SHOULD WE BORDER? A WITHIN-COUNTRY ANALYSIS OF IMMIGRANT PEER EFFECTS AND HOW THEY ARE AFFECTED BY INSTITUTIONAL ARRANGEMENTS

Abstract

This thesis studies how exposure to immigrant peers affects the academic achievement of native and immigrant students, and how these effects are affected by institutional arrangements. Data is drawn from TIMSS 1995, an international study providing information on the student achievement and immigrant status of students across cohorts. The main empirical approach utilises a within-country fixed effects analysis, where fluctuations in the proportion of immigrant students over adjacent cohorts are used to measure the impact of exposure to immigrant peers. A randomisation inference test is used as the main robustness check to substantiate the reliability of the findings. This study finds that although immigrant peers do not affect natives significantly, there are negative and statistically significant impacts on immigrant students. Moreover, immigrant peer effects are also influenced by educational institutions, with the placement of students into academic tracks exacerbating these effects the most.