|Page tools: Print Page Print All RSS Search this Product|
A History of the Australian Census of Population and Housing
A Census collector with Indigenous people in the Boonah-Fassifern district of Queensland in 1911
The Brisbane Courier 13 May 1911
In this first Census, the process of giving out and gathering back the forms met with the usual Australian difficulties, that is to say, ‘droughts’ (in South Australia) and ‘flooding rains’ (in Queensland). Once collected, the Census was compiled almost entirely by hand, with the help of only a few basic adding machines. About 250 temporary staff sorted over 4 million cards into piles according to the answers to particular questions. Each pile was totalled, then the cards were re-sorted according to the answers to the next question. The main work began in August 1911 and was completed by the end of June 1912.
In 1913, Charles Wickens was appointed Supervisor of the Census to oversee the remaining work of tabulating and analysing the data. Two volumes of tables were published in 1914, but the final Statistician’s Report was not released until 1917, six years after the Census. This was due to the outbreak of World War I, which meant changed arrangements within the Bureau. During the war, the Bureau conducted a ‘war census’ and a ‘wealth census’ under special legislation. Rather than true Censuses, they were registers of manpower and assets respectively, based on mail-back cards available from Post Offices. The war also affected the Bureau in an immediate way: of the 29 male staff in the Bureau in 1915, nine served in the military, four of whom died overseas.
Charles Edward Rudolph, one of four CBCS staff who died in World War I.
Australian War Memorial image no. DA12955
From War to Depression – the Censuses of 1921 and 1933
Preparatory work for the 1921 Census was done under the direction of Knibbs while he was still Commonwealth Statistician, with Charles Wickens directly in charge. When Knibbs left in 1921, Wickens was appointed the second Commonwealth Statistician.
In a new arrangement which lasted until the 1980s, the bulk of the collection work for the 1921 Census was undertaken by the Commonwealth Electoral Office, which had a network of staff located throughout Australia.
In the early 1920s Australia was concerned with post-war reconstruction, with some special schemes in place for returned servicemen and many delayed infrastructure projects going ahead. In line with this, preference was given to ex-servicemen in filling the temporary positions to tabulate the Census.
In a change from 1911, it was decided to hire automatic tabulating machines for use on the Census. These were Hollerith machines, similar to those developed for tabulating the US Census. The machines functioned through the use of cards which were punched with holes in specific places. Each position on the card related to a different question and response on the Census form. There were a number of different machines to cover a range of processes: punching the cards, verifying cards were punched correctly, punching multiple cards at once, sorting cards, counting cards and tabulating.
The next Census was expected to be held in 1931 but was cancelled in 1930. At that time Australia was in the grip of the Great Depression. The federal and state governments were struggling with debt repayments and introduced austerity measures curbing government spending. Eventually, the Census was run in 1933 after it was recognised that the Census was an opportunity to understand the impact of the depression and could also provide some temporary employment. Postponing the Census required an amendment to the Census and Statistics Act 1905 which had specified that Censuses must be held every ten years.
While the running, processing and tabulation of the 1933 Census occurred under the fourth Commonwealth Statistician, Edward McPhee, most of the preparation occurred while the Bureau was under the leadership of the academic and government economic advisor Lyndhurst Giblin. He was asked to step in when Charles Wickens suffered a stroke in 1931.
The 1933 Census was the only one run almost entirely from Canberra. The CBCS had moved there in 1928 and was located in West Block but there was not enough space there for the census staff. A Census Office was therefore established in the Jolimont Building in Alinga Street, a weatherboard building which had been relocated from Melbourne. In 1933 the mapping and tabulation were both undertaken in that building. Most of the temporary staff came to Canberra and lived in hostel accommodation for the duration of their employment, although some found accommodation in the nearby town of Queanbeyan. As in 1921, ex-servicemen were given preference in filling these jobs.
The Jolimont Building, 1929
National Archives of Australia
Several additional topics were added to the 1933 Census reflecting the concerns of the time regarding the welfare of the population. These included questions on war service, number of dependent children aged less than 16 years in each household, whether children were orphans and income. For the first time, information such as the number of adults earning no income or only a very low income whilst supporting dependent children could be obtained from the Census.
There was no Census for 14 years after the 1933 Census. In 1941, when a Census would otherwise have been held, Australia was at war. As in the First World War, registers of manpower and wealth were conducted under special legislation.
Post-war reconstruction Censuses - 1947 and 1954
The 1947 Census was planned and undertaken quickly after the end of the Second World War and was seen as essential for post-war reconstruction of the Australian economy. Its development was directed primarily by Roland Wilson, the youngest ever serving Commonwealth Statistician. He was first appointed in 1936 when he was aged just 32 years. Wilson was a former Rhodes Scholar who had been encouraged by Giblin to move into the Statistician’s Branch of the Treasury and while in that role, and then as Statistician, developed statistics which could inform economic policy.
In 1947, the employment questions were re-ordered and revised slightly. A question was used to identify all people undertaking what was termed ‘economic activity’ as opposed to people who were not, such as full-time students and housewives.
A significant number of questions about dwellings were added in 1947. These covered the dwelling’s gas, electricity and water supply (e.g. do you have flowing water?); its toilet, washing and cooking facilities (e.g. do you have a flush toilet?); and the date when the dwelling was built. Housing shortages, and shortages of building materials, had been a pressing issue both during and after the war. The federal and state governments were attempting various methods of increasing housing stock, including annual Commonwealth-State Housing Agreements from 1945, which directed funds at building public rental housing.
From 1947, Torres Strait Islander people were included in official Australian population totals from the Census, after lobbying by the Queensland State Government. They successfully argued that Torres Strait Islander peoples were not specifically listed for exclusion from population counts in the Constitution, which referred to ‘aboriginal natives’.
The war had highlighted the value of the radio for spreading information. Roland Wilson, in 1947, was the first to use that medium to promote the Census, through a series of radio talks explaining how to fill in the Census form. In addition, women’s magazines were used for promotion. In particular, the Australian Women’s Weekly ran a large piece on the Census focusing on the experiences of collectors. For the first time, the car was the primary method of transport for collectors not able to do their job on foot. However, the Women's Weekly still found collectors travelling by horse and sulky in 1947.
As the 1941 Census had been cancelled and a Census held in 1947, it was decided to run a Census in 1954 as a 'catch-up'. The 1954 Census would fall midway between 1947 and 1961, the year the next ten yearly Census was due, and bring Australia back into step with the timing of other censuses in the British Commonwealth. Stanley Carver was the Commonwealth Statistician at this time. The 1954 Census was notable in that a few years prior, the Bureau hosted the Conference of British Commonwealth Statisticians. This was considered to be a great learning experience for the host nation’s staff. While the agenda focused heavily on labour and industry statistics, there was also a session on lessons from recent censuses, with papers contributed by Canada, Ireland and India.
West Sydney Electoral Office staff answer questions about the Census.
Sydney Morning Herald 1954 29 June p. 3
For the 1954 Census, alternative ways of collecting the Census were trialled, including having people mail back Census forms. However, the traditional method of drop off/pick up by collectors was found to be the most successful and remains in use today. The 1954 Census also saw the processing of the Census forms divided between different State capitals, each centre employing temporary staff for the purpose. After the Census, assessments of processing found that this had led to variation in the way answers on Census forms were coded by the staff in different States, and a single centre has been the favoured option since then. This was the first Census in which coding was formally assessed.
In 1954, most of the new dwelling questions introduced in the previous Census were dropped and only a few other new questions were introduced, such as whether the dwelling included both a kitchen and a bathroom and whether the dwelling was a rural holding.
The Census in a decade of new perspectives - 1961 and 1966
The 1960s saw a new era with five yearly Censuses and new technology fundamentally changing Census-taking. Keith Archer was the Commonwealth Statistician for this period.
The 1961 Census was the last to use tabulating machines. Over time, tabulating machines had become more sophisticated and data could be processed faster. Large 'Census Trio' machines combined the processes of sorting, calculating totals, and producing summary punched cards, which had previously been performed by different machines.
George Crossman undertaking minor
maintenance on a Census Trio machine
The machines used in 1961 were so large that when delivered they could not fit through the door of the Census Office in Canberra. A large hole had to be cut in the side of the wooden building to install them.
The Census Trio machine being installed in the Jolimont Building
While in 1947 radio had been used for Census publicity for the first time, in 1961 a film advertised the Census in theatres and on television. Paid advertisements appeared in newspapers, including foreign language newspapers. Also, with so many new migrants in the country, a special statement was prepared in seven different languages and distributed to migrant clubs and other meeting places.
New questions included a question on professional qualifications for those employed or looking for work. The 1961 Census content also covered the take-up of new technology, with a question on whether households had a television.
As the Bureau’s new Labour Force Survey started around the same time as the 1961 Census, a comparison was made between the two collections. It proved so useful that such comparisons became standard after every Census, allowing the Bureau to identify and qualify differences in results between the two sources of information.
In 1966, for the first time since the colonial period, a Census was run five years after the previous Census, something statisticians had wanted since Federation. Australia has had five yearly Censuses since then, and this became mandatory after an amendment to the Census and Statistics Act 1905 in 1977. This amendment followed a High Court decision that there must be an electoral redistribution in the life of every parliament, something that had to be based on accurate population estimates.
The 1966 Census saw some really fundamental changes in process. Pilot testing was introduced to evaluate the draft Census schedule. This involved selected households being asked to fill in the schedule, and then to participate in a follow up interview, to identify any problems in design.
Also for the first time, users of Census data such as other government departments were approached to determine their requirements, prior to the Census. While only a few specific new requests were made, the number of questions did increase between 1961 and 1966. New topics included the number of motor vehicles at a dwelling as well as new questions on education.
The 1966 Census saw the introduction of computer technology to the processing of the Census. This utterly changed the way data was coded, edited, and even what it was possible to collect. In particular the new technology allowed family relationships to be far more comprehensively examined. The Bureau also began to spend considerably more time talking to clients to determine what their needs were before running and releasing tables. This Census also saw the cancelling of the Statistician's Report. Five yearly Censuses, combined with the long time needed to produce the report, made it less practicable, especially given the developing alternative of customised tables.
National Archives of Australia A2399, COD40/5
Computers used to process 1966 Census, and staff
National Archives of Australia: A2399. COD47/1
In 1966, the Bureau was asked to assist in conducting a census in the Territory of Papua New Guinea. Previously, censuses had only been run in the cities and towns. It was decided to run a 10% sample of the entire population, including remote villages. It proved difficult to obtain ages in remote areas where there had been no birth records kept, so some inventive techniques were used to extrapolate age. A list of notable dates was prepared for each district and each adult respondent was asked to identify an event they could remember as a child. For example, they might be asked ‘Can you remember when Father Ryan first arrived?’, an event that happened in 1912. They were then asked to point to a child who was around the same age as they had been when Father Ryan arrived. The collector then used the child’s age and the date of the event to get the approximate date of birth of the adult.
Census collector, Papua New Guinea, 1966
The other really significant event for the population Census in the 1960s occurred after the 1966 Census. Calls for Aboriginal and Torres Strait Islander people to have the same rights as other Australians, resulted in a referendum in 1967. The highest yes vote ever recorded in a referendum ensured that Section 127 of the Constitution was abolished and all Aboriginal and Torres Strait Islander people were to be included in population counts resulting from the Census.
FCAATSI poster, 1967 referendum
The Papers of Gordon Bryant (MS 8256); National Library of Australia
Privacy and change impacts the Census - 1971 and 1976
In 1970, Jack O’Neill took over as Commonwealth Statistician after Keith Archer retired from the position due to ill health. The 1970s was a time of great expansion in the Census topics. This followed seismic shifts in modern society and moves internationally to compile more social statistics to measure those changes. At the same time, the 1970s saw fresh public awareness of the issues of privacy and confidentiality and how these issues might relate to the Census.
There were substantial changes to Census questions for the 1971 Census. Following the changes to the Constitution, the race question was redesigned, to primarily focus on Aboriginal or Torres Strait Islander origin, and the term 'racial origin' replaced 'race'. This was a tick box question with options 'European', 'Aboriginal' 'Torres Strait Islander' and 'Other origin' (which was to be written in). People of more than one origin were asked to identify the group to which they considered themselves to belong. In subsequent Censuses, the question evolved further, to ask only whether people were of Aboriginal or Torres Strait Islander origin, and to enable people to identify themselves as of both these origins.
There were also a number of new questions, particularly about housing and geographic location. Questions included number of dwellings owned or being purchased, method of sewage disposal, number of bedrooms, birthplaces of parents, and address five years ago (which made it possible to study internal migration in Australia). One question covered in the early testing was a question on income but this was dropped before the final Census. There were also physical changes to the Census form. Instead of one large densely packed sheet as had been used since 1921, the schedule was redesigned into a booklet. It made the form larger but also much easier to follow.
Even prior to the 1971 Census there were protests and dissent over the Census, particularly in relation to privacy. This was stirred up by a television show on government privacy shown one month before the Census. Major newspapers and one political party questioned the inclusion of names on the Census form. However, these protests did not seem to have any impact on the quality of the Census data.
Newspaper headlines about the 1971 Census
Jack O’Neill oversaw the development of the Australian Bureau of Statistics Act 1975 which resulted in the CBCS being transformed into the Australian Bureau of Statistics (ABS) and O’Neill becoming the first Australian Statistician. He retired in the same year.
The 1976 Census included significant changes to content. New topics in this Census included income, handicaps, language use, holidays, childcare, life insurance, vehicle licences, and benefits received. Most of these questions have not been asked since while others have become standard.
To help understand the concerns raised in the media in 1971, the ABS commissioned a study into publicity prior to the 1976 Census. It recommended strategies and a substantial budget for a publicity campaign.
A story board expressing one of the ideas for a media campaign for the 1976 Census
Madison Research Pty Limited.
Late in 1975, the government changed with the election of the Fraser Coalition Government. The new government would have preferred to make changes to Census content but it was too late to change the schedule as printing was already underway. The new government was cutting expenditure in many departments at the time, and chose to cut Census expenditure including the publicity budget.
The 1976 Census encountered substantial negative publicity as the nation debated the value and purpose of a Census and the importance of privacy. Unfortunately, with a very limited budget the ABS was not able to combat all the misunderstandings (for example, that the Census was investigating mental health) and media concerns raised, particularly about whether names were really needed on the Census form. The new Australian Statistician, Bill Cole, only served from May to December 1976 but in this time strongly defended the Census. Nevertheless, it was estimated that there was a slightly higher undercount of the population for the 1976 Census than for any previous Census.
Newspaper headlines about the 1976 Census
Meanwhile, the government cut the Census budget further so the 1976 Census was unable to be fully processed. Only age, sex, marital status and birthplace were processed for all people. A 50% sample was processed for the other topics.
The debate surrounding the 1976 Census led to a review by the Law Reform Commission which, while making some recommendations for improvements, fundamentally supported the ABS’ methods and processes, including the requirement for names to be collected on the schedule. This was due to the clear evidence that names strongly encourage people to fill in the Census form and to be more accurate in their responses. Collecting names also makes it possible to undertake follow up surveys to subsets of households, to help assess the quality of Census data.
Asking the people – the Censuses of 1981 and 1986
In comparison to the 1970s, the Censuses of the 1980s were relatively trouble free. Roy Cameron was appointed the Australian Statistician in 1977 and in the lead up to the 1981 Census the ABS concentrated on improvements in the quality of the Census and on wider consultation. For the first time, the ABS advertised through major newspapers for public submissions on topics for the next Census. The ABS received over 1600 suggestions for topic inclusions (with 40 entirely new topics recommended) and only around 60 submissions for topics’ exclusion. Extensive testing was undertaken and topics found not suitable for the Census were often recommended for surveys.
In 1981, a question on the total number of children a woman had ever had was asked for the first time. While desirable for determining fertility levels and trends, it hadn’t been asked previously as it was seen as too sensitive. Instead, less useful questions had been used, asking only about children of marriage, and the question had been removed entirely from the 1933 Census form. By 1981, social mores had changed, and there had been lobbying by unmarried parents to include their children.
A new topic, on ancestry, was included for the first time in 1986. This topic had been developed by a committee specifically established to resolve the issue of a lack of sufficient information about ethnicity in the Census.
The goals for the 1981 Census reflected the difficulties of the 1970s with aims to reduce the size of the form and minimise the information required from the public. There was also a complementary goal of improving the quality of the data by reducing the number of people missed by the Census and the numbers of people not answering each question.
Roy Cameron retired in 1985 and Ian Castles was appointed Australian Statistician in 1986. Castles encouraged the ABS to return to more statistical analysis of the Census and during his time considerably more analytical Census publications were released.
In 1986, industrial action by staff of the Australian Electoral Commission in two states meant that the ABS took over managing the collection in those states.
The quality of the Census also came under the spotlight in 1986 with significant improvements to family data. More accurate counts of single parent families were made possible, through the coding of temporarily absent household members, while a change to the relationship question meant de facto and blended families could be identified for the first time.
In 1986, CD-ROMS were first introduced as a means of data release. In doing so, the ABS was using technology in advance of most Australian businesses at the time. As a result it found it necessary to sell CD-ROM readers as well.
Capturing the content – the Censuses of 1991 and 1996
The 1991 Census was the first Census since 1921 to be run completely by the ABS with the Australian Electoral Office no longer involved in collection management. From 1991, the date for the Census was moved from June to August to minimise the likelihood of the Census overlapping with school or public holidays in the states and territories.
In processing the forms, optical mark recognition was used for the first time, alongside computer-assisted coding. Some new tick boxes were included on the form and this had effects on the data. For those topics where tick boxes only listed the top responses from the previous Census (because the possible responses were too numerous to list), people showed a slight preference for ticking a box over writing in a response.
In 1994, Ian Castles retired and Bill McLennan was appointed the Australian Statistician in 1995. Bill McLennan had most recently been the Director of the UK Central Statistical Office and prior to that had had a long career in the ABS.
From 1996, a digital map database that encompassed every locality in Australia made the manual drafting of thousands of collector maps unnecessary. The same database was also used for the maps released as part of the data output. This Census was the first to release the data in two stages with data that was quicker and easier to process (such as age and sex) released earlier, and the more complex data (such as occupation) released later. This allowed much of the data to be released substantially earlier than in previous Censuses.
The Census in the 21st century - 2001 to 2011
Dennis Trewin was the Australian Statistician during the 2001 and 2006 Censuses, having been appointed in 2000. Following Dennis Trewin's retirement in 2007, Brian Pink was appointed Australian Statistician.
The first Censuses of the 21st century included a focus on new technology, with questions about the use of personal computers in the home and the use of the internet.
New technology also impacted in an unexpected way with a world-wide email/internet campaign based on the mistaken idea that if enough people recorded their religion as ‘Jedi Knight’, then Statistical Bureaux around the world would be forced to include ‘Jedi’ as a legitimate religion. This was untrue but did not stop people from trying. The 70,000 people who reported Jedi in the 2001 Census became part of the classification ‘religion – inadequately described.’
A big change this century was the decision in 2001 to allow each householder to choose to allow their name-identified information to be preserved for posterity, with the idea that in the future it will be of interest to genealogists and social historians. This followed the recommendation of the House of Representatives Standing Committee on Legal and Constitutional Affairs. The Committee took into account submissions from those who wanted all name-identified Census forms to be kept and concerns expressed, particularly by the ABS, that keeping name-identified information could compromise the quality of the data. This was because confidentiality and privacy assurances had been built based on the fact that name-identified information was not kept. For example, Census forms are pulped after processing and names are stripped from processed data. It was felt that keeping the forms without the public’s consent would seriously erode public confidence in the ABS’ privacy and confidentiality assurances. In 2006, over 50% of the population chose to have their Census forms kept for posterity. These microfilmed forms are to be kept confidential for 99 years by the Australian Archives before being released.
Australian Statistician Mr Brian Pink hands over the Time Capsule information from the
2006 Census to the Director-General of the National Archives of Australia, Mr Ross Gibbs.
There were also two new topics in 2006, on unpaid work and need for assistance. These questions reflect decades of effort to resolve long held client requests for such data. The 2011 Census will be conducted on a basis comparable to the 2006 Census, due to financial constraints within the ABS. As a consequence, all topics asked in the 2006 Census have been retained in 2011 and no new topics have been added.
Publicity photo 2006 Census
In 2006, many Australians completed their Census form on the internet. While it had been tested in 2001, the system was fully functioning in 2006 with nearly 10% of households submitting their forms this way. The eCensus system has been redeveloped for 2011 using web 2.0 technologies to create a faster, more efficient application. The ABS is aiming to have at least 30%, and perhaps even 40% of Australian households complete their Census form via the Internet. Though it’s a long way from the stiff cards and manual sorting of 1911, the Census of 9 August 2011 builds on the important work of those who organised the first national Census 100 years ago.
Endnote 3 Knibbs, G H & Commonwealth Bureau of Census and Statistics (Australia) c. 1908, Official Year Book of the Commonwealth of Australia No 1 - 1908, McCarron, Bird & Co, Melbourne.
Endnote 4 ‘Instructions to Collectors’ in Knibbs, GH c.1917, Census of the Commonwealth of Australia taken for the night between the 2nd and 3rd April, 1911 Volume 1: Statistician’s Report, CBCS, Melbourne. p. 465 Back to top
Archer, KM, c.1967, Census of the Commonwealth of Australia 30th June 1961, Volume VIII – Australia, Statistician's Report, CBCS, Canberra.
Australia, Parliament, House of Representatives Standing Committee on Legal and Constitutional Affairs 1998, Saving our census and preserving our history: a report on the inquiry into the treatment of name-identified census forms, (Kevin Andrews), House of Representatives Standing Committee on Legal and Constitutional Affairs.
Australia, House of Representatives, Census and Statistics Bill 1905, Second Reading Speech, 23 August, 1905.
Australia, Senate, Parliamentary Debates, 4 November 1911, pp. 5672-5679.
Australia, Senate, Parliamentary Debates, 11 November 1911, pp. 6013-6034.
Australian Bureau of Statistics (ABS), c.1986, Census 86 - How Australia Takes a Census, Census of Population and Housing, 30 June 1986, cat. no. 2102.0, ABS, Canberra.
ABS 1991, How Australia Takes a Census, Census of Population and Housing, 1991, cat. no. 2903.0, ABS, Canberra.
ABS 1996, How Australia Takes a Census, 1996 Census of Population and Housing, cat. no. 2903.0, ABS, Canberra.
ABS 2000, How Australia Takes a Census, 2001 Census of Population and Housing, cat. no. 2903.0, ABS, Canberra.
ABS 2006, How Australia Takes a Census, Australia, 2006, cat. no. 2903.0, ABS, Canberra.
ABS c. 1981, ‘History of the Australian Census’, 1981 Census of Population and Housing: Development Programme – Procedural Issues, Working Paper No. P10, unpublished paper.
ABS 2005, Informing a Nation: the evolution of the Australian Bureau of Statistics 1905-2005, cat. no. 1382.0, ABS, Canberra.
ABS 1983, Making Sense of Census, 1981, cat. no, 2140.0 & cat. no. 2103.0, ABS, Canberra.
ABS 2011, 'Aboriginal and Torres Strait Islander peoples and the Census of Population and Housing since 1967' in Reflecting a Nation: Stories from the 2011 Census, ABS, Canberra.
ABS 2011, 'One hundred years of working on the Census' in Reflecting a Nation: Stories from the 2011 Census, ABS, Canberra.
'Census takers expect shotgun or cup of tea', The Australian Women’s Weekly, Saturday 21 June, 1947, p. 13.
Carver, SR, c.1953, Census of the Commonwealth of Australia 30th June 1947, Statistician's Report, CBCS, Canberra.
Carver, SR, c. 1962, Census of the Commonwealth of Australia 30th June 1954, Volume VIII – Australia, Statistician's Report, CBCS, Canberra.
Commonwealth of Australia Constitution Act (UK) 1900 <<www.legislation.gov.uk>>
Forster C and C Hazelhurst, 1988, 'Australian Statisticians and the development of Official Statistics' in Year Book Australia 1988, cat. no.1301.0, ABS, Canberra.
Knibbs, GH c.1917, Census of the Commonwealth of Australia taken for the night between the 2nd and 3rd April, 1911 Volume 1: Statistician’s Report, CBCS, Melbourne.
Knibbs, GH & Commonwealth Bureau of Census and Statistics (Australia) 1911, The first Commonwealth census, 3rd April, 1911 : notes / by G.H. Knibbs, Commonwealth Statistician, Government Printer, Melbourne.
Madison Research Pty Limited 1975, Public Attitudes towards the Census - its form and function; Research report conducted for Fortune (Aust.) Pty. Ltd. on Behalf of The Australian Bureau of Statistics, Madison Research Pty. Limited, Sydney.
RM Walker 1975 'Centralisation and decentralisation considerations in locating the processing operation for large statistical collections with particular reference to the Australian population census', unpublished paper.
Wickens, Chas. H c.1927, Census of the Commonwealth of Australia taken for the night between the 3rd and 4th April 1921, Statistician's Report, CBCS, Melbourne.
Wilson, Roland c.1940, Census of the Commonwealth of Australia 30th June 1933, Statistician's Report, CBCS, Canberra.
These documents will be presented in a new window.