The missing ingredient in the knifing of Malcolm Turnbull: Votes

Malcolm Turnbull’s leadership is in a perilous state, adrift in a political white-out fomented by vengeful opponents and exacerbated by the PM's own panicked dissembling. Up resembles down.

Whether he faces a challenge in the next 48 hours from the conservative Peter Dutton, or some proxy, is unknowable.

A decade of roiling volatility in Canberra tells us that the absence of a visible horizon can cause even the more grounded of MPs to lose their bearings. Anything could happen.

Prime Minister Malcolm Turnbull on Monday.Credit: Dominic Lorrimer

Remember, no Australian prime minister has served out a full term since John Howard – he strung four together.

Yet in each of the bruising changes we have seen since Howard, there has been at least the semblance of a motivating logic – the allure of a poll bounce arising from change.

Julia Gillard’s 2010 move on the faltering Kevin Rudd capitalised on fears within the Labor caucus that Rudd’s star was waning fast and that Gillard’s then superior credibility would secure a second term.

While Rudd’s poll numbers look embarrassingly healthy in hindsight – Labor was still ahead - he was adjudged by his colleagues to be heading in the wrong direction and unwilling to change.

Gillard’s promotion did indeed lift that support, albeit only marginally and then only temporarily, thanks to some well targeted "friendly fire".

The case for switching back three years later was statistically more compelling with Rudd’s standing of 58 per cent, easily shading Gillard on just 32. More importantly, Labor’s primary vote had tanked to just 29 per cent against the Coalition on 47. Disaster beckoned unless Rudd was returned to the helm, which the Fairfax Nielson poll showed, would see the contest back to even at 40-42.

The pollster at the time, John Stirton, expressed caution in interpreting those raw numbers, describing what he termed ''a magical scenario'' in which Rudd acquired the leadership seamlessly, led a united party, and got a ''honeymoon'' that lasted “all the way to polling day”.

Magical was right. Labor lost the election a couple of months later but Rudd’s reprise had certainly limited the party’s losses.

Then came Turnbull’s September 2015 move on Tony Abbott. Unthinkable as it was, this Liberal re-run of Labor’s extraordinary leadership follies was nonetheless propelled by some pretty enticing maths.

Famously, Abbott’s government had suffered 30 consecutive losing Newspolls. A miserable first term failure seemed unavoidable. Whereas, the more urbane Turnbull, who had long out-polled Abbott as preferred Liberal leader, could offer his colleagues the promise of new life. Again, the subsequent reality was more sobering with Turnbull's 30 Newspoll losses flashing by months ago. But he remains his party's best asset.

This is where the current proposition - the case for dumping Turnbull - stalls. Here we see a not unpopular prime minister – leading Bill Shorten by 12 percentage points as preferred prime minister 48-36 – being apparently torn down by colleagues who favour a little-known colleague whose approval as alternative Liberal leader consistently languishes in the low single figures.

It is true a "Prime Minister" Dutton might well perform better than his current 2 or 3 per cent support suggests.

But nobody's even running that argument, which reveals this is more about dispatching Turnbull than the interests of the government.

It is being pushed even though electorally, Dutton represents a bigger risk than staying with Turnbull. Given the inevitable transaction costs of tearing down another elected prime minister, it's a gaping hole in the case for a change.