It’s just over a week now since Australia’s largest ever by-election, the re-run of the Western Australian half-Senate election. The Australian Electoral Commission is still some way off declaring the result – it can’t start distributing preferences until the deadline for receipt of postal votes has passed, which is still eight days off – but there’s no longer any doubt about the outcome. Western Australians have returned three Liberals, one ALP, one Green and one Palmer United.
That’s slightly different from both the original count last year and the recount, but in terms of its effect on Senate control it’s the same as both: three government, two Labor/Greens and one right-wing independent.
So it’s worth having a quick look at how that compares with what was expected beforehand, and specifically with my review of the preference dealing a month before the election. Here’s what I said:
Last year, when they ran the election the first time, the Coalition had fractionally over three quotas and the right-wing minors just over one quota. Labor had a bit less than two quotas, the Greens about two-thirds of a quota and the left-wing minors the remaining one-third. …
It’s generally assumed that for the re-election there’ll be a swing away from the government, so the overall Left should be favored to pick up a seat at the expense of the overall Right.
Unless the pattern of voting changes radically from last year, that would mean the second Labor candidate and the Greens’ Scott Ludlam would both get up, while on the Right there would be a close contest between the third Liberal and whoever came out at the head of the pack of right-wing minor parties.
There was indeed a swing away from the government, but it wasn’t quite enough to win the left its third seat. Although on the night it looked very close, with the counting of postal votes the third Liberal candidate has drawn away from the second Labor candidate and will probably beat her by something like 15,000 votes. (If you’re interested in the detail, Kevin Bonham’s calculations and commentary are invaluable.)
Generally in by-elections you expect two things to happen: a two-party swing to the opposition, and a shift away from the major parties to minor parties or independents. The second of those most certainly happened in Western Australia. The Coalition vote was down by 7% and Labor’s vote down almost 5%; the Greens, on the other hand, gained nearly 6% and the Palmer vote more than doubled, to 12.5%.
But the overall left-right balance didn’t shift very much. The combined Labor/Greens vote was up by about 1%, and left-wing microparties added about another 0.4%. On the right, Palmer’s gain more than balanced the Coalition’s loss, but the other right-wing minor parties between them lost about 2% – most of that attributable to the Liberal Democrats, whose vote almost halved.
The Coalition and the right-wing minors between them finished with almost exactly four quotas. Labor had one and a half, the Greens just over one, and the left-wing minors a bit over a third. Although it had nothing to spare, the government was able in essence to hold on to last year’s result, which was always the best it could hope for. It will still have to deal with Palmer in the Senate, but it could easily have been worse.
So preference aggregation for minor parties had no role to play. There was no prospect of a microparty being able to overtake Palmer, and no room on the left for anyone to get close to Labor or the Greens.
It would have been different if the overall swing to the left had been bigger, so that the third Liberal and the Palmer candidate were left fighting for the last seat instead of taking one each. In that case the preferences of microparties would have been critical. And if the Palmer vote had not ballooned so much, careful preference harvesting by one of the microparties might have gotten one of them into the lead.
The moral is that the problem caused by our bizarre automatic preferencing system of group ticket voting has not gone away, but it’s not quite as omnipresent as last year’s election might have led people to believe. To win as a microparty you don’t just need a good run of preferences, you also need a lot of luck in the pattern of the major parties’ vote.
Often, as in this case, the majors will divide up the vote in such a way as not to leave room for anyone else. But sometimes that won’t happen, and for those occasions the need for reform is still there.