Paper long abstract:

We examine how the transition to English-only instruction affects college students' academic outcomes in a non-English speaking country. Making use of a natural experiment at a selective university in Central Asia, we utilize a difference-in-differences strategy to estimate the causal effect of switching to English-only instruction on college outcomes. We find that the introduction of English-only instruction led to a decrease of GPAs and probability of graduation and an increase in the number of failed course credits. The effects were similar across gender, but heterogeneous across type of locality students come from and native language. Although negative, the effects were temporary. The findings suggest that at least in selective universities in non-English speaking countries, the switch to English-only instruction negatively affects college outcomes at the time of transition but does not necessarily imply longer-run negative effects.