Will they ever learn?

Twice in modern times (prior to last Saturday) the Victorian Liberal Party has lost in a landslide. The first was in 2002, when it fell to 33.9% of the vote (42.2% two-party-preferred) and won only 17 of the 88 seats (plus another seven held by the Nationals).

But it recovered from there. Not as much as it had hoped for: the next election, in 2006, was still seen as a disappointment. Nonetheless its vote rose to 34.4%, or 45.6% two-party-preferred – the primary vote was depressed by the emergence of Family First, which had 4.3% – and it picked up another six seats (plus another two for the Nationals). That put it within striking distance of victory, which it duly if narrowly achieved in 2010.

The second landslide was in 2018. There was debate at the time over just how similar to 2002 it was, but there’s no doubt it was at least broadly similar. The Liberal primary vote fell to 30.4% (42.7% two-party-preferred) and it won 21 seats (with another six for the Nationals). The question for Saturday was whether it could reproduce the limited success of 2006 and put itself in a plausible winning position for 2026.

And that is what it has failed to do. Its primary vote has actually fallen further, to 29.7% (it may come up slightly on postals), although again that is a little misleading due to the rise of far-right parties. More significantly, it has failed to improve on its seat total: with a few still doubtful, it looks like finishing on twenty, one fewer than last time (its best case is 22). The Nationals have picked up three.

The two-party-preferred vote looks a little more respectable – the ABC’s current estimate is 45.9%, a swing of 3.5%, almost identical to 2006. But from the Liberal Party’s point of view, the swing is in all the wrong places.

There are currently 73 seats showing a two-party-preferred swing between Labor and Coalition (another 14 have Greens or independents in the top two, and voting in Narracan has been deferred due to the death of a candidate).* The median swing in those 73 is 3.0% to the Coalition. But of the twenty most marginal Labor seats among them – the seats that the Coalition needs (and needed) to win – the median swing is only 0.3%. Nine of the twenty actually swung to Labor, including six of the nine most marginal.

The Liberals got some big swings, but they were nearly all in safe Labor seats: 15.6% in Greenvale, 14.1% in Mill Park, 12.3% in St Albans. That suggests the ALP has problems in its heartland, but it’s no help to the Liberals for the immediate future. In only three marginal seats did the Liberals get a swing of more than 6%: 7.0% to win Nepean, and 6.7% and 7.5% to hold Eildon and Brighton respectively.

The difference from 2006 is not huge; Labor then also did relatively well in its marginal seats. But it could easily be the difference between winning and losing next time. Why did the Liberals make up more ground then? Primarily, it seems, because they made some effort to show that they had learnt from defeat: they (eventually) chose a new leader and tried to present a modernised image.

This time around, the party changed leaders immediately, but then last year decided to go back to the old one, Matthew Guy. And instead of learning the lessons of previous Covid elections, they ran a negative campaign focused on premier Daniel Andrews and his responsibility for lockdowns – which many voters interpreted as an attack on public health. Add in the cultural obsessions of the News Corp media [link added] and the fundamentalist takeover of many Liberal Party branches and you had a recipe for disaster.

There were other features of interest on Saturday as well, including the poor result for independents and an equivocal result for the Greens. But they may look a bit different with later counting, so I’ll leave them for another post later in the week – by which time there may also be some meaningful results for the Legislative Council.

.

* These figures are all from the ABC’s website; a few are estimates because the appropriate two-party counts have not yet been done, and all are subject to slight change in later counting, but the general picture is clear. Kevin Bonham’s site is the best place to follow the detail of doubtful seats.