Determinants and Effects of WIC Rollout: Evidence from Newly Digitized Data from the National Archives
AbstractThe Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) program provides eligible low-income, nutritionally at-risk postpartum and pregnant women, infants, and children under age five with nutritious food. The WIC program was created during the War on Poverty in the 1970s, with the first clinic opening in January 1974.
Previous work by Hoynes, Page, and Stevens (2011) examined the effect of WIC adoption on birth outcomes using two-way fixed effects models. We have augmented their rollout data with extensive new data from the National Archives on State WIC plans through 1980, allowing us to detect with more granularity when-and-where WIC was implemented. We also bring new methods to deal with detecting pre-trends and obtaining estimates that are robust to treatment effect heterogeneity across places and over time.
We find robust evidence that access to WIC in a mother’s county of residence by the second trimester leads to higher birth weight and a reduction in the probability of low birthweight mainly for black mothers. We find that exposure to WIC increases birth weight by around 5.3 grams, while it decreases the probability of low birth weight by 1.5% of the mean. We roughly convert these to treatment on the treated estimates by inflating by the national participation rate in 1979, 8.6 percent of pregnant women obtained WIC. This suggests treatment on the treated effects on the order of 62 grams, which is clearly economically meaningful. These results hold whether we focus on standard two-way fixed-effect models or newer approaches to deal with treatment effect heterogeneity (e.g., Sun and Abraham, 2021). The sizes of the intent-to-treat effects are similar to those for Food Stamps for Blacks found by Almond, Hoynes, and Schanzenbach (2011). Effects for whites are smaller and in some specifications, not statistically significant. Ongoing work