Panel Paper: What Types of Families Do Dual Language Immersion Programs Attract: Causal Evidence from Lottery Data

Saturday, November 5, 2016 : 9:10 AM
Columbia 8 (Washington Hilton)

*Names in bold indicate Presenter

Trey Miller1, Jennifer L. Steele1, Gema Zamarro2, Kaitlin Anderson2, Robert O. Slater3 and Jennifer Li1, (1)RAND Corporation, (2)University of Arkansas, (3)American Councils for International Education


Dual-language immersion schools, which provide native English speakers and English language learners (ELLs) with general academic instruction in two languages from kindergarten onward, are proliferating rapidly in the United States.  A reason that is often cited for this rapid expansion is that dual language immersion programs are attractive to two very distinct types of families, both of whom are seeking high quality public schooling options for their kids:  relatively affluent and motivated families seeking a diverse environment that will develop their kids foreign language skills; as well as less-affluent and diverse parents of non-native English speaking students.  This “convergence of stakeholders,” along with rapidly emerging evidence that effectively implemented dual-language programs can improve student achievement, has helped educational administrators navigate the complicated landscape to make dual language programs available to students.  In this study, we use data from a large urban school district that allocates dual-language immersion slots by a lottery mechanism to investigate whether dual-language immersion programs help retain families within the school district.  Specifically, we investigate how characteristics of a lottery applicants’ catchment and choice school relate to their probability of enrolling in the district when they win the lottery or not.  We show that lottery applicants whose catchment school has relatively high test scores and / or a disproportionate share of white students are relatively less likely than other lottery applicants to enroll in the school district when they lose lottery and are not admitted to an immersion program.  Similarly, lottery applicants whose catchment school  has a relatively large share of ELLs, low income students, and Hispanic or black students are more likely than other lottery applicants to enroll in the district when they lose the lottery and are not admitted to an immersion program.  We also show that lottery applicants are more likely to stay in the district when they win a slot in a language immersion program at a school with a larger share of white students or students with high test scores; and are less likely to stay in district when they win a slot in a language immersion program at a school with a larger share of Hispanic, black, low income, and / or ELL students.  These results suggest that the presence of dual language immersion programs may help urban districts retain more affluent and white families, while also exposing less affluent ELL students to more affluent peers.