“..... in planning a better future we’re counting on you”.
(2001 Census, public awareness campaign theme)
On Tuesday 7 August 2001 the Australian Bureau of Statistics conducted Australia’s fourteenth national Census. The 2001 Census had particular historical significance, coinciding as it did with the Centenary of Federation and the start of a new century and millennium.
This article presents information on the background, lead up to, and processing of, the 2001 Census - the largest data collection operation ever undertaken by the ABS.
Planning and preparations for the Census started over six years ago and it will be the end of 2002 before the full array of statistics compiled from Census forms is available for release. The Census results will be used to compile the December quarter 2001 State/Territory population estimates. Detailed statistics for all geographic areas will become available from July 2002.
The main objective of the Census is to measure the number and key characteristics of people in Australia on Census night. This provides a reliable basis for estimating the population of each State, Territory and local government area, primarily for electoral purposes (including the determination of the number of seats allocated to each State and Territory in the House of Representatives) and for the distribution of government funds. About $35 billion per year will be distributed in Commonwealth financial assistance grants between the States and Territories using figures resulting from the 2001 Census.
USES OF CENSUS DATA
However, there are many other purposes served by the Census. At a time when there is growing demand for rural, regional and remote data, the five-yearly Census provides a comprehensive picture of Australia for small geographic areas. Detailed information is also available for small population groups. The Census complements the other sources of information gathered by the ABS from other surveys and collections for which the results are generally only available at national, state and broad regional levels. This wealth of information supports the planning, administration, policy development and evaluation activities by governments, businesses and community organisations.
The Australian Bureau of Statistics has handled over 100,000 requests for data from the 1996 Census, in addition to the access made through libraries and other organisations that disseminate Census data. It is anticipated that demand for 2001 Census data will be even greater.
2001 Census data will be available in many formats and through a variety of channels. 2001 Census data will be available from printed publications, CD-ROM, the Internet, the ABS website, the AusStats service and via email. Full details of the Census outputs are contained in the publication 2001 Census of Population and Housing, Proposed Products and Services (Cat. no. 2011.0).
There will be a significant increase in the amount of data from the 2001 Census that is available free of charge as a community service. As in the past, many of these products and publications will be available in most public libraries. In addition, for the 2001 Census much free data will be provided via the Internet including basic community profiles down to statistical local area level, and the main findings from the publications Selected Social and Housing Characteristics (Cat. nos. 2015.0-8), Selected Education and Labour Force Characteristics (Cat. nos. 2017.0-8), and Selected Characteristics for Urban Centres/Localities (Cat. nos. 2016.0-7). Extensive reference information about the Census will also be available on the Internet.
THE CENSUS TIME CAPSULE
A feature of the 2001 Census was the opportunity given to people to choose to have their name-identified Census information kept for release after 99 years. The ABS will microfilm the full Census records of those people who chose to have their information retained and then pass the microfilm to the National Archives of Australia for retention. On 7 August 2100 the records will be made available to the public. The name-identified records for those who do not choose to have their forms retained, or who do not answer the relevant question, will be destroyed. The project is known as the Census Time Capsule.
The Census Time Capsule project originated in 1999 with an inquiry into the treatment of Census forms by the House of Representatives Standing Committee on Legal and Constitutional Affairs. The Government agreed with the Committee’s recommendation that saving name-identified information from the 2001 Census ‘for future research, with appropriate safeguards, will make a valuable contribution to preserving Australia’s history for future generations’.
The ABS was concerned that the quality of the Census would be affected if all Census forms were retained even with appropriate safeguards. It was far more comfortable with the final decision to allow Australians a choice of whether their Census records should be retained or not.
The Census and Statistics Act 1905, and the Archives Act 1983 , have been amended to provide for the retention of information for those who elect to do so and to ensure that the name-identified Census information is retained from the 2001 Census, and that the name-identified information will not be available for any purpose within the 99 year closed access period, including access by a court or tribunal.
The Census question (question 50) which provides the opportunity for individuals to indicate that they wish to have their name-identified information retained, the associated supporting material, and the Census Time Capsule project procedures were developed in close consultation with the Federal Privacy Commissioner.
In collaboration with historical and genealogical organisations, the ABS developed an extensive education campaign, to be conducted as part of the 2001 Census communications campaign. The focus was on ensuring that people were aware of and understood all aspects of the Census Time Capsule project so that their decision on form retention was an informed one. While not advocating a particular position, the education and general public relations campaigns were designed to encourage people to participate. The Minister for Financial Services and Regulation and the Australian Federation of Family History Organisations participated in this campaign.
The ABS has endeavoured to make the project a success and, at the time of writing this article , the Census Time Capsule project has not caused significant privacy concerns. However, the real test of its success will be:
(a) whether sufficient Australians have elected to have their form retained to justify the cost of the project; and
(b) whether the quality of the Census data has been affected - the major concern is anecdotal evidence that some people have enhanced their ‘status’ as a result of the Time Capsule Project .
The ABS will be able to analyse whether there has been any significant deterioration in quality when Census outputs become available in 12-18 months' time.
PUBLIC AWARENESS AND ASSISTANCE
Prior to the 2001 Census the Australian Bureau of Statistics conducted a comprehensive communications campaign to ensure high levels of awareness and participation in the Census. As well as paid media advertising and briefings for community and other opinion leaders, the campaign included a comprehensive program of media contact to inform journalists and commentators about the Census and to encourage accurate and timely coverage of relevant issues.
An important aspect of the communications strategy was to raise awareness of procedures available to obtain help, especially for people who may have been disadvantaged through language or other difficulties (such as the deaf and hearing impaired, the sight impaired) in acquiring information about the Census or in completing the form. Extensive liaison was undertaken with relevant organisations and community resources identified as being able to assist people in understanding the Census or completing the form.
A principal avenue for assistance was the telephone Census Inquiry Service which operated throughout Australia during the Census period. The inquiry service answered inquiries ranging from questions about the Census itself, through delivery and collection issues, to requests for help in completing the Census form. Brochures providing information about the inquiry service were produced in 22 languages and a telephone interpreter service was available for those who required it.
A significant development in the 2001 Census has been the supplementation to the Census Inquiry Service with a separate Census website. This was the first time an externally developed and hosted website has supported an ABS activity and has provided the opportunity for learning in a wide range of areas within the ABS. In the week leading up to and including Census day 2001, 27,025 users sought advice/information on the 2001 Census or previous Censuses.
2001 CENSUS: INDIVIDUAL INQUIRIES HANDLED IN THE WEEK PRECEDING (AND INCLUDING) CENSUS DAY
These channels of assistance were additional to the Census Guide, a copy of which was distributed to every household along with their Census form.
As a result of the comprehensive communications campaign, the ABS found that nearly all individuals approached by Census collectors were aware of the 2001 Census. The coverage by the media was very extensive with saturation coverage achieved across Australia in the lead up to and on Census day. The coverage was very supportive to the ABS with around 95 per cent of messages either positive in tone or reporting key Census messages as news content.
Since the 1961 Census, Australia has had a Census every five years. A 1977 amendment to the Census and Statistics Act 1905 requires that ‘the Census shall be taken in the year 1981 and in every fifth year thereafter’.
All Census information is collected under the authority of the Census and Statistics Act 1905 and is, in turn, protected by its secrecy provisions.
In seeking Census information, the ABS relies on the willing cooperation of households in its initial approach. Traditionally, the ABS has always found widespread community support for the Census and most people participate willingly. The Census and Statistics Act 1905 does provide the Australian Statistician with the power to direct persons to provide the information sought in the Census but it has only been necessary to use these powers sparingly in the past - there were less than 50 prosecutions in the 1996 Census.
CHANGES IN CENSUS CONTENT
Consultation with users of Census statistics about the content of the 2001 Census commenced in March 1998 with the publication of an Information Paper 2001 Census: Australian Bureau of Statistics Views on Content and Procedures (Cat. no. 2007.0). Public consultation on possible topics was undertaken throughout 1998, with meetings in all States and Territories. Specialist user groups were established to examine in detail the ethnicity questions (such as language, birthplace and ancestry), transport questions (such as mode of transport to work and workplace address) and the possibility of including questions on disability. Extensive field and focus group testing on possible new questions, changes to current questions and overall form design was undertaken.
Valuable advice on possible topics was obtained from the Australian Statistics Advisory Council.
The outcome of the above process was that the 2001 Census comprised 50 questions, three more than in 1996. Two questions were included for the first time: use of the Internet; and home use of personal computers. A question on ancestry was included for the first time since 1986. The purpose of this question was to provide a better picture of the ethnic origins of the Australian population, particularly where this does not correspond with the person’s country of birth or parent’s country of birth. As a consequence, the 1996 question on country of birth of parents was able to be simplified in 2001 to ask only whether parents were born in Australia or overseas.
One topic, the number of children ever born, is justified only every ten years so it was omitted from the 2001 Census. Another topic from 1996, whether a dwelling is rented furnished or unfurnished, was not considered of sufficient priority for inclusion in 2001.
Despite a strong need for disability data, testing of a variety of disability questions showed that these did not produce reliable results using a Census methodology. This topic was, therefore, not included in the 2001 Census, however the ABS will continue to investigate the options for developing suitable questions for future censuses.
The Government announced the 2001 Census topics on 6 April 2000.
A relatively small group of permanent ABS staff undertake the planning of each Census, however the actual event calls for a very large temporary work force. For the 2001 Census nearly 23,000 collectors took to the roads and footpaths to deliver and pick up forms from every household in the country. In addition, several thousand special collectors were required to cover all those people who were staying in accommodation establishments and institutions, camping out or travelling overnight on the night of 7 August. The field operation was supervised by about 3,000 supervisors and coordinators employed for several months to organise recruitment, training, distribution and collection of forms. In total, over 850 tonnes of material were distributed by a convoy of trucks to the field operators.
Special attention was given during the Census to ensuring an accurate count of Indigenous people, whether living in Indigenous communities or in urban areas. In many communities, a specially designed form was used which was completed by an Indigenous interviewer. Indigenous collectors were also appointed in urban areas with high concentrations of Indigenous people. As well, Indigenous Assistants were available to assist collectors if required, bringing the proportion of Indigenous people in the Census work force to five per cent.
Another feature of the 2001 Census was the increased attention to identifying and counting the homeless, wherever they were on the night. The 2001 procedures were built on the pilot work undertaken in the 1996 Census. As in 1996, the 2001 Census data will be used to estimate the extent of homelessness in Australia.
The Census field operations always present a number of unusual situations. A special delivery of forms was sent with the Antarctic supply ship in late 2000 to ensure those spending the winter on Australian bases in Antarctica are counted. These forms were completed by expeditioners on Census night - 7 August, and will be returned with the first boat back.
About 10 million completed Census forms will be delivered to the Data Processing Centre (DPC) in Ultimo, New South Wales. Processing of the forms is being undertaken by a workforce of about 900, including 800 temporary staff recruited solely for this purpose.
The 2001 Census is being processed using the latest technology. Scanners are being used to take images of the forms which are being moved via the DPC’s computing network, to workstations as required. Answers to questions, both marks and numerical and alphabetic characters, are being automatically recognised by intelligent character recognition software and automatically coded by ABS developed software. Workstation operators will only be required to examine the images of forms where characters are either not recognised or when answers on the Census form cannot be automatically coded.
So that basic Census data can be released as soon as possible, processing is being split into two stages. In the first stage topics such as age, sex and religion, for which a high degree of automatic coding can be achieved, are processed. This covers eighty per cent of the topics on the Census form and first results for these topics will be released in July 2002. The second stage of processing deals with topics such as industry, qualifications and occupation, which require a large degree of manual intervention to code responses to the appropriate categories in the standard classification schemes. This will be done via Computer Assisted Coding developed by the ABS. Output products containing these more complex data items will be released progressively throughout the remainder of 2002.
Once all Census data are processed the paper forms will be pulped and converted to cardboard. Once microfilmed copies of the name-identified information for those consenting to retention are supplied to the National Archives of Australia, all ABS computer records containing names and addresses will be destroyed.
A number of innovations have been introduced in the 2001 Census:
|Telephone Census Inquiry Service calls|
|Internet - individual users |
- improvements were made to field procedures to ensure that collection activities were focused on the primary task of collecting a completed form from every household and that, at all stages, the best available practices and technologies were employed;
- the ABS spatial database was upgraded to a more advanced Geographic Information System (GIS) data model. This enabled the ABS to capitalise on its investment in GIS both in the Collection District (CD) design process and in the production of field maps for use by Census collectors. CD design was completed in a distributed GIS environment which fed seamlessly into the map production process. State of the art GIS technology was then used to produce two copies of each of the 40,200 full colour field maps. The map authoring and printing system enabled three to four people to produce up to 1,000 high quality maps per day;
- a new system was developed to simplify administration and payment of collection staff;
- in the Census Inquiry Service, Interactive Voice Recognition was used to provide a direct link between householders who have a concern about delivery or collection of their Census form and the Census field staff in their area. While this combined a number of existing technologies, the solution was a world first, which will have a number of other applications. For the Census, it greatly increased the capacity of the Census Inquiry Service so that it was able to handle a much larger volume of calls than was possible in 1996;
- an Internet site was established to provide information about the Census, and to allow householders to inform the ABS that they had either not received a Census form or the Census form had not been collected;
- imaging of Census forms will greatly reduce the volume of paper handling during processing;
- intelligent character recognition and automatic coding are all new applications enabling faster, more efficient and more accurate coding of responses provided by householders; and
A post-enumeration survey will be conducted two weeks after the Census to ascertain what proportion of the population was missed in the Census and what proportion was double counted. Historically, the net undercount in the Census has been of the order of 1.6 per cent, one of the lowest rates in the world.
An evaluation program has been designed as an integral part of the planning, development and operations of the 2001 Census. The objective of the evaluation program will be to ensure that all aspects of the Census perform as they have been designed to do, and that any lessons learned are captured and provide input to planning for the 2006 Census. The first phase, which will evaluate the development and collection phases of the Census, will be completed by late 2002. The remaining two phases of the evaluation, which will cover processing and dissemination, are scheduled to be completed in 2003 and 2004 respectively.
- much greater emphasis will be given to the dissemination, including free dissemination, of data via the Internet. In particular, a Census table specification service will be provided on the Internet which will allow users to design their own customised tables at times convenient to them.