The Journal

The Subjects

The Books






Housing and health: the nature of the connection

Alex Marsh

There has been surprisingly little scientific research on housing and health (Hunt, 1997, pp. 161-2).

Everyone with a grain of sense knows that it's bad for your health if you don't have somewhere decent to live (Frank Dobson, 1997).

Much of early public health policy in the UK was directed at the improvement of poor housing environments. With infectious disease both deadly and no respecter of social class, policy was as much inspired by the urge for self-preservation on the part of the middle classes as a concern for the welfare of the 'poor'. Improving housing was seen as central to improving public health, even if the mechanisms involved were only pinned down with any precision with the passage of time. The health dimensions of housing policy were prominent through to at least the slum clearance programmes of the 1960s.

After a prolonged absence, housing has resurfaced as an element in policy debates around public health and the improvement of the health of the nation. The issues of amenities, dampness, inadequate heating and energy efficiency are at the forefront of contemporary concerns (DoH, 1999, paras 4.28-4.31). There is ample evidence to suggest that these are features of the contemporary housing stock which can be deleterious. Action to address such problems with the existing stock is to be welcomed.

The renewed focus on the socio-economic determinants of health and the recognition that investment in the housing stock can play an important role in health improvement is a valuable change of emphasis in policy. Yet, arguably much needs to be done if we are to attain a satisfactory appreciation of the nature of the connection between housing and health. This paper draws upon DETR funded research looking into effects of poor housing upon health using the National Child Development Survey (NCDS) - a longitudinal dataset - involving myself and colleagues at the University of Bristol (Marsh et al., 1999). Four questions will be considered:

  • How secure is our understanding of the link between housing and health?
  • How should we be measuring housing deprivation?
  • Is it sufficient to focus on the direct pathways between poor housing and poor health?
  • Can we eliminate housing deprivation as a major influence upon health?

How secure is our understanding of the link between housing and health?

There is a well-established cross-sectional literature on the impact of various aspects of poor housing upon current health. The health risks posed by homelessness are also incontrovertible. Our knowledge of the longitudinal impact of housing upon health is much more limited. There is evidence that experience of poor housing environments in childhood carries implications for health status in adulthood, although the importance of childhood experiences relative to poor socio-economic circumstances in adulthood can be debated. Housing environments frequently do not feature in longitudinal analyses of health status and when they do single or simple indicators are typically employed. The move in epidemiological studies towards a life-course perspective on health risk presents the possibility of a greatly enhanced understanding of the determinants of health status. However, the role of housing both as a potential health risk in itself and as one of the intervening links in the chain which allows a particular health risk to develop into later illness has yet to be fully explored.

Our research using the NCDS strongly indicates that experience of housing deprivation has a substantial impact on the risk of cohort members suffering severe ill health or disability. We estimated that, after controlling for a range of other factors, experience of multiple housing deprivation increased the risk of severe ill-health or disability across the life course by 25% on average.

How should we be measuring housing deprivation?

Studies based on secondary data are constrained by the nature of the data collected. At a methodological level, one of the most important findings to emerge from our study using the NCDS was the need to look again at the indicators that we use to assess housing circumstances. Can they give us any insight into the nature of housing deprivation in contemporary society?

The housing information collected on a non-specialist survey such as the NCDS is largely concerned with longstanding issues around amenities. For example, the last sweep of NCDS, in 1991, contained the following variables that could potentially be used as part of a housing deprivation index:

  • Cohort member has been homeless
  • Dissatisfaction with the area they live in
  • Dissatisfaction with present accommodation
  • Accommodation has had serious problems of damp or mould
  • Overcrowding (more than one person per room)
  • Front door of the accommodation on or above the third floor of the building
  • Living in non self-contained accommodation (rooms, caravans, etc)
  • Not having sole access to a bath
  • Not having sole access to a kitchen
  • Lacking or having to share an indoor toilet.

Our analysis indicates that the reliability of a housing deprivation index based on these types of variable is reasonably satisfactory for the early sweeps of the NCDS (see Chapter 5 and Appendix 3). Yet, reliability plummets for later sweeps. Furthermore, it is clear that respondents' subjective assessments of their housing - both their dwelling and its broader locality - make a particularly important contribution to the indices and are strongly associated with health status.

These observations do not mark the NCDS as unusual. The results are, in fact, somewhat better than those reported for other non-housing specific surveys. We are drawn to the conclusion that, while questions focusing on issues of amenity and overcrowding may have been a good guide to the nature of housing deprivation up to and including the 1960s, there is now a need to examine alternative conceptions of deprivation which are more suitable for the 1990s and beyond.

Is it sufficient to focus on the direct pathways between poor housing and poor health?

A concern with lack of amenities, shared facilities and overcrowding is very much a concern with infectious disease, while damp and mould can cause asthma and other respiratory diseases. Overcrowding, damp and living above the ground floor can also carry with them implications for mental health. The discussion around housing and health thus tends to be concerned with elaborating upon the direct pathways from poor housing to health. There is much less consideration of the indirect effects of poor housing upon health. The increasing emphasis in epidemiological debate is upon the psycho-social origins of disease. In this context, the lack of detailed examination of the role of housing inequalities in the rise of the various ailments that dominate the current public health agenda is perhaps surprising. Following Wilkinson's (1996) emphasis upon inequality and social cohesion one might have thought that the role that the housing market plays in translating income inequalities into very visible signals of position within the social hierarchy would have attracted somewhat more attention. The potential of relative housing disadvantage to act as an indirect pathway to ill health, by inducing chronic stress and thereby inducing physical ailments, seems significant and warrants further exploration.

Housing is, to be sure, there in the background in current discussions: allusions to housing problems feature in the discussion as illustrations of the sort of issues involved when we think about the socio-economic determinants of health. But in a situation where housing deprivation is not a question of the absence of amenities or overcrowding, the question of how we conceive of housing deprivation remains. Equally the work needed to spell out precisely the mechanisms by which housing influences health remains to be done.

Housing deprivation will inevitably be broader than the lack of amenities, and incorporate considerations of neighbourhood. However, a nuanced view of the impact of housing on health would need to broaden the scope of the inquiry beyond poor housing in poor neighbourhoods. Ghodsian and Fogelman (1988) find, for example, that for NCDS cohort members living in the best off areas during childhood there was a marked difference in self-reported health between those with good amenities and those without. They argue that: 'it is reasonable to conclude that the disadvantages for subsequent feelings about ones health associated with having grown up with inadequate amenities appear to be greater if experienced in the contrasting setting of a well off neighbourhood' (Ghodsian and Fogelman 1988, p.75). Clearly, this suggests a strong subjective and relative component to housing deprivation. It could also be argued that the housing career or trajectory of individuals may determine whether particular housing circumstances are an indirect health risk or not: it may depend on whether climbing or descending the housing 'ladder' brought you to particular housing circumstances.

If greater emphasis is placed upon relative housing deprivation then the interpretation of the role of housing tenure needs to be considered. The owner/renter distinction is typically invoked as an indicator of differences in broader socio-economic inequalities and it clearly has some power to identify important cleavages within society. But that is not to argue that a relatively unfavourable tenure position, in and of itself, may be a cause of ill-health. Yet, with contemporary discourse referring to renters as 'damaged citizens' (Murie, 1998) and to the 'normalisation' of ownership (Gurney, 1999) one could construct an argument that residing in rented accommodation may - regardless of the quality of the housing environment - work through perceptions of social status to have an adverse effect upon residents' health. Clearly such an argument needs to be treated with some care and requires further elaboration.

Can we eliminate housing deprivation as a major influence upon health?

Emphasis upon remedying basic deficiencies in the housing stock - inadequate heating, overcrowding, etc. - carries with it the implication that we could, in principle at least, reach a point at which housing circumstances cease to be a significant public health concern. The move to a greater concern with the subjective assessments of residents and with housing inequalities opens the way - as it does with the broader move from a focus upon absolute to relative poverty - for housing circumstances to continue as a feature of health debate and policy. As Robson observed two decades ago: '[e]ven if, by absolute standards, the very worst conditions were &quotsolved", a new set of &quotvery worst" would automatically be created. Dealing as one is with relative and with rising expectations the problem of housing, like the poor, will ever be with us' (1979).

Linking housing and health

The socio-economic determinants of health have returned to policy debates and housing circumstances are identified as a key influence upon public health. Yet, while our knowledge of the link between housing and health is sufficient to underpin such action, I would suggest that our knowledge is less extensive and comprehensive than it might first appear. There are questions about the appropriate ways to both conceptualise and measure housing deprivation. There remains considerable work to elaborate fully the range of potential indirect impacts of housing circumstances upon health and to articulate the mechanisms which underlie these impacts. Beyond that task lies the need to test whether - and which - such impacts and mechanisms are significant in practice. Housing is likely to remain a component of the health policy debate for some time to come. If the suggestions presented above find support then housing is likely to be a permanent fixture of the debate. And a host of difficult questions for both housing and health policy will have to be faced.


DoH [Department of Health] (1999), Saving lives: Our Healthier Nation White Paper, Cm 4386, The Stationery Office: London.

Dobson, F. (1997), Healthy homes for healthy lives: Frank Dobson addresses National Housing Federation [16/10/1997], Department of Health Press Release 97/282.

Ghodsian, M. and Fogelman, K. (1988), A longitudinal study of housing circumstances in childhood and early adulthood, NCDS User Support Group Working Paper, 29.

Gurney, C. (1999), 'Pride and prejudice: Discourses of normalisation in public and private accounts of home ownership', Housing Studies, Vol. 14, no 2, pp. 163-183.

Hunt, S. (1997), 'Housing-related disorders', in J.Charlton and M. Murphy (eds.), The health of adult Britain 1941-1994, Vol.1, Decennial Supplement No. 12, HMSO: London.

Marsh, A., Gordon, D., Pantazis, C. and Heslop, P. (1999), Home sweet home? The impact of poor housing upon health, The Policy Press: Bristol.

Murie, A. (1998), 'Secure and contented citizens? Home ownership in Britain', in A. Marsh and D. Mullins (eds.) (1998), Housing and public policy, Open University Press: Buckingham.

Robson, B. (1979), 'Housing, empiricism and the state', in D. Herbert and D. Smith (eds.), Social problems and the city, Oxford University Press: Oxford.

Wilkinson, R. (1996), Unhealthy societies: The afflictions of inequality, Routledge: London.

Alex Marsh
Centre for Urban Studies
School for Policy Studies
University of Bristol
8 Priory Road

Tel: (0117) 954 5584
E-mail: alex.marsh@bristol.ac.uk


Journal 069 Index Top of page