When results are reported for the NSW upper house, generally there’s an expectation that the primary votes mostly tell us who will win. That’s mostly true, but at the last three elections there’s always been one seat that went to a party that benefited from stronger preference flows – and these stronger flows usually flow in one direction.
At the 2019 election, early figures suggested that Liberal Democrats candidate David Leyonhjelm, who had recently resigned from the Senate, was on track to win a seat. Yet when preferences were distributed, a strong flow of preferences saw the Animal Justice Party’s Emma Hurst pushed from 23rd place to win the 20th seat (out of a total of 21).
This chart is from my guide to the Legislative Council election, by the way.
Labor, Animal Justice and Keep Sydney Open did well in the preference flows, including from the Greens, Voluntary Euthanasia and Sustainable Australia. The Liberal Democrats had already been overtaken by Animal Justice before Keep Sydney Open was excluded, with their preferences pushing Labor level with the Nationals for the eighteenth and nineteenth seats, with Animal Justice on track for the 20th seat, and One Nation winning the final seat.
This trend has parallels with the results from 2015 and 2011. In 2015, Animal Justice candidate Mark Pearson was trailing No Land Tax candidate Peter Jones until the exclusion of the third Greens candidate, which pushed the AJP in front, and they went on to win.
In 2011, independent candidate Pauline Hanson was in twentieth place on primary votes, ahead of Nationals candidate Sarah Johnston (now Mitchell) and Greens candidate Jeremy Buckingham. There was only room for two out of these three to win. Buckingham and Johnston both gained on Hanson but she was still leading both of them right up until the point when Family First's Gordon Moyes was the last candidate eliminated. While his preferences mostly favoured his former Christian Democratic party, enough votes flowed to the Coalition and the Greens that Hanson missed out. Buckingham came 20th, with Mitchell outpolling Hanson by just 1306 votes.
We've had five elections under the current upper house electoral system. In 2003 and 2007, preferences played no role in changing the outcome, to the point where I remember thinking that they didn't particularly matter. But they have become increasingly important.
So I wanted to dive into an examination of what has been changing in how people actually vote. To do this, I looked at the raw preference data published by the NSW Electoral Commission which includes every preference indicated by every voter. Unfortunately they only publish this data dating back to 2011, so we don't have the same richness of data for 2003 or 2007. I've also examined the research papers Antony Green has published after every state election via the parliamentary library. His analysis of the 2019 Legislative Council election was recently published and he has a blog post about it.
They key statistic is the proportion of votes cast above the line with a single '1' and no further preferences. In NSWEC parlance this is a Single Above The Line (SATL) vote, as opposed to a Random Above The Line (RATL) or Below The Line (BTL) vote.
The SATL share was 82.2% in 2011, and increased slightly to 83% in 2015. In 2019 it dropped notably to just 69.7%. This means that prior to the last election, generally about one in six voters bothered to mark multiple preferences, but it's now more like three in ten.
But the story is not just about a general increase in preference rates - it's who is preferencing. So I grouped together parties into clear "left" and "right" groups. There's clearly a left-right dynamic in play in the 2019 results, with a variety of parties on the left helping AJP and Labor strengthen their final position. I classified Labor, the Greens, Animal Justice, Voluntary Euthanasia, Socialist Alliance, the Democrats, Keep Sydney Open and Jeremy Buckingham's independent ticket as "left", while the right includes the Shooters, Small Business, Christian Democrats, One Nation, the Liberal Democrats, the Coalition, the Australian Conservatives, Outdoor Recreation Party and Family First.
While the overall statistic was mostly steady in 2015, at the individual party level there was a shift. SATL rates went up for the CDP, Shooters, Coalition and right parties overall, while dropping for parties of the left. The Greens' SATL rate dropped from 78% in 2011 to 67% in 2015.
Then in 2019, the left maintained that edge, with the Greens having their SATL rate drop to just 53%. Less than half of Keep Sydney Open and the Socialist Alliance votes were SATL.
At the other end of the scale, One Nation, Lib Dems, Shooters and the Coalition all ranked in the top five groups for SATL votes.
With a lower overall SATL rate, that means there are a lot more preferences sloshing around, and it's easier for a party to chase down a rival and win one of the final seats. With the left outpacing the right in terms of preferences, that gives them a boost.
The Animal Justice Party would have won neither of their seats if results at the last two elections reflected the primary votes. Those seats would have instead gone to the Christian Democratic Party and No Land Tax.
The current Legislative Council has a slim right-left majority of 22-20. The "left" minority includes Labor, the Greens, Animal Justice and an ex-Green independent. But the parliament would have likely looked quite different if the split was 24-18.
It will be interesting to see how these trends progress in 2023. Does the SATL rate continue to decline, or possibly rebound? Can the parties of the right catch up with those of the left?
While we have some sense of what has been happening, we don't really know why. One compelling theory is that 2019 was the first NSW state election since the change in the federal Senate voting system, which in practice is similar to NSW but involves a strong encouragement to number at least 1 to 6 above the line (whereas NSW is more agnostic). Since 2019, we've had two more federal elections. Perhaps this will drive the SATL rate down further.
The rate could also be affected by parties making a strong effort to encourage preferencing, by emphasising their preferences on the how-to-vote card or by other communications. And I would expect they are influenced by the current political environment - preference flows from the Greens were much lower in 2011, when an outgoing Labor government was unlikely to inspire a spirit of cooperation.
Unfortunately we won't know these preference trends until well after the results are declared, but we will need to bear this in mind when we examine the first results in the days after the election - preferences really could make a difference.
Interesting and thanks for that analysis Ben.
I am not a maths nerd, so I won’t go into that detail. But I wonder what impact the Teals had / will have on this voting below the line. Maybe people saw the power in taking back their power and doing their own thing, instead of what the party dictates.
I know that at the last Federal election I voted strategically below the line for the senate, but I always knew I was in a small minority.
I would argue that one of the reasons for this is that the ‘new left’ parties are all essentially offshoots of the broader Greens movement, sometimes aligned with broader objectives but with a specific policy difference. So it makes a lot of sense to say vote for the smaller party then preference the ‘parent’ party.
On the right though the parties themselves, or at least a large proportion of the voters, are offshoots of the old Labor Right. They have little in common with not just the modern Labor party but also the traditional ‘patrician’ Libs, and even with the other small parties of the right. So there is no commonality of interest, therefore the preference flows go everywhere, or even none at all under OPV.
The Teals are interesting, because although culturally they are closer to the new left politically they are really old fashioned patrician (matrician?) Libs. Not sure what will happen with OPV if there are a significant number of them standing.
Labor voter, I think most of the teals generally run for the lower house only.
In terms of the upper house, a similar comparison could be to nick xenephons former party, he ran it like the old Australia democrats as a centrist movement similar to the teals.
the teals are Labor candidates in Liberal clothing.
Ben, whilst I do agree with you that many teal candidates are ex Labor members, many may be running because they are from the right/business friendly faction of the ALP. They feel the current Labor Party no longer represents their interests and they cannot fully switch over to the Liberal Party because they feel the Coalition is too socially conservative.
I think the teals would fit with centrist members who are economically conservative but socially progressive. Emmanuel Macron as French President fits this category, he even started a new centrist party that I believe holds a parliamentary majority. In the US, there are congressional members from both sides who were dissatisfied with the direction of their former parties and thus defected. Some examples include Justin Amash (Libertarian, anti-Trump Republican) and Kristen Sinema (Independent, conservative Democrat).
A bit late on this, but.
The Teals are clearly not Labor – or at least not what we used to think of as Labor. If the Teals had been members of the ALP, then that is both a point in favour and against the realighnment.
In favour – this group people (Uni educated, probably privately schooled, in high paying, or will be in high paying, roles) have nothing in common with high school educated lower paid manual labourers, so seeing themselves as part of Labor suggests Labor is not what it was.
Against – that this group of people are moving out of the ALP and becoming indi’s suggests that the ALP still has at the very least a residual element of class solidarity.
@mostly labor they are running as independants because they cant get elected under a labor banner
Comments are closed.