The Effect of WIC on Education and Health Outcomes

This dissertation examines the effect of WIC on education and health outcomes using policy variation in WIC created by Medicaid expansions. In the first chapter of the dissertation, “Increased WIC Eligibility and Birth Outcomes,” I employ the ‘Difference in differences (DID)’ estimation strategy to identify the effects of WIC and Medicaid separately. The results show that a one percentage point increase in WIC eligibility increases average birth weight by 0.495g, which corresponds to an estimated treatment on the treated effect of 14.8g. WIC also reduces the incidence of Small for Gestational Age (SGA) by 0.021 percentage points which corresponds to a treatment on the treated effect of 6.8%. For Medicaid, a one percentage point increase in Medicaid eligibility reduces the incidence of very low birth weight by 0.002 percentage points and preterm birth by 0.015 percentage points. I find little evidence that there is a strong causal link between WIC and low birth weight or preterm birth; however, most literature on WIC found a positive causation. In the second chapter of the dissertation, "How does increased WIC eligibility affect test scores?," I provide the first quasi-experimental evidence on the effect of WIC on test scores. DID estimator is used to separate out the effect of WIC from the effect of Medicaid. The result shows that even though Medicaid has a positive, statistically significant effect on 4th grade reading scores, WIC has no significant effect on educational outcomes. A ten percentage point increase in Medicaid eligibility is associated with 1.33 scaled point increase in reading scores for 4th grade; however, I find no effect of Medicaid on reading scores for 8th grade nor on math scores. Although this chapter finds WIC has no significant effect on test scores, adding WIC eligibility makes the effect of Medicaid statistically significant. It suggests that researchers need to be very careful when they study Medicaid or WIC since Medicaid expansions could have an effect on both Medicaid and WIC eligibility. In the third chapter of the dissertation, “The Effect of Increased WIC Eligibility on Educational Attainment," using the similar identification strategy, I find that the expansions in WIC eligibility are not associated with a significant effect on high school graduation or college enrollment rates. Consistent with past studies on Medicaid, I find that Medicaid has a positive and statistically significant effect on both high school graduation and college enrollment rates. A ten percentage point increase in Medicaid eligibility increases the High school completion rate by 0.19 percentage points, and college enrollment rate by 0.183 percentage points. Finally, I show that accounting for the linked eligibility of Medicaid and WIC does not change the estimated effect of Medicaid by much, presumably because the effect of WIC on these outcomes is limited.