Thirty years of the FOI Act: service, overhaul or refit?

John McMillan, Information Commissioner: Presentation to the Australian Corporate Lawyers’ seminar, Canberra, 29 November 2012

This talk gives me an opportunity to bring two themes together – the 30th anniversary of the FOI Act that occurs this week; and the review of the Freedom of Information Act 1982 and the Australian Information Commissioner Act 2010 by Dr Allan Hawke AC.

It is appropriate start by reflecting on why the FOI Act is fundamentally important in Australia. I can best do this by recounting some personal anecdotes that I shared at the 30th anniversary seminar that was hosted earlier this week by my office.[1]

In the 1970s, during the campaign for the FOI Act, I made over thirty requests to government agencies for reports and other documents referred to in the media. I was refused access to most.[2]

A common agency response, particularly to refuse requests for interdepartmental committee reports, was (to quote the Public Service Board response) that ‘the established practice is that all policy advice from officials to ministers should remain confidential’. Another request for details of the number of interdepartmental committees was rebuffed for the reason that ‘it is convention not to answer such a question’.

The Minister for Primary Industry, rejecting a request for a copy of a licence issued to a whaling station, advised that ‘Once a licence is issued it is the property of the company to which it is issued’. A subsequent request for the departmental copy of the licence was refused on the basis that ‘public access is not normally allowed to government records less than 30 years old’.

Requests for mundane documents were refused on spurious grounds. The Australian Bureau of Statistics, for example, responding to a request for a copy of the contract for pot plant hire, advised that before providing access ‘I am required to know your reasons for seeking the information’. Similarly, the Department of Business and Consumer Affairs not only refused a request for copies of submissions from industry groups (I was advised to approach each company for a copy of their submission) but declined to divulge the names of companies that had made submissions.

Those responses were received in the years following the Government announcement that Australia was to have an FOI Act, but before its enactment. They show why an FOI Act was needed, and how it changed the culture of government and the ability of people to obtain information that should be in the public arena in a functioning democracy. Since 1982, nearly 1 million requests have been made to Australian Government agencies and it is now routine for government agencies not only to release but to publish documents of the kind I was refused in the 1970s.

At the 30th anniversary seminar I also quoted from a brief message of support for the anniversary from legendary United States consumer advocate Ralph Nader:

The FOI Act is one of the most effective tools the public has to obtain government information … The 30th anniversary of [the Act is an occasion to] remind all who have used and benefited from this important law to challenge arbitrary government secrecy of its importance to a democracy. … We should constantly strive to use, strengthen and expand this wonderful law.

Let me now take up that challenge of how we improve the operation of Australia’s FOI Act. A submission from the Office of the Australian Information Commissioner (OAIC) to the Hawke review will propose a large range of drafting and procedural changes that I will not dwell on in this presentation. Quite simply, the FOI Act is not easy to interpret or administer, reflecting the fact that the substantial 2010 amendments were woven into an existing Act that was more than 25 years old.

On important issues such as calculating the processing period, granting extensions of time and third-party consultation, the OAIC has spent countless hours working out what the Act means in order both to clarify our compliance oversight role and to provide reliable guidance to a couple of hundred agencies that are struggling with similar issues.

I will outline three themes that will be prominent in the OAIC submission. Our primary aim is to raise these issues for broader discussion. The Hawke review will not report until April 2013, and there will be an opportunity for broader discussion about the merits of the ideas I will outline in this presentation and whether other approaches may be preferable.

FOI processing requirements

Difficulties in FOI processing stem from three important changes that were made in 2010 – removal of FOI access charges, allowing requests by email, and removing the ability of agencies to impose FOI charges if a decision is not made within the authorised processing period.

We support those changes and will not recommend they be wound back, other than a graduated reduction of access charges for delayed decisions.[3] However, the 2010 changes had downstream effects that do need to be addressed.

The FOI processing period should be calculated in working days, not calendar days. The processing period starts running when a request is received in the email inbox of an agency – which may be a weekend or public holiday, including Christmas Day before a long shutdown. It is inappropriate that the processing period can effectively be reduced in that uncontrollable manner.

We will also float the idea that at the front end of the FOI processing period there should be a short period within which an agency can talk informally to an applicant and clarify the scope of a request before the FOI clock starts running. This is not dissimilar to a proposal the OAIC made earlier this year in a report on FOI Charges. That report recommended that an agency be authorised to impose an access charge if an applicant does not first use an administrative access scheme that is published on an agency’s website.[4]

We were always uncomfortable about tying the development of administrative access to FOI charges, and there may be a better approach. The suggestion we will put forward is that upon receiving a request an agency can either accept it as a valid FOI request, and if so the clock starts running; or within a defined period (say, seven to ten days) contact the applicant to say ‘we want to talk about your request, its scope, its clarity, or whether we can provide information more suitably in another form’, and if so the clock will not start running until that defined period elapses. The earlier the agency contacts the applicant the more time it will have to discuss the request with the applicant.

There is the related problem of requests that are not really driven by a transparency objective, but are made by applicants who in truth want to insult the staff of an agency or make their work-life more difficult. This has become a problem. The only options at the moment in the FOI Act are to deny that it is a valid FOI request, which can be a risky strategy; use the charges or practical refusal mechanisms, which themselves can be formal and time consuming; or apply to the OAIC to have the applicant declared vexatious, which is a high bar to cross.

An option may be to give an agency an explicit power to reject an individual application on the ground that it is not a valid FOI request, is vexatious or frivolous, or lacks clarity. The applicant would then have two options – reframe the request, or appeal to the OAIC on an issue that is far simpler to review than existing options.

The FOI framework

The Hawke review provides an appropriate opportunity to reflect on the suitability of the FOI framework against the backdrop of other changes that have occurred in government and record keeping over the last thirty years.

Two examples of what I have in mind were raised earlier this year in the FOI Charges report. One was to build the concept of administrative access into the FOI scheme, so that an applicant need only rely on their formal legal rights to ask for a document when dissatisfied with an agency’s response to a less formal request for information or documents. I have already mentioned one way this could be done (by giving an agency a ten day window to discuss a request with an applicant). Since the Charges report the OAIC has also published general guidance on Administrative Access options.[5]

Another proposal was to place a cap of forty hours on an agency’s obligation to administer a person’s individual request. This has been the most controversial of the proposals in the Charges report, but let me give a simple example of why I think an hourly cap is a sensible and defensible approach.

A recent request published on a new Right to Know website[6] was from the Australian FOI commentator Peter Timmins, addressed to the Attorney-General’s Department asking for ‘a copy of the latest document assessing the arguments for and against Australia applying for membership of the Open Government Partnership’. He was given a document that is informative, but importantly contains enough information to enable him to ask further questions, or to make a supplementary FOI request.

By contrast, many requests to agencies adopt the more blanket approach of seeking ‘all documents about person X or on topic Z’. In some of the cases seen by the OAIC the agency has identified thousands of documents, that may include multiple copies of the same email, or documents that are uninformative or have been overtaken by other developments that are noted in later documents.

I query whether accepting and administering requests that are not specific or targeted strikes an appropriate balance between an applicant’s access rights and the administrative burden on agencies. The need to strike an appropriate balance is becoming greater in an age when agencies digitally record more information, there are more avenues available to the public to obtain information and documents from agencies, and agencies are generally more responsive in talking to people and providing information upon request. Importantly, a forty hour cap would not be the outer boundary for information access rights. It would be open to a person to make subsequent FOI requests informed by the information they have already learnt or discussions they have held with an agency.

Another example of how we can rethink the FOI scheme was put forward in a recent OAIC submission to a Senate inquiry on legislation to exempt Parliamentary Budget Office information from the FOI Act.[7] We proposed that the exemption operate for a defined period until after the next election (for example, one year after the next election, or four years following the date of creation of the document). There may be a good reason why a document should initially be exempt from disclosure at the time it is created, but that reason may not carry the same force for the next 20-30 years until the document reaches the open access period under the Archives Act 1983.

The same approach might be applied to incoming government briefs, and Senate estimates and parliamentary question time briefs. The character of those documents is such that they are framed as advice to one or two people, for an immediate, particular and temporary purpose. It may be that the document will be a very different document, and robbed of its very purpose, if the prospect of immediate FOI disclosure means that it is written instead as a public communication to an audience of many. Again, however, the strength of that reason for non-disclosure must surely erode after a period that is much shorter than the archival open access period. It may be simpler all round to have clear legislative rules on disclosure that balance public access rights against governmental interests.

Another issue that will be noted in the OAIC submission is the recent FOI Act amendment in Great Britain that will provide a right of access under the Act to unrefined electronic datasets.[8] The scope of this right is to be spelt out in a code issued under the Act that describes the type of datasets that can be requested, licensing and formatting datasets for re-use, and access charges. It may be that Australia is not ready for that change, but in a digital age in which the open data concept gains strength, we must consider whether it is time to transform or extend the FOI Act so that it applies to requests for information as well as documents.

Role of the OAIC

In the last two years the OAIC has finalised 478 IC reviews and 220 FOI complaints, in addition to publishing guidelines and fact sheets, arranging conferences and seminars, taking a prominent open government advocacy role, and discharging a privacy regulator role.

It is well-known, however, that we have a backlog of another 395 IC reviews and 68 complaints. Resources are clearly an issue, but our legislative framework is also a hindrance. For example, only the three Commissioners can resolve a case through a published decision, and we will recommend that that function be delegable to Assistant Commissioners. We will also recommend a more flexible range of powers to enable the OAIC to remit matters to agencies for further consideration, to resolve matters by alternative dispute resolution methods, or to dismiss matters that do not warrant formal review or investigation.

A larger issue that I invite others to consider in any submission to the Hawke review is this. The OAIC is a new model for promoting open government and for resolving disagreements between the public and agencies. We have a larger range of responsibilities than certainly other Australian and nearly all foreign access to information commissioners. Our FOI responsibilities and activities include complaint investigation, IC reviews, managing extensions of time, conducting audits and surveys, publishing guidelines and fact sheets, managing an Information Advisory Committee, hosting an Information Contact Officers Network, arranging discussion groups, making regular presentations to forums in and outside government, and general promotion of open government.

There are occasional comments that the OAIC has too many functions, and that some functions (such as training and advocacy) are incompatible with others (such as impartial merits review). Personally I disagree. I think the OAIC model represents an exciting new model in government oversight and regulation that goes beyond constrained legal models that are more traditional.

A related question is whether the OAIC is trying to do too much, particularly with its limited resources. Would it be better, for example, if the OAIC in its FOI function concentrated only on its merit review function, as the Administrative Appeals Tribunal has done? Or is it better, as in the OAIC model, to situate the review function within a broad range of activities?

I ask you to consider whether this extra work by the OAIC adds value to information management and policy, particularly in furthering public access rights to government information and documents. If you agree, please communicate your views to the Hawke review. By all means consider the alternative of recommending that the FOI scheme and the OAIC be wound back. Either way, I hope it has been helpful to be given this snapshot of some ideas the OAIC will raise for further public discussion in its submission to the Hawke review.