Part Two: How many have slept rough? and what does that tell us?

At last week’s Australian Social Policy Conference Chris Chamberlain and Guy Johnson, from the Centre for Applied Social Research, RMIT University, presented new research into homelessness in Australia. Today’s post is part two of an edited extract of this presentation. It looks at the number of people in Australia who have slept rough, and what these findings mean for policy making.

The second aim of this paper is to estimate how many people have slept rough. The best information on where homeless people stay is collected by the National Census of Population and Housing held every five years.

However, it could be argued that the Census undercounts the number of rough sleepers, because it defines the homeless population as everyone who is homeless on Census night. For other purposes, the homeless population can be thought about as everyone who becomes homeless during a given period of time (e.g. during the year that the Census takes place). If we think about the homeless population in this way (an annual count), then we can understand why some rough sleepers will not be counted on Census night. Some will have slept rough earlier in the year but were somewhere else on Census night (e.g. a boarding house, a friend’s place etc.). Others will have slept rough earlier in the year but were no longer homeless on Census night. Another group were housed on Census night but slept rough later in the year. For these reasons the number of rough sleepers will always be higher in an annual count than a census count. Taking this one step further, a lifetime count should always be higher than an annual count. Direct comparisons of the results of point-in-time, annual and lifetime counts need to bear this in mind.

We asked everyone in the survey , ‘Did you EVER sleep rough, stay in your car, stay in some other public place, or live in an empty building for any of the time that you were homeless?’ We found that 7.8 per cent (95% CI = 6.3%, 9.2%) of the sample (or about 1.4 million Australians) reported they had slept rough at some point in their lives. This means that of those who reported they had ever been homeless 59 per cent had slept rough, with 67 per cent of men and 50 per cent of women reporting that they had done so (Table 7). Earlier, we saw that approximately 1.35 million men and 1.0 million women had experienced homelessness. When these two pieces of information are combined, then it can inferred that about 900,000 men and 500,000 women have slept rough at some point during their lives.

Table 7: Replies to the question: ‘Did you EVER sleep rough, stay in your car, stay in some other public place, or live in an empty building for any of the time that you were homeless?’ (%)











It is usually assumed that people are more likely to sleep rough the longer they are homeless. Cameron Parsell (2014, pp.237-238) has pointed out that in the minds of many policy makers, rough sleeping has become inextricably linked with the idea of long-term homelessness. For example, the Labor Government’s White Paper, The Road Home (2008), noted that: ‘Rough sleepers and people who are chronically homeless are more likely to have complex needs such as mental health issues, substance abuse and disabilities’ (Australian Government 2008, p.8).

Table 8: Persons who slept rough during the time they were homeless, by duration (%)

        Duration of homelessness

 Less than 4 weeks (N=58)1 to 5 months (N=49)

6 months or more




% who slept rough most recent time

% slept rough ever









*Includes three people who did not report information on duration

Table 8 shows that rough sleeping is just as common amongst people who have a short-term problem with homelessness as it is amongst people who have a long-term problem. Fifty-two per cent of those with a long-term problem had slept rough the last time they were homeless, compared with 57 per cent of those with a short-term problem. Similarly, 54 per cent of respondents with a long-term problem had slept rough at least once in their lifetime, compared with 65 per cent of those who had a short-term problem.

Table 9: Persons who slept rough by age first homeless (%)

            Age first homeless






25 or older (N=64)



% who slept rough76604459

*Total includes four persons who became homeless before the age of 12

The results also suggest there is an association between the age people first become homeless and whether or not they sleep rough. Table 9 shows that 76 per cent of those who became homeless as teenagers (defined as young people aged 12 to 18) had slept rough, compared with 60 per cent of those who became homeless as young adults (aged 19 to 24), and 44 per cent of those who became homeless as adults (aged 25 or older). Moreover, one-third (32 per cent) of the formerly homeless were teenagers when this first happened. Thus, it is unlikely that our research has over-estimated the percentage of homeless people who have slept rough, because teenagers aged 12 to 17 were not included in the sample and they have higher rates of rough sleeping than adults.


This paper set out to investigate two issues: how many Australians have experienced homelessness during their lifetime and how many people have slept rough. The first Australian research into lifetime homelessness found that 13 per cent of adults had experienced a period of homelessness during their lifetime. This research confirmed that finding. Both results are much higher than reported in point-in-time accounts such as the Census, and lend credence to the view that point-in-time results underestimate the magnitude of the problem.

Our research also draws attention to the fact that 15 per cent of men have been homeless, compared with 11 per cent of women. This means that approximately 1.35 million men and 1.0 million women have experienced homelessness. While our findings confirm that men are more likely to experience homelessness, homelessness is by no means a uniquely male experience – indeed it is clear that a significant number of Australian women have experienced homelessness, and half (50 per cent) of them have slept rough.

It was also clear that women are more likely than men to have a short-term problem, and that for 70 per cent of women homelessness is a one-off event. The duration of homelessness, as much as the incidence, suggest that men and women experience homelessness differently. One possible explanation why women have shorter experiences of homelessness is that there are more services available to at risk and homeless women, relative to men. If service availability is a factor in the gender differences observed here, the key policy issue is not redistributing existing funding, but rather to consider options to broaden the availability of support services to at risk and homeless men.

The second aim of the paper was to estimate how many people have slept rough. Many people have concluded that rough sleeping is rare in Australia. However, for reasons already alluded to, census counts under-estimate the prevalence of rough sleeping. Our research revealed surprisingly high rates of rough sleeping – indeed nearly 1 in 10 Australians (7.8 per cent) had slept rough at some point in their lives. Not only do our findings suggest that the Census under-estimates the extent of rough sleeping, but also that rough sleeping is a relatively common experience among the homeless – we found that 59 per cent of the homeless have slept rough at some point in their lives, or approximately 1.4 million people.

An equally important finding was that many people who had a short-term experience of homelessness had also slept rough. In the policy realm and public imagination rough sleeping is often conflated with chronic homelessness. This makes intuitive sense as visible rough sleepers are often chronically homeless individuals who suffer from a range of behavioural and health issues. The visibility of this group informed the ABS special enumeration strategy which works closely with service providers to identify where rough sleepers ‘hang out’ so that census forms can be handed out. Many of these services (including food vans) work with the long-term homeless and are disproportionately located in the inner suburbs of the capital cities. People who have a short-term problem, particularly those from the outer suburbs, as well as in regional centres and country towns, often have no knowledge of these services and do not ‘hang out’ where the homeless congregate in the capital cities. People in the outer suburbs, regional centres and country towns are more likely to sleep at their local park, in their car, at the local football ground or in the bush on the edge of town.

Rough sleeping is more common than previously thought and it is not confined to chronically homeless individuals. Current policy settings are focused on reducing rough sleeping, but the emphasis is almost entirely on chronically homeless individuals. Reducing rough sleeping require programs that target the chronically homelessness, but there is also the need for services that emphasise early intervention to assist the short-term homeless resolve their problems before they begin to sleep rough.

Taken together our findings offer important insights into the extent of homelessness, the magnitude of rough sleeping and the temporal characteristics of rough sleepers in Australia.

Posted by Sarah Toohey