Association for Public Policy Analysis & Management

Researchers long struggled to develop a poverty measure that incorporates a need for health care or insurance and that values health insurance benefits as resources—a health inclusive poverty measure (HIPM)—but could find no valid approach (NAS 1995). Korenman and Remler (2016) showed that such a measure is made possible by guaranteed issue and community rating regulations, provided one accepts that health insurance as a basic need. Remler, Korenman and Hyson (2017) used the same HIPM framework to estimate the impacts of private and public health insurance, non-health means-tested benefits, social insurance and federal refundable tax credits on health inclusive poverty in the US under the ACA. Their estimates show that health insurance benefits are among the most important anti-poverty programs. Medicaid has a large impact on child poverty and reduces poverty greatly (nearly 20 percentage points) among Medicaid recipients.

Such estimates, however, neglect the care that the uninsured receive without payment or for limited payment. The Oregon Medicaid experiment found that the uninsured in the control group received free care with costs equal to about 60% of the costs of care received by those in the experimental group with Medicaid coverage (Finkelstein et al 2016, 2017).

We examine how explicitly accounting for the de facto insurance value of free care changes health inclusive poverty rates. We also estimate the direct impact of free care on HIPM poverty and contrast it to the impacts of Medicaid and other health insurance benefits. We estimated a free care value using the Oregon Medicaid experiment’s estimates for the de facto insurance value of free care provided to controls who applied for but were denied eligibility for Medicaid expansion benefits. Doing so reduced the child HIPM poverty rate nationally by almost 2 percentage points and reduced impact of Medicaid on HIPM child poverty nationally by 3 percentage points.

The uncompensated care provided to the uninsured is funded through a variety of means, including cross-subsidies from other payers. However, including such a value in the de facto insurance value of free care amount to double counting. Instead, we measure direct funding for free care to all the uninsured to calculate a de facto insurance value of free care for the uninsured. We age-rate the value of that insurance using the same age rating as the ACA marketplaces.

In order to get an accurate accounting of poverty impacts of expenditures on free care, we use New York State as a case study. New York is particularly interesting, because even under the ACA, an estimated $3 billion per year in uncompensated care is provided to uninsured New Yorkers (Buettgens et al. 2017). Indigent care pool and federally funded community health centers form the largest sources of funding for free care.