Errors in Survey Reporting and Imputation and Their Effects on Estimates of Food Stamp Program Participation

Receipt of Food Stamp Program (FSP) benefits by eligible people is often underreported in household surveys. Misreporting leads to biased estimates of program takeup and the effects of FSP on well-being and other outcomes. To understand the implications of underreporting, administrative data on FSP participation in two States are matched to American Community Survey (ACS) and Current Population Survey (CPS) data. Results indicate nearly 35 percent of true recipient households in the ACS and 50 percent in the CPS do not report receipt. Estimates of the determinants of program receipt using combined data show that using only survey data results in understated participation by single parents, nonwhites, and very low income households.