The Medicaid Cliff

The prospect of losing Medicaid coverage can discourage higher earnings.

• Although eligibility thresholds for Medicaid vary from state to state, nearly all beneficiaries face a steep “cliff” at which a small increase in income would make them ineligible for the program.

• The Medicaid cliff creates a large implicit tax rate on beneficiaries increasing their income, which weakens their incentives to accept higher-paying jobs or extend their work hours.

• Two of the Affordable Care Act’s provisions influence the Medicaid cliff. The expansion of Medicaid to 20 million working-age, able-bodied, mainly childless adults exacerbates the program’s negative effects on the labor market. On the other hand, the subsidization of private coverage for low-income individuals and families softens the Medicaid cliff by making the financial consequences of losing Medicaid coverage less severe.

Medicaid is the largest means-tested public assistance program in the U.S., with expenditures exceeding $728 billion in 2021 and enrollment surpassing 80 million people. It is also one of the few remaining programs for which people can be fully eligible at one income level and then completely ineligible if their earnings rise slightly. This phenomenon — the “Medicaid cliff” — creates perverse incentives.

States have considerable discretion in setting Medicaid income eligibility thresholds. In Utah, for example, young children qualify for Medicaid if their household income is below 144% of the federal poverty line (FPL, which is about $28,000 for a family of four). In Vermont, young children are eligible up to 317% of the FPL. To qualify for Medicaid in South Dakota, a pregnant woman’s household income must not exceed 138% of the FPL, while in neighboring Iowa she remains eligible up to 380% of the FPL.

No matter what eligibility limit a state sets, economic theory predicts that households will incorporate knowledge of this income threshold into their decisions about their total earnings in order to maintain Medicaid eligibility. Medicaid recipients just below the eligibility threshold are discouraged from accepting a higher-paying job or working more hours because the resulting loss of Medicaid coverage could, paradoxically, make them worse off than before the earnings increase. Choosing to earn less to remain on Medicaid is a rational choice for individuals and families, but it deprives society of their productivity while draining public resources.

The size of the Medicaid cliff depends on many factors. The number of people in the household enrolled in Medicaid, the availability of other sources of health coverage, the health needs of those who stand to lose coverage, and other considerations will influence individual decision-making. For a healthy, young adult in a state where provider participation in Medicaid is low and access to care is poor, coverage through Medicaid may not amount to much more than financial protection against catastrophic illness. Since the value of Medicaid is low for this person, the incentive to limit income to maintain eligibility is weak. On the other hand, for an individual with a disability requiring intensive treatment or someone who is highly risk-averse, the value of Medicaid coverage is higher — and the incentive to remain eligible is stronger.

The Affordable Care Act (ACA) instituted two reforms that affected the Medicaid cliff. The first was the expansion of Medicaid to cover a large group of non-elderly, able-bodied, mainly childless adults: roughly 20 million people, as of last year. Eligibility was capped at 138% of the FPL. Since this population tends to have a strong attachment to the labor market, Medicaid expansion may have prompted significant shifts in work-related behavior.

For people whose earnings are in the 138% threshold, slight adjustments — declining an extra shift here and there — may be sufficient to become (or remain) eligible for Medicaid. Consider a childless couple in their 20s. In the 38 states that have expanded Medicaid under the ACA, both individuals can qualify for the program if their annual household income does not exceed about $25,270 per year. Even at the federal minimum wage of $7.25 per hour, they would surpass this threshold if they both worked full-time ($29,000); at that wage rate, the Medicaid cliff provides a strong incentive against working full-time. If one is paid $12.50 per hour, they will earn $25,000 per year — narrowly below the Medicaid threshold. Their partner need not work at all.

The second reform put in place by the ACA is a system of premium tax credits that subsidize private health insurance for low-income people who don’t qualify for Medicaid. These subsidies blunt the effect of the Medicaid cliff by making the loss of Medicaid coverage less costly. Prior to the ACA, an individual without access to an employer-sponsored plan who narrowly exceeded the Medicaid eligibility threshold rarely had access to affordable health coverage; in the vast majority of cases, they were uninsured. Under the ACA, that individual in most states would qualify for substantial subsidies to purchase a private health insurance plan. The availability of affordable sources of coverage above the Medicaid eligibility threshold weakens the incentive to remain on Medicaid.

But the ACA’s subsidization of private coverage does not entirely eliminate the Medicaid cliff. While Medicaid (at least on paper) provides comprehensive, first-dollar coverage, private health plans can require significant cost sharing in the form of deductibles, copays, and coinsurance. If a 30-year-old single adult in an expansion state accepts a better paying job and sees her income rise from 137% to 180% of FPL (an annual increase of about $5,700), she will lose Medicaid coverage but qualify for ACA subsidies. Through the ACA exchanges, she can obtain a silver plan for $276 per year. But silver plans have an actuarial value of 70%, meaning that the typical person will wind up paying 30% of their medical expenses out-of-pocket. For many people, paying an extra $276 per year and 30% of one’s medical expenses to gain an additional $5,700 in income may be advantageous. For others, especially for those with significant health needs or high risk-aversion, forgoing the higher earnings and remaining on Medicaid will be the more attractive option.

A number of studies have tried to empirically quantify the effects of Medicaid on labor market outcomes. Some of the literature is mixed, but the evidence mostly supports the view that the Medicaid cliff discourages work.

Some of the best evidence available predates the ACA, drawing from earlier Medicaid expansions targeting low-income, working-age adults. For example, one study examined the effects of Tennessee’s decision in 2005 to discontinue its expansion of Medicaid, resulting in approximately 170,000 adults abruptly losing coverage over a three-month period. The authors found that the disenrollment caused a large and immediate increase in job search behavior and employment. Another paper found that Connecticut’s expansion of Medicaid in 2010 reduced the employment rate by 3.8 to 4.5 percentage points among low–income childless adults who gained coverage. Yet another analysis found that from 2014-16, the ACA’s Medicaid expansion reduced employment in expansion states by 1.6 percentage points.

A number of reforms could lessen the effects of the Medicaid cliff on labor market incentives.

There are two (trivially easy) ways to eliminate the Medicaid cliff: convert Medicaid to a universal entitlement, available to everyone regardless of income, or eliminate Medicaid entirely. I will not address the merits and demerits of those ideas here. I include them for completeness, but I suspect neither option will strike most readers as judicious.

Short of those radical steps, it is impossible to eliminate the Medicaid cliff altogether. Implicit penalties for higher earnings are inevitable when benefits are phased-out as income increases. 

Nonetheless, the most successful assistance programs (like the Earned Income Tax Credit) minimize these disincentives by tapering benefits off slowly to avoid “notches” where effective marginal taxes spike.

Accordingly, Congress could do more to smooth the transition from Medicaid to private insurance. Medicaid recipients could be permitted to remain on the program as their income rises, provided they contribute a portion of their earnings in premiums. Premiums would gradually rise as income increases until the value of Medicaid (with premiums) and private coverage (with ACA subsidies) are approximately equal. At that point, Medicaid eligibility would end and people would transition to private coverage.

Absent reforms, the Medicaid cliff’s negative effect on the labor market will only grow as the size and scope of Medicaid increases.