1001.0 - Australian Bureau of Statistics -- Annual Report, 2006-07  
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 12/10/2007   
   Page tools: Print Print Page  
Contents >> Section V - Performance Information >> Chapter 11 - Provider / respondent relationships

Section V - Performance Information

Chapter 11 - Provider/respondent relationships


The work of the ABS would not be possible without the continuing cooperation of those in business, householders and others who provide the data that form the basis of the statistics produced. The ABS works hard to secure the trust and confidence of data providers, without which the provision of a quality statistical service would be jeopardised. The ABS continues its endeavours to reduce the burden placed on data providers by expanding the use of administrative data.

ABS Service Charters

The ABS recognises its obligation to respondents in statistical collections, and has produced service charters relating to both business surveys and household surveys. In addition, the ABS has a Client Services Charter for users of its products and services. For more information on the Client Services Charter see Chapter 13 Dissemination of statistics. The charters include performance standards for the relationship between the ABS and its clients, and for service delivery. Performance against these standards is subject to ongoing review, as are the charters themselves.

Business Surveys Charter

The Business Surveys Charter sets out the relationship between the ABS and businesses which provide information for statistical purposes. The charter explains how businesses can seek help from the ABS, and also specifies how businesses can request a review of the handling of any complaints.

The charter is regularly reviewed and is available on this site. The charter is provided to new respondents in collections involving businesses, and is produced in English, Italian, Vietnamese, Greek, Chinese and Arabic.

Household Surveys Charter

The Household Surveys Charter sets out the relationship between the ABS and members of the public who provide information for statistical purposes about themselves and their household. The charter explains what people can expect when dealing with the ABS and ABS interviewers. The charter includes information on the way household surveys are conducted, the way complaints are handled, and invites comments and feedback.

The charter, available on this site, is produced in English, Italian, Vietnamese, Greek, Chinese and Arabic.


The ABS depends on the goodwill and cooperation of Australians, businesses and other organisations to provide information in response to its many data collections. A critical way of maintaining such cooperation and goodwill is by ensuring that the information supplied remains confidential. The ABS has an enviable reputation for the preservation of the secrecy of reported information, and for the protection of its statistical data holdings from unauthorised release.

The Census and Statistics Act 1905 requires the ABS to publish and disseminate compilations and analyses of statistical information, and to maintain the confidentiality of information collected under the Act. The ABS meets the confidentiality requirements of the Act by ensuring that information provided is:

    • securely maintained
    • only used for statistical purposes
    • not inadvertently revealed in any published statistics, and
    • used only in unidentifiable microdata files when supporting research and analysis.
A statement outlining the importance of confidentiality of data provided, and how the ABS protects the confidentiality of data, is available on this site under Survey Participant Information - How the ABS Keeps Your Information Confidential.

The Census of Population and Housing is undertaken every five years, and the fifteenth census was conducted on 8 August 2006. As with all collections undertaken by the ABS, members of the public were made aware of the confidentiality measures being taken and they were reassured that their privacy would be maintained.

In the 2006 Census, Australians were able to elect to have their name-identified census data microfilmed and stored by the National Archives of Australia for release in August 2105. More than 11 million people (or 57 per cent of respondents) elected to archive their information.


In order to fulfil its mission of providing information to support discussion, debate and decision making, the ABS is authorised to collect information about a range of issues from businesses and households. The ABS is conscious that the needs of users for information must be balanced against the extent of requests for information from businesses and households. The amount of time needed to complete ABS surveys, totalled for all collections and all respondents, is referred to as provider load.

The ABS endeavours to minimise provider load by promoting efficiency and effectiveness of ABS operations. Information is not requested unless the collection is of high priority for the national statistical program. To further reduce provider load, forms are carefully designed and thoroughly tested. For business surveys, there is close liaison with representative groups, in relation to both the survey program and the demands on businesses.

Provider load in business surveys

The ABS has been involved in the Australian Government’s Taskforce on Reducing the Regulatory Burden on Business, and has considered the recommendations. The taskforce report did not have any recommendations specific to the ABS, and acknowledged the value of the ABS’ role as a clearing house for Australian Government surveys. More information on the Statistical Clearing House can be found in Chapter 10 Engagement with users and producers of statistics.

As seen in Graph 11.1, following the recommendations made by the Australian Government’s 1996 Small Business Deregulation Task Force (which found that the ABS only accounted for about 1 per cent of total business compliance costs), in 2004–05, the total ABS provider load on businesses fell to its lowest level of 393 000 hours. In the last two years, overall provider load has increased, due, in part, to the conduct of irregular collections. For example, the Survey of Employee Earnings and Hours, the Natural Resource Management Survey, and the Agricultural Census, with a load of around 150 000 hours, are irregular collections.

The average total load on all businesses during the past five years has been around 450 000 hours per year—27 per cent lower than for 1996–97. For small businesses, the average total load during the past five years has been around 200 000 hours per year—29 per cent lower than in 1996–97.

Graph 11.1 shows the unadjusted provider load (measured in thousands of hours taken to complete statistical forms) imposed on businesses by the ABS for 1996–97 through to 2006–07.

Graph 11.1: Unadjusted provider load imposed on businesses by the ABS

Graph 11.1: Unadjusted provider load imposed on businesses by the ABS
(a) Defined as businesses with less than twenty employees or a derived estimate of employees of less than twenty
(b) Higher provider load estimates for 2001–02 and 2006–07 reflect the conduct of the five-yearly Agricultural Census

For the first time since 1997–98, provider load for small businesses increased in 2006–07 to a level higher than medium and large businesses combined. This was due to the conduct of the five-yearly Agricultural Census, where 90 per cent of the total load for that collection was undertaken by small businesses.

Provider load figures from 1997–98 to 2006–07 were smoothed using an eleven-term Henderson moving average to reveal the underlying direction. This smooths the effect of irregular and regular contributors to provider load, such as the five-yearly Agricultural Census. Since 2002–03, the smoothed provider load estimates have increased only slightly for businesses of all sizes, as shown in Graph 11.2.

Graph 11.2: Smoothed provider load imposed on businesses by the ABS

Graph 11.2: Smoothed provider load imposed on businesses by the ABS
(a) Defined as businesses with less than twenty employees or a derived estimate of employees of less than twenty

Provider load in household surveys

Table 11.1 shows household provider load over the last four years. The household survey program varies considerably from year to year, with many major surveys on 3–6 year cycles. The ABS recognises the need to balance the pressure to expand the household survey program to meet the need for information about the population, with the load on providers, and has a range of measures in place to minimise provider load.

All ABS household surveys are rigorously tested to ensure they can be completed as accurately and quickly as possible; the minimum sample size to achieve reliable results is used; and the sample design is such that dwellings cannot be selected in more than one survey in a five-year period. For example, a dwelling selected in the Adult Literacy and Lifeskills Survey cannot be selected in the Time Use Survey in the five-year period.

The ABS focuses on developing the professionalism of the ABS interviewers, and building their skills in obtaining information from residents, with the minimum inconvenience. ABS interviewers are trained to be flexible, organise interview times that are most suitable for respondents, and enter responses in a computer in an efficient manner.

Table 11.1: Household provider load—approximate financial year data
Achieved interview time (minutes)
Total fully responding households
Total provider load (household hours)

2003–04Monthly Population Survey (MPS)(a)(b)
363 592
40 601
Survey of Disability, Ageing and Carers
14 323
10 360
Household Expenditure Survey (HES)
12 940
Survey of Income and Housing (supplementary sample to HES)
389 276
67 623
363 018
42 352
Multi-Purpose Household Survey (MPHS)
14 522
National Health Survey
19 502
13 001
National Aboriginal and Torres Strait Islander Health Survey (NATSIHS) (Non-Remote)
400 367
60 614
367 705
38 609
14 234
General Social Survey
13 404
10 120
Personal Safety Survey
16 430
Survey of Education and Training
13 857
Survey of Income and Housing
10 158
435 814
70 299
2006–07MPS—July 06 to May 07(a)
345 093
40 951
MPHS—9 months only to May 07
12 483
Survey of Employment Arrangements, Retirement and Superannuation
Adult Literacy and Life Skills
12 578
Time Use Survey
Post Enumeration Survey
32 731
19 475
Longitudinal Survey of Australian Children
11 252
Survey of Income and Housing
10 156
442 587
112 457

(a) Includes the Labour Force Survey
(b) Full implementation of Computer Assisted Interviewing (CAI) for MPS occurred in August 2004, January to July figures include Pen and Paper Interviewing (PAPI) and CAI combined average

Provider load in the 2006 Census of Population and Housing

Design of the census form is a critical strategy for maintaining a minimum provider load in the census. Forms are designed to be completed by one or more residents in each household, and there is a focus on ensuring provider load is kept to a minimum by:

    • clearly showing the questions that are applicable to each member of the household, and
    • focus group testing questions to ensure they are suitable for self-enumeration.
The 2006 Census saw the introduction of the eCensus, an online facility which respondents could use to return their census information. To minimise provider load, the eCensus only displayed questions relevant to an individual, whilst indicating the questions that were being missed.

The 2006 Census form, when compared to the 2001 form, increased by ten questions, in response to high user demand for topics.

Table 11.2 shows the average time taken to complete either a paper form or electronic form. For the paper form, this represents findings from testing of the census form, as time taken to self-enumerate is not measured during the census itself.

Table 11.2: Census provider load
Time taken on average (minutes)
2001 Census
2006 Census

To complete a paper form
To complete an eCensus form

a) For the paper form, the time taken for the 2006 Census is based on results from the 2004 major test, 22.9 minutes, plus an estimated 7.5 minutes for the additional ten questions included after the major test
b) The time taken for the eCensus includes login times as well as the time taken to answer questions, across all sessions


The ABS has had consistently high response rates for both household and business surveys. Table 11.3 shows that response rates for selected business collections has been consistently maintained.

It is important to note that in regard to business surveys, follow up procedures tend to focus on the more significant businesses. That is, those with typically high sales or employment relative to the rest of the industry. For example, the response rate for businesses in the manufacturing survey might be 92 per cent, but the businesses that have responded may comprise 96 per cent of the total employment in the industry.

Table 11.3 also demonstrates the continuing high levels of response gained for household surveys. Maintaining high response rates is a major focus for the household survey program, particularly given the increasing number of interviews that are conducted during the limited evening hours (up to 8 pm).

Table 11.3: Response rates for selected surveys
Response rate
2005–06 (%)(a)
Response rate
2006–07 (%)(a)

Business surveys
Retail Business Survey
Economic Activity Survey
Quarterly Business Indicators Survey
Manufacturing Survey
Survey of Tourist Accommodation
Agricultural Census
Survey of International Trade in Services
Survey of Research and Experimental Development— Business
Household surveys
Labour Force Survey
Multi-purpose Household Survey(c)
Survey of Employment Arrangements, Retirement and Superannuation
Adult Literacy and Life Skills
Time Use Survey
Post Enumeration Survey
Longitudinal Survey of Australian Children

a) Average response rates over the year for sub-annual surveys. Response rates apply to the year in which the data is collected
b) The previous Agricultural Census was conducted in 2001
c) The results for the Multipurpose Household Survey are used in several ABS publications. The survey is enumerated monthly and the response rates are a simple average
d) The previous Survey of Employment Arrangements, Retirement and Superannuation was conducted as the Survey of Employment Arrangements and Superannuation in 2000
e) The previous Adult Literacy and Life Skills Survey was conducted as the Adult Literacy Survey in 1996
f) The previous Time Use Survey was conducted in 1997
g) The previous Post Enumeration Survey was conducted in 2001
h) This survey was first enumerated in 2006.

The census consistently has very high response rates. These response rates are generally maintained despite usual collection problems and some new difficulties faced in 2006, such as more access problems because of a greater number of secure premises, and issues surrounding the recruitment and retention of census collectors, who were more difficult to attract and retain in times of low unemployment.

The completeness and the accuracy of the census is measured by a separate survey conducted specifically for this purpose - the census Post Enumeration Survey (PES). For more information on this survey see Chapter 5 Population and Social Statistics Program.

The PES is a household survey conducted about 3 weeks after census night. The results of the PES are used to determine how many people were missed in the census, and how many were counted more than once.

A number of improvements were implemented in the 2006 PES. Some of these include:

    • expanding the scope of the PES to include remote areas and discrete Indigenous communities
    • the use of a computer assisted interviewing instrument to replace the paper questionnaire previously used for the PES
    • a more computerised match-and-search system for processing the PES data, and
    • an improved estimation method.
For the 2006 Census, the PES identified a net undercount of 2.7 per cent, which is an increase compared to the 1.8 per cent net undercount from the 2001 Census. Care should be taken when comparing 2006 estimates with previous years, due to changes made to PES estimation and because of the inclusion of remote areas and discrete Indigenous communities for the first time in the 2006 PES.

More information on the PES is available on this site in Information Paper: Measuring Net Undercount in the 2006 Population Census, 2007 (cat. no. 2940.0.55.001) and Census of Population and Housing - Undercount, 2006 (cat. no. 2940.0).


Over 200 000 businesses and around 400 000 households are selected in ABS surveys each year. Households or businesses have a number of avenues available if they have queries or complaints about being selected in a survey, or about the processes involved. Free call contact numbers are provided with all survey information (preliminary approach letters and brochures for household surveys; survey forms and brochures for business surveys). Respondents can call these numbers and, for the majority of cases, their concerns are resolved by the ABS officers receiving the calls.

The census involved every person in Australia, and contact numbers were provided in all census material. People called these numbers for answers to queries they had, and to discuss any concerns.

In addition, there are a range of other complaint mechanisms, which can be used by households and businesses, including writing to or emailing the ABS, writing to the ABS Complaints Review Officer, approaching a parliamentarian, or contacting the Commonwealth Ombudsman. Information about the first two avenues is shown in the Table 11.4.

Table 11.4: Written complaints from respondents in ABS surveys

Complaints to the ABS survey areas

Business surveys

Letters to politicians (ministerial correspondence)
Household surveys
Business surveys

(a) The Census of Population and Housing is conducted every five years
(b) Information is not available for 2001–02

In 2006–07, the number of ministerials received for household surveys was around the same level as the number received in 2005–06. As household surveys are conducted by interviewers, either by phone or in person, most questions and complaints are dealt with directly and informally by survey staff, and do not result in formal written complaints.

ABS business surveys are predominately mail-based and the number of written complaints is noticeably higher than for household surveys. In 2006–07, a total of 290 businesses wrote to the ABS with a complaint. Many written complaints received from businesses relate to requests for exemption from ABS business surveys. In 2006–07, forty-eight requests from businesses for temporary exemption from ABS surveys were granted. As outlined in the ABS Business Surveys Charter, every complaint is responded to, either in writing or by telephone. If the complainant is not satisfied with the ABS response, they can take the matter to the Complaints Review Officer. During 2006–07 there were no complaints made about business surveys to the ABS Complaints Review Officer, and there were three inquiries from the office of the Commonwealth Ombudsman.

The number of written complaints declined in 2006–07. This reduction is attributed to the high profile of the ABS, with the additional media coverage associated with the 2006 Census and the Agricultural Census, and an increased number of respondent concerns resolved by ABS officers on the telephone.


Under sections 10(4) and 11(2) of the Census and Statistics Act 1905, the Australian Statistician may, by notice in writing, direct a person to complete a form or answer a question. Under section 14 of the Census and Statistics Act 1905, a person commits an offence if they fail to comply with a notice of direction. The penalty for this is set at one penalty unit ($110 in 2006–07) in respect of each day until the person complies with the direction. The application of this penalty is at the discretion of the court.

These provisions are used sparingly, as the ABS prefers to seek the willing cooperation of respondents. However, it is sometimes necessary to use the legislative provisions, to ensure that high response rates are maintained and the data provided are high quality.

The number of notices of direction issued and the number of prosecution actions approved in recent years are shown in Table 11.5

In 2006–07, the number of notices of direction issued for household surveys has increased. This has resulted from an increased focus on non-compliance. The ABS has a set of criteria to consider when deciding to issue a notice of direction for a household survey. This approach is reducing the instances of non-compliance and none of the eighty-six notices of direction to households proceeded to prosecution.

Whilst there is a high level of cooperation from business survey providers, both the number of notices of direction issued and the number of prosecution actions approved increased in 2006–07 from 2005–06 levels, reflecting the conduct of the 2006 Census, and a focus by the ABS on further improving business compliance.

The number of notices of direction and prosecution actions approved is higher every five years, when the Census of Population and Housing is conducted. Table 11.5 shows a substantial increase for 2006–07, compared to 2001–02, when the 2001 Census was conducted. This increase came about as a result of improved processes and procedures following evaluation of the 2001 Census. These improvements included streamlining this aspect of census operation, more efficient notification systems for census field staff, and quicker turn-around times for taking action.

Table 11.5: Notices of direction issued and prosecution actions approved

Notices of direction issued

Census of Population and Housing
Population Statistics Group including household surveys
Economic Statistics Group including business censuses and surveys

Prosecution actions approved(a)

Census of Population and Housing
Population Statistics Group including household surveys
Economic Statistics Group including business censuses and surveys

(a) Approved by the Australian Statistician for referral to the relevant office of the Director of Public Prosecutions. Any particular prosecution approval may pertain to a number of notices of direction. Each prosecution action approval is counted in the year in which the corresponding notices of direction are issued. Not every prosecution action approved proceeds to court.

Previous PageNext Page