Something About How Washington Pays for the Uninsured
AI-generated Song: “Who Pays the Care”
Data: HCRIS
Why It Matters: As Washington debates how much the federal government should spend on health care during the shutdown, it revives a familiar question: How much should Washington subsidize the rising cost of hospital care for the poor and uninsured?
One flashpoint is Disproportionate Share Hospital (DSH) payments: funds that help hospitals cover costs not paid by patients or insurers. This blog focuses on Medicare’s Uncompensated Care (UC) DSH payments. While the shutdown itself doesn’t cut UC DSH, it underscores the broader budget debate over whether the federal government should continue to absorb these shortfalls or shift more responsibility to states and hospitals.
We examine a pivotal moment in 2014, when the UC DSH payment formula underwent a major redistribution. This episode offers perspective for today’s fiscal standoff, as Medicaid DSH faces automatic federal cuts unless Congress intervenes. Understanding how these programs balance fairness and incentives is key to evaluating the future of the hospital safety net.
Before FY 2014, Medicare DSH payments were an add-on to the inpatient PPS rate, essentially a per-discharge percentage adjustment based on a hospital’s share of low-income (Medicaid and SSI) patients. After FY 2014, the Affordable Care Act replaced most of this with an Uncompensated Care (UC) DSH pool system. Under the new rule, hospitals receive 25% of their old-formula DSH directly, while the remaining 75% forms a national pool that is redistributed annually:
Each hospital’s payment depends on its share of uncompensated care:
The 2014 reform did not merely coincide with the Medicaid expansion wave. Instead, it was fundamentally driven by it. First, as Medicaid expansion reduced the number of uninsured individuals, Congress reasoned that hospitals would face less uncompensated demand, so the national UC DSH pool was scaled down by the measured decline in the uninsured rate. Second, by tying redistribution to uncompensated care costs rather than low-income patient volume, the policy shifted focus from serving Medicaid enrollees to supporting hospitals that continued to bear the burden of caring for the remaining uninsured.
Data:
To begin with, the following figures show how uncompensated care, normalized by either total assets or net patient revenues, declines for both DSH-qualifying and non-DSH hospitals. These patterns suggest that the scaling-down factor in the UC DSH pool was well-motivated. Uncompensated care primarily reflects services for uninsured patients, and as Medicaid expansion reduced the uninsured population after 2014, the overall demand for uncompensated care fell. Naturally, both groups saw reductions, though the decline was steeper for DSH hospitals, which tend to operate in lower-income areas more directly affected by Medicaid expansion.
Next, let’s bring in the Medicaid expansion for DSH hospitals. The following figure reveals an interesting pattern. On average, uncompensated care among DSH hospitals declines after 2014, but this aggregate masks important heterogeneity. Once we separate by expansion status, the contrast is clear: DSH hospitals in Medicaid expansion states experienced steep declines in uncompensated care, while those in non-expansion states saw slight increases instead.
Remember, after 2014 each DSH hospital receives a share of the national UC DSH pool based on its reported uncompensated care. This figure illustrates a redistribution effect. When DSH hospitals in expansion states see UC drop (because uninsured patients get Medicaid), then:
The denominator shrinks.
Non-expansion DSH hospitals’ relative UC share rises.
Hence, more of the fixed national pool flows to them.
This is the redistribution effect: money shifts from expansion-state DSH hospitals to non-expansion DSH hospitals. Does this redistribution make sense? From a fairness perspective, this redistribution makes sense. Hospitals still facing high uninsured burdens should receive more federal support. The formula works as intended as it automatically redirects funds to where the need remains.
From an incentive perspective, however, the design is problematic. Non-expansion states are effectively rewarded for not expanding coverage, while expansion states are penalized for achieving it. Worse, the structure invites potential gaming. Hospitals in non-expansion states could inflate reported uncompensated care, either operationally by admitting more self-pay patients or accountingly by overstating charity care through inflated charges and partial payment write-offs. In fact, absent any endogenous response, we would expect uncompensated care in non-expansion states to remain roughly flat after 2014.
Lastly, consider Texas and Florida, two large non-expansion states that have become bottlenecks in the Medicaid expansion map. Both have high uninsured rates and a non-trivial share of DSH hospitals. These states face a built-in disincentive to expand: doing so would reduce their hospitals’ their UC DSH payments, effectively requiring state governments to replace the lost federal subsidies with state funds through Medicaid reimbursement.
But this creates a cross-state externality as a “lock-in” effect. By not expanding, Texas and Florida keep the national UC DSH pool larger, sustaining payments not only for their own hospitals but also for those in other non-expansion states. In a counterfactual world where both states were required to expand Medicaid, the uninsured population, and thus the national UC DSH pool, would shrink sharply. The incentive for any state to remain outside the expansion would largely disappear.