Saturday, October 15, 2016

How To Best Use Your Vote In ACT (and Tasmanian) Elections

This is a quick note because there are some incorrect myths doing the rounds about how people should vote in the ACT election today.  I was asked to write this piece because there have been some claims that voters who vote for independents should refuse to preference anyone but independents.  A claim to this effect was recently published in the Canberra Times in an otherwise generally good article.  It happens that the same advice I'm giving here for the ACT also applies to Tasmanian lower house elections, because both use the Hare-Clark system, albeit with some minor differences.

My basic recommendation for both ACT and Tasmanian elections is simple: If you want to make your vote as powerful as possible, number every square.  Don't even leave the final box blank.  If you don't feel you can number all the squares, you should number as many as you can.

This is the key incorrect claim from the Canberra Times article:

"But what if I don't like Labor or Liberal?

In this case you need to get strategic. The five-member electorates will almost certainly return two Labor and two Liberal politicians each, with the final seat in each of the five electorates up for grabs. These are the seats that will decide the election and where the minor parties are pinning their hopes. The reality is their chances are slim; and other than possibly the Greens, none will get a quota - 17 per cent of the vote - in their own right. So to get elected, they must rely on preferences, and a significant flow of them. A voter wanting to elect independents and would do best to confine their vote to the independent columns - and not give preferences, even distant ones, to Labor, Liberal or the Greens.

To see how this works, imagine Independent A gets 300 first-preference votes, the lowest of any in the seat. That person is eliminated and all of their votes redistributed to the candidate those voters numbered 2 on their list. If No 2 is from Labor, the Liberals or the Greens, the vote will head into those columns and likely stay there as more of the low-polling independents are eliminated. Independent B is likely eliminated next, with his or her No 2 votes distributed. You can see that if your vote stays entirely with independents and minor parties, and if there are enough people doing the same thing, an independent can stay in the count long enough to start getting preferences from the lowest-polling major party candidates as they are eliminated. This is a strategy to maximise the chances of independents, although given the magnitude of the vote required – about 8000 votes - it is still a tough ask."

The part in bold is incorrect.  A voter who prefers all the independents and minor parties over the big parties should indeed preference all the independents and minor parties at the top of their vote.  But after that, if they prefer one of the bigger parties to one of the others, they can and should safely preference the larger party that they prefer.

Why? Because provided such a voter preferences all independents and minor parties above the major parties, their vote can never reach any major party candidate until every independent and minor party candidate has been elected or excluded.

As I wrote about Senate elections:

To make best use of your vote, you should only stop when one of the following happens:

1. You could not care less which of the remaining candidates wins (assuming that at least one is elected).
2. You so strongly dislike all the remaining candidates that you feel morally opposed to even helping them beat each other.
3. Although you actually dislike one of the remaining parties less than one or more of the others, you want to exhaust your vote in protest to encourage that party to listen to your concerns.  (To make your point effectively, I suggest you send that party a letter after the election telling them you did this, since they won't be able to work it out from your vote.)

Indeed, much the same advice in general applies for Hare-Clark elections as applies for Senate elections.  But there is one extra twist.  In both the ACT and Tasmania, whether you number every square or just every square bar one cannot affect the result of the election.  Your vote will never reach your most disliked candidate while they are fighting for a seat against anyone else.  However - and perversely - putting a candidate last rather than leaving the box blank can in some cases help get your vote into the recount for that candidate's seat if that candidate gets elected then later dies or resigns.

I expect to have some, possibly sparse, live comments on the count for the ACT tonight, piggybacking off the ABC coverage and adding what insights I can based on my experience of Hare-Clark.  I don't have any forecasts for the election as in the total absence of polling I have nothing even remotely rigorous to base them on.

(For anyone interested in strategic voting, the comments in the Senate piece generally apply to this one too.)


  1. In either the ACT or Tasmania, have the relative preferences of Greens or Labor voters determined which of the Liberal candidates gets up? Or have Liberal preferences determined which Labor candidate gets up?

    1. Not aware of a case of a Liberal winning by overhauling another Liberal using Green or Labor preferences, but suspect it has happened. For preferences coming from Liberals deciding a Labor winner, yes that has happened - in Braddon 1989 Michael Weldon (ALP) defeated Greg Peart (ALP) by seven votes. Weldon received 16 more votes from Liberal preferences than Peart, so would have lost had votes received from the Liberals split evenly.