Impact of the South Australian redistribution

South Australia redraws its electoral map every four years, after every election. The most recent redistribution was the first held since the removal of the ‘fairness clause’ from the state constitution, and saw more than a sixth of South Australian voters moved into a different seat.

The Liberal Party won 25 seats in 2018, with Labor winning 19, and three independents: Geoff Brock in Frome, Frances Bedford in Florey and Troy Bell in Mount Gambier. The new boundaries drew one Liberal seat to have a very slim notional Labor majority, and also forced two of those incumbent independents to switch to a different seat.

I estimate that 18.1% of electors were moved into a different seat.

My estimates flipped the marginal Liberal seat of Newland from its 2.0% margin to 0.3% for Labor, thus bringing the Labor seat count up to 20 and Liberal down to 24. Antony’s estimates have Newland as 0.1% Liberal, but I should emphasise that none of these estimates are precise, so we should probably just think of Newland as a tied seat.

There are six seats where less than 60% of the voters in the new seat were already contained in that seat. Just 36.5% of voters in Frome were already there, and 48.8% in the neighbouring seat of Stuart, primarily due to the transfer of Port Pirie from Frome to Stuart. Florey, Reynell, Playford and Hurtle Vale are all at least 40% new voters.

In terms of voter margin, the biggest changes were:

• Stuart – Liberal 2PP margin dropped from 23.1% to 11.9%
• Reynell – Labor margin dropped from 14.5% to 9.4%
• Frome – Liberal 2PP margin increased from 11.1% to 17.8%
• Florey – Independent 2CP margin dropped from 6.1% to 1.7% and Labor 2PP margin increased from 11.0% to 13.4%
• Hurtle Vale – Labor margin increased from 5.3% to 8.3%
• Hammond – Liberal margin dropped from 19.5% to 16.8%
• Playford – Labor margin increased from 16.3% to 19.0%

It’s worth briefly mentioning the fates of independent MPs Frances Bedford and Geoff Brock. In both cases their electorates were radically redrawn, with their strongest areas moved into a neighbouring seat, and in both cases they have chosen to contest a different seat. I’m planning blog posts next week focusing in on those changes.

There are three seats that would normally be considered safe Liberal seats where the sitting MP is now an independent and is contesting the seat, but if you ignore those seats the Liberal Party needs to lose just one seat in addition to Newland to lose their majority. That would require a swing of just 0.7% in King. If you give Labor Newland (they have a slim notional 2PP majority) and Florey (the sitting independent has moved elsewhere) they hold 21 seats, and need three more for a majority. This requires a uniform swing of 2.1%.

I should also note that I wrote a blog post in December 2020 analysing the redistribution. At the time I did not have booth-level two-party-preferred data for seats where the final count was not Labor vs Liberal, but I acquired this data (and added it to the data repository) in 2021, and subsequently recalculated the margins. It didn’t change much, but did flip Newland from Liberal to Labor.

Finally, this map shows the changes in boundaries from 2018 to 2022, and is followed by a table comparing the margins before and after the redistribution.

Liked it? Take a second to support the Tally Room on Patreon!