English Learner Home Language Concentration and English Acquisition : Theory and Evidence from California

We consider the question of whether home language concentration among English learners, i.e. segregating English learners by home languages across schools, helps or hinders the acquisition of English. The sign of this effect is theoretically unclear, since there are at least two competing forces: first, economies of scale in language instruction, whereby the education system may more cheaply and effectively teach English to students from a more dominant non-English language, and second, network effects, whereby students who speak the same non-English language may coordinate on speaking that language, to the detriment of learning English. We present a simple game theory model borrowed from the network goods literature to illustrate these competing forces. We then use publicly available data from California on English learners and their instructors to probe the evidence for the theory. Our analysis provides some evidence for the economies of scale theory, since holding other important factors constant, schools with a greater language concentration are significantly more likely to employ bilingual instructors, during a policy regime when such a choice was not state-mandated. On the other hand, Spanish-to-English learners, the most concentrated and widespread language subgroup, acquire English significantly more slowly than most other common language subgroups, suggesting that the effects of concentration may be negative on net