EUKN Urban Review: Socio-Economic Segregation in EU Capital Cities

Machteld van Dijk

Part 1 of EUKN’s monthly interview with Mr. Tiit Tammaru, Professor of Urban and Population Geography and Head of the Centre for Migration and Urban Research at the Department of Geography of the University of Tartu, about his co-written book “Socio-Economic Segregation in European Capital Cities.”



1. You have investigated comparable data (2001 – 2011) of 13 European cities and made predictions about segregation levels in each of these cities. Poor and rich are living increasingly far apart, because people search a place to live that fits their own socio-economic position. Although the socio-economic segregation is relatively modest, the poor are increasingly segregated within the urban communities. What is the main reason? 

This is true that the main finding of our study reveals growing levels of segregation between top and bottom income groups across Europe on the one hand, but the levels of segregation are still modest in most of the 13 cities studies on the other hand. The term ‘segregation’ refers to the physical separation of two or more groups into different urban neighbourhoods, and this was the first study in Europe that explicitly compare levels of socio-economic segregation. There are many ways to measure segregation, but the best-known measure is the ‘Dissimilarity Index’ that indicates relative separation or integration of socio-economic groups across all neighbourhoods of a city. The dissimilarity index between the top and bottom income groups is mostly ranging between 20 and 40 in the studied cities. The value of the index indicates what is the proportion of one or another group (i.e. members of the top or bottom income groups) in a city would need to move to another neighbourhood to make them evenly distributed across all neighbourhoods. The values between 20 and 40 can be interpreted as modest. We further identify that in Madrid, Milan, London and Tallinn, dissimilarity index is more than 40 or more than 40% of the members of either the top or bottom income group need to change home across neighbourhoods to achieve their even distribution. The values of the index above 40 signal already high levels of socio-economic segregation. Still, all studied European capital cities are less socio-economically segregated than any largest city in the US.

 We start out by outlining four important structural factors that are generating segregation: globalization; level on inequality; type of the welfare regime; housing regime; and occupational structure. These factors are related to each other on the one hand, but do play an independent role as well. Based on the scores calculated these factors, we started out our study with a hypothetical ranking of the 13 cities by segregation. Our empirical findings did not fully support the hypothetical ranking. We conclude that the important ingredients of high segregation include high level of inequality or growing inequality; the overlap between socio-economic and ethnic inequalities; the role of markets in the housing sector and the availability of neighbourhoods with homogenous housing in the city. There is always a time-lag between growing social inequalities and spatial inequalities (segregation). The stronger the role of markets, the quicker is the sorting of people with different incomes into different neighbourhoods. We further find that often these are the new residential areas that tend to be more income-homogenous than old residential areas. This is so since higher-income groups are able to take the advantages of the new opportunities on the housing stock. Thus, when the public sector reduces social housing (this is called the residualisation of social housing) and it remains a common housing for the lowest income groups, and when the new housing construction booms and draws the highest income groups, the city becomes increasingly segregated. The  good examples of cities on housing considerable boom in the 2000s and quickly rising segregation include Madrid and Tallinn.


2. Together with your colleagues, you explained levels of segregation by an in-depth analysis of the unique local situation and policy in each city. Madrid and Milan turn out to be the most segregated where Oslo occupies the last place on this list. What are the main similarities and differences between cities and which lessons can be drawn? 

I will focus my answer to Madrid and Oslo since, unfortunately, we did not have new data for Milan to explicitly compare with the other cities. Based on the main structural factors outlined above — globalisation; level on inequality; type of the welfare regime; housing regime; and occupational structure — we predicted that Madrid would belong to the groups of most segregated cities and Oslo would belong to the groups of least —  segregated cities. Most importantly, Madrid is a more global city than Oslo, the levels of inequalities are higher and social-democratic welfare regime in Norway as opposed to the Mediterranean regime in Spain would lead to lower levels of segregation in Oslo compared to Madrid. Thus, broadly speaking, there are important structural differences that help to explain the differences in the levels of socio-economic segregation in these two cities. On top of that, the city-specific contextual factors further emphasize the differences. I discussed about the role of housing boom in Madrid above in facilitating quick rise in segregation. Norway is an outlier in Europe when it comes to the ability to redistribute the wealth because of the vast revenues from oil. Hence, despite  the fact that the housing system is strongly market based in Oslo, there are compensating welfare redistribution mechanism that help to keep both social inequalities and socio-economic segregation very low, with the Index of Dissimilarity being only somewhat higher than 20.

But low levels of segregation are supported also by urban planning that has tried to avoid to develop vast homogenous neighbourhoods in Oslo. Namely, not only our research shows that the spatial organisation of the urban housing market is one of the crucial factor shaping segregation—housing is at the heart of segregation. In many cities in Europe, large high rise housing estates were built during the 1960s and 1970s to make up for housing lost during WWII and to accommodate the post-WWII birth wave. Initially these were desirable places to live, but as alternatives became available in newly developed suburbs and when access to private vehicles became more widespread, the middle classes moved out of these high-rise apartments and low-income households, among whom immigrants were often overrepresented, moved in. Clearly, large low-cost housing areas are a major driver of segregation by socio-economic status, because housing affordability criteria dictate where low-income households can live. But there are less of such homogenous areas in Oslo compared to many other cities. The lesson we can draw thus is that it is not always money that is needed to tackle segregation. Segregation can be tackled also in ways that are less costly, i.e. by introducing smarter planning principles, especially for developing of the new housing areas that would not lead to such high segregation as we can observe both in Madrid and Tallinn.


3.  You and your colleagues argue that the increase of spatial segregation is related to increasing social inequality on the one hand, and withdrawing governments on the other hand. And immigration flows enhance this effect too.  How are these developments a danger for the social stability and competitive strength of cities? What could national EU governments do?

Our study focused on socio-economic segregation. On top of that, the authors of each study introduce topics that are most relevant to understanding socio-economic segregation in their study city. Many of them highlighted that there is an increasing overlap between ethnic segregation and socio-economic segregation taking place in Europe (although the ethnic groups differ from each other in terms of segregation, too). In short, social inequalities are rising, these inequalities are translating more and more into the urban space in the form of rising segregation, fragmenting cities into socio-ethnic islands. Partly, this process is natural in the sense that always the purchasing power of people varies and people like to live together with similar people. Urban segregation is often the highest among the more affluent groups, both in Europe and in the USA. The most affluent households have the option to choose neighbourhoods of their preference, often neighbourhoods with other affluent households, and this is not necessarily bad at all. In fact, most people prefer to live next to others who are similar to themselves, also low income households. Living among similar people can reduce conflict, and give people a sense of shared identity, comfort and safety, and it can foster social contacts and social networks. Living in enclaves with people with similar preferences, needs, and life styles can also have the benefit of shared services and facilities (such as shops, cultural and religious facilities).

Socio-economic segregation is mainly problematic if it is involuntary, prevents people to have access to services and social mobility, and when living in poverty neighbourhoods is transferred between generations. At certain point, the social, ethnic and spatial fragmentation of the cities can thus lead to problems. However, it is impossible to tell at what level of the Dissimilarity Index or any other segregation index the problems become that acute that they start to harm the social stability and economic competitiveness of the cities. It is still safe to argue that extremes tend to lead to problems, and especially for the people in the bottom-socio-economic ladder. The analysis of the riots in Paris, London and Stockholm all show that not only poverty but the spatial concentration of poverty is related one way or another with destructive behaviour even at our times of Internet and social media that has reduced the importance of the “tyranny of the space”.

Too much poverty and loss of hope into a bright future among people that live around you is a problem. Downgrading physical environment, lack of neighbourhood resource, all matter. In fact, the times of social media, differences between neighbourhoods can even grow larger as many social media initiatives, down to crowdfunding of local initiatives, can amplify the differences in the quality of life in different neighbourhoods. Thus, this is the role of public sector to deal with such imbalances, but it is often hard task in practice. For example, if a local community is viable and takes care about the home area, also by contributing financially, is it a reason for re-directing public resources to neighbourhoods that lack such initiatives, especially at time of contracting public resources for urban development. People in well-functioning neighbourhoods can feel punished for initiative in such situations. This feeds back to the economic performance of the neighbourhoods and the cities, and needs careful action for the public sector in balancing social problems and social initiatives in different neighbourhoods of the city.

Furthermore, attention to the physical attribute, services, as well as connections between neighbourhoods and services are important in harnessing the economic potential of the cities. However, what matters the most, still, is people. There is little evidence about negative neighbourhood effects on the labour market outcomes of people. For example, there is some evidence on neighbourhood based discrimination by employers in the hiring process and the social networks are partly neighbourhood based, too. The role of neighbourhood based networks in social networks differs between population groups, being most important for lower socio-economic groups, women, ethnic minorities and elderly. What we know less, though, is how the neighbourhood shapes the future lives of the kids or the intergenerational transmission of poverty. We just lack good data that would allow to provide conclusive evidence in this regard. What is important here is that schools are often neighbourhood based, and an important transmission of values and formation of social relations takes place at schools. Thus, in neighbourhoods of high poverty, children are cut from many important social networks. Again, if social and ethnic segregation overlaps, we have a problem. For example, the case of Tallinn shows that such overlap could have serious implications on the labour market even in the context where a third generation of immigrants has born. The neighbourhood and language based school system in Tallinn systematically sorts people members of the majority and minority population to different labour market careers with minorities being hugely over-represented in the lower end of the occupational ladder.

In short, both direct disruptive behaviours in the form of riots as well as more subtle and long-term effects through the inter-generational transmission of poverty can influence the competitive power of cities. And the solution would be a combination of neighbourhood focused and more importantly, people focused policies that help to overcome the socio-spatial isolation of the various social groups, and their children, facilitating positive social interaction and the formation of more inclusive socio-spatial networks.


In part 2 of the interview Professor Tammaru sheds light on ‘The spatial gap between poor and rich’ as well as ‘’the five structural factors that contribute to socio-economic spatial segregation'. Next week the final part of the interview will be published. 

This book is written by a strong international team of researches with expertise on urban development, inequality, poverty, segregation, deprived neighbourhoods, neighbourhood effects, an d other urban studies related topics. The professors involved are Maarten van Ham (Delft University of Technology & University of St Andrews), Szymon Marcińczak (University of
Łódź) and Sako Musterd (University of Amsterdam).

Visit the website for more information 

Write a comment