Consultative Processes Pay Dividends
The Case of Census 2006
Version of 24 March 2005
Published in Privacy Law & Policy Reporter 11, 6 (April 2005) 175-176
This document provides a summary of events of March 2005. Briefly, the Australian Bureau of Statistics (ABS) proposed very substantial changes to the Census. The Bureau conducted preliminary consultations with a small number of privacy and civil liberties organisations. The Australian Privacy Foundation (APF) made a speedy submission, and met with the Australian Statistician and his staff. As a result of that submission and meeting and other feedback received, the ABS decided to withdraw the proposals.
Until 1996, the Australian Census was always collected in an identified manner, but the identifying information, including address, was retained only as long as necessary to enable administration of the collection process. After capture and checking, the forms on which the data was collected were also destroyed.
In 2001, people were offered the option of having their record, including name and address, passed to the National Archives of Australia, to be released to the public in 99 years. The ABS did not retain name and address information. 54% opted in to this arrangement.
This approach corresponded to the position in the U.K. from the early 1800s until 1970, but with the added protection that retention was ‘opt-in’ / consent-based. The APF expressed caution about this option at the time; but because the retention was consensual, and the information provided about the option was sufficiently comprehensive, it did not oppose the new arrangements.
The Original Proposal for Census 2006
The ABS proposed substantial changes with effect from the 2006 census. As summarised in the Australian Privacy Foundation’s analysis, the changes were:
- Identified data would be kept on all people (which has never occurred before. Even in 2001, only about half the population ‘opted in’)
- Identified data would be kept indefinitely by the ABS (which has never occurred before. Even in 2001, identified data on the people who ‘opted in’ was sent to Archives, for release only after 99 years)
- Identified data would be used by the ABS, although names and addresses would be kept physically separate from the data
- Data from successive Censuses would be combined to build a ‘longitudinal census data-set’ for each person
- Data from many other sources would be added to it, subject to approval by a Privacy Reference Group whose decision criteria would include a strong public interest element
- This would start with other data held by ABS
- The database would include births and deaths register data (which ABS already has access to)
- The database would include data from the many disease registers (which would be a new source of data drawn on by the ABS)
- Uses of the ‘longitudinal census data-set’ would be transparent
The ABS was preparing to launch the proposal publicly. In the lead-up to the launch, it discussed the matter with Privacy Commissioners, and with some privacy and civil liberties organisations. The APF is not aware which organisations were consulted.
The APF received a copy of the proposal on the morning of Friday 4 March, together with an invitation to meet and discuss the proposal. On the evening of Tuesday 8 March, APF provided a submission to the ABS, which, due the seriousness of the matter, it submitted directly to the Australian Statistician, Dennis Trewin.
The submission contained an analysis of the changes that the proposal entailed. This concluded that:
- the proposal would effectiviely see the creation of a Central Population Register; and
- function creep was inevitable, because such centralisation of data would produce a ‘honey pot’ that agencies and corporations would be strongly attracted to.
The APF utterly rejected all elements of the proposal. It suggested that the Australian public would resoundingly reject it too, in particular the intention to centralise and retain permanently all personal data under government control, and the inevitability of an accompanying identifier.
The APF submitted that the proposal would seriously damage the hitherto strong reputation of the ABS. The APF recommended that the proposal be immediately withdrawn from consideration.
The ABS invited the authors of the APF submission, Roger Clarke and Chair Anna Johnston, to a meeting on Friday 11 March. The meeting was attended by the Australian Statistician, Dennis Trewin, the First Assistant Statistician, Siu-Ming Tam, and the Head of the Census and Geography Branch, Paul Williams.
The ABS explained that it believed that the changes would result in a data collection valuable for research. Fiona Stanley and other researchers were referred to as being supporters of the proposal. The APF explained its position, and clarified several aspects of its submission.
The ABS noted that legislation was not needed to authorise the changes, because the protections that were to be removed are ones of policy and practice, and are not statutory. However, the ABS stated that they intended asking for legislation to establish the additional protections that they envisaged.
At the conclusion of the meeting, it was clear that the ABS still intended to release a Discussion Paper, as originally planned, on 23 March. Although it would be modified in light of the discussions, it would still have the proposed features, at least as one option.
The APF Board commenced the preparation of communications to all privacy and civil liberties organisations, all Privacy Commissioners, and its substantial media list, intending to make a public statement to match the release of the Paper.
The AFR Article
On Wednesday 16 March, Rachel Lebihan, a senior journalist from The Australian Financial Review (AFR), became aware of the proposal.
She spoke to ABS, and then contacted Anna Johnston, in her APF role. As APF had not yet planned any media comment, Rachel was asked to explain what she had found out from the ABS. APF then provided limited comment in response to that information.
The AFR published a front-page article on Thursday 17 March, entitled ‘Census: they’ll know who you are‘. In addition to the APF, it quoted David Vaile, as Executive Director of the Baker & McKenzie Cyberspace Law and Policy Centre at the University of N.S.W. (rather than in his capacity as Vice-Chair of the APF), and Terry O’Gorman, President of the Australian Council for Civil Liberties.
Withdrawal of the Proposal by ABS
On the evening of Wednesday 16 March, Dennis Trewin contacted Anna Johnston and Roger Clarke, stating that the ABS had decided that retention of names and addresses was no longer an option, and would not be put forward in the Discussion Paper. He said that the ABS had accepted the arguments put forward by the Foundation and several civil liberties groups.
The decision had been made on the evening of 15 March, but the information reached the APF too late for a follow-up to Rachel Lebihan.
On Thursday 17 March, ABS published a Media Release, embargoed to 11:30am Canberra time. It stated that names and addresses would not be retained, and that the practice of destroying census records containing names and addresses following census processing would continue.
The AFR published a further article on Friday 18 March, headed ‘ABS axes plans to retain secret data‘. This again included quotations from Anna Johnston and David Vaile, and from Ian Cunliffe, a primary author of the Australian Law Reform Commission’s report on privacy published in 1983. A letter from the Australian Statistican was published in the same issue.
The APF believes that the ABS’s decision to perform a preliminary consultative round with key privacy and civil liberties organisations was highly valuable to all concerned.
Because of the long experience of its Board members, the APF was in a position to provide a comprehensive response very quickly. The comments of other privacy and civil liberties organisations clearly also conveyed alarm. The first newspaper article confirmed that this would be front-page news, and that the public response would be highly negative.
Early provision of sufficiently comprehensive information about the proposal enabled it to be re-evaluated before it did irreparable harm to public confidence in the ABS, and to the census.