Will larger ballot papers drive up informal voting?


This election will see a record number of candidates on the ballot paper for the House of Representatives, with 1203 candidates running, up from the previous record of 1188 in 2013.

In this post I’m analysing the relationship between the size of the ballot paper and the informal rate, in anticipation of voters having more difficulty filling out a formal ballot in 2022. Informal rates do become worse when voters are presented with more candidates, but that’s not the whole story.

Under the compulsory preferential voting (CPV) system used for the federal lower house, voters are required to fill out preferences for all candidates.

In practice this means you need to fill out consecutive numbers with no duplicates or missed numbers for every candidate (although you can skip the last candidate).

CPV, combined with compulsory voting, tends to produce a much higher informal rate than in other countries where casting a ballot is voluntary and easier.

The Australian Electoral Commission (AEC) used to publish reports after each election where they would survey informal ballots to categorise them based on why the vote was informal: was it blank, were the numbers incomplete, or were the numbers non-sequential. The last such report was published after the 2016 election.

Approximately half of the informal votes appear to be deliberate informal votes, including blank ballots and “scribbles, slogans and other protest vote marks”. Most of the rest appear to be accidental votes: incomplete numbering, non-sequential numbering or ticks and crosses. Most of these votes would be considered formal under the optional preferential voting (OPV) system used for New South Wales state elections.

The task of casting a formal vote gets steadily more difficult as the number of candidates increase, since you need to number more boxes without making a mistake. This is the theory underpinning my analysis: more candidates makes the task of voting more difficult, and thus results in more informal votes. Let’s find out if that’s true.

To start with, this first chart gives a sense of the number of candidates running in each contest over the last three decades.

You can see how many more electorates there are with double-digit nominations in 2022. 28 seats have 10+ candidates, compared to 6 in 2019, 26 in 2013, and just three in 2010.

So if higher candidate numbers increase the informal rate, that has implications for the informal rate at this year's election.

Next up, I've grabbed the number of candidates in each contest at the last four federal elections, and plotted that against the informal rate in each of those contests.

There's a clear trend of informal rates being higher with more candidates, but there are exceptions. The trend is most obvious at the lower end. While there are cases of informal rates around 2% where the number of candidates is 6 or less, you don't see those when there are ten or eleven candidates.

I've highlighted contests in New South Wales, and the stand out really clearly. While there are some high informal rates in contests with low candidate numbers, they are almost exclusively in New South Wales. Of 37 contests with an informal rate over 9%, 35 were in New South Wales.

I thought I'd simplify things here by just looking at the average informal rate for each state depending on the number of candidates running. There wasn't enough contests in the smaller jurisdictions, so the chart just features the four most populous states. I also limited the range of candidate numbers to exclude the rare 12-candidate contests and beyond.

The average informal rate in Victoria, Queensland and Western Australia steadily increases with more candidates, levelling out around 6-7% once you get to ten candidates.

That is not the case in New South Wales, where the trend line is pretty much flat. Contests with 5-7 candidates had similar rates of informality to those with ten candidates.

So at this point I return to the AEC's 2016 informal rate survey, which has quite a lot of detail about different types of informal voting.

The electorates with the highest informal rates, which are mostly in the suburbs of Sydney, tend to have higher proportions of informal votes that are assumed to be unintentional, compared to informal voting nationwide. These seats don't have particularly large numbers of candidates: Werriwa had just four candidates, yet ranked in the top ten, while five others had five or six candidates.

It appears that there are other factors going on in these Sydney seats. They have optional preferential voting at a state level, which can lead to confusion. They are also seats with a great deal of cultural diversity. Some voters may not understand English. Other voters may be perfectly capable of understanding English but just don't understand Australia's complex voting system. The 2016 report found a strong relationship between the ABS Index of Relative Socio-Economic Advantage and Disadvantage (SEIFA IRSED) and the informal rate amongst Sydney electorates.

So candidate numbers is not the only factor in the informal rate, but it appears to be a major factor outside of New South Wales.

So how are things looking in 2022?

We know that candidate numbers have hit a record high, with voters in 28 seats having to choose between ten or more candidates. This next chart shows the average number of candidates per seat for each state or territory at the last five elections.

The average number of candidates in New South Wales is one of the lowest in the country, even with Queensland and the Northern Territory on about 7.5 candidates per electorate. The ACT is lower, with just six candidates per seat.

South Australia and Tasmania have jumped from ranking quite low in recent elections to having well over 8 candidates per seat in 2022. Victoria had the most candidates per seat at the previous record level in 2013 (9.3 candidates per seat) and now have just 8.4. Western Australia took the lead in 2019, and has maintained its lead, with 9.4 candidates per seat this year.

This is possibly the worst outcome for the informal rate - the increase in candidate numbers is concentrated in states where it is a bigger factor in the informal rate, while in New South Wales the informal rate doesn't need increased candidate numbers to be relatively high.

Liked it? Take a second to support the Tally Room on Patreon!


  1. In my opinion the larger ballots won’t make a difference to the informal vote.

    Voters are already used to senate ballots which are 10 times bigger; I don’t think having 11 names instead of 8 on a lower house ballot would really make a difference.

  2. Following many of the political discussions on social media I am stunned at the poor level of political literacy out there. People do not understand our political system, the levels of government and their responsibilities. Of course it is clouded by several politicians who deliberately provide misleading information on the issue. The same goes for our preference voting system and techniques used by parties to harvest preferences. Many people are just frustrated by the complexity of it and just give up in frustration often voting informally. Clearly there is a case for better education on how the system works but I am not sure there is a political party who agrees.

  3. It just brings into question why we have Compulsory Preferential Voting when Optional Preferential would suffice for the lower house.
    If you now have to number 12 candidates instead of 4 or 5, or your vote doesn’t count that is kind of backward to democratic principles.

    Shrink the senate to 4 elected, 4 nominated by PM (like Canada and UK) and 4 by state. 2 up for election every 3 years- that will shrink the ballot

  4. @Trent, voters may be used to Senate ballots but they have never had to number 11 boxes consecutively for a formal vote for the Senate. A ‘1’ above the line has always been formal since 1984. Before that (and since then for BTL votes) the informal rate was much worse than in the House.

    @Greg, there’s some of that, for sure, but there’s also plenty of informal votes where people try to vote but fail to number enough boxes.

    @LJ, I’m talking about the House ballot here. The Senate ballot is shrinking anyway. 2-member elections are terrible, but smarter people have patiently explained that to you many times before. We have CPV because people in power don’t want to see preferences exhaust, it’s not hard to understand. I do support OPV with some of the measures we have in the Senate to softly encourage preferencing.

  5. @ Ben I think you’re the only one who has advocated why 2 member elections are terrible but that’s all good. I think you may confused both patience and acumen with ridicule and subjectivity. All systems have their positives and negatives, maybe it might make for more interesting podcasts if you had opinions on different to yours, particularly those coming from the academic sphere.

    If what you say is correct and CPV is around because in people in power don’t want to see preferences exhaust then why does OPV exist across many states including the largest in NSW?
    I also think voter exhaustion as it currently exists would only impact one side of politics more than the other being Labor because of the size of the Greens. It may be a problem for the Libs later down the track but parties like One Nation are not big enough to take sizeable portions of votes away in the way the Greens do of Labor.

  6. LJ Davidson, in the current environment I think it is more Labor who don’t want to lose compulsory preferences as they lose the bulk of Greens votes that flow to them. Recent changes from OPV back to compulsory preferences occurred under Labor governments (I believe this was the case in Queensland prior to 2017 and also NT before 2016)

  7. At this stage, only NSW uses OPV as Queensland and NT made the changes I referred to in the earlier post.

  8. @Yoh An that’s exactly what I was saying. Labor does not want to lose CPV because they would more than likely lose multiple seats both Federally and at a state level.
    There’s no benefit for Libs to retain it. Especially with the rise of independents. They can essentially run a FPTP campaign strategy while opposition parties attempt a disorganised and uncoordinated preference strategy. Much like what happened at the Willoughby by-election

  9. A few comments on this as I have looked at it in depth in material that was not published at the time because it was commissioned. I have found that because of the amount of noise between naturally high and low informality electorates (and the suspicion that some seats attract high candidate numbers that would otherwise have low informal votes, and vice versa), the most useful way to study the impact of candidate numbers is by looking at changes in the informal rate in seats where the number of candidates has changed from election to election.

    Since 2016 there has been a significant change caused by Senate reform which I refer to as the eight-candidate problem. If the number of candidates exceeds seven and a voter votes in the Reps as if voting for the Senate ATL, by voting 1-6 and stopping, then their vote becomes informal. (If there are seven candidates and they vote 1-6 leaving one box blank that is saved and formal.)

    On average if the number of candidates in a seat goes up there is an increase in informality and if it goes down ditto. But there is a noticeable “step” at eight candidates where, for instance, going up from 7 to 8 has a much bigger impact than going from 6 to 7 or 8 to 9. (And likewise for dropping below 8).

  10. I believe that CPV may, in many cases, lead to the perversion of a voter’s true intention.
    Take the common case when multiple “independents” need to be ticked off in order of preference, yet when those “independents” are blatantly part of a political movement, CPV makes a mockery of free choice.
    I firmly support OPV as being a system that is less open to abuse of “harvesting preferences”.
    The present chaotic state of our senate frustrating the house of representatives’ mandate is, in my mind, a clear indication that CPV is corrupting an otherwise workable system.
    The senate and its State proportion of senators were devised to promote and protect the interests of the State, it has ceased to be a State function and has become a purely Federal political movement.


Please enter your comment!
Please enter your name here