Gordon F. De Jong Deborah J. Graefe, Pennsylvania State University
This study examines whether devolution of welfare policy in the 1996 welfare reform legislation created new state benefits and rules inequalities that engender inter- and intra-state migration among distinct groups of welfare poor families, specifically African American and white single mothers and immigrant families. This study uses data from three sources-- the 2001-2004 panel of the Survey of Income and Program Participation (SIPP), the Urban Institute's Welfare Rules Database and a local labor market characteristics file created from 2003 American Community Survey census data. These databases will be merged to create a longitudinal, two-stage specification of welfare-benefit "push" and "pull" impacts on poor immigrant and single-mother-headed families' migration behavior. Using a state welfare policy inequality framework, the investigators will use factor analysis to develop measures from post-2000 textual policy manuals to operationalize multiple welfare benefit and eligibility rule dimensions and to test hypothesized state program effects on migration. A discrete-time event history analysis will be used to predict migration events (inter- and intra-state migration) in the SIPP data. The investigations will use a multi-level hierarchical modeling strategy, which is an integrated, and previously untested, micro-macro analysis of three determinant-of-migration hypotheses for welfare poor families. These tests evaluate effects of 1) time-varying state welfare policy characteristics; 2) individual and family characteristics, including fixed-time immigrant and marital status and race, from the information-rich SIPP files, and 3) local labor market-level economic opportunity structure indicators. Similar to previous studies, a separate analysis of push and pull migration effects of hypothesized co-variates will be completed first through, a "destination model" for identifying pull effects, and then, a "departure model" which identifies push effects for potential migrants' origin locations grouped by class of states according to state policy criteria. This two-stage model with state welfare policy, local labor market, and individual and household indicators will provide a strong test of the welfare policy impact on migration.
