Estimating Homelessness: Census methodology

Latest release
Reference period

This release provides estimates of the prevalence of homelessness from the 2021 Census of Population and Housing. As homelessness is not a characteristic that is directly measured in the Census, estimates of those experiencing homelessness are derived using analytical techniques based on the characteristics observed in the Census and statistical assumptions. 

Definition of homelessness

Estimates are based on the Australian Bureau of Statistics (ABS) Statistical definition of Homelessness: when a person does not have suitable accommodation alternatives, they are considered as experiencing homelessness if their current living arrangement:

  • is in a dwelling that is inadequate
  • 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 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.

For more information see Information Paper - A Statistical Definition of Homelessness.

The Parliamentary Inquiry into homelessness in Australia, conducted during 2020 and 2021, recommended the ABS review its statistical definition of homelessness, with particular focus on the circumstances in which people living in severely crowded dwellings and boarding houses should be categorised as homeless. The ABS supports a review of the definition of homelessness and plans to consider the Inquiry recommendations ahead of the 2026 Census. 

Data collection

Scope of the Census and homelessness estimates

The 2021 Census of Population and Housing aimed to count every person in Australia on Census night, 10 August 2021.

The Census includes:

  • visitors to Australia (regardless of how long they have been or plan to be in the country)
  • people in the six states, the Northern Territory, the Australian Capital Territory, Jervis Bay Territory, and the Territories of Christmas Island and Cocos (Keeling) Islands, Antarctica and Norfolk Island.

The Census excludes:

  • foreign diplomats and their families
  • Australian residents out of the country on Census night
  • Australian external territories (minor islands such as Heard Island and McDonald Island). 

While overseas visitors, and people in offshore, shipping, or migratory regions are included in the Census, they are excluded from the homelessness estimates.

Census data collection

Census data collection is outlined in detail in the following Census information pages:

Homelessness enumeration

A Homelessness Enumeration Strategy was used to ensure that all people experiencing homelessness could participate in the Census. The Strategy included additional collection and support mechanisms for people living in three broad situations on Census night:

  • ‘not in a dwelling’ (i.e. ‘people living in improvised dwellings, tents or sleeping out’ also known as ‘people sleeping rough’)
  • ‘in a private dwelling’ (i.e. ‘people staying temporarily with other households’ or ‘people living in 'severely' crowded dwellings’)
  • ‘in a non-private dwelling’ (i.e. ‘people in supported accommodation for the homeless’, ‘people living in boarding houses’ or ‘people in other temporary lodging’).

The Enumeration Strategy acknowledged the different experiences of people in each living situation and identified ways to maximise the quality of the homelessness count. Its success relied on the Homelessness Engagement Strategy, which focussed on early engagement and strong relationships with all levels of government, peak homelessness bodies and a wide range of service providers. These strong relationships assisted in the development and execution of the Enumeration Strategy including recruitment of field staff from the homelessness sector or with lived experience of homelessness; a key factor in a successful count.

Specific recruitment activities were undertaken to help recruit Aboriginal and Torres Strait Islander field staff whenever possible. These staff provided tailored communication to promote the shared benefits of the Census and encourage community-led action. Recruitment of Census engagement staff directly from Aboriginal and Torres Strait Islander communities meant that Census forms could be delivered and completed with the assistance of trusted members of local communities.

For more information on the inclusive strategies undertaken during the Census to enumerate people experiencing homelessness see Story 3: An inclusive Census.

Improvements in the 2021 Census homelessness enumeration

The following improvements were made to the 2021 Census that improved the quality of the estimates of people experiencing homelessness:

  • increased capacity of the online household form to collect information on up to 25 people, up from 10 people in 2016, making it easier for overcrowded households to fully participate.
  • adding extra information to the ‘front of form’ on: the importance of completing the Census and including everyone present in the household, including all visitors; how to request a Personal Form for people who preferred to complete their Census separately to the rest of the household; and the importance of including people who were experiencing homelessness and staying temporarily in the household.
  • revising the wording in the ‘usual address’ question in the Census Household and Personal Forms to list family conflict or eviction as reasons to report ‘None’ in the Suburb/Locality box, especially for people who were experiencing homelessness and staying temporarily with another household.

The ABS also increased the use of administrative lists from governments and homeless service providers to identify the location and nature of the homeless population. This included:

  • the Victorian rooming house register was used for the first time in 2021 Census to identify the location of boarding houses likely to provide accommodation for the homeless in Victoria.
  • several state and territory governments provided address lists of temporary accommodation (such as hotel and motels) where people stayed under brokerage or vouchers. This information helped allocate Census resources during enumeration, and in quality assuring Census forms when they were received. This improved the enumeration of people in other temporary lodgings and was particularly important given a number of state and territory governments provided additional brokerage funding to assist people to access emergency accommodation in hotels/motels during the COVID-19 pandemic.
  • more comprehensive lists of supported accommodation were received from each state and territory government. The information was received earlier than in 2016, which allowed for earlier engagement with the homelessness services to improve enumeration of people in supported accommodation for the homeless.

The improvements in data quality through greater use of administrative lists mean that estimates from the 2021 Census for people living in boarding houses in Victoria and people in other temporary lodgings nationally are not directly comparable with previous Censuses.

People not in a dwelling

As in previous censuses, the planned collection period for people in improvised dwellings, tents of sleeping out was over six days commencing from Census night (Tuesday 10 to Sunday 15 August 2021). However, adjustments to the timing were made in areas that were affected by COVID-19 restrictions, or when key services for people experiencing homelessness in a location operated only on a Monday. This population group were interviewed by specialist staff in locations where services are provided for people experiencing homelessness. The recruitment of staff from homelessness services to fill the specialist Community Field Officer roles was designed to ensure field staff could use their expertise to ensure an accurate count.

In most circumstances, people not in a dwelling were enumerated by interview using the paper Special Short Form (SSF) as was done in previous Censuses. The SSF is a shortened version of the Census Household Form (HHF) and is designed for the enumeration of people sleeping rough. The form enabled the collection of key characteristics in an efficient manner when a field staff member is needing to interview multiple people within a short window of time. This was the primary collection method for New South Wales, Victoria, Tasmania, South Australia and southern parts of Western Australia and Queensland.

In some cases, Census Field Officers interviewed people not in a dwelling using the paper or online Census Household Form (HHF). The use of the HHF allowed field staff to ask questions, including where a person usually lives, to help determine if the person was homeless or not. This was the primary collection method to enumerate people who were not in a dwelling in Northern Territory, and northern parts of Western Australia and Queensland. 

The ABS undertook a quality assurance process at the end of enumeration to remove any duplicate forms, given the mobility of some people experiencing homelessness and the extended enumeration period.

People in a private dwelling

Most people in private dwellings completed their Census using either the paper or online HHF. This includes people staying temporarily with other households and those who were living in 'severely' crowded dwellings.

The ABS worked with service providers who support youth to share communication and engagement materials that encouraged participation in the Census, and to specify ’None’ as their place of usual residence if they were experiencing homelessness and staying temporarily with other households.

People in a non-private dwelling

There are a range of non-private establishments that provide shelter and accommodation for people experiencing homelessness. These include:

  • crisis accommodation, shelters, and refuges,
  • short, medium, and long term supported accommodation,
  • registered and unregistered boarding houses; and
  • hotels and motels, under brokerage or vouchers.

In many cases the locations of accommodation for the homeless are not easily identifiable and the location of services may change location over time. As in previous Censuses, in 2021 the ABS used address lists of supported accommodation and boarding houses to supplement the information on the ABS Address Register. The lists of addresses were provided by government authorities, local councils and individual service providers.

The Census enumeration approach for non-private dwellings was determined by the nature of the service provided, including:

  • those that are publicly known, such as hostels for the homeless and registered boarding houses. People staying in these were enumerated like other types of non-private dwellings such as hospitals, and aged care.
  • those that were sensitive in nature, where the location and its status as accommodation for those experiencing homelessness were not known to the public. These establishments, as well as refuges for females and youth, were discreetly enumerated by specialist Census staff who had restricted access to Homelessness List information. Field staff were not advised of the location of these dwellings.
  • those that accepted brokerage or vouchers. These are mainly hotel/motel establishments where some guests are brokered to stay temporarily by homelessness service providers due to the lack of crisis accommodation elsewhere. People staying in these establishments were enumerated as per the standard procedures for these dwellings.

The ABS asked people who were in crisis accommodation to write 'none-crisis' instead of just 'none' as their place of usual residence, with the message passed on through homelessness service providers. People who reported 'none-crisis' as their place of usual residence and were enumerated in a private dwelling were estimated to be in supported accommodation, and if enumerated in a hotel or motel were estimated to be in other temporary lodgings. 

Impacts of the COVID-19 pandemic

State and territory government responses to the pandemic

State and territory governments introduced a number of responses to assist people experiencing homelessness during the pandemic. The responses, and the extent of their implementation, varied within and across jurisdictions. Measures included:

  • additional brokerage funding to assist people to access emergency accommodation in hotels/motels
  • additional short-term or emergency accommodation and additional outreach for people experiencing homelessness needing to apply physical distancing in shelters, self-isolation, or quarantine, especially during the lockdown periods
  • additional programs and accommodation for women with children at risk of, or experiencing homelessness, from family and domestic violence
  • supporting safe housing exits for individuals and families who received short-term hotel/motel accommodation during the pandemic.

The Northern Territory Government’s Return to Country program facilitated Aboriginal and Torres Strait Islander people returning to their home communities from urban regional centres.

A number of these measures would have temporarily affected peoples’ accommodation circumstances at the time of the Census, and will be reflected in the estimates presented in this release.

COVID-19 impacts on homelessness enumeration

The 2021 Census was undertaken during a unique time in our history, when many areas across Australia were operating under lockdown restrictions due to COVID-19. The ABS continuously monitored the changing COVID-19 environment to ensure that our safety measures were up-to-date and met the guidelines and advice from federal, state and territory governments.

Lockdown restrictions made it more difficult for Census Field Officers to enumerate people sleeping rough and those staying in non-private dwellings. In these circumstances, the method of enumeration was tailored to suit the specific lockdown conditions to maintain the safety of the public and our field staff. The following adaptations were made to the enumeration approach across the country:

  • in the Greater Sydney region and its surrounding areas, enumeration of people sleeping rough was not undertaken due to restrictions on movement across local government areas. The ABS used administrative data from NSW Department of Communities and Justice for counts and key demographic characteristics of people not in a dwelling on Census night, and people staying temporarily in brokered hotel or motel accommodation.
  • in Victoria, contactless enumeration was used in metropolitan Melbourne and was delayed by two weeks. Launch Housing, with the consent of key project partners, provided a de-identified list of people sleeping rough for the City of Port Phillip LGA as contactless enumeration was not possible in this area. Homes Victoria also provided administrative data of people staying temporarily in brokered hotel or motel accommodation. Each of these de-identified datasets were used for data substitution and supplementation purposes. 
  • while the Australian Capital Territory was placed into lockdown shortly after Census night, enumeration of people sleeping rough had been completed before then.
  • enumeration of people sleeping rough proceeded as normal in other states/territories, except for Cairns and Yarrabah in Queensland which were briefly delayed.

The ABS engaged extensively with service providers, including state and local governments, to identify the best option to ensure people experiencing homelessness were counted in the Census. The ABS gratefully acknowledges the assistance and support provided by the homelessness sector during enumeration in areas directly and indirectly impacted by COVID-19.

For more information on how Census data was collected during the pandemic see How the data is collected and COVID-19 and the Census.

Homelessness estimation

The homelessness estimates from the 2021 Census use the methodology outlined in Information Paper – Methodology for Estimating Homelessness from the Census of Population and Housing, 2012.

This methodology was developed in consultation with the ABS’ Homelessness Statistics Reference Group and builds on the review that the ABS conducted on the development of official estimates of homelessness. The ABS’ statistical definition of homelessness underpins this methodology. 

The methodology has been consistently applied with minor improvements to producing homelessness estimates for the 2021, 2016, 2011, 2006 and 2001 Censuses.

Homeless operational groups

The ABS uses six homeless operational groups for presenting estimates of people enumerated in the Census who were likely to have been homeless on Census night. These groups are:

  • people living in improvised dwellings, tents or sleeping out
  • people in supported accommodation for the homeless
  • people staying temporarily with other households
  • people living in boarding houses
  • people in other temporary lodgings
  • people living in ‘severely’ crowded dwellings.

Detailed information on the steps used to classify Census data for these groups are included in Information Paper – Methodology for Estimating Homelessness from the Census of Population and Housing, 2012.

Other marginal housing groups

This release also compiles estimates from Census data for the following three groups of people who may be marginally housed but are not classified as homeless: 

  • people living in other crowded dwellings
  • people in other improvised dwellings
  • people marginally housed in caravan parks.

Estimates for these groups can be found in the data downloads.

There are other forms of marginal housing such as housing with major structural problems or where residents are in constant threat of violence. These cannot be obtained from the Census and are therefore not included.

Humanitarian migrants

The majority of migrants who arrived in Australia within the Census year, were born overseas, and who reported having no usual address while in a private dwelling are not considered to be experiencing homelessness in the methodology.

The ABS identified a group of recent migrants which are likely to be humanitarian migrants, who in the methodology are classified as homeless. These are people who:

  • arrived in Australia within the Census year,
  • were born overseas,
  • reported having no usual address,
  • were in a private dwelling which was not an 'improvised dwelling, tent or sleeping out', and
  • report a country of birth which suggests they could be a recipient of a humanitarian visa.

According to the Department of Home Affairs, as extracted from departmental systems in July 2021, these countries were:

  • Iraq
  • Myanmar
  • Afghanistan
  • Syria
  • Pakistan
  • Eritrea
  • Iran
  • Democratic Republic of the Congo (DRC)
  • Venezuela
  • Tibet

As Tibet is not defined as a category within the Country of Birth of Person (BPLP) classification, nor is it feasible to separately identify from individual Census responses, only the top nine countries were considered in the 2021 Homelessness methodology.

Monetary cut-offs

The ABS applies monetary cut-offs for personal and household income, rental payments, and mortgage repayments as part of the homelessness estimation methodology. This ensures that individuals that have other means for “accommodation alternatives”, such as access to financial or housing resources, are not misclassified as experiencing homelessness. Information Paper – Methodology for Estimating Homelessness from the Census of Population and Housing provides more detail of these steps with links to the section for each operational group.

The values for monetary cut-offs are reconsidered each Census based on levels and movements in income, rents and mortgage payments as measured by relevant ABS data sources, as well as analysis of the Census data.

Personal and household income

For ‘People living in boarding houses’, an individual income cut-off of $800 per week was used in 2021, up from $650 per week used in 2016. For 2011 and 2006, an individual income cut-off of $600 per week was used and $400 per week in 2001. If 60% or more of the residents of a dwelling had incomes above this level the dwellings were not likely to be a boarding house.

For the 'People in other temporary lodging' homeless operational group, an individual income cut-off of $650 per week was used in 2021, up from $500 per week used in 2016. For 2011 and 2006, an individual income cut-off of $400 per week was used and $300 per week in 2001. People in these dwellings with incomes below this level and with certain other characteristics were classified as homeless.

A household income cut-off of $2,500 or more per week was used in 2021, in conjunction with certain tenure types and employment status to ascertain the likelihood of the occupants of a dwelling classified as improvised being construction workers, road workers etc. rather than as being homeless. The household income cut-off was $2,000 or more per week in the 2016, 2011 and 2006 Censuses and $1,594 per week for the 2001 Census.

Rental payments

The cut-off for rental payments has been increased in line with the intercensal Consumer Price Index increase for rents, and rounded to the nearest cut-off for Rent (weekly) ranges (RNTRD). The cut-off for rental payments was set at $500 per week in 2021, up from $450 per week in 2016 ($400 per week in 2011, $300 per week in 2006 and $265 per week in 2001).

Mortgage payments

As for income and rent, mortgage payment cut-offs were set to a level considered to indicative of accommodation alternatives and above a level of payment that could be afforded by people who were, on balance, most likely to be homeless.

The cut-off for mortgage payments was set at $1,800 per month in 2021, up from $1,600 per month in 2016 ($1,400 per month in 2011, $1,050 per month in 2006 and $845 per month in 2001).

Understanding data quality

Census data quality

Information about respondent error, processing error and undercount in the 2021 Census can be found at Managing Census quality. Information about dwelling and person response rates and item non-response rates for the 2021 Census can be found at Understanding data quality.

The Census of Population and Housing: Census Dictionary, 2021 provides a comprehensive reference for the 2021 Census of Population and Housing, providing information about the Census, the topics and classifications used, managing Census quality and a glossary of terms.

Post Enumeration Survey

The Census Post Enumeration Survey (PES) is run shortly after each Census to independently measure Census coverage. The PES results are used to determine how many people should have been counted in the Census, how many people were missed (undercount) and how many were counted more than once or in error (overcount).

However, the PES:

  • only covers people in private dwellings at the time of the PES, and therefore will not capture the homeless population staying in non-private dwellings at the time of the PES but who were missed on Census night. To ensure people in non-private dwellings on Census night (and other places not covered in the PES) were still represented within the target population, the PES applied an adjustment as part of the weighting step.
  • does not capture sufficient information to determine whether a person was experiencing homelessness at the time of the PES or at the time of the Census.

For all these reasons the PES cannot be used to estimate the undercount of people experiencing homelessness in the Census.

Use of administrative data

New sources of administrative data were used for the first time in 2021 Census. Since these were not available in previous censuses, there will be an increase in the total estimates due to the contributions from these data sources.

The use of the Victorian rooming house register identified a number of boarding houses that were counted as private dwellings in the 2016 Census. The characteristics and demographics of each boarding house were checked and those that did not provide accommodation to the homeless were removed from the estimates. The improved identification of boarding houses providing accommodation to the homeless will contribute to the increase in the estimate of ‘people living in boarding houses’. 

Administrative data was received from both the NSW and Victorian governments on people staying in temporary accommodation (such as hotel and motels) under brokerage or vouchers. They were quality checked and used for data supplementation purposes. Since these data sources were not available in previous censuses, the increase in the number of people in the operational group ‘People in other temporary lodgings’ is partly attributed to improved information in classifying this group. The increase is also likely to be partly attributed to COVID-19 related measures that provided temporary accommodation to people sleeping rough - or in crisis situations - to protect them, and the community, from the virus.

The improvements in data quality through greater use of administrative lists mean that estimates from the 2021 Census for people living in boarding houses in Victoria and people in other temporary lodgings nationally are not directly comparable with previous Censuses.

From the Homelessness Lists, additional information about occupancy, type of program and homeless cohort were used to inform data imputation of key demographic information for people in the homeless operational groups ‘people in supported accommodation for the homeless’, ‘people living in boarding houses’ and ‘people in other temporary lodgings’.

Population groups likely to be underestimated

Some population groups are underestimated in the Census and homelessness estimates. This incluides: youth experiencing homelessness; Aboriginal and Torres Strait Islander people; and people displaced from domestic and/or family violence.
To find out more information about these groups, see the following fact sheets and ABS research publications:

Item non-response rates

Most item non-response is attributable to the people who did not respond to the Census at all. These people have records created for them and while key demographic data items are imputed for them, the remaining data items are left as ‘not stated’.

The second and smaller contributor to item non-response is when people return a Census form but may not answer a particular question or questions. This is referred to as item non-response.

Item non-response rates in the Census are higher for those estimated to be experiencing homelessness, when compared to the general population. Item non-response also varies significantly across the homeless operational groups. This is a result of approaches used in the Census homeless enumeration strategy, particularly:

  • the Special Short Form used to enumerate people sleeping rough only asks a smaller set of Census questions. The questions that are not asked are coded as ‘Not stated’ and therefore contribute to item non-response
  • the use of administrative information and address lists to supplement the Census enumeration, which informed the imputation of key demographic data items. The remaining data items are left as ‘not stated.’

The table below compares the 2021 item non-response rates for person variables to 2016 rates. For item non-response rates of the general population see Understanding data quality.

Item non-response rates(a) for person variables, Total people experiencing homelessness, Australia(b), 2021
Person variable2016 Item non-response (%)2021 Item non-response (%)
Age (AGEP / IFAGEP)(c)



Australian Defence Force service status (ADFP)(d)



Core activity need for assistance (ASSNP)



Count of selected long-term health conditions (CLTHP)(d)



Country of birth of person (BPLP)



Educational attendance (TYSTAP)



Indigenous status (INGP)



Labour force status (LFSP)(e)



Proficiency in spoken English (ENGLP)(g)



Registered marital status (MSTP / IFMSTP)(c)(f)



Sex (SEXP / IFSEXP)(c)



Total personal income (weekly) (INCP)(f)



This table is based on place of usual residence(h).

(a) Includes imputed persons. Excludes overseas visitors (unless stated otherwise), Australian residents temporarily overseas and non-applicable persons.

(b) Includes Other Territories (Jervis Bay Territory, Territory of Christmas Island, Territory of the Cocos (Keeling) Islands and Norfolk Island).

(c) The variables, Imputation flag for age (IFAGEP), Imputation flag for Registered marital status (IFMSTP), and Imputation flag for Sex (IFSEXP) are calculated as 'imputation rates' rather than non-response rates. This is because 'not stated' is not a category within these variables. When a person has not responded to the Census at all, these key demographic variables are imputed. Imputation rates include overseas visitors.

(d) These variables were not available in 2016.

(e) Applicable to employed persons aged 15 years and over.

(f) Applicable to persons aged 15 years and over.

(g) Applicable to people who did not respond to both the Language used at home (LANP) and Proficiency in spoken English (ENGLP) questions.

(h) The place of usual residence refers to the count of people where they lived, or intend to live, for a total of six months or more during the Census year. Overseas visitors are excluded.

Data confrontation

The Australian Institute of Health and Welfare (AIHW) has provided ABS with data from the Specialist Homelessness Services (SHS) Collection, for validation purposes, which are preliminary estimates of the number of people reported by SHS providers to have been in SHS accommodation on Census.

Counts of people in supported accommodation estimates by state and territory, 2021









Persons in supported accommodation for the homeless (Census)









Number of SHS clients who were accommodated (AIHW SHSC) (a)









Census less SHSC









 na not avialble

(a) Clients are counted if they had a period of accommodation which crossed over the 10th August 2021 (derived using the accommodation start and end dates).

(b) In South Australia short-term and medium-term accommodation is often provided to SHS clients outside the SHS sector which is currently excluded from the AIHW data. Due to this, an accurate count for validation purposes is not currently available.

The estimates from the Census are higher in total for Australia than the AIHW SHS Collection estimates, and higher for all jurisdictions except for Tasmania. The Census estimates are generally expected to be higher than those for the SHS collection because of the inclusion of charities, church and community groups and other providers that are not SHS-funded, which are not in scope of the AIHW collection.

Use of Place of enumeration and Place of usual residence

The national and state-level homelessness estimates are presented on a ‘place of usual residence’ basis.

This is preferable for state and territory governments as it indicates the origin of those experiencing homelessness and where the services are deficient, rather than their location on Census night. It also supports comparison with the main Census outputs which are presented on ‘place of usual residence’ basis. For ‘Persons living in improvised dwellings, tents or sleeping out’ or ‘Persons staying temporarily with other households’ who reported ‘no usual address’ would have their place of usual residence imputed as the state or territory where they were enumerated.

The homelessness estimates at sub-state geography (i.e., Local Government Area, Statistical Area Level) are presented at ‘place of enumeration’ basis. The difference in the estimates between state-level estimates on ‘place of usual residence’ or a ‘place of enumeration’ basis is marginal.

Data release


For users who wish to undertake more detailed analysis of the data, the microdata will be released through TableBuilder: Census of Population and Housing: Estimating homelessness. Further information about ABS microdata, including conditions of use, and access is available via the Microdata and TableBuilder section on the ABS website.

There may be a small difference in Homeless operational groups (OPGP) counts between TableBuilder and the tables published in Estimating Homelessness: Census. This difference is due to the confidentiality technique applied that adjusts counts within Census products to prevent any identifiable information about individuals being released.

Custom tables

The ABS offers specialist consultancy services to assist data users with more complex statistical information needs. Users may wish to have the unit record data analysed according to their own needs or require tailored tables incorporating data items and populations as requested by them. Tables and other analytical outputs can be made available electronically or in printed form. As the level of detail or disaggregation increases with detailed requests, the number of contributors to data cells decreases. This may result in some requested information not being able to be released due to confidentiality. All specialist consultancy services attract a service charge, and clients will be provided with a quote before information is supplied.

For enquiries about these and related statistics, contact the Customer Assistance Service via the ABS website Contact Us page. The ABS Privacy Policy outlines how the ABS will handle any personal information that you provide to us.


The Census and Statistics Act 1905 provides the authority for the ABS to collect statistical information, and requires that statistical output shall not be published or disseminated in a manner that is likely to enable the identification of a particular person or organisation. This requirement means that the ABS must ensure that any statistical information about individuals cannot be derived from published data.

To minimise the risk of identifying individuals in aggregate statistics, a technique is used to randomly adjust cell values. This technique is called perturbation. Perturbation involves small random adjustment of the statistics and is considered the most satisfactory technique for avoiding the release of identifiable statistics while maximising the range of information that can be released. These adjustments have a negligible impact on the underlying pattern of the statistics.

After perturbation, a given published cell value will be consistent across all tables. However, adding up cell values to derive a total will not necessarily give the same result as published totals.

The introduction of perturbation in publications ensures that these statistics are consistent with statistics released via services such as TableBuilder.

As an Australian Government agency, ABS also complies with the Privacy Act 1988 and handles personal information in accordance with the Australian Privacy Principles (APPs). The ABS Privacy Policy outlines how the ABS will handle personal information for producing statistics.

Revisions to 2006 data

Estimates for the 2006 Census in this release have been updated using the same perturbation method applied to estimates for 2021, 2016 and 2011, and may differ slightly from the 2006 estimates in previous releases. Any differences are the result of the random adjustment through perturbation only. 


As estimates have been rounded, discrepancies may occur between sums of the component items and totals. As estimates have also been randomly adjusted to avoid the release of confidential information, discrepancies may occur in estimates appearing in more than one table.

Rates per 10,000 of the population

Population rates presented in this publication are presented as a rate per 10,000 of the total population. That is the number of people experiencing homelessness per 10,000 persons of the usual resident population in the Census excluding people, at sea, or in migratory and offshore regions.

The usual resident population is available in Census TableBuilder by subtracting people at sea, or in migratory and offshore regions from the population total of interest. For 2006 and 2011, usual residents of External Territories need to be excluded from the calculation of Australian population total.

ABS Geography

The Australian Statistical Geography Standard (ASGS) provides a framework of statistical areas used by the Australian Bureau of Statistics (ABS) and other organisations to enable the publication of statistics that are comparable and spatially integrated. The ASGS provides users with an integrated set of standard areas that can be used for analysing, visualising, and integrating statistics produced by the ABS and other organisations. See here for further information about the ASGS.

Further information about ABS statistical geography and a diagram of how the various ABS Structures interrelate can be found in the Australian Statistical Geography Standard (ASGS) Edition 3.

Boundaries of SA2, SA3 and SA4 regions may change over time. Geographic correspondences between 2016 and 2021 editions of statistical areas are available in the downloads section of the Australian Statistical Geography Standard (ASGS) Edition 3. These can be used to understand the relationships between 2016 and 2021 areas and to convert data from 2016 ASGS regions to 2021 regions.

The ASGS Local Government Areas are an ABS approximation of gazetted local government boundaries as defined by each State and Territory Local Government Department. Local Government Areas cover incorporated areas of Australia. Incorporated areas are legally designated parts of a State or Territory over which incorporated local governing bodies have responsibility. The major areas of Australia not administered by incorporated bodies are the northern parts of South Australia, and all of the Australian Capital Territory and the Other Territories. These regions are identified as ‘Unincorporated’ in the ASGS Local Government Areas structure. More information on Local government areas and changes between 2016 and 2021 boundaries can be found in Australian Statistical Geography Standard (ASGS) Edition 3.


ABS publications draw extensively on information provided freely by individuals, businesses, governments, and other organisations. Their continued cooperation is very much appreciated: without it, the wide range of statistics published by the ABS would not be available. Information received by the ABS is treated in strict confidence as required by the Census and Statistics Act 1905.

The ABS acknowledges and thanks the service providers and their staff for their ongoing contributions to help maximise the overall quality of the enumeration of people experiencing homelessness in the 2021 Census, some of whom serve on the Homelessness Statistics Reference Group (HSRG).

Homelessness Statistics Reference Group (HSRG)

The ABS would like to acknowledge and give thanks to the members of the HSRG who assisted in peer reviewing the homelessness estimates, and for their contributions and commitment to the process of analysing, understanding, and reporting on homelessness in Australia.

  • Australian Housing and Urban Research Institute
  • Australian Institute of Health and Welfare
  • Homes Tasmania (Tasmanian Government)
  • Council to Homeless Persons
  • Department of Communities (Government of Western Australia)
  • Department of Communities and Justice (New South Wales Government)
  • Department of Communities, Housing, Digital Economy (Queensland Government)
  • Department of Families, Fairness and Housing (Victorian State Government)
  • Department of Social Services (Australian Government)
  • Department of Territory Families, Housing and Communities (Northern Territory Government)
  • Gold Coast Youth Service
  • Homelessness NSW
  • Housing ACT (ACT Government)
  • Micah Projects Inc.
  • Mission Australia
  • National Aboriginal & Torres Strait Islander Housing Alliance
  • National Housing Finance and Investment Corporation
  • National Shelter
  • RMIT University
  • SA Housing Authority (Government of South Australia)
  • Salvation Army
  • University of Queensland / School of Social Science
  • University of Western Australia / Centre of Social Impact
  • Upstream Australia


Show all

Quality declaration

This quality declaration shows how the official estimates of the prevalence of homelessness from the ABS' Census of Population and Housing addresses each of the elements of the ABS’ quality framework, covering relevance, timeliness, accuracy, coherence, interpretability and accessibility.

Institutional environment








Show all

Back to top of the page