As the election campaign enters the home stretch, a consensus is hardening that Malcolm Turnbull will still be in The Lodge on Monday.
However, what’s even clearer is that he will have little or nothing to show for his strategy of resetting the Senate through a double dissolution held under a reformed electoral system.
While the new system does introduce an element of uncertainty, the available knowledge about voting intention can be combined with a few known knowns concerning voter behaviour at state and territory level to produce a set of reasonably educated guesses.
Ballot papers that direct voters to fill a minimum number of boxes are familiar from Tasmania (five boxes in the lower house, three in the upper), New South Wales (at least 15 boxes for below-the-line votes in the upper house) and the Australian Capital Territory (at least five boxes).
The ACT is particularly instructive, because a vote there is still formal even when the voter ignores the instructions and numbers fewer boxes than indicated, as will be the case in the Senate.
The picture that emerges from these precedents is clear: voters overwhelmingly do as much as the ballot paper asks of them, even if they technically don’t have to, and no more.
It can thus be stated with a high level of confidence that all but a small minority of Senate votes will consist of six boxes numbered above the line.
Polling on Senate voting intention is rarely conducted for the very sound reason that it has a dismal track record, but the first preference vote can be approximated by applying state-level swings from aggregated lower house polling to the results from 2013.
Picking the make-up of the “others” vote requires a mix of art and science, particularly in relation to the scattering of Palmer United Party support in Queensland, and the extent to which the Liberal Democrats vote will revert to type in New South Wales after David Leyonhjelm lucked out with first position on the nation’s largest ever federal ballot paper in 2013.
The distribution of preferences is even more of a wild card, as voters will overwhelmingly be making ad hoc decisions in the polling booth as they determine choices two through six.
However, a useful guide is available in the form of below-the-line voting data from the last election.
It’s true that below-the-line voters are a peculiar breed, but the results that emerge when this data is aggregated tell an intuitively compelling story, with supporters of religious parties overwhelmingly favouring other religious parties, leftist voters favouring other leftist parties, and social libertarians favouring other social libertarians (though interestingly, the Liberal Democrats did far better with conservatives than those who voted for the likes of the Australian Sex Party and Help End Marijuana Prohibition).
Then there are those who follow how-to-vote cards — a little less than half in the case of the major parties, around a quarter for the Greens, and a negligible figure for the micro-parties, which lack the mass base of volunteers needed to get cards into the hands of voters on polling day.
All concerned are directing their supporters to number exactly six boxes, and Labor in particular has adopted a strategy of directing preferences to almost any minor player who might conceivably keep a seat out of Coalition hands (although they have drawn the line at Pauline Hanson).
Combining the parties’ preference recommendations with the below-the-line data produces a plausible basis from which to project preference flows, and thereby to model the overall result.
So far as the big picture is concerned, this suggests the new Senate will be barely distinguishable from the old.
The crossbench looks set to be very slightly more manageable from a Coalition perspective, to the extent that the Greens should be down at least one seat and the Nick Xenophon Team will likely constitute a bloc of three seats, but there will ultimately be very little in it.
Of the existing micro-party senators, the only ones clearly headed for the exit are the last man standing for the PUP, Western Australian Senator Dio Wang, and the two Victorians, Ricky Muir and John Madigan.
It seems very likely they will be making way for Derryn Hinch, who has a national media profile, a populist message on law and order that sits will with an element of disaffected voters, and a spectacularly fortuitous position at the top of another enormous ballot paper.
Queensland is particularly fertile terrain for the anti-major party vote, and this should leave room for two micro-party senators, with the Coalition, Labor and the Greens having very little to spare after respectively electing their fifth, fourth and first candidates.
Insiders reckon it all but inevitable that Pauline Hanson will finally return to the political stage, and she looks set to do so without depriving Glenn Lazarus of his seat.
Three seats looks the most likely result for the Nick Xenophon Team in South Australia, and it is also thought likely that Family First Senator Bob Day will be able to mobilise at least 5% of the statewide vote, sufficient to get him over the line as well.
There have been suggestions over recent days that the Jacqui Lambie Network could even be in contention for two seats in Tasmania, but the published lower house polling on which the projection is based suggests a more modest statewide vote of around 6%, which is at least enough to get Lambie herself comfortably over the line.
In New South Wales, a game of musical chairs looms for the last two seats between David Leyonhjelm, a fifth Labor candidate and a second Green, with the latter to be left stranded when the music stops.
Only Western Australia seems to lack the conditions for a micro-party victory, with none of the contenders having the public profile needed to reach a critical mass.
That’s unless you count the Nationals, who are a particularly rebellious outfit in Western Australia, and have a once-in-a-generation opportunity to win a Senate seat thanks to the lower double dissolution quota.
The new electoral system should eventually produce its desired result of curbing the crossbench once two consecutive elections have been held under normal circumstances, with six senators elected per state from quotas of 14.3%.
However, that’s little comfort for Malcolm Turnbull, who stands to have whatever passes for his legislative agenda substantially thwarted over the coming term.
Worse still, the result will allow his conservative enemies to advance the case that he is no more sound tactically than he is ideologically.
*To read more from Crikey‘s William Bowe, visit The Poll Bludger