South Australia 2018 Archive


What happened to SA Best?

This is quite a long analysis of the SA Best performance at yesterday’s state election. If you stick with it I’ve included a chart comparing SA Best to the Nick Xenophon Team in 2016, and at the end there’s a map! Enjoy.

Expectations were very high for Nick Xenophon’s SA Best party, with earlier polling suggesting the party had the potential to break apart the two-party system in South Australia. Yet their support in the polls dropped away as we got to election day and did not manage to win any lower house seats.

It’s worth a reminder that the SA Best vote is pretty good for a minor party. The party is sitting on 13.7% in the House of Assembly, and 18.9% in the Legislative Council. That will be enough to win two seats in the Legislative Council, while the Greens will likely only manage one seat and Cory Bernardi’s Australian Conservatives will likely miss out entirely.

There was some polling in late 2017 which put SA Best neck-and-neck with the major parties, but it now appears that these polls were outliers. A Newspoll in late 2017 had the new party on 32% of the primary vote, while both major parties were stuck below 30%. One other Morgan poll put SA Best on 28.5%. If you exclude those two polls, no other poll had the party on anything more than 22%, which isn’t that much more than the final upper house vote of 19%.

The public narrative suggests that SA Best suffered from a polling collapse, but I’m not so sure. There definitely was some decline – there was a 4-point drop in the last Newspoll, and that was 3 points above the actual result – but I doubt the figures in the high 20s or low 30s, which implied SA Best wiping out a major party and taking over ten seats, were ever anything other than outliers.

Even though there was only a small amount of polling, SA Best had a lot of hype, which may have contributed to Nick Xenophon’s decision to resign from the Senate and contest the seat of Hartley, or the late surge in the number of SA Best candidates, until they were running in 36 out of 47 seats.

Read on below the fold for more about SA Best.

Become a Patron!

Prior to the election I used a breakdown of the NXT Senate vote at the 2016 federal election by state seat to attempt to predict the SA Best vote. So how did this perform?

The model worked reasonably well. This graph compares the vote in the 36 contested seats to the Senate vote, and the trend is very clear.

There were a few outliers, including Xenophon’s seat of Hartley, where he polled only 2% less than his party polled in the Senate in 2016, but in most areas there was a drop in the SA Best vote relative to 2016, and the drop was relatively consistent.

Nick Xenophon’s various political machines have tended to produce relatively even votes across the state. This gives him a broad base to win Senate and upper house seats, but it also means that he could poll quite well and struggle to win single-member electorates.

This trend hurt SA Best badly last night – the party’s vote of almost 14% was only enough to just crack 25% in three seats.

The party was very effective at breaking into the top two, but in most cases this didn’t give them a chance of winning. SA Best are currently sitting in second in twelve seats. In those twelve seats, the gap between the second-placed SA Best candidate and the leading candidate ranged from 14.8% in Hartley to 35.5% in MacKillop. Those are massive gaps to close on preferences. A more lopsided vote count would’ve allowed the party to come first on primary votes, or at least come a close second, in a handful of seats, while polling poorly in many other seats, or perhaps not even running.

This makes me wonder how much effort was put into concentrating SA Best’s vote in a few key seats. Apart from Xenophon himself, no other candidates stood out as leading figures, and I didn’t see any evidence (admittedly from afar) of the party picking a handful of seats to ensure the party at least won a few seats. Perhaps it was just hubris, believing the party had a real shot at winning, say, ten or twenty seats and forming the opposition. If you really believe that you’d want to spread out your resources, not concentrate them.

This is the only way I can explain Xenophon’s choice of Hartley. The seat ranked twentieth on the list of NXT seats from the 2016 election. I get that it was his own local electorate, but it put a lot of faith in his party’s ability to sustain a very high vote, or his ability to push his personal vote well ahead of his party. The chart above suggests that Xenophon did benefit from a substantial personal vote, something that would’ve been enough to win if he’d contested a seat like Finniss, Heysen or Chaffey, but it won’t be enough in Hartley.

The last thing to note is that this is Nick Xenophon’s first attempt at contesting a lower house seat. He won his first seat in the Legislative Council on a tiny vote in 1997 thanks to favourable preference deals. He was re-elected to the Legislative Council in 2006, and then to the Senate in 2007, 2013 and 2016, each time with a massive vote closer in scale to the major parties than to other minor parties.

Nick Xenophon was on the ballot across South Australia at his last four elections – not this time. I’ve long wondered how well his parties would perform without him on the ballot. Xenophon-endorsed tickets at the 2010 and 2014 state elections did much less well without his candidacy, and I suspect that is a key factor in the drop in vote. It could also reflect the fact that state lower house MPs have much more profile than upper house MPs, and the major parties prioritise putting those people forward. It’s easier to vote for Nick Xenophon over an anonymous Labor ticket (particularly when a favourite lead candidate won’t have trouble winning re-election) than to vote for an SA Best candidate over a known local MP.

Finally, here is a map showing the relative vote for SA Best in the 36 seats they contested:


SA election – Liberals win government in status quo result

There’s been a lot of chatter about SA Best and their impact on the SA election, and I will return to them in my next post, but I wanted to focus here on the actual result. The overall result saw a change of government, but with practically no seats changing hands compared to the notional redistribution analysis.

The last election saw the Liberal Party poll 53% of the two-party-preferred vote, but fall two seats short of a majority, with Labor gaining the support of an independent to take power for a fourth term.

The recent redistribution completely undid the map which had allowed for an even result off such a lopsided result.

The Liberal Party gained four seats off Labor in the redistribution, with others also experiencing quite large changes in margin.

Once you factor in that change, it appears that only one seat has changed hands between the major parties (according to the ABC), with the Liberal Party on track to win the seat of King. King is a successor seat to Napier, and the redistribution reduced the Labor margin from 9.1% to 1.4%.

It’s possible that Labor will retain the seat of Mawson despite the seat being redrawn as a Liberal seat (a notional Labor gain), and could still gain the seat of Adelaide, which did not experience much of a change in the redistribution and remained a marginal Liberal seat.

Each major party also lost one seat each to an independent MP who had been a member of their party at the 2014 election: Labor lost Florey and the Liberal Party lost Mount Gambier.

This adds up to a total result of 24 Liberal seats, 18 Labor seats, 2 more in doubt (with Labor and Liberal leading in one each) and 3 independents.

SA Best is currently in the top two in twelve seats, which could conceivably have massively disrupted the Labor-Liberal contest, but most of these seats are traditional safe seats. SA Best is only sitting in the top two in two marginal seats: the Labor seat of Giles and the marginal Liberal seat of Hartley, where Nick Xenophon is in danger of falling into third behind Labor. So the SA Best threat was mostly focused in traditional safe seats. This could have still had a big impact if the party had won some of these seats, but instead the major parties appear to have been completely successful at locking SA Best out of their heartland seats.

At this point we only have two-party-preferred counts in 31 seats. The ECSA made an incorrect selection in three other seats, so they will need to re-run a Labor-Liberal preference count, while in the other 15 seats an SA Best or independent candidate made the final two. We will eventually have two-party-preferred figures from every seat and every booth, and thus a statewide total, but that might take some time. When that’s available I’ll be sure to put together a map.

Since we lack this complete picture, we can’t precisely say what happened in terms of the contest between Labor and Liberal across the state, but the result in the two-thirds of the state where data is available suggests that Labor did gain ground on the Liberal Party.

The Liberal Party gained a positive swing in eight seats, while Labor gained a positive swing in twenty-two seats. There was no swing in the marginal Labor seat of Wright.

That’s it for now, I’ll be back later today with some more analysis of the SA Best vote.


Batman/SA election night live

Batman by-election results

CandidatePartyPrimaryPrimary %Swing2PP votes2PP %2PP Swing
Yvonne GentleRise Up Australia18952.562.56
Ged KearneyLabor3122642.228.013948353.382.1
Alex BhathalGreens3047041.195.573448346.62-2.1
Kevin BaileyConservatives47246.396.39
Tegan BurnsPeople’s Party3580.480.48
Debbie RobinsonLiberty Alliance9571.291.29
Teresa Van LieshoutIndependent8751.181.18
Adrian WhiteheadIndependent6170.830.83
Mark McDonaldSustainable Australia73711
Miranda SmithAnimal Justice21072.851.42

Primary booths reporting: 44/52
2CP booths reporting: 44/52
Last updated at: 23:06

11:23pm – As for South Australia, the Liberal Party has won government, almost certainly with a majority. There is, however, a handful of seats still in play. The Liberal Party could win as many as 26 seats, but is leading in 25.

The upper house will see the major parties and the Greens hold steady, while the rest of the crossbench shifts. The Dignity Party has lost their sole seat, while the Conservatives have lost one of the two seats they inherited from Family First, with both those seats going to SA Best. The new government will need either SA Best or the Greens to pass legislation, in addition to the other party, or one of the other two members. One of these members, John Darley, was allied with Nick Xenophon when he succeeded Xenophon in the upper house in 2007 and when he won re-election in 2014.

11:19pm – I’m going to finish up this liveblog now, with a summary of each election. Firstly, Batman.

I’ll be back in a few minutes with a separate post showing the maps for Batman, but the quick summary is that we are still waiting for some pre-poll votes, as well as quite a lot of postal votes, but they hold no hope for the Greens. On current numbers, Ged Kearney has gained a 2.1% swing to Labor.

There was a very clear trend where the Greens gained small swings in northern booths, while Labor gained large swings at the southern end, including a gobsmacking 34% swing in Northcote West. We just received the result from the Northcote pre-poll booth, which produced a 6.7% swing to Labor.

10:27pm – Current numbers suggest that Labor and Liberal will each maintain their four upper house seats, with two SA Best members and one Green. So that’s a loss for Dignity Party’s Kelly Vincent and the Conservatives’ Robert Brokenshire.

10:04pm – We’ve now got the votes in the Thornbury pre-poll in the centre of Batman, with Labor gaining a swing of 2.8% in that booth. This has properly shut the door on the Greens there.

10:01pm – So there are 23 seats which are locked on for the Liberal Party, with them looking good to win Newland. They are behind, but still with a shot, in Mawson, while Heysen is up in the air. So they could end up on as many as 26 seats. Even if they don’t win any of these seats, the independent MP in Mount Gambier should give them the numbers they need.

9:58pm – If you’ve found this analysis useful, tonight and throughout the campaign, maybe you could consider signing up as a patron?

Become a Patron!

9:31pm – I’ve been focusing on analysing data at the Guardian liveblog but here’s where things stand at 9:30:

  • Labor looks set to win the Batman by-election with a swing of over 1%. We’ve just started receiving pre-poll votes, but there’s no evidence of a shift in the vote.
  • The Liberal Party look set to win a majority in the South Australian election, with a few seats just firming up now. SA Best may not win any seats, with only one seat still in play for the third party.

7:41pm – Things don’t look good for Alex Bhathal and the Greens in Batman. Nearly every booth is swinging to Labor after preferences, with some large swings in southern booths.

7:11pm – After three booths, the swing to the Greens after preferences is 0.88%.

7:07pm – First three booths for the Greens in Batman include two large booths in Labor-friendly Reservoir. Overall the Greens vote is up 8.9%, while Labor is up 2.2%.

6:38pm – And polls have now closed in South Australia. Still no data from Batman.

6:00pm – Polls have just closed in the by-election for the federal seat of Batman in Melbourne’s inner north. Polls will close in half an hour for the South Australian state election, with polls closing in three hours in the lower-profile Cottesloe state by-election in Perth. I’ll be using AEDT times for my coverage tonight. I’ll also be dropping in to the Guardian’s election night liveblog.

For the Batman by-election I will be posting overall results (top of the page), which include swings based on the vote in 2016 at the booths reported so far. This can be extrapolated to determine projected final votes. Hopefully this is useful.


Super saturday – election day thread

Polls have now opened in the South Australian state election and the federal Batman by-election, and will be opening in the state by-election in Cottesloe within the next half-hour.

This post is mostly to provide a place to discuss today’s events, but I did want to touch briefly on the two statewide polls released last night for South Australia.

Both have very similar numbers, with the Liberal Party leading on 34%, Labor on 31%, and SA Best on 16% (Reachtel) or 17% (Newspoll). We know that Newspoll has dialled down their numbers for SA Best to reflect the fact that they were running everywhere (the last poll had them on a statewide figure of 27% but only 21% if you exclude voters who can’t vote for them), presumably Reachtel is similar.

A vote of 16-17% would be very good for a minor party, and will undoubtedly be enough to win some seats, but it’ll be a small bench and suggests that Nick Xenophon may struggle to win Hartley (as a local poll has emphasised).

I’ll be running a liveblog here while also contributing to the Guardian’s liveblog. The main thing I plan to include here is a model matching the booths in Batman. Of course, expect some booth results maps late in the night (this time I’m confident I’ll get access to the data on the night).

Good luck voting!


SA election – nominations close

The official list of nominations for lower house seats in the South Australian election was published late yesterday afternoon, and I have now updated all 47 seat profiles.

264 candidates nominated, an increase from 204 in 2014. This is due to an increasing number of minor parties running large numbers of candidates.

You can view the dataset here.

Six parties are running candidates in 30 seats or more. Labor, Liberal and the Greens are running full statewide tickets, while Nick Xenophon’s SA Best is running in 36 seats, Cory Bernardi’s Australian Conservatives are running in 33, and Kelly Vincent’s Dignity Party is running in 30. Three smaller parties are running nine candidates, along with fifteen independents.

There are only three candidates in Kaurna and Stuart, while there are nine candidates in Croydon.

There’ll be more about the SA nominations below the fold, but I also wanted to mention that nominations also closed last week for the federal Batman by-election and the state Cottesloe by-election, both due on the same day as South Australia. Ten candidates are running in Batman, and seven are running in Cottesloe.

Read the rest of this entry »


Nick Xenophon’s vote – does it hit Labor or Liberal harder?

Last week, I posted data showing the Nick Xenophon Team vote at the 2016 Senate election broken down by state electorate.

In the process of completing my guide to the South Australian state election, I noticed the trend in terms of which seats popped up with very high NXT votes. It appeared that numerous safe Liberal seats ranked highly in terms of NXT vote, while the safer Labor seats ranked more lowly.

It’s undoubtedly true that Xenophon takes votes from all parties – when his vote has been particularly high, both major parties have suffered big hits. But the vote does tend to be higher in electorates normally considered ‘safe Liberal’.

We saw this clearly at the 2016 federal election – the NXT vote was highest in Mayo, Barker and Grey, all places where the Liberal Party would normally walk all over Labor in a head-to-head fight.

And when I compared the 2016 NXT Senate data to the Liberal two-party-preferred vote in each new state seat, there is a trend where the NXT vote is higher in stronger Liberal seats.

It’s not a perfect trend, but the correlation is around 0.558.

So what does this suggest about the impact that Nick Xenophon’s SA Best could have on the next election?

We obviously don’t know how highly the party will poll – they have only announced a handful of candidates, so it’s conceivable they won’t run a full ticket. But it seems more likely that a large vote for SA Best will hit hardest in Liberal seats, particularly in seats where the Liberal Party would normally not need to campaign hard. In contrast, the safer Labor seats are likely to stay safe, unless SA Best polls particularly strongly.


Launching the guide to SA 2018

As we near the conclusion of 2017’s election season, with the Bennelong by-election this weekend, I have finished my first major guide for 2018: for the South Australian state election on March 17.

I have completed seat guides for all 47 electorates, as well as a guide to the Legislative Council and summaries of the key seats and the redistribution.

You can use the following pages to find your way to each seat’s profile, or click through on the following map.

If you find this guide useful, you can choose to sign up as a monthly patron to maintain and expand this website’s coverage.


SA election – how do we predict Nick Xenophon’s vote?

The recent Queensland election produced a dilemma for us electoral analysts: polls suggested One Nation would perform strongly, but the party had no recent history of contesting seats in most of Queensland at state elections.

We relied instead on the results of the 2016 federal Senate election, broken down by Queensland state electorate. This work was conducted by Alex Jago, who used data on which voters from each block voted at each polling place. He also used the entire preference dataset to distribute votes cast for minor candidates between Labor, the LNP, One Nation and the Greens.

The South Australian election has produced a similar dilemma. Nick Xenophon’s SA Best party has never contested a South Australian state election. Nick Xenophon previously ran for the state upper house in 1997 and 2006, polling quite highly in 2006, before running for the Senate with strong results at the 2007, 2013 and 2016 federal elections.

Thankfully Alex Jago performed a similar task for South Australia, distributing votes at a SA1 level between Labor, Liberal and the Nick Xenophon Team. He then gave it to me and I matched those SA1s to South Australian state electorates, to allow me to produce an estimate of the vote for the three biggest parties in each state seat. You can view this data here.

NXT polled well everywhere – the lowest NXT vote was 21.7% in Croydon, and the highest was 38.5% in Heysen. But there is a trend. The NXT vote was highest in the seats in the Adelaide Hills and to the south of Adelaide, as well as those in the north of the state, and it was lowest in centre and northern Adelaide.


Map update – South Australia and Tasmanian upper house

I’ve recently completed two new maps for download and use: the (kind of) final boundaries for the 2018 South Australian state election, and draft boundaries for the Tasmanian upper house.

South Australia’s state redistribution was overshadowed by the federal election last year. A final set of boundaries was released late last year, with some significant changes to the draft boundaries in southern Adelaide, but these boundaries are stuck pending a lawsuit by the SA Labor Party. Both the first draft and final draft can be downloaded from the maps page, and the map is embedded here.

I’ve also completed the draft boundaries for the Tasmanian Legislative Council. The Tasmanian upper house consists of fifteen single-member electorates, but its members are elected in a very odd way: only 2-3 seats are elected each year, with members serving a six year term. Boundaries are redistributed roughly once a decade, with the sitting members assigned to finish their term representing a new seat.

There have been some major changes to the boundaries along the east coast of Tasmania. The three Launceston-area seats have remained largely the same, as have the four Hobart-area seats and the two rural seats to the west of Hobart. The west coast seat of Murchison has undergone minor changes.

The east coast seat of Apsley has been chopped up, while the seat of Rumney in the south-eastern corner of the state has been pulled in closer to Hobart, losing Sorell and the Tasman peninsula. A new seat of Prosser stretches halfway up the east coast from the Tasman peninsula to Swansea, while the remainder of Apsley has been moved into a new seat of McIntyre.

The seat of Western Tiers has been chopped up, with the north-western seats of Montgomery and Mersey expanding south and the south-western seat of Derwent expanding north. The remainder of Western Tiers has joined the remainder of Apsley as McIntyre, a strangely-shaped seat curving around Launceston, stretching from Cradle Mountain to Flinders Island.

I would expect the final boundaries for the Tasmanian upper house to be determined later this year, and the new boundaries will be used for the first time in 2018.


SA state redistribution – draft boundary analysis

screen-shot-2016-10-19-at-8-22-18-pmWhile we were all recovering from the federal election, South Australia was undergoing a redrawing of its state electoral map for the 2018 state election. The draft electoral map was released in mid-August, and I blogged about the underlying statistics driving the redistribution at the time.

It’s taken me some time to get back to this redistribution, what with the many territory elections, council elections and by-elections taking place following the federal election, but it’s now complete.

You can download the 2018 draft map here.

The last election produced a result of 23 Labor seats, 22 Liberal seats and two independents. One of those independents, Geoff Brock, sided with Labor to give them a governing majority, while the other independent, Bob Such, went on sick leave soon after the election and later died of cancer. During Such’s absence, former Liberal leader Martin Hamilton-Smith resigned from the party to join Labor’s cabinet. Labor subsequently won the Fisher by-election, giving them a majority in the House of Assembly.

The Electoral Districts Boundaries Commission is required by law to consider the party-political impact of the redistribution, with the aim of producing a result which will give a majority of seats to the party that wins a majority of the two-party-preferred vote. Despite this requirement, Labor won a majority in 2010 despite a Liberal vote majority after preferences, and achieved government again despite losing the vote in 2014.

In line with their mandate, the EDBC has redrawn the boundaries to boost the Liberal position. Assuming no change from the 2014 election vote which gave the Liberal Party 53% after preferences, the new electorates would give the Liberal Party 24 seats and Labor 22. The last seat, Frome, is drawn as notionally Liberal but is held by an independent. Theoretically this should mean that the Liberal Party should be able to win a majority with no change in their vote (assuming they can win back Hamilton-Smith’s seat), although this theory did not work at the last two elections.

45 electorates remain notionally held by the party that won them in 2014 (either at the general election or, in the case of Labor and the seat of Fisher, at the by-election). The other two seats are Elder, in southern Adelaide, and Mawson, which has moved from being a southern Adelaide seat into a regional electorate by stretching out to take in Kangaroo Island and the Fleurieu Peninsula. Both seats are held by Labor MPs but are now notional Liberal seats.

The following map shows the new electoral map. Click on each seat to see the post-redistribution margin, and who held the seat before the redistribution.

This map is only a draft – we should be expecting a final version of the map to be released in November.