Let’s be clear from the outset: Australia will probably never again adopt first-past-the-post voting, whether at a federal or state level. Apart from the lack of appetite for it, other countries have looked to Australia’s system as something worth emulating.
But hypotheticals are fun! Having just moved to Canada, where federal politicians will face voters under a first-past-the-post system in about two months’ time, I was interested to see how Australia’s current parliament would look if we had used a similar method back in May.
This is similar to an article I wrote in 2016, shortly after that year’s federal election, looking at the results if first-past-the-post had been implemented.
What is first-past-the-post voting and what is preferential voting?
Others have described the differences between first-past-the-post and the single transferable (or preferential) system of Australia much better than I can (see CGP Grey’s video, this amazing cartoon or the official source), but I’ll give it a go.
Let’s say there are 10 votes in an electorate. They are distributed among four candidates like so:
- Candidate A: 5
- Candidate B: 3
- Candidate C: 2
- Candidate D: 0
In this situation, both first-past-the-post and preferential voting work exactly the same. More than 50 per cent of the electorate has decided Candidate A should represent them, and so they are duly elected.
But now let’s change the results:
- Candidate A: 4
- Candidate B: 3
- Candidate C: 2
- Candidate D: 1
Candidate A still received more votes than any single one of their rivals, but they did not take the majority. This is where the systems differ.
Had this been an election for a seat in the national lower houses of Canada, the United Kingdom or the United States, Candidate A wins, even though most of the electorate voted for somebody else.
In Australia’s system, however, a candidate for the House of Representatives cannot be elected unless they receive the blessing of 50 per cent plus one of the voters in their electorate.
Rather than ticking the box next to their preferred candidate, ignoring the others and slipping their voting paper into the ballot boxes, Australian voters at a federal level must number every candidate’s box in the order of their preference (their top choice with ‘1’, their next choice with ‘2’, all the way down the paper).
If no candidate receives 50 per cent plus one of the boxes marked ‘1’ on ballot papers, the candidate with the fewest votes is eliminated and their votes are distributed to the ‘second’ preference on those ballots. In our scenario, Candidate D is out of the race, and their single vote goes to Candidate B.
- Candidate A: 4
- Candidate B: 3 + 1 (4)
- Candidate C: 2
Candidate D: 1
So Candidate B has picked up an extra vote, but nobody has reached that magic majority. Now it’s Candidate C’s turn to say farewell, and their second preferences are distributed to the remaining two candidates.
- Candidate A: 4
- Candidate B: 3 + 1 + 2 (6)
Candidate C: 2 Candidate D: 1
And there we have it: a majority! Candidate A may have received the highest number of first preferences, but the majority of the electorate decided that they ultimately wanted Candidate B to represent them.
After preferences are distributed, the final count between the winning candidate and immediate runner-up is known as the two-party-preferred or two-candidate-preferred statistic. So in the above instance, Candidate B would win over Candidate A 60-40 per cent on a two-candidate preferred basis.
On May 18 this year, the Liberal-National Coalition was returned to power at the federal level in Australia, forming a majority government with 77 of the 151 seats in the House of Representatives.
While the incumbents received only 41.44 per cent share of the primary vote, their share of seats was boosted through preferences to eventually command a two-party-preferred majority over the opposition Labor party, 51.3 per cent to 48.47. Similarly, Labor achieved 33.34 per cent of the primary vote, but came to control 45 per cent of seats through preferences.
These charts show the outcome of the election by seats won:
But what would happen if preferential voting had been taken away at the election, and replaced by first-past-the-post? A disclaimer: this assumes voters would have made the exact same choice under first-past-the-post as in reality, which would almost certainly not be the case. The following charts show first-past-the-post results by seats won:
As you can see, the Coalition’s majority has blown out from two seats to a comfortable 14. The biggest loser is the opposition Labor party. Labor did not achieve the highest primary vote in 10 of the seats they ultimately won. Through preference flows from the Greens and other minor parties, however, they were able to achieve more than 50 per cent of the vote on a two-candidate-preferred basis.
This would not have happened under first-past-the-post. The below table shows the seats that would have elected a different representative, had first-past-the-post been used, as well as the primary vote on which these new candidates would have won.
Although the majority of voters in these electorates ultimately preferred the Labor, independent or Centre Alliance candidate over the Coalition, the latter would win under first-past-the-post with a “plurality” (but minority) of votes.
So why have preferential voting?
The main motivation for preferential voting is to ensure that no single vote is “wasted”. It also, theoretically, allows voters to make their first choice for the party that best represents them and eliminates the need for “strategic” voting.
Let’s say you’re voting in the Canadian election later this year. There are three candidates vying for election in your riding: Conservative, Liberal, and NDP.
You might ideologically identify most with the NDP, but their candidate isn’t very popular, and you know that the riding has traditionally been a contest between the Conservatives and Liberals. You don’t want your vote to be “wasted” on someone coming in third-place, so despite your misgivings about some of their policies, you place your single vote with the Liberals as they are more ideologically aligned with you than the Conservatives.
This strengthens the power of major political parties at the expense of smaller ones, and tends to artificially inflate the number of seats a winning party (and even the primary opposition) will take.
In the 2018 Ontario election (Canada’s most populous province), under first-past-the-post, the Progressive Conservatives achieved 40.5 per cent of the popular vote, but took 61 per cent of the seats. Of the 124 seats in Ontario’s Legislative Assembly, 74 were won by candidates with less than 50 per cent of the vote (44 Progressive Conservative, 22 NDP, all seven of the Liberal seats and the sole Green seat).
Meanwhile, in the 2019 New South Wales election (Australia’s most populous state), while the Liberal-National coalition achieved a 41.58 per cent share of the popular vote, but a two-party preferred vote of 52.02 per cent, translating to 51.6 per cent of seats.
It’s entirely possible that, under a preferential system, Ontario would have ended up with the exact same parliament. In any single-member system in which constituencies are determined by geography, the popular vote is an imperfect metric, and one that ignores regional preferences.
But without the single transferable vote, we ultimately have no idea where the 59.5 per cent of the electorate that did not vote for the governing party would have directed preferences beyond their first choice.
And why go without preferential voting?
It’s far easier to explain first-past-the-post to prospective voters: tick a box next to your preferred candidate; if they get more votes than any of their opponents, they win.
The tallying of votes is also much simpler under first-past-the-post. Tight elections can be called earlier, as the ballots would need to be counted only once. In countries without the compulsory voting system Australia maintains, creating a potentially confusing new system could lead to a rise in informal voting, or simply low voter turnout.
Plus, with the exception of very tight election races, preferential voting would be unlikely to change the result of who would form government. As seen in the table, the Coalition still won most seats in a preferential system, but that working majority is far smaller than under first-past-the-post. Preferential voting in Australia still tends to grant majority government, but without that artificially inflated boost in seats.
Unlike in systems with proportional representation, in which majority government is rare and coalitions must be formed among many parties, preferential voting tends to favour the major parties. The Australian Greens, for example, were granted just one seat despite a 10 per cent primary vote. The below tables illustrate this: here are the first preference percentages:
And this is the actual share of seats in the House of Representatives:
As Antony Green says in a reply to his analysis of the 2015 Canadian federal election results:
“It is true that some minor party voters would no longer feel the need to strategic vote. However I think it is safe to assume people going back to their preferred party with their first preference would give their second preference to the party they would otherwise have strategically voted for.”
As such, the push for proportional representation, rather than preferential voting, has gained popularity in some quarters as a way to ensure a more representative voice in parliament, as well as granting more of a say to those who vote for minor parties.