The Racial Achievement Gap and Causality: Using ECLS-K to Consider the Impact of Social Programs

Abstract

Despite many attempts to close it, the racial achievement gap is a sustained fact in American public education. Attempts to study the gap may use causal inference on the basis of race. I echo concerns of previous scholars and caution that such designs will introduce unexpected bias into estimates. Furthermore I call attention to proceeding with extreme care when explaining descriptive results in order to ensure that no reader unacclimated to the causal inference literature will jump to causal conclusions.I furthermore make the argument that the racial achievement gap is a institutionalized remnant of the disparities of slavery and estimate the effects of social programs on student achievement. In order to do so, I use matching to estimate the effect of the Special Supplemental Nutrition Program for Women, Infants, and Children; and two-way fixed effects to isolate the effect of the Supplemental Nutrition Assistance Program on student achievement using data from the kindergarten cohort of the Early Childhood Longitudinal Study.In both cases I find small causal effects with relatively large accompanying standard errors. However, in both cases, the estimates for African-American students are better, that is less hurtful in the case of WIC and more beneficial in the case of SNAP, than those for white students. I thus conclude that, while there still exists much ground to cover in truly closing the achievement gap, large-scale social programs such as these have potential to help mitigate differences in student achievement.