Food Stamp Program Participation Behavior - An Econometric Explanation

Abstract

The effect of selected socioeconomic factors on household participation in the Food Stamp Program was estimated. The demand relation for food stamp bonus is conceptually derived via conventional utility theory. The sample encompasses eligible participants and nonparticipants. Thus, bonus, which quantifies extent of participation as the dependent variable, is zero for nonparticipants. For this reason the Tobit model is used in estimating the demand relation in lieu of ordinary least squares. Fair's computational procedure, which is reportedly faster than Newton's, was used. Impact of family size and level of income on extent of bonus use, was significant. Other economic factors which have a similar impact are home ownership and type of income such as wage and social security. Age and sex of the household head were also important variables. Applications of the statistical relation include participation prediction, elasticities, and probability changes associated with isolated exogenous variable changes. The estimation procedure for the model framework yields encouraging results. Thus, a more general application of the framework may be of interest in evaluating participation in other types of programs.