Ethnic Educational Inequality: The Role of Neighbourhood Contexts
Theoretical background and objectives
Comparative research on ethnic educational inequalities has repeatedly shown that Belgium is one of the countries with the most severe ethnic penalties in education (Marks, 2005; OECD, 2006, 2007). This project tests contextual explanations for educational disadvantages among the second generation in Belgium, since common explanations from a comparative stratification perspective in terms of social background or family resources could not fully explain lower attainment levels among the Turkish and Moroccan second generation in Belgium (Phalet, Deboosere and Bastiaenssen). The first study looks at the role of ethnic composition in the municipality as a proxy for ethnic capital and compares different concepts of ethnic composition both theoretically and empirically. The second study asks to what extent the positive ethnic density effect found in the first study is conditional on neighbourhood structure, arguing that dense social networks with co-ethnics are more easily developed and maintained and yield more returns for educational success in more stable, higher quality neighbourhoods.
Research design, data and methodology
Multilevel analysis is applied to individual data from the 1991 Belgian Census and contextual data of all Belgian municipalities in order to examine contextual influences on ethnic educational inequality.
The first study shows that the percentage of ethnic minorities goes together with lower school completion rates among the Belgian majority population, but the effects are inconsistent among the three largest ethnic minority groups in Belgium, the Italian, Moroccan and Turkish second generation. In contrast, ethnic density, i.e., the presence of co-ethnics in the municipality, is positively associated with school completion in minority as well as majority groups. The second study reveals that ethnic density and neighbourhood quality overlap strongly among majority Belgians and the Moroccan second generation, albeit with opposite sign, such that the absence of ethnic minorities goes together with more favourable neighbourhood structure, while concentrations of Moroccans occur in the least attractive neighbourhoods. Among the Turkish and Italian second generation, we find, as expected, that the effects of ethnic density depend on neighbourhood stability and quality. These results confirm differential trajectories of integration of the three largest ethnic minorities in Belgium.