The debate over Medicaid work requirements has been captured by a false choice: either the requirements are morally essential and empirically proven, or they are cruel and demonstrably counterproductive. Neither caricature serves the millions of low-income Americans whose long-term health and economic prospects depend on getting the policy right.
The Congressional Budget Office projects the Medicaid work requirement provisions in the One Big Beautiful Bill Act (OBBBA)—dubbed “community engagement”—will reduce federal spending by $326 billion over ten years, reflecting the dramatic Medicaid expansion since 2014. The question now is not whether to implement these requirements, but how to implement them in a way that advances the policy’s underlying goals: more employment, less dependency, and stronger communities.
The Centers for Medicare and Medicaid Services (CMS) has framed community engagement as having the potential to “empower Medicaid beneficiaries through employment, education, or volunteer service so they can escape isolation and dependency, build confidence, and achieve self-sufficiency.” That framing is not simply rhetorical cover; it reflects a coherent theory about what long-term well-being actually requires. Health insurance is a means to an end. The end is a thriving life, which cannot not be built on government dependency alone. A person who is working, connected to their community, and investing in education or skills development is almost certainly healthier, more resilient, and less likely to need acute care than one who is enrolled in Medicaid but disengaged from civic and economic life. The requirement makes that connection explicit.
What the evidence shows—and what it doesn’t
Critics lean heavily on the Georgia Pathways experience and the Arkansas pilot to argue the community engagement fails on its own terms. These cases deserve a fair reading, but they prove less than the critics claim.
In Georgia’s first year, just over 4,200 people enrolled in Medicaid through the Pathways program, far below the state’s initial projection of 100,000. Georgia’s stringent documentation and reporting requirements likely contributed to low enrollment. Arkansas, which ran requirements for just nine months before a court order ended the program, saw approximately 18,000 people lose coverage, a number frequently cited as evidence of harm. What this proves is that poorly designed reporting infrastructure will disenroll eligible people. It does not prove that the underlying policy is wrong.
The broader literature on work requirements across analogous programs is instructive. The CBO’s analysis of Temporary Aid for Needy Families (TANF) found that making benefit receipt contingent on working substantially increased employment during the year recipients entered the program: a finding that gets regularly omitted from critiques of work requirements. TANF’s work requirements are now nearly three decades old. The employment norms they helped establish among low-income families are part of what made the post-welfare-reform era one of the largest reductions in poverty rates in modern American history. The policy worked. Why its success hasn’t transferred cleanly to Medicaid has far more to do with implementation than with the idea itself.
Research on the Supplemental Nutrition Assistance Program’s work requirements finds that many studies measure employment outcomes using only unemployment insurance-covered wages, which miss informal and self-employment. That missing data may reflect 20 to 50 percent of prime-age adults’ work activity. This matters enormously for Medicaid: expansion enrollees disproportionately work in informal sectors, gig arrangements, and seasonal jobs that generate real economic output but leave few administrative footprints. A reporting system designed around formal W-2 employment will systematically misidentify compliant workers as non-compliant. That is not a problem with work requirements; it is a problem with a reporting architecture built for a 1990s labor market. Thus, the primary risk of OBBBA’s work requirements is not that people who should lose coverage will keep it; rather, it’s that people who meet the spirit of the requirement will lose coverage because the verification systems aren’t built to recognize how they actually work.
Building on what the RHTP got right
In my September analysis of the Rural Health Transformation Program (RHTP), I argued that the OBBBA’s rural provisions represent a meaningful federal commitment to communities that have been structurally disadvantaged by decades of consolidation and underinvestment. The RHTP allocates $50 billion in grants over five years, empowers states to apply for funds without requiring a state match, and directs resources toward emergency care, labor and delivery, behavioral health, workforce development, and infrastructure upgrades. The work requirement provisions and the RHTP are not separate policies; they are complementary limbs of the same body.
Consider the rural enrollee who works seasonally in agriculture, logging, or construction. He works hard, but his employment is episodic, and his employer doesn’t file quarterly unemployment insurance records in any format that an automated Medicaid eligibility system will catch. He is exactly the person the RHTP is trying to serve with better rural health care infrastructure. He is also exactly the person at highest risk of procedural disenrollment under a poorly built reporting system. Getting the implementation right for him is not a concession to the opponents of work requirements, it is the definition of implementing the policy faithfully.
The implementation agenda: Five priorities for CMS and states
CMS is expected to issue its final implementation rule by June 1, 2026, with states required to comply by January 1, 2027, though states may request a good-faith extension to December 31, 2028. That nine-to-twenty-one month window is tight. Here is what states and CMS should prioritize.
1. Build data-matching infrastructure before enforcement begins. The single most important lesson from Arkansas and Georgia is that passive reporting systems—which require enrollees to affirmatively document compliance each month—fail at scale. In both Arkansas and Georgia, operational challenges including system glitches, unclear reporting processes, and limited support options created significant access barriers, and states underestimated the staff and training needed to support implementation. CMS should require, not merely encourage, states to pursue automated data matching with state workforce agencies, SNAP employment and training programs, and IRS earnings records before their enforcement dates. Verification should happen behind the scenes wherever possible.
2. Recognize informal and gig work through self-attestation with audit sampling. States should be permitted and actively encouraged to accept self-attestation of work activity for categories of workers whose employment is structurally unlikely to appear in administrative records. This is how the earned income tax credit handles gig income. There is no policy reason Medicaid compliance should be held to a stricter evidentiary standard than federal tax administration. States can implement random audit samples to deter fraud.
3. Pair work requirements with the RHTP workforce agenda. The RHTP explicitly includes workforce development as an allowable use of grant funds. States should channel a meaningful share of their RHTP allocations into sectoral training pipelines—in particular health care, construction, and manufacturing—that generate both workforce supply and qualifying activity credit for Medicaid enrollees simultaneously. A rural Medicaid enrollee who completes a certified nursing assistant training program funded by the RHTP satisfies both the community engagement requirement and the RHTP’s workforce mandate in a single intervention.
4. Expand the exemption for high-unemployment areas. Under OBBBA, states have the option to request exemptions for areas with declared emergencies, disasters, or high unemployment. This is wise policy. Work requirements make sense in labor markets where work is available, but they impose unnecessary hardship in areas with sustained joblessness. CMS should make the high-unemployment exemption administratively easy to invoke, with automatic triggering when county unemployment exceeds 1.5 times the national rate for two consecutive quarters. This is not a backdoor opt-out; it is a precision tool.
5. Design re-enrollment to be frictionless. One of the underappreciated features of OBBBA is that it makes re-enrollment harder than prior waiver-based programs. The law’s work requirement policies are more stringent than previous policies considered by Congress, making it harder to gain coverage and to re-enroll, and does not exempt older adults from requirements. For people who lose coverage due to administrative failures—not genuine non-compliance—re-enrollment friction is counterproductive. States should establish 90-day re-enrollment windows with presumptive eligibility for recent disenrolled individuals who can document retroactive compliance.
The stakes are high and the policy is worth defending
The conservative case for work requirements does not rest on indifference to coverage losses. It rests on a belief that the structure of dependency itself produces worse health outcomes over time, that work and civic engagement are genuine goods, and that a well-designed program can expand both. Evidence from the CBO and others confirms that TANF’s requirements substantially increased employment—not perfectly, not for everyone, but meaningfully and durably. Medicaid requirements can do the same if states treat the implementation phase as the policy itself, not as an afterthought.