There's a robust body of research that has examined how the various forms of English-language instruction impact the ability of ELLs to acquire English and achieve academically, but a group of researchers is taking a completely different look at this question.
How, they ask, do bilingual education programs—in which some instruction is delivered in an ELL's native language—spill over to impact the performance of students who are not learning English?
The researchers use Texas—home to the second largest concentration of ELLs in the nation and where Spanish bilingual education programs still exist in public schools across the state—to try and answer this question. Their findings were released in a working paper today from the National Bureau of Economic Research.
What they found is sure to be the equivalent of a "Hey Martha," story, the term my old city editor at The Sacramento Bee used for those surprising or provocative front page stories that would get everyone talking.
This is a rather technical study for an academic layperson like me, but the findings are clear and fascinating. Essentially, the researchers compared student outcomes in school districts that offered bilingual education to ELLs with those for districts that use ESL for English-learners. They drew on publicly-available data from Texas elementary schools that enrolled slightly fewer or slightly more than 20 ELL students in a given first grade cohort.
Here's the upshot:
In districts offering bilingual education, the test scores of non-ELLs whose home language is not Spanish were raised "significantly." Because these kids were not Spanish speakers, they never would have been candidates to participate in bilingual education, therefore, researchers conclude, there were "program spillover effects." How, you ask?
The researchers offer some possible explanations, many of which can be boiled down to differences between the composition of classrooms in bilingual programs versus those in ESL programs.
When ELLs are in bilingual programs, they are typically grouped together in a separate class and tend not to be in mainstream classes with non-ELLs(ELLs in ESL programs are more often in mainstream classes and are pulled out for English-language instruction). With fewer or no ELLs in their classes, non-ELLs perhaps benefit from teachers not having to focus extra help and attention on English-learners and from not having as much exposure to ELL students who may need lower levels of instruction in order to understand content.
On the other hand, the researchers point out, there are positive peer effects for non-ELLs who are exposed to English-learners. Non-ELLs can also benefit from bilingual programs because of the additional resources that such programs tend to bring to a school, the researchers said.
The study also found that bilingual education had "generally positive," but less significant effects on the achievement of Spanish-speaking ELLs than for their non-ELL peers.
Aimee Chin from the University of Houston, N. Meltem Daysal from Tilburg University in the Netherlands, and Scott A. Imberman of Michigan State University are the authors.