| During the last two decades there has been a rapid rise in the number of students identified as limited English proficient (LEP) in the U.S. school system. Schools have responded to this increase by instituting a variety of bilingual education programs with the primary objective of improving the English language skills of students whose native tongue is not English and whose level of English proficiency is low. Despite widespread agreement on the goals of improving English proficiency among LEP students, there is considerable disagreement over whether bilingual education programs help achieve that goal, or whether they actually slow down the acquisition of English language skills specifically, and educational progress more generally.;This dissertation investigates to what extent participation in a bilingual education program affects educational attainment in the short and long runs, and ultimately labor market performance. If language deficiencies inhibit educational progress, then one would expect LEP students who have received no English language assistance to be held back a grade at higher rates, to drop out at higher rates, and in the long run, to earn lower wages and obtain fewer years of education than native English speakers.;I address this relationship by first delineating a control group to serve as a counter-factual for the program participants. Together, the control and treatment groups constitute an "at risk" group. Within this "at risk" sample, students are non-randomly assigned to participate in the program. Using three similar strategies to address the problem of selection, I find that overall, exposure to bilingual education, exclusive of the timing of exposure, produces statistically insignificant to small but negative and statistically significant effects on educational achievement. However, this result masks the importance of timing of a student's first exposure to bilingual education. The National Education Longitudinal Study of 1988 data suggests that earlier exposure produces no difference in outcome relative to a control group, while in the High School and Beyond data, earlier exposure leads to lower levels of educational attainment but no difference in labor market performance relative to a control group. |
