|Page tools: Print Page|
Statistical Developments in 2004-05
The Hon. Chris Pearce, MP, opening the 55th Session of the International Statistical Institute
The need for the ABS to establish a program of research into new and innovative methods of collecting, analysing and exploiting statistical information has led to the formation of a partnership between the ABS and the University of Wollongong to establish a Chair in Statistical Methodology. The new Chair, a research position as a Professorial Fellow, is located within the Centre for Statistical and Survey Methodology. The Centre will undertake fundamental research and industry focused research with an emphasis on research relevant to official statistics. It will also undertake contract research projects involving statistical or survey methodology. A program of training courses in statistical methodology targeted at ABS staff will also be developed.
THE CENSUS OF POPULATION AND HOUSING
In 2004–05 significant effort has been committed to planning Australia’s fifteenth national Census of Population and Housing to be conducted on 8 August 2006. Numerous discussion papers have been released and extensive user consultation processes undertaken to ensure that the output from the census will reflect the current information needs of Australians.
Following the release of Information Paper: Census of Population and Housing, ABS Views on Content and Procedures, 2006 (cat. no. 2007.0) and an extensive user consultation process, the Australian Government agreed to include four new topics in the 2006 Census incorporating questions on the number of children ever born, need for assistance (disability), unpaid work, and household access to the Internet. As a consequence, the 2006 Census questionnaire will be one page longer. Extensive testing has indicated that the extra length of the questionnaire is not a concern to householders.
The Discussion Paper: Enhancing the Population Census: Developing a Longitudinal View (cat. no. 2060.0) presents a proposal to enhance the value of the census. The paper outlines how statistical methods could be used to combine the information provided in the 2006 Census with information provided in future censuses to create a Statistical Longitudinal Census Dataset (SLCD). An extensive consultation process was conducted, with many public submissions received. A decision was made to base the SLCD on a 5 per cent sample using statistical matching techniques. Name and address information will continue to be destroyed once census processing is complete, as has been the case for previous censuses.
Susan Linacre, Deputy Australian Statistician (seated, second left), chairing the Census Data Enhancement Steering Committee.
Information Paper: Census of Population and Housing, ABS Views on Census Output Strategy, 2006 (cat. no. 2009.0), was released to inform census users of the proposed strategies for 2006 Census products and services and to seek their views.
With the continued development in technology, particularly the Internet, and the growing sophistication of the user community, the ABS intends to recast its range of products and services for the 2006 Census. Most electronic products (including CDATA 2006) will only be available via the Internet. A range of new Internet based products will be developed which will make it easier for users of varying levels of sophistication to obtain the data they require. While the range of census publications will be reduced, the range of data that will be made available as standard census output will be expanded. It is proposed that all the profile tables that were released for the 2001 Census will be made available again for the 2006 Census.
Salmat Ltd, a leader in customer communications in Australasia, was chosen to provide call centre infrastructure to handle the 800,000 calls expected for the 2006 Census. Salmat, an Australian owned company, won the $4.5 million contract through an open tender process. Salmat will partner with Telstra to deliver technical aspects of the project, which will include the despatch of approximately 1.4 million SMS messages to census collectors, as well as the use of a Natural Language Speech Recognition technology assistant in the management of inbound calls.
IBM Global Services, one of the world’s largest consulting services organisations, has been chosen as the industry partner to develop the eCensus system for the 2006 Census.
The inaugural State Accounts User Group met to discuss state and territory level economic measurement issues. The State Accounts User Group was formed primarily to provide strategic and conceptual advice on proposed methodologies for the development of enhanced state account statistics, including production based gross state product chain volume estimates. The group also provides a mechanism for promoting discussion to directly address any issues or concerns that key State Accounts’ users have with existing methodologies and associated published data, and to identify gaps in the State Accounts dataset and advise on how these gaps might be closed.
Enhanced seasonal adjustment methodologies were introduced into balance of payments, housing, and other lending finance statistics by using concurrent seasonal analysis. These methodologies use available data at the current reference period to estimate seasonal factors for the current and previous months.
A new quarterly Survey of Superannuation was developed based on data from the Australian Prudential Regulation Authority. The ABS uses the data in the sectoral compilation of Australian National Accounts: Financial Accounts (cat. no. 5232.0) and also in the compilation of Managed Funds, Australia (cat. no. 5655.0). The Australian Prudential Regulation Authority uses the data to derive measures of the size of the superannuation industry and to monitor trends within the industry.
There was an issue of major concern during the year when, on 18 May 2005, the ABS released revised monthly retail trade statistics for the period July 2004 to March 2005. This followed a number of criticisms of the accuracy of the retail statistics. The ABS took these criticisms seriously and undertook a number of investigations into the accuracy of these statistics. Unfortunately the error was not detected until May 2005. On finding this error, the ABS published the revised statistics as soon as practicable, together with the impact on Gross Domestic Product.
An independent review was commissioned which identified several areas where the ABS could improve procedures. The ABS will implement the necessary changes.
The ABS released an experimental index measuring price changes for financial services acquired by households in Information Paper: Experimental Price Indexes for Financial Services, 1998 to 2003 (cat. no. 6413.0). The experimental indexes will be incorporated into the Consumer Price Index, Australia (cat. no. 6401.0) from September quarter 2005 in conjunction with the introduction of new weights as part of the 15th Series Review of the Consumer Price Index.
Information Paper: The Introduction of Hedonic Price Indexes for Personal Computers (cat. no. 6458.0) was released to present the methodology that the ABS proposes to use to adjust computer prices for quality change, and its plans for implementation. The ABS has long been aware of the need for improvement in price indexes for personal computers, and progress in this regard has been discussed in the ABS Economic Statistics User Group meetings as well as other forums. When faced with measuring prices for products that undergo rapid quality change, international best practice is to develop hedonic price indexes provided suitable source data are available.
The ABS has been actively assisting researchers and policy makers in key current projects aimed at improving the information base for children and youth, such as the Longitudinal Study of Australia’s Children and the Australian Research Alliance for Children and Youth. A series of information papers were released during 2004–05 on children and youth statistics.
During 2004–05 the ABS released Measures of Australia’s Progress: Summary Indicators, 2005 (cat. no. 1383.0.55.001), and Measures of Australia’s Progress: At A Glance, 2005 (cat. no. 1383.0.55.002). Extensive user consultation was undertaken to reduce the number of indicators to assist readers in gaining a quick overview of national progress. A suite of 15 headline indicators were chosen by the ABS based on statistical grounds, not on their relative values. Some readers have tried to infer an ABS view about the relative importance of the different aspects of Australian life from the number of aspects discussed under the social, economic and environmental headings, or from the number of headline indicators, or the number of indicators overall. No such inference can or should be drawn. It is not for the national statistical agency to determine what relative importance should be accorded to particular measures.
Enumeration for the 2004–05 National Health Survey (NHS) was completed. The survey collected information from approximately
16,800 households with some additional user funded sample supplementation. Topics collected included health status, risk factors, health related actions, and demography. Information from the survey is expected to be released in early 2006.
In conjunction with the NHS, the 2004–05 National Aboriginal and Torres Strait Islander Health Survey was undertaken. In the future, it is expected this survey will be conducted six-yearly to coincide with every second NHS. This survey builds on past ABS surveys which have provided Indigenous health information, including the 1995 NHS and the Indigenous supplement to the 2001 NHS. The survey was developed in consultation with representatives from a wide range of organisations representing Aboriginal and Torres Strait Islander interests. The information collected will assist in formulating, targeting, and monitoring the effectiveness of health policies and programs to improve the health outcomes of Indigenous Australians.
Changes were made to the processes used to code industry and occupation data in the monthly Labour Force Survey. These changes resulted in an improvement in the quality of estimates classified by industry and occupation. The classifications used and underlying coding methodology remained unchanged. Aggregate estimates of employment and unemployment were unaffected, and there was no meaningful change in the level of employment classified to any industry Division or occupation Major Group.
The ABS implemented Computer Assisted Interviewing (CAI) in the monthly Labour Force Survey. The use of CAI will increase interviewer productivity through improved survey field collection structures and systems, and will result in more flexible and cost effective household survey data collection.
There continues to be strong demand for high quality environment and energy statistics to inform policy debate and decision making. The ABS has established a Centre for Environment and Energy Statistics to: provide leadership in the development of environment and energy statistics; coordinate statistical activities including assisting with the development of frameworks and standards and improve data quality, comparability and coverage; and improve the understanding of trends and current issues, particularly across sectors, through its own analysis, and by supporting analysis undertaken by others.
The ABS is now responding to the increase in demand for regional level output environmental and land management statistics. The ABS has commenced work on a biennial Natural Resource Management (NRM) Survey beginning with survey content, the development of survey forms, and survey testing in readiness for a late 2005 despatch.
In parallel with the development of the NRM Survey, the ABS has undertaken development work on an alternative survey methodology for its agricultural collections based on a land parcel frame. This approach involves contacting the land owners of specific parcels of land, of a known location, and asking them to identify the activities occurring on that land, and any other land under their ownership or management. This is the opposite of what currently occurs where agricultural businesses are surveyed and then the business activities are attributed back to ABS defined areas, for example Statistical Local Areas.
The ABS released Innovation in Australian Business (cat. no. 8158.0) presenting results from the Innovation Survey 2003. This survey differed significantly from previous ABS surveys on innovation, with the scope extended to cover most Australian and New Zealand Standard Industrial Classification (ANZSIC) divisions and the use of a broader definition of innovation than currently used internationally. This broader definition is expected to be adopted in international standards from mid-2005.
The ABS continued its support in the development of international statistical standards for science and technology indicators. The ABS participated: as a member of the Organisation for Economic Cooperation and Development task force on Fields of Science classification for research and development statistics, and as a member of the working party on Indicators for the Information Society; as a participant in the June 2005 National Experts in Science and Technology Indicators meeting; and, through presentations to the biennial Asia–Pacific Information and Communication Technology Technical meeting.
Information Paper: ANZSIC 2006 Development, 2004 (cat. no. 1294.0) was released during 2004–05 outlining a joint project between the ABS and Statistics New Zealand to review the existing ANZSIC 1993, and develop a revised classification, ANZSIC 2006. This classification is widely used in both countries for the production and analysis of industry statistics. As well as industry based surveys, results from the 2006 Census of Population and Housing will be published based on the 2006 ANZSIC classification.
Preparation is underway for the 2006 edition of Australian Standard Geographic Classification (ASGC) — Electronic Publication (cat. no. 1216.0), which will be used for the 2006 Census of Population and Housing. The basic framework for dividing Australia into relevant geographic areas is provided in the ASGC, a hierarchical classification with a number of different types of areas to satisfy different statistical purposes. The ASGC facilitates the standardisation of terminology and the comparability of data across collections. Census data is available for a wide range of geographic areas ranging from the whole of Australia through to a Collection District of a few hundred households.
The ABS has developed a new geographical unit known as Mesh Blocks. These units will be four to five times smaller than the current smallest spatial unit — the Collection District. It is planned that Mesh Blocks will become the basic building block for all statistical, political and administrative regions in Australia. Currently there is a wide range of geographic units in use in Australia and many organisations have adopted their own geographical units to suit their needs, often without reference to units that are used by others. As a result, data cannot be readily integrated and exchanged between organisations and the development of Mesh Blocks will help address this problem. Mesh Blocks will be able to aggregate to any geographical region, resulting in more accurate demographic analysis, which in turn will lead to improved government policy formulation and service delivery. It is estimated Australia will be divided into around 200,000 Mesh Blocks compared to 37,209 Collection Districts. Facilities are being developed to enable an address coder to be used in conjunction with the Geocoded National Address File that will allow addresses to be coded to Mesh Blocks.
Developments in technology and methodology enable the ABS to improve its statistical activities in a number of ways — improving productivity, improving quality, better managing the relationship with data providers, and development of new statistical outputs. For the last three years the ABS has had a Business Statistics Innovation Program in place.
It was completed towards the end of 2004–05 and resulted in a number of important improvements, including greatly enhanced productivity. Another program is being put in place to build on the gains. Similarly, changes are being made to our household survey systems to enable productivity improvements, better quality and more timely statistics.
NATIONAL STATISTICAL SERVICE
During 2004–05 the Australian Taxation Office installed an ABS developed ANZSIC autocoder which has improved the speed and accuracy of ANZSIC coding of Australian Business Register records.
The National Data Network (NDN) is an ABS initiative that will provide infrastructure, protocols, standards, and services to support acquiring, sharing and integration of data across Australia. It is currently being developed to increase the availability, accessibility, and useability of information sources relevant to policy analysis and research — particularly key administrative and survey datasets held by the Australian Government and state and territory agencies.
During 2004–05 the NDN Interim Governing Board held its inaugural meeting. Key outcomes from this meeting included: agreement on the NDN exposure draft content to provide information to interested members and users; discussion of the ‘Draft NDN Guiding Principles’; endorsement of the ‘Terms of Reference’ for the Board; and articulating future directions for the development of the Network. In addition, the ABS released a series of papers relating to the protocols, guidelines and directions of the NDN leading up to the first release of the demonstration phase of the Network in April 2005. For more information about this initiative visit the new NDN web site on www.nationaldatanetwork.org.
An Information Development Plan for Crime and Justice Statistics was released and endorsed by the Boards of Management of each of the three statistical units within the National Centre for Crime and Justice Statistics. Each of the three Boards report through a National Officer Group to a Ministerial Council (Australasian Police Ministers’ Council, Standing Committee of Attorneys-General, or Corrective Services Ministers’ Conference). The plan seeks to formally develop strong partnerships and, through the establishment of a Steering Committee, to facilitate and guide development activities over the three year life of the plan. The Steering Committee has representation from all lead groups and will assist coordination and collaboration of the 62 activities identified in the plan, involving 20 groups from within and across different subject fields.