Over the past fifteen years, the U.S. government has implemented a series of changes in the social safety net that dramatically altered the provision of public health insurance to low-income women and children. Beginning in the late 1980s and early 1990s, laws were passed that expanded Medicaid eligibility for poor pregnant women and for poor children. Welfare system changes implemented during the mid-1990s encouraged welfare recipients to work and leave welfare, but assured them that public health insurance benefits would continue. More recently, the commitment to publicly-provided health insurance was expanded further with the passage of legislation authorizing the largest increase in public spending on insurance for children in three decades. Despite interest by policymakers and previous researchers in examining the effects of these changes, important questions remain about their implications. To improve understanding of the impacts of publicly-provided health insurance, this proposed research has three specific aims. First, to examine the factors determining the extent to which eligible children enroll in Medicaid. Second, to examine how changes in the relationship between cash assistance and public health insurance for single mothers affect the health insurance coverage of single mothers, particularly the fraction of time spent with private insurance, public insurance, and uninsured. The effects of these changes on the welfare participation and employment of single women will also be examined. Third, to estimate the impact of expanded public health insurance for pregnant women and children on fertility. Detailed panel data on insurance, income, employment, welfare participation, and fertility from the Survey of Income and Program Participation (SIPP) will be used. The Survey of Program Dynamics, the 1990 Census microdata, and a survey of closed welfare cases from Ohio will also be used for some analyses. The data will be analyzed using a variety of estimation techniques appropriate to panel data, including descriptive analyses, static models, switching models and duration models. As eligibility for Medicaid may be endogenous, exogenous variation in the impact of federally-mandated expansions by state, time, and age as well as business-cycle variation will be used to identify the models.
