Have electronic benefits cards improved food access for food stamp recipients?

This paper investigates Electronic Benefits Transfer (EBT) card reforms in California’s food stamp program, and the impact on food insecurity. We test the hypothesis that EBT cards reduce food insecurity by reducing the food costs associated with loss and theft of benefits, as well as by decreasing fraudulent sales of benefits. We use a natural experiment in the form of the time-varying roll-out of EBT card reforms across California counties in conjunction with multiple food security measures as measured by the California Health Interview Survey, to conduct a nonparametric event study. Our findings suggest no evidence for a causal decrease in food insecurity stemming from implementation of EBT card reforms. We do, however, find causal evidence of a transitory increase in food insecurity immediately following implementation of EBT reforms. Reforms increase the likelihood of food insecurity by about 3% for at least 2 months. The result is distinguishable from zero, and highly robust to changes in specification, inclusion of controls, and measurement choices. We find evidence that the increase was due to informational frictions in the transition to EBT card systems. Namely, program participants were ill prepared for the transition, and in many cases did not believe the announced dates for program reform. Our findings are not supportive of many policy makers’ claim that a positive externality of EBT benefits delivery is a lasting reduction in food insecurity. Moreover, our findings suggest that policy makers must consider the transparency and credibility of proposed changes to social welfare programs in order to avoid deleterious outcomes stemming from imperfect information.