2049.0 - Census of Population and Housing: Estimating homelessness, 2006  
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 11/09/2012   
   Page tools: Print Print Page Print all pages in this productPrint All

HOMELESSNESS STATISTICS

The ABS has developed a statistical definition of homelessness. Details of the definition can be found in Information Paper: A Statistical Definition of Homelessness (cat. no. 4922.0).

The ABS has also finalised the methodology for estimating homelessness using data from the Census of Population and Housing. Details of the methodology can be found in Information Paper - Methodology for Estimating Homelessness from the Census of Population and Housing (cat. no. 2049.0.55.001).

Below are commonly asked questions relating to homelessness statistics and answers to these questions.

The ABS has also prepared Factsheets on key areas of interest relating to homeless:


The ABS published final official homeless estimates from the 2006 and 2001 Censuses on Tuesday 11 September 2012 in the publication: Census of Population and Housing: Estimating Homelessness, 2006 (cat. no. 2049.0). Estimates from the 2011 Census will be published on 12 November under the same catalogue number.

QUESTIONS AND ANSWERS


WHAT IS THE ABS' NEW DEFINITION OF HOMELESSNESS?


The ABS has never before had a definition of homelessness.

In brief, the ABS statistical definition is that:

When a person does not have suitable accommodation alternatives they are considered homeless if their current living arrangement:

    • is in a dwelling that is inadequate; or
    • has no tenure, or if their initial tenure is short and not extendable; or
    • does not allow them to have control of, and access to space for social relations.

The ABS definition of homelessness is informed by an understanding of homelessness as 'home'lessness, not rooflessness. It emphasises the core elements of 'home' in Anglo American and European interpretations of the meaning of home as identified in research evidence (Mallett, 2004). These elements may include: a sense of security, stability, privacy, safety, and the ability to control living space. Homelessness is therefore a lack of one or more of the elements that represent 'home'.

The definition has been constructed from a conceptual framework centred around the following elements:

    • Adequacy of the dwelling;
    • Security of tenure in the dwelling; and
    • Control of, and access to space for social relations.
See Information Paper: A Statistical Definition of Homelessness (cat. no. 4922.0) for more information on the ABS' definition of homelessness.

WHY DID THE ABS DEVELOP A DEFINITION OF HOMELESSNESS?


The ABS has never before had a definition of homelessness. However, ABS did recognise the importance of homelessness for Australian governments and society more generally, and in 2009 commenced development of modules for the collection of past periods of homelessness in its household surveys. This development, coupled with the potential to develop official estimates of homelessness using data from the Census of Population and Housing, required a conceptual definition of homelessness to underpin operationalisation of the measurement of homelessness on a broadly consistent basis across these emerging datasets. Not only do comparable quality statistics, over time and across data sources, require a clear conceptual framework and definition to underpin their operationalisation in multiple collections, but the conceptual clarity guides ongoing fine tuning of those datasets to better serve that measurement purpose.

The ABS definition of homelessness was developed in consultation with the ABS' Homelessness Statistics Reference Group. Future ABS surveys will adopt this definition.
WHY DIDN'T THE ABS USE CHAMBERLAIN AND MACKENZIE'S 'CULTURAL' DEFINITION OF HOMELESSNESS?


Key international and national definitions of homelessness (including Chamberlain and MacKenzie's 'cultural' definition) must demonstrate that a rigorous, evidence informed, process had been undertaken to develop the conceptual framework /definition with core elements underpinning it. The purpose of the definition is to advise on the conceptual elements of a definition that could be operationalised across a range of datasets, and not just be a summary method of developing a view of homelessness from the data currently collected currently in the Census of Population and Housing. There has been no empirical validation of the 'cultural definition' in terms of its assumptions about a minimum shared community standard. Moreover, views have evolved over the past 20 years or so, suggesting that this standard is historically contingent.

The new ABS definition of homelessness was tested against the categorisation of a large list of living situations and addresses much of the ambiguity in the operationalisation of Chamberlain and MacKenzie's 'cultural' definition.
WHAT PROCESS DID THE ABS GO THROUGH TO DEVELOP THEIR DEFINITION OF HOMELESSNESS?


The ABS established the Homelessness Statistics Reference Group (HSRG) to advise the ABS on the development, collection, compilation, production and dissemination of robust statistics for use in analysing, understanding and reporting on homelessness in Australia. HSRG is made up of key stakeholders from state and federal governments, peak bodies, service providers and researchers. The ABS convened a HSRG sub-group to advise HSRG and ABS on the key elements of a conceptual framework that could underpin the ABS definition of homelessness for use in compiling official statistics in Australia. The work of the definition sub-group was framed by tight timelines linked to specified ABS release dates for ABS official estimates of homelessness from the 2001, 2006 and 2011 Censuses of Population and Housing. Because of the tight timelines the work of the sub-group was divided into two phases. The first phase, covered in the Information Paper: A Statistical Definition of Homelessness (cat. no. 4922.0), does not address any specific cultural definition issues relating to Aboriginal and Torres Strait Islander peoples understandings of homelessness. The second phase, relating to Aboriginal and Torres Strait Islander homelessness, commenced in August 2012.

The Definition subgroup first met in December 2011. To complete its first phase work it met on eight occasions, including three all day face to face meetings and five teleconferences between December 2011 and June 2012.

The subgroup developed its thinking about the key elements of the definition through both a top down and a bottom up approach including:
    • consideration of national and international statistical definitions of homelessness (summarised in a separate paper on the ABS statistical definition of homelessness); and
    • review and analysis of over 300 housing and personal circumstances in Australia to test whether homelessness, as understood by researchers, service practitioners and policy and program officers, could be mapped to a basic set of key elements.

At key points in this process subgroup members were invited to seek broader stakeholder feedback about the various living situations under review and the implications for key elements of a homelessness definition.

Following critical analysis of the definitions and living situations ABS articulated a range of possible elements to support a definition of homelessness. Members discussed the pros and cons of these elements, and refined the elements over time. While consensus was not reached in the sub-group, the ABS definition largely reflects a majority view of members of the sub-group.
HOW WILL THE ABS USE THIS NEW DEFINITION OF HOMELESSNESS?


The ABS definition of homelessness will be used to produce statistics on homelessness from a range of ABS collections. This includes prevalence estimates of homelessness from the five-yearly Census of Population and Housing, and from household surveys such as the General Social Survey, Survey of Disability, Ageing and Carers, Personal Safety Survey, National Aboriginal and Torres Strait Islander Social Survey, and other surveys, as appropriate.

This definition can also be used by other organisations to collect and output their statistics in line with the ABS definition and ABS statistical outputs.
HOW DOES THE DEFINITION OF HOMELESSNESS RELATE TO THE METHODOLOGY FOR ESTIMATING HOMELESSNESS FROM THE CENSUS?


Comparable quality statistics over time and across data sources, require a clear conceptual framework and definition that underpins operationalisation of that definition in multiple collections, including fine tuning those datasets for that purpose.

The methodology for use in the Census can only partially operationalise the definition because although the Census is designed for many purposes, it is limited in the nature of the questions it can ask that will reflect on homelessness. While the information derivable from the Census for homelessness measurement will improve over time, nevertheless some proxies will always need to be developed for some elements of the definition for some of the different homeless operational groups that can be output from the Census. Details on how each of the homeless groups relate to the definition can be found in Information Paper - Methodology for Estimating Homelessness from the Census of Population and Housing (cat. no. 2049.0.55.001).WHY DID THE ABS DEVELOP ITS OWN OFFICIAL METHODOLOGY FOR ESTIMATING HOMELESSNESS FROM THE CENSUS?

In 2008 following wide-spread discussion in Australia about the meaning and measurement of social inclusion and exclusion, the ABS recognised the need to develop robust and transparent homelessness statistics across a range of ABS datasets. This decision coincided with the release of the Federal Government White Paper on Homelessness (The Road Home), which highlighted homelessness as an important social issue in Australia and identified the need to "turn of the tap" , "break the cycle" and arrest chronic homelessness.

To this point the ABS did not provide official estimates of homelessness through the Census. Estimates were produced by academics, Professors Chamberlain and Mackenzie, who estimated the numbers of homeless people in Australia using the 1996, 2001 and 2006 Census of Population and Housing (Chamberlain, 1999; Chamberlain & MacKenzie, 2003; Chamberlain & MacKenzie, 2008). This estimation work was underpinned by the cultural definition of homelessness developed by Chamberlain and Mackenzie (Chamberlain & MacKenzie, 2008).

Following the decision to develop official ABS homelessness statistics, the ABS began developmental work in this area by first reviewing the methodology employed by Chamberlain and Mackenzie to estimate homelessness through the Census of Population and Housing. During this review the ABS identified the need to develop a robust, defensible and evidence informed definition of homelessness for statistical purposes. The ABS began this work by reviewing key national (SAAP and Cultural) and international (ETHOS and NZ statistics) definitions.

A key outcome of the review, and the homelessness sector consultation that followed, was the establishment of the Homelessness Statistics Reference Group (HSRG). The HSRG was established to provide advice to the ABS about the development of homelessness statistics. A definition sub-group of the HSRG was established to inform the work on the homelessness definition.
WHAT PROCESS DID THE ABS GO THROUGH TO DEVELOP THEIR METHODOLOGY OF HOMELESSNESS FROM THE CENSUS?


ABS initiated a methodological review of Chamberlain and MacKenzie's Counting the Homeless (ABS cat. no. 2050.0) by engaging with a range of stakeholders, including researchers and the homelessness services sector, and with the advice of a Steering Committee comprising representatives from the Australian Government Department of Families, Housing, Community Services and Indigenous Affairs, the Australian Institute of Health and Welfare, and from three states (New South Wales, Queensland and South Australia) represented on the inter-jurisdictional Housing and Homelessness Information Management Group (reporting to the Housing Ministers' Advisory Committee).

Following an initiating review workshop on 21 October 2009, with Professors Chamberlain and MacKenzie, and representatives from Homelessness Australia, as well as from Commonwealth, state/territory and local government organisations, the nature of the ABS’s concerns with the Counting the Homeless 2006 methodology were outlined in Issues in estimating the number of homeless in Australia: A paper to inform a review of Counting the Homeless methodology. This paper was made available on the National Homelessness Information Clearinghouse website in October 2009, and submissions were sought. Submissions were received from government organisations, academics and eight homelessness services sector organisations. Workshops to progress the review, which involved Professors Chamberlain and MacKenzie, were held in May 2010 and October 2010.

ABS's initial findings from the methodological review were published on 31 March 2011 in the Discussion Paper: Methodological Review of Counting the Homeless, 2006 (ABS cat. no. 2050.0.55.001). That Discussion Paper announced a public submissions process and a series of advertised public forums in each capital city. The Discussion Paper noted the importance of the issue of homelessness for society and governments, and the need for quality data for decision making purposes, particularly for measuring change over time. In that context, the Discussion Paper described a methodology that had been previously used, and proposed a range of methodological changes that would be needed before consistent, transparent and repeatable official estimates could be made of the number of people enumerated in the Census who were likely to have been homeless on Census night.

To maximise exposure to the review findings, the advertised public forum details were also emailed directly to many stakeholders, and the forums were held in each capital city through April and May 2011. Over 150 people attended the public forums, and many more attended sector or jurisdictional specific discussions with the ABS. The ABS was also invited to participate in discussions about the review findings in a range of meetings. There were 35 written submissions in response to the Discussion paper.

Consultation on the review findings confirmed the Discussion Paper's emphasis on the significance of the areas of likely underestimation of homelessness in relation to youth, Aboriginal and Torres Strait Islander people and people escaping domestic and/or family violence. However, no alternative estimation methods were proposed that could address the issues of underestimation for homelessness in these groups.

As a result of the consultation, the ABS released Position Paper - ABS Review of Counting the Homeless Methodology, Aug 2011 (cat. no. 2050.0.55.002) in August 2011 that outlined
-the key themes identified from the consultation and the submissions received;
    • addressed some of the issues raised in the consultation;
    • reiterated and further articulated the concerns expressed in the Discussion Paper about the underestimation of key groups of homeless people;
    • noted the establishment of the Homelessness Statistics Reference Group; and
    • described future developments in homelessness estimation, including but not limited to a proposed quality study about homeless school students.

A key outcome of the review, and the homelessness sector consultation that followed, was the establishment of the Homelessness Statistics Reference Group (HSRG). The HSRG was established to provide advice to the ABS about the development of homelessness statistics. A methodology sub-group of the HSRG was established to inform the ABS' work on the official estimates of homelessness from the Census.

The purpose of the sub-group was to advise the HSRG and the ABS on the methodology used for estimating the prevalence of homelessness using data from the Censuses of Population and Housing in 2001, 2006 and 2011, taking into consideration the limitations of the Census and the ABS definition of homelessness. The work of the definition sub-group was framed by tight timelines linked to specified ABS release dates for ABS official estimates of homelessness from the 2001, 2006 and 2011 Censuses of Population and Housing.

The Methodology subgroup first met in April 2012. To complete its work it met on six occasions including one two day face to face meeting and five teleconferences between April 2012 and July 2012.

The subgroup developed its thinking about the methodology by reviewing and analysing each of the individual homeless operational groups including consideration of:
    • the wording and intent of the Census questions;
    • the field procedures of the Census;
    • the limitations of the Census for enumerating some population groups;
    • the limitations of the Census data for estimating some population groups;

At key points in this process subgroup members were invited to seek broader stakeholder feedback about the many living conditions that may constitute homelessness and their implications for key aspects of the methodology.

The ABS considered the input it received from members in refining the methodology for official estimates of homelessness from the Census, and undertook additional analysis to further consider the recommendations made by members. Balanced with member suggestions was how the methodology: needed to be aligned with the new ABS definition of homelessness; reflected the wording and intent of the Census questions and field procedures; and that the methodology was consistent and repeatable across Censuses.
HOW DOES THE FINAL METHODOLOGY FOR ESTIMATING OFFICIAL ESTIMATES OF HOMELESSNESS FROM THE CENSUS COMPARE TO THE REVIEW METHODOLOGY?


The bulk of the Review methodology of estimates of homelessness from the Census has been carried forward into the official methodology.

The following changes were made to the Review methodology:
    1. The removal of imputed estimates, as appropriate, throughout the methodology;
    2. The removal from the homeless grouping 'Persons in temporary lodging' of people who were in those types of institutions on Census night that are specific exclusions from the official definition of homelessness (as there excluded from the 'cultural definition'; and
    3. The addition of a new homeless operational group: 'Persons living in severely crowded dwellings'.

In addition, the ABS presents, alongside the homeless estimates, three additional categories of marginal housing that can be identified as homeless:
    • Persons living in other crowded dwellings;
    • Persons in other improvised dwellings; and
    • Persons marginally housed in caravan parks.
These categories of living arrangements provide some insight into those who may be at risk of homelessness.

The reasons for the changes to the methodology were:
    1. There is no evidence that many of the people who were imputed in homeless situations were homeless. See Information Paper - Methodology for Estimating Homelessness from the Census of Population and Housing (cat. no. 2049.0.55.001) for more details on both which imputed records were incorporated in the official homelessness estimates, and the reasons for the exclusion of others.
    2. The new ABS definition of homelessness outlined specific exclusions from homelessness as people may have chosen to live in these circumstances and have accommodation alternatives; or they are required by law to live in these circumstances; or they are in acceptable temporary living arrangements (such as student halls of residence); or or they are living in temporary arrangements that are essential for their broader health and wellbeing. The specific exclusions include: people confined in prisons, detention centres and other institutions such as juvenile correctional facilities or hospitals; students living in halls of residence; and members of religious orders such as monks and nuns living in seminaries and nunneries and similar establishments. The category: 'Persons in temporary lodging' in the Review estimates included some people in these circumstances which have been removed in the final estimates.
    3. Under a definition of homelessness that is centred around the concept of 'home' those people who are living in severely crowded dwellings meet the definition of homeless. International definitions, and key stakeholder feedback confirmed this dimension of homeless.
WHEN WILL THE 2006 AND 2001 ESTIMATES OF HOMELESSNESS FROM THE CENSUS BE PUBLISHED?

The ABS published final official homeless estimates from the 2006 and 2001 Censuses on Tuesday 11 September 2012 in the publication: Census of Population and Housing: Estimating Homelessness, 2006 (cat. no. 2049.0). Estimates from the 2011 Census will be published in November under the same catalogue number.
WHEN WILL THE 2011 ESTIMATES OF HOMELESSNESS FROM THE CENSUS BE PUBLISHED?


The ABS will be publishing estimates from the 2011 Census on Monday 12 November 2012 in the publication: Census of Population and Housing: Estimating Homelessness, 2006 (cat. no. 2049.0). The methodology uses Census second release variables, which will be published on 30 October 2012.

The same ABS methodology will be used for the 2011 as the 2006 estimates, aside from any changes required due to the changes in estimation cut offs applied when considering income, mortgage and rent (see Key differences between methodology employed for estimating homelessness for different Census years - Income, mortgage and rent cut offs in Information Paper - Methodology for Estimating Homelessness from the Census of Population and Housing (cat. no. 2049.0.55.001) for more details) and any changes that result from changes in the field procedures carried out in 2011. Any such changes will be outlined in the 2011 publication.
WHAT OTHER DATA CAN I USE TO HELP ME TO UNDERSTAND HOMELESSNESS?


The ABS will publish an Information Paper: Guide to Homelessness Statistics (cat. no. 4923.0) in November 2012 to assist users with analysing the multiple data sources available to obtain a more complete picture of homelessness. The guide will outline which parts of the homeless definition ABS collections can, or cannot capture.

In addition to prevalence estimates of homelessness from the five-yearly Census of Population and Housing, the ABS has collected previous experiences of homelessness from the 2010 General Social Survey (GSS), published in September 2011. In March 2012, the ABS released an article titled Life after homelessness in the publication Australian Social Trends (cat. no. 4102.0) drawing on those GSS results. The article examines a range of socio-economic indicators of those who had experienced at least one episode of homelessness in the 10 years prior to the survey, but were no longer homeless.

An improved homelessness module has also been included in the 2012 Survey of Disability, Ageing and Carers. The ABS expects to also include this module, further developed, in the 2014 General Social Survey. The ABS will consider developing a culturally appropriate module for the National Aboriginal and Torres Strait Islander Social Survey. The ABS will also be collecting experiences of homelessness for people who leave a violent partner through the Personal Safety Survey.

The ABS will also investigate using the 5% Statistical Longitudinal Census Dataset (SLCD) to undertake longitudinal analysis of the circumstances of those who have been identified as likely to be homeless. The circumstances of people identified as likely to be homeless on the 2011 SLCD can then be compared with their circumstances in 2006, and into the future it should be possible to report on repeat periods of homelessness and long term outcomes as seen in the SLCD. It will also be possible to compare these results, for those likely to be homeless, with the rest of the population.

There are also non-ABS sources of information about homelessness, such as the Australian Institute of Health and Welfare (AIHW) Specialist Homeless Services collection, and the FaHCSIA funded Melbourne Institute Study: Journeys Home: Longitudinal Study of Factors Affecting Housing Stability.
WHAT IS THE DIFFERENCE BETWEEN ENUMERATION AND ESTIMATION AND HOW DO THESE EFFECT ESTIMATING HOMELESSNESS FROM THE CENSUS?


In summary, the main limitations with using Census data to estimate homelessness can be summarised as:
    • under / over estimation - people were enumerated in the Census but the data collected about them is not sufficient to be certain about whether or not they were homeless on Census night;
    • under-enumeration - people who were not enumerated in the Census.

Observing homeless people in any data collection is a challenge, and their homeless circumstance may mean that these people are not captured at all in datasets used to count people generally. And not all homeless people will be enumerated in data sets of those homeless people accessing particular services for the homeless. The 2010 ABS General Social Survey 2010 found that of those who had had an experience of homelessness in the last ten years and who were no longer homeless at the time of interview, only 40% had sought assistance of formal services. While data on people who access services are very important data sources for understanding people who access services, they cannot provide an estimate of the total number of homeless people at one point in time. Only the Census offers the opportunity to estimate for most people who were likely to have been homeless at one point in time.

However, there is an inherent imprecision in estimating homelessness using the Census of Population and Housing because the Census is not designed to classify people according to whether or not they were homeless on Census night. Variables collected in the Census that were designed for other purposes must be interpreted as proxies for likely homelessness. The ABS methodology includes in its homelessness estimates groups of people who, on balance, were most likely to have been homeless on Census night.

While it may be tempting to overestimate homelessness in some groups to compensate for both under-enumeration and likely under-estimation for some other groups, such an approach would result in estimates of characteristics that did not reflect those of the homeless population, including but not limited to their geographic spread. This may result in the misdirection of policy, funds and services. And while a balance between unavoidable under-estimation and deliberate over-estimation may result, this is unlikely. It is also very likely to be very different with each Census, destroying the capacity to monitor change over time. Recognising which groups of homeless people are underestimated in the Census, and using supplementary data sources to understand these groups, will both better address the needs of homeless people, and allow for assessments of change over time in the level of homelessness.

ABS recognises that some groups of people are more likely to be under enumerated in the Census. Aboriginal and Torres Strait Islander people are more likely to be both under enumerated and over-represented in the homeless population. ABS has developed strategies for each Census aimed at maximising the enumeration of Indigenous persons.

So called rough sleepers and people staying in supported accommodation for the homeless are also at risk of being under enumerated in the Census. The ABS develops a homeless enumeration strategy for each Census, and works with homeless service providers to maximise the enumeration of these groups on Census night.

The ABS Post Enumeration Survey (PES) is used to estimate for the under enumeration of the Australian population in the Census. However it only covers people in private dwellings at the time of the PES, and therefore will not capture homeless people living in non-private dwellings but who were missed on Census night.
IF THE CENSUS UNDER-ENUMERATES AND UNDER-ESTIMATES THE NUMBER OF HOMELESS PEOPLE, WHY DO THE ABS REMOVE ANY OVERESTIMATES OF HOMELESSNESS?


The ABS methodology for estimating homelessness from the Census ensures that over-estimation is minimised. Such over-estimation has the potential to obscure the true size of the homeless population, mask where homeless people are located, misrepresent the characteristics of those who are homeless, and prevent valid measurement of progress over time.

Overall, estimating homelessness from the Census is done by identifying those groups who, on balance, are most likely to be homeless based on a number of characteristics that identify different 'elements' of the definition of homelessness. These estimates cannot include those people who were never enumerated in the Census.

While it may be tempting to overestimate homelessness in some groups to compensate for both under-enumeration and likely under-estimation for some other groups, such an approach would result in estimates of characteristics that did not reflect those of the homeless population, including but not limited to their geographic spread. This may result in the misdirection of policy, funds and services. And while a balance between unavoidable under-estimation and deliberate over-estimation may result, this is unlikely. It is also very likely to be very different with each Census, destroying the capacity to monitor change over time. Recognising which groups of homeless people are underestimated in the Census, and using supplementary data sources to understand these groups, will both better address the needs of homeless people, and allow for assessments of change over time in the level of homelessness.
WHICH GROUPS ARE MOST LIKELY TO BE UNDERESTIMATED IN THE ESTIMATES OF HOMELESSNESS FROM THE CENSUS?


The ABS has previously identified in its publications: Discussion Paper: Methodological Review of Counting the Homeless, 2006 (cat. no. 2050.0.55.001) and Position Paper - ABS Review of Counting the Homeless Methodology, August 2011 (cat. no. 2050.0.55.001) that some key population groups will be underestimated in estimates of homelessness from the Census: homeless youth, homeless Aboriginal and Torres Strait Islanders, and people fleeing domestic and/or family violence.

To find out more information about these groups, see the following fact sheets:
THE ESTIMATE OF YOUTH IS LOWER THAN WAS FOUND IN CHAMBERLAIN AND MACKENZIE'S METHODOLOGY FOR ESTIMATING HOMELESSNESS. WHY IS THAT?


The ABS has previously identified in its publications: Discussion Paper: Methodological Review of Counting the Homeless, 2006 (cat. no. 2050.0.55.001) and Position Paper - ABS Review of Counting the Homeless Methodology, August 2011 (cat. no. 2050.0.55.001) that youth homeless will be underestimated from the Census. Chamberlain and MacKenzie used their collection and a range of assumptions to estimate the number of homeless youth who would have been staying with friends and relatives on Census night. ABS has determined that the methodology used is not robust for use in official estimates of homelessness. See below for more information.



Youth (sometimes referred to as 12-18 years or 12-24 years) who are homeless and 'couch surfing', but for whom a usual residence is reported in the Census, are masked because their characteristics look no different to other youth who are not homeless but visiting on Census night. A usual address may be reported either because the young person doesn't want to disclose to the people they are staying with that they are unable to go home, or the person who fills out the Census form on behalf of the young person assumes that the youth will return to their home and reports the youth's previous address. Homeless youth will be underestimated with the group: 'Persons staying temporarily with other households'. Analysis of Census data shows that estimates of the level of youth homelessness from the Census are likely to be low, but the scale of any hidden youth homelessness cannot be assessed with any currently available data.

Some commentators reflect on the numbers of homeless youth who contact services over time and use this to gauge the scale of youth homelessness. However, evidence from the former Supported Accommodation Assistance Program statistics shows that the number of youth who experience homelessness over a year is a high multiple of the number who are homeless at a point in time (the Census prevalence measure). There is likely to be an even higher number of approaches to services than the numbers assisted over time. So far, none of these service related data sources provides a fix on the prevalence of homelessness and ABS surveys still provide the best option for a retrospective view through understanding past homelessness.

ABS has not yet been able to implement any reliable way of estimating current homelessness among youth staying with other households and for whom a usual address is reported in the Census. Service providers and researchers indicate that the low number of homeless youth staying with other households does not concord with their knowledge about youth homelessness.

Guided by its Homelessness Statistics Reference Group, the ABS is continuing to undertake research and development to improve the estimation of homelessness, including youth homelessness. In particular, the ABS has been undertaking a quality study to inform the potential development of a nationally representative homeless school students survey.

Until a robust methodology is developed to measure the level of youth homelessness, ABS will focus on producing transparent, consistent and repeatable estimates that can be used to monitor change over time. Because the ABS methods are transparent, users can assess whether there is any evidence to suggest that the components of homelessness that cannot yet be estimated reliably are likely to be moving differently over time to those elements that can be measured.

For analysis on youth homelessness from the Census see the ABS publications: Discussion Paper: Methodological Review of Counting the Homeless, 2006 (cat. no. 2050.0.55.001) and Position Paper - ABS Review of Counting the Homeless Methodology, August 2011 (cat. no. 2050.0.55.001). These papers include comparisons with the Chamberlain and MacKenzie methods.

In their analysis, Chamberlain and MacKenzie classified 6,378 youth aged 12 to 18 years old who were enumerated in the Census as homeless. However, a separate estimate of youth homelessness was derived by Chamberlain and MacKenzie (21,940 people) using sources other than the Census. The main source that was used was their National Census of Homeless School Students (NCHSS), adjusted using assumptions about the ratio of school to non-school homeless youth aged 12 to 18 years. In their estimate the difference between the Chamberlain and MacKenzie Census measure and their separate estimate is assumed to relate to those youth who may be homeless who were staying with another household on Census night but had a usual address reported for them. They conclude "…we think the missing 15,562 are hidden within this category". Pages 18 to 20 explains this methodology in more detail.

ABS has concluded that the Chamberlain and MacKenzie estimate of 21,940 homeless youth is not sufficiently robust to inform on the numbers of 12-18 year old youth who were experiencing homelessness on Census night. The Chamberlain and MacKenzie estimate is based on: school reports which capture youth homelessness over a week rather than on Census night; an undercount adjustment methodology that is inconsistent between Censuses; and it applies an extrapolation to the non-school youth population that is both inconsistent with the stated methodology and which overstates the estimate. See pages 46 to 48 in the ABS' Discussion paper for more information.

In analysing the characteristics of 12-18 year olds who had a usual address reported and were visitors on Census night, no characteristics have yet been identified that differentiate between those who were homeless and those who were visiting for other reasons. Additional analysis on the characteristics of the possible relationship status of the 39,966 12-18 year old usual residents visiting on Census night was provided in the Discussion Paper to try to understand any characteristics that can be used to distinguish potential homeless groups. This analysis used age, sex and Census District of usual residence to attempt to understand possible family units travelling together.