Locked-in vs. Locked-out: Can Detracked Classes Increase Education Equality?

Do detracked classes affect students from different socio-economic backgrounds differently? In the Swiss education system, students are assigned to one of two tracks based on prior achievements at age twelve: approximately 70% are placed in an advanced track and roughly 30% in a basic track. After this assignment, students may either be grouped into classes based on their track or placed in mixed classes with students from both tracks. While tracking is common in many countries, the evidence on its impact remains inconclusive. Understanding this impact is crucial for optimizing school systems to improve students' labor market outcomes later in life. To evaluate the effect of detracked classes, I exploit a unique detracking reform in one Swiss canton, using a difference-in-differences design. This reform, implemented in 2015, changed only how students were grouped into classes, while track assignments remained the same. Before 2015, classes were tracked, meaning they contained only students from either the advanced or basic track. After the reform, classes were detracked, meaning students from both tracks were placed together, while tracks were still assigned. Using individual-level register data for the entire population of Swiss students from 2012 to 2022, I show that the reform dramatically altered class compositions in terms of peers' background characteristics. Since track assignment is correlated with socio-economic background, advanced track students, on average, had for example fewer native speakers in their classes after the reform, and vice versa for basic track students. The likelihood of being assigned to further education, which enables students to pursue tertiary education, increased for the average student due to detracking. My heterogeneity analysis reveals that the overall positive effects were concentrated among socio-economically disadvantaged students. For students whose parents are not tertiary educated and who are not native in the regional language, the probability of further academic education nearly doubled, while more advantaged students did not experience any negative effects. I can rule out changes in curricula, teacher quantity and quality, and motivational factors as mechanisms for these findings, and interpret my main estimates as the causal effects of detracked classes.