Counting in the Victorian election – or at least the lower house, the Legislative Assembly – finally wrapped up yesterday after exactly two weeks. My friend Louise Staley, Liberal MP for Ripon, held her seat after a recount by the razor-thin margin of 15 votes.
That leaves Labor with 55 seats, a gain of eight from 2014 and a majority of 22 against all comers. The Liberals will have 21 (down nine), with six Nationals (down two), three Greens (up one) and three independents (up two). You can find all the official figures here.
Those numbers, however, don’t bear much relationship to the votes cast. We’re used to seeing tables like this, but we mustn’t get desensitised to how strange they are:
|Party||% of vote||Seats||% of seats|
Labor won five-eighths of the seats with only three-sevenths of the vote, and the Nationals, with less than half as many votes as the Greens, won twice as many seats.
If Victoria had, for example, the democratic electoral system used in nearby New Zealand, Labor would have won 40 seats as against 28 Liberals, ten Greens, six independents and four Nationals. It would have had to negotiate with the Greens and/or independents to secure a majority.
This lack of proportionality would be less of a disgrace if Victoria had an upper house where party strengths reflected their actual support. But while counting for the Legislative Council won’t be finished before Tuesday, we already know that it will do a poor job on that front as well.
What about preferences? Although the Greens don’t win many seats (and those they did win were all very close), their votes do help Labor to win seats. Preferences had to be distributed in 47 seats, and in twelve of them the eventual winner had trailed on primaries: two independents, one Green, and nine Labor candidates.
With all preferences distributed, Labor had, by my count, 57.7% of the two-party-preferred vote, a swing of 5.7%.* That’s a bit better than expected (a week and a half ago I said it would be “in the neighborhood of 5%”), reflecting the fact that Labor did relatively well out of postal and absentee votes; it leaves it as close as doesn’t matter to its high point of 57.8% from 2002.
Of the 88 seats, all but six swung to Labor. The biggest swing was 13.6%, in Footscray. The four seats along the Frankston line, famous in recent years for making and breaking governments, all swung by more than 9%.
It was a very poor result for the betting market. Going by Sportsbet’s odds from the eve of the election, 14 candidates went in as favorites but were nonetheless defeated: one Labor, two Greens, two Nationals and nine Liberals.
Although Labor did well in the late counting, the Liberals still won the greater share of the very close seats. Of those with margins below 2%, eight are on the Liberal side of the pendulum (Ripon, Caulfield, Sandringham, Gembrook, Hastings, Brighton, Forest Hill and Ferntree Gully) and only four Labor (Bayswater, Hawthorn, Nepean and Mount Waverley).
(The three Greens seats and two of the three independent seats also have very small margins, but only Morwell is at all close if counted out as Coalition vs ALP.)
At the other end of the scale it is very different. Labor is, if not competitive, at least respectable throughout the state. In no seat did it have less than a quarter of the vote (two-party-preferred), and in only three (Murray Plains, Lowan and Gippsland East) did it have less than a third.
But the Coalition is below a third of the vote in 22 seats – one in every four – and in nine of those it’s below 25%.
The details of safe and marginal seats will change before the next election, with a redistribution of boundaries to be held in 2020-21. But there’s every chance that that will make things worse for the Coalition rather than better.
That’s because the fastest-growing areas of the state, where new seats will have to be created, mostly vote Labor (or Greens, in the case of the inner city). Of the ten seats that had the most voters last month, Labor won nine and the Liberals one. However, Labor also won seven of the ten smallest, so there is a potential downside for it as well.
* There are a few uncertainties in this figure. The Liberal candidate for Yan Yean was disendorsed before the election, but was still identified as a Liberal on the ballot paper; I have included her in the Coalition total. The electoral commission has not produced a two-party count for two Labor vs Greens seats, Melbourne and Preston, so I have used estimates of preference flows based on the similar neighboring seats of Brunswick and Northcote. And there was no Liberal candidate in Richmond; I have used the vote for Kevin Tran, a Liberal Party member standing as an independent, and estimated a preference flow to him on the same basis.
ADDENDUM, 17 December: The electoral commission has now done the two-party-count in Melbourne and Preston, and it’s not quite as strong for Labor as the estimates I was using; taking comparable numbers for Richmond as well, that brings the statewide Labor vote down a tenth of a point to 57.6% two-party-preferred. Kevin Bonham now has an excellent summary post covering that and much else besides.