Re-counting the homeless

Counting the Homeless reports provide the most comprehensive picture of homelessness in Australia today. However, recently the Australian Bureau of Statistics (ABS) released a discussion paper that calls into question the methodology underpinning Counting the Homeless figures. In the paper, the ABS proposes a new methodology for estimating the number of homeless Australians which, if it were implemented, would result in a statistically significant reduction in the estimate of homeless Australians.

For example, applying the ABS’s proposed methodology to 2006 census data reduces the number of homeless people by around 40 per cent, from 104 676 to 63 472 people. Using the methodology on 2001 census data reduces the homeless figure for that year by around 35 per cent, from 99 000 to 65 384. Application of the new methodology suggests that there has been a decrease in homelessness between 2001 and 2006 of 2.9 per cent, rather than an increase of 4.8 per cent, as is otherwise indicated by Counting the Homeless figures.

Not surprisingly, the outcome of the current discussion over how homeless Australians are counted will be of some consequence. For one thing, if the new methodology was adopted, it would mean that the Government’s goal of halving homelessness by 2020 would have been all but achieved. A new methodology could also have implications for the ways in which homelessness is tackled, as the Australian Government relies on Counting the Homeless data in developing homelessness policy and programs and allocating resources.

So, how do we currently count the homeless? Every five years the ABS conducts a census in Australia. As a part of the census process, the ABS uses particular methods and specially trained collectors to arrive at the most accurate count of the homeless population possible. Supplementary information on homeless school students and people staying in Supported Accommodation Assistance Program (SAAP) emergency or transitional accommodation is then used by Counting the Homeless researchers to adjust the raw census figures.

There are a number of issues associated with deriving homelessness figures from census data. Homeless people are inherently difficult to count in a census which relies on dwelling-based enumeration. This situation is exacerbated by the transient nature of the homeless population, with homeless people frequently moving from one form of temporary accommodation to another. The ABS attempts to deal with these problems in a number of ways. It undertakes its census count over three or four days, rather than on census night alone, works closely with providers of homeless services and uses people who have experience working with homeless Australians, in order to count rough sleepers.

While methodological issues such as the above are clearly important, it is, arguably, more important to agree on precisely who should be counted as homeless before undertaking a census of Australia’s homeless population. What is required, then, is an operational and generally agreed-upon definition of homelessness, one that enables the collection of meaningful data that can be used to inform policy responses to homelessness.

The ABS employs the definition of homelessness that has been developed by homelessness researchers, Professors Chris Chamberlain and David MacKenzie. This cultural definition of homelessness is based on ‘shared community standards about the minimum housing that people have the right to expect, in order to live according to the conventions and expectations of a particular culture’. The minimum community standard is, in their view, ‘a small rental flat—with a bedroom, living room, kitchen, bathroom and an element of security of tenure’.

Thus, the cultural definition of homelessness encompasses not only those people who are literally homeless (those who do not have a roof over their heads), but also those who are living in circumstances that fall beneath the minimum community standard.

Some commentators have criticised the Chamberlain and McKenzie definition of homelessness on the grounds that it is too broad and thus potentially overstates the number of homeless Australians. That said, it is important to bear in mind that Australia is one of the few developed nations that has a workable definition of homelessness. This definition has been used for some years, and the ABS has indicated in its discussion paper that it has no intention of changing it.

Nevertheless, through its proposed revisions to the Counting the Homeless methodology, it could be argued that the ABS is challenging aspects of the cultural definition of homelessness. The ABS proposes to revise Counting the Homeless assumptions in relation to: people reporting no usual address but counted in private dwellings on census night; young people staying in a place that was not their usual address, but citing a usual address; and, people counted in improvised dwellings. It has also more strictly applied the Counting the Homeless methodology with regard to people in boarding houses and staying in SAAP accommodation. Applying each of these revisions results in reduced numbers of homeless people being recorded.

It should be noted that the ABS review is still underway, and that many researchers and workers in the homelessness sector have raised concerns around some of the assumptions underpinning the ABS’ proposed new methodology. For example, the national peak homelessness body, Homelessness Australia has taken issue with many of the above proposed revisions. Based on the views of those who work in the homelessness sector, it has argued that the Counting the Homeless reports are more likely to under-estimate rather than over-estimate the number of Australians who are homeless on any given night. Similarly, George Seymour, the president of a youth homelessness shelter in Queensland, has pointed out that some of the assessments of young people’s experiences ‘are sharply at odds with the experiences and views of the sector and researchers’.

Without discounting the importance of developing a methodology for counting the homeless that is rigorous and consistent, it is unlikely that universal agreement will be reached on technical issues such as definitions and census methodologies. In any case, arguably what is most important is that we have enough of a grasp of the scale and nature of homelessness in Australia to develop and implement effective policy responses. As Seymour puts it:

Whatever the outcome of the ABS review, whether it is deemed that we have 105,000, 63,000 or some other number of people homeless on any one night, the moral imperative remains. We should never turn our backs on the most disadvantaged members of our society. A society is only as good as the circumstances it allows its most unfortunate and dispossessed members to find themselves in. We may not have the ability to count the homeless with full accuracy, to precisely quantify the problem, but as a people we have the resources, the compassion and the ingenuity to fix it.


Flagpost is a blog on current issues of interest to members of the Australian Parliament

Logo - Parliamentary Library Department of Parliamentary Services

Filter by



Tag cloud