The United States has now enjoyed two decades of dual-language immersion (DLI) education programs in its public schools. The pedagogical model promises to improve the education outcomes of students for whom English is not their first or most proficient language while helping children from all language backgrounds become bilingual. Done well, DLI programs can also break down racial, ethnic, and socioeconomic segregation.

But it’s no simple thing to do this well. Concern dominated a recent convening at UCLA of leaders in the field, which was followed by a report that synthesized the findings and discussion: in practice, it looks like too many DLI programs may be actually exacerbating the very equity issues they were designed to address.

How did the DLI community get here? And what can this trend tell us about how to implement the model—and how not to implement it?

A Political Solution for Bilingual Education…

Odd as it may seem now, as recently as the late 1990s, bilingual education was in full retreat across the country. Anti-immigrant movements in Arizona, California, and elsewhere coalesced into a series of referenda in which voters decided to impose English-only mandates on public schools. In most cases, these made it nearly impossible—indeed, illegal—for schools to serve English learners in bilingual classroom settings.

This was nothing new, necessarily: bilingual education’s roots in the United States stretch at least a century deep, and this was hardly the first wane driven by English-dominant communities’ cultural anxieties.

Bilingual education in these states had been largely segregated. For the most part, the programs were designed for native speakers of non-English languages (predominantly Spanish). This narrow academic purpose also limited their political constituency. Specifically, native English speakers generally had no direct stake in them, which left many voters (at best) ambivalent about them, and left the programs politically vulnerable.

DLI could, in essence, resuscitate bilingual education by sharing its fruits with a wider range of students.

Dual-language immersion arose from the resulting defeats. They offered a way out: these classrooms would enroll students of diverse linguistic backgrounds, English-dominant and English-learning students alike. Ideally, this linguistic integration would drive stronger academic outcomes, richer multilingual learning, and some broader political investment. DLI could, in essence, resuscitate bilingual education by sharing its fruits with a wider range of students. What’s more, by including native speakers of each language in bilingual classrooms, two-way DLI programs could launch truly bilingual learning environments that would foster better linguistic and academic development across the board.

After twenty years of promoting themselves thus, dual-language immersion (DLI) programs have grown from an occasional oddity in a handful of public schools to a movement with momentum. A new report out of UCLA’s Civil Rights Project outlines DLI’s tantalizing promise: “There is limited but growing evidence that supports the argument that instruction in two languages from early grades produces higher academic achievement in core academic content…tested in English, especially for English language learners.”

…With Pedagogical Consequences

With two decades of DLI program development and growth underneath the country’s belt, some of the consequences of its political strategy are now coming clear. Last December, a group of DLI researchers, advocates, and educators convened at UCLA to take stock of how well DLI is delivering on the equity promises that accelerated its rise. As a researcher who has long been curious about DLI’s potential, I was eager to attend.

Last month, the forum’s lead conveners, the UCLA Civil Rights Project’s Patricia Gándara and National Dual Language Immersion Research Alliance’s Robert Slater, published the aforementioned report, summarizing the forum’s two days of discussions.
The paper makes clear that access concerns are the primary equity concern for DLI going forward. In other words, DLI’s political advantages have become pedagogical liabilities.

Here’s how that’s happened: the programs tried to build a new (English-dominant) political constituency for bilingual education under the banner of “Multilingualism for All,” only to discover that, in practice, this increasingly means providing only English immersion for English learners (ELs) and multilingual DLI instruction for English-dominant students. Instead of ushering in an era when English-learning students have more access to schools that can help them maintain and grow their emerging bilingualism, DLI programs may be launching—or, worse, may have already launched—an era that primarily offers multilingualism to English-dominant and higher-privilege communities.

A bird’s eye view of the DLI landscape hints at this apparent trend. Much of DLI’s growth in the past twenty years has been in the establishment of “one-way” programs that primarily serve English-dominant children. For instance, Utah has launched nearly 200 new DLI programs since 2008; but the large majority of these programs are one-way programs. In other cities and states, DLI schools begin as two-way programs, but gradually shift as interest from English-dominant families increases. In Washington, D.C., where two-way DLI programs have been steadily growing more English-dominant for years, a new bill before the DC Council would launch eight new DLI schools while specifically prohibiting leaders from considering whether the new programs would serve ELs. Each new school year brings new examples like these.

The problem is fundamentally structural. U.S. public schools customarily determine school enrollment and access by means of real estate markets. Given that EL students and their families are more likely than English-only households to live below the poverty line, they often have less power to choose or retain access to particular schools—including those that offer DLI. In many cases, then, English-dominant families simply have more material resources that they can wield to displace ELs and their families from neighborhoods and schools. Even in other systems that determine school enrollment, privileged, English-dominant families frequently find ways to tilt access towards their children.

In essence, as the UCLA report puts it, “DLI offers the promise not only to level the playing field for students but to transform schools and school districts through ‘purposeful integration’ of partner language speakers (immigrant and U.S.-born) and English speakers…perhaps even shifting populations that might help reverse trends that include increased segregation of Latinx and immigrant populations.” And yet, that potential is nothing certain. The report warns, “The ideal [two-way DLI] model is a balance of 50% English and 50% partner language speakers…The effort to enroll and appease parents of English speakers by allowing programs to become unbalanced, or the practice of tipping the enrollment balance toward speakers of the partner (e.g. Spanish) language significantly impacts the power of the two-way model.”

Components Limiting DLI Growth

To be sure, “Multilingualism for All” is a laudable educational goal. It only poses an equity problem when 1) there’s a shortage of available multilingual school seats, and 2) the available seats are allocated to members of historically privileged communities at the cost of access for EL students (who, research shows, uniquely benefit from learning in their native language as they learn English).

Unfortunately, though, at present, there are real structural limits on the growth of quality DLI programs. The UCLA paper aptly summarizes some of the forum’s discussions on this score.

First and foremost, DLI growth is limited by shortages of qualified bilingual teachers. The U.S. teaching force is largely monolingual and English-dominant. Precise data on this are scarce, but we do know that only around one in eight U.S. teachers speaks a non-English language at home, compared to nearly one-quarter of U.S. children.

There are many reasons for this, but U.S. teacher licensure policies play a role. In most states, linguistically-diverse teacher candidates face multiple obstacles to becoming licensed. Most states require teacher-candidates to pass a licensure assessment in English, even if they will be teaching and working predominantly in another language. State licensure systems do not always recognize credentials that immigrant educators have earned overseas.

There are workarounds—DLI schools can hire foreign teachers. But these teachers are generally only permitted to work in the country for three years. Constantly replacing them is costly. Also, teachers trained overseas in other countries’ education systems often face a steep learning curve when engaging with the U.S. public education system.

Schools can also hire non-native speakers of the non-English target language (English-dominant adults who studied Mandarin in college, for instance). But these teachers may not have the depth of language proficiency and/or breadth of vocabulary to effectively model and transmit language proficiency for all students.

There are other, more permanent solutions. For instance, districts can build alternative teacher training pathways designed to help bilingual teacher-candidates become licensed. Furthermore, traditional teacher training programs at colleges and universities can build partnerships with their institutions’ foreign language departments. Indeed, there is scarce evidence that teacher licensure systems provide meaningful quality control for determining the quality of states’ bilingual teaching forces. The UCLA report paraphrases a forum participant thus: “There is no consensus in the field about how exactly to best prepare bilingual teachers to address the burgeoning need.”

Expanding bilingual teacher training pipelines is only the first step we need to take to advance equity in DLI, however. The UCLA paper identifies a range of next-order equity problems for these schools. High-quality, culturally rich multilingual curricula are rare. Unproven DLI educational materials—especially education technology—are common. Teacher and administrator linguistic and cultural biases too often shape the tracks that DLI programs take.

These challenges suggest that DLI’s many goals—political and pedagogical—are in tension with one another.

Above all, these challenges suggest that DLI’s many goals—political and pedagogical—are in tension with one another. Consider the daunting list of proposed DLI goals. Dual-language immersion programs promise to advance English learners’ linguistic and academic development, help develop ELs’ pride in their cultural identities, allow a diverse range of students to become bilingual and biliterate, support school integration, close achievement gaps, aid in immigrant integration processes, diversify the teaching force, and improve bilingual education’s political fortunes.

The report notes all of these DLI priorities, and highlights how this breadth of priorities hampers efforts to improve the actual quality and equity of DLI programs. For instance, policies to foster linguistic, racial, and socioeconomic integration may not always be perfectly compatible with policies to improve ELs’ linguistic development and academic outcomes. What’s more, none of the foregoing may work smoothly with efforts to raise the political appeal of DLI. Slater and Gándara, the report’s authors, noted that, in a survey of thirty-two dual-language immersion administrators (across twenty-three states) conducted to accompany their report, they found “a broad recognition of the critical importance of equity, but also a degree of uncertainty as to the depth and breadth of equity issues and their potential impact on DLI programs.”

The paper, and the forum it synthesizes, are encouraging first steps towards clarifying what DLI is for, and how programs can work towards that objective. But if DLI programs are to deliver on their many political promises without undermining their pedagogical purposes, it’s important that those steps don’t end the field’s journey.