Projects
Entry, Exit, and the Changing Composition of the Welfare CaseloadProject Investigator/Key Staff: Steven Haider (PI), Jacob Klerman (Co-PI)
Sponsor: Assistant Secretary for Policy and Evaluation, U.S. Deparment of Health and Human Services
Time Period: September 15, 2000 - September 14, 2001
The welfare caseload evolves through flows onto and off of welfare that can be represented by a Markov structure. This paper explores the dynamic properties implicit in such a system. In particular, we examine steady states, the speed of convergence, and the relative importance of entry and exit to changes in the caseload. In addition, we examine the sensitivity of our findings to different models, including a simple two-state model, a model with duration dependence in the hazard for continuing on welfare, and a model with duration dependence for continuing on and re-entering welfare. We implement these ideas with individual level data for California. The results suggest that adjustment is far from instantaneous: if the entry and exit rates observed in California during 1998 were to persist indefinitely, we would expect the welfare caseload to decline an additional 30 percent more. Moreover, the speed of adjustment is sluggish, with the implicit half-life in the system being over 20 months. Finally, changes in the entry rate are empirically much more important than changes in the exit rate and re-entry rate for explaining changes in the welfare caseload itself; thus, empirical research that focuses only on the factors that explain exit (and possibly re-entry) ignores a substantial share of the actual caseload changes.
Status: OngoingRelated Projects: Project is a continuation of "Insights into the Dynamic Determinants of the Welfare Caseload"

Top