Australia’s new pendulum

We’re still waiting for the electoral commission to release detailed figures on preference flows from Australia’s 21 May federal election. But from the abundance of results already available both Antony Green and Kevin Bonham have produced versions of the post-election pendulum, showing the swings now required for seats to change hands.

Bonham’s post in particular includes a lot of interesting commentary on the results, including a discussion of why he thinks his way of presenting the pendulum is better than its rivals. He also explains in some detail, with nice graphs, why the two-party-preferred vote (and swing expressed in those terms) remains a good predictor of results, despite the greatly increased vote for minor parties.

The new pendulum isn’t that significant in itself, because redistributions will change some of the figures before the next election – New South Wales’s boundaries have to be redrawn next year, and movements in population mean that Victoria will almost certainly lose a seat and Western Australia gain one. But it tells us some interesting things about this year’s result that are otherwise not obvious.

The median seat in two-party-preferred terms was Ryan, in Brisbane. It was actually won by the Greens, but when counted out as Labor vs Liberal, Labor had 52.4%. Since Labor won 52.1% of the two-party-preferred vote nationally, that means the boundaries worked slightly in its favor, but a difference of just 0.3% is pretty good.

A uniform swing of 2.4% would therefore give the Coalition an “underlying” majority – that is, it would win the two-party-preferred vote in 76 of the 151 seats, a gain of nine. But that doesn’t mean it would win an actual majority. The 76 include 11 seats currently held by crossbench MPs (one each of Greens, Centre Alliance and Katter’s Australian Party, plus eight independents), and there’s no guarantee that the Coalition would win them back even with a fairly substantial two-party swing.

Mayo, one of the nine notional gains, illustrates the problem. At the previous election, in 2019, the Liberals won the two-party-preferred vote there with 52.5%. But they didn’t win the seat: Centre Alliance’s Rebekha Sharkie had a much better preference flow than Labor would have if it had been in the top two, and she won the seat with 55.1%. This time she improved that to 62.3%; it would require a mammoth swing to unseat her, much more than the 1.6% required to flip back the two-party-preferred vote.

Of the crossbench MPs who sit for what would otherwise be Coalition seats, only Bob Katter could be counted as a reasonably solid vote for a prospective Coalition government. The others, including the “teal” independents who won a swathe of inner suburban seats, would at the very least demand a high price for their support. And so far Labor has much less of a problem on its side – although the victory of Dai Le, a Liberal-leaning independent who won the previously safe seat of Fowler, could be a sign of trouble to come.

The other thing worth looking at is the overall proportionality of the result. Here, using the commission’s national aggregate figures (but consolidating the Coalition vote), are the top scoring parties in the House of Representatives, together with the seats won, the percentage of seats won, and the seats they would have won according to two different systems of proportional representation, Sainte-Laguë and D’Hondt (here’s a similar calculation in 2019):

Party% voteSeats won% seatsSt-Lag.D’Hondt
The Greens12.3%42.6%1919
One Nation5.0%00.0%87
Liberal Democrats1.7%00.0%32
Animal Justice0.6%00.0%10
Katter’s Australian Party0.4%10.7%10
Australian Federation Party0.4%00.0%10
Centre Alliance0.3%10.7%00
All others1.7%00.0%00

As you can see, both the Coalition and the independents won seats roughly in proportion to their vote.* But Labor is massively over-represented, and both the Greens and the smaller minor parties (most of them on the far right) are correspondingly under-represented. On a Sainte-Laguë count, as used for example in New Zealand (although without its high threshold), the combination of Labor, Greens and independents would have won the same total as Labor actually won on its own.

Neither of our major parties has shown the slightest interest in moving towards a more democratic electoral system, a fact that may not be unconnected with their steadily declining share of the vote. But even with single-member districts it’s becoming more and more difficult for either of them to win a majority; it’s only a matter of time before power-sharing of some sort will be required, and that will inevitably put electoral reform on the agenda.

We do, however, have another chamber, the Senate, in which equal representation of states is balanced by proportional representation within each of them. Somewhat in spite of itself, it does quite a good job of providing fair representation, but there are some quirks. We’ll look at that in another post.


* Grouping the independent vote as a single total assumes that in a proportional system they would have co-operated to form a nationwide list, which of course would not have happened without complications. But the fact that most of the successful independents were in fact part of a loose alliance makes the assumption more realistic than usual.


5 thoughts on “Australia’s new pendulum

  1. A parliament without a majority would not inevitably put electoral reform onto the table. None of this is to argue shoulds. If I had my druthers, it would be STV with ungrouped Irish style ballots. But I just don’t think the evidence suggests a crossbench with balance of power is a motivator for electoral reform. Historically it hasn’t been. Electoral reform would not necessarily be in the obvious self-interest of the crossbench (who got voted in on personal appeals to local voters). It would probably be perceived by most people in parliament as in the partisan interest of the Greens (and therefore hard for them to press in coalition negotiations – would they really trade a certain cabinet seat for a promise to introduce a bill that might not get through the Labor caucus or the senate and which will get dropped as soon as someone mentions a constitutional concern?). It would be against the interest of senators (who would go from being more representative in one way and less in another, to being less representative in both ways – and thereby lose all bargaining power). And in a country that seems like it’s becoming more focused on local and less focused on national – especially through the distrust of national media and party leaders – it’s not obvious that voters are going to start thinking of the vote/seat relationship nationally in a way that delegitimises electorate wins, especially since a majority of people in overwhelmingly most electorates will have an MP who they thought was reasonable even if it wasn’t their number 1.

    To expand on the history – electoral reform came onto the agenda in New Zealand at a time of majority governments – because people weren’t represented at all, not because they were represented badly. Canada and the UK have had three or more parties in parliament for most of a century, including various minority governments and coalitions, and they’ve been able to fend off calls for democratic improvement. Generally, to the extent that democratic improvement has come onto the agenda in any of those countries, it was because the wrong side won elections – a majority of people voted for parties on one side and the other side formed government. At an earlier time, the Netherlands and Belgium introduced proportional representation to prevent labor/socialist majorities rather than to resolve problems with parliamentary minorities.

    Electoral reform doesn’t come because some in parliament argue that there’s a better electoral system; it is necessary that the current electoral system should seem illegitimate to a majority of voters or parliament. So probably a result like 1998 is more likely to put it on the agenda than a result like 2022. Maybe 1998+2022 would do it. Another cause could be seeing successful switches in Canada and the UK in quick succession – I doubt that seeing a shift and being a laggard would change enough minds, but it might.


    1. Thanks Felix! Yes, I fear you’re mostly right there. I do think a parliament with no major party majority would see the subject raised, but I’m far from confident that it would end up going anywhere. Reform in New Zealand was probably only possible because of the lack of an upper house (or states, for that matter), leading to governments that were all-powerful as well as unrepresentative. Perhaps a 1998 but with the Coalition on the losing end might make a difference.


  2. Using the Senate as a likely barometre of how people would have voted in the Reps had the House been PR, the 2019 federal election would have resulted in a weak Liberal-Nation government dependent on One Nation.

    The moral of the story?

    Australia is NOT going to adopt PR.


    1. Thanks Patricia – good point! I think there’s a chicken-and-egg problem; until we get more responsible-looking minor parties (particularly in the centre), it will look unsafe to move to a system where minor parties might have more power, but it’s the move to that sort of system that is likely to produce more responsible-looking minor parties. But perhaps if the teals turn themselves into a party then things will look different.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.