A Liberal in Melbourne?

- First published in the Australian 26 July 2012

It has been reported that the Victorian branch of the Liberal Party is very likely to preference Labor ahead of Adam Bandt in the federal seat of Melbourne at the next election.

The expectation in some quarters is that this will send the sole Greens House of Representatives MP packing. One “senior Liberal” has told the Australian she or he would be “absolutely blown away if the party didn’t end Bandt’s career.” It’s true that Liberal how-to-vote cards helped deliver Bandt victory in 2010. If the Liberal candidate had “preferenced” Labor’s Cath Bowtell ahead of Bandt, instead of vice versa, Bowtell would now be the Member for Melbourne.

But that was then. Putting the Labor candidate ahead of Bandt on the Liberal how-to-vote card next time is not likely to see him off.

Now he’s there, Bandt won’t be that easy to move.

The logic of the “Bandt is doomed” narrative seems to be that an expected increase in support for the Liberals in Victoria at the next federal election will give them many more preferences to direct.

The flaw in this logic is that nearly all of this increase will come from the ALP, and the Libs can only return a portion of them in preferences.

Bandt is quoted as saying he is “aiming to win the federal seat of Melbourne on primary votes”. If he means he aims to top the primary vote, it is close to certain he will do that. If he means he aims to get more than 50 per cent of the primary vote, it is close to certain he won’t.

But he won’t need to. Bandt is looking very safe in Melbourne.

At the 2010 federal election., Bowtell received 38.1 per cent of the primary vote, Bandt got 36.2 and the Liberal candidate 21.0.

Thanks to the Liberal HTV card, 80.0 per cent of the Libs’ 21.0 flowed to Bandt and 20.0 to Bowtell (from AEC data) and so Bandt won the two-candidate-preferred vote 56.0 to 44.0.

In a recent post I compared the results in the state seat of Melbourne in 2006 when Libs preferenced Greens, and 2010 when they preferenced Labor.

When the Liberal candidate dropped out of the count in 2006, 74.4 per cent of those votes flowed to the Greens and 25.6 to Labor. When the Liberal candidate dropped out in 2010, 33.6 per cent went to Greens and 66.4 to Labor. The big difference, of course, is mostly due to HTV cards.

(These state numbers are not quite the equivalent of the federal 80.0 20.0 quoted above. The federal number represents what happened to all votes with a ‘1’ next to the Lib, while the state numbers represent what happened to votes that were with the Lib when they dropped out of the count. But they are almost equivalent.)

How might federal preferences have flowed in 2010 if Libs had put Bowtell ahead of Bandt on their cards? We might subtract 33.6 from 74.4, and take that number off 80.0. Then we would say 39.2 per cent of their preferences would have gone to Bandt and 60.8 to Bowtell. But let’s err on the side of Liberal preferencing power for the sake of argument and apply the 33.6 to 66.4.

If we distribute Lib preferences 33.6 to Bandt and 66.4 to Bowtell at the 2010 federal election, Bowtell wins 53.7 to 46.3 two-candidate-preferred.

One way to contemplate whether Bandt will survive next time is to ask whether we think he would have built a personal vote of 3.7 per cent or more, to overcome this deficit.

It’s very likely he would have.

But it gets worse for Labor. As we’ve noted, there is the expectation that Labor’s vote in Victoria will be down next time and the Libs will be up. Of every vote that goes from Labor to the Libs, most of it will go back to Labor but a large minority will go to Bandt.

We might say Bandt will get 39.2 per cent of them, or 33.6.

So the worse the ALP does against the Libs, the worse they do against Bandt.

In 2010 Bowtell got 38.1 while the state-wide Labor vote was 42.8 per cent against the Coalition’s 39.6 and Greens’ 12.7.

The latest Newspoll quarterly had Labor on 34 and Coalition on 42 and Greens on 15 in the state. Whichever way you look at it, Bandt increases his support on those numbers, both from the higher Greens vote (which may or may not eventuate at the ballot box) and the higher Coalition vote, about a third of which flows to him.

And that’s without taking account of Bandt’s personal vote, his “sophomore surge”.

The ALP faces an uphill battle taking Melbourne back while Bandt is there, unless its vote in the state improves to a higher level than in 2010, which is very unlikely.

In fact, if Labor does very badly, and Bandt’s vote holds up, then the Liberal candidate may get into second place. If Labor preferences the Liberals, as some in the party believe they should, then the Liberal might win the seat.

Perhaps that’s what the Victorian Liberals have in mind.