In 2022 I wrote a blog post where I explained the phenomenon of the AEC needing to do more distribution of preferences before declaring election results, and thus needing longer to complete this process. We now have updated information for 2025, and can see that this process has become even more complex this year. When you look at how this has changed over two decades, it is a remarkable transformation.
The AEC is required to officially declare the winner of each seat before eventually returning the writs. They will eventually produce full distribution of preferences for each seat, along with two-candidate-preferred, two-party-preferred figures and various preference flow data, at the level of the Division and the polling place. But they are free to declare the winner once there is mathematical certainty.
They can do this even before the final deadline for votes to be returned, if the result is clear enough. Apparently the seats of Cook and Maranoa are ready to be declared tomorrow. In Maranoa, the sitting MP David Littleproud is leading on the primary vote by enough that it is not mathematically possible on the potential remaining votes for him to lose. In Cook, Simon Kennedy does not have a majority of the primary vote but it is not mathematically possible for the Labor-Liberal two-candidate-preferred count to be changed, and he is winning that count by a sufficient margin.
A result can be declared on primary votes if there are not enough outstanding votes for the leading candidate to fall below 50%, or on the two-candidate-preferred count if there is no possibility of a third-placed candidate making it into the top two on the preferences of other candidates.
But where there is some possibility – however slight – that one of the top two could be overtaken by the remaining candidates, then a full distribution is necessary before declaring the result.
This is a different standard to the standard applied by those of us “calling” results. We can do that even when there is a mathematical possibility of the result changing, but when we feel confident that it is very unlikely to change. There is only one seat where we are waiting on a full DoP to know who has won – Calwell – and a handful of others where a 3CP is necessary to know who has made the 2CP and thus who has won – Ryan, Grey, Fisher, Flinders come to mind.
The AEC’s media briefing earlier today reported that they expect to need to conduct a full DoP before declaring in 93 seats. This compares to 76 in 2022 and 34 in 2019. In 2004, there were just four seats with this status.
The number of seats where the winner has won a majority of the primary vote has crashed, down to 15 last time and right now on track for 13 seats in 2025.
Initially, as the number of seats where the winner had a primary vote majority started falling, there was an increase in the number of seats where there was a mathematically clear top two. Yet that number is falling now as well.
This is all a consequence of how voters are changing how they vote. The dropping vote for the major parties means the top two is less and less clear, and it is becoming increasingly rare for an MP to have received the first preference of a majority of their voters. This is the cause of a lot of the extra complexity we’ve seen this year, and I see nothing to tell me we’ve yet reached the end of this trend.