The Cathie Marsh Centre for 
Census and Survey Research

Why Focus On Ethnic Composition? It’s Deprivation That Matters!

As segregation reappeared as a subject of high profile discussion on the political agenda, it prompted considerable debate regarding its nature, causes and consequences. It is segregation that was attributed as the cause of the 2001 riots in the Northwest of England and is regarded as a fundamental obstacle to realising community cohesion. In this political context, segregation has come to symbolise cultural difference. Moreover, the separation of cultures and lack of inter-cultural contact have been the focus of attention amongst policy makers at the expense of other issues such as housing allocation practices and unemployment.

While there has been much quantitative analysis of residential segregation, such approaches often assume ethnic minority concentration to be a causal factor, laying emphasis on ethnic minority groups rather than structural factors, such as housing, employment and income deprivation, which all have material impacts on individual life chances. We know, for example, that Pakistani and Bangladeshi communities are disproportionately concentrated in the most deprived neighbourhoods across the country.

In light of this, using data from the 2001 Census and the 2005 Citizenship Survey, this project set out to examine whether neighbourhood deprivation was more strongly associated with education, employment and health outcomes for South Asian ethnic groups than the ethnic composition of the neighbourhood in which they live. Neighbourhood deprivation was measured using a neighbourhood income deprivation score and ethnic composition was measured by the concentration of an individual’s ethnic group within their neighbourhood.

The results show neighbourhood deprivation to be more important in explaining poor outcomes in education, employment, and health than levels of co-ethnic concentration. Furthermore, the findings showed that the negative association that is often reported between ethnic minority concentration and education and employment outcomes became insignificant when levels of neighbourhood deprivation were considered. Figure 2 shows that there is little relationship

between probability of having no qualifications and co-ethnic concentration for three ethnic categories when levels of neighbourhood deprivation are taken into account.

In contrast, Figure 3 shows that for the same three categories the levels of neighbourhood deprivation are clearly related to the probability of having no qualifications after adjusting for co-ethnic group density.