So the Labour government will not introduce a Freedom of Information bill in the 1997 session. We have to await a White Paper first. It is not clear why. The issues are well known, both within government and outside. Freedom of information (FOI) legislation exists in Australia, New Zealand, and the USA, and its strengths and limitations are well documented. The Campaign for Freedom of Information has sponsored Acts and Bills covering several areas, including Mark Fisher's excellent Right to Know Bill, which was talked out in 1993. Mark Fisher is now a member of the Government. A streamlined version of his bill is available off the shelf. It need not be talked out again.

Academics have a more pressing interest in FOI than anybody else. Government is the biggest single source of information in Britain, and maintains enormous holdings relevant to science, economics, and society generally. Whole disciplines revolve around the study of government. But excessive secrecy means that both natural and social scientists are cut off from information that would enrich their research, and make their fellow-citizens safer and wealthier.

Yet academics have not made the running that one might have expected. Historians neither requested nor seem to have made much use of the recent liberalisation of the '30-year rule' at the Public Record Office. Since 1992, as part of the Conservatives' 'Open Government' initiative, as many as 50,000 documents which would not otherwise have been released are now in the public domain. They include some less than 30 years old, enabling analysts to probe such recent events as the 1966 Aberfan disaster (See I. McLean, 'It's not too late to say sorry', THES, 17.1.97). It seems that historians are content for the 'bow-wave of historical research' to foam onward just one year at a time in a stately progress. And neither political scientists nor social policy analysts seem to have been curious enough about the policy process in the UK. Natural scientists are often frustrated by the conditions of confidentiality and controls over publication that government departments attach to contracts. But in these hard times, few can reject them, even though they breach the canons of scientific research. Government departments sometimes hide information even from the scientists working for them (as when Professor Hugh Pennington was not shown the Meat Hygiene Service's devastating report on abattoir hygiene). Science policy academics go to the USA and get, under their FOI legislation, information on food safety in Britain which the Ministry of Agriculture will not release. The greatest step towards FOI in Britain so far was forced upon government not by academics but by a judge, Sir Richard Scott.

FOI policy involves both 'freedom of information about me' and 'freedom of information about other people'. The academic interest in FOI is in the latter. But the two may conflict. The right to see information about myself is controlled by the Data Protection Act 1984. In the 1997 Queen's Speech, the Government promised a Bill extending the 1984 Act to cover manually-held records. But this right may legitimately override the right for other people to see information about me. In consequence, people who wish to block third-party requests for FOI, such as those discussed above, may use data protection as an excuse. It is important, therefore, to dovetail data protection and FOI legislation.

Civil service advice to ministers is a controversial area of FOI. Some fear that the quality of this advice would suffer if it were to be made public. But in Australia the quality of both advice and academic investigation into government has improved since their FOI Act was passed in 1982. When the former Australian Prime Minister Paul Keating left office in 1996, he took with him ten filing cabinets of material. This was subsequently turned into a book by Peter Edwards, an adviser turned academic. Edwards' account of Keating as Treasurer and Prime Minster - laced with extracts from cabinet documents and ministerial correspondence - was judged to have provided important insights, rather than harmed government. Similarly, in Britain ever since the Scott report civil servants have had to presume that whatever they write may well be made public. Yet the system of government has been strengthened, not weakened, by Scott's revelations. If policymakers know that what they write may soon enter the public domain, they have an incentive to transparency and propriety when taking decisions. FOI represents a credible commitment to procedural faithfulness.. A system like New Zealand's, where government increasingly releases background material and internal advice once a policy decision has been taken, would allow natural scientists to review the bases, and social scientists the judgments, upon which decisions are made.

But freedom of information requires more than merely interest in its introduction. Success or failure is in the detail. In America FOI was first introduced in 1966. Four years later Ralph Nader complained that 'the FOI Act, which came in on a wave of liberating rhetoric, is being undercut by a riptide of agency ingenuity'. Costs (particularly crippling to research), delays, and the creative interpretation of exemptions, were being used to restrict access. The solution elsewhere has commonly been to create 'FOI Commissioners', whose role is to be an advocate for openness. The same would be needed in the UK.

Academics, then, need to turn their thoughts beyond the usual terms of debate, covering just access to information. Institutions matter. More basically, it is critical that information be both created in the first place, and retained once it is no longer needed by government. Thoughts should also be turning to how FOI will cope with the electronic age. Email traffic is not being retained adequately, and outsourcing of data handling reduces government control over its own data. FOI should concern academics as professionals and as citizens. It will not mean much if it relates to a world that has passed.

Iain McLean is professor of politics, Oxford University, and an affiliate of NEXUS. He is writing in a personal capacity. Greg Terrill is a visiting Australian academic at Nuffield College, Oxford.