What Works for the Children? What We Know and Don't Know About Bilingual Education
September, 2002
By The Civil Rights Project
Executive Summary by Jacinta Ma
English Language Learners (ELLs), also known as limited English proficient (LEP) students face significant barriers to obtaining quality education. While research findings have not yet indicated how best to address the issues ELLs face, schools must provide some type of support to help them overcome these barriers. Current policy debates about how to improve education for ELLs have turned into battles over whether to implement one-year English immersion programs or bilingual education programs. In 1998, California voters passed Proposition 227, the so-called "Unz Initiative," which had the goal of replacing bilingual education programs with one-year English immersion programs. Unz Initiative proponents have heralded their approach and Proposition 227 as a success, proven by dramatic increases in ELL test scores. This paper reviews existing analyses of California test score data on which these claims are based and finds no serious evidence that the Unz Initiative resulted in gains for ELLs.
Generally, important research studies conclude that bilingual education programs are helpful in increasing children's academic achievement. Education research generally indicates that no one has identified any single program or approach that is a perfect model, or even simply the most effective for all ELLs. Several studies find that it takes ELLs a number of years - close to five or more - before they can develop language skills and academic knowledge equal to that of a native English speaker of the same age and grade. There is no basis in sound education research for supporting arbitrary one-year limits on language support for ELLs. However, the data regarding bilingual education and the most effective strategies for educating ELLs are limited. Because much is still not known, at a minimum, policymakers should allow parents and educators to select from a variety of sound language support programs that are likely to work for the children in their schools.
Analyses of ELL test scores in California are inconclusive and suggest that achievement gaps may be widening. While Ron Unz and opponents of bilingual education have claimed that ELLs made dramatic increases on Stanford-9 test scores, some researchers have found that students in bilingual education programs did better than those in English immersion programs. Other researchers have found that bilingual education programs were no worse than English immersion programs. One study found that the achievement gap between ELLs and English proficient students is growing. Moreover, most researchers find that using Stanford-9 test scores to make conclusions about increases in language proficiency are inappropriate because the test was designed to measure differences in academic achievement of native speakers of English.
This paper also examines other effects of Proposition 227. It looks at the change in the rates ELLs were "redesignated" as English proficient and ready to enter general education classes. During the campaign to pass Proposition 227 in California, Ron Unz suggested that after the implementation of one-year English immersion programs many, if not most, ELLs would know enough English to enter general education classes. The paper concludes, however, that a 3% increase in the redesignation rate does not support claims of success. It also looks at Proposition 227's effect on teachers and finds that a number of teachers have been demoralized by it and find that it has had a negative affect on the ELLs they teach.
Federal laws and constitutional provisions require that school districts provide ELLs with equal educational opportunities to learn and participate. Mandating one-year English immersion programs for all students would likely violate ELLs' rights under these laws.
While there are limitations to using existing standardized tests to measure ELL progress, evaluative testing of ELLs is imperative. Appropriate tests need to be developed to determine how ELLs are progressing in learning English and other subjects. The decision of whether or not to assess a child in English or to move a child to general education classes should depend on whether a child has reached certain levels of knowledge rather than on whether the child has participated in a certain number of years of instruction.
These findings suggest that more effort should be spent on improving the education that ELLs receive than on debating the best type of language support program to implement in all of our schools.
The paper makes the following policy recommendations:
- Make language support programs one part of an overall strategy for improving the quality of education for ELL children.
- Give parents and schools the flexibility to implement different types of language support programs.
- Clarify the goals of language support programs.
- Support the use of academic or literacy measures to determine when children should be moved from bilingual education or other language support programs to general education classes rather than through the use of time limits.
- Hold schools and school districts accountable for ensuring improved academic achievement of ELLs measured by appropriate knowledge of English, consistent academic achievement over time, improvement in diagnostic test scores, and improvements in graduation rates.
- Provide additional high quality instruction for ELLS.
- Provide additional funding to improve the quality of education for ELLs.
- Support additional research to evaluate language support programs, to develop appropriate tests to measure ELL student achievement, and to evaluate effective testing accommodations for ELLs.
- Use sound education research to influence policymakers through advocacy and litigation.
In compliance with the UC Open Access Policy, this report has been made available on eScholarship:
