Both ethnic and social inequalities are being almost universally found throughout western European school systems. Using administrative microdata on the entire population of students in public schools in Denmark, we zoom in on such inequalities in a specific evaluation situation where teacher evaluate the preparedness for upper secondary educational tracks after the first nine or ten years of initial schooling. As a specificity of the Danish system, there are two such evaluations: one in grade 8, which can be seen as low stake and mainly of pedagogical nature since it is preliminary, and one in grade 9, which is definitely a high stakes evaluation since it is the final disposition and a positive evaluation is required to enter different upper secondary tracks (both, vocational and general upper secondary education).

Descriptive statistics from our data show an enormous gap between native Danish students and different minorities. To exemplify, while about three quarters of Danish students without migrant background are being evaluated as "qualified" already in grade 8, this applies to only roughly half of all students with Turkish, Middle Eastern or Subsaharan African background.

There are different mechanisms that can be expected to explain these differences. The most critical one would be discrimination against these groups, which is a possibility given the experimental results that have been found on this question (see e.g. Wenz & Hoenig, 2020). Another possible mechanism is the lower performance of ethnic minority students throughout the school career which could then mirror in the evaluation, since the evaluation in question is based on both scholastic performance as well as on the personal and social prepardness for upper secondary education. This, again, might be traced back to different reasons which we will not be able to investigate in our study. It could e.g. be discrimination in earlier years in the school system or language barriers. Finally, a third mechanism lies in the strong relationship between social origins and migration background. To a large degree, parents of children with migrant background are working in routine jobs in the low skilled sector of the labor market, which is known to be related with school related disadvantages of children.

We can study implications for these three mechanisms in our data. This helps explaining the disadvantages of typical labour migrant and asylumn seeker groups in western Europe in a particular evaluation system instead of broadly in summary measures of educational success. The evaluation be further differentiated into a high-stake and a low-stake evaluation; and there are separate evaluations for the vocational and the general academic upper secondary track (gymnasium). With a large scale dataset at hand, one of our contributions is also that we can hold constant for both scholastic performance and parental social status in a particularly fine-grained manner. 

We use microdata from administrative registers linked with data on scholastic perfomance (nationwide school tests, final grades in high school) and the evaluation regarding preparedness for the different upper secondary track. Parental live courses can be linked via unique identifiers. We keep both scholastic performance and parental social status (measured using the European Socio-Economic Classification ESeC) constant applying coarsened exact matching. This rather rigid form of control is possible due to the large datasets in population wide register data. Regarding migration background, we analyze solely children who were born in Denmark or migrated before turning five.

Our preliminary results show enormous gross gaps between students without a migrant background and most ethnic minority students, especially those from the largest refugee and labour migration groups. Gaps for migrants from European and East Asian countries are clearly smaller or even positive as in the case of migrants from Western Europe. Controlling for scholastic performance does not close the gap, which implies that migrant children on average - at the same performance in standardized school tests - have a lower chance of getting being evaluated prepared for upper secondary education. However, once taking the social status of the parents constant, these gaps become fairly small or even disappear. This shows that the manifest disadvantage of minority is related to the lower social status position of parents. At the same time, it contradicts the discrimination hypothesis - at least for discrimination against ethnic minorities in the specific evaluation.

Wenz, S. E., & Hoenig, K. (2020). Ethnic and Social Class Discrimination in Education: Experimental Evidence from Germany. Research in Social Stratification and Mobility, 65. https://doi.org/10.1016/j.rssm.2019.100461