Poverty, housing tenure and social exclusion

Peter Lee, Alan Murie

This research examined the link between patterns of multiple deprivation and housing tenure.

It concludes that housing policy and regeneration activity targeted exclusively at areas of council housing would exclude some disadvantage groups and that any national policy needs to be able to accommodate different local circumstances.

Out of print and PDF not available. Please see html version of summary below.



Newly completed research has explicitly examined the association between patterns of multiple deprivation and housing tenure. The researchers, Peter Lee and Alan Murie of Birmingham University, used the 1991 Census to explore how far national patterns are replicated at a local level and how far the policy solutions which would appear to be appropriate from national data are likely to be appropriate locally. The research demonstrates that:

  • The national trend for disadvantaged households to gravitate towards council housing and the social rented sector generally applies at a local level.
  • Disadvantaged groups are not exclusively housed in the social rented sector or in council housing.
  • In particular, those experiencing long-term illness are not so heavily concentrated in council housing. Economically disadvantaged households from minority ethnic groups are less concentrated in council housing.
  • The associations between housing tenure and disadvantaged households differ between different cities.
  • In the period 1981 to 1991 the gap between the affluent areas in cities and the most deprived areas widened.
  • The most deprived areas of cities are not exclusively areas of council housing.
  • Housing policy and regeneration activity targeted exclusively at areas of council housing would exclude some disadvantaged groups.
  • The researchers conclude that the national framework for policy needs to acknowledge that particular local circumstances require locally specific strategies.

Social exclusion and housing

The term 'social exclusion' is relatively new to the housing debate in Britain. However, it is particularly relevant in a period in which there is increased evidence that housing circumstances relate to and contribute to problems of social disadvantage more generally. Housing situations are not simply products of poverty but themselves contribute to the difficulties facing households and affect social integration. There is increasing awareness of the ways in which homelessness or living in deprived neighbourhoods contribute to social exclusion. Rather than assisting people to participate fully in society and increasing their opportunities and life chances, housing and housing policy are increasingly regarded as contributing to the processes which disadvantage people. The experience of homelessness and of living in disadvantaged neighbourhoods are key elements in the process through which people are disadvantaged.

This project sought to explore these issues and particularly to focus upon the extent to which there are direct associations between different housing tenures and disadvantage.

This debate grows out of a long-established discussion concerned with the residualisation of council housing. National data have demonstrated for a considerable period that there is a tendency for households with fewest resources to gravitate towards the council housing sector. This analysis demonstrates that this effect generally applies locally as well. However, it is not equally pronounced in all areas and patterns vary. In some cities deprived groups are more exclusively concentrated in council housing than in others. This suggests there is a need to be sensitive to these local differences, rather than to accept the national pattern as one which applies everywhere.

Mapping deprivation

A key element in the research carried out has been a review of approaches to measuring and mapping deprivation using the 1991 Census. The full results of this review have been published (see 'Further information'). While all approaches to measuring and mapping deprivation using Census data are flawed, the research has evaluated different approaches and identified key issues related to the construction of indices, double counting and the weights attached to different elements within indices. From this evaluation an index based on the Breadline Britain survey with recalculated weights appropriate to data available at enumeration district level was used to identify areas of deprivation. The Breadline Britain approach considers deprivation as being the absence of goods and services which large proportions of the population believe to be essential. The pattern of deprivation which emerges from this is different from that which emerges at a local level using other indices. Whilst the government's Index of Local Conditions may be the most appropriate index in terms of urban policy, it is not necessarily the best index for all purposes. The patterns of deprivation indicated by the Breadline Britain index should be considered in policies designed to target poor areas.

Deprived areas and housing tenure

At a national level, the pattern of deprivation indicated through the Breadline Britain index or other indices does not coincide neatly with patterns of housing tenure. Council housing is much more significant in Scotland and the variations regionally and locally do not relate to variations in levels of deprivation. However, the most important aspect of this relates to analysis at a local level, at ward and neighbourhood level.

The most detailed analysis carried out in this project focused upon five local authorities: Birmingham, Bradford, Edinburgh, Liverpool and the London Borough of Tower Hamlets. It involved mapping and analysing patterns of housing tenure and patterns of deprivation and assessing the relationship between the two. The research demonstrates that at the neighbourhood level there has been a polarisation within each of these cities. This is most clearly related to unemployment. The gap between the best and the worst areas in terms of unemployment increased between 1981 and 1991.

The general direction within each of the cities is that the most disadvantaged areas are most often areas of council housing. However, many of the most disadvantaged areas are areas of mixed tenure or tenures other than council housing. There are major differences in the spatial distribution of poverty and its relationship with housing tenure in different cities. This analysis has implications for policies concerned with area regeneration and the extent to which this should focus on council housing. It suggests that there is a great danger in assuming that targeting resources on mass council estates would deal with problems of deprivation in cities. In some cases this would be appropriate. In Edinburgh, targeting council areas would be an effective way of targeting deprivation. However, even in Edinburgh there is a question about which council estates to target. This question becomes more profound in Tower Hamlets where an indiscriminate identity of council housing with deprivation is inappropriate and there needs to be some more accurate identification of which council estates have the greatest concentrations of deprivation.

Looking at Birmingham or Liverpool or Bradford, targeting resources purely upon council estates would seriously neglect the problems of deprivation which exist in other tenures. Some council estates have lower levels of deprivation than areas of older owner-occupied and privately rented housing. Table 1 below summarises how the map of deprivation relates to council housing.

Table 1 The spatial incidence of deprivation and council housing
City/District Location of deprived areas Location of council housing
Birmingham Deeply rooted in the inner city Polarised structure - periphery and centre locations
Bradford Concentrated predominantly in the inner city Mosaic - small concentrations scattered around the district
Edinburgh In peripheral areas Highly peripheral
Liverpool Scattered but tending towards the centre Mixed - concentrations at centre and periphery
Tower Hamlets Scattered Widespread - estates throughout the district

It is also apparent that while deprivation in some cities remains concentrated in the inner city, in other cases deprivation is more dispersed or is greatest in peripheral areas. The concentration of resources in areas of council housing or in either inner or outer areas of cities would not always target the most deprived areas. Moreover it would also discriminate in terms of other attributes.

The differences between cities are explored further at a household level using the Sample of Anonymised Records from the population census. This analysis shows again that the most deprived households are most likely to live in council housing and are least likely to be households in the process of purchasing dwellings.

The association between council housing and deprivation is strongest when reference is made to indicators of income or economic deprivation. The pattern is less strong when reference is made to long-term illness. Even in relation to economic deprivation there are high levels of deprivation in other tenures and in some cities the difference between council housing and other tenures is the same everywhere.

While the greatest concentration of disadvantage amongst lone parents is in council housing, there are many such households in other tenures and the overall pattern varies between cities. The most striking example which raises questions over assumptions about where concentrations of deprivation are, and the one with the most profound implications for policy, relates to minority ethnic households. For both non-white and white households in these five cities, the highest levels of deprivation are found in the council housing sector. However, it is clear that the association between deprivation and council housing is much weaker among the non-white population. Stated simply it is much truer to argue that deprived households from the white community are heavily concentrated in council housing than it is for the non-white population. In Birmingham, Bradford and Liverpool, for example, the non-white population in all of the deprived categories (unemployed, no car, no income etc.) is less likely to be living in council housing. For example, in Bradford 35 per cent of white and less than 10 per cent of non-white headed households without a car live in council housing (Table 2).

Proportion of households without a car living in council housing
White headed households Non-white headed households
Birmingham 45.7 33.7
Bradford 35.2 9.5
Liverpool 42.3 30.3
Tower Hamlets 66.4 79.8

The implications of this are that targeting council housing is much more likely to be effective in targeting deprived groups within the white population than it is in targeting deprived groups in the non-white population.


The analysis carried out in this research is an important complement to recent work on residualisation which has generally been carried out at a national level and makes little reference to local variations. In confirming that the greatest concentrations of deprived households are in the council sector, the research emphasises the need to develop policies targeted on the council sector, to improve that sector and to improve the opportunities of people living within it. Regeneration activities focused on council estates are an important part of future housing and urban policy.

The researchers conclude that this work provides an important qualification to this and shows that it is essential that approaches to regeneration do not focus exclusively on council housing. The implication for policy is the need to develop local strategies which are based on detailed analysis of local circumstances and which will enable the different elements, both in patterns of deprivation and patterns of housing, to be taken into account in decisions about resource allocation, policy and practice.

About the study

Details of the methodology are given in the section 'Mapping deprivation'.