The new structure is due to be reviewed later this year. How is it travelling? This question was the subject of a recent seminar sponsored by the Institute of Public Administration Australia's ACT division, with the Information Commissioner, Professor John McMillan, as lead speaker.

McMillan's office can point to significant activity, in its internal procedures and within agencies, which have needed to establish their own information publication schemes and FOI disclosure logs. The overall objective is to encourage an attitude of proactive disclosure, in which the agencies take the initiate in publishing information. In this way, for instance, requests for information under the FOI Act would become the avenue of last, rather than first, resort. The publication of post-election briefs to incoming ministers and the Defence Department's publication of all ''hot issue'' briefs to its minister are significant examples of the new approach.

The aim is ambitious and long-term, amounting to a significant shift in public service culture, which has been instinctively secretive and suspicious of disclosure. As McMillan emphasised, his office also faces more immediate practical challenges, particularly because of mounting costs. New request and appeal procedures inevitably generate more work for hard-pressed staff, both in the office itself and in agencies.

Paradoxically, the change from written to electronic communication and documents has tended to complicate some aspects of information disclosure. While publication itself via the internet is much simpler and cheaper, the process of retrieving information has often become more complex. Instead of laying their hands on a specified document or documents in a file, staff must now follow a complex electronic trail of emails, links and attachments.

As staff resources are squeezed within the Australian Public Service as a whole, meeting the growing demand for public information will become increasingly problematic. Hard-pressed agency heads, forced to do more with less, will look askance at their growing information budgets. Some form of rationing, whether through charges or caps on numbers of requests, appears almost inevitable.

How far the new regime has actually changed public service attitudes to disclosure is debatable. Optimists can certainly point to some significant innovations. But many journalists and others who use FOI procedures regularly report little substantial change, particularly in relation to information that could be seen as politically sensitive. The amended Act was supposed to remove the blanket exemptions for confidential public service advice to ministers, including the use of conclusive certificates to prevent disclosure. The public interest test was deliberately tightened to favour disclosure of departmental documents that could be used to criticise government policy.

The amended act goes out of its way to specify factors that must not be taken into account in deciding whether access to the document would, on balance, be contrary to the public interest. These include whether ''access to the document could result in embarrassment to the Commonwealth government, or a loss of confidence in the Commonwealth government''. Experience to date, however, suggests little change on this front. Public servants are still resisting the release of documents, particularly adverse advice to ministers, that could provide ammunition to the government's media critics and political opponents.

Arguably, however, criticism of the public service on this score is ill-judged because the amended act itself is problematic. The original FOI Act was passed in 1982 as one of a set of laws aimed at improving government accountability and transparency. These changes became known as the new administrative law, which also included the Administrative Appeals Tribunal Act, the Ombudsman Act and the Administrative Decisions (Judicial Review) Act. To them we could also add the development of the Senate committee system in the 1970s as well as the auditor-general's adoption of regular performance auditing in the 1980s, which also opened up the executive, particularly public servants, to greater scrutiny.

At the time of their introduction, all of these reforms were criticised by traditionalists as being incompatible with ministerial responsibility. Strictly speaking, the criticisms were correct because the reforms certainly reduced ministers' status as sole spokespersons for their departments. However, for the most part, the new procedures were careful not to transgress on the core of ministerial responsibility; that is, ministers' right to determine policy directions and public servants' public loyalty to ministers. In Senate committees, for example, public servants answer questions about details of administration, in which ministers are not directly involved. But they are careful to avoid comment on matters of policy, for which ministers remain responsible.

A similar distinction between policy and administration underlies the respective jurisdictions of executive watchdogs, such as the ombudsman and the auditor-general, both of whom consciously avoid criticism of matters of government policy. The distinction is notoriously slippery and not subject to precise definition. It needs to be applied in context and requires a fine sense of political judgment about what will be seen as criticism of the elected government. In particular, it requires avoiding matters that are of partisan controversy, where disagreement with ministers will immediately give comfort to the government's political opponents. Drawing the line between administration and policy is difficult but officials certainly feel the heat if they cross it, as the previous ombudsman, Allan Asher, found to his cost.

In this way, the new mechanisms of accountability have been successful because they have acted as ancillary supplements to traditional ministerial responsibility without attempting to dilute its main principles. Ministers still remain responsible for policy while other members of the executive branch, including public servants, do not publicly question their political masters. Over the years, both ministers and public servants have learned to live with the greater openness of Senate committees and audit inquiries, without any threat to their professional relationships.

FOI has always been a cuckoo in this particular nest. Alone among the accountability innovations of the 1970s and '80s, the original FOI Act challenged the right of ministers to expect loyal support from public servants in public. In principle, it allowed members of the public, including journalists and the opposition, to request departmental documents that included analysis and advice potentially critical of government policy.

Unsurprisingly, public servants resisted such a frontal attack on their professional ethos. Their job, as they saw it, was to protect the standing of their ministers, not to connive at their public embarrassment. Public servants willingly used all available means, including the notorious conclusive certificates, to help frustrate this aim of FOI.

In response, the FOI lobby, dominated by media interests, regrouped and led a successful campaign to toughen the act. The amended act spells out the implicit aim of the original act: that departmental analysis and recommendations, other than those related to immediate discussions with ministers and cabinet, should be available to the public. In contrast to other accountability mechanisms that supplement ministerial responsibility, there is no general exemption for matters of policy, understood as matters that are politically controversial on which public comment by public servants could embarrass ministers and give comfort to the opposition. Indeed, it is precisely these matters that are the target of FOI campaigners. Predictably, many senior public servants have been reluctant to embrace these aspects of the new act and have continued to drag their feet.

Conflict over the release of politically sensitive information will continue. Journalists and oppositions will keep pressing for documents that will make the headlines while public servants will strive to protect their frank exchanges with ministers. Over time, as with other mechanisms of accountability and transparency, both ministers and public servants will adjust to greater openness. But FOI campaigners should not be surprised if the citadel of government secrecy does not fall at the first attack. They need to settle in for a long siege.

Meanwhile, supporters of a stronger information regime should be more immediately concerned about whether the current advances, modest though they may be, are wound back through lack of political support. Politicians of all parties show little inclination to justify spending on transparency as a core service to the public. The present information policy is largely due to the efforts of Labor senator John Faulkner, who, quite exceptionally, devoted his considerable political authority to the cause of open and ethical government. Since his departure, information policy has lacked an effective champion in cabinet and, after the last ministerial reshuffle, no longer has a minister with specific responsibility for it.

Unlike its Conservative counterpart in Britain, which has launched an impressive transparency drive, Australia's Coalition has so far revealed little interest in the issue. If elected, it will look hard for suggestions of how to cut the budget further. Top of the hit list will be the ALP's pet projects, including any new agencies for which the Howard government saw no need. Will the information commissioner's office survive? If it were abolished or radically restructured and downgraded into another agency, the open government agenda would certainly suffer. But would any minister (or any secretary) care?

Richard Mulgan is an emeritus professor with the Crawford School of Public Policy at the Australian National University.

richard.mulgan@anu.edu.au