How to appoint a prime minister

In one of those nice political coincidences, Australia is debating a hypothetical about the appointment of a prime minister just at the time when Britain is gearing up to deal with the same problem in reality.

Britain’s problem is that Boris Johnson is (barring some unimaginable event) about to be elected leader of the Conservative Party, with a resulting claim to be appointed as prime minister. But it’s not clear whether or not he can command a majority in the House of Commons, and it’s not clear exactly how that affects his prospects.

We can maybe get some guidance, though, from the question that might easily have arisen in Australia. If Peter Dutton had won the Liberal leadership ballot last year but Malcolm Turnbull had refused – as he apparently contemplated doing – to recommend his appointment as prime minister, what would have happened?

A couple of preliminary points first. Britain and Australia are not in exactly the same constitutional position: not because Britain’s constitution is unwritten (the key understandings are not written down in Australia either, or indeed in most parliamentary systems), but because although both share the same head of state, she acts personally in Britain but through a representative, the governor-general, in Australia.

That means that Australia has some of the characteristics of a parliamentary republic. The governor-general is supposed to be an impartial arbiter, but he or she can act personally in a way that the queen, with the weight of a dynasty on her shoulders, would not. A governor-general who makes the wrong decision can resign or be dismissed and the system would carry on.

On the other hand, because the governor-general lacks the queen’s security of tenure, they can sometimes be forced into a hasty action by the fear of dismissal – as John Kerr was in 1975.

The other point is that appointment of a prime minister only becomes an issue when there is a vacancy. If Turnbull had really wanted to test the system (and his own support) last year he would have done so not by resigning and then offering the governor-general unorthodox advice, but simply by not resigning at all, and daring his opponents to vote against him in parliament.

So, with that background, what would the governor-general’s options have been if Dutton had won the leadership ballot but Turnbull had advised him to send for someone else – say, then deputy Liberal leader Julie Bishop?

When in office, a prime minister is entitled to have their advice accepted unless the governor-general is prepared to countenance their resignation or dismissal. But once the prime minister has resigned, that convention no longer applies.

The governor-general may still, as a matter of form, ask the prime minister’s advice on their successor, but there is no requirement to do so. (Queen Victoria, for example, famously refused to consult Gladstone on who should succeed him in 1894.) And in the vast majority of cases, the choice is obvious and the advice adds nothing to what the governor-general would do anyway.

But sometimes, particularly when the numbers in parliament are finely balanced, there is some doubt. When Kevin Rudd succeeded Julia Gillard in 2013, the governor-general apparently had some concerns about his ability to command a parliamentary majority, and sought an assurance from him that he would “announce his appointment at the first possible opportunity to the House of Representatives in order to give the House the opportunity for whatever, if any, action it chooses to take.” He duly did so.

Writing the other day in the Conversation, Anne Twomey explains that the Commonwealth solicitor-general at the time apparently took the view that, while the governor-general could ask for such an assurance, she had no right to insist on it and was bound to appoint Rudd in any case on the basis of Gillard’s advice.

I agree – and indeed said at the time – that in the circumstances “it would have been grossly improper” to reject Gillard’s advice. Similarly, it would have been wrong for the governor-general last year to go against Turnbull’s advice to send for Scott Morrison – or for Dutton, in the event that he had won the ballot.

But knowing that an outgoing prime minister should be followed when they give the obvious, safe advice is no guide to what to do if they advise something different. My view is that if Turnbull had advised the governor-general to send for Bishop he should have done so and asked her some difficult questions about whether she could command the confidence of the House of Representatives.

He could then have sent for Dutton (and probably for opposition leader Bill Shorten as well) and asked the same questions, and then exercised his judgement. That is, fundamentally, his job. But he would ultimately be constrained by what parliament wanted, and a hostile vote in the House of Representatives would have forced a different choice – or, if no-one could command a majority, an election.

Could the governor-general have commissioned one of the contenders as prime minister on a provisional basis and told them to go and test it forthwith on the floor of the House? There seems no particular reason why not. It would be an unusual proceeding and certainly not appropriate in most cases, but it would be consistent with the workings of our constitution and would uphold the principle of parliamentary sovereignty.

Which brings us to Britain. At first sight (and despite the difference in the method of choosing the leader, which is really a side issue), this is a case like 2013: the party chooses a new leader, the outgoing prime minister recommends that person as her successor, so provided parliament has an early opportunity to pass judgement on him, there is no reason he should not be appointed.

And I think that is most probably what will happen. But unlike 2013 in Australia, there is a real possibility that there will be enough internal dissent within the Conservative Party to give the opposition a majority on a no-confidence vote. And because the process is so drawn out, that may become clear before Johnson can be sworn in.

There is also much more at stake. Rudd in 2013 was about to go to an election anyway; if it had been forced a couple of months early that would not have made much difference to anything. But the British parliament otherwise has nearly three years to run: an election now, with the party system in chaos, would be a very big deal.

Hence the queen, who has been waiting for this moment for 67 years, may have some serious decisions to make. Constitutional professors Robert Hazell and Meg Russell put it like this, as reported by Michael Savage in the Guardian:

“One possible scenario is that a group of Conservative MPs is so concerned about the winning candidate that they declare their withdrawal of support immediately the result of the leadership contest is known – ie, before the new PM is appointed. This would pose a serious dilemma for the Queen and those advising her, because it would not be clear that the new Conservative leader could command confidence.” …

In their analysis, Hazell and Russell conclude that the Queen could make the new Tory leader a “provisional appointment” as prime minister, conditional on him demonstrating he has the confidence of enough MPs. “Alternatively, Theresa May could remain in place and facilitate a process in parliament to demonstrate that the winning candidate – or indeed an alternative candidate – can win a confidence vote, before recommending that person to the Queen.”

For Johnson, the choice may well come down to defeat in a couple of months time versus defeat before he has even started. But with the clock running down to the Brexit deadline of 31 October, even that could make a big difference.