Daily updated post tracking the rates of postal and pre-poll voting compared to previous elections.
State and territory elections over the last two years have seen a huge increase in both pre-poll and postal voting. I’ve written a number of posts on trends at those elections. I’ll include some relevant links at the end of this post.
I’ve also published a second post with a sortable table and graph showing the rates of pre-poll and postal voting by division.
The time between close of nominations and polling day is one week longer in 2022 compared to 2019, four weeks versus three weeks. In contrast, a change in the law means that pre-poll voting will be one week shorter in 2022, confined to only two weeks rather than the three weeks allowed at previous elections. The change means that where in 2019 pre-poll voting and postal voting started at the same time, in 2022 there have been two weeks for parties to flood the electorate with postal vote applications before the start of pre-poll voting for two weeks.
Summary of Early Voting statistics at the end of the pre-polling day period Friday 20 May –
- (Postals for Friday not yet available) With the closing date for postal vote applications now past, a total of 2,731,060 postal vote applications had been received representing 15.9% of enrolment. This compares to 1,538,139 in 2019 or 9.4% of enrolment.
- 1,644,061 postal votes have been returned representing 9.5% of enrolment or 55.0% of dispatched postal vote packs. There have now been more postal votes returned in 2022 than the total number of postal applications in 2019.
- In 2019 84.0% of postal votes dispatched were returned, though 3% postals did not make it through scrutiny so only 81.0% of postal votes dispatched made it into the count.
- Postal votes admitted to the count in 2019 represented 7.6% of enrolment, or 8.2% of votes. (Based on House ballot papers admitted.)
Summary Pre-Poll Statistice –
- Final Pre-polls to total is 5,541,757 to a total of 4,778,856 in 2019 in the reported daily figures. There were just under a million votes taken on Friday, 923,852 in total and 208,000 more than on the final day in 2019. There have been five fewer days of pre-polling in 2022 because of the change in the law, but the number of pre-polls has passed the number taken in the longer period in 2019. See graphs below. Pre-polls represent 32.2% of enrolled voters compared to 29.9 in 2019.
- Pre-poll figures for both 2019 and 2022 are of all pre-polls issued at early voting centres so includes both within district ordinary pre-polls and out of district pre-poll declaration. Based on 2019 experience, most pre-polls will be within district pre-polls.
- There were 4,908,831 pre-poll votes in 2019 representing 29.9% of enrolment or 32.5% of votes counted. This is the final audited pre-poll figure which is slightly higher than the reported daily totals used above in comparison to the daily totals in 2022.
- In 2019 there were 4,288,451 House votes cast as pre-poll ordinaries, that is at a polling place for the voter’s home division. These represented 28.4% of votes counted.
- There were another 620,380 House votes cast as pre-poll declaration, largely pre-poll absents cast outside of division, representing 4.1% of votes counted.
Early Votes Taken to Date
The chart below shows the Pre-poll and Postal Vote applications by state as a percentage of enrolment.
Comparing Pre-poll Voting at the 2019 and 2022 Elections
The first chart shows the number of votes per day. The blue dots are 2019 figures with the number increasing daily through the campaign with the fewest number taken in the first week. The green dots are 2022 daily figures by day.
The next chart shows the cumulative numbers as a percentage of enrolled voters. The percentage will be higher as a proportion of votes cast. The horizontal axis shows days of pre-polling over the three weeks in 2019 and the same for the two weeks in 2022.
Comparing Postal Voting at the 2019 and 2022 Elections
Graphs comparing postal votes can be a little complex. There is always a big surge of postal votes added to the AEC statistics with the close of nominations. This is caused by the inclusion of permanent postal voters to the tally with the close of nominations. With the 2022 election closing nominations a week earlier than 2022, the comparison is complicated. I've decided just to plot the last three weeks of statistics as shown in the chart below. I'm also including the rate of postal votes returned.
Past Record of Pre-Poll and Postal Voting
The chart below shows percentage of votes by category at elections since 1993. It shows clearly the rise in early voting over the last 15 years.
Until 2007 all pre-poll votes were declaration votes, and like postal votes, required a voter to tick one of a category of reasons why the voter was casting a pre-poll or postal vote. Scrutineers were entitled to challenge votes on the basis of the reason. The inclusion of reason on the declaration was removed after the 2007 election. Pre-poll voting within division was converted to voters making an oral declaration, having their name marked off the electoral roll, and being issued ballot papers which were placed directly into a ballot box on completion. Pre-poll ordinary votes are available for counting on election night.
I go into more detail on the background to the change in pre-poll voting in a previous post, Should we Count Pre-poll votes before 6pm on Election Day?. The idea of counting pre-polls early was rejected, but the AEC is now permitted to open pre-poll ballot boxes from 4pm on election day to sort and unfold ballot papers. Counting won't start until 6pm, but the extra time for unfolding
Two-Party results by Vote Category
The chart below shows the two-party vote for categories of votes cast at the 2019 election. The percentage of votes in each category is shown on the left.
As shown, Labor won the vote on polling day with ordinary and absent votes, but the Coalition's victory was built upon stronger performance with pre-poll voting and especially with postal votes.
Note that "Other" votes includes, hospital, remote and other forms of mobile voting.
Previous Posts on Early Voting
2022 South Australian Election
2021 Western Australian Election
Hi Antony: Off topic but on election night will the ABC computer/new break out the national vote share of the ‘Teal’ independents? Whilst not a party in an electoral law sense who is a ‘Teal’ is a closed class (they are listed on the Climate 200 website) and it is pretty clear a Teal will be in the last two of at least six electorates. It would be helpful to have that as a top line when considering how far the LNP primary vote may drop at the end of the day.
COMMENT: We’ve considered and decided not to. The Independents have argued consistently they are Independents and not a political party. It is not for the ABC to start re-classifying them as a pseudo-party. We are classifying candidates as they nominated and as they appear on the ballot paper. The true impact of the Independents will be in individual seats and their vote will not be very high on a national level. Plus there are other Independents with a chance of election who are not associated with Climate 200. They will be treated as the Independents they say they are.
Thanks for your analysis, Antony. Which parties sent out postal vote applications with their promotional material to residents. And is this a new thing? Has it assisted to increase the number of postal votes for this election?
COMMENT: The major parties mass mailed postal vote applications. The fact there were two weeks between close of nominations and the start of pre-poll voting, combined with mass application mail-outs, has played an important part in the postal vote increase. I was always of the view that shortening the pre-poll period would would cause parties to do mass mail-outs and increase postal voting.
Requiring all postal vote applications to be returned directly to the AEC instead of to party campaigns would reduce the attractiveness of postal vote mailouts. People who really need a postal vote would not be disenfranchised, and it’s really easy to apply online and vote delivery speeds are fast.
Thanks Anthony. One donor may donate to multiple people but it’s not right that they grouped together because of the donation. If there is an agreement in place then they should make that public but since they have stated that there aren’t any binding agreements then donation alone is not sufficient.
Hi Antony, will the AEC be counting any postal votes on election night?
COMMENT: They count within district pre-polls on the night but they don’t start counting postals until the next day.
What is your view of the likelihood of this being the first federal election with a minority of polling day votes? Seems a distinct possibility.
COMMENT: It is likely as has happened at other elections over the last two years.
Hi Antony, what is the minimum number of preferences you must complete on both the HoR and Senate (above and below the line) ballots papers for the ballot to be “saved” and therefore counted as valid?
ANSWER: In the House ballot paper you must number all the squares with a correct sequence of preferences. The only savings provision is you don’t have to complete the last square but the best message is to number all the squares.
On the Senate ballot paper the instructions are to complete a minimum 6 preferences above the line or a minimum 12 below. Above the line any ballot paper with at least a first preference is formal. Below the line there must be a correct sequence of 1 to 6 preferences as a minimum for a formal below the line vote.
What do you mean by “last square”? The last (bottom) box, or for example only having to number any 1 to 5 on a 6 candidate ballot?
COMMENT: There is a savings provision that means on a ballot paper with ‘n’ candidates, only (n-1) preferences are required. There are two reasons why it is allowed. First, a ballot paper can never reach the final preference so there is no need for it. Second, if (n-1) preferences are shown, then it is clear the last square can be imputed as the final number.
Hi Antony. What is the logic (if there is one) for the need to correctly sequence 1 to 6 if your senate vote is below the line, but a single 1 is formal if above the line? If you inadvertently number 1 to (say) 24, but forget to number 6, why is that less valid than selecting 1 above the line for a minor party who may only have 2 candidates?
COMMENT: The original ATL proposal was 1 plus further preferences optional, as has been introduced for the NSW, SA and WA Legislative Councils. When Labor withdrew support for the original proposal, the Coalition had to negotiate with the cross bench who wanted the instructions to state a minimum number of preferences. As a single ‘1’ had been the method of ATL voting from 1984 to 2013, it was kept as a savings provision so that no vote that was formal before 2016 would become informal with the new rules.
The BTL proposal was more complex. The original proposal did not suggest a minimum number of preferences. Negotiations in 2016 had Labor suggesting it would support the changes as long as BTL voting still required full preferences. The idea lacked any logic and was opposed by the cross bench and somewhere 12 became the suggestion. It was the equivalent to a 6 preference ATL vote for 6 parties with two candidates. I suggested the savings provision which meant anybody using the ATL instructions below the line would be casting a formal vote.
Hi Paul. Your sequence with a missing 6 means that the voters intention is not clear after the number 5. The numbers 7 – 24 must be discarded. You now only have a sequence of 5.
As Antony says in another answer above, “Below the line there must be a correct sequence of 1 to 6 preferences as a minimum for a formal below the line vote.” These are the rules set by parliament.
Didn’t there used to be a savings provision in the old system that you could make one numbering mistake without the ballot being informal? Why couldn’t that still apply?
COMMENT: It could have applied except it was removed and replaced by the new savings provisions. The old savings provision was built around full preferential voting. Yes you could make an error and the vote would be saved, but you still had to fill all the squares. Well, 90% of them to be exact.
The new provision is effectively optional preferential voting beyond six preferences.
Thanks Robert but that wasn’t my point.
My point was that an ATL vote for just a single party (with 2 candidates) is valid, yet a BTL vote that only goes to 5 is not. It was not a question on the legality, but rather the logic. Antony gave us the answer – in terms of history even though it doesn’t (can’t) address the logic.
Hi Antony,
I noticed that my name was checked (marked off??) on computer at pre-poll on monday.
Is this a measure brought in to reduce voter fraud by allowing names to be marked off simultaneously at all booths?
ANSWER: No, they’ve just been rolling out electronic roll mark off in pre-poll centres. The volume of votes is so much larger than for polling places so roll mark off is a huge advantage when it comes to reconciling ballot papers issued to ballot papers in the ballot box. .
Hi Antony
Given that NSW has more urban seats (including provincial cities) than Vic and Qld, and this is where where the greatest numbers of postal votes applications emanate from (not rural seats), can you suggest reasons why such applications are so much higher in Qld and Vic as a percentage of enrolled voters statewide than in similarly east coast NSW?
COMMENT: I can’t offer an explanation other than postal voting is much higher at Queensland state elections than other states, and NSW has the lowest rate of postal voting at state elections. Some of those state differences come down to different state laws.
I notice in a pre poll how to vote for a seat with 5 candidates, one ticket is numbered has 1, 3,5,6,7. Would this make such a vote informal after the No1 was counted.
COMMENT: It would make the vote informal.
Is there official AEC guidance on this issue?
The requirement under the Electoral Act is that you express a preference for every candidate. I don’t see why using non-sequential numbers would fail this requirement. As long as their is a clear and unambiguous ranking of the candidates, I would think the vote would be counted.
COMMENT: Non-sequential numbers are informal. There is official AEC guidance on this. See Page 30 of the Scrutineer’s Handbook.
Hey Antony, do you know if the data is public anywhere that has prepoll numbers broken down by electorate?
ANSWER: The data is all available on the AEC’s download and statistics page.
What is the % of votes if you are number 1 on a rep paper? Therefore why not go for Robson rotation?
COMMENT: There is an advantage of maybe 1% through appearing at the top of the ballot paper. Most participants would rather put up with this than have random ballot papers that would nullify the use use of how-to-vote material.
Probably counteracted by the people who number 1 from the bottom and then work up the ballot paper (the reverse donkey)
I have vague recollection of Antony writing about the donkey vote, the reverse donkey and the circular donkey previously.
COMMENT: There’s lots of evidence for donkey voting, and for drift of preferences back to the top of the ballot paper and then down, but not much evidence for reverse donkeys.
When are the silent elector votes counted? Will they be with the pre-poll or with the postal?
COMMENT: I presume they will be counted as they are cast but can’t tell you for sure.
Silent electors cast a declaration vote via the Officer-in-Charge in a polling booth (they cannot be marked off the roll by an ordinary issuing officer and therefore cannot be given an ordinary vote that goes in the normal ballot box), so their votes are sealed in an envelope and go into the declaration ballot box, and are processed/counted at the same time as other declaration/provisional votes after election day.
Silent electors are managed as a declaration vote so will be counted along with the absent and other provisional votes.
I noticed today at Pre Poll that the AEC staff were using laptops to mark my name off the roll. Is this a new measure so that the AEC can pick up multi voting breaches?
COMMENT: No. It’s more about making it easier to reconcile names marked off the roll against ballot papers issued when counting starts. Each days markoff can be reconciled against ballot papers in the box used that day. It is also about minimising the number of incorrect pre-poll absent votes issued.
Hi Antony – Thanks for all your years of great work.
Something that has bothered me for years is this…
1. Why aren’t simple Senate [1] how-to-vote cards required to state on them what they actually translate to below the line?
2. Where is that information publicly displayed – is it in every polling station and how hard is it to find/read?
My desire has always been to see what order my favourite parties put the micro-parties in, so I can be guided by those how-to-vote cards but still decide for myself. In the upper house (Fed and State), this never happens. I find it disenfranchising and condescending.
COMMENT: A simple ‘1’ how-to-vote has preferences only for the candidates of that party and no other party. The lodged preference tickets that used to exist were abolished in the 2016 reforms. An above the line how-to-vote showing party preferences translates into the candidates for each party as they are printed below the line on the ballot paper. So the vote goes for the candidates of the first choice party in the order on the ballot paper, then the candidates of the second party, and so on.
On (2), it’s for candidates to distribute their how-to-votes. The Electoral Commission plays no role.
Thanks Antony ! Yes I was referring to the lodged preference tickets. I had forgotten they no longer exist. That’s much better. (So probably I was remembering only very old Senate how-to-votes, in the days when just a “1” was provided by the major parties, with the rest of the meaning implied by the lodged preference tickets. I feel better now.)
Why was the pre-voting period shorted from 3 to 2 weeks?
COMMENT: The Parliament agreed to amend the Electoral Act to shorten the period.
Is the actual candidate nomination forms lodged listed on the aec website I cannot find it
COMMENT: The nomination forms are not. The candidate declarations are.
With the greater number of early voters I think the early votes should be counted at around 5 o clock. Not many people vote after 5 o clock so the vast majority of votes will have been cast. It would give us a clearer picture earlier as to what way the election is headed. Mind you, you did call the last WA election correctly after 45 minutes.
COMMENT: There is no way they will allow counting of votes at a venue where voting is also taking place.
It doesn’t have to be at the polling places.
There are quite a few last minute voters between 5pm and 6 pm. Between 2pm and 4 pm is the really quiet time.
When will the senate prospects for SA and TAS get added in the senate preview
COMMENT: When I get a chance.
Thank you for informing us of the electoral process. The scrutineers handbook is entertaining reading. Vote last Donald Duck.
My question is can I vote 42 43 44 45 46 47 48 49 in the house of representatives.
After all 42 is the answer to life the universe and everything.
COMMENT: No. Your vote must have a valid first preference followed by a sequence of consecutive numbers up to the number of candidates on the ballot paper. If you don’t have a first preference, or miss a preference, or duplicate a preference, your vote is informal.
It still strikes me that young people in this country need to show ID to buy a pot of beer, but on Election Day, they can walk into as many polling stations as they like, and cast one vote after another, without ever being required to show ID, or be marked off a central voter roll for his electorate.
COMMENT: The rate of multiple voting is tiny, under 2,000 instance in 2019 and many of those were clerical error.
Hey Antony, do you know if the data is public anywhere that has prepoll numbers broken down by electorate?
Thank you
ANSWER: The data files are on the AEC downloads page.
The Postal data is by electorate, the Pre-Poll data by district, pre-poll centre and date of issue.
Why pick on “young people”. Old people can do those things as well.
It’s not the young that are committing electoral fraud it’s the middle aged and elderly.
All the cases in the US I’ve seen the miscreants are in the older age bracket and forging dead partners signatures on absentee ballots.
Hi all
On the diagram: “Postal Vote Application and Return Rates“ it would be good if the return rates were included from 2019 along the horizontal time horizon axis not just applications from 2019, as it’s not clear if the return rate is running on par (noting that returns can come after the election) or whether a lot of people have rewarded postals but then won’t use them. This data may not be available I suppose, but if it is it would be great to plot it in.
Thanks
COMMENT: I will switch the display to returned votes later this week. They were still low when I did the page based on Monday’s numbers.
Why don’t we have to show ID when we go to vote?
COMMENT: Apart from one Queensland election in 2015, I don’t think voters have required ID since early last century. There’s not a lot of evidence of multiple voting.
Hi Antony
Thanks for all your work over the years. The amount of work you put in is amazing.
Is there a way to determine whether the 2019 pre-poll percentages for TPP will be the same in 2022?
Thanks
Jordan
COMMENT: Not really. The best way is to check it afterwards.
Thanks very much for supplying the raw numbers underlying your Ordinary Prepoll graphs – I had been estimating numbers from low resolution graphics in AEC press releases. I can’t find daily prepoll figures on the AEC’s TallyRoom or in the Downloads for past elections. Are the figures part of the media feed FTP site?
The figures were published on a daily basis during the campiagn so you had to keep track of them on a daily basis.