Electronic Publications from the Food Assistance & Nutrition Research Program No. (EFAN-04013) 46 pp

December 2004

Estimating Longrun Food Stamp Program Caseloads

by J. Stephen Clarke, J. William Levedahl, and Albert Reed

This study analyzes the relationship between Food Stamp Program (FSP) caseloads and the macroeconomy using annual State-level panel data for 1980-99. It is the first study to link the time-series properties of the data to an interpretation of public assistance program policy. A longrun relationship involving FSP caseload equation and the macroeconomy is estimated but requires that Aid to Families with Dependent Children/Temporary Assistance for Needy Families (AFDC/TANF) caseloads be included as an explanatory variable. The resulting equation that connects trends in the economy with the trend in FSP caseloads indicates that the economic expansion in the 1990s permanently lowered future FSP caseloads below what they would be otherwise. The potential for the economy to change the trend in FSP caseloads is in contrast to its role presented in previous studies in which the economic expansion of the 1990s is seen as causing only a temporary decrease in FSP caseloads that dies out over time. Tests of the estimated equation also indicate that the common practice of including year effects (annual dummy variables) or State-specific time trend in the FSP caseload equation may overcontrol for omitted variables. Instead, trends in the regressors should be allowed to explain the trend in FSP caseloads.

Keywords: Food Stamp Program, FSP, caseloads, Aid to Families with Dependent Children, AFDC, Temporary Assistance for Needy Families, TANF, State-level data, Food Assistance and Nutrition Research Program, FANRP

In this publication...