Australia

It's about time the AEC adopts a 3CP count on election night for 3-way seats.

For those unaware, on election night, AEC workers count the 1st preferences of votes, and then conduct a 2CP count to the 2 candidates the Commissioner (or the Division Returning Officer, not sure which...but it's someone high up in the permanent AEC staff) has decided are most likely to win. This is great in most cases, but if the AEC gets it wrong it can lead to wild results where the person who was believed to have won on the night has actually lost, and either the 2CP loser on the night actually wins, or even the winner can be someone who made no appearance in the 2CP on the night.

~~I don't know who the 2CP was between on election night in Melbourne (I was too busy doing that counting in Ryan—ours was between Greens and LNP),~~* but the need to "re-throw" between Labor and Greens implies it was probably Greens/LNP? If so that seems strangely out of alignment with previous results so I don't understand why they did that. If they did in fact 2CP between Labor and Greens I wonder why the recount would be so far off of what was declared on the night.

Doing a 3CP count on the night, only in seats where this kind of thing is considered likely, would give a much better indication. Yes, it would be more anticlimactic because you'd lose the ability to confidently declare who did win, but it would at least mean the numbers you're seeing on the ABC (or your media source of choice) are definitely accurate and unlikely to change by large amounts over the coming week. And you can just make an estimate of how preferences will flow from the 3rd in 3CP to the other two. (Before anyone asks, it would definitely not be viable to do a 3CP and then a 2CP on the night. I didn't leave the booth until 11:30 pm last night as it was; extending it too much more than that would be unreasonable. Besides, you couldn't start a 2CP until every booth had done its 3CP, including the postal votes and prepolls. And that's just not how it works. Each booth does their own thing based on guidance set out ahead of time.)

3CP would actually speed up the result, allowing workers to get home earlier and the media to get reliable answers sooner. Each ballot would take a bit longer to count, but the number of ballots to be counted is absolutely decimated (in the modern, not Roman, sense). Instead of counting nearly 700 2CPs, we'd be counting less than 250 3CPs at my booth. The disadvantage is the potentially higher error rate. (A less important disadvantage is the lack of ability to use the 3CP to find errors in the 1st preference results...but you only lose this ability in the 1 candidate that would have undergone 2CP redistribution but is now part of the 3CP...in my booth our redistribution of the Labor candidate results meant we noticed we had undercounted her 1st preference results initially by 1.)

So in summary:

• 3CP would reduce post-election-night surprises
• 3CP would give the media accurate, if incomplete, results that can be used to make informed speculation about the final result
• 3CP would speed up the count on election night, saving the AEC money, their workers sleep, and giving the media information faster

I don't know if this would require legislative change, a directive from some Minister, or just an internal AEC policy change. But whatever it is, it needs to happen.

Ok I found the ABC liveblog from last night which said this:

the electoral commission was counting preferences between the Liberals and the Greens because they were the final two parties in 2022

I don't understand this, because the AEC Tallyroom website says the final two were Labor and Greens. But at least I now know who they were counting...unless the ABC was wrong on both points.

edit:

* Ok I found the ABC liveblog from last night which said this:

the electoral commission was counting preferences between the Liberals and the Greens because they were the final two parties in 2022

I don't understand this, because the AEC Tallyroom website says the final two were Labor and Greens. But at least I now know who they were counting...unless the ABC was wrong on both points.