1 Introduction

Welcome to the website of the Summer-School project BIGSSS - Segregation and Polarization!

The literature on how the ethnic and socio-economic composition of our living environments impacts social cohesion between and within groups is vast. With larger outgroup sizes, competition over scarce economic resources would become more intense and this would subsequently lead to hostility between groups. Increased diversity of our living environment would reduce the predictability of the behaviour and opinions of our fellow neighbourhood residents and hence would fuel feelings of anomie and ultimately cause the deterioration of generalized trust. Segregation, along ethnic, economic or other key social dimensions, would not only hamper inter-group contact but would, at the same time, also increase the visibility of outgroups, and thereby feelings of cultural/identity threat. But notwithstanding the long-lasting scholarly attention to this research area, up till now, it has not been possible to distill a ‘social law’ from these studies on how the composition - and other characteristics of our living environments - impacts social cohesion.

The aim of this project is to discover this law and we will focus on one of the most important indicator of social cohesion, namely political polarization.

Challenge 1. Describing our living environment

One of the challenges the field faces is that the composition of our living environments can be described in a myriad of ways. Is it group sizes that matter, levels of diversity, segregation or a combination thereof? And at which geographic scale, which social dimensions and categories to use, and which measurement instrument would best fit our theoretical needs?

Challenge 2. Polarization as indicator of social cohesion
Concurrently, there are many indicators of social cohesion. While the literature has recently moved away from purely individual-level indicators of social cohesion (e.g., undertaking voluntary work) and has recognized that social cohesion is by definition a relational concept (e.g., trust of a specific ego in a specific alter/target), studies on the consequences for macro-level indicators of cohesion have been relatively scarce.

A theoretically informed data-driven exploration
To move the field forward, this project will take a (theoretically informed) data-driven approach. First, we will discuss how to define, identify and visualize different patterns of segregation. We will pay special attention to issues of scale – and by extension to spatial measures of segregation –, acknowledging that both patterns of segregation and consequences of segregation may be scale dependent. We will use techniques of spatial analysis and social network analysis and take full advantage of the available fine-grained spatial data of Statistics Netherlands. Second, we will investigate how these (many different) identified micro- and macro-segregation patterns are related to a true macro-level indicator of social cohesion, namely polarized voting behavior. Our data source for the latter will be election results at the polling station level. The final step in this project will be to assess whether the ‘discovered’ patterns of segregation that are related to macro-level polarized voting behavior also predict political attitudes and voting behavior at the individual-level. The aim is to use the third wave of the NEtherlands Lifecouse Study wave 3, to be collected early summer in 2022 among approximately 5000 native Dutch and 5000 Moroccan-Dutch and Turkish-Dutch citizens. To improve our theoretical and methodological understanding of neighbourhood and spatial effects, we will compare results from traditional multi-level models with spatial regression models.

2 Getting started

During the course, students will journal their work and assignments in their custom lab journal. A template lab journal can be found on GitHub. Here, you find how to get started.

3 Intranet, google drive and ZOOM

If we want to share files, we will use this google drive

You can find the BIGSSS intranet page here. See mail send 1-7-2022 for password.

And this is the place to be in ZOOM

4 Program

You can find the summer school program here.

The program of this project will be as follows:

4.1 Week 1

On Monday, we will discuss definitions of and functions to measure polarization.
On Tuesday, we will discuss approaches to determining party positions.
On Wednesday, we discuss and calculate the level of polarization at the polling station level, and the spatial distribution of polarization.
On Thursday, we discuss definitions of and functions to measure residential segregation.
On Friday, we work with the spatial data of CBS.

4.2 Week 2 (preliminary)

On Monday, we estimate one simple OLS model with one segregation measure and one polarization measure.
On Tuesday, students choose their own measures; we also compare OLS with spatial regression models.
On Wednesday, we add complexity with respect to neighbourhoods; and we combine multi-level modeling (i.e., units nested in polling stations, and covariates at both levels) with spatial regression.
On Thursday, students present their final model, and we try to come to a final ‘representative model’.
On Friday, we estimate a final model using the NELLS data on individual-level voting behavior.

Political polarization and Residential Segregation

AUTHORS:

Jochem Tolsma - Radboud University / University of Groningen, the Netherlands

Thomas Feliciani

Rob Franken