Low participation rates among low-income people eligible for food stamp benefits have prompted a number of outreach and public education efforts. In 2002, the Food and Nutrition Service awarded $5 million in grants to community-based organizations in 15 States to investigate how to increase participation among people eligible for food stamp benefits. The evaluation of these grants describes the features and outcomes of these 18 projects.

Most discussion of payment accuracy in the Food Stamp Program focuses on the overall level and cost of payment errors. Rarely does the discussion focus on the impact of payment errors on individual households affected. This analysis – based on 2003 food stamp quality control data – leads to two broad conclusions. First, virtually all households receiving food stamps are eligible. Thus, the problem of erroneous payments is not so much one of determining eligibility, but rather one of attempting to finely target benefits to the complicated and changing circumstances of low-income households. Second, most overpayments to eligible households are small relative to household income and official poverty standards. As a result, most food stamp households are poor, and they remain poor even when overpaid.

A fundamental issue in the design of the Food Stamp Program (FSP) is the form benefits should take. Advocates of the current coupon system argue that coupons are a direct and inexpensive way to ensure that food stamp benefits are used to purchase food; that, despite some evidence of fraud and benefit diversion under the current system, the unauthorized use of food stamps is relatively limited; and that coupons provide some measure of protection to food budgets from other demands on limited household resources. Advocates of replacing coupons with cash argue that the current system limits the food purchasing choices of participants, places a stigma on participation; and entails excessive costs for coupon issuance, transaction, and redemption.

This study is an analysis of very low birthweight among Medicaid newborns and the effect of prenatal WIC participation on the likelihood of very low birthweight.