Rethinking Medicaid work requirements: From compliance to quality

The disconnect between Medicaid eligibility and work requirements

Starting January 1, 2027, states will need to calibrate the new systems put in place to handle Medicaid work requirements under the One Big Beautiful Bill Act (OBBBA). These systems have to track who needs to work to keep their Medicaid, monitor specific requirements for different people, verify who qualifies for exemptions, and report everything to the Centers for Medicare & Medicaid Services (CMS). The problem is that the federal law only requires proof that someone is working; not that they have adequate or stable work. Right now, the system doesn't connect people to quality job opportunities or help them find work that actually reduces their need for Medicaid. Instead, it often just creates paperwork headaches and coverage gaps. Because the deadline is tight, most states and vendors are scrambling to meet the basic admin requirements. The early data shows that only focusing on the compliance system in not enough. States also need to help people actually search for jobs and understand what benefits they might qualify for if they want this adaptation to work smoothly.

How do work requirements impact Medicaid recipients?

The information drawn from the small number of states that have already implemented a Medicaid work requirement shows nearly one in 4 adults subject to this Medicaid policy reform measure lost coverage, including people who were already working and others managing serious health conditions, with no measurable increase in employment opportunity.¹ More than 95 percent of the people subject to the requirement already met it or should have qualified for an exemption but lost their coverage anyway.² What this illustrates is that coverage lapses are occurring not because recipients refuse to work but more likely due to the reporting process being difficult to navigate.

Another state that moved first under the new federal law this year is showing early signs of a similar pattern.³ Outreach by email and text reportedly failed to reach more than half of the people affected in the initial rollout, and state estimates now suggest tens of thousands could lose coverage. Local hospital and advocacy groups have flagged that as a meaningful risk worth watching closely as more states follow.

How state Medicaid work requirements are being executed

Part of why states keep landing on the compliance-focused model, even when they know it will not produce good outcomes, is that the market has framed this as a choice between two expensive extremes.

Option one: Blind automation. Buy the right interface, automate data matching, and utilize the right portal. It’s an appealing pitch because it requires no judgment about local labor markets and no ongoing accountability for what happens to the person at the other end of the transaction. This model, by design, measures itself on uptime and throughput, not outcomes. This helps explain why the states described above saw disappointing results.

An automation system relies on a blind referral to the state's workforce agency who have their own funding constraints, performance expectations, and enrollment prioritization rules. The system can generate a referral efficiently, can pull back activity data to determine compliance, but it is primarily a blind exchange and does not ensure that the participant is connecting with the correct employment on the other end. It is focused on activity measurement rather than quality outcomes. Automation is well suited for the population of people who are already working and meeting the requirements through their current job. They need no job search assistance, only a reliable way to verify hours and wages they are already earning.

Option two: Expensive and intensive case management. Dedicated coaches, individualized plans, barrier remediation, ongoing caseloads, the kind of model many states already know is costly from their own Temporary Assistance for Needy Families (TANF) experience. Across the country, states routinely spend millions of dollars a year on TANF work programs and still see participation rates in the range of only 15 to 20%.⁴ States are still able to meet their federal TANF performance requirements at rates that are low because of caseload reduction credits, not necessarily for getting people into stable jobs. Scaled to a Medicaid expansion population that can run into hundreds of thousands per state, that same intensive, per-case model is simply not affordable under H.R. 1's limited implementation dollars. States know it, and default back to a basic compliance model rather than face that price tag.

However, a state does not need to choose between an automated compliance shell, and a case management build it cannot afford. What it needs is a right-sized local design effort focused on self-directed job search, layered on top of whatever compliance technology it has already bought. This can scale to each county’s economic conditions and partner capacity rather than be applied generically and uniformly across the state.

Medicaid compliance vs. accessibility: What a quality program requires

• Local design. A statewide tool and a statewide agency partnership, however well built, cannot capture what is different about a rural county with 3 major employers and limited transit. Design should happen at the county level, starting from what already exists; the work is based on current infrastructure instead of duplicating it.
• Evaluation. Federal compliance reporting will show CMS who is exempt, compliant, or not eligible. It will not show whether a job search program was any good or how many people found employment. Programs should build evaluation frameworks around the outcomes that actually matter: placements, retention, employer health insurance, and income. An exit from related benefit often aligns to the same performance indicators the workforce system already uses, so states are not building a new measurement language from scratch.
• Workload and cost. Compliance reporting shows whether a case was processed, not what processing actually took. Quality review should also track how managing the work requirement is affecting renewal workflows, caseworker workloads, case processing timeliness, overall caseload changes, and program costs, so a state can see the operational toll of the requirement alongside its effect on participants.

Proof from Pennsylvania

Pennsylvania has built an award-winning technology solution that automated compliance reporting for non-custodial parents' employment participation through a direct interface with the state's workforce system. That automation was based on the states' experience running the highest-performing child support program in the country. This tool made verified job search data available to caseworkers and meant the child support program could rely on a referral to the workforce system rather than building and duplicating its own job search infrastructure.

Once counties received the support to design how that process worked at the local level, they were able to match it to their own job market and to the specific workflows of their county offices and community partners. That local design layer, not the interface alone, is what turned a well-built automated tool into a program that moved people into jobs.

That same pattern, a strong statewide tool paired with a local design and implementation support, is precisely what Medicaid work requirements will need, at a scale many times larger than any state’s child support or TANF caseload. It is also precisely where a short, inexpensive, county-specific design engagement closes the distance between a working compliance system and a job seeker’s actual experience.

Medicaid compliance challenges: The choice in front of every state

States that treat the Medicaid work requirement purely as a compliance build risk the same pattern seen by some of the early adopters: administrative churn, coverage loss concentrated among people who were mostly already compliant or exempt, and no employment gain to show for it. A state that invests in quality, local design, evaluation, and performance, at a cost sized to what it can sustain, has a real chance to reduce its own long-term Medicaid, SNAP, and TANF caseloads.

Getting a quality program does not require the expensive, case-management-intensive model described earlier. It can remain a self-directed job search, but one designed to minimize staff workload, match local conditions and existing workflows, and improve outcomes. Costs stay contained through careful design and by ensuring the program does not duplicate what is already in place. A mix of automation, referral, and targeted local partnerships have the potential to maximize the Medicaid work requirement eligibility impact for participants entering into these services as a result of the work requirement.

That is the choice every state is making right now, whether it has explicitly weighed automation against a case-managed job search model or simply defaulted into one.

CAI’s deep expertise in guiding health and human services agencies to the resolutions necessary for their most complex policy and program challenges will be a great asset when it comes to crafting a right-sized, county-level design effort for these upcoming changes. To learn how CAI has helped address a variety of complex shifts in the current landscape, complete the form below to discover more.