The two-party-preferred vote is the simplest statistic we have in Australian elections – just two numbers that always add up to 100%. While it has reduced relevance for calculating the result in some seats, it still has predictive power and also plays a role as a sort of barometer of the relative popularity of the two major parties. Despite the increase in the vote for minor parties and independents, the choice of who forms government is still binary.
For today’s post, I am looking at what makes up that figure today. How preferences flow between Labor and the Coalition, both in aggregate and from particular minor parties, and once we calculate this we can look at which votes actually make up the 2PP vote for each of the major parties.
Let’s start with preference flows. The AEC now calculates 2PP preference flows for each party – ie how many Greens votes flow to Labor or Coalition. But this data has only been collected since 1984. To go back further, we can just calculate how the total pool of all preferences flowed between Labor and the Coalition, by comparing the primary vote to the 2PP.
The Coalition used to do best out of preferences, but since the 1980s they have favoured Labor.
There was a lot of talk about preferences becoming more favourable to the Coalition in 2025, but that didn’t happen. There was almost no difference, with preferences shifting in Labor’s favour by 0.03%.
The story is a bit more interesting if we look at the flow of preferences for particular parties.
It may be hard to believe, but the rate of Greens preferences flowing to the ALP continues to climb, up from 85.7% to 88.2%.
On the other hand, One Nation preferences have become substantially more favourable to the Coalition, increasing from 64.3% to 74.5%.
It appears that, while the major party vote has continued to drop, there is increasing polarisation of voters into two groups – Greens voters overwhelmingly favouring Labor, and One Nation favouring the Coalition. These two parties polled more than half of the minor party and independent vote between them.
Meanwhile independent preferences continue to become more Labor-friendly in their make-up, with 67% preferencing Labor over the Coalition. This vote category had given a majority of preferences to the Coalition up to 2010, but the wipe-out of rural independents in 2013 seems to have flipped the ratio. The independent vote now has a strong teal tinge, with a lot of centre-left independent candidates who take votes that would have otherwise gone to Labor or the Coalition.
While not every independent fits the mould, there is an increasing number who can be described as “teal-adjacent”. Using Kevin Bonham’s classifications, I found that this group had polled 5.4% at the 2025 election, which is just under three quarters of the total independent vote. Unfortunately 2PP preference flows are only published by party bloc at the state level, so it’s not possible to isolate particular independents. However I think it’s likely that this bloc preferences Labor at an even higher rate, so combined with the Greens and One Nation it suggests an increasing polarisation of preference flows for minor parties and independents.
Since we can split the votes for all the other groups between Labor and the Coalition, we can see what votes make up each of the 2PP coalitions.
When I did this analysis in 2022, there was a more stark difference, with the Coalition’s primary vote making up a bigger part of a smaller pile than for Labor. But the dropping Coalition vote means that it is starting to look similar.
Labor’s primary vote ended up on 34.6%, out of a total 2PP of 55.3%. That means that 20.7% of the electorate voted for someone else but preferenced Labor over the Coalition.
For the Coalition, their primary vote of 31.8% made up just over 71% of their total 2PP of 44.7%. So almost 13% of the electorate voted for someone else but preferenced the Coalition.
I finally decided to calculate the proportion of each party’s 2PP that came from preferences, dating right back to 1958.
Since the late 1980s, Labor’s 2PP has had a more significant contribution from preferences, but the number has gone up for both parties.
Interestingly the proportion of Labor’s 2PP that came from preferences dipped slightly this year, but the Coalition’s figure reached a new height. The Coalition’s 2PP is now as influenced by preferences as Labor’s was in the 2013-19 era.
That’s it for today. Still to come: what made up the 2CP in key crossbench seats, and the distribution of 3CP.
I well remember in my teens, first becoming involved in politics and some Labor people strongly advocating ‘first past the post’ while Liberals and Nationals (then called the Country Party) strongly defending compulsory preferences. This changed, and your chart nicely shows exactly when it changed – after 1977 when the DLP influence ended and the Australian Democrats began to rise. What would be interesting would be to do this analysis state-by-state, so that we could see whether some states are different from the average. For example, in regional Queensland it might be the case that stronger votes for One Nation and weaker votes for the Greens have the LNP ‘advantaged’ by preferences. In Flynn, the LNP got 61% of preferences, just one case in point. Of course in reality there’s no advantage or disadvantage built into the preferential system, it’s the fair and reasonable expression of the voter’s will. And the voters make up their own minds about preferences.
Interesting analysis Ben
I wanted to point out Family First they ran for the first time in 2004 and came back in 2025 they did not run in the 2019 and 2022 elections. There preference flow has been consistantly 2:1 in favour of the Coalition so not much change
Good job Ben.
I wonder about this.
If you’ve already done it, let me know.
What is the split of the 150 Seats: How many seats are led by either – ALP // L/NP // Others? (Primary vote only).
And how do the trends look on that from prior elections.
In terms of 3CP.
That’s it for today. Still to come: what made up the 2CP in key crossbench seats, and the distribution of 3CP.
I guess you could create a 3CP for each electorate if you have the distributions – would all seats end up as ALP // L/NP // Other?
Are there any seats in which ALP or L/NP did not make the 3CP?
And how does this compare to prior elections as well.
Obviously the simplest way to do a 3-Way Breakdown is just to use the Primary votes of ALP // L/NP // Others and compare that back to prior elections.
Using the redistributions you could do a trend on 3CP going back elections by going seat-by-seat. A bit more work.
Why were One Nation preferences to the Coalition relatively weak between 1998 and 2016?
Yes I have 3CP data for every electorate, dating back to 2004. Basically you just need to work out which count is the second-last count. You can either do that by calculating the number of candidates running and subtracting two, or by identifying what the final round of the count was in each seat and subtracting one. Then you just take that count round for each seat from the DoP files.
I’ll be coming back to this, probably on Friday, but there are 2 seats where Labor didn’t make the 3CP. In the seats of Indi and Calare, it seems to be the first time since at least 2004 where two non-major party candidates made the 3CP – every other time when Labor hasn’t made the 3CP it’s because it is Lib vs Nat vs Ind. That has only happened three times from 2004 to 2022.
And yes you could convert 3CP data to reflect redistributions, like we do with 2PP and primary, but I haven’t done that.
@Julian – I believe in Calare Labor missed out on the 3CP. It was made up of the Nationals, Andrew Gee and Kate Hook.