Association for Public Policy Analysis & Management

Dual language classrooms give a significant portion of instruction in a non-English language. Two-way classrooms typically contain an approximately equal number of students from two different language backgrounds and provide instruction in both languages. This provides an opportunity for English Language Learners (ELLs) to receive instruction in their native language as they transition to English fluency. Participation in a dual language program can replace or complement more traditional services, and could allow ELLs to build a better foundation in core subjects and lead to improved academic outcomes. One-way classrooms, on the other hand, typically focus on providing a larger proportion of instruction in a second language to English first language speakers. Both classroom types present options for English first language students to receive instruction in a foreign language. Some might expect this to hinder a child’s progress, but it could also promote development through a number of potential mechanisms such that it improves overall academic achievement.

This leads to important questions about the effect of participating in dual language education on the achievement of ELLs and non-ELLs. The number and proportion of ELLs more than doubled from 1990 to 2008 (Goldenberg, 2008), yet there has been a shift away from home language instruction for ELLs (Zehler et al, 2003). Positive findings might suggest a potentially effective path in a longstanding effort to improve the performance of a struggling population. Dramatic growth in the number of one- and two-way programs (Howard and Sugarman, 2001; Center for Applied Linguistics, 2011) suggests that there is increasing demand among English speaking students for this style of education despite a lack of concrete evidence on the impact, positive or negative, on achievement. Even a null effect would indicate that English first language speakers are unharmed by receiving instruction in a second language, despite taking end of grade exams in English. Meanwhile, they presumably become fluent in the second language, a desirable outcome in and of itself.

Much of the prior research has focused on an apparent test score advantage for participants in dual language classrooms without necessarily addressing selection bias, although recent research has paid more attention to estimating causal relationships. This study examines dual language education and student achievement using administrative data from Charlotte-Mecklenburg Schools (CMS). Admission lotteries provide exogenous variation used to estimate the effects of participation separately for students who do and do not speak English as their first language. Early analyses show large gains in math test scores for students who do not speak English as their first language. Furthermore, there is no strong evidence of any negative effects on math or reading test scores for either subgroup.