In September 2018, Arkansas terminated Sarah Martinez’s Medicaid coverage. She worked 35 hours weekly as a nursing home aide in Little Rock, earning $11.50 an hour caring for elderly patients. The state required monthly online reporting to maintain coverage. She had no home computer. The nursing home’s shared staff terminal crashed frequently. The public library closed before her evening shift ended. Over three months she tried to report her hours. Portal timeouts. Password reset failures. System errors. Arkansas saw non-compliance. Federal courts later saw documentation failure among working people. Martinez was one of 18,164 Arkansans who lost coverage in ten months, most of them working or exempt but unable to navigate verification systems designed to catch non-workers.
Seven years later, Arkansas, Georgia, Kentucky, Indiana, Ohio, Kansas, Wisconsin, New Hampshire, and Utah share uncomfortable distinction. All pursued Section 1115 work requirement waivers between 2018 and 2020, building verification systems, establishing exemption processes, creating enforcement protocols. Some implemented. Most didn’t before litigation blocked them. When H.R.1 mandated work requirements nationally in July 2025, these states possessed implementation history their peers lacked. They had evidence, lessons, failures documented in federal court decisions and GAO reports. They knew what didn’t work. December 2026 deadline finds them not starting fresh but confronting harder reality: the infrastructure they built for waiver populations under federal encouragement fundamentally mismatches the infrastructure needed for statutory mandates affecting vastly larger populations under compressed timelines.
The Waiver Legacy: What They Built and Why It Fails Now#
These nine states invested in work requirement infrastructure when implementation was voluntary, populations manageable, and federal flexibility assured. Arkansas built online portals for monthly reporting by 240,000 expansion adults. Georgia constructed Pathways to Coverage for partial expansion reaching roughly 100,000 people at 100% FPL. Kentucky’s waiver, blocked before implementation, developed verification architecture for 500,000 expansion adults. Each design reflected assumptions about state control over enrollment pace, 90% federal match continuation, and CMS permission to experiment with strict verification.
The infrastructure they built emphasized member self-reporting through digital portals, monthly verification cycles, limited navigation support, and verification-before-enrollment models keeping people off coverage until work was proven. Ohio developed sophisticated data-matching systems expecting to automatically verify most workers through wage records. Indiana built on HIP 2.0’s premium payment infrastructure, adding work verification to existing conditionality systems. New Hampshire created employer verification protocols assuming stable employment relationships.
These assumptions no longer hold. Federal mandate eliminates state control over who complies. All expansion adults face requirements regardless of state preferences. Semi-annual redetermination cycles already burden state systems; adding work verification creates compounding administrative load. The 90% enhanced match begins phasing down in 2026, meaning states pay increasing shares of expansion costs while building expensive verification systems. CMS guidance constrains rather than enables experimentation, establishing federal floors states cannot fall below while leaving many details unresolved.
The populations have changed. Waiver-era systems expected to screen applicants before enrollment. Statutory mandates apply to existing enrollees already receiving coverage. Arkansas’s 2018 approach deterred applications; federal requirements cannot deter people already enrolled. Georgia’s Pathways kept enrollment minimal by design (4,100 people in year one); federal mandate affects Georgia’s full 600,000+ expansion population whether the state wants broader enrollment or not.
The evidence from actual implementation makes replication politically toxic. Arkansas’s 18,164 terminations generated federal court rebuke and national media attention. Judge James Boasberg’s decisions found that CMS failed to consider coverage loss when approving waivers. GAO documented that terminated Arkansans were working or exempt; verification system failure, not employment failure, drove coverage losses. Georgia’s Pathways spent more on administration (primarily Deloitte contracts) than healthcare for enrolled members through first 15 months. These outcomes cannot be repeated under federal mandate applying to millions nationally.
The infrastructure they built thus becomes burden rather than advantage. Ohio cannot simply scale its automation approach from planning stages to 770,000 enrollees; the data infrastructure supporting automation requires years of modernization December 2026 doesn’t provide. Kentucky cannot implement county-level hardship exemptions its blocked waiver proposed without federal guidance permitting geographic differentiation federal law may not allow. Wisconsin cannot maintain work requirements only for childless adults when federal law covers all expansion adults. Arkansas cannot use monthly reporting after federal courts found it created unconstitutional barriers.
Arkansas: Learning From Disaster#
Arkansas implemented aggressively in June 2018. Monthly online reporting. Narrow exemptions. Strict enforcement. By March 2019, 18,164 people had lost coverage. Federal courts intervened. Judge Boasberg found the state and CMS had ignored predictable coverage losses among working people. The terminated weren’t refusing to work; they couldn’t navigate digital verification systems that presumed computer access, internet literacy, and stable schedules compatible with monthly reporting deadlines.
Benjamin Sommers and colleagues documented in the New England Journal of Medicine that employment didn’t increase while coverage plummeted. The requirement didn’t change behavior; it changed documentation. People working informally, people with episodic employment crossing 80-hour thresholds unpredictably, people lacking digital access, people with limited English proficiency, people whose cognitive or mental health conditions made monthly bureaucratic compliance difficult despite working, all lost coverage not because they weren’t working but because systems couldn’t see their work.
Arkansas learned. The state’s April 2025 Pathway to Prosperity waiver proposal looks nothing like 2018. Annual reporting replaces monthly. “Success Coaching” through data matching to identify struggling members. But the proposal includes no exemptions beyond federal minimums, suggesting the state still hasn’t fully internalized that documentation burden affects exempt populations as heavily as working populations. A medically frail person struggling with chronic illness may work zero hours but also struggle to navigate exemption documentation proving medical frailty.
The federal mandate forces Arkansas into paradox. Cannot repeat 2018’s strict enforcement after court rebuke. Cannot maintain Pathway to Prosperity’s narrow exemptions when federal guidance may require broader protections. Must build systems serving 350,000+ expansion adults, far exceeding any prior implementation scale. The state’s institutional memory becomes trauma: we know this fails, we have evidence it harms people, we were sued and lost, but federal law now requires we do it anyway.
Georgia: From Deterrence to Mandate#
Georgia solved Arkansas’s problem by creating different problem. Georgia Pathways to Coverage enrolled 4,100 people in year one against projections of 25,000. The system worked as designed, not as described. Make application sufficiently onerous that few people enroll, avoiding coverage loss stories that damaged Arkansas. Annual reporting with work verification. Member Reward Accounts creating financial barriers. Partial expansion to 100% FPL rather than 138%, leaving coverage gap for people between 100-138% ineligible for marketplace subsidies.
GAO’s September 2024 report revealed Georgia spent $110 million on Pathways through first 15 months. $73 million went to administrative costs, primarily Deloitte contracts. $37 million went to healthcare for enrolled members. The state spent more managing 4,100 people than treating them. An additional $10.7 million in American Rescue Plan funds paid for advertising promoting a program that successfully kept people away.
Georgia’s January 2025 waiver extension through December 31, 2026 modified the approach. Monthly reporting becomes annual. SNAP work requirement compliance now counts as qualifying activity. Caregivers of children under six gain exemptions. These changes acknowledge that monthly reporting proved too burdensome and narrow exemptions created hardships, but they come after the deterrence-based design succeeded at limiting enrollment.
Now federal mandate eliminates deterrence option. People are enrolled. Requirements apply. The 600,000+ Georgia expansion adults (if the state pursued full expansion, which it hasn’t) cannot be scared away like Pathways applicants. The state must pivot from making enrollment difficult to managing compliance at scale. Georgia’s actual challenge looks different: managing dual populations under different rules. Pathways enrollees continue under waiver requirements through 2026. If Georgia expanded fully, new expansion adults would face federal requirements from January 2027. This creates two parallel verification systems for similar populations following different procedures.
Georgia’s lesson: deterrence-based systems that succeed at limiting enrollment fail when enrollment is mandated. Annual reporting works for 4,100 people; whether it scales to hundreds of thousands remains untested. MCO infrastructure exists but has never managed work requirement verification. Whether Centene, Peach State Health Plan, and other Georgia MCOs can add verification to existing functions or whether the state must build separate systems affects implementation costs and member experience.
Kentucky: Veto Override and Eastern Kentucky#
The Kentucky General Assembly overrode Governor Andy Beshear’s veto of HB 695 in March 2025, enacting work requirement legislation directing CHFS to pursue waivers. The Democratic governor opposed requirements; Republican supermajority passed them anyway. The June 2025 waiver application targets able-bodied adults without dependents enrolled 12+ months, creating tiered system where new enrollees face immediate requirements while existing members gain one year grace period.
Kentucky’s blocked 2018 waiver included county-level hardship exemptions recognizing eastern Kentucky coalfield counties lack employment work requirements assume. Breathitt County unemployment exceeds 10%. Leslie County has population 10,000 spread across mountain hollows with no public transportation. Formal employment barely exists. You cannot require people to work in counties where work isn’t available, but exempting entire counties undermines reciprocity framework justifying requirements elsewhere.
This geographic tension persists. Kentucky HEALTH data showed proposed exemptions would have covered approximately 70% of expansion population, leaving 30% subject to requirements. But that 30% distributes unevenly. Louisville has jobs. Appalachian counties don’t. Uniform requirements applied statewide create structurally impossible compliance in regions without employment base supporting verification assumptions. Geographic exemptions create politically difficult explanations: why do Breathitt County residents avoid requirements Lexington residents face?
Federal guidance resolves or exacerbates this tension depending on how CMS interprets geographic hardship provisions. If federal regulations permit county-level exemptions based on labor market conditions, Kentucky implements something resembling its original waiver design. If federal law requires uniform statewide requirements, Kentucky faces building verification systems guaranteed to fail in significant geographic areas producing predictable coverage losses among populations facing structural unemployment rather than individual failure.
Kentucky also confronts divided government implementation. Governor Beshear opposed HB 695, but his veto was overridden. CHFS operates under gubernatorial direction but implements legislative mandate the governor opposed. This creates ambiguous signals about implementation priorities. Does Kentucky build rigorous enforcement systems or protection-oriented systems? The answer depends partly on whether administration follows legislative intent or gubernatorial preference when federal guidance leaves discretion.
Ohio: Automation’s Promise and Limits#
Ohio represents the automation dream. ODM November 2025 webinars revealed detailed implementation planning. Of 770,000 expansion enrollees, 43% are already known working through wage record matches. Another 18% meet medical frailty or disability criteria through SSDI/SSI verification. That leaves approximately 170,000 people (22%) requiring further documentation or exemption processing. Automation could handle the majority; human systems manage the minority.
This sounds efficient until you examine what automation actually means. Ohio’s wage record system updates quarterly. Work requirements verify monthly. Temporal mismatch creates gaps. Someone working steadily shows in wage records three months late, potentially losing coverage before verification catches up. Part-time workers crossing 80-hour thresholds sporadically may show some months but not others. Gig workers, independent contractors, informal economy participants may not appear in wage records at all despite working.
Cross-state workers create additional challenges. Someone living in Cincinnati working in Kentucky generates Kentucky wage records, not Ohio’s. Indiana border workers face similar issues. Ohio’s system can’t verify what other states’ systems capture. Multi-state data sharing agreements could solve this, but building them requires coordination Ohio doesn’t control and other states may not prioritize before December 2026.
The 18% showing disability or medical frailty through federal program enrollment looks clear until you consider people whose conditions qualify them for exemptions but who aren’t SSDI/SSI recipients. Mental health conditions, episodic illnesses, chronic pain conditions, functional limitations not meeting SSI criteria but preventing 80 hours monthly work. These people require medical exemption processing through provider attestation, exactly the human-intensive verification automation was supposed to avoid.
Ohio’s SNAP error rate compounds pressure. The state’s 9.13% rate in 2025 must reach 6% by October 2027 or the state faces fiscal penalties. SNAP work requirements overlap with Medicaid requirements, creating opportunity for deemed compliance if federal agencies permit it. But high SNAP error rates suggest verification systems struggle with accuracy. Applying those same systems to Medicaid risks importing errors. The alternative is building separate Medicaid verification divorced from SNAP, eliminating efficiency gains from cross-program coordination.
The synthesis lesson: automation succeeds only when underlying data infrastructure supports it. Ohio’s infrastructure requires years of modernization December 2026 doesn’t provide. The state will implement partial automation knowing it creates coverage losses for populations systems cannot see, or maintain manual verification alongside automation, doubling burden rather than eliminating it.
Indiana, Kansas, Wisconsin: The Conservative Implementation Spectrum#
Indiana, Kansas, and Wisconsin all pursued waivers under first Trump administration. All have Republican-controlled legislatures. All enacted legislation directing Medicaid agencies toward work requirements. But their approaches diverge based on administrative capacity, political culture, and prior experience with Medicaid conditionality.
Indiana’s HIP 2.0 provides foundation and warning. The state required monthly POWER account payments (small premiums) from expansion adults earning above 100% FPL, with coverage lockout periods for non-payment. This created verification infrastructure for financial conditionality. Adding work verification builds on existing systems. But HIP’s lockout provisions generated coverage losses that undermined health outcomes the program sought to promote. Work requirements compound this problem.
Indiana’s SB 2 (enacted 2025) directs implementation building on HIP experience. The state’s statewide MCO infrastructure provides delegation capacity. Anthem, CareSource, MDwise, and MHS already manage risk adjustment, care coordination, and member communications. Adding work requirement verification to MCO responsibilities distributes administrative burden. But H.R.1’s conflict of interest provisions prevent MCOs from conducting compliance determinations if they have financial interest in terminations. This forces Indiana to maintain state-operated verification separate from MCO service delivery, eliminating administrative efficiency MCO delegation was supposed to provide.
Kansas faces divided government complications. Democratic Governor Laura Kelly consistently vetoed Medicaid restrictions during her terms. The Republican legislature passed work requirement legislation but couldn’t override vetoes. Kelly is term-limited in 2026. If Republicans win the governorship, unified conservative government could pursue aggressive implementation. If Democrats retain, divided government continues moderating legislative preferences.
Kansas’s challenge is building systems during political uncertainty. Start building strict enforcement infrastructure and it may be dismantled if Democrats win. Build protection-oriented systems and they may be deemed insufficiently rigorous if Republicans win. The timing (December 2026 deadline, November 2026 election) means Kansas implements during transition between governors, creating institutional confusion about priorities.
Wisconsin operates work requirements under its 1115 demonstration rather than standard expansion, making federal requirements apply even though the state hasn’t technically “expanded” Medicaid. Wisconsin’s partial coverage (childless adults 0-100% FPL) creates smaller affected population than full expansion states, approximately 100,000 enrollees. Smaller scale makes manual verification more feasible than automation-dependent approaches Ohio requires. But partial expansion politics (Act 10 legacy, Republican legislature resistant to full expansion) means implementation occurs in contentious political environment.
These three states illuminate how conservative political culture produces different implementation approaches based on administrative capacity and political stability. Indiana builds on conditionality infrastructure but faces capacity constraints from conflict of interest provisions. Kansas faces political uncertainty shaping system design. Wisconsin manages smaller population under unique demonstration authority. All three implement under conservative political preference for rigorous verification, but their capacity to achieve it varies substantially.
New Hampshire and Utah: Small State Dynamics#
New Hampshire and Utah both pursued waivers, both have Republican governors and legislatures, both have relatively small expansion populations (approximately 50,000 New Hampshire, 150,000 Utah). Small scale creates different implementation dynamics than large states face.
New Hampshire can theoretically provide personalized navigation to every expansion enrollee. The state’s partnership with MCOs could enable relationship-based compliance support if resources were allocated. But small scale also means limited state IT capacity and limited contractor market. Large verification system vendors may not prioritize New Hampshire’s small contract over larger states. Small Medicaid staff must build verification infrastructure while managing existing program administration without proportional capacity increases.
New Hampshire’s waiver design emphasized employer verification for workers with stable employment. This works when someone has single full-time employer willing to provide documentation. It fails for multiple part-time jobs, gig work, informal employment, seasonal work. The state’s tourism economy creates significant seasonal employment (ski resorts winter, lakes region summer) that monthly verification poorly accommodates. Whether federal guidance permits annualized hour calculations or requires monthly thresholds determines how many seasonal workers face inappropriate non-compliance determinations.
Utah pursued partial expansion before shifting to full expansion, creating layered eligibility landscape where some adults qualified at 100% FPL while others qualified at 138% FPL. The state has strong MCO infrastructure (SelectHealth, Molina, Healthy U) but conservative political culture emphasizing personal responsibility. Utah’s July 2025 waiver submission proposed work requirements with target July 2026 effective date, before federal mandate. The state planned to be early implementer by choice; federal law made it mandatory implementer by requirement.
Utah faces religious community dynamics other states avoid. LDS Church emphasis on self-reliance aligns with work requirement philosophy, but Church welfare system also emphasizes community support for members in need. How work requirements interact with religious community support networks affects implementation in uniquely Utah ways. Whether bishops’ assistance counts as employment, whether Church employment programs qualify as job training, whether religious volunteer service counts toward hours, all present questions Utah must navigate that Texas or Massachusetts don’t face.
Small state scale means both personalized potential and capacity constraints. Whether small becomes advantage (relationship-based navigation) or disadvantage (limited IT capacity) depends on how states choose to leverage their size and what resources they invest.
What Federal Mandate Does to Waiver Lessons#
The transition from waivers to mandate changes everything about what implementation experience means. Under waivers, states controlled timing, populations, and could withdraw if implementation failed. Under mandate, federal law dictates timeline, coverage, and states cannot opt out. Early adopters learned lessons optimized for voluntary implementation under federal flexibility. Those lessons transfer imperfectly to mandatory implementation under federal constraints.
Lesson one was monthly reporting creates documentation burdens driving coverage losses among working people. But federal law requires verification at application and semi-annual redetermination. States must verify, whether monthly or semi-annually. The lesson isn’t “don’t verify,” it’s “verification fails when it assumes digital access and bureaucratic capacity vulnerable populations lack.” This lesson transfers: build accessible verification systems regardless of frequency.
Lesson two was deterrence-based systems that limit enrollment avoid coverage loss stories. But federal mandate applies to existing enrollees who cannot be deterred. The lesson doesn’t transfer directly but transforms: if you cannot deter enrollment, you must design verification that accommodates rather than punishes complexity. Recognition-based systems replace deterrence as the mechanism avoiding coverage losses, but building recognition systems requires infrastructure early adopters didn’t create.
Lesson three was MCO delegation distributes administrative burden. But conflict of interest provisions prohibit MCO compliance determinations. The lesson partially transfers: MCOs can provide navigation and support, but verification must occur separately. This creates hybrid systems more complex than either full delegation or full state operation.
Lesson four was narrow exemptions reduce administrative burden from processing exemption applications. But narrow exemptions also drive coverage losses among people who should be protected. Federal guidance may require broader exemptions than states planned. The lesson becomes: exemptions are not administrative burden to minimize but protection essential to avoiding courts striking down implementation as arbitrary.
The synthesis insight is that early adopter experience provides knowledge about what fails rather than templates for what succeeds. These nine states know pitfalls their peers can only theorize about. But knowing what doesn’t work doesn’t automatically reveal what does. Federal mandate requires building new infrastructure informed by but not duplicating what they built before. Experience becomes burden when it teaches you how hard the problem is without showing you how to solve it under different constraints. December 2026 deadline approaches with these states knowing more about failure modes than success paths.