Proportional representation electoral systems are used in Australia to elect candidates to the Senate, the upper houses of NSW, Victoria, South Australia, and Western Australia, the Lower House of Tasmania, the ACT Legislative Assembly and many Local Government Councils.
Proportional Representation (PR) is the term which describes a group of electoral systems used to elect candidates in multi-member electorates. Under PR, parties, groups and independent candidates are elected to the Parliament in proportion to the number of votes they receive.
The composition of a Legislature where members are elected using PR usually better reflects the proportions of votes received by candidates on a State or Territory-wide basis than houses where members are elected to single seat electorates.
There are three main types of PR electoral systems:
- list systems;
- mixed-member proportional systems; and
- single transferable vote (STV) systems.
All Australian PR electoral systems are STV systems.
Under STV electoral systems, each vote can be transferred between candidates in the order of the voter's preferences.
How is a candidate elected?
A candidate is elected when his or her total number of votes equals or exceeds the quota. In some circumstances, a candidate can be elected with less than a quota (see How votes are counted to elect candidates).
The quota is the number of votes a candidate needs to be certain of election. The quota is calculated using the formula:
(total number of formal votes / (number of candidates to be elected + 1)) + 1 (disregarding any remainder or fraction)
For example, if there were a total of 10 000 formal votes and 4 candidates to be elected, the quota would be:
(10 000 / (4 + 1)) + 1 = 2001
The quota used for all STV systems in Australia is called the 'Droop' formula first published in 1868 by mathematician and lawyer, Henry R Droop.
Ballot papers used in Australian STV elections vary depending on the legislative requirements of the particular system. Variations include:
- Directions on how to cast a valid vote;
- How candidates' names are listed (usually in party, group or ungrouped columns);
- How the order of columns is decided (usually drawn by lot, with ungrouped candidates listed in the right-hand columns);
- Whether party or group ticket voting squares appear on the ballot paper;
- How candidate names are ordered in a column (e.g. alphabetical, drawn by lot, Robson Rotation or party list); and
- Whether group ticket voting square appear above the line or beside the line.
In elections for the Senate and the Upper Houses of New South Wales, Victoria, Western Australia and South Australia, in addition to voting for individual candidates the elector has the option to vote for a particular party or group ticket. A separate section at the top or at the side of the ballot is paper is provided for the elector to mark a single preference for a party or group. This is known as "above the line", "beside the line" or "ticket" voting.
Each party or group can register a specific order of preferences to apply to ballot papers. This order of preferences is known as a "ticket". Except in Western Australia, these "tickets" are required to be displayed at each polling place for electors to inspect. By marking a ticket square, the voter's preferences are taken to be the same as the ticket registered by a party, group or independents.
For elections for the Senate and the Upper houses in NSW, Victoria and SA, those parties, candidates or non-party groups entitled to register tickets have the option of registering more than one ticket – up to 3 tickets for the Senate, NSW and Victoria, and up to 2 tickets in SA. Where 2 or 3 tickets are lodged, 1/2 or 1/3 of the votes for that ticket (as the case may be) are taken to follow each ticket.
Robson Rotation is a process of rotating candidate names within a column so that favoured (top and bottom) positions are shared equally between all candidates. Neil Robson, a member of the Tasmanian House of Assembly, introduced the process to the Tasmanian Parliament in 1977. The ACT Legislative Assembly adopted Robson Rotation in 1995.
The rules for casting a formal vote vary between Australian jurisdictions and are set out in the individual jurisdiction pages below.
Where "above the line", "beside the line" or "ticket" voting is available, voters have the choice of voting for tickets or for individual candidates. If ticket voting is not available, voters can only vote for individual candidates.
A formal vote for a ticket usually consists of a single first preference for one ticket.
There are 3 basic types of formal voting for individual candidates:
- full preferential, where the voter is required to show a preference for every candidate;
- optional preferential, where the voter is only required to show a preference for 1 candidate, but has the option of showing further preferences; and
- minimum-length preferential, where the voter is required to show a preference for at least the number of candidates to be elected, but has the option of showing further preferences.
In general, a ballot paper can be classed as informal if:
- the voter's intention is not clear;
- it contains unacceptable mistakes (see Ballot Papers);
- the returning officer is not satisfied that the ballot paper is authentic (that is, the ballot paper may be a forgery or incorrectly issued); or
- the voter has placed his or her name on the ballot paper (violating the secrecy of the ballot).
The first step is to identify all formal ballot papers and distribute them to candidates according to each ballot paper's first preference. All informal ballot papers are set aside. The quota is calculated from the total formal vote.
If any candidate receives exactly a quota of votes, he or she is elected and his or her ballot papers are set aside.
If any candidate receives more votes than the quota, he or she is elected, and the excess (or surplus) votes are passed on to continuing candidates according to voters' preferences. (See Distributing a surplus for more detail on this process.) Following the distribution of each surplus, any candidate who has reached the quota is elected and any resulting surplus again passed on.
If more than one candidate is elected at the same stage (or 'count') in the scrutiny, each surplus is distributed as a separate count. The candidate with the largest surplus is dealt with first, the candidate with the second largest surplus is dealt with second, and so on.
Once all surplus votes have been distributed, the candidate with the fewest votes is excluded and all of his or her votes passed on to continuing candidates according to the voters' preferences. (See Exclusion of a candidate for more detail). Further candidates are excluded in the same manner until another candidate reaches the quota.
The process of distributing surplus votes and excluding the candidate with the fewest votes continues until the required number of candidates is elected. In some cases the final candidate(s) may be elected without reaching the quota where all other candidates have been either elected or excluded.
One of the main differences between Australian STV systems is the way in which surplus votes are distributed. The method of distributing a surplus depends on two considerations:
- which ballot papers are redistributed or used to transfer the surplus; and
- how the new transfer value for these ballot papers is calculated.
The Senate, Victorian, Western Australian and South Australian systems use all of the ballot papers the elected candidate has received to transfer the surplus. In Tasmania and the ACT only the ballot papers the elected candidate received in the count at which he or she was elected are used to transfer the surplus (sometimes called the 'last parcel' method). The New South Wales system randomly selects a proportion of the ballot papers to transfer the surplus.
In most jurisdictions, ballot papers distributed in a surplus are assigned a 'fractional transfer value'. While each ballot paper is worth a single vote, under STV a fraction of that vote can be used to elect one candidate, and the remaining fraction of the vote can be distributed to other candidates as part of the elected candidate's surplus.
For example, if 1000 ballot papers are used to transfer a surplus of 500 votes, each ballot paper is given the new fractional transfer value of 0.5 of one vote.
When a candidate is excluded, all ballot papers received by the excluded candidate are transferred to continuing candidates at the transfer value at which they were received (except NSW Upper House). Ballot papers received by an excluded candidate as first preference votes have a transfer value of 1 vote. Ballot papers received by an excluded candidate from an elected candidate's surplus will usually have a fractional transfer value.
In most systems, candidates are excluded one at a time. Only the Senate allows for more than one candidate to be excluded at one time. This is called "bulk exclusion". Under bulk exclusion more than one candidate can be excluded at the same time if it can be proven that none of the votes of these candidates could possibly affect the order of election of the next candidate to be elected.
Where the contest for the last seat in an election is close, it is common for the final two continuing candidates to both have less than a quota. In this case, the continuing candidate with the highest number of votes is elected. Most systems elect the highest candidate without distributing the votes of the losing candidate. In the ACT, the votes of the last excluded candidate are distributed, and as a result the last elected candidate usually achieves a quota as well.
The final remaining continuing candidate(s) in a scrutiny can also be elected without a quota where significant numbers of votes become 'exhausted' during the scrutiny. A vote is exhausted if it does not have a preference marked next to any of the continuing candidates. (This cannot occur in full preferential systems that do not allow voters to make mistakes, as exhausted votes are not possible.)
Under multi-member PR systems, it is generally considered unfair to fill casual vacancies by holding by-elections, because the vacating member was elected to represent a proportion of the electorate, not a majority of the electorate.
Holding a by-election for one vacancy, where the majority would choose the vacating member's replacement, could result in an unfair shift in the political balance in the Parliament.
In Australia, casual vacancies in Parliaments elected under STV are usually filled either by nomination or by recount.
Senate vacancies are filled by a joint sitting of the State or Territory Parliament nominating a new Senator who must be of the same political affiliation as the vacating Senator at the time of his or her election. In the NSW, Victorian and South Australian upper houses a casual vacancy is filled by a nomination approved by both Houses of Parliament. In NSW, a party member is generally replaced by a member of the same party. In Victoria and South Australia, a party member must be replaced by a member of the same party.
In Western Australia, Tasmania and the ACT a casual vacancy is generally filled by a recount process using votes cast at the previous general election. Only unsuccessful candidates at the general election are eligible to contest the recount. The ballot papers which were used to elect the vacating members are the only ones distributed during the recount. In the event that it is not possible to fill the vacancy by recount (for example if no unsuccessful candidates wish to contest the vacancy), casual vacancies can be filled by appointment except in Western Australia where an election would be held and in Tasmania where if there are no available candidates who belong to the same registered party as the vacating member the parliamentary leader of that party may request that a by-election be held.