Until recently, states were permitted to have different “new entrant” and “continuing recipient” income limits for parental Medicaid eligibility by implementing income disregards that changed with spell length. Some states utilized this option—either tightening income limits for the same family over time or loosening them.

In this article, the authors construct a theoretical model of utility-maximizing workers facing different time-dependent eligibility thresholds to predict the Medicaid participation and employment behavior of workers with varying wage levels. The model reveals some inter-temporally perverse incentives created by linking eligibility thresholds to Medicaid duration. Then, they empirically test these predictions using the Survey of Income and Program Participation and a unique compilation of state-by-family size Medicaid thresholds for both new and continuing recipients.

The authors find that patterns of Medicaid participation and spell duration are consistent with the predictions of their model. There is also evidence that the individuals predicted by their model to lower their work hours may supply fewer hours of labor. As of January 2014, the Affordable Care Act disallows time-varying income disregards; the authors' findings suggest that states previously using this strategy will experience an adjustment in Medicaid caseloads and possibly labor market outcomes because of the change.