We show that the German early tracking system reinforces existing socio-economic disadvantages of children through much stronger social multiplier effects. We estimate social multipliers in math and German language for 9th-grade students through a Conditional Quasi-Maximum Likelihood approach: we identify the average effect of classmates' performance from the reduced form using classroom size variation and treat a student's potential (self-)selection in a classroom as an omitted variables problem. We find that, first, a 1 point decrease in peer average performance in math (German language) leads to a 5.6 (4.4) points decrease in classroom performance for vocational training (Hauptschule) students but only to a 2.2 (2.5) points decrease for the university-path (Gymnasium) students; second, although the native-to-native and immigrant-to-immigrant average peer influences are similar with native-to-immigrant and immigrant-to-native influences in the Hauptschule, the respective same-type effects are larger than the cross-type effects for the Gymnasium; and, third, the average effect of natives is always larger than the average effect of immigrant students on the whole classroom. We conclude that a native student's performance is more important - either in a beneficial or detrimental way - than an immigrant's performance especially for the Gymnasium classrooms.



Subscribe to SOFI-AME seminars using Google calendar: 


Subscribe to the SOFI-AME mailing list:


Follow the SOFI-AME seminar updates on Twitter: @SOFI_AME