Association for Public Policy Analysis & Management

Panel: Language-of-Instruction Policies and Their Effects on Learning

(Education)

Even as economies become more globally interdependent and science demonstrates the cognitive advantages of bilingualism, access to instruction in languages other than English remains contested in U.S. classrooms. In the past two years, California and Massachusetts each overturned longstanding ballot measures requiring English-only instruction for English learners (ELs) in schools, but these efforts were vigorously debated, and English-only instruction remains the law for ELs in Arizona (New America Foundation, 2016; Sanchez, 2016; Vaznis, 2017). Meanwhile, meta-analyses that have examined effects of bilingual education on English Learners' performance in English have demonstrated positive achievement effects relative to English-only instruction (Francis, Lesaux, & August, 2006; Greene, 1998; Slavin & Cheung, 2005), especially when students were exposed to reading instruction in both languages within the school day. Recent longitudinal studies that have rigorously adjusted for parents' school preferences and other student characteristics have found neutral to positive effects of bilingual and dual language education on the English proficiency and English language arts achievement of ELs (Steele et al., 2017; Umansky & Reardon, 2014; Valentino & Reardon, 2015), and on the English language arts proficiency of native English speakers (Steele et al., 2017; Watzinger-Tharp, Swenson, & Mayne, 2016). Policies around language of instruction, though potentially innovative, may be especially fraught because language is so closely entwined with questions of culture and identity (Valdés, 1997), and also because access to bilingual or dual language education in schools has resource implications for who is hired and for which curricula are purchased (Lara-Alecio, Galloway, Mason, Irby, & Brown, 2004; Parrish, 1994).

This panel brings new research to bear on the ways in which the classroom language of instruction (monolingual or bilingual) causally influences students' academic achievement in core content areas. Paper 1 leverages regional variation in EL classification rules across the U.S. to examine how EL classification status affects teachers' expectations and student achievement, and the extent to which teachers' expectations of ELs are moderated by the classroom language of instruction under a bilingual versus monolingual English instructional model. Paper 2 focuses attention on dual language immersion programs, which promote bilingualism and biliteracy for both native English speakers and ELs. It leverages district immersion lottery data to estimate causal effects of dual language immersion on the math and reading achievement of native English speakers as well as ELs. Finally, Paper 3 leverages the rapid expansion of dual language immersion programs statewide in Utah to estimate causal effects on the math and reading achievement of native English speakers and ELs net of other stable school characteristics.

By examining the ways in which bilingual and dual language models influence the academic performance of ELs and native English speakers and moderate teachers' perceptions of ELs, the proposed session aims to inform the evolution of language education policies in the U.S.