Quantifying segregation on a small scale: how and where locality determines student compositions and outcomes taking Hamburg, Germany, as an example

ABSTRACT

Increased social and academic segregation are known side effects of school choice policies in market-driven environments that facilitate competition amongst schools. Aiming at complementing foundational knowledge in quantifying segregation, this study first defines school markets (i.e., geographical context) based on student transitions from primary school to secondary school in Hamburg, Germany. Second, genuine spatial measures of segregation are applied to generate differentiated in-situ insights. In general, social segregation appears evident between school markets, school types, and individual schools and, thus, shapes social compositions of secondary schools. The pattern of student transfers across the city confirms that parents are selecting particular schools for their children, resulting in different schools servicing different composition of students and so markets. Furthermore, the findings suggest that school markets in both very affluent and very deprived areas are spatially isolated and hence persistently reproduce wealth and affluence as well as poverty and disadvantage.

KEYWORDS:

• Market models
• school choice
• segregation
• socioeconomic status
• student transitions

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes on contributors

Sebastian Leist is a PhD candidate in the field of Educational Research. He worked for 8 years with the Institute for Educational Monitoring and Quality Development (IfBQ) in Hamburg, Germany, and afterwards for more than 2 years with the Department of Education and Training (DET) in Victoria, Australia. Currently, he works as a Senior Analytics Consultant with WorkSafe Victoria, Australia. His research covers educational geographies, in particular modelling geographical context factors of schools, school segregation in market-driven environments, equity, student admission, and related organizational behaviours. Research interests are based on exploring the interplay between these factors and how they affect school effectiveness, school improvement, and student outcomes.

Laura Perry is Associate Professor of education policy and comparative education. She conducts research about educational disadvantage and inequalities, especially as they appear between schools, and the systems, structures, and policies that shape them. Specific research interests include educational marketization, school segregation, school funding, and school stratification.

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Notes

1 Network data that include objects and measurements between objects: “data whose properties cannot be reduced to the attribute of the individuals (nodes) involved. The relation, or ‘tie’, is the object of (and unit of) analysis” (Handcock, Hunter, Butts, Goodreau, & Morris, 2008, p. 1).

2 If a student has not received admission to the preferred school, the same process applies to the second best option and, subsequently, to the third best option that were indicated by the time the family of the student has submitted their preferred options of schooling.

3 In the following sections, special schools are considered in the data model but are not further described due to their specific clientele and therefore minor role in school choice processes in Hamburg.

4 We use inverted raw scores for ease of interpretation of related content in the paper. Otherwise, negative raw scores would have meant a high level of advantage and vice versa.

5 Data of the Programme Social Monitoring Framework for Integrated Urban District Development are publicly accessible (Behörde für Stadtentwicklung und Wohnen, 2015).

6 Effect size interpretation as initially suggested by Cohen (1988), expanded by Sawilowsky (2009): 0.01 < d < 0.20 very small, 0.20 < d < 0.50 small, 0.50 < d < 0.80 medium, 0.80 < d < 1.20 large, 1.20 < d < 2.00 very large, d > 2.00 huge. For example, an effect size of 0.52 means that the average social status of a student on School Market A is 0.52 standard deviations below the average social status of a student leaving this school market. Thus, the average student retained has a lower social status than 69% of those students leaving the market.

7 Eighty-five students enter this school market coming from Walddoerfer with 40 of these attending comprehensive schools.

8 The weight is equal to the proportion of school i’s intake that is drawn from the primary schools shared with school j, multiplied by the proportion of school type j’s intake that is drawn from the same. The weights are then scaled (row-standardized) so that the sum of the weights for any school is equal to 1 (Harris, 2011a).

9 Discussed differences appear substantial, given Lücken et al. (2014) state that 30 points or more difference at the group level (class size or greater) is interpreted as being meaningful.