Rates of early voting have been increasing for a long time, but reached a record high level of 2022, in the wake of the COVID-19 pandemic. I most recently summarised these statistics at a federal level in this blog post.
For this blog post I’m going to summarise the statistics on how many people are voting early as of the end of the short 3-day week between the Easter and Anzac Day long weekends. The short answer: early voting seems to be holding steady at the record high levels seen in 2022.
The early vote crossed over and made up a majority of votes cast for the first time in 2022. Generally postal voting had been fairly steady while pre-poll voting was continuing to grow, but postal voting also experienced a big surge in 2022, growing from 8.5% of the total to 14.6%.
Similar trends have been evident at state elections. The vote categories vary between states, so it's easiest to measure the share of voters who cast a normal election day vote.
The examples from Tasmania and Queensland suggest that the collapse of election day voting has mostly slowed down, but it has not reversed, and indeed has slightly increased for those states to have held a second post-COVID election. Unfortunately the WAEC has made it harder to distinguish election day votes from pre-poll votes as of 2025, but I will try to make an estimate after the federal election.
So what about the federal election?
The first thing to note is that this period has been significantly disturbed by three public holidays being held during the key campaigning period, two of which are on days when pre-poll voting would have been expected. It also means that this week has been a peak time for people to take off work, as three days of leave can give you ten days off. This makes it a bit hard to compare like-for-like with 2022, when the election was held about three weeks later.
If we look first at pre-poll voting, it's worth noting that up until 2019 there was three weeks of pre-poll, but it has been two weeks since 2022.
The rate of pre-poll voting always starts slow and speeds up, but this year we've had a running start. Over 500,000 people voted on Tuesday, and this reached almost 650,000 people on Thursday. Overall 1.78 million people voted this week. There have only been four other days of pre-poll to ever register more votes than yesterday: the last three days in 2022, and the Friday before the 2019 election.
But the voting period has been compressed, and it is possible this simply reflects people who would otherwise have chosen to vote today or on Monday being squeezed into three days. Once you factor in the zero number of votes to be cast today, the total cast by the end of the Friday eight days out from the election will be slightly higher than 2022 (by about 140,000 votes) and slightly lower than 2019 (by about 143,000 votes).
Overall I think this suggests that pre-poll voting will remain at the high levels we saw in 2022, and could end up a bit higher, but it would be a mistake to assume that the numbers being significantly higher day-by-day this week means that the numbers will be much bigger than in 2022.
The other major source of early voting is postal voting.
For postal voting we get two statistics - the number of people who request a postal vote (PVA) and the number who return the postal vote. The number of returned postal votes won't be final until two weeks after the election, but some have started to come in. PVAs will close on Apriil 30, and they have already slowed to a trickle. Last-minute applications may not turn around quickly enough for the vote to be counted.
It's also worth noting that some PVAs may never be returned. It is perfectly legal to request a postal vote and then decide to use some other method of voting.
The number of PVAs is very similar to 2022. It was a bit ahead of 2022, but fell behind at the start of this week.
The number of actual votes is a long way behind 2022. I assume this at least partly due to the long weekends, and we will see the rate catch up this week, but it's also possible that more postal vote applicants have decided to use a different voting method.
Overall I think we are on track for a similar outcome to 2022, when a slight majority of the country voted early. But I'll do another update next week.
Ben I think that pre-poll will increase on 2022 based on observations & figures so far. Normally tomorrow would be a biggy but with people away probably less so. The big day is set to be Monday.Departing from previous history it is not an RDO so trades people will be back.
@Roger Roughead
16 day break on Brisbane sites. I’d say that would increase rather than decrease PrePoll turnout.
14/4 [monday] was an Industry RDO, 15,16,17 were EBA RDOs, 18,19,20, 21 were Public holidays, 22,23, 24 were EBA RDOs and today is Anzac Day, back to work om Monday the 28th April.
Nice work if you can get it …
Wills, Watson, Blaxland, Reid among lowest PrePoll turnout in the urban seats. Issues with accessibility or just voter disinterest in these divisions.
Also seen in The Australian, AEC has investigated unauthorised letter box drop in Wentworth, came up with figure of 47,000 pamphlets, perp isn’t politically aligned, according to AEC, it’s declining to divulge more information, case closed.
How can they not be politically aligned given the content of what they were putting out? One would think by definition, they are politically aligned. Or does the AEC mean not aligned with a specific political party or entity?
There are too many variables
At Manly, the traffic past the pre-poll comes in waves, synched with the ferry timetables and different on weekdays from weekend days
Some of these visitors dash into the booth on their way to the beach
We have an army of about 30 to 40 hand-out crews most of the time
They perceive different things; such “yesterday was busier that today”
They aren’t (so far)
Next Friday afternoon will settle the question: it’s always “Bigger than Ben Hur”.
@JM From Qld:
Who knows what the content was, Spender hasn’t specifically detailed it, has any third party [apart from the AEC] sighted the material?
Where do I find the figures broken down by polling stations ?
You can find them here.
Firstly, one thing needs to be considered here.
In 2022, there was something like 16 days of pre-poll. In 2025, there are just 9 days.
So its fair to say that this record (as per average) will be broken on current trends.
Which leads me to this point. I was at the Gordon Pre-Poll in Bradfield and it’s fair to say that from what I saw, indicates that.
You can see in the above graph that there was not 16 days of pre-poll in 2022, there was 11 days. So yes there is slightly fewer days but the big drop in voting days was between 2019 and 2022.
So there is absolutely no doubt that the average number of votes per day will set a new record, to the point where it isn’t really worth discussing. The question is whether the total volume of votes will be higher and right now that looks likely.
I wonder why Friday of week 2 is the busiest and why Saturday of week 1 is quieter than most weekdays of week 2.
Generally pre-poll voting gets more popular as you get closer to election day and week 2 is closer than the Saturday of week 1. But also if it’s easier for you to vote on a Saturday you can just vote on election day. A big advantage of pre-poll voting is being able to vote on all the other days of the week.
Got it. There are some who vote to get it out of the way or they’re unavailable on election day or to avoid queues. Ironically, the queue is longer on Friday of week 2 than on election day depending on where you are.
I noticed that Fowler has by far the most prepoll votes or 23% of enrolled voters. I wonder why.
Not sure about that Ben – passed a polling station in Holt on Saturday morning and it was busier than I have ever seen on election day.
Well there are much more booths on election day than at pre-poll, so they can be busier without the total volume being higher.
Anyway you can rely on your gut but I’m relying on hard data – the Saturday before the election is not as busy as the final week, and certainly not as busy as election day!
Yeah true, just wanted to mention how busy the early Saturday was.
I am at the stage where it is now quicker to simply wait and vote on the day
Vote at 5pm on election day – really quiet!
High Street
In fact there is normally a surge at the end that comtinues untill the doors are closed. AEC procedures are that at close of polls an AEC officer is directed to stand behind the last person in the Queue. All of those in the Queue ahead of the designated officer are able to vote. That still does not stop there being people hammering on the door attempting to vote but being denied entry due to the doors being locked once the designated officer reaches the front of the queue.
My legs and the legs of all other volunteers are looking forward to that occurring.
It always amazes me that the candidates have parties after polls close. I just want to go home and put my legs up.
I’ve been handing out on the late shift on polling day across multiple booths over the years. I can agree Andrew about a last spike in people, but it’s more like the last half hour. 5pm is usually still quiet. 4pm and 430pm really quiet.
How’s the Rennick handig out going?
I’ve been very surprised by the scale and strength of the Rennick operation. There were three volunteers in Hotham yesterday, quite incredible, and being listened to too.