Dual Language Immersion Programs within LAUSD Elementary Schools: Who Do They Enroll?

• In Event: Dual Language Immersion Programs in LAUSD and the Potential for Equitable Learning Opportunities

Abstract

Two-way dual language instruction (DLI) integrates English speakers and speakers of a partner language (e.g., Spanish) in the same classroom to receive content instruction in both languages. Stated goals include bilingualism and biliteracy, high academic achievement, and intercultural competence. In segregated districts, DLI programs can also integrate students from diverse linguistic, cultural, and racial backgrounds (Gándara & Aldana, 2014; Kotok & DeMatthews, 2018; Uzzell & Ayscue, 2021). Despite interest, there is little systematic information about whether DLI program enrollments enroll diverse student populations, especially in segregated urban school environments. DLI programs are often strand programs of varying size that students opt into. This paper adds to the literature by using student-level enrollment data to quantitatively describe enrollment and diversity in elementary DLI programs, comparing them to non-DLI strands within their schools).

Research Questions
1. To what extent do elementary DLI programs enroll linguistic minority students and students traditionally underrepresented in dual-language immersion (e.g. African-American students)?
2. To what extent do elementary DLI program enrollments and diversity differ from the composition of their host schools?

Methods
From LAUSD, we obtained a list of all DLI programs, host school, and language of instruction. We combined this with three years of student-level enrollment data, accessed through an agreement with Stanford’s Policy Analysis for California Education (PACE), containing each student’s grade level, DLI program participation, race/ethnicity, EL status, and poverty status.

To answer the first research question, we compare the numbers and proportions of students from different groups enrolled in DLI programs, examining changes over time and between DLI programs depending upon program characteristics (e.g., length of time established, partner language, etc.). To answer the second research question, we compare the racial, linguistic, and economic diversity (entropy) of DLI program enrollments to that of the rest of a host school. We also calculate segregation (Theil’s H, exposure/isolation indices, and concentration), decomposing measures into the segregation between schools and segregation within schools (i.e., between DLI and non-DLI strands) (Clotfelter et al., 2021; Theil, 1972).

Findings
School-level data show that enrollments in LAUSD elementary schools with DLI programs are disproportionately Latinx and have slightly higher percentages of low-income students and EL students than do elementary schools without DLI programs (authors, 2025). In recent years, schools with DLI programs have diverging trends of increasing enrollment in DLI strands for most student demographic groups but decreasing enrollment in the non-DLI strands within the same schools. Descriptive patterns vary by DLI language focus. Preliminary results for research question 2 illustrate lower racial diversity in DLI strands compared to non-DLI strands in the same school.

Implications
This paper systematically examines DLI program enrollments in a large, diverse yet segregated district. Our findings describe whether within-school stratification limits access to DLI programs and whether they provide integrated learning environments within schools, depriving students of intended benefits. As DLI programs expand, understanding student participation of varying linguistic, racial, and economic characteristics in these programs is essential to ensuring DLI benefit our most marginalized students.

Authors

• Erica Frankenberg, Pennsylvania State University

• Sarah Asson, Education Northwest

• Clemence Darriet, University of California - Los Angeles

• Lucrecia Santibanez, University of California - Los Angeles

• Claudia G. Cervantes-Soon, Arizona State University

• Francesca López, University of Wisconsin - Madison