This paper addresses the possibility of biased estimates of the determinants of food expenditures of households eligible for the food stamp program. The source of such bias, sample partitioning, is describes and a correction procedure is utilized. For eligible nonparticipants' food expenditure...

This paper demonstrates that income survey data can be very informative in explaining the variation across households in the incidence and severity of absolute poverty using a rural household income data set for Mozambique. Results from regression analysis of the sources of variation are used to...