Bilingual Education: The Failed Experiment?

The Latino Education Crisis: The Consequences of Failed Social Policies

In this excerpt from their book The Latino Education Crisis: The Consequences of Failed Social Policies (Harvard University Press, 2009), Drs. Patricia Gándara and Frances Contreras discuss the history and implications of the debate around bilingual education in the U.S. and offer an in-depth look at the research reviews that have frequently been cited in those conversations.


Teaching children in two languages so that they can be competent learners and successfully acquire English has become over time such a hot-button issue that many people are reluctant to use the term "bilingual education" for fear of inciting scorn or stopping a conversation cold. Researchers even counsel each other to avoid the phrase so as not to prejudice readers against their research.

Yet while critics in the United States claim that bilingual education is a "failed experiment," most other modern nations consider it the norm and cannot imagine why Americans would prefer an education in only one language. What about the history of bilingual education in the United States has led to the emotional controversy surrounding the teaching of students in two languages?

Historical Background

For the last thirty years, with only brief exceptions, the only government-sanctioned justification for bilingual education in the United States has been as a means to transition students as rapidly as possible into an English-only school experience. The idea that a good bilingual education might actually produce students who are literate in two languages gained initial acceptance in the late 1960s and early 1970s.

Ever since, however, programs have had to defend themselves against the accusation that they were "maintenance" programs, trying to maintain students' primary language as they became fluent in English, as though this were a terrible thing to do to a child. Like "busing," bilingual education has become a lightning rod for those who oppose policies designed to equalize (or enhance) educational opportunity for minority children. Social conservatives have very effectively branded as failures these policies that an abundance of research evidence shows have provided many black and brown children with educational opportunities that had been foreclosed to them.15

Research: Bilingual or English immersion?

Educational evaluation was used very effectively in the late 1970s and early 1980s to undermine bilingual education as a pedagogical strategy. Two expensive, multiyear studies were commissioned by the federal government to answer the perennial (and ultimately unanswerable) question: Which is more effective, bilingual education or English immersion? (One reason this question is so difficult to answer is because there is great confusion about the goals of these programs — for example, should the goal be rapid transition to English? Grade-level academic achievement?

In addition, the nature of the instruction is seldom carefully specified — it is often unclear whether it should entail, for instance, fully qualified bilingual teachers providing rigorous curriculum in two languages, bilingual aides helping to translate some portions of an English curriculum, or sink-or-swim immersion into English.) The first major study, released in 1978, found that "there had been no consistent significant impact" of Title VII bilingual education on English learners.16

Critics of this study noted that it was impossible to know what the "treatment" had been in these bilingual and English-immersion classrooms; the researchers had simply taken at face value the label applied to the classes. In 1991 a second federally sponsored longitudinal study was released with similarly ambiguous results.17 In this study, great care was taken to observe actual instruction in the classrooms. But although the principal investigator, David Ramírez, argued that the trajectories of student achievement strongly favored late-exit ("maintenance") bilingual programs, he conceded that the four-year duration of the study was insufficient to draw definitive conclusions. Importantly, the study also noted that in several cases programs had shifted their language emphasis and no longer looked distinctly different from other categories of programs.

Criticisms of both studies focused on the insurmountable methodological challenges of trying to compare students and programs that are not truly comparable:

  • Schools assign different types of students to English-only and bilingual programs (usually placing in the bilingual programs the students who are least English proficient because they are viewed as needing the most primary language assistance);
  • teachers tend to rely on the language they feel most comfortable in rather than the language mix that is mandated by the program; and
  • students in low-income schools come and go so quickly that few are still in the same school or program after only a few years, making long-term assessments of outcomes unreliable.18

Moreover, those students who do not move are likely to be more advantaged than those students who are highly mobile, which also biases study outcomes.19

There has been no major empirical study of dual-language or "two-way" programs, those that mix English speakers and non-English speakers in roughly equal proportions with the objective of fostering full bilingualism and biliteracy in both groups. Smaller studies, however, suggest that this is the most effective strategy for educating both groups of students to become competent bilinguals without sacrificing English development.20

In a major meta-analysis funded by the U.S. Department of Education that looked at bilingual and two-way programs compared to English-only programs for English learners, researchers concluded:

Evaluations conducted in the early years of a program (Grades K-3) typically reveal that students in bilingual education scored below grade level (and sometimes very low) and performed either lower than or equivalent to their comparison group peers (i.e., ELL [English-language learner] students in mainstream English, SEI [structured English immersion]/ESL [English as a second language], or EO [English-only] students in mainstream classrooms).
Almost all evaluation of students at the end of elementary school and in middle and high school show that the educational outcomes of bilingually educated students, especially those in late-exit and two-way programs, were at least comparable to and usually higher than their comparison peers. There was no study of middle school or high school students that found that bilingually educated students were less successful than their comparison peers.21

These findings are important because they may explain results of short-term evaluation studies in which students in bilingual programs sometimes score lower than others in English-only programs when tested early on and only in English. It seems that students require more time to become competent in both English and the primary language. Dual-language, or two-way, programs have the added advantage of helping to reduce the linguistic isolation that so many Latino students experience.

The strategy has its own challenges, including the need to cluster the appropriate numbers of both groups and protect against the social power imbalances that can occur in these settings, but their potential for increasing the academic success of English learners (as well as English speakers) would appear to far outweigh these potential pitfalls.22

Research: Meta-analyses on effectiveness

The federal government has also sponsored research syntheses. The first of these, requested in 1980 by the U.S. Department of Education's Office of Planning and Budget, was charged with reviewing the research literature on the effectiveness of bilingual education.23 Keith Baker and Adriana de Kanter conducted a simple meta-analysis on twenty-eight studies of bilingual education that they considered sufficiently methodologically sound, and concluded that there was not enough evidence in favor of transitional bilingual education to mandate it as the favored approach for educating English learners.

The study was cited widely for many years afterward, as bilingual education came under increasing attack during the Reagan administration. The Baker and de Kanter study itself, however, was widely criticized for being biased in the studies selected and methodologically weak in its simple "up or down vote" methods.

Questions about their meta-analysis gave rise to a series of re-analyses by other researchers, most of which concluded that when strict methodological criteria were applied to the selection of the studies, such as only including studies that had well-defined control groups, bilingual programs tended to show better outcomes than English-only programs.24

In the mid 1990s, in an effort to find definitive answers, the National Research Council commissioned a major study of what was known about educating language-minority students.25 It concluded, "When socioeconomic status is controlled, bilingualism shows no negative effects on the overall linguistic, cognitive, or social development of children, and may even provide general advantages in these areas of mental functioning." In addition, "use of the child's native language does not impede the acquisition of English."26

Research: Reading in the native language

A subsequent study also commissioned by the National Research Council on preventing reading difficulties in young children concurred: "If language minority children arrive at school with no proficiency in English but speaking a language for which there are instructional guides, learning materials, and locally available proficient teachers, they should be taught how to read in their native language."27

More recently, a synthesis study commissioned by the U.S. Department of Education and the National Institute of Child Health and Human Development also concluded that it was generally preferable to teach Spanish-speaking students to read in Spanish where possible. The Bush administration, however, refused to release it with its imprimatur in spite of two peer reviews that concurred with its findings.28

Thus while the debates over primary language — or bilingual — instruction have continued unabated, the major research syntheses that have been commissioned by the federal government have increasingly concluded that use of the primary language in instruction probably holds certain benefits, and at a minimum does not impede English learners' achievement in English. These findings suggest that English-only efforts are likely based on something other than a simple concern for these students' academic welfare.

Policy Implications

The cost of bilingual education

Policymakers have presumed that if bilingual education is not shown to be systematically superior to English-only instruction with respect to achievement outcomes measured in English, then preference should automatically be given either to English-only programs, or to transitional programs that incorporate the least primary language for the shortest period of time. On its face, there is no particular logic to this conclusion. If both programs had more or less equal results in English, but one had the added benefit of teaching some competence in another language, shouldn't that program be considered the superior one? For the most part the answer to this question has been framed as one of cost. Critics of bilingual instruction have long assailed the programs as being a "costly" waste of the taxpayers' money.

Consider the preamble to Proposition 227, California's antibilingual ballot initiative that largely prohibited bilingual education:

(d) Whereas, The public schools of California currently do a poor job of educating immigrant children, wasting financial resources on costly experimental language programs.29

In fact, there has not been a great deal of research on the costs of bilingual education as compared to any other kind of intervention for students who are limited in English. The two studies that have been conducted, however — both by large, reputable research organizations — came to similar conclusions: bilingual education is in fact generally among the least expensive models for meeting the needs of English learners because it requires only that one teacher in a classroom be able to provide instruction in two languages (or two teachers who trade classes for part of the day).

Other models, such as separate ESL pull-out programs or primary language aides who provide educational support, require additional personnel, which increases these programs' cost. Moreover, bilingual teachers tend to be less senior than other teachers, having been hired in larger numbers more recently, and so their pay level, on average, is lower.30 Program costs, then, appear to be more of a red herring than a real issue in the debates over language education.

Program models and instruction

Because earlier studies of language instructional models had not been definitive, proponents and opponents of the different models used these earlier evaluations to make whatever point they chose. On one side are those who assert that English-only instruction not only is more efficient in teaching English learners because students are exposed to more English, but also does not waste valuable time instructing in a language that they do not "need to know." The primary argument for this position is that more "time on task" (learning English) will result in greater and more rapid acquisition of English.31

It is an argument with intuitive appeal. On the other side of this debate, however, are researchers who argue that a fundamental principle of learning is that students learn best when the teacher builds on frameworks of knowledge — or "schemata" — that students already possess. That is, teaching is most effective when new learning is tied to what students already know.32 Thus by teaching students concepts in a language they understand, one can advance the learning in other subject areas while also teaching students English. In fact, as noted, studies comparing bilingual education with English-only approaches consistently find little or no difference in the rate at which students acquire fluency in English.

Reading skills (in English), however, do appear to be enhanced by initial instruction in the primary language, suggesting that learning that occurs in the native language is not "wasted." Hence the earlier mentioned study commissioned by the U.S. Department of Education and the National Institute of Child Health and Human Development found: "Where differences were observed [in studies comparing English-only and bilingual instruction] on average they favored the students in the bilingual program. The meta-analytic results clearly suggest a positive effect [on reading scores] for bilingual instruction that is moderate in size."33

Although well-implemented bilingual programs can claim a "moderate" advantage over English-immersion instruction, no language intervention has erased the gap between English speakers and English learners, or between native-English white children and Latinos. Clearly, other factors are essential to the academic achievement of English learners. Unfortunately our cultural obsession with whether to pursue English-only versus bilingual education has obscured the more critical social and pedagogical issues that need to be studied and understood.


Our policy section is made possible by a generous grant from the Carnegie Corporation. The statements and views expressed are solely the responsibility of the authors.


Gándara, P. and Contreras, F. The Latino Education Crisis: The Consequences of Failed Social Policies. Excerpt from Chapter 4, "Is Language the Problem?" Pps. 128-135. Cambridge, MA: Harvard University Press, 2009.


August, D., and K. Hakuta. 1997. Improving Schooling for Language Minority Children: A Research Agenda. Washington, D.C.: National Research Council, Institute of Medicine.

August, D., and T. Shanahan, eds. 2006. Developing Literacy in Second-Language Learners: Report of the National Literacy Panel on Language-Minority Children and Youth. Mahwah, N.J.: Lawrence Erlbaum.

Baker, K., and A. de Kanter. 1981. Effectiveness of Bilingual Education: A Review of the Literature. Washington, D.C.: U.S. Department of Education, Office of Planning, Budget, and Evaluation.

-------. 1983. "Federal Policy and the Effectiveness of Bilingual Education," In Baker and de Kanter, eds., Bilingual Education: A Reappraisal of Federal Policy. Lexington, Mass.: Lexington Books, pp. 33-86.

Carpenter-Huffman, P., and M. Samulon. 1981. Case Studies of Delivery of Bilingual Education. Santa Monica: RAND Corporation.

Chambers, J., and T. Parrish. 1992. Meeting the Challenge of Diversity: An Evaluation of Programs for Pupils with Limited Proficiency in English, vol. 4: Cost of Programs and Services for LEP Students. Berkeley, Calif.: BW Associates.

Danoff, M., B. Arias, G. Coles, and B. Everett. 1978. Evaluation of the Impact of ESEA Title VII Spanish/English Bilingual Education Program. Palo Alto, Calif.: American Institutes of Research.

Genesee, F., K. Lindholm-Leary, W. Saunders, and D. Christian. 2006. Educating English Language Learners: A Synthesis of Research Evidence. New York: Cambridge University Press.

Gersten, R. 1985. "Structured Immersion for Language Minority Students: Results of a Longitudinal Evaluation." Educational Evaluation and Policy Analysis 7: 187-196.

Greene, J. P. 1998. A Meta-Analysis of the Effectiveness of Bilingual Education. Claremont, Calif.: Thomas Rivera Policy Institute.

National Research Council, Panel on High Risk Youth. 1993. Losing Generation: Adolescents in High-Risk Settings. Washington, D.C.: National Academy Press.

Orfield, G., and S. Eaton, eds. 1996. Dismantling Desegregation: The Quiet Reversal of "Brown v. Board of Education". New York: New Press.

Preamble to California Proposition 227. 1998. Codified as sec. 1, chap. 3, art. 1 of the California Education Code.

Ramírez, J., S. Yuen, D. Ramey, and D. Pasta. 1991. Final Report: Longitudinal Study of Structured English Immersion Strategy, Early-Exit and Late-Exit Bilingual Education Programs for Language-Minority Children. Paper 300-87-0156 prepared for U.S. Department of Education. San Mateo, Calif.: Aguirre International.

Ream, R. 2004. Uprooting Children: Mobility, Social Capital, and Mexican American Underachievement. New York: LFB Scholarly Publishing.

Rolstad, K., K. Mahoney, and G. V. Glass. 2005. "The Big Picture: A Meta-Analysis of Program Effectiveness Research on English Language Learners." Educational Policy, 19: 572-594.

Rossell, C., and K. Baker. 1996. Bilingual Education in Massachusetts: The Emperor Has No Clothes. Boston: Pioneer Institute.

Rumberger, R. 2003. "The Causes and Consequences of Student Mobility." Journal of Negro Education 72: 6-21.

Snow, C., M. Burns, and P. Griffin, eds. 1998. Preventing Reading Difficulties in Young Children. Washington, D.C.: National Academy Press.

Valdés, G. 1997. "Dual-Language Immersion Programs: A Cautionary Note concerning the Education of Language-Minority Students." Harvard Educational Review.

Willig, A. 1985. "A Meta-Analysis of Selected Studies on the Effectiveness of Bilingual Education." Review of Educational Research 55: 269-317.


15 For evidence on the positive outcomes of busing for African American (and white) students, see Orfield and Eaton 1996.

16 Danoff et al. 1976

17 Ramírez et al. 1991.

18 Because most bilingual teachers are primary English speakers, English more often predominates over Spanish in bilingual classrooms. See ibid.

19 Ream 2004; Rumberger 2003.

20 Genesee et al. 2006.

21 Ibid. Bilingual is defined as a program in which English learners are taught through two languages, and two-way is a program in which English learners and English speakers are taught through two languages together. ELL is the acronym for English language learner, a common variation of English learner, which we prefer; SEI is the acronym for structured English immersion, an English-only curriculum with structured adaptation to the needs of English learners; ESL is the acronym for English as second language instruction; and EO refers to English only.

22 See Valdes 1997 for discussion of power imbalances and unintended consequences.

23 Baker and de Kanter 1981.

24 Willig 1985; Greene 1998; Rolstad, Mahoney, and Glass 2005.

25 August and Hakuta 1997.

26 Ibid., p. 28.

27 Snow, Burns, and Griffin 1998, p. 235.

28 August and Shanahan 2006.

29 Preamble to California Proposition 227 of 1998. (accessed May 10, 2008).

30 Carptenter-Huffman and Samulon 1981; Chambers and Parrish 1992.

31 Proponents of this position included prominently Baker and De Kanter 1983; Rossell and Baker 1996; and Gersten 1985.

32 National Research Council 1993.

33 Ibid., p. 397.


For any reprint requests, please contact the author or publisher listed.

More by this author


Add new comment

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.