One in five working Medicaid enrollees could fail work reporting by working the wrong week

Most of the argument about Medicaid work-reporting requirements has been an argument about paperwork: whether enrollees who are working will lose coverage because they cannot navigate the reporting system. A cross-sectional study published this month in JAMA Health Forum makes a different and narrower point. Even setting paperwork aside, a substantial minority of working enrollees do not work a schedule the requirement can reliably recognise [s1].

What the requirement is

House Resolution 1 of the 119th US Congress, known as the One Big Beautiful Bill Act, created national work-reporting requirements for adult Medicaid expansion enrollees, to begin by 1 January 2027 [s1]. The threshold the study evaluates is 20 hours per week [s1].

The authors' stated motivation is that the expected driver of disenrollment has been administrative burden, while the effect of variable hours or part-time work on compliance among low-income working adults was unknown [s1].

What the study did

The analysis used the 2023 to 2025 Current Population Survey to describe employment among adults aged 18 to 64 who were plausible Medicaid-expansion enrollees, then estimated the prevalence of being at risk of non-compliance and used modified Poisson regression to test whether that risk differed by demographic characteristics [s1]. Data were analysed from February through June 2026 [s1].

The analytic sample was 19,476 observations, representing a mean annual weighted population of more than 16 million people: mean age 41.1 years (95% CI 40.8–41.3), 49.8% female and 50.2% male, 71.8% unmarried, 47.9% non-Hispanic White, and 41.0% with a high school education or equivalent [s1].

"At risk of non-compliance" was defined in two ways, which overlap: usually working near the hours threshold, or reporting variable hours [s1].

What it found

Among those potentially subject to work-reporting requirements and working, an estimated 19.8% were at risk of non-compliance — 13.6% by usually working near the hours threshold, 7.6% by reporting variable hours [s1].

Risk was not evenly distributed. Female respondents had a 22% higher probability of being at risk (adjusted risk ratio 1.22, 95% CI 1.11–1.35), while married respondents had an 18% lower probability (aRR 0.82, 0.74–0.92) [s1]. The authors also report significant variation by race and ethnicity and by education level [s1].

The study's own summary is worth quoting for its balance: it found near full-time mean hours per week among working adults enrolled in or likely eligible for Medicaid expansion, and nearly one-fifth at risk of non-compliance from insufficient or variable hours [s1]. Both halves are the finding. Most of this population works close to full time. A fifth of them work close to full time in a way a monthly hours test can miss.

Why the distinction matters

An hours threshold assumes hours are a stable property of a job. For a substantial part of the low-wage labour market they are a scheduling output: retail and food service shifts assigned week to week, hours cut in a slow month, a second job that fills some months and not others. Someone averaging just above 20 hours across a year can fall below it in any given month without changing employer, occupation or effort.

That is a different failure mode from not filing paperwork, and it is not fixed by better paperwork. A worker who correctly reports fewer hours than the 20-hour weekly threshold in a given month [s1] is compliant with the reporting system and non-compliant with the requirement.

The sex difference in the estimates is consistent with the same mechanism rather than being an independent finding: part-time and variable-hours work is not evenly distributed between men and women, and the study reports women at higher risk on both counts [s1].

What the study does not establish

This is a cross-sectional estimate of who is at risk, not a projection of how many people will lose coverage. It does not model exemptions. It does not model state variation in how hours are verified, or the use of data matching that could confirm employment without an enrollee filing anything. It draws on survey-reported hours, which carry their own error, and it identifies "plausible" expansion enrollees rather than actual ones. Any of those could move the estimate in either direction.

The other half of the burden

A companion analysis in the same journal quantifies the administrative side directly. Using natural language processing on care-coordination encounter notes from a community-based programme covering Medicaid managed care beneficiaries in Washington, Virginia and Ohio between January 2023 and November 2025, it classified four documented burden types: scheduling difficulties, transportation problems, paperwork and documentation requirements, and prior authorisation delays [s2]. Of 142,473 beneficiaries enrolled in the four participating managed care plans, 49,282 (34.6%) completed at least one care coordination encounter [s2]. Patient-level prevalence of documented transportation burden alone was 6.1% (95% CI 5.9–6.3) [s2]. The study valued patient time at three rates — the federal minimum wage of $7.25 an hour, a state-weighted living wage of $22.00, and the Resource-Based Relative Value Scale conversion factor of $33.40 — precisely because that time is invisible in claims data [s2].

What to watch

The requirement begins nationally by 1 January 2027 [s1]. The measurable question in the first year is whether disenrollments concentrate among people who were working — which is what a variable-hours failure mode, rather than a paperwork failure mode, would produce.

Sources

• Medicaid Work-Reporting Requirements Under HR 1 and Insufficient or Inconsistent Work Hours, JAMA Health Forum, August 2026
• Administrative Burden Documented in Medicaid Care Coordination, JAMA Health Forum, August 2026