Bilingual education affects performance of students NOT in bilingual education

Interesting new study in press at the Journal of Public Economics used regression discontinuity to address how Texas elementary school students performed when separated by districts that implemented Spanish-English bilingual education programs that those that didn't. In Texas, if a district enrolls 20 or more students from a common language background (in this case, Spanish), then that district is required to provide bilingual education. Less than 20, no bilingual education, but rather English-as-a-second-language (ESL). The researchers looked at districts that enrolled between 8 and 39 students from Spanish-speaking backgrounds. They looked at student performance on state test indicators and found, generally, that kids who attend schools in bilingual education districts did significantly better on achievement outcomes than kids enrolled in ESL districts. However, those significant differences were driven by the performance of the kids who did not receive language services, that is, the English proficient students! There were no significant differences between the native Spanish speakers enrolled in bilingual or ESL districts.

The findings are interesting, but the real takeaway appears to be that keeping LEP kids away from EP kids is likely beneficial to the EP kids but a wash for the LEP children. This was one of the arguments used by pro-bilingual education forces in Colorado when a measure proposing the elimination of bilingual education was on the ballot in 2002 (see Escamilla, Shannon, Carlos, & García, 2003).