Investigating social frontiers in the Netherlands: a data analytics approach

Introduction: The impact and presence of segregation has long been a topic of
interest in social science (Massey and Duncan, 1985). Yet transitions between
communities are vastly underrepresented (Legewie and Schaeffer, 2016; Dean et
al., 2019), with segregation measures often bypassing the steep transitions. Dean et
al. (2019) coined the term ‘social frontiers’ to describe these steep boundaries,
highlighting that they could have large consequences on society through increasing
crime and conflict, impacting household mobility (Olner et al., 2024), reducing access
to resources, and heightening feelings of relative deprivation (Iyer and Pryce, 2024).

Objectives: This thesis aims to address some of the gaps in the relatively new
social frontier literature, making large contributions to the conceptual,
methodological, and empirical findings. The thesis especially focuses on the
frontiers’ prevalence, evolution, and impact in the Netherlands.

Methodology: This thesis implements a range of techniques. I apply four different
identification methods including the established Lee and Mitchell (2013) method.
Additionally, the thesis uses multiple measures to describe the prevalence of social
frontiers and discuss their evolution. Finally, applying causal matching methods to
evaluate the impact of living close to social frontiers.

Results: The thesis has several key findings which form an important contribution to
the literature: the establishment of the Partitioning Around Medoid (PAM) clustering
method for identifying frontiers; growing rates of social frontiers in Amsterdam and
the Hague; the contrasting prevalence of single dimension and intersectional social
frontiers; and the negative association between living close to social frontiers and
socioeconomic outcomes.

Conclusion: The thesis makes significant methodological, empirical, and theoretical
contributions to the literature. It highlights the importance of social frontiers and the
need for further research into them. The findings of all the chapters have the
potential to drive future research, as well as having large policy implications for the
Netherlands.