The Success of Local Independents – Kooyong

A key story of the last three federal elections has been the Liberal Party’s loss of once safe heartland seats to ‘teal’ Independents.

The list is a catalogue of seats once held by Prime Ministers and senior Liberal Ministers. In NSW there’s Wentworth, North Sydney (now abolished), Warringah, Mackellar and Bradfield. In Victoria it’s Goldstein (recovered in 2025) and Kooyong, and in Western Australia, Curtin.

These electorates top lists of the most affluent in the country. They encompass some of the most desirable and expensive real estate in each state. As recently as a decade ago, it was unthinkable they could leave the Liberal fold.

There are no Queensland or South Australian seats in the above list, but two could be. The Adelaide Hills seat of Mayo, held by Rebekha Sharkie since 2016, is effectively an independent as the only member of the Centre Alliance. Ryan, Brisbane’s leafiest and once safest Liberal seat, has been held by the Greens since 2022.

As well as being Liberal Party losses, these seats have seen a dramatic decline in their Liberal two-party preferred margins based on alternative preference counts versus Labor.

Indeed, Wentworth, Warringah, Ryan and Mayo recorded Labor two-party preferred majorities by alternative counts in 2025. Had a full two-party preferred count been completed in Bradfield, this once safest of safe Liberal seats could also have turned red.

The paradox is that Labor is achieving record two-party preferred results in these seats despite the party’s primary vote collapsing. The decline in Liberal two-party preferred results in these seats says more about Liberal support in its once safe seats than it says about Labor prospects.

Support for both sides of politics has declined since the high point of two-party politics in 1975. The decline has accelerated in the last two decades. Labor’s support is down sharply from its 2007 victory under Kevin Rudd. The Coalition’s collapse is an even steeper decline from a higher starting point at the 2013 election.

The loss for Labor has been in large part matched by rising support for the Greens. Much of Labor’s fall has also been in seats it doesn’t win, such as rural seats and Coalition seats contested by high profile Independents. This causes Labor’s loss of primary vote to have less impact on the number of seats it wins.

For the Coalition, and more particularly the Liberal Party, the destination of lost primary votes is less obvious. The Liberal Party has lost some support to Labor and to several conservative minor parties. But it is the loss of support in a limited number of seats to ‘teal’ and other centrist Independents that has damaged the Liberal Party by costing it once safe seats.

In this post I’m going to look at the loss of Liberal vote to the centre, to urban Independents. To illustrate the trend, I’m going to analyse one ‘teal’ seat in depth, the inner Melbourne seat of Kooyong.

The Kooyong Electorate

Kooyong is an affluent inner Melbourne electorate covering suburbs east of the Melbourne CBD beyond the Yarra. It includes what has long been a slice of upper middle class Melbourne. On its current boundaries Kooyong includes Hawthorn, Toorak, Malvern, Kew, Balwyn, Canterbury, Camberwell and parts of Mont Albert and Surrey Hills.

The seat has existed since Federation and has had just eight members in 124 years. The first seven were from the conservative side of politics before the seat was won by Independent Monique Ryan in 2022. Ryan was re-elected in 2025 overcoming significant boundary changes due to a redistribution.

The list of former members includes three conservative leaders, Sir John Latham (1922-34), Sir Robert Menzies (1934-66) and Andrew Peacock (1966-94). Peacock was succeeded by Petro Georgiou at a 1994 by-election, and he held the seat until retiring at the 2010 election. He was succeeded by Josh Frydenberg who served as Treasurer in the Morrison government. Had Frydenberg not been defeated by Monique Ryan in 2022, he would have been a contender for the Liberal leadership.

Past Results

The chart below shows Liberal two-party preferred results in Kooyong (blue line) since 1984 compared to the Victorian Liberal two-party preferred result (solid black line). The chart shows that Kooyong has long been one of the Liberal Party’s safest Melbourne seats, recording results well above the party’s state-wide result. Kooyong has normally recorded results 10-15% above the Liberal state-wide result, though the figure has slipped under 10% at the last three elections.

More relevant for the last three elections is the black dashed line of recent Liberal two-candidate preferred results. The final two points for 2022 and 2025 drop below 50% and show Monique Ryan’s victories.

The real story of what has happened in Kooyong comes from the next chart showing first preference support by party. The Liberal Party was never in danger of losing Kooyong while its primary vote support was above 50%. It dropped below that level for the first time in 2019 before falling below 45% for defeat in 2022 and 2025.

Labor’s support bobbed around 30% until 2007, one notable departure from trend being in 1990 when the Australian Democrats had a record vote. Between 2007 and 2019, Labor’s vote almost halved, matched by rising support for the Greens. At the last two elections, primary vote support for Labor and the Greens collapsed with the majority of vote for both parties decamping to Monique Ryan.

When I’ve been asked about how Independents win in safe seats, I’ve always mentioned a 45-30-80 rule of thumb. The holding party must be forced under 45% to ease the way for victory from second place on preferences. The Independent should poll 30% to insure they finish second. From that point achieving 80% of preferences from excluded candidates allows the second placed candidate to achieve victory.

If the leading candidate falls short of 45%, or the second candidate polls more than 30%, then the flow of preferences required for victory is lower.

How that played out in 2025 is shown in the chart of preference flows below. The Liberal primary vote was 43.7%, Ryan 33.9%, the vote for all other candidates 23.0%. With a smaller gap to close, Ryan required fewer preferences.

Ryan attracted 73.0% of preference from voters supporting other candidates. 76.1% of Labor voters gave preferences to Ryan ahead of the Liberals and an even higher 89.3% of Greens voters did the same. Voters for three right of centre parties delivered three-quarters of preferences to the Liberal candidates, but the three parties recorded only 3.3% of the primary vote between them.

What About Labor’s Rising Two-Party Preferred Result?

The earlier chart of primary votes showed that Labor’s primary vote had declined over four decades from 33.3% to 11.9%. Over the same period, Labor’s two-party preferred results rose from 37.0% to 43.7%. Labor was excluded from the final preference count at the last three federal elections, but a two-party preferred count based on an alternative preference throw by the AEC is available.

The diverging primary and two-party preferred Labor results are explained by the chart below. The grey line shows the total vote for Others, that is non-Labor and non-Liberal primary vote. The dark red line shows what proportion of voters for Other candidates directed their preferences to Labor, the line plotted as a percentage of the formal vote.

The orange line shows the Other voters flow of preferences to Labor as a percentage of the total Other vote, so it is using a different scale. As the Labor primary vote has declined, and as the vote for Others has risen, so has the flow of preferences to Labor risen. Labor’s loss of primary vote has been more than matched by the increasing flow of preferences from an ever growing pool of Other voters.

Another way of saying it is that the Liberal Party has lost primary support at recent elections and failed to compensate for the loss by attracting enough of those votes back as preferences.

The next chart is based on the alternate two-party preferred preference count for Kooyong and shows two-party preference flows for candidates including Independent Ryan. The chart is based on previously unreleased two-party preference flow data. It highlights how badly the Liberal Party has done in attracting preferences from voters casting a first preference for Monique Ryan.

But first, knowing this data is certain to be mis-represented as something Monique Ryan rather than voters have done, I need to make a couple of points.

First, the data played NO part in Ryan’s victory and NO part in the re-election of the Albanese government. The distribution of Ryan’s preferences was not required and was done for information purposes to allow the calculation of a national two-party preferred result. Ryan was helped to victory by the preferences of Labor voters, not the other way round.

Second, if anyone uses this data to say Monique Ryan sent 83.3% of her preferences to Labor, they either have zero knowledge of how preferential voting works, or they are deliberately mis-representing the data to spread mis-information.

The preference flows are accumulated and turned into percentages based on the following vote data. A total of 38,955 first preference votes were cast for Monique Ryan. Of those votes, 32,463 were completed with preferences that listed the Labor candidate higher than the Liberal candidate, and 6,492 listed Liberal before Labor.

Ryan issued an open how-to-vote asking for a first preference vote and left it to each voter on how they would complete preferences. For people who repeat the nonsense that voters blindly copy how-to-vote preference recommendations, Ryan’s preferences show that voters can make up their own minds.

What was the Preference Flow from Former Liberal Voters?

As the primary vote chart showed clearly, Ryan’s vote has been boosted by Labor and Greens voters voting strategically with a first preference for Ryan. It can be safely assumed that the overwhelmong majority of these votes then directed preferences to Labor.

But what about preference flows from former Liberal voters?

With secret ballots, it is of course impossible to know this number, but you can produce a rough estimate based on changes in primary vote.

Between 2016 and 2025 the changes in primary vote were –

• Liberal support declined from 58.2% to 43.1%, down roughly 15%.
• Labor support fell from 19.8% to 11.9%, so down roughly 8%.
• Green support declined from 18.9% to 7.7%, down roughly 11%.
• Monique Ryan polled nothing in 2016 and 33.9% in 2025, say rounded to 34%.
• Others was roughly unchanged, 3.1% in 2016, 3.3% in 2025.

I’m including a lot of ‘roughlies’ here because there were changes in boundaries ahead of the 2025 election, and comparing vote over nine years does not account for turnover of voters and changes in demographics.

So let me start with the rounded number 34% for Ryan derived from 19% Labor and Greens voters in 2016, and 15% from Liberal voters in 2016.

Let me assume a high 95% of switching former Labor and Greens voters gave preferences to Labor in 2025. What percentage of former Liberal voters who switched to Ryan between 2016 and 2025 gave preferences to Labor?

With a little algebra you can work out that around 70% of Liberal defectors gave preferences to Labor. Or alternatively, around 30% of Liberal deserters gave their preferences back to the Liberal Party.

Which roughly explains the two-party preferred swing between 2016 and 2025. The two-party swing was 10.9%, the loss of Liberal primary vote 15.1%. So about 28% of the Liberal primary loss comes back as preferences to lower the two-party preferred swing. That’s around the same estimate of how many Liberal defectors to Ryan gave preferences to the Liberal Party.

So What Does This Mean?

What all this shows is that the two-party preferred swing in Koyoong in recent years has been in line with the national two-party preferred swing. In a two-party contest, the Liberal Party would still have won Kooyong.

But as has been clear at recent elections, the Liberal party is not facing off against Labor candidates in its safe urban seats. It is having to compete with Independents to the right of Labor, creating a contest between two candidates putting forward different versions of centre-right politics.

By eating into primary vote support for the Liberal Party, the Liberals are being forced to preferences in once safe seats. And their Independent opponents then benefit from preferences from Labor and the Greens.

The Independents are drawing from both sides of politics, but by forcing the Liberal Party to preferences, Independents benefit in two ways. First from Labor and Green voters strategically switching their primary vote to the Independent, and secondly from the preferences of Labor and Green voters who stayed with their preferred party.

Does this mean Independents are ‘stealing’ seats from the Liberal Party. No, it shows that Independents are constructing majority support based on defectors from all parties that better suits the revealed preferences of the electorate.

While I have made a number of comments about Labor two-party preferred results in these seats, I still don’t believe that Labor would win these seats in the absence of an Independent.

In the absence of an Independent, voters in these safe Liberals seats would be faced with a choice between a candidate representing Liberal policy, and a candidate representing Labor policy. Two ends of the political spectrum. With an Independent in the field, there is a three-way choice and a majority coalesces around the centre in a way that would not happen for parties of the left in a two-party contest.

Also, a Liberal-Labor contest in Kooyong would be a considerably lower key race. The Liberal Party and Independent Monique Ryan have thrown huge resources at the seat at the last two elections. In my opinion, the scale and political volume of the Liberal campaign may have aided the Independent and assisted them on preference flows.

A Labor-Liberal contest in Kooyong would be lower profile, convert the contest into one between parties rather than candidates, and make it unlikely that Labor could do as well as appears possible by the published alternate two-party preferred counts.

The Australian party system is changing, struggling to adapt to society that does not have the simpler political cleavages of a century ago.

The argument that Labor can win safe Liberal seats is the obverse to the dispute in the Liberal Party about whether to adopt more populist policies that might appeal to voters in seats where Labor has traditionally been strong.

Both sides of politics are struggling to adjust to the end of a bi-polar political contest. Seats like Kooyong are an example of that change.

So far Labor has been better at tapping into the changing cleavages and alignments of more complex national politics. It has been shedding support in rural and regional seats it can’t possibly win. Labor’s supporters have been happy to swing behind both rural and urban Independents, with particular success in safe Liberal seats.

With the Coalition left holding only a handful of urban seats, it now finds itself under threat from a populist wave of One Nation support in the seats it continues to hold, almost all in rural and regional Australia.

How it battles against rural populism while crafting a message to attract voters in the nation’s capital cities is likely to determine who wins the next Federal election.

Much will be revealed by the next round of state elections, South Australia in March 2026, Victoria in November 2026, and New South Wales in March 2027.