Preference data that could inform a hung parliament

6

There is a lot of speculation about the potential for a hung parliament in NSW, legitimately in my opinion.

In today’s post, the last examining preference flow data from 2019, I look at a number of datapoints that might tell us something about how the crossbenchers elected in 2019 might jump in the circumstances of a hung parliament.

Most of the data in this blog post is based on preference flow data that will likely take some time to become available following this election. So if there is a hung parliament we likely won’t have the updated version in time. It would definitely be preferable to use the 2023 data for this question – there may be new members of the crossbench, and voters may shift their preferences in the changed political context – but I still think it has value to examine some trends from 2019.

So let’s start with preference flows from voters for these nine MPs. This first chart looks at how primary voters for these nine crossbench MPs preferenced Labor and the Coalition (two-party-preferred preference flows).

Unsurprisingly Greens voters in Greens seats strongly favour Labor over Coalition, more than Greens voters in other seats. We already knew that.

Piper voters slightly favour Labor over Coalition (but the vast majority exhaust). More than twice as many Greenwich voters preference Labor over Coalition, but it's not the kind of large numbers you see for Greens voters. I suspect this reflects how-to-votes, as Greenwich just issued a 1-only HTV.

McGirr's preferences also slightly favour Labor.

Amongst the Shooters, Fishers and Farmers candidates (all now independents) there's an obvious difference between Roy Butler in Barwon and the other two. Preferences for SFF voters in the other two seats slightly favour the Coalition with a very high exhaust rate, but preferences to Labor in Barwon are substantially higher - although still almost two thirds exhaust.

We can generalise this data a bit more by looking at the total two-party-preferred vote in each seat. So now we're not just looking at primary votes for these candidates but the total vote for all candidates in each seat.

Labor wins large 2PP majorities in the Greens seats. They also win a majority in Lake Macquarie, but with two fifths exhausting.

The Coalition slightly pips Labor at the post in Sydney, which is interesting. I have a theory about what is happening here which I will get to later in this post.

The other four crossbench seats all have Coalition 2PP majorities. The strongest are in Murray and Orange. Labor actually comes quite close in Barwon.

This next dataset is not based on preference flows but I think is also similarly useful in understanding the political positioning of the voters in these seats, particularly in the independent seats. This next chart shows how voters in these seats voted for the upper house. This dataset should be available in the decision-making period after an upcoming election.

The Greens and Shooters unsurprisingly maintain their strong position in the upper house (although usually with a slightly lower vote than in the lower house). But independents can't run for the upper house.

Wagga Wagga clearly has a conservative lean, with close to 60% voting for the Coalition, One Nation or the Shooters.

Lake Macquarie has more of a leftward lean, with over 42% voting for Labor, Greens or Animal Justice and 45% voting for Coalition, One Nation and Shooters.

Sydney is very interesting. Over 40% vote for Labor or the Greens, with another 10% for Keep Sydney Open, whose policy program was centred on issues affecting this seat. The vote for the right-wing minor parties was tiny and the Coalition polled under 34%.

My theory in the case of Sydney is that this seat shows a conservative two-party-preferred majority despite having a progressive majority because the nature of the contest means independent and Greens voters who would otherwise preference Labor just don't bother.

For Greenwich voters, there is no how-to-vote card recommendation to preference Labor. For Greens voters, the party recommended a third preference after Greenwich, but the preference flows were not impressive compared to seats with similar demographics. I ranked the proportion of Greens votes that flowed to Labor on the 2PP by seat, and Sydney came 35th in the state, while Balmain and Newtown ranked near the top.

I suspect for these voters it was enough to preference Greenwich and then exhaust. Whereas Coalition voters count to the 2PP at full value. So I suspect that Greenwich's seat looks more conservative than it is thanks to other progressive voters not bothering to preference Labor. This could also be a small factor in Lake Macquarie - a proportion of progressive voters who vote 1 for Piper would exhaust, but every conservative voter for the Coalition counts to the 2PP.

Beyond Sydney, what do these trends show us?

Well, the Greens seats are very left-wing and those voters are comfortable in giving Labor a second preference. The Greens vote in the upper house is a bit lower in these seats but the Labor-Greens combined majority is still quite large. It comes as no surprise to say that it would be extremely damaging to the Greens in these seats to be seen to be helping a Coalition government, which Labor has been trying to exploit in Balmain.

Lake Macquarie has some lean to the left but it is not a strong one. I think it could hurt Piper if he backed a Coalition government but it is not so clearcut. In Sydney I do think Greenwich would face a backlash from backing the Coalition that exceeds the 2PP.

For the four rural independent seats, the Coalition clearly has the edge. Labor does respectably in Barwon and Wagga Wagga, but not so in Orange and Murray. I think it could be risky for the independents in Barwon and Wagga Wagga to be seen to be backing a Labor government, but it would be much worse for those in Murray and Orange. Of course there may be a difference between choosing Labor when the parliament has first been elected and then choosing to work with that government once it's in place. Labor may still find that it is useful to work with the more conservative crossbenchers than the Greens from time to time.

This is the final part of my series on preference flows. I'm going to do some analysis of who is running later this week but will also return at some point with an analysis of the potential dynamics in a hung parliament.

But finally for this blog post, this chart is a bit more complex but should be interesting. Just like how in yesterday's post I broke down the two-party-preferred vote by primary vote, for this chart I've broken down which primary votes made up the two-candidate-preferred vote for the crossbench MP and their major party opposition in these nine seats.

In most of these seats the sitting MP is in a very solid position just on primary votes. It's worth noting that Greenwich does receive a substantial boost from Labor and the Greens (over 10% of the formal vote). Roy Butler also gets a lot more Labor preferences in Barwon than the other ex-Shooters. There really is only one seat where the primary vote was not the basis for winning. Tamara Smith in Ballina won despite being outpolled by over 5% on the primary vote, with a strong preference flow from Labor putting her over the top.

Ballina is a good example of wanting to think through who actually helped an MP get elected. While Ballina has a conservative history and a strong Nationals vote, ultimately it's the Labor voters who give the Greens the boost they need to win there, and I'm sure she'll be keenly aware of that fact if we end up in a hung parliament.

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

6 COMMENTS

  1. Interesting data. Hopefully you will do this for the 16 crossbenchers elected at the 2022 federal election before the next federal election.

  2. I wonder if the exhaustion rate will go down since the federal election is still relatively fresh in our minds and only 10 months ago, we had to number every box.

  3. When did NSW switch to OPV from CPV? According to Wikipedia the 1941 state election used CPV, so what year was the switch, and what was the motivation to change it?

Comments are closed.