Time to talk about PR

There’s been a lot of chatter about the low primary vote for the major parties, and Labor in particular polling their lowest primary vote in many decades.

There are plenty of excuses to explain why the vote is so low, such as strategic voting for “teal independents”, but they only explain some of the decline. Ultimately the teal movement is not the reason why a government was elected with the lowest primary vote under our modern party system.

And yes it is true that we have a voting system that (in most seats) makes it easy for someone to, say, vote Greens first and give a second preference to Labor. Some of them may have chosen just for Labor in different circumstances, but there’s also plenty of evidence that there are voters who choose to vote Greens when they can win, but otherwise vote Labor. I cannot say which group is bigger, but what we do know is those voters could have voted Labor and chose not to do so.

So we need to talk about the voting system. Is it really a fair result for a party that wins less than a third of the vote to win total control of government?

There have been some silly suggestions, such as that from Resolve pollster Jim Reed who advocated for first-past-the-post. I think that is the wrong way to move. While the Coalition outpolled Labor, it is clear that a majority of the country would prefer a Labor-led government to a Coalition-led government. I just don’t think they necessarily voted for that Labor government to have a single-party majority. In a first-past-the-post system, people would definitely change how they vote. It seems likely a large proportion of Greens voters would have simply voted Labor, particularly in marginal seats, giving Labor government, but it puts the burden of thinking strategically back on the voter and leads to even more erratic outcomes.

Instead, I want us to start talking about a move to a proportional system.

In this post I’m not arguing for any particular system, although I think if we were to apply a proportional voting system at a federal level without constitutional change it would only be possible to switch to multi-member districts using the single transferable vote, ideally alongside an expansion of the size of the parliament. I don’t think a party list or MMP system is currently constitutional. Still, this argument applies to proportional systems in general.

This election was objectively the most disproportionate House result we’ve had in at least three quarters of a century. The Gallagher index, or the least squares index, measures disproportionality. Very proportional voting systems tend to produce Gallagher scores close to zero, usually below 5. Most Australian elections during my lifetime have produced Gallagher scores between 8 and 12, although the 1975 and 1977 elections scored above 14.

The 2022 score is 16.5.

The Greens and the Liberal Party are significantly under-represented in the current Parliament, while Labor and the LNP (who for this historic dataset had been treated as a separate party, so I've been consistent) are significantly over-represented, with the independents and the Nationals slightly overrepresented.

If you amalgamate the Coalition as a single unit, then the Coalition, Labor, independents and the Katter and Centre Alliance parties between them polled 22.7% less of the primary vote than their seat share would suggest. The Greens polled 9.2% more than their seat share suggests, with the three largest minor right-wing parties (ON, UAP and LDP) under-represented by 10.6%.

So I've always been an advocate for proportionality. I just don't believe that the supposed "stability" arguments for majoritarian voting systems justify putting power in the hands of minority parties (and for the purposes of this essay I define the Coalition as a single party). I think we should elect a parliament that represents the true diversity of the people, and then allow the parties to come to agreements to represent that diversity, rather than hiding that diversity away.

Of course that's not to say we need to have pure proportionality. I wouldn't support an Israeli or Dutch style system with a national PR list. I think geographic representation has some value, if not as much value as the current system places in it, and that you can achieve a middle-range outcome with more than two strong parties but without a totally fragmented political system. Indeed this paper suggests that PR systems with a low magnitude district size can gain most of the proportionality benefits while still electing relatively stable parliaments without too many parties and with accountability for government decisions. The trade-offs are not linear. You can have stability and proportionality.

But even if you don't buy my arguments about proportionality, I think we are now entering a stage in Australia's party system where the electoral system will no longer deliver the stability that is promised, and indeed a proportional system may be the best way to make outcomes more predictable and stable.

Single-member electorates have a tendency to exaggerate swings. A party can win a bunch of seats by slim margins, and thus outperform their proportional share. This doesn't just apply to the major parties. In Sydney and Melbourne, it looks like the teals had a devastating impact on the Liberal Party in certain parts of the city which exceed the actual shift in vote. They won four seats in Sydney, and almost won five. In a proportional result that wouldn't have been so dramatic.

Indeed we saw a similar outcome for the Greens in inner city Brisbane, winning three seats in a small area with no more than 35% in any of those three seats. The preferential system makes this quite possible when the vote is split three ways (as would a first past the post system).

This might be the fair outcome for a single electorate, but when you scale it up to dozens of seats it can be quite erratic and produce lopsided outcomes. You need only look at the late counting in seats like Macnamara, Richmond and Brisbane. A small change in the vote share switches the winner, and you could imagine this sort of outcome happening in more seats if the major party vote continues to decline.

I think it's unlikely the Greens would get to a point where they win more seats than their proportional seat share, but independents and Greens are getting better at taking advantage of these peculiarities of the single-member system and I think we may see more local disproportionalities in their favour, as they wipe out the major parties in one area.

In contrast, a proportional system would probably elect more Greens (and if the teals decided to go in the direction of forming a genuine centrist liberal party, they could elect more too), but spread across a wider range of the country, while the Liberal Party would still have safe seats in places like the North Shore.

Which brings me to my next point: protection of party talent. The Coalition was hit hard by losing a generation of potential future leaders (in the broader sense of the term). Labor was also unable to find an appropriate seat for frontbencher Kristina Keneally, which ended in her losing a previously safe Labor seat.

While there wouldn't be "safe electorates" in the way there are now under proportional representation, the major parties would be assured of a certain number of seats in each area, and prominent members of the party would be able to hold on to those seats with more ease. Exactly how much control the party would have over who holds these seats depends on the voting system - under Hare-Clark it's possible a seat could be "safe for Labor" while not being safe for any individual Labor candidate, but generally prominent senior party figures are able to win even under that system, let alone a mostly-closed list system like that we see for the Senate.

And more generally, a PR system produces less erratic and more stable seat counts. You don't see the kind of major party wipe-outs we saw in Queensland in 2012 or more often in Canada over recent decades. We saw a localised version of that for the Liberal Party in northern Sydney and eastern Melbourne in 2022, and such an outcome could happen nationally. A proportional system tends to result in seat changes that are proportional to vote changes, while a majoritarian system can exaggerate seat changes off small vote changes.

I think we have also reached the point where we can no longer argue that the majoritarian system protects the system of single-party majority government. Yes Labor barely scraped across the line this time, but the range of 2PP outcomes that would produce a hung parliament will end up being much wider than it has been at the past. If the vote for minor parties and independents remains as strong as it was in 2022, hung parliaments will likely be more common than single-party majority governments. So if we will end up with minor parties and independents in the balance of power, the question is how they are elected. Do they win in an unpredictable and erratic manner in particular single-member electorates, or are they proportionally elected according to their strength across the country?

And if we are going to end up with common hung parliaments, it's probably better to give up on the pretence that a government of a single party is better than shared power. Labor could probably govern in current circumstances with less than 76 seats, but I don't think it's a good thing.

Australia's party system has generally worked pretty well with its electoral system, but that isn't the case everywhere. Take a look at Canada, where there are three major parties, hung parliaments are more common, but governments tend to govern in minority. It's quite common for the party that is less preferred overall to stay in government because they happen to have more seats. That sort of thing could easily happen under our preferential voting system. Canada also has a history of parties winning large landslides in seat terms while not doing that well in terms of votes. That was unlikely as long as the seat contests were dominated by the same two parties, but as we get more diversity in types of contests, the possibility of lopsided outcomes becomes more possible.

If we're electing minor party and independent MPs, they should have the opportunity to share power, but also the accountability of being responsible for decisions.

I suspect the Coalition will have a great deal of trouble defeating the independents who beat them in their former heartlands in part because those MPs won't be in any governmental role in the new parliament. While the Coalition will be able to criticise government decisions in seats they are trying to win from Labor, that doesn't apply here.

Wouldn't it be healthier if a bloc of independents like those elected were to take on some ministerial responsibility, take some decisions, and then be accountable for that in 2025?

I don't think we are about to switch to proportional representation, but let's have the conversation. Not just because it's unfair to give total majority power to a party polling under 33% (although it is), but also because the majoritarian voting system is erratic, produces unusual outcomes, and will be a less stable counterpart to a more heterogenous party system.

I'm going to return, hopefully tomorrow, with some analysis of potential models for proportional representation and how they would play out on the results of the 2019 and 2022 elections.

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