Bilingual Education May Shortchange Some Students, Studies Find

But Prop. 227 said to further limit parental choices

Californians will be armed more with prejudice than evidence as they vote this June on Proposition 227, which bars most bilingual instruction in public schools.

Bilingual education — the teaching of non-English-speakers in their native language — has long aroused political passions. In 1993, the state’s Little Hoover Commission condemned it as “divisive, wasteful and unproductive.” Many advocates insist just as adamantly that English-only instruction lets too many students sink rather than swim.

But after several decades of experience with such programs, scandalously little quality research exists on whether bilingual education actually works.

Two careful new studies of bilingual education by economist Mark Lopez at the University of Maryland thus make an important contribution to our limited knowledge. They offer troubling evidence that such programs may hold kids back in school and hinder their future success in the job market.

That’s not what Lopez set out to find. He was looking for ways to help Latino students improve their English-language ability, get more education and ultimately earn more money. “My hope was to find positive effects (of bilingual education),” he said. “I assumed the program must be working.”

In one paper finished last month, Lopez looked at two large national samples of students, gathered in 1980 and 1988. Some of his evidence raises real concerns:

— Among the 1988 sample of eighth-graders, students who had previously gone through bilingual programs were 3.4 percentage points more likely to drop out of school by 10th grade than similar students taught only in English.

— Among high-school sophomores in 1980, those who had been in bilingual programs were 3.6 percentage points less likely to obtain a high school diploma.

— Among those 1980 sophomores, the probability of bilingual education participants going on to earn a bachelor’s degree was 13.5 percentage points lower.

In another recent paper, Lopez and a colleague, economist Marie Mora at New Mexico State University, reported that many bilingual education students go on to earn especially low wages as adults, suggesting that educational deficiencies hold them back on the job.

Latino immigrant students who were sophomores in 1980 and who had been educated in bilingual programs made 37 percent lower earnings by 1990-91 than similar students who did not go through special programs. Latino students born in the United States to immigrant parents had 29 percent lower earnings after attending bilingual programs.

Lopez interprets his findings cautiously. He notes that the results for the 1980 sample reflect the effects of bilingual education programs in the 1970s, and may not apply today. And despite his (and Mora’s) best statistical controls, some of the negative effects of bilingual programs may really reflect the disadvantaged background of the students themselves. But even if bilingual education has no adverse effects, taxpayers should be concerned. Lopez and Mora report that bilingual education in U.S. schools costs somewhere between $2 billion and $3 billion a year, no small change. A recent review of previous research on bilingual education by Jay Greene at the University of Texas found only 11 studies that passed scientific muster. Most other studies flunked Statistics 101 — for example, by failing to control for the socioeconomic differences between students who attended bilingual programs and those who were taught only in English.

Greene concluded from the handful of good studies issued before the work of Lopez and Mora that bilingual education can work. The earlier studies showed clear gains in English and reading for students in bilingual programs, compared with students in English-only programs. Over a two- year period, the gain was equal to about three additional months of instruction.

“This has policy implications for Prop. 227, that fundamentally there’s nothing wrong with native- language instruction,” Greene said.

Greene nonetheless applauds the careful new research by Lopez and Mora. “I think their work is good and should give people pause in interpeting the long-term effects of bilingual education,” Greene said.

Greene reviewed mostly small- scale studies of a handful of classrooms, not studies based on broad samples of all bilingual education students. The research by Lopez and Mora may thus give a better picture of what happens to bilingual-program students in general.

In the real world, students often encounter not model classrooms, but a variety of poorly administered programs, sometimes taught by teachers who have little familiarity with the languages or cultures of their diverse students. Synthesizing these strands of evidence, Greene argues sensibly for more educational choice to suit the individual needs of students. Until now, almost all school programs designed for non-English speaking students in California have followed the bilingual model. Prop. 227 offers no more choice, insisting instead on English-language instruction.

“I’d prefer to get parents more involved, so they can choose from a variety of programs,” Greene said. “Some students need no native-language instruction; some may need a lot. The people who are in the best position to know are in the family.” Is anybody listening?



Comments are closed.