Social Mix and the Neighbourhood Effect. Policy Ambitions and Empirical Evidence

Abstract

Segregation is a central concept in both academic and policy debates on urban issues. It has been argued that the process of globalisation results in increased social polarisation and subsequently sharper spatial segregation. Indeed, many politicians express a fear of rising segregation, envisioning the emergence of 'ghettos' or as it is called in the Netherlands 'income neighbourhoods'. In order to prevent concentrations of poverty from forming, a new area-based policy was formulated which aimed to restructure the urban housing market at the neighbourhood level and mix low-quality with high-quality houses. Such a concern with social mix has become common in a number of developed countries. In this regard the analysis has a wide relevance. This paper explores these ideas both by discussing the theoretical framework underpinning the policy, and by examining empirical support for it. Since the policy of housing-quality mixing is still in the first phase of implementation, relevant longitudinal data is not yet available. As a consequence our evaluation addresses present poverty concentrations and housing stock (mix) characteristics in the city of Amsterdam. By comparing neighbourhoods that already have a 'mixed' housing stock to homogeneous neighbourhoods, it has been possible to see whether mixing really does correspond to significantly lower poverty rates. It turns out that the empirical facts are quite different from the expected results: mixing does not in fact reduce poverty. It is concluded that the policy lacks an empirical basis. Housing-mix policy requires substantial budgets, while the goal of reducing poverty cannot be reached. As an alternative, we suggest that poverty is a personal characteristic and that it is therefore preferable to approach poverty directly instead of hoping for the results of a dubious 'neighbourhood effect'.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 177 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.