Bilingual Blame Game 

Response to the Washington Post by James Crawford
August 2001

Census data should indeed help to “sharpen the debate on bilingual education” – a welcome change – rather than provide more blunt instruments to assault the program. Unfortunately, the Post takes the latter approach [“Teach English,” August 9]. If a growing number of children in this country are limited in English, it asserts, then bilingual education must be to blame. 

Let’s look at what the data actually show. 

The Census 2000 Supplementary Survey serves mainly to confirm the obvious: the United States is becoming more linguistically diverse. Since 1990, the number of U.S. residents who speak a language other than English at home increased by 41 percent, while the number who report some difficulty with English increased by 40 percent. 

These trends are impressive, but hardly surprising in a decade when our foreign-born population expanded by 54 percent. It’s no secret that today's immigrants come primarily from non-English-speaking countries. Although a growing percentage of new Americans are highly educated on arrival, most need a few years to develop their English skills. That goes for children as well as adults. 

According to the survey, about 3 million persons aged 5-17 speak English less than “very well.” This is roughly equivalent to state-reported figures on the number of students who need help in overcoming language barriers. The only surprise here is a pleasant one. While school-age children from non-English backgrounds increased by 55 percent over the decade, those with limited English increased by a mere 25 percent. Bilingual fluency is on the rise. 

Nevertheless, the Post finds the Census numbers “shameful.” Dismissing the demographic factors, it prefers to blame the public schools. Since “children pick up languages with relative ease,” it reasons, “the school system ought to be able to deliver near universal fluency. But bilingual programs often involve teaching mainly in Spanish, with rather desultory efforts to teach English.” The editorial goes on to recommend the all-English “immersion” approach recently mandated by voters in two states. Finally, it cites “encouraging” achievement test scores for English learners after California “did away with bilingual education.” 

There is little evidence to support any of these claims, and much to refute them: 

• Because of teacher shortages and limited funding, bilingual education is available to only a minority of eligible children. Before the California vote, for example, 29 percent of limited-English-proficient students were in bilingual classrooms. In most other states, the proportion is lower. 

• While bilingual programs vary in design, survey after survey has shown that English tends to predominate as the language of instruction. 

• In literally hundreds of studies, researchers have documented the effectiveness of bilingual program models. And not just liberal or Latino researchers, mind you. An influential 1998 report was authored by Jay Greene of the Manhattan Institute. 

• So far, not a single scientific study has supported the effectiveness of the one-year, “structured English immersion” model adopted in California and Arizona. Data favoring any kind of all-English approach remain extremely limited. 

• Since its passage three years ago, the California law has failed to deliver on its promise of overnight fluency in English. Oceanside Unified, the “poster district” for immersion proponents, reclassified only 4 percent of its students as English-proficient in 2000 – about half the statewide average. 

• Meanwhile, a Stanford research team found that, thanks to intensive test preparation and a variety of reforms, California test scores have been rising for virtually all groups of students. These include not only English learners transferred to immersion programs, but also the 170,000 children who remain in bilingual education because their parents demanded it. 

• Children learning a second language do tend to “pick up” conversational English quickly. But research shows that it takes considerably longer to acquire academic English – the more complex language skills they need for school. While nobody has found a way to speed up this process, bilingual education is a proven method to facilitate it. 

A House-Senate conference committee is now attempting to reconcile different visions of how to serve these children. Legislation approved by the House would effectively repeal the 1968 Bilingual Education Act and turn its grant programs over to the states. Federal money would be spread thinner and schools would be held less accountable – except for an arbitrary, three-year deadline to “mainstream” students. Worst of all, teacher-training programs would be decimated. 

By contrast, the Senate bill would retain the strongest features of current law, such as quality control and support services, and increase funding to keep pace with immigrant enrollments. The reality is that most children who need help with English today are getting far less than they need – whatever the teaching method. 

One can only hope that, in its deliberations, Congress will move beyond the wild indictments and hasty verdicts about bilingual education and actually look at the evidence. 

Copyright © 2001 by James Crawford. Permission is hereby granted to reproduce this page for free, noncommercial distribution, provided that credit is given and this notice is included. Requests for permission to reproduce in any other form should be emailed to before writing, please read my permissions FAQ