STUDY GAUGES EFFICIENCY OF AFFORDABLE CARE ACT AT REDUCING UNINSURED RATES

The following information was released by the University of Chicago:

UChicago economist finds public coverage helped poorest households while often replacing employer-sponsored insurance at higher incomes

Aug 14, 2026

The Patient Protection and Affordable Care Act of 2010known now mostly as the ACA or Obamacaredramatically expanded publicly supported health insurance coverage across the United States. But how efficiently did those gains translate into fewer people being uninsured?

Anew paperby University of Chicago Research ProfessorRobert Kaestnerwith Anuj Gangopadhyaya of Loyola University Chicago examines that question by looking not only at whether more people gained coverage, but where that coverage came from.

The ACA's efficiency, the authors find, depends heavily on who is being covered. For the lowest-income adults, the ACA's coverage provisions translated almost directly into fewer uninsured people. As income rises, the relationship becomes weaker, and public support is more likely to replace existing employer-sponsored insurance rather than insure someone who previously lacked coverage.

"This research sheds light on a key aspect of the Affordable Care Act, which is the law's ability to increase access to health insurance," said Kaestner, an economist at the Harris School of Public Policy.

"While the law did reduce uninsured rates among the lowest-income adults, the research also shows that it shifted a significant number of people who already had health insurance to new plans under the ACAcertainly not the intent of the policymakers, who sought an efficient mechanism for lifting the number of insured people in the United States."

Using data from the 2008-2024 American Community Survey, the authors study changes in uninsured rates, Medicaid and Marketplace coverage, and employer-sponsored insurance among adults ages 19 to 64. Their analysis focuses on how those relationships varied by income level, across states and over time.

Eligibility for public support is linked to the federal poverty level, an income measure used to determine eligibility for programs and subsidies. These include Medicaid and financial assistance for Marketplace health insurance plans.

The dollar amount changes by household size and is updated annually.

In 2026, for instance, the federal poverty level is $15,960 for an individual and $33,000 for a family of four.

People with incomes up to four times the federal poverty level qualify in all states for premium tax credits that lower monthly premiums for Marketplace plans. People with incomes up to almost 1.4 times the poverty level may qualify for Medicaid, if they live in a state that has expanded it.

People with incomes below the poverty level generally do not qualify for Marketplace savings. For adults in households below 150% of the federal poverty level, or $49,500 for a family of four, nearly all the increase in public coverage via the ACA came from people who had previously been uninsured.

That pattern changed at higher income levels. For adults with incomes between 151 and 400% of the federal poverty level, $132,000 for a family of four, each percentage point increase in public coverage was associated with about a 0.6 percentage point decrease in uninsured rates and about a 0.4 percentage point decrease in employer-sponsored insurance.

The authors describe this shift as "crowd-out": public coverage did not always newly insure someone, but in some cases replaced coverage that had previously come through an employer.

The paper argues this distinction matters for evaluating the ACA's efficiency. If public subsidies extend coverage to people who were uninsured, the policy directly reduces the uninsured rate. But if publicly supported coverage replaces employer-sponsored insurance, the number of uninsured people does not fallinstead, the payer simply changes.

The authors find a clear income gradient. Among adults below 150% of the federal poverty level, public coverage gains translated almost directly into fewer uninsured people. But among those with incomes above 250% of the federal poverty level, the tradeoff was closer to 1.5 people receiving publicly supported coverage for every one-person reduction in the uninsured rate.

The paper also finds that crowd-out was larger among groups more likely to have employer-sponsored insurance before the ACA, including parents and married adults. For parents and married adults with incomes between 250 and 500% of the federal poverty level, the authors estimate the ACA supported approximately three people for each newly insured person.

"This finding gets at the central question of policy efficiency," Kaestner said. "The government was often subsidizing a change in who paid for coverage, rather than newly insuring someone who otherwise would have gone without."

The study's broader approach is important, the authors argue, because much of the existing research on the ACA has focused primarily on Medicaid expansion. By contrast, the authors estimate the total effect of the ACA's major coverage provisions, including both Medicaid expansion and Marketplace coverage.

That broader view changes the interpretation of crowd-out.

Looking only at Medicaid expansion can overstate crowd-out among low-income adults and understate crowd-out among higher-income adults. Among higher-income households, the authors find that much of the estimated shift away from employer-sponsored insurance comes from Marketplace coverage rather than Medicaid expansion.

Some states did not expand Medicaid. The authors argue that analyses focused only on the difference between expansion and non-expansion states miss important parts of the ACA's overall impact.

The paper shows the ACA affected both expansion and non-expansion states. Medicaid expansion states saw large increases in Medicaid coverage. But non-expansion states also experienced meaningful coverage gains through the Marketplace, along with substantial reductions in uninsured rates compared with pre-ACA levels.

The timing of coverage changes also matters.

The authors identify a weakening relationship after 2021 between increases in public coverage and decreases in uninsured rates, a period that coincided with changes to Medicaid eligibility rules during the COVID-19 pandemic and enhanced Marketplace tax credits under the American Rescue Plan Act. Their results suggest more generous subsidies and eligibility rules may have increased public coverage in ways that did not always produce proportional reductions in the uninsured rate.

If the goal is to reduce the number of uninsured people as efficiently as possible, the authors find that policies targeted at low-income populations are likely to produce larger reductions per public dollar spent. They point specifically to expanding coverage among low-income adults in non-expansion states as a more efficient approach than subsidizing coverage for higher-income groups.

Conversely, the paper suggests that subsidies aimed at households above 250% of the federal poverty level are less efficient at reducing uninsured rates, because a larger share of public coverage gains among those groups appears to come from people moving out of employer-sponsored insurance.

The authors set the efficiency question against a clear gain in enrollment.

"The ACA dramatically expanded publicly supported insurance coverage across all states," the authors conclude.