SA's Voter Choice Bill Could Be Better

Earlier this year, the Australian federal parliament successfully passed Senate voting reform, which abolished the Group Voting Tickets that had trashed the 2013 election, and returned control over preferences to the voters.  I've had plenty to say already about how successful that was (performance review part 1, part 2, JSCEM sub (PDF) etc).  However, while Group Ticket Voting has been put in the bin at federal level, hopefully to remain there for good, it remains in place in the upper houses of Victoria, South Australia and Western Australia. (For completeness, NSW fixed its system after the 1999 debacle, Tasmania has a single-member-per-seat Upper House and Queensland has no Upper House at all).

I'm not going to expend a huge amount of energy on trying to get state upper house systems reformed in states other than my own.  But with South Australia the first state where a move away from GVTs is the subject of legislation since the 2016 Senate outcome, I think it's interesting to have a look at what is being proposed.  In attempting to get rid of Group Ticket Voting, the SA Labor government has come up with a near-polar opposite.  The proposed alternative, while still much better than keeping GVTs (as is just about anything really) is so inferior to the Senate and NSW systems that I wondered for a moment if it was built to be voted down.  It appears this isn't the case, so hopefully parties in the SA Upper House (where the Government holds just eight out of 22 seats) will be able to amend the legislation to improve it.  An interview with the Attorney-General in July did suggest he was open to different models provided that preference harvesting got the chop, which is commendable.



The name of the bill is the Electoral (Legislative Council Voting) (Voter Choice) Amendment Bill.  It is worthwhile quoting at some length from Attorney-General John Rau's introductory speech:

-----------------------------------------------------------------------------------------------------------------
The Electoral (Legislative Council Voting)(Voter Choice) Amendment Bill 2016 proposes to change the voting system used for Legislative Council elections and implement what is to be known as the 'voter choice' method of voting. Voter choice is a variant on the current system of voting used for Legislative Council elections. Voter choice would essentially work as follows: there would no longer be voting tickets in Legislative Council elections.

As is currently the case, voters would be able to vote '1' above the line for the party or group of their choice. This would be known as a 'group vote'. Unlike the current system, where the vote above the line is interpreted in accordance with the voting ticket lodged by the particular party or group, a group vote would be a vote for each of the candidates in that party or group in the order nominated by that party or group.

Below the line voting, or an individual vote, would be largely unchanged, although there is provision relating to the interpretation of ballot papers so that where a person just votes '1' for the lead candidate of a party or group below the line that would be interpreted as a vote for the party or group above the line. In other respects, the voting system for Legislative Council elections would remain largely unchanged. The methods of calculating the quota and transferring surplus votes remain the same.

The voter choice method of voting would limit the potential for parties to secure Legislative Council seats through preference harvesting. The proposal is intended to make it easier for people to understand the implications of their vote and to have control over their vote and preferences. Voters who cast a group vote above the line will be casting a vote for the members of that group or party and not for all candidates in the election in the order of the group's voting ticket, as is currently the case. Voters who cast an individual vote below the line will continue to be required to indicate a preference for all candidates.
---------------------------------------------------------------------------------------------------------------------

What is not so clear from the above is that if a person votes for multiple parties above the line, as they now can in a Senate election, then their preferences are simply discarded!  In the bill text see part 12 and the proposed new section 92 (3):

(3) If a voter marks a ballot paper in accordance with subsection (2) and also places other numbers in other group voting squares (but does not indicate preferences for individual candidates in accordance with subsection (4)), the voter will be taken to have recorded their vote in accordance with subsection (2) and all other purported indications of preferences will be disregarded.

Currently, the voter can choose to vote according to a group voting ticket or they can choose to vote below the line, numbering all boxes, to distribute their own preferences.  Under the Bill, the voter will be able to vote for just one party (with their vote then exhausting once all that party's candidates are excluded or elected) or vote below the line as before.  While the new system means the voter's preference can't be gamed by preference harvesting, it's really not much of a choice.

There are quite a few points to be made about this proposed system.

The first is that federal Labor got aboard the Bob Day train in complaining about the potential for exhaust under the new Senate system.  Concerns about high exhaust were often based on spurious comparisons with New South Wales, where preferencing above the line is possible but not specifically encouraged.  The new Senate system directed above-the-line voters to number at least six boxes, leading to a high rate of preferencing and rather low rates of exhaust and just-vote-1s. Now SA Labor comes along with a system that doesn't instruct, encourage or even allow above the line preferencing, but bans it.  (This isn't completely news, by the way; the SA government previously explored introducing the Sainte-Lague method without any preferences at all).  So where is the crusade against exhaustion now?

The second is that the existence of differing above-the-line systems between state and federal parliaments can create issues.  At the 2016 federal election, the just-vote-1 rate in NSW was almost twice that of the next highest state, and while Ray Hadley had something to do with that, there were substantial differences for voters of all parties.  The NSW upper house system makes no attempt to dissuade voters from just voting 1 above the line, and most do.  If South Australia introduces the proposed system it is highly likely that some voters will attempt to preference above the line (thinking that it is like the Senate) and thereby imagine that they are expressing preferences when they are not.  This could cause some voters to unintentionally waste their vote.  It is also likely that over time the proportion of South Australians voting just 1 ATL in the Senate will increase.  (This isn't a massive issue though;  NSW's just-vote-1 rate of 4.7% of all Senate votes is nowhere near being a scandal, and is much less than just about everyone expected.)

The third is that a voter who wants to strongly control their own preferences can do that no more or less easily than they already could.  South Australia has radical savings provisions for its lower house, under which a vote that would otherwise be informal can become a group ticket vote lodged by a party.  But for the upper house its savings provisions for below-the-line voting are almost nonexistent.  A voter can get away with leaving the last square blank, but any repetition or omission makes the vote informal.  An interesting result of this is that in the current SA Legislative Council system, votes don't exhaust - at all (though some vote-values are still lost to fractions).

Voting below the line in the SA Legislative Council is so difficult that at the last election only 4.2% of all voters even attempted it, but among those diehards who did, 7.4% had their votes declared informal.  (Some of these may just have been hopeless cases.  In any system like this some voters will attempt to number the candidates in order within every column, producing a BTL vote with numerous number 1s, for example.)

It might even be argued that some voters will have less control over their preferences under the new system.  Under the existing system some voters may have perused the GVTs and decided there was one that acceptably aligned with their own views.  They may have then considered that voting 1 for that party above the line was better than numbering all boxes below the line, with the risk of casting an informal vote if they made a single mistake.  Under the proposed "voter choice" system the voter can only choose to vote below-the-line (and get it perfectly right or lose their vote), or else they can choose not to preference at all.  It is true that the choice to vote for just one party wasn't present in the old system, but it's not a choice that anyone should cherish, except maybe as an alternative to having their electoral system trashed by group tickets and preference harvesting.

Something else worth noting is an oddly limited savings provision that is contained within the Bill.  If the voter votes 1 below the line for the candidate who is at the head of a party ticket, and marks no other boxes, then their vote will be treated as a 1 ATL vote for the party.  However, if a party has six candidates, and the voter votes 1-6 for those candidates and stops (a vote meaning exactly the same thing as an ATL #1, and which would be saved as formal under the new Senate system), then that vote isn't saved.  Indeed the only BTL votes that will be saved will be those that most strongly differ from the instructions, but a vote that is less divergent and more clearly in line with the solution will not.  That is odd.

(Federally, under the new Senate system, 2.3% of all informal votes, or 0.09% of all votes cast, had just a single 1 below the line, though I do not know what proportion were for the lead candidate of a ticket.  Four and a half times as many votes were rendered informal by omissions or repetitions affecting the first six numbers alone. Again, some of these would have been hopeless cases.)

It still beats GVTs though

As limited as this legislation is in providing "voter choice", it would still be better to pass it than to allow Group Voting Tickets to continue.  Among other reasons, Group Voting Tickets are democratically illegitimate because they allow for the potential election of candidates with no real public support, and because they massively inflate the risk that small errors in the electoral process will cause elections to be voided.  At least under the proposed Bill, parties will need to poll a significant percentage of a quota (a quota being 8.33%) to be in the running.  At most recent elections, about 4% would have been needed had the same votes been cast under the new system.  It's possible in some cases that 2-3% will do it for one of 11 seats in a given year.

Some might say that electoral wonks in Australia are a bit precious about the whole idea of preferencing in upper houses.  Globally, proportional representation systems are common, and generally they don't allow preferences. Our determination to graft preferencing onto PR might be seen as a legacy of our early adoption of preferencing in lower house elections, which was driven by the conservative parties wanting to avoid three-cornered contests.  However, that determination has its advantages.  For one thing, it allows similar parties to compete with each other without both destroying each other's chances - something that could be a problem with the proposed new system.  Australian voters are also used to the idea that you can vote 1 for whoever you like without wasting your vote, because you can preference another party.  They will take some getting used to the idea that a 1 vote for a micro-party is in fact a wasted vote and that they need to either think strategically or number all the boxes.

Looking at recent SA Legislative Council results as they might apply under the new system, the main difference is simply that preference-snowballing is stopped.  On the votes cast, a Liberal would have won instead of Dignity 4 Disability (who polled just 1.2%) in 2010 and the Liberals would have also beaten that obscure No Pokies dude who got up in 1997. As with the Senate reform debate, the issue is not about the quality of MPs who happen to have been elected under the current system; it is about whether they have a real mandate to be there (however good they might be) and the potential for the system to elect more or less any random MP who might be completely terrible.

Some other results might or might not have switched depending on the the patterns of below-the-line preferencing.  As with the Senate system the reform is designed to target preference-harvesters and the parties that most obviously suffer will be preference-harvesters.  As with the Senate system, micro-parties can get merging or go home, and with the ridiculous fragmentation of micro-parties we see in SA as everywhere else (63 candidates on 25 tickets in 2014) that can only be a good thing.

The questions remains - why isn't this bill better?  Why create a new system that, apparently unnecessarily, differs from the Senate system and will therefore cause avoidable confusion?  Smashing group ticket voting is good, but why smash it with the bluntest instrument in the room when we know there are better ways?  Why say it makes it easier for a voter to control their preferences when it doesn't?

Perhaps the answer is a purely practical one, that the increased data entry workload a Senate-style system would involve would be just too demanding for the Electoral Commission SA. However, if that is the problem it doesn't seem to have been ventilated.

The upper house parties in South Australia should explore amendments to at least allow for preferencing above the line, or alternatively allow for much more relaxed below-the-line voting.  Better still, both.  After all, the only even arguably sane point of SA's current shortage of effective below-the-line savings provisions is to prevent exhaust.  But if ATL votes are all votes for just one party, then exhaust will be a massive feature of the new system anyway so below-the-line voting may as well be liberalised.

I will continue to watch with interest. At least SA might be doing something about group ticket voting.  Last state to discard it is a rotten egg!

Further Reading: see also Henry Schlechta's post on this.  And Ben Raue has given the proposal the thumbs down.