Series 14: State Implementation of Work Requirements
On November 7 and 12, 2025, the Ohio Department of Medicaid hosted a pair of webinars that offered the most detailed picture yet of how any large state plans to operationalize Medicaid work requirements. Patrick Beatty, the department’s Deputy Director and Chief Policy Officer, walked through a framework built around a simple insight that Ohio had arrived at years earlier: with nearly 770,000 expansion adults, the state cannot process individual compliance determinations through human review. The math does not allow it. Whatever Ohio builds must be automated first and manual second, or it will not work at all.
This was not a new conclusion. Ohio had reached it during the design of its 2019 Section 1115 waiver, which proposed community engagement requirements verified primarily through administrative data matching. That waiver was approved during the first Trump administration but never implemented because the COVID-19 pandemic intervened and the Biden administration later withdrew approval. The legislature renewed its directive in the FY 2024-2025 state budget (House Bill 33), requiring ODM to pursue a work requirement waiver structured around a 20-hour-per-week standard. ODM submitted its application to CMS on February 28, 2025, the waiver passed its completeness review and entered a federal public comment period ending April 7, and the state waited for approval.
Then the landscape shifted. The One Big Beautiful Bill Act, signed July 4, 2025, established a nationwide Work and Community Engagement Requirement that superseded Ohio’s pending waiver. The federal mandate covers all nonexempt expansion adults ages 19 through 64, requires 80 hours monthly of work, education, training, or qualifying community engagement, imposes semi-annual redeterminations, and sets a hard implementation deadline of January 1, 2027. Ohio’s carefully designed state approach was overtaken by federal requirements that are broader in age applicability, more prescriptive in exemption categories, and paired with enforcement mechanisms that include marketplace exclusion for noncompliance.
ODM’s November webinars acknowledged this new reality. The department’s earlier waiver had proposed annual reporting at redetermination for most members, with quarterly reporting only for those not verified through automated channels. The federal law demands semi-annual verification for everyone. The waiver had proposed requirements for expansion adults under 55; the federal mandate covers adults through 64. The design philosophy remained intact, but the scale and frequency of the challenge had grown.
The Population and the Mathematics#
Ohio’s expansion adult population of approximately 770,000 makes it among the five largest in the nation. ODM’s own analysis, presented during the November webinars, offers a useful decomposition of this population that illustrates why automation is not merely a preference but an operational necessity.
Roughly 43% of expansion adults are known to be working, identifiable through unemployment insurance wage data and other employment records. An additional 18% meet the definition of medical frailty or disability, verifiable through Social Security Administration records, claims data, and diagnosis codes. When all identifiable exemptions are accounted for, roughly 17% can be confirmed as exempt through existing data sources. That leaves approximately 22%, or about 170,000 individuals, who will require additional assessment or documentation to determine whether they meet an exemption or need to demonstrate qualifying activities.
The Center for Community Solutions, a nonpartisan research organization, puts it plainly: those 170,000 individuals are the ones “who have to in some other way show that they’re either meeting the requirement or are exempt.” They represent the zone where Ohio’s automation strategy ends and human verification begins. Whether that transition from automated recognition to manual processing functions smoothly or creates the kind of barriers that caused 18,164 coverage losses in Arkansas in 2018 will determine whether Ohio’s approach succeeds.
The Urban Institute, in its public comments on Ohio’s waiver application, estimated that more than 200,000 Ohioans could lose health coverage. ODM’s own estimate was more conservative at roughly 62,000. The gap between these projections reflects different assumptions about how effectively automation will identify exempt and compliant individuals, and how many of the 170,000 requiring assessment will successfully navigate the documentation process.
The Data-First Architecture#
Ohio’s verification design establishes a hierarchy of automated checks that must be exhausted before any member is asked to provide documentation. This is the architecture’s central principle, and it distinguishes Ohio’s approach from the systems that failed in other states.
Unemployment insurance wage data comes first. Quarterly wage reports from the Ohio Department of Job and Family Services identify employed members. Anyone showing wages consistent with the 80-hour monthly threshold is automatically deemed compliant without any action required. Social Security Administration disability data identifies members qualifying for disability exemptions. SNAP and TANF compliance status provides automatic deemed compliance: members already meeting work requirements under other programs are automatically compliant for Medicaid. Vital records and dependent relationships identify caregivers of young children for exemption purposes. New-hire databases, incarceration data, and Medicare enrollment records fill additional automated verification channels.
Only after these automated checks fail to confirm compliance or exemption does a member enter active verification workflows. At that point, the member receives a verification request accompanied by a detailed checklist specifying exactly what documentation the state needs. Members have 30 days to respond. Failure to provide adequate verification results in denial or termination, though the federal 30-day cure period provides an additional buffer before coverage actually ends.
The promise of this design is substantial. If automation can verify 60-70% of the expansion population without member action, Ohio dramatically reduces the number of people facing active reporting burdens. The risk is equally substantial. If automation covers only 50% rather than 70%, the exception processing system faces significantly higher volumes than planned. And the 170,000 individuals requiring assessment include precisely the populations whose circumstances least fit administrative data categories: gig workers whose income comes through 1099 rather than W-2, home health aides employed directly by families, people caring for elderly parents without formal documentation, and workers in cash economies whose labor is real but whose paper trail is thin.
The Three Ohios#
Ohio’s implementation challenge is not singular. It is at least three distinct challenges playing out simultaneously across geographies that share a state government but little else in terms of economic structure, service infrastructure, or population characteristics.
Metropolitan Ohio, centered on the Cleveland, Columbus, and Cincinnati corridors, contains roughly 60% of the expansion population and the most favorable conditions for automated verification. Formal economy employment is common. Employer-based wage data captures most workers. Service infrastructure is dense. MCOs have established care coordination networks. The metropolitan challenge is primarily about gig workers, informal employment, and the refugee and immigrant populations concentrated in Columbus and Cleveland whose employment patterns may not register in standard data systems. Columbus alone has significant Somali, Bhutanese, and Congolese refugee communities whose workforce participation is high but whose employment documentation may be incomplete.
Small city Ohio presents different terrain. Dayton, Toledo, Akron, Youngstown, and similar legacy manufacturing centers face elevated poverty and unemployment, strained service infrastructure, and economies that have been contracting for decades. The remaining manufacturing jobs are increasingly automated. Service sector employment is precarious. Healthcare systems like Kettering Health in Dayton or ProMedica in Toledo are among the largest employers in their communities, creating a circular dynamic where the hospitals that serve the Medicaid population also depend on it for revenue. Automation will identify fewer compliant members here. More will need exception processing. The question is whether workforce development resources and qualifying activity options exist in sufficient density.
Appalachian Ohio is something else entirely. The 32 counties in southeastern Ohio that fall within the Appalachian Regional Commission’s designation share more with West Virginia and eastern Kentucky than with Columbus or Cleveland. These counties have among the highest overdose death rates in the nation. Employment is scarce, with several counties where the school district and county government are the largest employers. Multi-generational poverty coexists with educational attainment significantly below the state average. Public transportation is essentially nonexistent. Broadband penetration remains low enough that online verification options available in Columbus may be inaccessible in Vinton County.
The substance use disorder treatment exemption will be heavily utilized in Appalachian Ohio, but treatment availability constrains access. Some counties have no medication-assisted treatment providers. The counties with the greatest need for SUD exemptions are the same counties with the least capacity to document and process those exemptions.
County Administration: 88 Variations#
Unlike states with centralized eligibility systems, Ohio delegates Medicaid eligibility determination to 88 county Job and Family Services offices. These same offices handle SNAP eligibility, TANF administration, and workforce development referrals through OhioMeansJobs, creating an integrated local infrastructure that no centralized system can replicate. A caseworker in Athens County may know personally that a particular family’s circumstances make standard compliance pathways impossible and can document exemptions accordingly.
But county administration creates 88 potential variations in how requirements are interpreted, how aggressively compliance is pursued, and how generously exemptions are evaluated. A strict interpretation in one county could cause coverage losses while a flexible interpretation next door maintains coverage for people in identical circumstances. The experience of SNAP ABAWD work requirements offers a preview: county implementation has varied enough that geographic location can meaningfully affect outcomes for people with the same characteristics.
The state faces a genuine tension between respecting county operational autonomy and ensuring statewide consistency in member treatment. ODM’s training, monitoring, and appeals oversight infrastructure must bridge this gap, but the department has never operated anything at this scale. Training materials and partner packets are expected in 2026 as CMS guidance becomes available, leaving limited runway between guidance and the January 2027 deadline.
Ohio’s county structure also creates compounding administrative pressures. The OBBBA simultaneously expands SNAP work requirements by removing exemptions for several population categories, imposes new SNAP error rate penalties, and requires semi-annual Medicaid redeterminations. County JFS offices must implement all of these changes concurrently. Ohio’s SNAP error rate stood at 9.13% in 2025. Under the OBBBA, states with error rates above 6% by October 2027 face penalties requiring them to pay a share of total SNAP benefit costs proportional to their excess error rate. For Ohio, DJFS Director Matt Damschroder has estimated this exposure at more than $300 million annually if the rate does not come down. County offices building new Medicaid work requirement capacity are simultaneously under pressure to reduce errors in the very programs whose verification infrastructure they would leverage for Medicaid compliance.
Managed Care Infrastructure and the MCO Role#
Ohio’s Medicaid managed care program provides a second layer of implementation capacity through five contracted MCOs: CareSource, Molina Healthcare, Buckeye Health Plan (Centene), AmeriHealth Caritas Ohio, and UnitedHealthcare Community Plan. These organizations have existing relationships with expansion adults, established care coordination staff, and data systems that already track member demographics and service utilization.
ODM’s design integrates work requirement support into MCO responsibilities. Care coordinators will identify members at compliance risk during routine contacts. MCOs will provide referrals to workforce development resources. Automated compliance alerts will trigger outreach to members needing assistance. Performance metrics may eventually incorporate compliance rates.
This integration leverages infrastructure that already exists rather than building parallel systems. But it also means MCOs must add work requirement functions to care coordination capacity that was designed for clinical purposes. The semi-annual redetermination cycle doubles the frequency of eligibility disruption that MCOs must manage. Member churn between plans, combined with potential coverage gaps during verification periods, threatens the continuity of care that managed care was designed to provide. MCOs serving members with complex behavioral health needs or chronic conditions face particular challenges: the members most vulnerable to compliance failure are often the same members requiring the most intensive and continuous clinical management.
Ohio’s major healthcare systems, including the Cleveland Clinic, Ohio State Wexner Medical Center, University Hospitals, OhioHealth, and Cincinnati Children’s, have substantial financial interest in coverage maintenance. Hospital assessments constitute a significant portion of Ohio’s Medicaid financing. If work requirements cause substantial coverage losses, hospitals serving newly uninsured populations face increased uncompensated care while continuing to fund the Medicaid program through those same assessments.
The Amish Question and Other Special Populations#
Ohio has the largest Amish population of any state, with approximately 82,000 individuals concentrated in Holmes, Wayne, and Tuscarawas counties. Amish communities present unique verification challenges. Many Amish work in agriculture, construction, and small manufacturing enterprises within their communities, often in employment relationships that do not generate standard wage documentation. Religious convictions may limit interaction with government systems. The verification infrastructure designed for metropolitan workers simply does not map onto Amish economic life.
Ohio’s formerly incarcerated population adds another dimension. The state prison system releases approximately 25,000 individuals annually into communities where employment prospects are limited and reentry services are uneven. These individuals frequently qualify for Medicaid expansion coverage and may also qualify for exemptions, but their circumstances require human assessment that automated systems cannot provide. The transition from incarceration to community, already one of the most vulnerable periods in a person’s life, now includes navigating a work requirement verification system within the same timeframe.
Refugee populations in Columbus and Cleveland present similar challenges. High workforce participation rates coexist with employment documentation that may not appear in standard data systems. Language barriers compound verification difficulties. The communities most likely to be working may be among the hardest to verify as compliant through automated channels.
The SNAP Convergence Problem#
Ohio’s implementation planning cannot be understood in isolation from the SNAP changes occurring simultaneously. The OBBBA removed work requirement exemptions for several SNAP recipient categories, including parents with children ages 7 to 17, individuals ages 55 to 64, and homeless individuals. These changes take effect in 2026 and are processed through the same county JFS infrastructure handling Medicaid work requirements.
The convergence creates compounding administrative pressure. County offices that managed SNAP ABAWD requirements for a relatively small population must now extend work verification to a much larger SNAP cohort while simultaneously building Medicaid work requirement capacity. The 9.13% SNAP error rate looming over $300 million in potential penalties creates institutional pressure to prioritize accuracy over speed, potentially slowing the very processing that Medicaid semi-annual redeterminations demand.
There is also a potential synergy. Members meeting SNAP work requirements are automatically deemed compliant for Medicaid. The cross-program recognition reduces duplicative burden and leverages existing verification. But this synergy depends on data systems communicating effectively between SNAP and Medicaid eligibility determinations, a coordination that works well in theory but that 88 different county offices must implement in practice.
What Remains Unknown#
As of February 2026, Ohio’s framework is best understood as a work in progress. Several critical components depend on federal regulations that CMS will not finalize until June 2026.
How CMS defines “serious medical condition” and what documentation will be required remains unsettled. The length of the look-back window for verifying pre-application compliance is currently a state-level decision, but CMS may prescribe parameters. Whether Ohio’s original January 2026 implementation target (from its 1115 waiver) gives way entirely to the federal January 2027 deadline, or whether the state attempts some earlier rollout, depends on CMS negotiation outcomes. The interaction between Ohio’s pending waiver and the superseding federal requirements creates administrative ambiguity that ODM continues to work through with CMS.
ODM has indicated it will begin member outreach and education several months before implementation, using mailed notices, website updates, text and phone outreach, and collaboration with stakeholder organizations. Partner packets and draft materials are expected in 2026. But the sequence is compressed: federal guidance in June, state operational design in the summer, county training in the fall, and full implementation by January 2027 leaves a margin for error that experienced administrators describe as thin.
Ohio’s outcomes will be watched closely by every large state building automation-centered verification systems. The state has the analytical depth, the administrative infrastructure, and the institutional memory from its 2019 design process. What it has never had is the opportunity to test whether the theory works at scale with real populations facing real consequences. The difference between 60% automation coverage and 50% is not merely statistical; it is the difference between a system that serves most of its population automatically and one that overwhelms the human infrastructure handling exceptions. Whether Ohio’s arithmetic holds is the question that will shape the national conversation about what automation can and cannot accomplish in compliance systems.