Is Australia’s preferential voting system a mystery to you – a tangled web of wheeling and dealing and scary mathematics too complex to comprehend?
First of all, remember political parties don’t decide the order in which you number the candidates on your ballot paper. You do. The outcomes of party preference deals are plain to see by the way they suggest you should number your preferences. And the maths really is quite simple. The only scary part is that unless you understand how preferences work you risk wasting much of the power of your vote.
In Senate elections more than 95% of people vote “above the line” which means their preferences are distributed automatically according to their chosen candidate’s wishes. It is reasonable to assume the same applies to lower house elections. Others take the greatest care to vote 1 for the candidate they want to win then give very little thought to the order in which they number the other candidates. Either way, many people don’t realise the importance of their second, third and fourth preferences. Yet more often than not it is preferences that make the difference between who wins and who loses.
That’s why political parties are so keen to get you to number your preferences in the order they think is to their best advantage. They show this on their “How to vote” cards handed out at the polling booths. But the decision is up to you.
So how does it work?
The easiest way to see how it works is to start by considering the simplest system of all – first past the post – where the candidate with the most votes wins.
This seems not only simple but fair, until we ask: What if a candidate wins the most votes yet still has less than half of the vote?
In the example above more people voted for Andy than for any other candidate but more than half didn’t vote for him. Is Andy really the best candidate when 60% of voters didn’t want him?
That’s why in some first past the post systems, if no candidate wins more than 50% of the vote, a second or run-off election is held between the two top contenders.
In the example above there would be a run-off election between Andy and Jill. The 350 people who voted for Lee and Joe in the first election would then have to choose between Andy and Jill in the second. Many voters would still not get the candidate they wanted but more than 50% would get the most preferred of the two.
How preferences work - House of Representatives
An election using the Australian preferential system is just like holding several run-off elections until one candidate is the clear winner, preferred by the majority of the voters. But it’s all done in one election on one ballot paper by simply listing our preferences 1, 2, 3 and so on.
Australia has 150 House of Representatives (Lower House) electorates, each represented by one Member of Parliament (MP). Each electorate has about 80,000 to 90,000 voters and as many candidates as care to stand. However for the sake of clarity we’ll use the same 4 candidates and the same 1000 voters to see how the election above might have turned out using preferential voting.
First count: After the first count Andy leads and Joe has the fewest votes so he is excluded. Joe is out of the race but watch how the secret power of his preferences lives on.
Second count: Joe’s 150 votes are now distributed according to their second preference. 100 of the people who voted 1 for Joe voted 2 for Lee (their second preference) and this has pushed Lee into second place with 300 votes.
Andy has received 20 of Joe’s second preferences bringing his total to 420 and Jill has received 30 bringing her total to 280.
No candidate yet has a majority (at least 50% + 1) so Jill, with the fewest votes, will be the next to be excluded.
Third count: Jill is out of the election … or is she? Watch how her preferences now decides who wins.
240 of the people who voted 1 for Jill voted 2 for Lee (their second preference) and these preferences have pushed Lee over the line with 54% of the vote. She is the winner even though she started with just 20% of the primary vote! And she owes her win to the secret power of Joe’s and Jill’s preferences.
Note that Jill’s 250 primary votes were distributed according to their second preference. The 30 votes she aquired as Joe’s second preference, were distributed according to their third preference.
None of Jill’s second preferences were directed to Joe but if any had been they too would have been distributed according to their third preference. That’s because Joe was already excluded and his second preferences had already been distributed.
The aim of the preferential voting system is to ensure that the candidate most preferred by the most voters does, in fact, get elected.
Whether or not you agree that it achieves that result, hopefully you will now understand the secret power of preferences and how vitally important it is to choose them very carefully.
How preferences work - The Senate
Some people will tell you that if you think the House of Representatives preference system is complicated don’t even try to understand how our Senators are elected.
It is true that understanding why it works requires some mental gymnastics but understanding how it works requires only simple mathematics.
That’s a good thing because preferences play an even more important role in deciding Senate elections and our Senators play an extremely important role in how Australia is governed.
In the Senate (also called the Upper House) each of the 6 states is represented by 12 Senators and the Australian Capital Territory (ACT) and Northern Territory (NT) are represented by 2 each – a total of 76 Senators.
Their job is to ensure the interests of the states and territories are represented equally and to review the proposals and decisions of the House of Representatives and the Government.
The state Senators are elected for 6 years (2 parliaments) and the territory Senators are elected for 3 years. The state Senators’ terms overlap so that every 3 years 6 from each state are due for election. This is called a normal or half-Senate election and coincides with the House of Representatives election. If the Senate rejects a piece of legislation twice the Government can call a double dissolution in which case an election is held for all 76 Senate seats.
So how does it work?
Because there are 2, 6 or 12 Senators to be elected to represent one state or territory (as explained above) a quota system is used whereby the first candidates to achieve the quota of votes are elected.
Those with the fewest votes are progressively excluded and their votes are distributed to the other candidates according to voter preferences. The aim is to ensure parties are represented in the Senate in a similar proportion to their proportion of the vote.
Each Senate electorate covers a whole state or territory and has from several hundred thousand to several million voters. Often there are 50 or more candidates vying for the vacant Senate seats.
In the example below we’ll assume it is a half-Senate election for 6 seats in one of the states. For the sake of clarity we’ll assume there are only 10 candidates and 1000 voters.
First we need to determine the quota. It is calculated as follows: Formal (or valid) votes cast divided by [vacancies + one] + 1
So the quota in our 6 seat example is (1000/7) + 1 = 143.857. The fraction is disregarded giving us a quota of 143.
First count: Peter and Judy each have more votes than the quota and are elected at the first count. They each have a surplus of votes that will be distributed according to their preferences.
But how can preferences be distributed without knowing which votes are part of the quota and which are part of the surplus?
Fortunately the same result is achieved by distributing all of a candidate’s votes but at a reduced value. This is called the transfer value and is calculated by dividing a candidate’s surplus votes by her or his total votes.
Second count: Peter’s 157 surplus votes divided by his 300 total votes gives a transfer value of 0.523. (157/300 = 0.523) All 300 of his votes are distributed according to their second preferences but at the reduced transfer value of 0.523. (The transfer value is actually calculated to 8 decimal places.)
Peter’s ballot papers had no second preferences for Judy but if any had their third preference would have been used because Judy is already elected.
In this simple example there is no need to track the type of preferences (2nd, 3rd, 4th etc) involved in every transfer of votes, but you’ll appreciate that in a real election by the 100th count, for example, each transfer would contain many types of preferences that will influence the outcome. (That’s the secret power of preferences again.)
Notice that the total number of votes distributed at the reduced transfer value (157) is the same as the total of Peter’s surplus votes at their full value (157).
Peter’s preferences have pushed Keith over the quota. Keith is elected and now has a surplus of 88 votes. But Judy reached the quota first so her preferences are next to be distributed.
Third count: Judy’s transfer value is calculated (58/201 = 0.289) All 201 of her votes are distributed according to their preferences but at the reduced transfer value of 0.289. Most go to Alice (50 votes) giving her the quota plus a surplus of 9 votes. Alice is elected. But Keith reached the quota first so his preferences are next to be distributed.
Fourth count: Keith’s transfer value is calculated (88/231 = 0.381) All 231 of his votes (his 91 primary votes plus the 140 aquired from Peter) are distributed according to their preferences but at the reduced transfer value of 0.381. Although most go to Mahmoud (53 votes) the 33 votes that go to Linda bring her up to the quota. Linda is elected. But Alice reached the quota first so her preferences are next to be distributed.
Fifth count: Alice’s transfer value is calculated (9/152 = 0.059) All 152 of her votes (her 102 primary votes plus the 50 aquired from Judy) are distributed according to their preferences but at the reduced transfer value of 0.059. They all go to Mahmoud but still leave him short of the quota.
Sixth count: Now no candidate has a surplus, so Steve, the candidate with the fewest votes, is excluded and his preferences are distributed. They all go to Mahmoud but he is still 2 votes short of the quota.
Note that Steve doesn’t have a surplus so there is no need to apply a transfer value. Preferences from excluded candidates are distributed at the value at which they were received. Therefore preferences from Steve’s 15 primary votes each have a value of 1 vote and the 4 preferences he aquired from Peter also each have a value of 1 vote.
Seventh count: Katrina is now the candidate with the fewest votes, so she is excluded and her preferences are distributed. Like Steve, she doesn’t have a surplus so there is no need to apply a transfer value and her preferences are distributed at the value at which they were received. Luckily for Mahmoud they all flow to him. Mahmoud is elected.
All 6 Senate positions are filled. In this simplified example it took just 7 counts. In reality, while the first 4 seats are filled quickly by the big parties, the last two seats can take several hundred counts (and several weeks) to be decided.
Notice that Mahmoud and Robena started with an almost identical number of primary votes – 60 and 61 respectively. Yet after the distribution of preferences Mahmoud had more than twice as many. He was elected and Robena missed out. It all came down to preferences.
Do small parties have a chance?
Did you notice that Mahmoud was elected with only 6% of the primary vote and just 14.3% with preferences? Compare this with the House of Representatives where he would have needed more than 50% to be elected.
That’s why small parties and independents have a far better chance of winning a seat in the Senate than in the House of Representatives – even moreso in a double dissolution election when the quota works out to be just 7.7%
Of course this makes it easier for the big parties, too, and they quickly snap up most of the seats. But small parties would always be kept out in the cold except for the secret power of preferences.
No related links for this post.
