Reviewed by
Andreas Koch
University of Salzburg
“Measuring residential segregation only matters because residential segregation matters”. Indeed, there are substantial reasons to assume that segregation has an impact on people, communities and neighbourhoods. “It matters to people’s life chances […]. Where you live matters to your future.” “It also matters with regard to lifestyle and development of the ideologies and attitudes that underpin how people behave”. “Segregation matters, too, because it is frequently a source […] of individual and, especially, group identity and difference within society that can generate tensions and occasional conflict, especially where the differences are linked with economic, social and political disadvantage” (p. 14f). All these quotes are drawn from the second chapter of the book Social-Spatial Segregation. Concepts, Processes and Outcomes, edited by Christopher Lloyd, Ian Shuttleworth and David Wong, and they express a strong claim for research in this field.The remarkable issue with segregation lies in the relationship between social and spatial facts. The vertical and horizontal structuring of a society, access to markets such as labour, housing, and education markets, or the realisation of one’s own capabilities within communities, are all determined – or at least influenced – by spatial configuration and composition in its material, functional and temporal manifestations. Thus, if societal problems are analysed and discussed in science and politics one has to incorporate the spatial dimension in order to make these analyses and discussions an adequate endeavour. As Danny Dorling in chapter 15 of the book asks: “So what is it to be ‘normal’, to live in a ‘normal’ area? Well, there is nothing normal about being normal in a highly class-segregated society [like the British society]. Most people are not normal but it is normal to be either better-off than average or worst-off – polarisation is the norm” (p. 368).
This peculiar relationship gives the book its particular meaning. The editors put a lot of effort into framing the topic of segregation coherently, which has been done by the three core notions of “concepts”, “processes”, and “outcomes”. The section on “concepts” is divided into seven chapters covering such topics as “spatial pattern statistics”, “application of spatial interaction data”, “historical data”, “individualised neighbourhood data”, and “international comparability of ethnicity and collective identity”. Although a strong focus on geographical methodology is obviously given in most of the chapters, attempts to rethink and revise social attributes in segregation research also exist to some extent (the contribution by Pablo Mateos on “The international comparability of ethnicity and collective identity: implications for segregation studies” is the one within the concept section). Interested readers of these social attributes, with an emphasis on social interactions, might be interested in Ioannides’s (2013) book From Neighborhoods to Nations.
Progress in geographical methodology is actually much needed because different geographies of units determine concepts of segregation measures differently. It is not only the well and longly known modifiable areal unit problem (MAUP) that requires alternative spatial approaches to segregation studies, but also the fact that “traditional segregation indices were aspatial in nature because their formulations assume that people across enumeration unit boundaries, regardless of the nature of the boundaries, cannot and do not interact” (p. 47f) as Wong stresses in his contribution. Though these problems of spatial representation in models and statistical analyses are doubtlessly worth further investigation the same is true for the social attributes such as ethnicity, status, class, cultural identity, and the like.
The section of “processes” is divided into five chapters. It is important to highlight that the definitions of processes used in these chapters are different from those that are usually used in the agent-based modelling and simulation community. Comparatively little attention is paid to the meaning(s) of time(s) in segregation research, for example, with respect to changes in population composition due to migration (dynamic) and natural development (evolutionary). The reference to agent-based modelling occurred only once in the book, namely in the concluding chapter, by emphasising its value in creating almost realistic representations of social-spatial relationships instead of random distributions of households and neighbourhoods. The value of this section, however, rests upon approaches that deal with the difficult use of cross-sectional and cross-temporal data, i.e. the linkage of social (demographic) phenomena at specific moments in time and place – without much knowledge of the processes in between. This restricted knowledge, however, will be enriched by historical knowledge that helps to better understand the results at the respective time points. For example, Maarten van Ham and David Manley in their contribution about “Segregation, choice-based letting and social housing” (chapter 11) discuss how social housing policy can affect the segregation process, thus what the effects an intermingling of social, political and spatial processes can have on segregation.
The “outcomes” section is divided into four chapters and addresses questions of the implications of segregation. The analytical matrix here is structured along a social and geographical scale in one dimension and the number of addressed problems in the other. While, for example, Emma Catney in her article about “’Religious’ concentration and health outcomes in Northern Ireland” considers segregation at a geographically local and socially specific level, Danny Dorling in his contribution about “Class segregation” creates an image of class-based social segregation in Britain mainly at the macro-level (however, he also included an interesting cartogram of social class residential segregation in Britain taking parliamentary constituencies as spatial units into account).
One strength of the book is found with its clear focus on methodological challenges in segregation research. Thorough debates about which measures can contribute what to correctly represent issues of segregation – which is itself a notion that can be referred to as concentration, centralisation, clustering, exposure and evenness – may help researchers in the domain of social simulation modelling to develop valid simulation models in terms of the social attributes, the spatial representations, and the temporal scales used. Of course, it would have been advantageous if there had been a chapter included which utilised agent-based modelling in segregation analysis (see, for example, Heppenstall et al., 2012).
Another strength lies in the deliberate self-limitation of the content presented. The book addresses conceptual and methodological problems of segregation mostly by presenting empirical studies from the UK, Ireland, the US, and Sweden. There could have been other national studies included from other regions (e.g. Asia and Africa), but it would likely not have enhanced the general awareness of contemporary challenges that socio-spatial segregation has to deal with. A reasonable number of national case studies allows the reader to compare empirical results across different methods. This restriction of empiricism opens up the opportunity to extend analyses towards implications by putting political questions into the arena. In doing so, segregation research may contribute to surmount mutual ignorance of this problem due to the fact that more often the rich and the poor live in – socially and spatially – distinct worlds.
The book therefore is highly recommended to modellers who are interested in coping with geographical conceptualisations of space, tailored measures of segregation, and applications for a non-academic audience, because they may wish to expand their models in these directions.
ReferencesHEPPENSTALL, A., Crooks, A. T., See, L. M., Batty, M. (eds.) (2012): Agent-Based Models of Geographical Systems. Springer, Dordrecht Heidelberg London New York.
Return to Contents
of this issue