Georgia Medicaid Work Requirement Linked to Worse Mental Health

Key Takeaways

• Implementation of Pathways was associated with worse self-reported mental health among adults aged 18 to 64 years with household income at or below 100% of the federal poverty level.
• Respondents were more likely to report ever being told they had a depressive disorder after implementation than similar adults in neighboring nonexpansion states.
• Event-study estimates generally supported the main analysis, and robustness checks were either consistent with the main findings or null where no association was expected, while the authors noted limits related to self-report, unobserved coverage transitions, and possible residual confounding.

A JAMA Network Open economic evaluation associated Georgia’s Pathways to Coverage with 4.0 additional days per month when mental health was not good among low-income adults, compared with neighboring nonexpansion states. Reporting of ever being told one had a depressive disorder also increased after implementation.

Researchers conducted a cross-sectional difference-in-differences analysis using Behavioral Risk Factor Surveillance System data. The sample included adults aged 18 to 64 years with household income at or below 100% of the federal poverty level in Georgia, Alabama, Mississippi, Tennessee, and South Carolina. Pregnant respondents and caregivers were excluded, but disability-related exemptions could not be fully identified. The dataset included 8138 respondents, with analytic samples of 7071 to 7259, a mean age of 46.4 years, and 21.6% residing in Georgia. Pathways began on July 1, 2023, and the postperiod started July 11 after March 29 through July 10 was removed for the continuous-enrollment unwind.

The analysis examined past-month days when mental health was not good and whether respondents had ever been told they had a depressive disorder. In adjusted difference-in-differences models, Pathways implementation was associated with 4.0 additional days per month for poor mental health (95% CI, 2.0-6.0; P < .001). It was also associated with a 7.2 percentage point increase in ever being told one had a depressive disorder (95% CI, 2.4-12.0; P = .003). Overall, the pattern was consistent with worse self-reported mental health after implementation.

Event-study estimates before the policy were not jointly different from zero, although a small number of prepolicy coefficients were statistically significant in isolated periods. Comparison analyses among respondents above 100% of the federal poverty level, those with private insurance, and those assigned a July 2019 placebo date did not show significant associations. Results were also similar when the preperiod was limited to January 2020 onward, and Supplement 1 showed comparable findings with midpoint-income coding. As policy context, Georgia required at least 80 hours each month of work or other qualifying activities, with exemptions for pregnancy, school or job training, caregivers, and disability-related circumstances. Early enrollment was 4231 against an initial projection of 100000, and the supplementary analyses largely matched the main pattern.

The measures were self-reported, and the depressive-disorder outcome reflected lifetime history rather than recent symptoms. Behavioral Risk Factor Surveillance System data could not capture individual Medicaid coverage transitions, and the analysis did not assess mental health care use or events such as suicide. Some exempt individuals may have remained in the sample because disability-related exemptions could not be fully identified.

The authors also noted possible residual confounding from concurrent state-level policy changes. They concluded that the findings suggest Pathways may have worsened mental health among low-income adults, with potential implications for population health and equity.