Skip to content

Housing banner

Quality Frameworks for Homelessness and Related Services – Literature Review and Environmental Scan

1 Homelessness: the current context - Australia

1.1 Introduction

As part of the Australian Government's White Paper - The Road Home: A National Approach to Reducing Homelessness, the Government is planning to develop agreed national accreditation service standards and service charters for organisations delivering services to homeless people.2 It is intended that by developing a national service charter for mainstream and specialist homelessness services, progress towards protecting the rights of homeless people will be achieved by providing them with safer and higher quality services.

This environmental scan and literature review aims to contribute to the knowledge base available to inform decisions relating to the development of the new legislation to support the above initiatives.

This section sets the context for this work by providing an overview of the current population of Australia which is considered to be homeless on the basis of current agreed definitions. Counting the homeless is an imprecise science due to the difficulties inherent in capturing what is a diverse population living in a diverse range of accommodations. In recognition of this, this section discusses the range of definitions of homelessness currently in use, namely:

On the basis of 2001 and 2006 Census data, this section also provides an overview of the profile of the homeless population within Australia in relation to characteristics such as age, sex, and household composition (i.e. families, couples, single people).

It also provides an overview of rates of homelessness, both nationally and across individual states and territories, and highlights changes which have occurred to this population during the five years between the two censuses.

[ top ]

1.2 Homelessness definitions and demand for services

1.2.1 The Road Home: approach of the White Paper on Homelessness

According to estimates contained in the White Paper on Homelessness, around 105,000 people are homeless across Australia every night. This includes people staying in 'temporary, unstable or substandard accommodation' as well as those sleeping rough.3

The White Paper applies the widely accepted cultural definition of homelessness, developed by Chamberlain and MacKenzie and utilised by the Australian Bureau of Statistics (ABS) in Census-counting of the homeless population, discussed in Section 1.2.2 below.4 Estimates of the homeless population contained in the White Paper are based on data provided by the ABS five-yearly Census, as well as data collected from services operating under the Supported Accommodation Assistance Program (SAAP).

The White Paper notes that the profile of the homeless population has changed over the past five years and today includes rising numbers of children, families and older people, along with declining numbers of teenagers.5 This changing profile is discussed further in Section 1.2.3, below.

On the basis of this research, the White Paper's overarching objectives are to:

[ top ]

1.2.2 Defining and counting the homeless population

The Supported Accommodation Assistance Act 1994 (SAA Act) defines homelessness as: 'A person is homeless if, and only if, he or she has inadequate access to safe and secure housing'.6

The legislation outlines factors which may contribute to a lack of access to 'safe and secure housing'.7 These include situations where the only housing to which a person has access:

Whilst this definition identified the target group for service provision under the Supported Accommodation Assistance Program (SAAP), a wider definition is considered to be a 'culturally accepted' definition of homelessness developed by researchers Chamberlain and MacKenzie.8

Cultural definition of homelessness

In October 2000, the SAAP National Coordination and Development Committee (CAD) commissioned research aimed at providing an estimate of the extent of homelessness in Australia. This involved academics, government and non-government representatives and researchers.

This initiative led to support for a definition of homelessness based on the work of Chamberlain and MacKenzie,9 which defined homelessness across three levels:

[ top ]

The CAD supported this definition, rephrased as:

Limitations to the SAA Act 1994 definition have also been pointed out. Thompson11 emphasises limitations in SAAP data collection methods, particularly in relation to calculating 'unmet demands for accommodation'. These include people needing accommodation who cannot be housed by SAAP agencies. Figures on unmet demands are collected biannually during one week in November/December and one week in May.

'Unmet demand' figures exclude three categories of people:

[ top ]

According to Thompson, not including these categories of people seriously underestimates the turn-away rate from SAAP. For example, if these categories of people were included in 2003/04 the turn-away rate would have been four times that of the official rate. In this context, Thompson asserts that SAAP fell a long way short of accommodating the needs of Australia's homeless population. This was not the fault of the SAAP system 'which on the whole [was] well conceived and well managed, and helps many people in various ways'.12 Instead, Thompson maintains the fault lies with the government allocation of resources and the lack of funding directed at dealing with the problem of homelessness.

Thompson concludes that the published figures on homelessness are not an ideal source of information on the exact extent of homelessness in Australia. For most purposes, the Census and SAAP data are not comparable. The Census takes a cross-section of Australian society at one point in time, whereas the SAAP data on the total numbers of people assisted relates to the whole year. The Census also attempts to cover the whole population, whereas the SAAP clientele is unrepresentative and largely self-selected. SAAP figures on unmet demand understate the extent to which SAAP was failing to deal with homelessness; Census figures give a fuller picture, but they too are limited.

Homelessness among Aboriginal and Torres Strait Islander people

It has been recognised that the definition of homelessness among particular cultural groups, including Aboriginal and Torres Strait Islander people, may need to be widened to include people in 'grossly overcrowded' accommodation and people in 'impoverished dwellings'.

In 2005 Urbis Keys Young13 consulted extensively with Aboriginal and Torres Strait Islander people - a project which led to a definition of five types of homelessness among these particular cultural groups:

The report emphasises the importance of an approach to these cultural groups which is sensitive to their particular culture and 'outside the legacy of colonialisation'.

[ top ]

ABS Census counting methodology

Historically, the Census methodology for counting homeless people has consisted of a 'Special Enumeration Strategy' based on the fact that homeless people do not live in houses - the usual basis for collecting Census data. Therefore, special measures are needed to contact those who have no fixed address.

The Australian Bureau of Statistics' (ABS) strategy involves including a category of 'no usual address' on the form for visitors to households as well as a shortened version of the form for people who might be intimidated by the longer form, enlisting the help of service providers and using their local knowledge, and recruiting homeless people themselves as Census collectors.

According to Thompson,15 the ABS strategy for counting the homeless has a number of limitations, including:

[ top ]

1.2.3 Profile of the homeless population

A comparison of 2001 and 2006 Census data indicates that although the number of homeless people has increased over this five year period - from 99,900 homeless people counted in the 2001 Census to 104,676 in 200617 - the rate of homelessness has remained steady. There was a rate of 53 homeless people per 10,000 of the population at both Census periods (a figure which should be read in recognition of the fact that changes from 2001 to 2006 may not be entirely accurate due to possible differences in counting rules, sampling methods, and collection rates).

Rates of homelessness in each state and territory also varied little between 2001 and 2006. In New South Wales, Victoria, and the Australian Capital Territory, there were 42 homeless people per 10,000 in 2006, similar to the 2001 rates. South Australia and Tasmania had a rate of 53 per 10,000 in 2006, again similar to 2001. The rates of homelessness in the other states were higher: in Western Australia and Queensland there were, respectively, 68 and 69 per 10,000 at the time of the 2006 Census. In the Northern Territory there were more than three times as many: 248 homeless people per 10 000 in 2006 and 288 per 10,000 in 2001 (refer to Table 1 below).18 In specific regions within the territory, the rate is even higher. In the Kimberley region, for example, there were an estimated 1,870 homeless people, at a rate of 638 people per 10,000.19

This significant differential in rates of homelessness found in the Northern Territory, a rate which is more than five times the national average, is considered by Chamberlain to partly reflect issues such as 'itinerant people on working holidays during the Kimberley dry season'20 and 'issues of Indigenous transience and inadequate housing, although non-Indigenous Australians also experience higher rates of homelessness in remote locations'. Yet even this high figure is expected by the academics to be based on an undercounting of the Indigenous homeless population within the Northern Territory.


Table Description
Table 1 - Homeless rates per 10,000 by state and territory - 2001 and 2006
      NSW      VIC      QLD      WA      SA      TAS      NT      ACT      Aust.     
2001 42 44 70 64 52 52 288 40 53
2006 42 42 69 68 53 53 248 42 53

Source: Chamberlain C., and MacKenzie, D., June 2009, 'Counting the Homeless, 2006, Northern Territory,' Australian Institute of Health and Welfare, p.25, viewed at http://www.aihw.gov.au/publications/hou/hou-210-10761/hou-210-10761-c04.pdf

[ top ]

The profile of the nation's homeless is heterogeneous. Homeless people come from a broad range of ethnic backgrounds and age bands and exhibit a variety of personal circumstances and problems.

Today a homeless person is increasingly likely to be younger than at the time of the 2001 Census, i.e. less than 35 years of age. Although there are still higher numbers of homeless males than females, the proportions vary significantly across the various homeless accommodation options. For instance, although more males than females utilise boarding houses across the country (ratio of 72:28) females outnumber males in accessing crisis accommodation (ratio of 53:47). Although homelessness is still more likely to be experienced by lone individuals, there are increasing numbers of homeless or single parents with children as well as a growing proportion of homeless families.21

The Counting the Homeless report of 2006 provides an estimate of the composition of the homeless population and a comparative compositional analysis with 2001 figures.

Of the 104,676 people who were counted as homeless in the 2006 Census, there were:

In addition:

[ top ]

Indigenous persons are over-represented amongst the homeless. The 2006 Census indicates that while only 2.4% of the national population was identified as Indigenous, 9.9% of the total homeless population were Aboriginal or Torres Strait Islander persons.23. Indigenous persons comprised approximately 16% of those requiring crisis and emergency housing under SAAP, and approximately 19% of those who were 'sleeping rough'.24 Previous work by Urbis25 has revealed that Indigenous people experience considerable difficulty in accessing private rental market housing (often due to discrimination) and accordingly are heavily reliant on public housing and crisis accommodation, which itself is in increasingly short supply.

Census data on homelessness can be problematic because it does not include people in institutions, such as aged care facilities or prisons. Further, if the criterion for homelessness is expanded to include those permanently housed in caravan parks, the total number of homeless people at the 2001 Census would increase to around 122,770.26

The important message emerging is that the 'visible homeless' on the streets and in impermanent locations scattered throughout Australia's cities represents only part of the problem. A much larger volume of people is also without permanent shelter, yet their plight is not as immediately obvious. For example Chamberlain and MacKenzie added a fourth category to the three traditional categories (primary, secondary and tertiary homelessness) to cover people who live in caravan parks on a long-term basis because it is the only housing they can afford.

[ top ]

Sectors of the homeless population

As the table below illustrates, since 2001 the numbers of homeless people living with friends and relatives and in boarding houses have decreased, while the proportion of homeless people living in SAAP accommodation and in improvised dwellings, tents and sleepers-out has increased.

In both 2001 and 2006, the highest numbers of homeless people were living with friends and relatives (45%) followed by boarding houses (20%).


Table Description
Table 2 - People in different sectors of the homeless population (Australia 2001 and 2006)27
      2001         %         2006         %         Difference in no. of homeless people
2001 - 2006 (%)
Improvised dwellings, tents and sleepers-out 14,158 14 16,375 16 +15.7
Friends and relatives 48 614 49 46, 856 45 -3.6
SAAP accommodation 14,251 14 19,849 19 +39.2
Boarding houses 22 877 23 21, 596 20 -5.6
Total 99,900 100 104,676 100 +4.8

[ top ]

Households

The following table reveals that there were more homeless households in 2006 compared with 2001 (74,825 and 74,281 respectively). The numbers of persons in each homelessness category did not change significantly between the 2001 and 2006 censuses. In both 2001 and 2006, a little more than three-quarters of the homeless population was single people.


Table Description
Table 3 - Homeless single person-, couple- and family-households (Australia)
  2001         %         2006         %         Difference 2001 - 2006 (%)
Single persons 58,116 78 57,182 76 -1.6
Couple only 9,420 13 10,160 14 +7.9
Family with children 6,745 9 7,483 10 +10.9
Total 74,281 100 74,825 100 +0.8
Age

Data from the 2006 Census shows that 58% of the homeless population was less than 35 years old. Furthermore, the highest proportion of homeless people was in the age bracket 12-18 years old (21%). It is clear that the age profile of the population is now much younger.

[ top ]

These issues are illustrated in Table 4, below.


Table Description
Table 4 - Age of people experiencing homelessness in 2006 (Australia)
Age                 Number                 %                
Less than 12 12,133 12
12-18 21,940 21
19-24 10,504 10
25-34 15,804 15
35-44 13,981 13
45-54 12,206 12
55-64 10,708 10
65+ 7,740 7
Total 104,676 100

[ top ]

Sex

In 2006, there were more male homeless people (56%) than female homeless people (44%) in Australia. The proportion of males and females in different homelessness sectors translated to:

In 2006, there were more males in all homelessness sectors except for in SAAP. The biggest difference between males and females was in the 'boarding house residents' category, which was strongly dominated by males.

Indigenous and non-Indigenous

Indigenous people are more likely to experience homelessness when compared with other Australians. In 2006, 2.4% of the Australian population was identified as Indigenous while 9.9% of the total homeless population was Aboriginal or Torres Strait Islander people. Indigenous people were overrepresented in all homelessness sectors, with SAAP clients being the most significant (17% of SAAP clients were of Aboriginal or Torres Strait Islander origin compared to an Australian population base of 2%).

[ top ]

1.3 Conclusions

This section highlights the difficulties inherent in obtaining an accurate estimate of the homeless population of Australia on any given night, due to challenges of counting a population which in many cases may be 'hidden' in private sector accommodation or staying with friends and family. Indigenous people in particular are considered to be undercounted in homelessness figures.

In addition to challenges in 'counting the homeless,' figures which are available through ABS Census data is also qualified in that it is based on a specific definition of homelessness, developed by the work of Chamberlain and MacKenzie.28 This definition differs from that utilised by the Supported Accommodation Assistance Act 1994 (SAA Act), which is based on a person's 'access to safe and secure housing' and which does not take into account the high proportion of the homeless who may be temporarily housed with friends or relatives or housed in caravan parks which offer no security of tenure, for example.

Aside from those who are 'sleeping rough', homeless people reside in a diverse range of accommodation, including staying with other households, the latter of which accounted for 45% of people counted as 'homeless' on Census night in 2006.

Taking into account these challenges, Census data for 2001 and 2006 paints a useful picture of the current composition of the homeless population and changes therein, along with the varying rates of homelessness across states and territories.

[ top ]

The data highlights the diversity of the homeless population with regard to population characteristics such as age, sex and household composition, for example. It also highlights the changing profile of this population, with current trends suggesting rising numbers of homeless children, families and older people and declining numbers of homeless teenagers - the latter trend which is attributed to the success of youth-focused early intervention services.29

Overall, national rates of homeless have remained relatively steady during the five years between the 2001 and 2006 censuses, both nationally and within individual states and territories.30 The 2006 Census data indicates that Queensland has the second highest number of homeless people after New South Wales. By comparison, the ACT, followed by Tasmania, have the lowest number of homeless people. The Northern Territory stands out as having a far higher rate of homelessness per head of population - more than five times the national average. This figure is explained by a range of factors including itinerant workers migrating to and living within the territory and also the high Indigenous population of this region.

Indigenous people are significantly over-represented amongst the homeless. According to 2006 Census figures, whilst just 2.4% of the national population was identified as Indigenous, 9.9% of the total homeless population was Aboriginal or Torres Strait Islander people.31. Indigenous people comprised an estimated 16% of those requiring crisis and emergency housing through SAAP accommodation, and approximately 19% of those who were 'sleeping rough'.32

The current profile of the homeless population highlights the sectors where attention must be focused in an effort to reduce homelessness rates. It also highlights the need to tailor appropriate approaches to particular population groups, such as families or Indigenous people, to achieve effective outcomes. The potential effectiveness of such tailored approaches is suggested by declining homelessness among teenagers during the five years to 2006.

Overall, the relative stability of homelessness rates across Australian states and territories between 2001 and 2006 must not be seen as a source of comfort, but rather an indication that more needs to be done to tackle homelessness with the aim of effecting a significant reduction in the homeless population.

[ top ]

  1. Australian Government, 2008, White Paper - The Road Home: A National Approach to Reducing Homelessness, p. 43
  2. Ibid., p.3
  3. Chamberlain, C., and MacKenzie, D., 1992, 'Understanding contemporary homelessness: Issues of definition and meaning', Australian Journal of Social Issues, vol. 27, issue 4, p.291 in Chamberlain C., and MacKenzie, D., 'Counting the Homeless, Australia', Australian Bureau of Statistics, Census 2006
  4. Australian Government, 2008, White Paper - The Road Home: A National Approach to Reducing Homelessness, p. 2
  5. Australian Government, Supported Accommodation Assistance Act 1994, Section 4(1), viewed at www.austlii.edu.au/au/legis/cth/consol_act/saaa1994359/s4.html
  6. Ibid., Section 4(2)
  7. Chamberlain C., and MacKenzie, D., 'Counting the homeless, Australia', Australian Bureau of Statistics, Census 2006
  8. Ibid.
  9. Australian Government Department of Health and Ageing, Homelessness and Mental Health Linkages: Review of National and International Literature, 2005
  10. Thompson, D., 'What do the published figures tell us about homelessness in Australia?' in the Australian Journal of Social Issues, vol. 32, no.3, 2007, p.358
  11. Ibid., p.364
  12. Urbis Keys Young, 1998, Homelessness in the Aboriginal and Torres Strait Islander Context and its Implications for the Supported Accommodation Assistance Program (SAAP), prepared for Family Services Branch Commonwealth Department of Health and Family Services; and, Urbis Keys Young, 2005, National Consultations on Aboriginal and Torres Strait Islander Homelessness, prepared for The Department of Family and Community Services
  13. Memmot, P., et al., Categories of Indigenous 'Homeless' people and good practice responses to their needs. 2003, Australian Housing and Urban Research Institute, in Clark, J., and D'Onise, K., No Pulgi, Nunkuwarrin Yunti - Urban Indigenous Homelessness and its Effect on Health, Report of the Sixth National Indigenous Homelessness Conference, Australian Government Department of Health and Ageing, accessed at: http://health.gov.au/internet/main/publishing.nsf/Content/natsieh-publicat.htm~natsienh-publicat-ch2.htm~natsienh-publicat-ch2-10.htm
  14. Thompson, D., 'What do the published figures tell us about homelessness in Australia?' in the Australian Journal of Social Issues, Vol. 32, No.3, 2007
  15. Chamberlain C., and MacKenzie, D., 'Counting the Homeless, Australia'' Australian Bureau of Statistics, Census 2006
  16. Chamberlain C., and MacKenzie, D., June 2009, 'Counting the Homeless, 2006, NSW', Australian Institute of Health and Welfare, p.6, Canberra: AIHW, viewed at www.aihw.gov.au/publications/hou/hou-204-10755/hou-204-10755-c00.pdf
  17. Ibid., p.27
  18. Chamberlain C., and MacKenzie, D., June 2009, 'Counting the Homeless, 2006, Northern Territory', Australian Institute of Health and Welfare, p.25, Canberra: AIHW, viewed at http://www.aihw.gov.au/publications/index.cfm/title/10761
  19. Thomson, C., 2006 (July 9), Kimberley is Australia's homeless capital, in WA Today, viewed athttp://www.watoday.com.au/wa-news/kimberley-is-australias-homeless-capital-20090708-dd7h.html
  20. Council to Homeless Persons, 2007, An Australian Blueprint to Reduce and Eliminate Homelessness, viewed at www.chp.org.au
  21. National Youth Commission, 2008, Australia's Homeless Youth: a report of the national youth commission inquiry into youth homelessness, viewed at www.nyc.net.au
  22. Australian Institute of Health and Welfare, 2006, Demand for SAAP assistance by homeless people 2003-04: A report from the SAAP National Data Collection, AIHW cat. no. HOU143, Canberra: AIHW
  23. Ibid.
  24. Urbis Keys Young, 1998, Homelessness in the Aboriginal and Torres Strait Islander Context and its Implications for the Supported Accommodation Assistance Program (SAAP), prepared for Family Services Branch Commonwealth Department of Health and Family Services; and, Urbis Keys Young (2005) National Consultations on Aboriginal and Torres Strait Islander Homelessness, prepared for The Department of Family and Community Services
  25. Australian Institute of Health and Welfare, 2005b, p. 318-325
  26. It should be noted that there is a question about whether ABS comparisons between 2001 and 2006 are reliable, or whether the differences may be an artefact linked to differences in counting rules, sampling differences or to differences in collection rates and other factors
  27. Ibid.
  28. Australian Government, 2008, White Paper - The Road Home: A National Approach to Reducing Homelessness, p. 2
  29. Ibid., p.27
  30. Australian Institute of Health and Welfare, 2006, Demand for SAAP assistance by homeless people 2003-04: A report from the SAAP National Data Collection, AIHW cat. no. HOU143, Canberra: AIHW
  31. Ibid.

[ top ]