SA2026 – huge win for Labor but a warning in the result

The result of the recent South Australian election was clear with Labor the biggest winner, the Liberal Party the biggest loser, and Pauline Hanson’s One Nation the big improver.

In the aftermath Labor can bask in its massive majority and the smiting of its political opponents.

Division on the other side of politics eased Labor’s path to re-election, but the strong performance of One Nation in Labor’s safest seats has delivered Labor a serious warning.

Hidden in the detail of the results is that One Nation turned Labor’s side of the electoral pendulum upside down. Liberal seats and Labor’s most marginal seats swung to Labor, but many of Labor’s safest seats swung strongly to One Nation.

Much of this comes about because the South Australian election was not a two-party contest. Only 13 of the state’s 47 lower house seats finished as Labor-Liberal contests. The majority of seats, 25, finished as Labor-One Nation contests.

The result is the Electoral Commission of South Australia (ECSA) has abandoned past practice of undertaking alternative two-party preferred counts in contests that weren’t Labor-Liberal finishes. In 2026 the Commission will do a two-party re-count in seats where Labor did not make the final pairing, but in every other seat report a count that is Labor versus whoever finished second.

In the past there has been Coalition criticism of academics for analysing Australian elections as Labor versus Non-Labor. For the 2026 South Australian election, there is no other alternative.

It creates a post-election pendulum that is a mix-and-match of party contests, perhaps a portent of where Australian politics is heading. All swings since 2022 will compare a Labor-Liberal finish with whatever pairing is produced for 2026.

Which is where the Labor-One Nation swings in safe seats are startling. Sometimes a change in final pairing can create an artificially large swing caused by the vagaries of exclusion order and unexpected preference flows.

That’s not the case in South Australia. The two-candidate swing against Labor with these mix-and-match results correlates with the decline in Labor’s primary vote and the rise in support for One Nation.

The chart below shows the relationship between Labor pre-election margin and 2026 swing. Labor’s safest seats are to the right of the chart, and the array of points below the horizontal axis show how many of Labor’s safest seats swung strongly towards One Nation. The blue points for Labor-Liberal contests are all to the left and show a clear pattern of swings to Labor. But the swing against Labor increases as Labor’s pre-election margin increases.

The swing was to Labor in areas where the Liberal vote held up, but against Labor where the Liberal Party slipped to third and the swing is measured against One Nation. And this swing is heavily correlated with the decline in Labor’s primary vote and strong support for One Nation.

In safe Labor seats, there was a switch in primary vote from Labor to One Nation. There are safe Labor seats in northern Adelaide where One Nation polled above 30%, levels the Liberal hardly ever achieves.

The trend in the above chart stands out as starkly in the tables below which aggregate results for Labor seats based on pre-election margin. Labor came out of the election with a higher primary vote and after preferences vote in the seats it defended on margins under 15% than it recorded in the seats with margins above 15%. There was a 5.9% swing to Labor in seats with margins under 15%, but 6.7% away in very safe seats with margins above 15%.

Again, this is a mixed swing based on One Nation replacing the Liberal Party as the second placed in seats. But the swing occurred on primary votes as well as two-candidate preferred votes, so its not an artifact of the change in final pairing.

At a normal election both the above tables would swing in the same direction. But in South Australia in 2026, marginal seats swung to Labor and very safe seats against Labor. The seats of Elizabeth, Light and Taylor had margins above 19% before the election, but after the election are three of Labor’s four most marginal seats under 5%.

There’s more on the results inside the post with a link to a draft analysis I’ve prepared. It includes results for all seats, tables of preference flows and a new electoral pendulum.

But I’ll start with some comments on the looming Farrer by-election.

Looking Towards Farrer

In the next few weeks, political attention will switch to the Farrer by-election as the nation’s next electoral test.

In my opinion, the above analysis of Adelaide seats delivers a more rounded picture of where national electoral politics is going than can be gleaned from the result in Farrer.

Farrer is one of the most rural seats in the country. It has been in Coalition hands since its first contest in 1949. It has never come close to being won by Labor, Labor has fallen short of 20% at recent elections, and the party isn’t even contesting the by-election.

As the Coalition holds only a handful of metropolitan seats across the country, all the Farrer by-election can reveal is whether the Coalition can retain its deepest heartland seats in the face of One Nation’s resurgence. If the Coalition can’t hold Farrer, or in a worst case scenario not even finish second, what are its hopes of winning the next Federal election?

Could the Farrer by-election result trigger another change in Liberal Leadership and change in policy direction?

If One Nation is eviscerating the Liberal and National Parties in their remaining redoubts, what hope is their for the Coalition to be truly competitive against Labor at the next election? Farrer might tell us how One Nation is doing in Liberal and National regional heartland, but it tells us little about trends where the election will be decided, in city seats.

Which is why there should be more attention to what happened in metropolitan Adelaide in March. The results showed that One Nation can eat into support for a popular Labor government in its safest seats. It begs the question – how would One Nation go against an unpopular Labor government? The Victorian election in November may tell us.

Perhaps in the face of an expected easy victory for Peter Malinauskas, some Labor voters in Labor heartland felt safe in casting a protest vote over the course and content of current politics.

For Labor the luxury of a protest vote is something it must convince its supporters to resist.

For the Coalition, that One Nation replaced the Liberal Party as Labor’s major opponent in large parts of Adelaide is a more existential threat.

More on the South Australian Result

Based on preliminary results I’ve prepared a summary and analysis of the SA House of Assembly election that you can find at this link.

Moving on from the Federal implications, let me outline more on the winners and losers of the election and on how preferences flowed.

The biggest loser of the election was clearly the Liberal Party. Its primary vote almost halved, polling just 18.9% compared to 35.7% in 2022. The Liberal seat numbers crashed from 16 seats in 2022 to five in 2026. After decades of being in a two-party race with Labor, the Liberal Party slipped to third third behind One Nation. Only 13 of the 47 seats finished as traditional Labor-Liberal contests. The Liberal Party led the primary vote in only six seats, was placed second on primaries in 12 seats, third in 15, fourth in 12 and fifth in two seats. The Liberal primary vote fell to single digits in 13 Adelaide seats.

For the big improver, One Nation’s 22.9% was even higher than the 22.7% the party polled at its first outing, the 1998 Queensland election. One Nation had trailed the combined Coalition vote in 1998, but in South Australia the party finished second ahead of the Liberal Party. One Nation won four seats while another four seats were won by Independents. One Nation recorded the highest primary vote in four seats, was second in 29, and passed 30% in nine seats.

Labor significantly increased its majority despite a decline in first preference support. Labor’s primary vote was 37.5%, down from 40.0% in 2022. But Labor had the highest primary vote in 35 of the 47 seats, winning 34, a gain of seven seats compared to 2022. Labor’s victory was built on winning 32 of the 33 metropolitan seats. In regional areas Labor retained the Whyalla based seat of Giles, and also retained Mawson to the south of Adelaide.

Even though One Nation supplanted the Liberal Party across outer Adelaide, the metropolitan contests still had a rough left-right alignment. In regional seats the contest was more fractured with the presence of Independents. Of the 14 regional seats, the Liberal Party won four, One Nation four, Independents four and Labor two.

Preferences

45 of the 47 districts were won by the candidate leading on first preferences. Independent Lou Nicholson won Finniss from fourth place on 18.1% of the vote, and Independent Matt Schultz won Kavel from second place with 20.9%.

Labor won 34 of the 35 seats where it led on first preferences, including four seats where it recorded a majority on first preferences. Labor reached a majority after a partial distribution of preferences in 19 seats with a full distribution required in the other 11 seats won. The Liberal Party won Bragg with a partial distribution of preferences but all other Liberal, One Nation and Independent victories required a full distribution to determine the winner.

Preference flow data is not yet available, so the following statistics are drawn from the published distributions of preferences. Green preferences generally flowed 80-85% to Labor. One Nation preferences flowed 65-70% to Liberal despite One Nation not having recommended preferences on how-to-vote material. The Liberal Party did recommend preferences to One Nation but generally the preference flows were weaker at around 55-65%. Both Labor and Liberal preferences assisted Independents to win seats ahead of One Nation. (See preference summaries page 32)

State-wide Result After Preferences

As already pointed out, there was a clear swing towards Labor in seats that were Labor-Liberal contests, but against Labor in seats where One Nation finished second. There were 13 Labor-Liberal contests, 25 Labor-One Nation races, a Labor-Green contests where a Labor-One Nation count is available and of more use in calculating state wide totals. One other seat finished as Labor versus Independent.

In the remaining seven seats where Labor was excluded in the distribution of preferences, ECSA is conducting an alternative Labor versus Liberal preference count.

Using estimates for these seven seats, and adding together all the other seats, produces a Labor versus Non-Labor result of Labor 55.7%, Non-Labor 44.3%, a swing to Labor of 1.1%. I’m not sure how useful these numbers are compared to decades of available Labor-Liberal state-wide results.

If you add up the final counts in each seat to produce a state-wide two-candidate preferred figure, you arrive at Labor 50.5%, One Nation 26.1%, Liberal 17.9%, Greens 0.5% and Independents 4.9%. This is not a useful statistic.

New Electoral Pendulum

For all its faults, arranging electorates by margin still tends to be the most useful way of displaying an electoral contest.

Below is my best estimate of new margins for all 47 seats based on the results of the 2026 South Australian election. It would take the loss of 11 seats on a uniform swing of 8.5% for Labor to lose its majority at the next election. That is if anyone thinks it useful to talk about uniform swing given the 2026 results.