What causes public assistance caseloads to grow?

This paper uses state panel data to investigate changes in public assistanc e caseloads through the end of the AFDC program in 1996, with particular at tention to the rapid increase in caseloads between 1990 and 1994. Previous research has focused on total caseloads, with attention to economic and pol icy variables, and does a relatively poor job of explaining this caseload i ncrease. This paper utilizes a much richer set of control variables to inve stigate the causes of caseload change: it separates AFDC caseloads into thr ee subcomponent programs, separately investigating changes within the AFDC- UP program, AFDC child-only cases, and the remaining "core" AFDC cases (wit h benefits paid to single mothers and their children); and it explores whet her this caseload rise was driven by changes in takeup rates versus in elig ibility. The results indicate a large unexplained rise in total AFDC caselo ads, even with a very rich specification, A good share of this is due to sh arp increases in child-only cases, driven by program and demographic shifts . To a lesser extent, this rise was caused by the expansion of AFDC-UP to a ll stares. These two factors explain half of the overall rise, and all of t he unexplained rise in AFDC caseloads. The remaining increase in "core" AFD C cases-benefits received by single mothers and their children-is well expl ained within the model, and is the result of economic, demographic, politic al and policy changes. These variables appear to have increased eligibility among the core AFDC population in the early 1990s during the economic slow down. Takeup rates also increased during this rime period but nor by a larg e amount.