P F Short1, V A Freedman. 1. Center for Health Policy Research, Pennsylvania State University, University Park 16802, USA.
Abstract
OBJECTIVE: To investigate transitions in and out of Medicaid for a cohort of single adult women of childbearing age in order to address questions that arise as policymakers try to encourage transitions from welfare to work. DATA SOURCES: Longitudinal data from Waves 2 through 8 of the 1990 panel of the Survey of Income and Program Participation, a nationally representative survey of American adults covering May 1990-1992. STUDY DESIGN: We estimate a series of discrete-time logit models with duration dependence to obtain transition probabilities among Medicaid, privately insured, and uninsured spells. Explanatory variables in the models include prior insurance history, income limits on Medicaid by state, and important socioeconomic and demographic characteristics. We use these models to characterize insurance spells for a cohort of single women. PRINCIPAL FINDINGS: Most Medicaid spells are relatively short. Over half end in a year or less; only one spell out of seven lasts longer than five years. Two-thirds of Medicaid disenrollees become uninsured. Former welfare recipients are prone to frequent changes in insurance status. In states with more generous income limits for AFDC, women stay on Medicaid longer, but they do not move into the program at a faster rate. CONCLUSIONS: Imposing time limits on Medicaid eligibility would affect only a small proportion of Medicaid spells but would eliminate a significant proportion of the caseload at a point in time. In considering changes in Medicaid that would encourage transitions from welfare to work and would alter the dynamics of Medicaid, policymakers need to consider how transitions both in and out of private insurance and Medicaid would be affected.
OBJECTIVE: To investigate transitions in and out of Medicaid for a cohort of single adult women of childbearing age in order to address questions that arise as policymakers try to encourage transitions from welfare to work. DATA SOURCES: Longitudinal data from Waves 2 through 8 of the 1990 panel of the Survey of Income and Program Participation, a nationally representative survey of American adults covering May 1990-1992. STUDY DESIGN: We estimate a series of discrete-time logit models with duration dependence to obtain transition probabilities among Medicaid, privately insured, and uninsured spells. Explanatory variables in the models include prior insurance history, income limits on Medicaid by state, and important socioeconomic and demographic characteristics. We use these models to characterize insurance spells for a cohort of single women. PRINCIPAL FINDINGS: Most Medicaid spells are relatively short. Over half end in a year or less; only one spell out of seven lasts longer than five years. Two-thirds of Medicaid disenrollees become uninsured. Former welfare recipients are prone to frequent changes in insurance status. In states with more generous income limits for AFDC, women stay on Medicaid longer, but they do not move into the program at a faster rate. CONCLUSIONS: Imposing time limits on Medicaid eligibility would affect only a small proportion of Medicaid spells but would eliminate a significant proportion of the caseload at a point in time. In considering changes in Medicaid that would encourage transitions from welfare to work and would alter the dynamics of Medicaid, policymakers need to consider how transitions both in and out of private insurance and Medicaid would be affected.