A tale of two systems

Last Saturday we saw state elections in Australia’s two smallest states. Both states have been governed by the Labor Party for a number of terms and saw a resurgent Liberal Party threaten the ALP’s hold on power. In both states, we saw a swing away from the ALP. That’s where the comparisons end, because South Australia’s election was conducted using a single-member preferential voting system, while Tasmania uses the single transferable vote proportional representation system (known locally as Hare-Clark).

Last weekend stands as a perfect comparison between the two broad options in western democracy about how we organise our elections: do you go for a system of single-member electorates, or do you aim for a system that closely reflects each party’s vote in the seats in the Parliament?

In South Australia, we saw the Liberal Party poll 41%, with the ALP polling 38%. A further 7.8% voted for the Greens and 5.3% voted for Family First. As I write, the ALP is on track for 25 seats out of 47, with only 18 for the Liberal Party, and no seats won by either of the two significant minor parties. In addition, Antony Green has estimated that the ALP won approximately 48% to 48.5% of the two-party preferred vote. Despite this decisive win for the Liberal Party on both primary votes and two-party preferred vote, the ALP’s majority should allow them to govern without much trouble for the next four years.

In contrast, in Tasmania, the Liberal Party polled 39%, the ALP 37% and the Greens 21%. While the last few seats have not been determined, the collective wisdom has settled on a result of 10 Labor, 10 Liberal, 5 Greens, although a 10-11-4 or a 10-9-6 result is entirely possible. In contrast to  South Australia, the Tasmanian result very strongly correlates to the vote numbers.

It’s common to see mainstream media political reporters and commentators, as well as major party figures, bemoaning the prospect of a hung parliament and the ‘instability’ of minority government. Yet the Tasmanian result reflects the reality of Tasmanian politics: no-one won the support of the voters. No-one managed to get over 40%. In a single-member system, one of the parties would have managed to construct a majority of seats in the Assembly while 60% of voters had voted against them.

In contrast, the South Australian system constructs an artificial majority in Parliament for a party that didn’t receive the votes of 62% of voters. A single-member electoral system prizes giving total power to one political party over any questions of whether that reflects the voters’ wishes. In addition, the South Australian result gave power to the major party preferred by less voters, even in a two-horse race. It demonstrates that, in a system of single-member electorates, it is impossible to construct a result that can be considered ‘fair’.

South Australian electoral law goes to a lot of trouble to construct a ‘fair’ set of boundaries. A redistribution is held every four years and redraws the boundaries to ensure that a two-party preferred vote of 50% results in an even number of seats, supposedly to ensure that a party with a majority of the two-party preferred vote wins a majority of seats. The system was established following the 1989 election, when the Labor government held on with a one-seat majority despite polling 48.1% 2PP.

Last weekend’s result demonstrates the folly in trying to introduce the concept of fairness into a system that fundamentally ignores fairness. The 2010 boundaries were drawn based on the 2006 result, and assumed a uniform swing. Instead, the ALP lost a lot of votes in safe seats, not in marginal seats, as Possum has examined. This has resulted in a similar result as in 1989, with a long-term Labor government holding on while a majority of voters express  a preference for the opposition.

No electoral system can exactly reflect the voters wishes. It is possible that the ALP will win more seats than the Liberal Party in Tasmania. The key difference is that, in a single-member system, a distortion results in a majority government ruling while being opposed by a majority. In contrast, the Tasmanian system will mean that, regardless of the individual seat results, a majority in the Parliament will need to be made up of two of the three parties, and any of those combinations adds up to a solid majority of votes cast at the election.

If you want a system that actually reflects voters wishes in any way, you can’t use a system of single-member electorates which has such a checkered history of electing governments who are opposed by a majority of their constituents.