Found in Translation? The Effects of Dual Language Immersion Programs (Job Market Paper)

Abstract: Dual Language Immersion (DLI) programs are educational models where students learn core academic subjects in both English and a partner language. They aim to support the needs of English learners (ELs) by promoting bilingualism in English and their native language. Despite their growing popularity, there is little causal evidence of their effects on students' short- and long-run outcomes. Using longitudinal student-level data from Texas, I exploit the gradual expansion of DLI programs across schools in a staggered difference-in-differences framework and estimate their effect on short and long-run educational outcomes including test scores, EL reclassification, college preparation and enrollment. Students drawn to DLI programs would have counterfactually been served by transitional bilingual programs, which use less native language instruction, serve ELs only in all-EL classrooms, and aim to graduate students quickly into English-only instruction. I find that DLI programs raise ELs’ reading scores and reclassification rates, with gains concentrated in schools where ELs are integrated with non-ELs. However, these gains do not translate into higher long-term educational attainment for ELs. Non-ELs, by contrast, see gains in both test scores and college application and enrollment. Overall, my findings indicate that DLI programs improve ELs' academic achievement and English proficiency but suggest that improving English proficiency and academic achievement alone may not be sufficient in increasing college enrollment among ELs.