Denying Medicaid Crowd-Out Doesn’t Make It Go Away
A White House report claims that Medicaid enrollment gains have not been offset by a decrease in private insurance coverage. They have.
The passage of the American Rescue Plan in 2021 temporarily increased federal reimbursement for hold-out states that agree to extend Medicaid coverage to all low-income adults. Soon after the bill was signed, the White House’s Council of Economic Advisers (CEA) released an issue brief touting the positive effects of Medicaid expansion. The brief makes many fair points about the impact of expansion on coverage and access to care, but also distorts the truth by implying that Medicaid doesn’t crowd-out private insurance.
Crowd-out occurs when public funds replace private dollars that otherwise would have been spent on healthcare or health insurance. By extending Medicaid eligibility to all adults with incomes below 138% of the federal poverty level (about $20,100 in 2023), Medicaid expansion states have created a strong incentive for those earning less than the income threshold to drop whatever private coverage they have and enroll in Medicaid. Even those earning a bit more than the income threshold might find it advantageous to strategically reduce their earnings in order to qualify for Medicaid (e.g., someone earning $23,000 per year and spending $3,000 on a private insurance plan could work a bit less, earn only $20,000 per year, qualify for Medicaid, and have more non-health insurance-related disposable income.)
Crowd-out exacerbates Medicaid’s unsustainable fiscal trajectory. Official projections now predict Medicaid expenditures will exceed $1 trillion by 2028. By adding more people to a program already struggling to care for its neediest beneficiaries, crowd-out may undermine Medicaid’s central mission of providing access to timely, high-quality medical services.
Yet the White House seems oblivious to these concerns. In discussing Medicaid expansion’s effect on insurance coverage, the CEA report argues, “Importantly, studies have found that increases in Medicaid coverage have not been offset by a decrease in private insurance (Frean et al., 2017).” This statement is, at best, highly misleading.
If by “offset” the White House means “entirely compensated for,” then the claim is strictly true. Medicaid expansion has led to an increase in overall rates of health coverage. But a more natural reading of the sentence – and the message that an average reader would likely take away – is that Medicaid coverage expansions have not caused declines in private insurance uptake. The research literature says otherwise.
The seminal study on this topic is Cutler and Gruber (1996), which examines expansions of Medicaid to pregnant women and children over the 1987-1992 period. They estimate that approximately 50% of the increase in Medicaid coverage was associated with a reduction in private insurance coverage — meaning that for every 100 new Medicaid recipients, there were 50 fewer individuals on private plans. Of course, the Affordable Care Act (ACA)’s expansion of Medicaid targeted different demographic groups and was implemented in a different historical and policy context. But more recent studies have echoed those findings.
Using American Community Survey and Current Population Survey data and a difference-in-differences approach, Leung and Mas (2016) analyze the ACA’s Medicaid expansion and report: “The total ‘crowdout’ implied by our estimates is 42 percent, with about 11 percent coming from employer-sponsored insurance and 31 percent from direct purchase insurance.”
In 2017, Kaestner et al. used similar empirical methods and found that Medicaid expansion was associated with a 4.6 percentage point increase in Medicaid; a 2.7 percentage point decrease in uninsured; and a 1.6 percentage point decrease in private insurance among low-income adults — implying a Medicaid crowd-out rate of 35%.
The latest evidence, from a 2022 working paper, calculates a 43% crowd-out rate from Medicaid expansion.
To support its claim, the White House cites a single study, Frean et al. (2017), which found that after controlling for other provisions of the ACA, including subsidies to help low-income households purchase plans on the individual market, private insurance coverage actually went up slightly in states that expanded Medicaid. The paper’s authors argue, plausibly, that other research had not adequately considered the effects of premium tax credits under the ACA. While people in non-expansion states earning between 100% and 138% of the federal poverty level are eligible for these subsidies, in expansion states the subsidies are only available to those above 138% of the federal poverty level.
The authors point out that ignoring this discrepancy could overstate the extent of Medicaid crowd-out because the subsidy would be expected to disproportionately increase private coverage for low-income people in non-expansion states. Maybe the studies finding large amounts of crowd-out were really picking up the effects of smaller private insurance subsidies in expansion states, not the effect of Medicaid expansion itself.
However, other choices made by Frean et al. (2017) might have made it more difficult to detect crowd-out. For example, their analysis was based on a sample of all non-elderly individuals residing in the US, the vast majority of whom were not eligible for Medicaid and whose inclusion likely diluted any crowd-out signal in the data.
Although flaws can be found in any study, the evidence mostly confirms what common sense suggests: That expanding eligibility for Medicaid causes at least a moderate reduction in private insurance coverage. With a single study, the White House attempts to ignore the vast majority of published research on this topic. If the Biden administration wants to persuade state policymakers to expand Medicaid, it would do well to confront the reality of crowd-out rather than denying its existence.