The monthly food stamp cycle: Shopping frequency and food intake decisions in an endogenous switching regression framework

This dissertation for the first time uses nationally representative data to describe and measure the monthly cycle in food expenditure and food intake for U.S. food stamp recipients. Mean food spending peaks sharply in the first few days of the food stamp month, counting from the day that benefits are received. Mean food intake exhibits a more modest monthly cycle, which depends on the frequency of food shopping. For households that conduct major grocery trips more than once per month, mean food energy intake appears smooth over the whole month. For households that shop less frequently, mean food energy intake is significantly lower in week four of the food stamp month than in week one.;For frequent shoppers in both halves of the month, and for infrequent shoppers in the first half of the month only, food energy intake reaches a relatively high level, and further increases in food stamp benefits have no significant marginal effect on food energy intake. However, food energy intake is lower for some recipients: (1) for most recipients with very low benefit levels, and (2) for infrequent shoppers in the second half of the month, at any level of food stamp benefits.;Two simulations indicate the policy implications of the econometric results. The first suggests that different policies for changing food stamp benefit levels may have different effects on expected food intake, even holding constant the change in program costs. The second simulation suggests, for households with "typical" food stamp benefit levels (in the middle two quartiles), that policies to increase the incentive to shop frequently could reduce or eliminate the monthly cycle in food energy intake.;These descriptive patterns motivate an economic model that simultaneously accounts for the shopping frequency decision and the food intake decision in two halves of the food stamp month. An econometric model, based on this theoretical framework, accounts for both the endogeneity of the shopping regime choice and the potential heteroskedasticity in the disturbances of the food intake functions. Similar results are derived using two distinct functional forms for the effect of food stamps on food energy intake.