Voter Mandates and Bilingual Education

[Harold F. Schiffman]

New York Times, October 23, 2002

Voter Mandates and Bilingual Education

By RICHARD ROTHSTEIN

    CARROLLTON, Tex.

IN November, voters in Colorado and Massachusetts will decide on
initiatives that would ban bilingual education. Arizona and California
have already adopted similar measures. Proponents, who want all
instruction in English, rely on a claim that early-20th-century immigrants
succeeded by that method, called English immersion. But the claim is
largely false. A century ago, dropout and failure rates were much higher
among the many immigrants from illiterate backgrounds than they are today.

Secretary of Education Rod Paige opposes the proposals, saying decisions
about the proportion of English and a child's native language should be
made at the "point of instruction." That is the approach here in the
affluent Dallas suburb of Carrollton, which is mostly non-Hispanic but has
a growing population of Hispanic immigrants. Decisions about how much
English a child should have at any time are made by teachers, with
parents' consent. The factors they weigh show why a flat ban on bilingual
education is unlikely to improve immigrants' chances of success.

Carrollton's providing some instruction in Spanish is intended to help
students reach grade level as rapidly as possible. Children learn English
for part of the day but study other subjects in Spanish. Then, when they
are ready to join regular classes, they don't start off behind in math,
science, social studies and literary skills. Even English-speaking
children will often do poorly if their parents had a poor education and
are unfamiliar with academic culture. When those handicaps are compounded
by trying to learn in a language that the student does not comprehend,
success is even less likely.

So when children from Spanish-speaking homes in Carrollton enter school,
teachers assess their fluency in both English and Spanish. Those who are
stronger in Spanish are put in bilingual classes where math, science and
social studies are taught for half the day in Spanish, with the other half
in English. Some immigrants may be stronger in English, though far behind
their peers in both languages, if their home literacy in Spanish is poor.
This gives teachers little on which to build in Spanish, so such children
get classes taught mainly in English.

Most language experts say it usually takes Spanish-speaking children five
to seven years of bilingual instruction to be ready for mainstream English
classes. But Carrollton administrators usually move children to regular
classes after three years, provided they pass the Texas minimum-skills
test in English. Annette Griffin, superintendent of schools, says her
staff balances several factors in deciding how much Spanish and English
each child should have. Bilingual teachers are in short supply, so the
district concentrates them where they are most needed, in the early
grades. After three years in bilingual classes, many children have enough
English fluency that regular teachers can give whatever extra help they
need.

Although such children are not as English-fluent after three years of
bilingual education as most American-born peers, they may benefit from the
influence of English-speaking classmates. The social value of an English
environment has to be weighed against the instructional value of more
Spanish teaching. (This consideration is a luxury that districts can't
indulge if they have few nonimmigrant peers with whom the immigrants can
integrate.) Support of "point of instruction" decisions on the teaching of
immigrants is a rare case where the Bush administration wants to defer to
teachers and professional educators. Perhaps the reason is that Texas, the
president's home state, has a policy of teaching in native languages and
introducing English gradually. That bilingual approach has proved more
successful than English-only.

Some districts have had success by using even more Spanish than
Carrollton. A study of the schools in Houston, where Mr. Paige was
previously superintendent, found that when Hispanic immigrant children had
more instruction in Spanish, their English scores were higher than those
of other immigrants and they were less likely to drop out. Houston usually
keeps children in bilingual education longer than the three years
Carrollton aims for. In fact, the Texas Education Agency recently told
Carrollton schools to increase their emphasis on Spanish.

Perhaps Carrollton's effort to speed the transition to English has been
unwise, and academic gains from more Spanish instruction outweigh social
gains from integrating native and non-native English speakers. Perhaps
scarce bilingual teachers should be spread among more grade levels.
Perhaps regular teachers give poor support to English learners in busy
classrooms. Or perhaps future research will show that immersing immigrants
in English-speaking classes has benefits that have yet been undetected.

But one thing is certain: The worst way to resolve these issues is by
voter mandates that prevent the decisions from being made at the "point of
instruction."

E-mail: rrothstein at nytimes.com