By-election wash


Election results on Saturday have largely been interpreted as some sort of endorsement of new Liberal leader Tony Abbott, yet looking at the figures the result is highly unremarkable for the Liberal Party.

Bradfield and Higgins are both incredibly safe seats for the Liberal Party. Bradfield is the second-safest Liberal seat in the country, and neither seat has ever gone to preferences. The Greens achieved a strong result in a right-wing heartland seat while running an explicitly left-wing campaign.

Ever since the Cunningham by-election, conventional wisdom always assumes that the Greens have a chance of winning any by-election in a safe seat, and build it up so that it becomes a disappointment when the Greens don’t achieve some dramatic victory. Of course, some of these races are genuine opportunities for the Greens, such as those in Marrickville, Mayo and Fremantle.

Bradfield and Higgins were never going to be seats that the Greens could win. The absence of an ALP candidate gave the Greens a shot at increasing their profile and winning more votes, but these extra voters would never come close to the numbers needed to win the seat, and the Liberals still had a solid majority of support in both seats.

Clive Hamilton polled the highest achieved by any Greens candidate in a federal lower house by-election, while Susie Gemmell polled higher than in the Cunningham or Mayo by-elections on primary votes. That was the achievement for the Greens. Ultimately the goal of the Greens in areas like Bradfield and Higgins is to bite into the ALP vote and strengthen their Senate vote. The absence of the ALP made the Greens the only progressive alternative and gave the party an opportunity as ALP voters seriously considered and voted for the Greens for the first time. If half of those voters stick with the Greens in the Senate in 2010, it will be a big step towards getting Lee Rhiannon and Richard di Natale elected.

Even if you agree with the assumption that the by-elections were a ‘referendum on climate change’, it is madness to suggest that the result in two blue-ribbon Liberal seats has national implications in terms of the climate change debate. It is also not necessarily an endorsement of climate change policy for voters who have loyally voted Liberal for decades to continue voting Liberal.

No-one has ever suggested Tony Abbott would have trouble holding on to his party’s base in their safe seats. His problem lies in marginal seats where the ALP will attack him on climate change and other issues. Those are the seats that keep the ALP in government and Abbott needs to win them to bring his party forward. Last weekend’s by-elections do not offer a shred of evidence as to how Abbott will play in the key seats, nor how those seats will react to campaigns around climate change.

It’s also worth remembering that the Greens ran on a principled, explicitly left-wing platform in both seats, focusing on strong action on climate change. Yet they managed to achieve the same vote the ALP had achieved in 2007. It is not reasonable to assume that all ALP voters would automatically switch to a party running substantially to the left in a conservative seat.

The Greens were never going to win either seat at this election. Both by-elections were great  opportunities for the party to continue it’s goal of supplanting the ALP amongst progressive voters, and the Greens achieved this. In particular, the Greens have an opportunity to almost entirely push the ALP out of the north shore of Sydney over the next decade. The Greens have already come second at a state level in the seat of North Shore and beaten the ALP in Pittwater, and came close to beating them in a number of other north shore seats. By being the only progressive option in northern Sydney, it may not be enough for the Greens to win lower house seats, but it substantially increases the party’s presence and chances of electing Senators and MLCs. Bradfield helps move the party closer to turning the north shore into a series of Liberal-Green contests.

The Greens aren’t a centrist party, and aren’t going to be winning seats in conservative areas for a long time. Most of the party’s chances at a state and federal level lie in inner city areas in Sydney, Melbourne and Perth.

Liked it? Take a second to support the Tally Room on Patreon!


  1. People can say that, in general, the Greens failed to take votes off the Liberals, but taking votes off the Liberals, regardless of the circumstances, isn’t something that is very easy for the Greens, especially the rusted-on Liberal voters in these heartland seats. It’s not simply a matter of the left-right linear scale metaphor (which we need to stop using), but also a matter of identity. Because we have a class-based party system, both major parties, but mostly Labor of course, have progressive-leaning elements of their supporter base – the Greens find it easier to attract the progressive Labor voters, but the drop in the Liberal vote in many of their strongest booths in Bradfield in particular might indicate that we are doing better at picking up those harder to get progressive Liberal voters as well.

    Whilst I agree with you Ben, there’s a couple of other thoughts which crossed my mind. The comparison with Mayo, a genuine chance? The 2007 primary vote shares for Liberals, Labor and Greens in Mayo and Higgins were about the same. In Mayo, solid Liberal territory, but also the small-l liberal heartland of the breakaway LM and subsequently Democrats, the previous threat to the Liberals had come from the Democrats, who were able to attract progressive-leaning Liberals, but there weren’t enough of them to win. For the Greens to have won Mayo, they had to get the progressive Liberals, but likewise also needed more conservative Liberals, which is also what would’ve needed to have happened in Higgins or Bradfield. What made Mayo more interesting was that the field included a few other strong candidates who helped break up the Liberal vote, thus giving the Greens a chance to gather more preferences and get closer, but otherwise, was it really that much of a ‘genuine’ chance to rank closer to Marrickville and Fremantle than to Higgins? (Mayo was a very good result, and bodes well for the future of the Greens in SA)

    And whilst Higgins and Bradfield might look like solidly right-wing conservative areas (because they vote Liberal), they both voted ‘Yes’ in the 1999 republic referendum by large margins – 64% in Higgins, 56% in Bradfield (compared for instance with only 49% Yes in Mayo).

    Abbott can take comfort from the fact that the result wasn’t a disaster, which should at least mean that if the party can stabilise, the next election won’t be a total disaster either, but to claim it as an endorsement of his leadership, vindication of their climate policy, or anything else, just doesn’t add up.

  2. ” The absence of the ALP made the Greens the only progressive alternative…”

    You forget the Sex Party, non?

    Having said that, their preference went 2:1 (The Greens: Libs) where I was… what a deliciously seedy underbelly the Libs have in their base.

  3. After the disasterous last few weeks for the Libs, and with climate change absolutely centre stage as the MAIN public policy issue, they actually improved their vote by something like 3% 2PP in Higgins, despite the retirement of the popular Costello. There would have to be no more than a half dozen seats nationally that are more favourable to the Green message on climate change than Higgins (Melbourne & Melb Ports in Vic can anyone name another? …if this is the result in Higgins, imagine what could happen (to the ALP) in LaTrobe or Holt or Issacs. Hardheads in the ALP will be taking a close look.

  4. Peter K – In respect of seats that might be responsive to a climate change message try Fremantle, Curtin, Sydney, Grayndler, Wentworth – and thats the ones I know. Alot depends on the mix of voters and their demographics. I for one would not have picked a Liberal seat as the one to switch to a Green seat – where it has happened so far (Cunningham & Fremantle) its been an ALP seat. Certainly when asked by Lib vollies in Bradfield what the likely outcome would be I suggested that they would get over 50% of primaries – because although climate change was an issue, with Abbott being the newly installed leader the nexus between the ALP, Libs and climate change had been broken.

    Essentially, the older (tribal?) divisions between the ALP & Libs was able to reassert itself. Voters would switch back to base issues to vote on. Being the kind of seats these were I would have expected some of the ALP vote to come to the Greens but never all (there are plenty of conservative materialistic ALP voters out there) so the vote The Greens got was quite respectable in those circumstances. And while climate change might have been an issue in these by-elections, the way the ALP handled it (ie; completely failing to sell the ETS and spending alot of time on making deals with the Libs) it was more an issue of confusion than conviction.

  5. I just checked up on past results. This by-election gave the Liberals their worst primary vote ever in any contest in Bradfield. It was also within 1.3% of their worst ever result in Higgins. Both without a Labor candidate present.

  6. Peter K, what you’re asking (about what could happen to the ALP in La Trobe or Holt or Issacs) is just pure speculation. The ALP didn’t stand a candidate in Higgins or Bradfield, and the fact that some of their voters preferred the Liberals over the other alternatives is not surprising. It’s also a fact that both the Liberals and the Greens have voted the same way regarding the Government’s ETS (even though it’s for very different reasons), so it’s wrong to some how equate the Greens climate change campaign as being a stand-in for support for the ALP’s ETS, no matter how much some people want to say ALP=Greens. Some in the ALP have argued that they may have come close or even won Higgins if they had run there – and if you look at the booth by booth breakdown, they have a point. It is doubtful that the Libs would have gotten the significant swings to them that they did on the stronger ALP booths if there was an actual ALP candidate for those voters to support. However, the big swings the Libs received on many of their strong booths would probably have still occurred, which would have made it a very close contest.

    As for how the ETS will play in the outer suburbs – I think it will be much harder for the Libs to run an effective ETS scare campaign then it was for them to run the scare campaign about interest rates in 2004 (which saw seats like Holt swing hard to the Libs & briefly become marginal). People can easily workout the impact of increasing interest rates on their mortgages, and the Libs took advantage of the ALP having Mark Latham as leader. Kevin Rudd is both popular and respected. The Howard Government lost a lot of respect and trust they had built up over the increase in interest rates after 2004, and even more for introducing Work Choices. The ALP will certainly remind the voters of these things. The ALP can also run a “think of the future of your kids” campaign, which could be extremely successful in those seats. It will be interesting to see how it all plays out in the election next year, but I think on the current evidence, the ALP is so strong at the moment that they’ll easily be re-elected.

  7. Stewart,
    To that list you could add North Sydney, Adelaide, Denison and perhaps Kingsford-Smith, Brisbane, Perth & Batman but it only proves the point that Higgins is demographically in the top 10-12 seats nationally for sensitivity to Green issues, and top 3 for Vic – yet they couldnt manage a protest vote (ie anti-Lib swing) even there on the CC issue. You have to read this result in that light.
    Btw I’m talking about demographic inclination to green issues, not chances of the AG actually winning the seat, which is different. I don’t think Cunningham (or Mayo) would rate particularly highly.
    As for Bradfield, the Lib vote has declined significantly in recent years – the TPP used to be above 70% even in opposition – due to changes in boundaries and ethnic composition. Having the worst primary vote ever dosent mean much then, especially given the circumstances of the by-election. This time their TPP held up.

  8. Let’s not get too lost in the clouds here. The Libs won two by-elections with swings to them. That is a solid result for them. The good voters of Higgins and Bradfield were given the option to stick the boot into Abbott and ultimately did not. The Greens did OK, eating up most of the Labor vote but not recording a swing ala Mayo. That it was their best primary vote is a bit of a redundant point given there was no ALP candidate, and is thus hard to judge actual Green support in the electorates. Equally, that the Libs recorded their lowest vote is again redundant because voters use by-elections to punish major parties and also that there were 22 candidates.

    I think the main lesson to learn here is that local issues, rather than leadership (Abbott) or international issues (CC) is probably the best way for the Greens to win by-elections as it alienates far fewer ALP voters (or Lib voters in the case of Labor/Green contests).

  9. Um, not Kingsford Smith – the likely bits of the eastern suburbs to be Green inclined are all in the north of the seat. The rest is solid old-school ALP. As to Perth, people have often suggested it, but I would consider the core of the Green vote to be too small – and alot of the Green vote you might otherwise expect to find in Perth tends to head down towards Freo, or is spread more thinly around the city. Denison – yes, I will be interested to see what happens this time round with Denison with no Duncan Kerr, but again the problem is that the concentration is relatively low. As well, most of the seats you’ve named are ALP not Liberal seats (North Sydney notwithstanding).

    And I’d add that the literature on post-materialist idea transmission would suggest that having a Green party increases the chance of those ideas being transmitted, but this doesn’t show up as a 1:1 vote rise – from the seats where Greens have done well there is a need for a pre-existing increase. Even if you compare Higgins to Mayo, the Greens only doubled their primary in Mayo, but had the presence of high profile Independent (Di Bell) soaking up the ALP primary, as well as FFP (Bob Day) bringing down the Lib vote. This doesn’t occur in Higgins, where the Greens tripled their previous primary. On that basis it was an resounding win – although I notice Greens not claiming a “win” out of this.

    Lastly, it became reasonably obvious towards the end of the campaign that both O’Dwyer & Fletcher are more likely to be moderates than hard right Lib MPs. This actually provides the Lib voters with a chance to send a message to their own party about keeping moderates in to counter the right (ie; maybe Turnbull would have won the leadership if both had been in Parlt the week before?).

  10. Good point re: Moderates in the Libs.

    To copy and paste from Antony Green’s blog:

    “What-if anything, does the Greens vote suggest regarding their potential senate vote at an upcoming election? Do the tealeaves give us any likely trends? ie should they be disappointed with the vote given the pre-poll predictions from various pundits or should they think its a reasonable set of numbers?

    COMMENT: I don’t think there is anything to read into these results.”

  11. The Libs won two by-elections with swings to them.

    Bradfield (so far has) swung away from the Liberal candidate on primaries by 3.33%. The “swing” on 2PP is close to meaningless in this case, but that was away from them in Bradfield as well (3.37%).

    After postals it may turn into a very small swing to Liberal, which will negate my “worst ever result” remark. I guess we’ll all be bored of the topic by then though.

  12. Primary vote doesn’t really matter much though does it? I mean, the Greens didn’t win the primary vote in Cunningham way back when. You also can’t discount that 22 candidates would lower the major party’s primary vote.

    I don’t think it can be fairly said that this is anything but a solid result for the Libs.

    And, from what I understand, the 2PP is marginally to the Libs in Bradfield (0.8%, 3.1% in Higgins).

  13. All I can say is that it was excruciatingly annoying to see ALP and Liberal spinners coming out saying that Greens were in with a good chance. It was clearly a set-up to make the Greens look like failures when the results were through.

    This (non-)strategy that the ALP takes in by-elections may be “clever” in the short-term but it is eroding their base in the long-term. The Greens will continue to grow, slowly and at the margins, while ever the ALP let us.

    It reminds me of a phrase I have heard many hacks say: “The ALP are outsourcing its left wing to the Greens”.

  14. @Hamish Coffee

    I’m looking at the AEC data, Antony Green’s data can contain projected figures and it’s not always clear when it does. Since I last commented the postals have been counted and it’s still swinging away from the Liberals on primaries of 2.64% and 2PP swing away of 2.74%.

    Only provisional votes remain to count and there are only a maximum of 466 of those, so the (extremely small) swing away from the Liberals in Bradfield is pretty much definite now.

  15. Where do you get this 2.74% 2PP swing from in the AEC data? The current Liberal 2CP is 64.8%. The 2PP result in 2007 was 63.5%. So the Liberal vote after preferencesd is up 1.3%, not down.

  16. I goofed up on the 2PP, I accidentally used the “last election” column on the 2007 results for Bradfield (67.55 – 64.81 = 2.74). I don’t imagine the swing between 2004 to 2009 means much 🙂

    The primary figure was right though, not that it matters.

Comments are closed.