Dual-language immersion (DLI) programs are increasingly popular in communities and states across the country. Their number is growing largely on the strength of two key shifts: one in the research on bilingual education’s efficacy and a subsequent change in public demand for multilingualism. These programs can be a powerful way for education leaders to support English learners’ success while also growing access to diverse learning environments for all students. This report explores the DLI ecosystem in the San Francisco Bay Area, a region blessed with extraordinary linguistic, racial, ethnic, and cultural diversity, but also a region wrestling with gentrification and significant wealth inequality.1

There’s no question that DLI programs can be a powerful lever for advancing equity—particularly for English learners (ELs). A series of studies have shown that, when designed and implemented well, DLI is the most effective way to advance ELs’ linguistic and academic development.2 This is a significant shift from the late twentieth century, when schools were encouraged to abandon bilingual education because of fears that continued development of ELs’ home languages would hinder their development of English. That outdated monolingual approach was frequently marshalled to support segregated, remedial, English-only instructional programs for English learners.3 In California, this culminated in the 1998 passage of Proposition 227, which largely banned ELs from bilingual education settings across the state. After nearly twenty years as a so-called “English-only” state, California voters changed course in 2016 with the passage of Proposition 58.4 This new law grants local education leaders more discretion to determine the best instructional models for ELs in their communities—and, in particular, permits them to launch or relaunch bilingual and DLI programs. 

There’s also some evidence that DLI programs can contribute to efforts to enroll linguistically, ethnically, and/or socioeconomically diverse student bodies at the K–12 schools which host them.5 English-dominant families in many communities are newly interested in enrolling their children in public schools that provide bilingual instruction. These families are disproportionately likely to be white and to come from higher socioeconomic privilege than families that speak non-English languages at home.6 It appears that DLI programs can help some privileged families see the value of diverse learning settings. This shift is particularly exciting, since many privileged families struggle to recognize, appreciate, and act on the fact that diverse schools are good for all children.7 Meanwhile, their interest in DLI schools can be valuable for all students, since emerging research suggests that these programs rely upon linguistic diversity to work best.8 Furthermore, those families, who typically have greater access to social and political capital, can lend DLI advocates political support when local school districts are considering whether to launch new DLI programs. 

This family demand dynamic is clearly operative in the Bay Area. In a 2017 article titled “How San Francisco Paved the Way for California to Embrace Bilingual Education,” journalist Jeremy Adam Smith noted that, “The timing of Proposition 58 could not be better for the San Francisco Unified School District. Demand for language education is rising…because so many parents want their U.S.-born children to learn another language.”9 Given that the Bay Area’s wealth gaps are much more pronounced than those in many other U.S. communities, demand for DLI programs from more privileged families could make it difficult for students from historically marginalized communities—particularly ELs—to retain fair access.10

In sum, the challenge for DLI programs is to protect access for those children who stand to gain most from enrolling—ELs—while recognizing that the very best DLI programs are linguistically integrated “two-way” models. These appear to have two key mechanisms driving their success. First, research shows strong linguistic and academic benefits to bilingual learning for all children, and particularly for ELs.11

Second, as explored below, research indicates that students gain advantages from “peer effects” that accrue from linguistically, racially, and/or socioeconomically diverse schools.12 By enrolling roughly equal numbers of native speakers of English and native speakers of the non-English partner language, two-way DLI programs create a naturally multilingual—and often multiracial—public school setting. Ideally, all students contribute their home language skills to the classroom environment. This practice reduces stigma related to English-language learning by converting every child’s home language into a valuable educational asset, which is good for students’ linguistic and academic development—and good, more broadly, for social cohesion in communities and the country. 

By contrast, traditional bilingual education models and one-way DLI programs that serve native speakers of just one language are generally designed to educate linguistically segregated classrooms in two languages. That is, they are organized to bilingually educate classes of 100-percent English-dominant or 100-percent Spanish-dominant (or Korean-dominant, or Vietnamese-dominant, or Welsh-dominant, etc.) students. The bilingualism of these classrooms is more fully contingent upon the language abilities of the teacher (or teachers). In many one-way DLI programs, for instance, each classroom’s only native speaker of the non-English partner language is a teacher (or, conversely, the only native English speaker in a classroom of English learners may be a teacher). 

So while DLI programs can potentially advance integration, local leaders must be attentive to ways that different forms of student diversity may or may not intersect in their communities. For instance, linguistic integration of a school is not always and automatically coterminous with racial and/or socioeconomic integration. Not all English-dominant families are privileged—white, wealthy, highly-educated, and so forth. Similarly, not all families of ELs come from historically underserved backgrounds. As a result, well-intentioned policies that aim solely at linguistic integration within a DLI program may inadvertently exclude a number of historically underserved communities from that same program. 

For DLI to support schools with truly integrated enrollment that advance better outcomes for ELs and other students, educators and policymakers must make intentional choices that safeguard equitable access to DLI for ELs, while also growing sufficient numbers of DLI seats to reach a wide range of linguistically, racially, ethnically, and/or socioeconomically diverse communities.

In sum, this combination—new research on DLI’s potential contributions educational equity and growing support from privileged families—makes DLI programs a uniquely interesting vehicle for advancing conversations about school diversity, integration, and better outcomes for ELs in the Bay Area (and beyond). But DLI programs should not be seen—or treated—as an easy workaround for difficult policy reform work to advance those projects. For DLI to support schools with truly integrated enrollment that advance better outcomes for ELs and other students, educators and policymakers must make intentional choices that safeguard equitable access to DLI for ELs, while also growing sufficient numbers of DLI seats to reach a wide range of linguistically, racially, ethnically, and/or socioeconomically diverse communities.13

The Research Base: Why and How Dual-Language Immersion Works (Particularly for ELs)

Until recently, there were few methodologically rigorous studies on the efficacy of DLI programs. The field has been growing, however. One Stanford University study, “Effectiveness of Four Instructional Programs Designed to Serve English Learners,” explored the performance of various EL-focused program types active in San Francisco Unified School District (SFUSD): DLI, transitional bilingual, developmental bilingual, and English-only. It found that, while English-only programs generated better English language arts (ELA) outcomes for EL students in the early grades, EL students in DLI programs showed significantly higher growth in ensuing years: “although in the early years of attendance [DLI] programs may have a negative effect on performance in ELA, in the long term, the short-term negative effects are more than overturned by the positive effects in test score growth.”14 ELs’ math performance was similar in English-only and DLI programs. Interestingly, they found that Latino EL students seemed to gain unique benefits in academic performance from enrolling in DLI, as opposed to English-only instruction. ELs who speak a Chinese language did not see nearly as strong an academic advantage from enrolling in DLI programs: “While Chinese ELs do best longitudinally in ELA and math when enrolled in [DLI], they also do very well in [English-only].”15 The researchers hypothesized that this finding could be the result of the greater degree of linguistic overlap between Spanish and English, as opposed to Mandarin and English or Cantonese and English. That is, Spanish-speaking ELs may gain more transferable benefits in English from more deeply developing their home language skills in DLI programs than Mandarin- or Cantonese-speaking ELs will. 

A second Stanford University study, “Reclassification Patterns Among Latino English Learner Students in Bilingual, Dual Immersion, and English Immersion Classrooms,” explored the efficacy of SFUSD’s DLI programs for helping ELs reach full academic English proficiency. Similarly, it found that English-only programs initially outperformed DLI programs in the early grades, but that, over time, ELs in DLI programs did better. Specifically, “More students are reclassified in the early years in English immersion, but as students reach middle school, reclassification in the two-language programs catches up with and (in maintenance bilingual and dual immersion) surpasses that of English immersion.”16

These results align with the findings of other methodologically rigorous studies of DLI programs. Researchers studying Portland Public Schools’ DLI programs found that students (of all backgrounds) enrolled in DLI “outperform their peers on state accountability tests in reading by about seven months of learning in Grade 5 and nine months of learning in Grade 8.”17 It also found that ELs enrolled in DLI programs were more likely to reach full English proficiency than ELs in other programs—particularly if they were native speakers of the DLI program’s non-English partner language. 

These studies only scratch the surface of DLI’s potential equity benefits, however.18 There are assuredly other benefits from enrolling students of many backgrounds in diverse, multilingual, multicultural schools that foreground the value of diversity in their pedagogical approach. Diverse, integrated schools benefit all students—particularly historically marginalized students. Given this range of benefits, it’s perhaps unsurprising that DLI programs appear to be growing in diverse communities across the country, including the Bay Area.19

Data and Methodology

This report’s DLI database gathers school-level demographic data for campuses identified as hosting DLI programs in Alameda, Contra Costa, Marin, San Francisco, San Mateo, and Santa Clara counties. There is no up-to-date directory of DLI schools and/or programs in the Bay Area (or the country), so schools were identified for the database via publicly available online records, including publications from county offices of education, school districts, EL master plans, local control and accountability plans, and other sources.20 To that end, it includes schools that host one-way or two-way dual-language immersion programs, but it does not incorporate the region’s transitional bilingual education programs.21

The student-level data from schools in the report come from state and federal education databases that aggregate district-level data on the number of ELs enrolled in different types of Language Instruction Educational Programs (LIEPs). Federal law mandates that, in return for hundreds of millions of federal education dollars, districts provide ELs with LIEPs that help students develop linguistically and academically. Districts track the number of their ELs enrolled in each LIEP they offer. These data are then published at the state and federal levels. 

Notably, however, it was not possible to collect classroom-level demographic data for each school’s DLI program. That is, many DLI schools contain multiple language pedagogy “strands”—e.g. one strand that offers an English-only educational model, and another that offers a multilingual DLI model, sometimes known as “a school within a school” model. This means that most of the demographic data on DLI access gathered here reflect enrollment data of the schools that host DLI programs, but do not necessarily reflect the specific enrollments of the DLI classrooms within those schools. 

This distinction matters because strand programs can differ significantly in their demographics. For instance, in some cases, DLI programs are treated as schools of choice within neighborhood schools, with correspondingly different enrollment targets and rules for enrollment. As such, demographic differences between a school’s two instructional strands may be consequential for DLI’s ability to facilitate integration. This dynamic has periodically produced concern in communities where DLI programs come to be viewed as educational programs that encourage gentrification and in-school segregation.22

To supplement these data, the appendix provides district-level data on the number of ELs in DLI, bilingual, and other instructional programs. These provide a different angle for measuring ELs’ access to multilingual instruction in the region. 

The report’s approach has several other limitations. Data collected for this memo represent a snapshot in time. They capture the state of DLI in the Bay Area in the 2019–20 academic year, just before the COVID-19 pandemic shuttered schools and scrambled enrollment demographics across the region (and country). As such, they cannot illuminate demographic trendlines. Further, the sheer number and diversity of districts in the Bay Area make it difficult to provide conclusive analysis of how various DLI programs have affected—or may be affecting—particular schools’ or districts’ enrollment patterns. That is, DLI’s impact in a housing-scarce, high-density city like San Francisco is assuredly different from its impact in the region’s smaller, rural districts. 

DLI Programs in the Bay Area

In 2019–2020, the six counties studied in this report offer a number of DLI programs:

  • San Francisco Unified School District ran sixteen DLI programs, which provided Spanish–English, Cantonese–English, Korean–English, or Mandarin–English instruction. The county also contained a Spanish–English DLI charter school. 
  • Contra Costa County contained thirteen DLI programs across five school districts. It also contained a Spanish–English DLI charter school. Its DLI programs provided Spanish–English instruction or Mandarin–English instruction. 
  • In Alameda County, there were twenty-five DLI programs across eight school districts. Its DLI programs provided Spanish–English instruction or Mandarin–English instruction, along with two charter schools, one offering a French–English program and the other a Mandarin–English program. Oakland Unified School District ran eight Spanish–English DLI programs of its own. 
  • In Santa Clara County, there were twenty DLI programs across ten school districts. Its DLI programs provided Spanish–English instruction, Mandarin–English instruction, or (in one school) Vietnamese–English instruction. There were also three Spanish–English DLI charter schools. 
  • In San Mateo County, there were six DLI programs across four school districts: five were Spanish–English DLI and one was Mandarin–English. 
  • In Marin County, one district—Novato Unified—offered Spanish–English DLI at two campuses. 

In sum, the six counties contained twenty-nine school districts offering eighty-eight DLI programs. Spanish predominated amongst the partner languages: there were sixty-nine Spanish-English programs. Mandarin was the next most prevalent, with twelve programs. There were four Cantonese programs, and one each offering French, Korean, or Vietnamese immersion.

Demographics of Bay Area DLI Programs

School-Level Demographics

The following analyses compare the demographics of schools that run DLI programs with the demographics of their respective counties as a whole. This provides a point of reference: are schools with DLI programs generally tracking the demographic balance of their communities? Do they tend to overenroll or underenroll students from particular groups? To what degree are their DLI programs equitably accessible to their EL populations, who—research indicates—are likely to garner unique linguistic and academic benefits? 

First, aggregate data on Bay Area schools with DLI programs (Table 1) suggest that these campuses loosely track the region’s demographics—with some exceptions. Schools with DLI programs oversample for Latino students, students from low-income families, and English learners, and they undersample for white students. Given that the bulk of the region’s DLI programs offer instruction in Spanish and English, and some programs reserve as many as half of their available seats for native speakers of the non-English partner language, it is not entirely surprising to see higher enrollment numbers for Latino students—who make up more than 80 percent of California’s EL population and are more likely to speak Spanish at home than students of other ethnicities.23 Meanwhile, African-American students and Asian students enroll in DLI schools at roughly their rates of enrollment in the region’s schools at large. 

Table 1

Bay Area DLI School Demographics Compared with Bay Area Enrollment Demographics
African-American Latino White Asian Low-Income EL
Demographics of Schools with DLI Enrollment 5.19% 47.72% 16.22% 22.57% 43.97% 27.90%
Demographics of Bay Area School Enrollment 5.56% 36.12% 22.04% 23.16% 38.12% 20.02%
Source: TCF Bay Area DLI Database, data collected from Ed-Data, Education Data Partnership, California Department of Education, EdSource, Fiscal Crisis and Management Assistance Team/California School Information Services, https://www.ed-data.org/state/CA.

Of course, because of the diversity of schools and school districts in the region, these regional demographics offer an unnuanced picture. They combine, for instance, schools that run Mandarin and Cantonese immersion programs in San Francisco with schools that run Spanish immersion in Mountain View.

Examining the county-level demographics of schools hosting DLI programs can provide a somewhat more useful picture. In San Francisco (Table 2), which offers language programs in Spanish, Cantonese, Mandarin, and Korean, DLI schools enroll African-American students, white students, and Asian students at lower rates than the district at large. These schools oversample Latino and EL families and slightly oversample for low-income families. While San Francisco DLI schools enroll roughly one-eighth of the community’s students, they enroll one-sixth of the community’s Latino students, one-sixth of its ELs, and about one-ninth of Asian students.

Table 2

SFUSD DLI School Demographics Compared with Total Enrollment Demographics
African-American Latino White Asian Low-income EL
Demographics of San Francisco DLI Schools 4.87% 41.67% 11.56% 26.55% 51.29% 36.98%
Demographics of San Francisco County Students 7.79% 32.36% 14.55% 29.37% 48.19% 27.72%
Percent of Subgroup Enrolled in DLI 7.58% 16.09% 9.85% 11.41% 13.31% 16.77%
Note: As San Francisco Unified School District is the predominant district in San Francisco County, we have not included a separate row of data comparing the demographics of “San Francisco Districts Offering DLI.”

In Contra Costa County (Table 3), the patterns tilt differently. African-American, Latino, and Asian students enroll in DLI schools at higher rates than they enroll in their districts or the broader county. The same is true for ELs and students from low-income families. Only white students enroll in these schools at lower rates than their share of district enrollment.

Districts in Contra Costa County enroll significantly smaller shares of all students in DLI schools compared with San Francisco. Not quite 5 percent of Contra Costa County students are enrolled in DLI schools, and no student group has more than 7 percent of their countywide enrollment in those programs.

Table 3

Contra Costa DLI School Demographics Compared with Contra Costa Total Enrollment Demographics
African-American Latino White Asian Low-income EL
Demographics of Contra Costa DLI Schools 12.37% 43.40% 18.18% 19.12% 47.64% 23.15%
Demographics of Contra Costa Districts Offering DLI 9.81% 40.04% 22.18% 15.39% 45.19% 19.23%
Demographics of Contra Costa County Students 8.59% 36.72% 28.67% 13.07% 39.43% 15.95%
Percent of Contra Costa County Subgroup Enrolled in DLI 6.71% 5.51% 2.96% 6.82% 5.63% 6.77%

In Alameda County (Table 4), meanwhile, African-American students enroll in schools with DLI at a lower rate than they enroll in the county or in the county’s districts that offer DLI programming. Latino students and ELs are overrepresented in DLI programs relative to their share of the countywide population. White, Asian, and low-income students enroll in DLI schools at roughly the same rates as they enroll in Alameda County schools or in Alameda County school districts offering DLI. Notably, Alameda County manages to enroll more than one in ten of its ELs in a single DLI school.

Table 4

Alameda DLI School Demographics Compared with Alameda County Total Enrollment Demographics
African-American Latino White Asian Low-income EL
Demographics of Alameda DLI Schools 5.93% 43.51% 16.68% 26.19% 42.52% 28.03%
Demographics of Alameda Districts Offering DLI 10.16% 35.82% 16.05% 26.26% 45.41% 21.08%
Demographics of Alameda County Students 9.18% 34.03% 17.23% 26.67% 41.78% 19.53%
Percent of Alameda County Subgroup Enrolled in DLI 4.89% 9.68% 7.33% 7.44% 7.70% 10.87%

In Santa Clara County (Table 5), African-American students, Latino students, ELs, and students from low-income families all enroll in schools with DLI at higher rates than their enrollment in the district, while white students and Asian students enroll in these schools at lower rates. Patterns are largely the same when comparing the demographics of schools with DLI to the demographics of Santa Clara County as a whole: African-American students, Latino students, ELs, and students from low-income families enroll at higher rates in DLI schools than in the broader county.

Table 5

Santa Clara DLI School Demographics Compared with Santa Clara County Total Enrollment Demographics
African-American Latino White Asian Low-income EL
Demographics of Santa Clara DLI Schools 1.99% 58.03% 14.56% 19.61% 43.19% 26.11%
Demographics of Santa Clara County Districts Offering DLI 1.72% 43.65% 17.27% 28.10% 39.56% 24.62%
Demographics of Santa Clara County Students 1.80% 38.28% 18.69% 30.40% 34.88% 21.74%
Percent of Santa Clara County Subgroup Enrolled in DLI 6.75% 8.74% 4.91% 3.08% 6.87% 6.33%

In San Mateo county (Table 6), Latino students and EL students enroll in DLI schools at higher rates than they enroll in the broader county—but they enroll in DLI schools at lower rates than they enroll in DLI-offering districts. Low-income students and African-American students are comparatively underrepresented in DLI schools, relative to the county and to DLI-offering districts. White students are overrepresented in DLI schools relative to their share of the county population and the DLI-offering districts population, while Asian students are slightly overrepresented.

Table 6

San Mateo DLI School Demographics Compared with San Mateo County Total Enrollment Demographics
African-American Latino White Asian Low-income EL
Demographics of San Mateo DLI Schools 0.56% 43.02% 27.23% 18.28% 30.12% 27.01%
Demographics of San Mateo County Districts Offering DLI 1.38% 49.90% 21.74% 15.33% 42.53% 30.56%
Demographics of San Mateo County Students 1.31% 38.07% 23.87% 16.07% 32.71% 20.85%
Percent of San Mateo County Subgroup Enrolled in DLI 1.63% 4.31% 4.35% 4.34% 3.51% 4.94%

There are only a couple DLI schools in Marin County, and they oversample for Latino students, ELs, and students from low-income families, relative to DLI-offering districts and the county. They underenroll white students, relative to both districts and county. Asian students and African-American students are a smaller population here, but their DLI school enrollment share roughly tracks with their county enrollment share. Asian students are slightly overrepresented in DLI schools.

Table 7

Marin DLI School Demographics Compared with Marin County Total Enrollment Demographics
African-American Latino White Asian Low-income EL
Demographics of Marin DLI Schools 2.29% 55.72% 27.26% 6.67% 54.53% 23.98%
Demographics of Marin County Districts Offering DLI 2.36% 38.67% 44.78% 5.31% 38.48% 15.47%
Demographics of Marin County Students 1.66% 31.50% 54.10% 4.63% 28.63% 15.17%
Percent of Marin County Subgroup Enrolled in DLI 4.14% 5.31% 1.51% 4.32% 5.71% 4.74%

There is also some evidence that Bay Area DLI schools’ demographic patterns vary according to the specific non-English partner language they offer. For instance, nearly 53 percent of the students enrolled in Bay Area schools with Spanish–English DLI programs are Latino, but not quite 12 percent of the students enrolled in schools with Mandarin–English DLI programs are. By contrast, nearly 62 percent of the students enrolled in Bay Area schools with Mandarin–English DLI programs are Asian, while just 14.5 percent of the students enrolled in schools with Spanish–English DLI programs are.

Spanish–English DLI schools sample more heavily for ELs, white students, and for students from low-income families. Nearly one-third of students in schools with Spanish–English DLI programs are ELs, while just 18 percent of students in schools with Mandarin–English DLI programs are. Nearly one-quarter of students in Spanish–English DLI programs are white, while under one in ten children in Mandarin–English DLI schools is white. Similarly, nearly 45 percent of students in Spanish–English DLI schools are low-income, while only 18 percent of students in Mandarin–English DLI programs are.

Table 8

Demographics in Bay Area Schools with DLI Programs, by Non-English Partner Language
African-American Latino White Asian Low-income EL
Spanish–English Demographic Percentage 4.83% 52.72% 23.36% 14.52% 44.73% 31.98%
Mandarin–English Demographic Percentage 4.03% 11.64% 9.41% 61.69% 18.34% 16.20%

Discussion: The Present State of Bilingual Learning in the Bay Area

These data show some encouraging trends. In general, Bay Area DLI schools oversample for Latino students and for ELs—particularly in Spanish–English DLI schools. This tracks the research on what works best for these students, and indicates that, aggregated to the regional level, DLI access patterns are reasonably equitable.

Unfortunately, children do not attend schools according to aggregated large-scale regional data. They attend particular schools in particular districts located in particular communities.

Unfortunately, children do not attend schools according to aggregated large-scale regional data. They attend particular schools in particular districts located in particular communities. And at these levels, the data indicate that local education leaders have much more work to do. For instance, San Mateo County districts only hosted six DLI schools in 2019–20, and these underenrolled ELs and Latinos relative to the demographics of districts offering DLI. While this is concerning on its face, it likely reflects a deeper problem: San Mateo County only enrolled about 4 percent (3,566) of its total student body (93,554) in DLI programs. Given that there were more than 35,000 Latino students and nearly 20,000 ELs in the county that year, San Mateo’s primary DLI equity challenge is simply the county’s scarcity of programs.

There are assuredly other patterns to be gleaned from these data. Additional research should, for instance, seek out more nuance within the linguistic integration of these programs. For example, in many cases, Bay Area DLI programs reserve seats for native speakers of the non-English partner language. In many cases, these students are also classified as English learners by the state of California’s English-Language Proficiency Assessments for California (ELPAC) system.

However, these two things—native usage of a non-English language and still-developing proficiency in using English—are not identical. That is, an EL student who speaks Spanish at home would not qualify for a native Mandarin speaker seat in a Mandarin–English DLI program, while a bilingual non-EL student who speaks both Mandarin and English at home perhaps would. Given that the number of ELs in a DLI program is a relatively standard metric for measuring equitable access, policies seeking to use DLI to further integration should address these sorts of complexities. For instance, how will DLI programs identify native speakers of each language for the purpose of allocating seats?

Indeed, these sorts of allocations can be fraught, given the intense demand for these programs from many English-dominant families. As noted above, in some U.S. communities, this demand can make it difficult to maintain linguistic integration—let alone socioeconomic or racial integration.24 In a city as plagued by income, asset, and housing inequalities as San Francisco, these demand disparities will sit uneasily amidst limited local DLI supply. At some point, privileged English-dominant families will likely seek—and find—ways to circumvent or undermine linguistic integration. 

There is already some evidence of this dynamic in the city, where public school enrollment is largely open and lottery-based. In 2019, San Francisco Unified School District published a document outlining family demand for seats in its various DLI programs. In program after program, demand for the English-dominant seats in DLI was larger than demand for the Spanish- (or Cantonese-, Mandarin-, etc. dominant seats). For example, for every kindergarten seat held for non-native Korean speakers (i.e. English-dominant) in Lilienthal K–8’s Korean–English DLI program, there were 28.11 children seeking that slot. For each native Korean-speaking seat, there were 1.08 children requesting the seat. Lilienthal is no exception. At Alvarado Elementary, there were 27.11 children requesting each non-native Spanish speaker kindergarten seat, and just 3.92 children for each native Spanish speaker kindergarten seat. At Ortega Elementary, there were 23 children seeking each non-native Mandarin speaker kindergarten seat, and 6.54 seeking each native Mandarin speaker seat. And at Clarendon Elementary, there were 15.44 children pursuing each non-native Japanese speaker kindergarten seat, and two children pursuing each native Japanese speaker kindergarten seat.25 These disparities likely reflect both the growing number of privileged, English-dominant families in San Francisco during a period of gentrification, as well as increased interest in DLI from these families. 

Again, enthusiasm for DLI from privileged English-dominant families can make these schools useful tools for growing diverse learning environments. But delivering on that potential requires setting clear guidelines and objectives for DLI programs—and supporting them with enrollment policies that protect access for ELs and other historically marginalized students. Notably, San Francisco Unified explicitly builds its core DLI programs around linguistic integration. SFUSD describes its DLI enrollment system thus: “[the] district will assign an appropriate balance of students who are proficient in the pathway language (2/3 to 1/2) and students who have not demonstrated proficiency in the pathway language (1/3 to 1/2).”26 It justifies this decision on the basis of “DLI research,” which presumably refers to the growing number of studies indicating the value of peer effects, where native speakers of each language learn from one another.27 That is, SFUSD holds seats in each DLI program for proficient speakers of English and proficient speakers of the partner language because that reflects best practices in DLI program design—which happens to summarily produce more diverse programs. 

Other districts have also set explicit, appropriate diversity goals for their DLI programs. For example, West Contra Costa Unified School District’s Washington Elementary describes its DLI program as a model “where English and Spanish instruction is provided to English learners and Spanish learners so that both learn to read, write and speak in both languages as well as feel positive about themselves, their school, and people of other cultures.”28 Hayward Unified School District describes its DLI programs as 

bring[ing] together native English speakers and native speakers of another language in a unique opportunity for both groups to learn a second language. Students are taught in English for 50% of the day and in Spanish or Mandarin for the other 50% of the day. Interactive lessons encourage students from both language groups to help each other understand, speak, read, and write one another’s home language. All students gain a second language in this enriched educational environment.29

The challenge, of course, is for leaders in these—and all—communities to back these DLI goals with firm, thoughtfully implemented policies that make them meaningful for all students—particularly ELs.

Next Steps for DLI in the Bay Area

While the data show some enrollment patterns that could indicate that Bay Area DLI programs are not as equitably accessible as they might be, most of the evidence here is encouraging. Latinos make up 36 percent of Bay Area enrollment, but nearly 53 percent of enrollment in Spanish–English DLI schools (Table 1 and Table 8). ELs make up 20 percent of the Bay Area student body, but they’re 32 percent of enrollment in Spanish–English DLI schools (Table 1 and Table 8). Asian students, meanwhile, make up 23 percent of the region’s enrollment, but are 62 percent of enrollment at Mandarin–English DLI schools.

Bay Area schools still have DLI equity issues to address. First among these is the challenge of growing supply to meet diverse families’ demand for bilingual learning. The twenty-nine districts analyzed here enrolled under 17,000 of their 101,775 ELs in DLI (see Appendix). For comparison, that same year, Dallas Independent School District enrolled 30,803 of its 64,217 ELs in DLI programs. Districts’ self-reported data on ELs’ language instructional programs (see Appendix) indicate that no Bay Area district managed to enroll half of its ELs in some form of bilingual or DLI programming. Hayward Unified School District reported enrolling 40 percent of its ELs in DLI, Redwood City Elementary School District reported 34 percent, and San Jose Unified School District hit 27 percent.

In other words, it’s not that other student groups are crowding ELs out: there simply aren’t enough seats for any district to enroll particularly large shares of any of their student groups. For instance, San Francisco outpaces all other counties by enrolling just 16 percent of its Latino students in DLI schools—no other county breaks double-digits for that student group (Tables 2–7). San Francisco also leads all counties by enrolling 13 percent of its low-income students in DLI schools; Alameda is second, with just short of 8 percent of its low-income students enrolled (Tables 2–7).

Bay Area schools should increase their supply of DLI schools to: 1) offer more ELs access to the DLI programs that research says is best for them and 2) grow enough DLI programs so that demand from privileged families does not crowd out historically marginalized students. To do so, they can follow the steps discussed below.

Recommendations

These recommendations draw heavily on prior research published by TCF authors, including these reports:

1. Identify priority campuses for new DLI programs.

Districts should, whenever possible, establish new DLI programs on campuses with significant numbers of ELs and/or native speakers of the potential DLI program’s non-English partner language. The precise EL enrollment threshold may vary: districts lucky enough to have high rates of linguistic diversity might prioritize launching DLI in schools where it’s possible to reach a target of enrolling, say, 50 percent Vietnamese-dominant students in the new program. Districts with fewer ELs—or ELs who are spread across a large number of schools—may need to set lower thresholds. 

Once districts have established priority campuses for new DLI programs based on linguistic equity, they should assess whether their current and proposed DLI programs will exclude any other student groups—particularly historically marginalized ethnic or socioeconomic groups. 

2. Set clear enrollment rules that protect equitable access for ELs.

Districts should establish clear enrollment targets that reserve equal numbers of seats for native speakers of English and native speakers of the program’s non-English partner language. Districts should also weight any DLI enrollment lotteries towards ELs or native speakers of the program’s partner language. 

If current—or new—DLI programs appear to be shifting housing patterns in ways that allow privileged families to disproportionately purchase DLI access through the housing market, districts should reduce or eliminate the number of seats assigned by students’ addresses. Moreover, they should make investments in school transportation options to ensure that students from all linguistic, racial, and socioeconomic backgrounds have fair access to these programs, and they should prioritize “backfilling” DLI seats in older grades with native speakers of the non-English partner language.

3. Recruit for DLI programs in linguistically and culturally competent ways.

Districts should ensure that all public communications about new and existing DLI programs are translated or interpreted. Applications should be made available in all languages prevalent in the district community. 

Districts should also review all DLI application procedures to verify that they are equitably accessible to families of all backgrounds. For instance, districts should avoid policies that require interested families to have home internet access, apply within a narrow time window, or travel to particular locations to demonstrate their interest. 

4. Focus on bilingual teacher pathways.

Districts should establish a short-term DLI staffing plan that:

  1. draws on existing local teacher training pathways whenever possible—and avoids reliance on guest teacher visas; and that 
  2. commits district resources—and utilizes state resources for supporting bilingual teacher candidates—to create pathways that link existing bilingual paraprofessionals and classified staff with teacher training programs.30

Districts should make long-term DLI staffing plans that:

  1. increase their ability to produce bilingual and biliterate graduates from their own K–12 campuses, and that
  2. connect their own bilingual and biliterate graduates with area teacher training pathways, at both traditional programs at institutions of higher education and alternative pathways.

Acknowledgements

The data gathering and analysis for this report were made possible through generous support from the W.K. Kellogg Foundation, Sobrato Philanthropies, and the Heising-Simons Foundation. The core analysis of this report was conducted as part of a collaboration with the Poverty and Race Research Action Council and the Evelyn and Walter Haas, Jr. Fund. This project was significantly advanced by research support from Jonathan Zabala.

Appendix

The demographic data in this appendix cover all Bay Area districts that offered DLI programming during the 2019–2020 academic year. These data provide a district-level snapshot of which local communities are most successful at providing their ELs with access to bilingual learning opportunities.

Note that these data have limits. They only capture the demographics of Bay Area school districts that offer DLI as an instructional option. Further, districts sometimes report students multiple times. For instance, an EL enrolled in a transitional bilingual program will often also receive ESL services; many districts will report that EL as enrolled in both language instruction educational programs (LIEPs). Indeed, this appears to be happening in most Bay Area districts, since nearly all report 100 percent of their ELs enrolled in ESL while also reporting that some ELs are enrolled in other LIEPs. Similarly, in some cases (e.g. Mount Pleasant Elementary School District and, Campbell Union School District, et al), districts reported identical numbers of ELs enrolled in transitional bilingual programs and DLI programs—it is likely that these are the same ELs, counted twice.

Share of ELs Enrolled in Transitional Bilingual Education, or Dual Language Immersion, and/or ESL/ELD Programs in the Bay Area, by School District, 2019–20
Reported Enrollment in Transitional Bilingual or Early-Exit Bilingual Education Programs Reported Enrollment in Dual- Language Immersion Programs Reported Enrollment in ESL or ELD Programs (see note 21) Total Reported Number of ELs
San Francisco Unified School District 27% 25% 100% 15,060
Oakland Unified School District 17% 21% 100% 11,874
Antioch Unified School District 3% 3% 100% 3,051
Mt. Diablo Unified School District 18% 18% 100% 6,185
Pittsburg Unified School District 14% 14% 100% 2,850
San Ramon Valley Unified School District 0% 0% 100% 1,446
West Contra Costa Unified School District 12% 12% 100% 8,904
Berkeley Unified School District 21% 21% 100% 813
Fremont Unified School District 4% 4% 100% 4,460
Hayward Unified School District 40% 40% 100% 5,954
Livermore Valley Joint Unified School District 1% 9% 98% 1,453
New Haven Unified School District 0% 3% 100% 2,218
Pleasanton Unified School District 7% 8% 100% 1,701
San Leandro Unified School District 14% 13% 100% 2,406
Alum Rock Union Elementary School District 9% 9% 100% 3,361
Berryessa Union Elementary School District 0% 2% 100% 2,071
Cupertino Union School District 0% 0% 100% 2,648
Campbell Union School District 14% 14% 100% 1,812
Franklin-McKinley School District 0% 1% 100% 3,041
Gilroy Unified School District 7% 12% 100% 2,535
Mount Pleasant Elementary School District 2% 2% 100% 969
Mountain View Whisman School District 14% 15% 100% 1,110
Palo Alto Unified School District 8% 8% 100% 1,099
San Jose Unified School District 27% 28% 100% 6,500
Burlingame Elementary School District 12% 12% 100% 504
Ravenswood City Elementary School District 18% 18% 100% 1,212
Redwood City Elementary 34% 34% 99% 2,542
San Mateo-Foster City 10% 10% 100% 2,794
Novato Unified School District 6% 6% 100% 1,202
All Districts Running DLI in the Bay Area 16% 16% Total ELs: 101,775
Source: U.S. Department of Education, National Center for Education Statistics, EDFacts file 141, Data Group 678 and EdFacts file 116, Data Group 849, 2019–20.

These data make clear that some Bay Area districts have made bilingual learning much more accessible to their EL students than others. First, as noted above, dozens of the region’s districts are not included in this table, as they offer no DLI programming at all. Second, by contrast, districts like Hayward, San Francisco, Redwood City, and San Jose have invested sufficient resources to enroll substantial shares of their ELs in some form of bilingual or DLI instruction. However, these data do not provide context about the racial, ethnic, and socioeconomic demographics of DLI program enrollments and how equitably diverse groups are able to access them.

Notes

  1. Jason Richardson, Bruce Mitchell, and Jad Edlebi, Gentrification and Disinvestment 2020 (Washington, DC: National Community Reinvestment Coalition, 2020), https://ncrc.org/gentrification20/.
  2. Conor P. Williams, Xigrid Soto-Boykin, Jonathan Zabala, and Shantel Meek, Why We Need to Cultivate America’s Multilingual, Multicultural Assets (Washington, DC: The Century Foundation, June 14, 2023), https://tcf.org/content/report/why-we-need-to-cultivate-americas-multilingual-multicultural-assets/.
  3. Cf. Patricia Gándara and Megan Hopkins’ Forbidden Language: English Learners and Restrictive Language Policies (New York: Teachers College Press, 2010).
  4. Lillian Mongeau Hughes, “California Voters Overturn English-only Instruction Law,” The Hechinger Report, November 7, 2016, https://hechingerreport.org/california-voters-poised-gut-english-instruction-law/.
  5. Andrew Bibler, “Dual Language Education and Student Achievement” Education Finance and Policy 16, no. 4 (Fall 2021): 634–58, 652, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3192647.
  6. Conor P. Williams and Jonathan Zabala, “How to Grow Bilingual Teacher Pathways: Making the Most of U.S. Linguistic and Cultural Diversity” (Washington, DC: The Century Foundation, August 28, 2023), https://tcf.org/content/report/how-to-grow-bilingual-teacher-pathways-making-the-most-of-u-s-linguistic-and-cultural-diversity/.
  7. Amy Stuart Wells, Lauren Fox, and Diana Cordova-Cobo, How Racially Diverse Schools and Classrooms Can Benefit All Students (Washington, DC: The Century Foundation, February 9, 2016), https://tcf.org/content/report/how-racially-diverse-schools-and-classrooms-can-benefit-all-students/.
  8. Jennifer L. Steele, Johanna Watzinger-Tharp, Robert O. Slater, Gregg Roberts, and Karl Bowman, “Achievement Effects of Dual Language Immersion in One-Way and Two-Way Programs: Evidence from a State Scale-Up in Utah,” working paper, https://jensteele1.github.io/files/Utah_2021April26.pdf.
  9. Jeremy Adam Smith, “How San Francisco Paved the Way for California to Embrace Bilingual Education,” San Francisco Public Press, January 25, 2017, https://sfpublicpress.org/news/bilingualschools/2017-01/how-san-francisco-paved-the-way-for-california-to-embrace-bilingual-education
  10. Alejando Lazo, “Silicon Valley’s Vast Wealth Disparity Deepens as Poverty Increased,” CALMatters, February 17, 2023, https://calmatters.org/california-divide/2023/02/silicon-valley-inequality/; Conor P. Williams, Shantel Meek, Maggie Marcus, and Jonathan Zabala, Ensuring Equitable Access to Dual-Language Immersion Programs: Supporting English Learners’ Emerging Bilingualism (Washington, DC: The Century Foundation, May 15, 2023), https://tcf.org/content/report/ensuring-equitable-access-to-dual-language-immersion-programs-supporting-english-learners-emerging-bilingualism.
  11. Conor P. Williams, Xigrid Soto-Boykin, Jonathan Zabala, and Shantel Meek, Why We Need to Cultivate America’s Multilingual, Multicultural Assets (Washington, DC: The Century Foundation, June 14, 2023), https://tcf.org/content/report/why-we-need-to-cultivate-americas-multilingual-multicultural-assets/.
  12. Amy Stuart Wells, Lauren Fox, and Diana Cordova-Cobo, How Racially Diverse Schools and Classrooms Can Benefit All Students (Washington, DC: The Century Foundation, February 9, 2016), https://tcf.org/content/report/how-racially-diverse-schools-and-classrooms-can-benefit-all-students/; Jennifer L. Steele, Johanna Watzinger-Tharp, Robert O. Slater, Gregg Roberts, and Karl Bowman, “Achievement Effects of Dual Language Immersion in One-Way and Two-Way Programs: Evidence from a State Scale-Up in Utah,” working paper, https://jensteele1.github.io/files/Utah_2021April26.pdf; Richard D. Kahlenberg, Peter W. Cookson, Jr., Susan Shaffer, and Charo Basterra, “Socioeconomic Integration from an Equity Perspective” (Bethesda, MD: The Center for Education Equity at the Mid-Atlantic Equity Consortium, 2017), https://files.eric.ed.gov/fulltext/ED585403.pdf.
  13. Cf. Stephen Kotok and David DeMatthews, “Challenging School Segregation in the Twenty-first Century: How Districts Can Leverage Dual Language Education to Increase School and Classroom Diversity,” The Clearing House: A Journal of Educational Strategies, Issues and Ideas 91, no. 1 (2018): 1–6, DOI: 10.1080/00098655.2017.1336405.
  14. Rachel Valentino and Sean Reardon, “Effectiveness of Four Instructional Programs Designed to Serve English Learners: Variation by Ethnicity and Initial English Proficiency.” Educational Evaluation and Policy Analysis 37, no. 4 (2015): 626, https://doi.org/10.3102/0162373715573310.
  15. Rachel Valentino and Sean Reardon, “Effectiveness of Four Instructional Programs Designed to Serve English Learners: Variation by Ethnicity and Initial English Proficiency.” Educational Evaluation and Policy Analysis 37, no. 4 (2015): 630, https://doi.org/10.3102/0162373715573310.
  16. Ilana Umansky and Sean Reardon, “Reclassification Patterns Among Latino English Learner Students in Bilingual, Dual Immersion, and English Immersion Classrooms.” American Educational Research Journal 51, no. 5 (2014): 895, https://doi.org/10.3102/0002831214545110.
  17. Jennifer L. Steele, Robert Slater, Gema Zamarro, Trey Miller, Jennifer Li, Susan Burkhauser, and Michael Bacon, “Effects of Dual-Language Immersion Programs on Student Achievement: Evidence From Lottery Data.” American Educational Research Journal 54, No: 1 (2017): 302, https://doi.org/10.3102/0002831216634463.
  18. To read the most recent field consensus on these, and other, questions related to ELs’ linguistic and academic development, Cf. National Academies of Sciences, Engineering, and Medicine 2017. Promoting the Educational Success of Children and Youth Learning English: Promising Futures. Washington, DC: The National Academies Press, https://www.nap.edu/read/24677/chapter/1.
  19. Conor P. Williams, Shantel Meek, Maggie Marcus, and Jonathan Zabala, Ensuring Equitable Access to Dual-Language Immersion Programs: Supporting English Learners’ Emerging Bilingualism (Washington, DC: The Century Foundation, May 15, 2023), https://tcf.org/content/report/ensuring-equitable-access-to-dual-language-immersion-programs-supporting-english-learners-emerging-bilingualism.
  20. The Century Foundation has published a report analyzing the demographics of hundreds of DLI programs around the country, however. Cf. Conor Williams, Shantel Meek, Maggie Marcus, and Jonathan Zabala, Ensuring Equitable Access to Dual-Language Immersion Programs: Supporting English Learners’ Emerging Bilingualism (Washington, DC: The Century Foundation, May 15, 2023), https://tcf.org/content/report/ensuring-equitable-access-to-dual-language-immersion-programs-supporting-english-learners-emerging-bilingualism/.
  21. Note: myriad bilingual education models exist in the United States outside the world of DLI. Transitional bilingual education programs have historically been the most common. These usually provide bilingual instruction to a class of native speakers of a non-English language with the goal of moving towards English-only instruction by third grade (early-exit transitional bilingual education) or fifth or sixth grade (late-exit transitional bilingual education). Similarly, U.S. schools engage multiple forms of English-only language instruction for ELs. English as a Second Language (ESL) models are usually 1) “push-in” instruction, in which ESL instructors join ELs in mainstream classrooms to support their language learning; or 2) “pull-out” instruction, in which ESL teachers work with ELs outside of mainstream classrooms for lessons focused on learning the English language. California’s “integrated” and “designated” English language development classrooms roughly track this distinction, respectively.
  22. Perry Stein, “Are Dual Language Programs in Urban Schools a Sign of Gentrification?” Washington Post, July 3, 2018, https://www.washingtonpost.com/local/education/are-dual-language-programs-in-urban-schools-a-sign-of-gentrification/2018/07/03/926c4a42-68c2-11e8-9e38-24e693b38637_story.html.
  23. U.S. Department of Education, National Center for Education Statistics, EDFacts file 141, Data Group 678, 2019–20, https://eddataexpress.ed.gov/download/data-builder/data-download-tool?f%5B0%5D=data_group_id%3A678&f%5B1%5D=level%3AState%20Education%20Agency&f%5B2%5D=population%3AEnglish%20Learners&f%5B3%5D=program%3ATitle%20III&f%5B4%5D=school_year%3A2019-2020&f%5B5%5D=state_name%3ACALIFORNIA.
  24. Conor Williams, “The Intrusion of White Families into Bilingual Schools,” The Atlantic, December 28, 2017, https://www.theatlantic.com/education/archive/2017/12/the-middle-class-takeover-of-bilingual-schools/549278/.
  25. “Program Requests per Seat,” 2019, San Francisco Unified School District, https://web.archive.org/web/20190828230208/http://www.sfusd.edu/en/assets/sfusd-staff/enroll/files/2019-20/March_2019_Program_Requests_per_Seat.pdf; Note that 2024 SFUSD data reflects similar patterns. “2024–25 March Requests Per Seat,” San Francisco Unified School District Enrollment Center, https://drive.google.com/file/d/1eVnaiKkFs4Qg69ZP3xcBYK07ZMiQZ9j9/view.
  26. “Kindergarten-Grade 5 Mandarin Dual Language Immersion,” 2024, San Francisco Unified School District, https://www.sfusd.edu/learning/language-pathways-language/mandarin-language-programs/kindergarten-grade-5-mandarin-dual-language-immersion. See also “Kindergarten-5th Grade (K-5) Spanish Dual Language Immersion,” 2024, San Francisco Unified School District, https://www.sfusd.edu/learning/language-pathways-by-language/spanish/k-5-dual-language.
  27. Conor P. Williams, Dr. Xigrid Soto-Boykin, Jonathan Zabala and Dr. Shantel Meek, “Why We Need to Cultivate America’s Multilingual, Multicultural Assets” (Washington, DC: The Century Foundation, June 14, 2023), https://tcf.org/content/report/why-we-need-to-cultivate-americas-multilingual-multicultural-assets/.
  28. “Dual Immersion Spanish/English: What is it?” Washington Elementary School, West Contra Costa Unified School District, https://www.wccusd.net/Page/8180.
  29. “Programs,” Hayward Unified School District, 2024, https://www.husd.us/about-us/programs.
  30. Conor P. Williams and Jonathan Zabala, Moving from Vision to Reality: Establishing California as a National Bilingual Education and Dual-Language Immersion Leader (Washington, DC: The Century Foundation, October 23, 2023), https://tcf.org/content/report/moving-from-vision-to-reality-establishing-california-as-a-national-bilingual-education-and-dual-language-immersion-leader/.

Tags: , ,