Election preview: South Australia

Tomorrow’s state election in South Australia no longer looks as exciting as many commentators were saying a month or two ago. But it’s still going to be interesting, and it provides an opportunity to review some of South Australia’s remarkable electoral history.

Since 1944, there have been 23 state elections – this will be the 24th. All have been primarily contests between the Liberal (centre-right) and Labor (centre-left) parties; the Country Party was absorbed by the Liberals in 1932 (for a time they were known as the Liberal and Country League), and although it later re-formed it was never a major player.

In twelve of those elections, the party that won the two-party-preferred vote went on to form government. In the other eleven, it didn’t: the opposing party did.

Just let that sink in. South Australians could have saved themselves all the trouble and expense of elections to determine their government and tossed a coin every three or four years instead. It would have been just as likely to get the right result.

Both parties have missed out about equally (Labor six times, the Liberals five). Nor is there any correlation between margin of victory and likelihood of it being implemented: in three of Labor’s six biggest wins (in terms of votes), the Liberals ended up in government. Similarly, the Liberals missed out in three of their five biggest wins.

I’d be grateful for suggestions, but I’m not aware of any other jurisdiction that has so systematically and impartially frustrated the will of its voters.

The symmetry, however, is only superficial; there has been a major change over time. In the nine elections from 1944 to 1968, Labor won a majority of the vote seven times but only formed government once. Since then, however, the system has benefited Labor: it has won five of the last seven elections, but on only one of those occasions (in 2006) was it the choice of the majority of voters.

What happened, as readers might have guessed, is that a particularly gross malapportionment kept Labor mostly locked out until it was abolished in 1969. But fair electoral boundaries have not delivered fairness, partly because Liberal voters are heavily concentrated in a few (mostly rural) safe seats. Labor, with its voters more spread out, is consistently able to win more seats than its share of the vote would suggest.

A glance at the pendulum shows the problem: the Liberals have seven seats with margins above 15%, while Labor has only two.

Attempts have been made to overcome this; in 1991 the state’s constitution was amended to specify electoral fairness as a criterion for redistributions – the only provision of its kind in Australia. It has manifestly not worked. Antony Green takes up the story:

As it has applied since, the Electoral Districts Boundaries Commission (EDBC) tries to draw boundaries so that the major party winning the majority of the state-wide two-party preferred vote will also win the majority of the seats in the House of Assembly.

Experience has shown that the fairness provision has two problems. First, boundaries are always drawn based on past voting patterns, not the patterns at the next election. Second, as the Liberal Party keeps discovering, the fairness provision cannot stop an Independent winning a notional Liberal seat and choosing to back a Labor government.

This time around, the EDBC has approached its task more aggressively, redrawing boundaries in such a way as to put the Labor government of premier Jay Weatherill six seats behind its Liberal opposition. It needs a uniform swing of 3.0% to win a majority, even if it can count on the independent member for Frome, Geoff Brock (who sits in the Weatherill cabinet but holds what would otherwise be a safe Liberal seat).

Labor’s real position, however, is probably better than that. Three of the notionally Liberal-held seats have Labor sitting members, who will have the advantage of incumbency. And another two have sitting members who have resigned from the Liberal Party and sit as independents – and who, if re-elected, could potentially be lured to support Labor. (One Labor seat, Florey, is in the converse position, but Labor does not have the same history of its members being poached by the other side.)

And then, of course, there is the wild card of Nick Xenophon and his SA Best party. Earlier this year there was much talk to the effect that they could win a dozen or more of the 47 seats, perhaps even enough to leverage Xenophon into office. But those hopes have faded; SA Best has declined in the polls and has been treated badly by other parties on preferences. The betting market now predicts it will win only one or two seats; its candidate is narrowly favored in Heysen, and Xenophon and a few others are within striking distance.

Most of SA Best’s target seats are Liberal-held, so Labor probably has more to gain if it does well. But Xenophon himself comes from a Liberal background, and in a situation where the Liberals (again) win a majority of the two-party-preferred vote but SA Best holds the balance of power, it is reasonably likely that he would reach an agreement to make Liberal leader Steven Marshall premier.

Having been in power for 16 years, it would be an impressive achievement for Labor to win re-election. But with the national mood apparently still running in its favor, it is certainly in with a chance.

Polls close at 6.30pm, eastern Australian time. Check out Antony Green and William Bowe (the Poll Bludger) for both previews and live results.

One thought on “Election preview: South Australia

  1. I can’t recall an Australian election in modern times where the government has changed hands solely or primarily as a result of boundary changes. That appears to be exactly what has happened in South Australia. I think it’s clear now that the electoral commission got the redistribution for the 2014 election badly wrong.

    Interestingly the debate over the importance of partisan fairness in electoral boundaries has recently played out in the US. Pennsylvania’s supreme court has seen fit to rectify that state’s grossly gerrymandered congressional districts. But what’s the antidote to gerrymandering? Is it a strictly apolitical set of boundaries or is it a politically balanced map? It would appear that whilst the court has used compactness/community-of-interest as the primary criteria for the new boundaries, they have also made a concerted effort to achieve partisan balance.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.