The effects of third-grade retention on multilingual students: A gateway or a gatekeeper?

This paper estimates the effect of test-based grade retention on multilingual students classified as English Learners (ML-ELs). This policy could provide an opportunity for ML-ELs to develop English language proficiency and master academic content or put them at increased risk of worse academic and labor market outcomes. I use a regression discontinuity design leveraging Texas’s test-based grade retention policy to estimate the effects of this policy on ML-ELs' short- and long-term outcomes. Under Texas policy, students scoring below the reading test cutoff are required to repeat third grade and receive additional reading support. I find that, for students close to the test score cutoff, grade retention and supplemental services have a positive effect on math and reading achievement, as well as increasing the probability of reclassification. These positive effects do not, however, translate into higher post-secondary enrollment or earnings. I find small and not statistically significant effects of grade retention on ML-ELs’ probability of enrollment in a post-secondary institution. I also find that earnings for marginally retained students are statistically the same as those for promoted students in young adulthood. Finally, I find suggestive evidence that the effects of retention are larger in schools with higher expenditures on retained students or on bilingual education. This paper underscore the importance of analyzing how education policies designed with all students in mind affect ML-ELs in the short- and long-term.