Arkansas showed how people fall off a Medicaid work requirement. Georgia's Pathways to Coverage, launched in 2023, shows a related failure: how a high-friction front door keeps eligible people from ever getting on. Both lessons matter for states building toward 2027, because the same design choices drive both outcomes.
Enrollment far below projection
Georgia projected that tens of thousands of people would enroll through Pathways, which ties coverage to documented qualifying activity. Actual enrollment came in dramatically lower than initial projections, a small fraction of the eligible population in the program's early phase. The gap was not a lack of eligible people. It was the difficulty of proving and continually re-proving activity to get and keep coverage.
The structural lesson is that a requirement to document hours, applied at the front door, functions as a barrier even for people who are genuinely working. If the proof process is burdensome, eligible people simply do not complete it.
Administrative cost is the quiet line item
The second Georgia lesson is about cost. Reporting on Pathways indicated that a large share of program spending went to administration and systems rather than to member medical care. When a program requires ongoing verification of activity, someone has to build and run the machinery to collect, check, and adjudicate that information, and that machinery is expensive.
For states implementing H.R.1 requirements, this reframes the budget conversation. The cost is not only the eligibility-system build. It is the recurring operational cost of processing documentation, handling appeals, staffing call centers, and producing notices, repeated every reporting cycle for every subject member.
The takeaway is not that requirements cannot be implemented well. It is that friction is the enemy of both coverage and cost control. Every step that makes reporting harder for the member simultaneously lowers enrollment and raises administrative spend. States that want better numbers on both fronts should design for the lowest-friction path that still meets the federal standard, pre-fill what they can, default to automatic exemption where data supports it, and reserve manual processes for genuine edge cases.