The Influence of the Educational Composition of the Neighborhood on the Transition from Primary to Secondary School
Bearbeitung: Natalie Backes
Projektbeschreibung:
Sociological research has identified numerous factors contributing to educational inequality, including parental financial, social, and cultural resources. Recently, increasing attention has been directed towards the role of neighborhood socio-economic composition in shaping children’s educational outcomes. This master’s thesis examines how educational neighborhood composition in Germany influences transitions from primary school to the academic track (‘Gymnasium’) and whether these effects vary by social background.
This thesis makes several contributions. First, unlike prior studies using broad neighborhood definitions such as 1km grid cells or post code levels, I utilize fine-grained data aggregating characteristics only of nearby households, better reflecting the importance of close neighbors. Second, additionally to adult neighbor aggregates, I explore the influence of adolescents already attending secondary school in the neighborhood, as they could serve as important role models, an overlooked factor in existing research. Third, while neighborhood studies on Germany’s stratified education system has primarily focused on major cities like Berlin and Bremen, this research uses nationally representative data, improving the generalizability of findings. It also enhances the control of confounding household and neighborhood characteristics and provides separate effect estimates for different social backgrounds, delivering broader and more nuanced insights. Fourth, while the thesis tests whether neighborhood effects follow linear or non-linear patterns by employing models testing both.
Using the 2012–2015 and 2016–2019 German Microcensus Panels, a nationally representative dataset, this thesis measures neighborhood educational composition through the share of high school graduates in the neighborhood, the share of neighbors with lowest or no school certificates, and the share of adolescents in academic tracks among approximately nine nearest neighboring households.
The methodological approach incrementally improves causal interpretability through four steps: first, cross-sectional linear regression models with both the treatment and outcome variables measured in fifth grade; second, models including treatment variables measured in fourth grade to ensure temporal precedence; third, adding controls for third-grade neighborhood composition as a proxy for neighborhood desirability, thereby addressing unobserved characteristics influencing selection into certain neighborhoods; and fourth, sibling fixed-effects models to control for time-constant confounders at the neighborhood and household-level. Each step examines both linear and non-linear neighborhood effects.
Findings indicate that neighborhood educational composition significantly influences Gymnasium transitions, with disadvantaged children experiencing stronger effects. Higher shares of highly educated neighbors positively affect Gymnasium transitions, with significant effects starting at 23.5% and becoming pronounced above 40%. In contrast, neighborhoods with over 50% low-educated residents strongly reduce Gymnasium transitions, especially for disadvantaged children. Advantaged children benefit from "big fish little pond" effects in low-educated neighborhoods but face negative effects when low-education shares exceed 50%. Surprisingly, disadvantaged children benefit more from neighborhoods with non-Gymnasium students than from those without any secondary school students, highlighting the role of peer dynamics. Negative neighborhood effects are more pronounced than positive ones, and the stepwise modeling approach reveals that earlier studies likely underestimated these effects due to insufficient control for selection bias. These findings emphasize the importance of small-scale neighborhood effects in shaping educational trajectories and provide new insights into the interplay between neighborhood composition, social background, and educational inequality.