Australia has used preferential voting for over one hundred years, and in that time the party system has changed quite a bit. Since the 1950s, there has been an increasing trend of minor parties picking up votes, but until recently these votes were not enough to actually win many seats. So their main contribution, in the House of Representatives at least, was in having their votes flow as preferences to one of the major parties.
Even in the current system, about 5/6ths of all seats end up as Labor vs Coalition contests. So for today’s post I wanted to explore how the flows of preferences have changed over time, in general and then for particular minor parties, and how they flowed in 2022.
This first chart starts in 1958 – prior elections had uncontested seats which reduces the value of the data. It shows what share of the vote for non-Labor and non-Coalition votes ended up flowing to Labor or the Coalition as two-party-preferred votes.
It’s worth noting that this chart was calculated at the national level, so it doesn’t capture preferences that may have flowed from one Coalition candidate to another.
In the 1950s and 1960s, the main source of preferences was the Democratic Labor Party, which despite their name favoured the Coalition with preferences.
They faded from the political landscape in the early 1970s, and the next dominant minor party was the Australian Democrats. There were also other contenders like the Nuclear Disarmament Party in 1984. Labor gained a majority of preferences for the first time in 1980.
Preference rates for Labor have continued to climb, but more slowly, with a dip in 1996 and 1998. Labor gained almost 65% of preferences at the peak in 2010, and the figure in 2022 was 61.54%.
Some of the same dynamics can also be seen in this next chart. It shows the number of seats where the successful candidate was not leading on the primary vote, and the colour represents the party of the winning candidate.
Prior to 1990, come-from-behind wins were overwhelmingly won by Coalition candidates. Labor had just one come from behind per decade in the 1960s and 1970s, and three wins at the 1980 election.
This was a combination of two factors: DLP preferences favouring the Coalition, and the Liberal and Country parties running candidates against each other and defeating Labor on the other Coalition party’s preferences.
The 1990s in hindsight look like a transitional period. The Coalition still won from behind on occasion, but Labor started having these wins regularly.
But for the last two decades, the Coalition hasn’t had a single come-from-behind win.
There has also been an increasing number of come-from-behind wins in recent years, although the total number of cases hasn’t gone far above the previous peaks in 1972 and 1984. There were 16 come from behind wins in 2016 and 2022.
The final twist was in 2022, when there were nine come-from-behind wins by independents and minor parties. Most of these were still at the expense of the Coalition, but Labor did lose from ahead in Fowler.
All of this analysis just treats the minor party and independent vote as one big block but of course they are not. You can see the overall trend shifting as the DLP was replaced by the Democrats, then the initial rise of One Nation and finally the rise of the Greens.
This next chart shows the specific preference flows to Labor on the 2PP for primary votes for the Greens, Democrats, One Nation and Palmer United/United Australia since 1984.
It’s worth noting this does include some elections where these parties lost their status as significant minor parties. One Nation only ran in 15-35 seats at elections from 2007 to 2016, and have only come close to running in every seat at the 2022 and 2025 federal elections. The Democrats still ran in a majority of seats in 2007, but polled just 0.7%. The Greens only ran in 36 seats. They ran in almost every seat in 2001 and have run in every seat since 2004.
The Democrats may have been founded by an ex-Liberal MP but by the mid-1980s they were giving a majority of their preferences to Labor, and it slightly increased over time. Their peak flow of preferences to Labor was in 2001, the last election before they collapsed.
The Greens have seen a steady increase in their share of preferences flowing to Labor, from 66% in 1990 to 86% in 2022.
One Nation unsurprisingly has always had preferences that favour the Coalition, but never by as much as the Greens favour Labor. The initial incarnation of One Nation gave just 54% of preferences to the Coalition, and it stayed around that level until 2016, when they ran in just 15 seats and gave just 50.5% of preferences to the Coalition.
One Nation massively outperformed expectations in 2016, winning four Senate seats. After three years back in parliament and a rebuilding of the party, their preference flows changed dramatically in 2019, with almost two thirds going to the Coalition. The result was similar in 2022.
Finally there are less elections to use to analyse Clive Palmer’s various parties, but there was a dramatic shift between the 2013 Palmer United Party, with 54% of preferences flowing to the Coalition, compared to the 2019-22 United Australia Party, with 62-65% of preferences flowing to the Coalition. This reflects his shift from positioning his party as a centrist alternative to being a party clearly on the minor right.
Preference flows from independents are fascinating, because they have moved around. Traditionally they favoured the Coalition, although there was a period in the early 1990s when Labor won a majority of independent preferences. I suspect that this reflects the numbers of green independents running at the time, and those forces were gradually absorbed into the Greens, but I have not fully investigated that question.
The big shift in the independent preference flows took place at the 2013 election. MPs like Tony Windsor and Rob Oakeshott retired. This may reflect a change in who was running, or that Coalition-leaning voters who had given their primary vote to independents fled while Labor-leaning voters stayed. Labor now gained almost 64% of independent preferences, a very different picture to the pre-2013 situation.
Understanding preference flows is crucial to conducting opinion polls. A given primary vote will produce quite different outcomes if preference flows change, as they can do from time to time. Typically pollsters will use the ‘last election’ estimates when calculating preference flows, but this can be difficult when the range of options in a poll don’t match the previous election results. Some other pollsters instead just ask the voter and use those preference flows.
It’s worth noting that there was an uptick in One Nation’s preference flows to the Liberal National Party at the 2024 Queensland election, compared both to the 2022 federal election and the 2020 state election. This, combined with the data from their own polls, led Newspoll to tweak their preference flow model to give more One Nation preferences to the Coalition. Peter Brent summarised the situation for Inside Story in February.
Preference flows will be fascinating to watch at this year’s election. Can the Greens preference flows to Labor go any higher, or are they at a ceiling. Does the surge in teal-ish independents lead to even stronger preference flows to Labor? Do the big minor parties of the right continue to polarise further and give the Coalition stronger preference flows? If we see right-wing parties start to give preferences to the Coalition at the rate that the Greens flow to Labor, and see teal voters become more Labor-friendly, it could indicate a growing polarisation in Australian politics. Not just a two-party system, but two blocs, with preferences staying within each bloc. Even if the parties don’t want to be grouped together, the voters appear to be doing this for them.
It’s weird that Labor bag the Greens so hard when Greens voters increasingly preference Labor ahead of Coalition. Do they just want the $3/vote, or is there something else at play in their antipathy?
Ben, I take it you haven’t looked at Senate numbers?
Josh, the ALP’s bagging of the Greens is even weirder when one considers a significant number of ALP seats are won *because* of Greens preferences. If the Greens stopped preferencing the ALP, they’d lose a lot of seats. What’s the saying about not biting the hand that feeds you?
Unfortunately we are heading down the southern European political system where an inability to form a majority will lead to a lack of consensus or reform and increasing debt which can never be repaid. In this country that probably means more regular centre left governments as they will traditionally benefit from greater preference flows despite the main centre right coalition winning significantly more primary votes. Until such time the centre right can provide a compelling and holistic economic reform story (along the lines of suggestions by Allegra Spender and Chris Richardson) we will head down the path of an increasing structural deficit as the economic pie is redistributed rather than grown and expenditure growth exceeds GDP. The coalition do not deserve to win this election based on their economic platform, notwithstanding an incredibly weak government economically and an intellectually inferior PM to any other in Australia’s recent history. There is no need at this stage to change tune for the majority when the unemployment rate is 4%. And the Treasurer can sleep well at night in the knowledge that those that have worked hard and sacrificed to build wealth (how dare they) will need to eventually pay the price at some point – the exact people he holds disdain for. My tip is that the credit ratings agencies will downgrade Australia’s sovereign credit rating over the next 3 years. The economic platforms of both parties will necessitate this.