policy Implications: This is the first evidence about the relative benefits of receiving care from trained workers as part of home health care agencies compared to informal – unpaid and likely untrained – home care. Our findings suggest that further examination of the benefits of care is warranted before further policy pushes towards informal care are made.