21 April 2022
PDF version [222 KB]
Politics and Public Administration
When a general election or by-election
is imminent the subject of trading preferences, or who is directing their
preferences to whom, is always a hot topic of media discussion. In practice,
however, the process of trading preferences is less concrete than it might
What is a preference?
All lower and upper house elections at the federal level and
in each state and territory either require or encourage voters to allocate
preferences to candidates on the ballot paper.
At the federal level in the House of Representatives
preferential voting means that if a candidate gets over 50 per cent of the vote
in first preferences they are declared elected. If no candidate gets above 50
per cent of the votes on first preferences, the candidate with the lowest number
of first preference votes is excluded and the preferences on those ballot
papers are distributed to remaining candidates. When one candidate has at least
50 per cent of the vote from their own first preference votes and preferences transferred
from excluded candidates, that candidate is declared elected.
One consequence of this process is that the final two
candidates in each contest (generally referred to in Australian elections as
the ‘two candidate preferred’ or ‘TCP’ candidates) will not have their
preferences distributed. Under full preferential voting, as in a federal House
of Representatives election, all the ballot papers that indicate a first
preference for a candidate other than the final two will be counted towards one
or the other of these two TCP candidates.
There is no way to cast a formal ballot so that it does not
count towards one or the other TCP candidate, as all candidates on the ballot
paper must be numbered so each ballot must preference one of the two TCP
candidates over the other. If the TCP candidates are Labor and Liberal, for
example, then all formal ballots will count towards either the Labor or Liberal
candidate’s total, no matter who the voter preferenced or in what order.
This also means that it does not make sense to say that ‘candidate
A beat candidate B on the preference of candidate C’. Candidate C did not
allocate the preferences: the voters did. And the votes for all the candidates
must have counted against either candidate A or candidate B (the TCP
candidates) due to the nature of compulsory full preferential voting.
In the majority of contests in a federal election (136
divisions in the 2019 federal election) one of the TCP candidates is from
Labor and one is from the Liberal/National Coalition. In these divisions, which
are referred to as ‘classic’ divisions, the preferences from the Labor and
Coalition candidates are not distributed, and so are not counted by the Australian
Electoral Commission (AEC). So we tend not to know how the preferences of most
of the people who vote for Labor or the Coalition with their first preference are
The distribution of preferences for Senate elections is more
complex and the preferences for a winning candidate may be distributed. In
Senate elections voters can also allocate preferences above or below the line.
A detailed discussion of how the count in Senate elections works can be found
in a prior
How do candidates direct
Technically candidates do not allocate preferences: voters
do. The only way a candidate can direct preferences is to communicate to their
supporters how they would like them to order the candidates on their ballot
papers. The way that this is typically done in Australian elections is through
producing a How To Vote (HTV) card that indicates the candidate’s preferred
preference distribution, and getting those HTV cards into the hands of their
Many candidates have an interest as to which candidate
‘their’ votes (that is, a ballot paper that gives them a first preference) ends
up counting towards. For example, a candidate who does not have sufficient
support to be one of the final two candidates in an election might want to
support another candidate in the election who is ideologically aligned, and
have their first preference votes count against that candidate.
In some cases a candidate’s allocation of preferences might
be purely for show or signalling, such as a candidate who is likely to be one
of the two final candidates stating that they are preferencing another
candidate last, despite the fact that their preferences will likely not be
The success of a HTV card in directing preferences depends
on two factors: whether the supporters receive the HTV card, and whether they
Distributing HTV cards
The easiest way for a candidate to ensure that their voters
get one of their HTV cards is by handing them one as they walk into the polling
place. Some states require parties to register HTV cards for state elections
with the state electoral commission, and the electoral commission hosts those
HTV cards on a website, creating a central repository for voters looking to
follow a candidate’s HTV card (this is the case in New
South Wales, Queensland
Australia, for example). South Australia even goes so far as printing the
HTV cards on the voting screens in the polling place. However, none of these
options are available for federal elections.
Distributing HTV cards at polling places in an election is
usually done by party members and volunteers. Smaller parties may have trouble
mobilising sufficient volunteers to cover all polling places, particularly in a
general election. Media reports stated that in 2013 a Palmer United Party
candidate was paying
backpackers to hand out HTV cards, and in 2016 the Australian Equality
Party was reported to have used job
outsourcing website Airtasker to hire people to hand out HTV cards. However,
at around $100 per person, adequately covering around 7,000 polling places at a
federal election (including the pre-poll centres full time for two weeks) would
As a rule, then, smaller and newer parties that do not have
a large grass-roots membership base, or the financial resources to hire people,
will struggle to direct their preferences unless they restrict themselves to a
manageable number of seats. Statements by these parties in relation to who they
are preferencing in an election may be more aspirational than likely to affect
actual preference flows.
Following HTV cards
While a HTV card can indicate to a voter how their preferred
candidate would like that voter to distribute their preferences, preference
allocation is still up to the voter.
Due to the anonymity of voting it is not possible to determine whether people
who take HTV cards actually follow them, however some Australian research has
examined voter use of HTV cards.
The Australian Election Study (AES) asks respondents whether
they followed a HTV card when they marked their preferences. As shown in Figure
1, the 2016 and 2019 results show that HTV card use varies considerably from
party to party, with almost half of all Liberal voters following HTV cards, but
less than 5 per cent of One Nation voters. Note that for some of the
smaller parties the sample size is quite small, so these figures should not be
assumed to be representative of all voters for those parties.
Figure 1 Voters who used a HTV card
in 2016 and 2019, by party voted for in the House of Representatives
Source: Data from the Australian Election
Note: Data for some smaller parties is based on a very small
number of respondents and is not necessarily representative.
Looking at the actual votes and their preferences is another
way to measure HTV card compliance, however it is complicated by the inability
to determine between those who followed a HTV card and those who just happened
to allocate their preferences in the same order as a HTV card.
As noted above, Labor and Coalition preferences are typically
not distributed in federal elections, so it is more difficult to determine
preference compliance for the major parties. ABC election analyst Antony Green
looked at a number of seats that were TCP
contests between the Greens and Labor, meaning Liberal preferences were
distributed. In 2010 the Liberals preferenced the Greens above Labor, and in
the divisions of Batman (since renamed to Cooper) and Melbourne around 80 per
cent of Liberal preferences went to Greens. In 2013 the Liberals preferenced Labor
over the Greens, and 32.6 per cent of Liberal preferences went to the Greens in
Batman and 33.7 per cent went to the Greens in Melbourne. In 2019 Liberal
33.9 per cent to the Greens in Cooper (formerly known as Batman) on the basis
of Liberal HTV cards also putting Labor over the Greens (the Liberals were a
TCP candidate in Melbourne in 2019 and so their preferences were not
distributed). This suggests a relatively high level of Liberal voter HTV card compliance
in those seats.
Some state electoral commissions conduct audits of HTV card
compliance in state elections. In 2014
in South Australia, for example, an audit of 76 divisions found 42.4 per
cent of formal ballot papers followed a HTV card. The audit found that
divisions that had 5 or more candidates showed below average HTV card compliance,
and those with 4 candidates showed above average HTV card compliance. For the
2018 state election the South
Australian Electoral Commission found 37.7 per cent of all formal
ballot papers followed the HTV cards. As mentioned above, polling places in
South Australia have the HTV cards printed on the voting screen, making them
accessible to voters at the time of voting, so it would be expected that SA
would have relatively high HTV card compliance.
Following the 2019 NSW state election a survey
was conducted which asked about HTV card usage, finding that 63 per cent of
voters ignored HTV cards completely. Of the reported usage by party, 53 per
cent of Coalition voters ignored HTV cards, as did 80 per cent of Greens
voters. These figures are largely consistent with the AES findings. The report
also noted that HTV card usage was highest among ‘rusted on voters’ (83 per
cent), however it is unclear how this was defined.
The Parliamentary Library looked at distribution
of preferences in the 2016 Senate election and found highly variable HTV card
compliance. In NSW, for example, almost a third of Coalition voters preferenced
in accordance with the party’s HTV card to the third preference, whereas only
1.4 per cent of Labor voters did. In Queensland, where One Nation
polled strongly in the Senate, only 10.36 per cent of One Nation
voters gave their second preference to the Australian Liberty Alliance, which
was second on One Nation’s Senate HTV card. Almost as many One Nation voters
(9.95 per cent) gave their second preference to the Liberal National
Party of Queensland instead.
It is worth noting that 2016 was the first time in recent
memory that HTV cards have been used for the Senate, and it is possible voters
may take time to get used to using them. It is also true that people are
reporting less usage of HTV cards over time, according to the AES. In 1996 more
than half of all respondents reported using a HTV card for their House of
Representative vote, however by 2016 that had dropped to only a third of
respondents (Figure 2).
Figure 2 Followed ‘How to Vote’
card for House of Representatives
Source: Australian Election Study
HTV cards and formality
While HTV cards are useful for directing how a candidate’s
supporters should vote, they also have a role in supporting voters to cast a
formal vote—that is, a vote which is filled in correctly and can be counted in the
analyst Kevin Bonham argues that formality is the main purpose of HTV cards).
This is particularly true for contests in seats where many electors do not speak
English as their first language, or in contests where there are large numbers
of candidates and voters risk making numbering errors.
by the AEC into informal voting at the 2013 federal election found that
areas with high levels of social exclusion tended to have higher rates of
informal voting and suggested that
factors such as educational attainment and English
proficiency could mean that many voters submitted blank ballots due to
linguistic problems or a lack of understanding of the Australian electoral
The same research found that when the number of candidates
on the ballot paper increased, the informality rate from non-sequential or
incomplete numbering increased. If the number of candidates changed between
elections, this was also associated with increased informality.
In 2013 Labor MP for the ACT electorate of Fraser (since
renamed to Fenner) Dr Andrew Leigh issued a HTV card which indicated a first
preference for himself, with the remaining preferences in order down the ballot
paper, resulting in an order of preferencing which favoured right-wing parties
such as Rise Up Australia over the Greens. Dr Leigh argued that as Fraser was a
safe Labor seat his preferences were unlikely to be distributed (which turned
out to be
correct), so he opted for the simplest numbering approach. He
stated that the ‘priority for us was formality, to minimise informality
rather than to worry about the extremely small chance that Labor preferences
would be distributed’.
Providing a HTV card is an easy way to ensure that, as a
candidate, your supporters’ votes will be formal and entered into the count, as
even a relatively numerate elector might make an error allocating as many as 22
How important are preferences?
As voters are becoming less likely to use HTV cards, the
requirement to count preferences in federal elections is generally increasing.
Preferences are only distributed when none of the candidates has received 50
per cent or more of the first preference vote, and the proportion of seats in
federal elections where the winning candidate has won with less than 50 per
cent of the first preference votes has been increasing over time (Figure 3).
Figure 3 Seats in federal
elections where preferences were counted to determine the winner
Source: Parliamentary Library analysis of AEC election results
Candidates who are ahead on first preferences still
generally win the election after preferences are distributed, however.
According to Library calculations, in only around 1 in 10 divisions in recent
federal elections did preferences result in the candidate leading on first
preferences losing the election. Analysis
by Antony Green on House of Representatives elections since 1949 found that
generally less than 10 per cent of seats had the results changed by
preferences. So while preferences are increasingly being counted, they do not
generally make large differences to election results.
Preferences may also be important to candidates and parties
to the extent to which they can be ‘traded’. A minor party may do a deal with a
major party that each will preference the other prominently on their HTV cards
(or a major party might preference a minor party and hope that the minor party
will reciprocate, without making a formal deal). While the major party’s
preferences may never be distributed, and the minor party’s success in
directing preferences might be negligible, featuring as a major party’s highly
ranked preference provides some free publicity that the minor party might not
otherwise be able to achieve. They may also help boost the party’s vote,
allowing the party to receive more per-vote
public funding. Such deals may turn out to be more effective for
ideological signalling than they are for strategic electoral advantage,
although this does not stop them from being aggressively pursued when an
Regardless of the effectiveness of HTV cards for directing
preferences (as opposed to encouraging formal voting), their value in terms of
both signalling and the small number of additional directed preferences they
may secure means that preferencing debate is unlikely to diminish in the near
future. In an election every vote counts, and the actual or perceived benefit
from securing deals and distributing HTV cards likely well outweighs the
financial costs of producing them.
This was not the case for the Senate system as it operated between 1984 and
2014 as voters who voted above the line had their preferences distributed in
accordance with a group voting ticket provided by the party to the AEC. Voters
could only direct their own preferences if they elected to number every box
below the line.
For copyright reasons some linked items are only available to members of Parliament.
© Commonwealth of Australia
With the exception of the Commonwealth Coat
of Arms, and to the extent that copyright subsists in a third party, this
publication, its logo and front page design are licensed under a Creative Commons
Attribution-NonCommercial-NoDerivs 3.0 Australia licence.
In essence, you are free to copy and communicate this work in its current form for all non-commercial purposes, as long as you attribute the work to the author and abide by the other licence terms. The work cannot be adapted or modified in any way. Content from this publication should be attributed in the following way: Author(s), Title of publication, Series Name and No, Publisher, Date.
To the extent that copyright subsists in third party quotes it remains with the original owner and permission may be required to reuse the material.
Inquiries regarding the licence and any use of the publication are welcome to email@example.com.
This work has been prepared to support the work of the Australian Parliament using information available at the time of production. The views expressed do not reflect an official position of the Parliamentary Library, nor do they constitute professional legal opinion.
Any concerns or complaints should be directed to the Parliamentary Librarian. Parliamentary Library staff are available to discuss the contents of publications with Senators and Members and their staff. To access this service, clients may contact the author or the Library‘s Central Entry Point for referral.