Abstract
The signing of the Bilingual Education Act in 1968 presumed that the federal government had a role to play in the equitable education of immigrant and English learner students, who had been largely invisible to most of the country. Initial language of the Act was intended to build on these students’ assets. Nonetheless, the language that survived in the BEA limited its effectiveness and created ongoing challenges for educators, including an ever-changing definition of the goals and purposes of funding; a deficit rather than an asset-based orientation that cast English learners as “remedial students”; unresolved tensions between the goals of desegregation and bilingual education; and fluctuating and inadequate attention to the capacity development needs of the field. The latest iteration of the ESEA removed the BEA from federal legislation altogether, failed to resolve any of the ongoing issues, and reinforced the remedial framing of ELs, arguably placing them at even greater educational risk.
In 1968, two and a half years after President Lyndon Johnson signed the Elementary and Secondary Education Act (ESEA) into law, it was amended to add Title VII, the Bilingual Education Act (BEA). This was the first time that the federal government had acknowledged that English learners—at that time termed limited English-speaking (LES) or limited English proficient (LEP) students1—experienced unique challenges in meeting the same educational goals as English-speaking children, and that the federal government should play a role in helping them meet those challenges. Thus not only had the federal government stepped up to help equalize resources for children with significant economic need, but it had also declared through the new Title VII that it would also address the needs of children who did not speak English or did not speak it well. The passage of the BEA was an important turning point in the role of the federal government, yet several aspects of the new law and the debates that shaped it limited its effectiveness and created ongoing challenges for educators. Those challenges include an ever-changing definition of the goals and purposes of the funding, making it difficult to sustain bilingual programs over time; a deficit rather than an asset-based orientation that cast English learners (ELs) as remedial students; unresolved tensions between the goals of desegregation and bilingual education; and fluctuating and inadequate attention to the capacity development needs of the field.
Because the BEA was framed as targeting poverty as much as language needs, only children living in homes with an income below $3,000 and from homes where English was not spoken were allowed to receive the benefits, and the targeted children could not include English speakers. Thus, if programs were to be designed to benefit only LEP students who were also poor, it required that the English learners be separated from their nonpoor and English-speaking peers. Hence the funds were generally restricted to linguistically and economically segregated schools (Moran 1988). The law thus perhaps inadvertently supported the segregation of EL students. This would set the stage for an ongoing tension between segregation and bilingual education that would resonate through the years, and be revisited each time Title VII was reauthorized.
Arguments for serving only LEP students included that the impact of the limited funding should not be diluted by serving children who did not need language assistance. In the early years of the BEA, the benefits were especially modest: no funds were appropriated in 1968, and in the following year only $7.5 million were provided. Rachel Moran asserts that between 1968 and 1973 no more than $5 to $6 per child in need was appropriated (1988). Thus it was understandable that many educators would want to preserve these modest funds for children with the greatest need. Of course, arguments in favor of including non-LEP students primarily focused on reducing the linguistic isolation and segregation of EL students, and on the role they could serve as important English-language models. Research since that time has tended to support this belief, though education policy has not generally reflected it (Gass, Mackey, and Pica 1998; Gifford and Valdés 2006).
The first iteration of the BEA only sought to provide very modest grant in aid funds to “carry out new and imaginative elementary and secondary school programs designed to meet these special educational needs” (Title VII, 1968, Sec. 702, 81 stat. at 816). Thus it was intentionally vague in its goals and purposes, and carefully crafted to not appear to usurp local authority. Although titled the Bilingual Education Act, the law actually skirted any definition of bilingual education. Senator Yarborough of Texas, the bill’s chief sponsor, had gone on record in 1967 as wanting “the creation of bilingual-bicultural programs, the teaching of Spanish as a native language” “designed to impart to Spanish-speaking students a knowledge and pride in their culture” (Schneider 1976, 22). This was a far cry from the compensatory education frame of the ESEA, which sought to remediate the deficits wrought by poverty. But, as Natalia Mehlman Petrzela recounts, passage of the bill depended on its fitting into the overall objectives of the ESEA and not challenging the popular notion of the melting pot into which immigrants were expected to relinquish their distinctive cultural features (2010). As Moran notes, “The vague statement of purpose masked fundamental differences over whether the programs were designed to promote assimilation by overcoming a language ‘deficiency’ or were intended to foster pluralism by acknowledging a linguistic asset” (1988, 1273). The definition and goals of the program under the ESEA would continue to be debated in Congress and to create confusion and instability in the funding of programs at the state level.
Finally, in good part because the goals and purposes of bilingual education were left ill defined, it was assumed that good educators could simply come up with a strong program out of whole cloth to support their English learners. But in fact teachers had not been trained to provide bilingual instruction, good materials and curriculum had not been developed, and both basic and applied research on effective practice was just beginning in the United States. In other words, no capacity had been built to mount the programs in the early years of the act, and the meager funds appropriated were supposed to be used for developing “innovative” programs. Yet there was little to no infrastructure with which to innovate.
REAUTHORIZATIONS AND THE EVOLVING LEGISLATION
The first reauthorization of the BEA came in 1974 and clarified some of the issues left vague in the initial writing of the law. It specifically incorporated language to address equal educational opportunity and linked it to bilingual education programs: “the Congress declares it the policy of the United States to establish equal educational opportunity for all children (a) to encourage the establishment and operation… of education programs using bilingual education practices, techniques, and methods” (Title VII, 1974, Sec. 702[a]). Bilingual education was defined as “instruction given in, and study of, English, and, to the extent necessary to allow a child to progress effectively through the educational system, the native language” (Wiese and Garcia 1998, 5 [emphasis added]). This would set the tone for two decades, defining bilingual instruction as transitional, use of the first language being only a temporary means to another end. The 1974 reauthorization also importantly removed the poverty requirement and allowed English-speaking children to enroll in bilingual education programs to “acquire an understanding of the cultural heritage of the children of limited English-speaking ability” (Title VII, 1974, Sec. 703 [a][4][B]). Thus, this version of the BEA also attempted to promote cultural understanding and to reduce the segregation implied in the implementation of a program that could only serve poor LEP children. Another addition to the BEA in 1974 was significant: a graduate fellowship program for study in the field of training teachers for bilingual education programs, a clear acknowledgment that educators were expected to mount programs without any pipeline of qualified candidates to implement them. The 1974 reauthorization also included a program to support the development of materials to be used in bilingual programs.
This honeymoon for bilingual education, however, was to be short lived. The civil rights era was coming to a close and conservative forces sought to restore the old order. The programs would quickly come under attack just as desegregation efforts were also being turned back by increasingly conservative courts and presidential administrations (Gándara, Moran, and Garcia 2004). In the 1978 reauthorization, support for bilingual education would begin to evaporate. Language was added to the act to emphasize acquisition of English skills over bilingualism, and by the 1984 reauthorization the legislation made clear that only transitional bilingual education was favored—75 percent of all funds had to be expended on such programs. The goal of the Bilingual Education Act was now to provide “structured English-language instruction, and, to the extent necessary to allow a child to achieve competence in the English language, instruction in the child’s native language” (Title VII, 1984, Sec. 703 [a][4][A]). That is, the native language was to be used only insofar as it furthered the goal of English acquisition. It is also important that because English was the goal, little was mentioned and nothing measured with respect to primary language achievement. Integration of these students into the mainstream was assumed as a by-product of becoming more Americanized through acquisition of English.
In a seminal 1984 article describing the orientations or philosophies of policymakers regarding language policies, Richard Ruiz (1984) contends three basic orientations: language as a problem, something that needs to be fixed; language as a right, something that must be legally protected; and language as a resource, something that is an asset. By creating legislation that casts language as a problem, it results in practices that focus neither on social justice (a right), nor on asset development (a resource). Clearly, by 1984, if not before, languages other than English were seen as a problem.
In 1984, the door was also opened for purported bilingual programs that did not incorporate a primary language at all: Special Alternative Instructional Programs (SAIPs) were written into the law and 4 percent of funds were to go to these programs. In the 1988 reauthorization, the door was opened further to nonbilingual programs, increasing the funding to 25 percent for SAIPs and limiting the time that a student could spend in a bilingual program: “No student may be enrolled in a bilingual program… for a period of more than 3 years” (Title VII, 1988, Sec. 7021 [d][3][A]). This effectively ended the notion that bilingual education could be used for the purpose of developing bilingual individuals because the program was designed to end as soon as enough English was acquired. No research supported the idea that enough English to keep pace with native English speakers in academic subjects could be acquired within a maximum of three years. In fact, the research has long since concluded otherwise (Weise and Garcia 1998). Many factors determine how quickly a student becomes proficient in a second language, including prior schooling, age, parents’ socioeconomic status, contact with the language outside of school, and quality of education, but as a general rule researchers have converged on the finding that it takes at least five to seven years to develop mastery of academic English (Hakuta, Butler, and Witt 2000).
During the Clinton (Democratic) administration, the march away from primary language instruction took an about face. The 1994 reauthorization of the ESEA, under the title Improving America’s Schools Act (IASA), retained Title VII but changed its focus considerably. This time an emphasis on bilingualism was written into the law. Although it actually allowed the 25 percent cap on funding of SAIPs to be lifted if districts could show they were not able to mount a bilingual program, this version of Title VII gave “priority to applications which provide for the development of bilingual proficiency both in English and another language for all participating students” (Title VII, 1994, Sec. 7116[i][1]). By including “all participating students,” the law invoked the possibility that ELs could be studying alongside English speakers, breaking down their isolation and providing a potential asset for all students. With support to build a cadre of well-prepared teachers, instructional materials, and pedagogical strategies that could equalize education for English learners, it appeared that support for bilingual instruction was back. However this was not to be, as the pendulum would again swing back with the next reauthorization of the ESEA in 2001.
Although support for the BEA fluctuated, its passage probably helped spur the development of policy in the states. Whereas before 1968 no state had a pro-bilingual education policy on the books (Moran 1988), by 1983 all fifty permitted bilingual education and nine required some form of dual language instruction (Ovando and Collier 1985).
HISTORIC TRENDS IN EDUCATION OF ENGLISH LEARNERS
Bilingual education was not new to the United States at the time of the passage of the BEA. In fact, during the history of the nation it had thrived in many parts of the country, and especially in the Midwest. By the end of the nineteenth century, about a dozen states had passed pro-bilingual laws (Kloss 1998). However, an unprecedented wave of immigration to the country and a war with Germany (the language group with the most bilingual schools in the United States) in the first decades of the twentieth century brought a swift end to pro-bilingual education policies. Americanization became the policy regarding children of immigrants and included transitioning them to English as rapidly as possible. Speaking a language other than English was inconsistent with the American melting pot.
The Immigration and Nationality Act of 1965, which was signed into law the same year as the ESEA, would launch a stream of immigration that would begin to change the face of the nation. There had long been large pockets of children who did not speak English along the southern border and throughout the Southwest, a legacy of the war with Mexico in 1848, which had incorporated a third of the landmass of Mexico and its citizens into the United States. These children, however, remained invisible to most of the country.
The only immigrants that were highly visible in the late 1950s and early 1960s were Cubans fleeing the Castro regime. They were warmly received by the American public in the anticommunist tenor of the times and legislation was quickly passed to aid their integration into American society. The Cuban migration to South Florida had an important impact on the way that bilingual instruction came to be viewed. Unlike the Mexicans of the Southwest in almost every way (such as wealth, status, education, race) except language, the Cubans established bilingual schools where their children could learn in two languages while they waited to return to the Spanish-speaking island as soon as Castro was deposed. The Coral Way School, the first established to meet the needs of the Cuban children, became a model of bilingual education for the nation, and it clearly supported maintenance of the Spanish language.
A BRIEF OVERVIEW OF THE RESEARCH ON PROGRAMS FOR ENGLISH LEARNERS
Bilingual programs has often been used as an umbrella term for all programs serving ELs, even those in which no primary language is used (Lessow-Hurley 2004). Thus the first study commissioned by the Department of Education to examine the effectiveness of bilingual education was conducted by American Institutes for Research in 1977 and 1978 (Danoff 1978). It compared students in thirty-eight Title VII programs with similar students in English as a Second Language (presumably English only) classrooms and found no particular impact of the bilingual programs on test scores. The study was criticized by many researchers for including programs in the two groups solely on the basis of program labels without examining the actual educational treatment provided or controlling for differences in the students assigned to the programs (August and Hakuta 1998; Willig 1985). This and other methodological problems left the findings of the study in significant dispute. Perhaps most important, however, the study paid little attention to the fundamental questions of teacher skill and preparedness, curriculum, materials, and pedagogy, and it tested differences after only a few months of exposure to the programs.
A second, large-scale comparative study was commissioned, again by the Department of Education, about a decade later. Conducted by David Ramirez and his colleagues (1991), it was much more complex and involved a four-year comparison of English immersion, early-exit transitional bilingual (usually lasting no more than three years), and late-exit bilingual programs (normally continued through the end of elementary grades and focusing on biliteracy) on various achievement outcomes in both English and Spanish. The researchers were careful to examine the instruction provided in each, and the time dedicated to instructing in each language and in eliciting language from students, as well as teacher characteristics and pedagogical strategies. Unfortunately, the researchers found significant differences in the students assigned to each program type, late-exit students being much more low income and having a significantly lesser chance of having attended preschool—all characteristics not controlled in the nonrandomly assigned classrooms. Also, programs could not usually be compared with others in the same district or school (because they did not exist) so that school and district effects were likely powerful contributors to uncontrolled differences among the groups (Meyer and Fienberg 1992). Moreover, very heavy attrition of students called many of the results into question. Still, the researchers found a small positive difference in first grade reading outcomes for the early-exit bilingual model over the English immersion, but overall no significantly different outcomes for the three groups of students. The researchers did, however, note that the trend lines for test scores were in a steep upward trajectory for the bilingually educated students at the point of termination of the study at fourth grade. Longitudinal studies since that time find that bilingual and dual language program test scores tend to exceed those of English-only programs at about fifth grade (see, for example, Genesee et al. 2006; Umansky and Reardon 2014; Valentino and Reardon 2015). Among the more influential smaller studies during the period was the one Keith Baker and Adriana de Kanter conducted in 1981, in which they reviewed twenty-eight studies that met sufficient methodological rigor to be included in their qualitative analysis of the programs: yes, the evaluation found positive effects for bilingual instruction or, no, it did not. No attempt was made to quantify the degree of effectiveness, and they had no firsthand knowledge of the “treatments.” This very widely cited study found that “the case for the effectiveness of transitional bilingual education is so weak that exclusive reliance on this instructional method is clearly not justified” (Baker and de Kanter 1981, 1).2 In other words, they did not find a definitively superior outcome for either of the two methods tested. Yet this study, coupled with the American Institutes for Research (AIR) study, had a very important impact on the language of the reauthorization of Title VII in 1984. The new amendments included language that specifically opened up funding for English-only programs; it was no longer necessary to incorporate the students’ primary language to gain funding under the Bilingual Education Act.
A NEW GENERATION OF RESEARCH ON BILINGUALISM AND ITS BENEFITS
Both the knowledge base regarding language acquisition and methodological techniques for studying it have developed substantially over the last several decades. Newer studies address many of the limitations of earlier research. In a best evidence meta-analytic study, Robert Slavin and Alan Cheung find that among the seventeen studies that met their strict methodological criteria for inclusion, thirteen favored bilingual programs (all Spanish-English), and four found no differences (2005).3 This study, in contrast to the earlier Baker and de Kanter study, incorporated quantitative methods to determine the actual effect sizes of the treatments. The effect size for the averaged score differences was between 0.33 and 0.45, indicating a medium positive effect. Across both the Slavin and Cheung study and four other rigorous meta-analyses reviewed by Diane August and her colleagues, the researchers find “differences in favor of native-language instruction, with effect sizes ranges from small to moderate” (August, Goldenberg, and Rueda 2010, 143). They also note that the better the technical quality of the studies, the larger were the effect sizes. In a synthesis of the most rigorous research on reading instructional approaches for English learners, Claude Goldenberg also concludes that “teaching students to read in their first language promotes higher levels of reading achievement in English” (2008, 14). This finding is often thought to be counterintuitive, though it is well supported by theory as well as by data (Durgunoğlu, Nagy, and Hancin-Bhatt 1993; Verhoeven 1994). The theories underlying this finding are that of transfer, knowledge acquired in one language is transferred to additional languages as they are acquired, and comprehensible input, individuals learn more efficiently when they can understand at least part of what is being communicated (Cummins 1981; Krashen 1987).
Most evaluation research on bilingual education has focused narrowly on short-term outcomes for reading and sometimes math in English only. Very little attention has been paid to longer-term effects or to other potential outcomes. In fact, many of the studies that have found no difference or less positive effects for bilingual instruction have been based on very short-term analyses. Fred Genesee and his colleagues, reporting on a synthesis of research on English learners, note thatEvaluations conducted in the early years of a program (Grades K-3) typically reveal that students in bilingual education scored below grade level… [but] Almost all evaluations 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. (Genesee et al. 2006, 201)
A recent study that followed thousands of students in one large school district in transitional bilingual (aka early exit), dual language bilingual (longer term, incorporating English speakers and English learners), and English-only programs beginning in kindergarten and following them into high school found that the EL students who had remained in bilingual instruction, and especially dual language bilingual programs, outperformed the students in English only instruction on all measures. Specifically, they ultimately reclassified to English proficient at higher rates and scored higher on both English-language arts and measures of English proficiency (Umansky and Reardon 2014).
With respect to outcomes other than test scores or English proficiency, the body of research on a host of outcomes is large and growing. Ellen Bialystok and her colleagues find in a series of innovative studies that bilingually educated students have greater cognitive flexibility, working memory, and executive functioning, such as concentration (Bialystok 2001; Bialystok and Majumder 1998; Bialystok and Craik 2010). Alejandro Portes and Lingxin Hao find that bilingual students have more cohesive family relations and fewer behavior problems in school (2002). They attribute this, as have others, to greater communication and parental authority fostered by parents and children communicating in the same language. Lucrecia Santibañez and Maria Estela Zárate, analyzing longitudinal data from the Educational Longitudinal Study (ELS) of the Department of Education, find that students from bilingual homes who maintain their bilingualism into high school are more likely to go to college than those who lose the home language, and that bilingual Latinos are more likely to go to four-year colleges (2014).4 Rubén Rumbaut, in analyses of two longitudinal data sets with more than six thousand subjects from the southern California region, finds that those students from immigrant backgrounds who maintain bilingual skills are less likely to drop out of high school and more likely to secure higher level positions in the workforce, and to earn more at those jobs than monolinguals (2014). Most of these studies also find that the benefits of bilingualism increase with the level of fluency the individual has in both languages (Gándara 2015). These findings call attention to the need to be more specific about the goals of instructional programs for English learners when comparing outcomes. If the goal is simply oral English proficiency, it may not matter greatly which program is provided as long as the quality is high. However, if educators are concerned about a host of other potential outcomes, research suggests that bilingual instruction may be desirable, at least as an option.
LEGAL RIGHTS OF ENGLISH LEARNERS
Educational policies for the instruction of English learners are created in the courts and through administrative regulations as well as by acts of Congress, and the BEA has proved a catalyst for some of these rights. (Sometimes they are also created at the ballot box, as in the anti-bilingual initiatives that were voted on in California, Arizona, and Massachusetts between 1998 and 2002.) The Office for Civil Rights (OCR) has played an active role in advocating for the rights of EL students to receive appropriate services. It first entered the fray in interpreting the national origin clause of the Title VI of the Civil Rights Act of 1964 as prohibiting discrimination based on language. This stance remains the law of the land today. Following on the passage of the BEA, in 1970 OCR issued a memorandum that came to be known as the May 25th Memorandum, putting school districts on notice that they “must take affirmative steps to rectify the language deficiency in order to open its instructional program to these [LEP] students” (Office for Civil Rights 1970). The memorandum also included notice that districts would be reviewed for compliance and would have to prove they had such a program in place and that it did not operate as an educational dead end. OCR was highly conscious of the segregative potential of programs that grouped ELs separately and so included that the ELs’ needs should be met “as soon as possible.”
In 1973, the Supreme Court ruled in Keyes v. School District No.1 (13 U.S. 189) that the Denver schools must provide relief from segregation for “Hispano” students as well as for blacks. In 1975 the district court, to which Keyes had been remanded to develop a desegregation plan, clarified the guarantee of a bilingual education program for the Latino students in its ruling (Moran 2013). The next critical event in the development of legal rights for English learners occurred in 1974 with the Supreme Court ruling, Lau v. Nichols (414 U.S. 563), in which 1,856 Chinese-speaking children in San Francisco argued that they were being denied an equal education because they could not understand the classroom instruction and no accommodations were made for their language difference. The Court ruled that the school district had to take affirmative steps to provide access to the same curriculum that English-speaking students received, but did not instruct the schools about how this should happen.
Days after the Lau ruling, Congress passed the Equal Educational Opportunities Act (EEOA), which helped clarify Lau, requiring school districts to “take appropriate action to overcome language barriers that impede equal participation by its students in its instructional programs.” Appropriate action was defined by the Fifth Circuit in the 1981 Castañeda v. Pickard decision (648 F.2d 989), setting the three-prong standard that has survived, at least in theory, to the present and includes a program based on recognized theory; faithfully implemented according to the theory, including adequate resources for implementation; and that demonstrated effectiveness over time.
Finally, in 1982, the Supreme Court ruled against the state of Texas, which had passed a statute requiring undocumented children to pay the state for the costs of their public education. The Court found that “Public education has a pivotal role in maintaining the fabric of our society” and “the deprivation of education takes an inestimable toll on the social, economic, intellectual, and psychological well-being of the individual” (Plyler v. Doe, 457 U.S. 203). In other words, because the children found themselves in the United States through no fault of their own, to deprive them of an education because their parents could not afford to pay for it served no rational purpose. Although Plyler did not direct itself to the language education of the children, it clarified that all children within U.S. borders were to be provided with a public education at the state’s expense, and it had to be an education that would overcome language barriers that impeded their equal participation in school, as already stipulated in Lau v. Nichols and in the EEOA.
THE UNDOING OF LEGAL RIGHTS
As Congress was redefining the BEA with each reauthorization, the rights for English learners that had been hard won through a series of court decisions were also being redefined. Congressional actions and court rulings would gradually come into alignment to undermine the protections that existed. Although Lau v. Nichols appeared to require that school districts provide access to the regular curriculum for all EL students, the right to sue a district for not providing this access was taken away from individuals in a 1983 Supreme Court decision. The Court found that Title VI of the Civil Rights Act authorized compensatory relief only for purposeful wrongs, not actions involving adverse effects, or “disparate impact.” Lau suffered another blow in 2001 when the Court decided in Alexander v. Sandoval (532 U.S. 275) that there is no private right of action under Title VI disparate impact regulations. Private plaintiffs can sue only for intentional discrimination, which is virtually impossible to prove. As a result, if federal agencies are too overburdened, or not interested in filing an action, children are left without recourse under Title VI.
The loss of Title VI protections made the EEOA the best alternative to seek redress because it allows for private right of action. That is, an individual or a group of individuals can bring a case against a school district for failing to meet its obligations to provide equal access to the curriculum for its English learners. This is exactly what Miriam Flores did in 1992 in Arizona, claiming that the state did not invest enough funding in her education to make true access possible. In 2009, the case ended up in the Supreme Court, which ruled in Horne v. Flores (557 U.S. 433) that the federal court had overstepped its bounds in ordering the school district to increase its investment in the education of its English learners, as there was no rational relationship between the amount of funds expended and the quality of an instructional program. The Court effectively obviated the second prong of Castañeda requiring that schools provide sufficient resources to ensure faithful implementation of the program. The Court also added in dicta that in imposing a statewide requirement that schools serving English learners provide Structured English Immersion (SEI), Arizona had implemented a program that was “significantly more effective than bilingual education.” This appeared to suggest that the state had also met the third prong—that a program must demonstrate its effectiveness over time. However, the Court’s claim that SEI is more effective than bilingual instruction was not warranted by research (Martinez-Wenzl, Pérez, and Gándara 2012). Although the courts have attempted to rein in the legal protections for EL students, OCR has once again engaged the battle. On January 7, 2015, OCR and the Department of Justice released guidance for schools and districts on their civil rights obligations to EL students, reiterating the legal requirements under Lau and the EEOA.
A NEW ERA OF ESEA AND THE DISAPPEARANCE OF THE BEA
The last reauthorization of the ESEA came in 2001 with a bipartisan effort to set high standards for America’s schools and to place accountability at the center of the framework. Titled the No Child Left Behind Act (NCLB), it declared that all children would be proficient in basic academic standards by the year 2014. Of course, because it has never been possible for all children to meet any particular goal, many educators immediately predicted its failure (Sunderman et al. 2004). This time, the reauthorization did another about face and eliminated the Bilingual Education Act altogether. In its place was a new Title III, renamed Language Instruction for Limited English Proficient and Immigrant Students. The Office of Bilingual Education and Minority Language Affairs established in 1974 as the arm of the Department of Health, Education, and Welfare (HEW) to deal with the implementation of the law disappeared. It was replaced by the Office of English Language Acquisition, Language Enhancement, and Academic Achievement for Limited English Proficient Students. This new name reflected the shift away from bilingualism as a goal and emphasized the perspective that EL students’ defining characteristic is their lack of English proficiency.
Title III of NCLB states its purpose as “to help ensure that children who are limited English proficient, including immigrant children and youth, attain English proficiency, develop high levels of academic attainment in English, and meet the same challenging State academic content and student academic achievement standards as all children are expected to meet” (Title III, 2001, Sec. 3102 (1)). It aims to do this by providing “flexibility to implement language instruction educational programs, based on scientifically based research on teaching limited English proficient children, that the agencies believe to be the most effective for teaching English” (Title III, 2001, Sec. 3102 (9)). The focus of Title III is entirely on English proficiency outcomes and schools may use any method they “believe” is most effective for this purpose. Moreover, in the strict accountability system of NCLB, schools have had to show progress on three Annual Measurable Achievement Objectives (AMAOs) for their ELs to receive formula grants from the federal government. That is, unlike past practices in Title VII when funds were competitive, Title III provides funds on a formula basis, but the funds depend on achieving specific accountability goals.
AMAO 1 requires that districts and schools show measurable improvement in the percentage of ELs who “make progress” in learning English. Progress is to be defined by the state and varies widely among states. The 2010 evaluation of Title III accountability found that the range was between 20 percent and 85 percent of students achieving the state standard (Boyle et al. 2010). AMAO 2 requires that EL students make annual progress toward achieving English proficiency, which is generally described as the percent of students who achieve English proficiency. However, this measurement is also fraught with variability, because states can adopt their assessments and set their own standards for proficiency, so that as in AMAO 1, outcomes are dependent on where the states set the bar for proficiency. Finally, AMAO 3 requires that a set percentage of EL students achieve proficiency in math and English language arts, and now science using the same tests that all students are tested on under Title I. Thus, with respect to assessment requirements, ELs are governed by both Title I and Title III. Given that by definition ELs do not have enough command of English to be instructed in an English only setting without some kind of language support, it seems inconsistent to test them in English and have those scores count toward an evaluation of a school’s or district’s academic program. It is well established that tests given in English to students who do not have a good command of English are likely to be both unreliable and invalid (Abedi 2004; Abedi and Gándara 2006).
Title III acknowledged this problem:[ELs] shall be assessed in a valid and reliable manner and provided reasonable accommodations on assessments administered to such students under this paragraph, including, to the extent practicable, assessments in the language and form most likely to yield accurate data on what such students know and can do in academic content areas, until such students have achieved English language proficiency. (Title III. Sec. 1111[C][III])
In the following subsection, it is noted that this practice should normally continue for at least three years and on a case-by-case basis for up to five years. Yet this language has never been enforced. Across the nation, EL students have been subjected to tests in English as soon as a matter of months after enrolling in school (Menken 2010). With high stakes attached to these testing outcomes, some schools and districts have found themselves in a no-win situation.
In June 2005, nine school districts in California—the state that accounts for at least one-third of all English learners—and three statewide nonprofit organizations banded together under the lead district, Coachella Unified School District, to sue the state of California for failing to comply with the provisions of NCLB that ELs should be assessed in a valid and reliable manner. A survey of districts conducted for the court found that only 3 percent of students were actually provided with accommodations, and most accommodations provided to EL students were not of a kind that have been shown to have any particular impact on their ability to perform on a test they do not fully understand (Haertel 2007). It was also relevant that the state had already developed a Spanish version of the statewide achievement exam, which made “to the extent practicable” very possible at least for this one language group that comprises about 85 percent of the ELs in the state. Moreover, the expert who oversaw the development of the California Standards Test (CST) asserted in a declaration to the court that the test should not be used with students who did not speak English because it was specifically not designed for their use (Haertel 2007). Despite these arguments, in 2007 the plaintiffs lost the case. California has continued to test in English-only even though one-fourth of all students in the state are English learners, and predictably those schools with large numbers of EL students routinely fail to meet adequate yearly progress (AYP) (Bryant et al. 2008).
The consequences of low test scores that result in the subsequent sanctioning of schools with large numbers of EL students have pushed many districts to simply focus on quick-fix English drill programs that ignore the students’ broader learning needs in an effort to raise their test scores in English (Menken 2010; Ravitch 2011). Maryann Zehr reported in 2007 that many schools were abandoning bilingual programs in favor of English-only instruction in order to meet the pressures of NCLB. In Arizona, state policy became four hours of English drill daily to the exclusion of other coursework such as science, math, social studies and other courses offered to mainstream students, in what is an apparent challenge to Lau. Some high school ELs have been unable to graduate with their class because of the requirement to take several hours of English drill daily (Gándara and Orfield 2012; Lillie et al. 2012).
NCLB’s impact has not been entirely negative, however. It is widely believed that its focus on subgroups, and holding schools accountable for their achievement as well as that of their more advantaged students, was and is an important step forward in achieving more equitable educational opportunities. Without a light shining on the problems, no one was likely to pay attention. However, it is obvious that overall EL students have been more hurt than helped by the law. According to National Assessment of Educational Progress (NAEP) scores, they have made no appreciable progress toward closing the gaps with their non-EL peers since before NCLB was instituted and have often been stigmatized as the reason their schools have carried the label of failing (Novak and Fuller 2003).
Nationally, since 1996, the first year for which NAEP shows the gap trend lines for ELs, the gap has not closed for fourth grade math; in fact, over the last decade of NCLB, the gaps have begun to widen between ELs and all others. In 2003, the gap between English learners and English speakers in fourth grade mathematics was 23 points, in 2013 the gap had grown to 25 points; fourth grade reading showed a similar widening, 3 points, over the same period. Gaps at the eighth grade had grown even larger (NAEP 2014). At least from the perspective of math and reading score gaps nationally, educational achievement has not improved for English learners.
THE CURRENT STATE OF EDUCATION POLICY FOR ENGLISH LEARNERS
The legacy of the BEA remains even though the act itself has disappeared. The vagueness about the purposes of bilingual education, the language-as-problem orientation of the law, the failure to adequately address the capacity needs of schools, and the increasing segregation of English learners (programmatically as well in the schools they attend) continue to challenge the field. Ironically, as the research has converged on the many benefits of bilingualism, both for academic as well as for noncognitive outcomes, education policy appears to have moved in the opposite direction. Even while the secretary of education touts the importance of bilingualism—saying, it “is clearly an asset that these kids are coming to school with” that should be “maintained” and “that our kids don’t grow up [bilingual] puts them at a competitive disadvantage”—the federal government actually has no policy to foster bilingualism and maintains no office dedicated to this goal (Maxwell 2013). The focus of the ESEA (NCLB) continues to be on the acquisition of only English, and as quickly as possible. This focus is embodied in the test-driven accountability of NCLB that holds children accountable on tests given in English before they actually know the language.
A group of the nation’s foremost researchers on English learner education formed in 2010 to provide advice to Congress on the reauthorization of Title III of ESEA and those recommendations were updated in 2015.5 A primary problem that the ELL Working Group identified was the lack of longitudinal data on EL students. Once ELs are redesignated as proficient in English, the school is required to track them for two years to ensure that they are progressing adequately without additional services, but no particular intervention is required if they are not, and after two years the students are absorbed into the mainstream. We do not know whether they flounder later, though indications are that this may be so (Slama 2014; Robinson 2011). Another area of concern the ELL Working Group highlighted was the need for enhanced training of teachers who serve EL students. Although NCLB requires that all children have a highly qualified or highly effective teacher, it is silent on what constitutes high qualifications for teachers of EL students. Because it does not, no policy focuses on recruiting teachers with specialized skills for the classrooms serving these students. Surveys suggest that this is the one area of instruction teachers feel most inadequately prepared to undertake (Editorial Projects in Education Research Center 2013; Gándara, Maxwell-Jolly, and Driscoll 2005). Some research suggests that the best qualified teachers are those who meet all the standard definitions of highly qualified and additionally are bilingual (Loeb, Soland, and Fox 2014; Hopkins 2013; de Jong and Harper 2005), but both research and policy are clearly lacking in this area. Moreover, financial support to train the individuals who prepare these highly qualified teachers has not been restored in the current version of the ESEA.
Because the emphasis at the federal level has been solely on the acquisition of English, the great majority of students are placed in temporary programs dedicated to that goal. Thus, a major debate has been raging across the country with respect to when to exit English learners from these special programs designed to teach them English. Some have argued that lowering the bar to program exit (reducing the number and level of test scores required to demonstrate English or subject matter proficiency) benefits these students by introducing them into the mainstream earlier (Hill, Weston, and Hayes 2014; Flores et al. 2009). Others have argued that the evidence that students actually gain greater access to high level coursework on exit is inconsistent (Robinson 2011), and that in fact many fall behind because they exit too early (Slama 2014). It would seem, however, that this debate has missed the point. If programs for English learners were truly strong, there should be no rush to exit them, and if the programs were actually additive, producing high-level skills in two languages, perhaps the students should never exit. Instead the debate has centered on how to more quickly remove students from programs under the assumption that they are stigmatizing, limit access to appropriate coursework, and retard student progress. This may all be true, and some evidence does suggest that it is (Callahan, Wilkinson, and Muller 2010), but one must ask: why not focus on strengthening the programs rather than avoiding them?
THE DEMOGRAPHIC IMPERATIVE
When the BEA was first signed into law, no accurate count of how many students needed services had been taken. Schools did not collect these data and actually had no standard or way to assess them. However, the Congress estimated, based on testimony by Bruce Gaarder at the Office of Education, that even in this period of historically low immigration, approximately three million students used a primary language other than English and needed such services (Moran 1988). Much has changed in the ensuing years. Today the population of ELs has mushroomed, in part because of the changes in immigration law. Since 1980, the number of people five years old and older who speak a language other than English at home in the United States has nearly tripled. Today more than sixty million people, some 20 percent of the total population, use another language at home. Two-thirds of these individuals speak Spanish, the next most common languages being Chinese, French, Tagalog, Vietnamese, and Korean (Ryan 2013). One in five students in American public schools comes from a home in which English is not the primary language, and about 10 percent of all students are designated as English learners at any given time (NCES 2014). Moreover, the overwhelming majority (estimated now as high as 90 percent) of these students are native-born Americans. This is no longer an issue that can be put on the back burner of education policy. But it can either be framed as a problem or a tremendous opportunity for this nation in a globalizing world.
LOOKING FORWARD
EL students are now found in virtually every state and in all the major cities of America, and they are the least likely of any subgroup to graduate high school (Callahan 2013) and rarely are they prepared to enroll in college (Martinez-Wenzl 2014). Today the challenges that ELs face have become a central issue for most large school districts in the country. The stakes are very high. Some studies have predicted that per capita income will decline substantially in the next decade in those states with high EL populations because of the failure to adequately educate these and other underrepresented students (National Center for Public Policy and Higher Education 2005; Kelly and Strawn 2011). California alone is predicted to be one million bachelor’s degrees short of meeting its labor force needs in the next decade (Johnson and Sengupta 2009).
When the BEA was first conceived, Congress evidently believed that these students, who were faring so poorly in the nation’s public schools, could be brought into the mainstream if their language “problem” were remediated. More recently, Title III of NCLB was fashioned in the belief that simply holding these students and their teachers accountable to higher standards would force improvement. But laws have been made and remade without much attention to the research on what these students actually do need. A new ESEA should incorporate what we have learned over the nearly fifty years since the BEA was designed. First, it should revisit the initial impetus for the Bilingual Education Act as articulated by Senator Yarborough, and it should define these students as having both linguistic and cultural assets on which to build. Because overwhelmingly, English learners are also low-income students, Title I should continue to provide support to ameliorate these students’ social and economic disadvantages, which most ELs have in abundance, due in part to the increasingly unequal distribution of income and wealth in the country (Desilver 2013). The successor to Title III, however, could best focus on these students’ assets while supporting their acquisition of academic English.
A new ESEA should identify the students who speak a language other than English with assessments that are standardized across the states. It should also support the states in developing assessment measures that allow them to validly and reliably chart their progress. The students should be assessed in all languages in which they are being taught and progress in all languages should be “counted” as meaningful educational achievements. To track the performance of these students and ensure that adequate services are provided for the period they are needed, EL students’ academic performance should be monitored throughout their K–12 careers. There has been pushback on this by some groups that worry that the label of EL is itself stigmatizing and therefore these students should no longer be identified after exiting a language assistance program. But an education program oriented toward building on assets should not be stigmatizing, and therefore monitoring student performance over time should not invite concern.
Given that EL students who manage to maintain strong dual language skills graduate and go on to college at higher rates than their monolingual peers, and that all ELs benefit from having a teacher who can communicate with them and their families (Hopkins 2013), the new ESEA should provide the wherewithal to train highly skilled bilingual teachers and principals. This would also allow for opening many more dual language and international baccalaureate programs that incorporate both English learners and English speakers. Such programs help reduce the increasing segregation of these students so that they have the opportunity to learn alongside students from different backgrounds, different language groups, and different socioeconomic statuses. In this way they can become much more aware of the opportunities that exist for them in school and in the broader society. Advanced placement (AP), international baccalaureate (IB), Gifted and Talented, and other high-level curricular offerings should be expanded for these students, helping them to stretch their limits. The Seal of Biliteracy is another way of rewarding EL students for building on their assets. It is now awarded by eight states, from Washington to New York, and several other states are actively considering it. The seal on a student’s diploma is earned by demonstrating proficiency in all four modalities (speaking, understanding, reading, and writing) in two or more languages. Research we have conducted suggests that many employers would value this designation in their hiring practices and a new ESEA could encourage these opportunities (Porras, Ee, and Gándara 2014).
Certainly a new ESEA should learn from past errors and avoid the high-stakes accountability system that has had a particularly negative impact on English learners. A new and improved ESEA would emphasize formative assessment that helps teachers better meet their EL students’ needs. Accountability should be achieved through a system that is more sensitive to both the unique challenges and resources that different schools and districts experience. A system such as the education inspectorate used in some European countries can help schools analyze their strengths and weaknesses and find meaningful solutions to their problems. This does not obviate accountability, it simply makes it more sensitive to local circumstances and makes more clear what needs to be changed internally in a school, and can more easily differentiate the particular needs of students, such as English learners (Grubb 2000).
The Common Core State Standards (CCSS) are sweeping the land as forty-three states endeavor to put into place either a national set of common core standards or something like them (some states have rebelled against the homogenization of curriculum but have created something similar but with a different name). The CCSS have particular relevance for EL students because, depending on how a new ESEA interprets accountability for ELs, the CCSS can encourage teachers to focus more of their instruction on language development broadly or can place greater emphasis on English acquisition in an effort to comply with test accountability in English. In the latter case, the testing regimen is likely to place English learners at a greater disadvantage. (Early returns from CCSS testing of ELs in New York, for example, have shown drastic declines in scores for ELs.) If the ESEA continues to pursue the path of high stakes, English-only testing, the potential benefit of CCSS for ELs can be lost and will most likely further erode support for bilingual education at the same time that the research has achieved consensus on its many benefits.
To clarify, I propose a reauthorized ESEA that strengthens the assets that ELs bring to school while attending to their socioeconomic needs both in school and out through Title I (or something like it). This would remove the stigma of being an English learner, invest more in the development of “highly effective” educators for ELs, and provide incentives to create many more dual language programs and integrate ELs with other high-performing peers. In sum, these changes would address the problems outlined in the original BEA and its successors. However, moving beyond merely ameliorating problematic aspects of past legislation to actually valuing those students who bring other languages and cultures to the classroom as assets to the nation could result in a new ESEA much more aligned with twenty-first-century reality.
FOOTNOTES
↵1. Labels for students who are not proficient in English have changed over time, including LES (limited English-speaking), LEP (limited English proficient), EL (English learners), ELL (English-language learners), Emerging Bilinguals, among others. Here, EL and LEP are used interchangeably to refer to this population, as many government documents used and continue to use LEP, though the term is no longer preferred in the field.
↵2. A Google search returns 324 citations to this study in other published articles.
↵3. A best evidence meta-analytic study is one that uses strict methodological criteria for inclusion, eliminating those studies that do not meet these standards.
↵4. ELS data collection began in 2002 with students in the tenth grade and has now followed these young people to their mid-twenties, with the latest data collection occurring in 2012. Data used for this study were from the 2006 follow-up of approximately sixteen thousand students
↵5. See The Working Group on ELL Policy 2015, available at: http://www.ellpolicy.org (accessed July 27, 2015).
- Copyright © 2015 by Russell Sage Foundation. All rights reserved. Printed in the United States of America. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the publisher. Reproduction by the United States Government in whole or in part is permitted for any purpose. Direct correspondence to: Patricia Gándara, gandara{at}gseis.ucla.edu, University of California, Los Angeles, Graduate School of Education & Information Studies, 3329 Moore Hall, Box 951521, Los Angeles, CA 90095.
Open Access Policy: RSF: The Russell Sage Foundation Journal of the Social Sciences is an open access journal. This article is published under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.