The big ethical lesson for all journalists from the Peter Costello dinner affair is this: be sure about the rules of engagement and never agree to change them retrospectively unless there is a substantial public interest in doing so.

Three senior Canberra press gallery journalists, Paul Daley, Tony Wright and Michael Brissenden, created for themselves an impossible ethical dilemma when they agreed retrospectively to treat a dinner conversation with Treasurer Peter Costello as “off the record” rather than as “background”.

These conventions – “on the record”, “off the record” and “background” – are well understood by journalists, politicians and press secretaries. “On the record” means that anything said or done in the presence of the journalist may be published. “Off the record” material may not be published at all. “Background” material may be published but not attributed in a way that reveals where it came from.

It appears from Tony Wright’s account on the website of The Age, where he is now National Affairs Editor, that the journalists assumed the dinner conversation with Costello could be treated as “background”. Only later did Costello’s press secretary at the time, David Alexander, insist that it was “off the record”.

Wright says he, Daley and Brissenden reluctantly agreed to give Costello the benefit of the doubt.

By agreeing to treat the conversation as “off the record”, the journalists were placing themselves under the obligation set out in clause three of their professional code of ethics:

Where confidences are accepted, respect them in all circumstances.

At the same time, in agreeing retrospectively to treat as unpublishable information that they originally considered publishable, they were breaching clause one of their professional code:

Do not suppress relevant available facts.

That three experienced and talented journalists could impale themselves so painfully on the horns of this dilemma reflects poorly on the attention given to ethical training in the profession.

In this case, it was their failure to clarify the rules of engagement and their subsequent retrospective agreement that caused the problem.

There are circumstances where it is ethically justifiable to agree to go “off the record” retrospectively, but it involves a public-interest test.

For example, Wallace Crouch, a senior writer with The Sydney Morning Herald, was interviewing on the record the general manager of Qantas, when a phone call came through to say a man calling himself “Mr Brown” was attempting to extort money from the airline by threatening to blow up two aeroplanes.

Strictly speaking, Crouch would have been within his rights to report what he heard of that initial phone call.

In fact, immediately after putting down the phone, the general manager asked that the matter be treated as off the record. During the rest of the afternoon Crouch was made privy to much else, including discussions between Qantas and the police, all of which of course was certainly off the record.

It became apparent within a few hours that the original threat was a hoax because no explosion occurred at the specified time and because of detectable technical flaws inherent in the threat. Although there was a threat to a second aircraft, the police and Qantas technicians assessed that to be a hoax as well.

Crouch and the Herald had to weigh two competing public interest considerations: playing into the hands of the extortioner by publicising the threat, and warning the public of a possible though improbable threat to the safety of an aircraft.

They took the decision to treat the entire extortion-related material as off the record and the paper therefore did not publish it until the extortioner – or hoaxer, as it turned out — had been caught and convicted.

The Peter Costello case fails the public-interest test.

No public interest was served by letting him off the hook. There was, if anything, a clear public interest in publishing because the conversation went to the heart of the functioning of the national government.

Moreover, when they are dealing with media-savvy people like politicians and their minders, journalists should regard everything as publishable unless there is a specific agreement to the contrary.

The Bulletin’s Paul Daley did finally find a way to publish the most damning of the Costello quotes: “He (Howard) can’t win. I can. We can, but he can’t.” He did so, according to the editor-in-chief John Lehmann, by sourcing it to “confidants” of Costello, that is, by not using the dinner conversation as the source.

This points up one of the many problems of going “off the record”: what if you get the same material in publishable form somewhere else? You publish, of course, usually to the chagrin of the person who spoke to you off the record in the first place.

Another problem is, what do you do if – as here – the source denies what he said, casts doubt on your professional integrity and, as it seems to you, misleads the public? Ethically, you remain bound by the undertaking of confidentiality, no matter how reprehensible you think it is that he should get away with it.

Yet in this case the outraged journalists then broke their undertaking of confidentiality, thus catching themselves on both horns of the ethical dilemma.

To paraphrase Stephen Fry in Absolute Power:

Never trust a politician, never shake hands with a gynaecologist, and never give a press secretary the benefit of the doubt.

*Dr Muller is a Visiting Fellow in the Centre for Applied Philosophy and Public Ethics at the University of Melbourne.