Saturday, 6 April 2013

Australian Elections and Me - A Love-Hate Relationship

In the immortal words of Eddie Izzard "So... Yeah..."

There hasn't been much to fire up my imagination this week in the realm of electoral blogging. Everyone has been distracted by North Korea's latest Kim-Jong-Un-ism.

There are a couple of pet topics I've been saving up as fillers for just such an occasion. The first is economics and the second is the voting system. Both deserve a little more statistical analysis than I have been able to afford this week. So what you get is a disappointingly/mercifully short article on the latter that will be backed up by a glorious/tedious statistical supplement/page of indecipherable graphs.

How Voting Works - Lower House

For those who never bothered to learn how your vote is counted, this section is probably not of any interest to you. For everyone else this is old news rehashed as blog filler. But I'm writing this anyway because this is my blog and I call the shots around here.

When you go to your local voting booth you will receive a small green form and a big white form. The green form is the lower house form and is your opportunity to vote for the seat for your member. You fill in the boxes next to each candidates name with ascending numbers, where 1 is the person you dislike the least and the highest number is the one you dislike the most.

These votes get counted and assigned to the candidates ranked 1 on each ballot:

In the Australian electoral system, the winner is not simply decided based on who gets the most primary votes. Let us imagine this election is largely focused on the controversial new chicken wings policy. The first three parties above are anti-chicken wing, and the final party is pro-chicken wing. While the majority of voters (68%) voted anti-chicken wing, the lone pro-chicken wing party has come out on top because their opposition's vote is divided.

Instead the lowest ranking party has its votes redistributed according to their second preferences (i.e. go to the candidate with a 2 next to their name).

The process is repeated until only two candidates remain. This means all of the red votes are redistributed according to their second preference, and all the yellow votes that flowed on to the red stack go to their third preference.

The result is close, but the X-Wing Party comes out ahead and claims the seat.

Because certain parties have similar policies to other parties, you can often predict where the majority of votes are going to flow. This is increased by 'how to vote' cards that encourage voters to preference some parties over others. For example, the Liberal and National parties normally preference each other on 'how to vote' cards and their similar positions on many issues mean that the majority of redistributed Liberal votes go to the Nationals and vice versa.

This allows us to make more accurate predictions from polling data and ensures that every vote counts, so that the person who wins is not just the one with the most first-place votes - they also need relatively few last-place votes.

How Voting Works - Upper House

The white ballot elects senators for your state or territory. It is larger than the green ballot, but there are three ways to fill it out.

Firstly, you can fill in all the boxes below the line, which then works much the same way as the green ballot. Secondly, you can fill in one box above the line. This saves you considerable research time, because your second and later preferences follow a pre-determined path set by your first choice. Of course, it pays to know what this path is before voting, so that you can make an informed choice and not have your second preference going to your least favourite party.

The third option is to do both. It is only marginally longer than filling in every box below the line, but it means if you make a mistake below the line and your vote would otherwise become invalid, they revert to your above-the-line preference flow pathway.

The advantage of voting above the line is chiefly one of speed and simplicity. It does meant, however, that you vote for a party rather than for a particular candidate. Once the nominees are finalized in August this blog will provide links to your senate nominees preference flow paths, so you know exactly where your above the line vote will go.

The advantage of voting below the line is the control you have over your vote. This blog will also contain as many resources as I can muster in order to allow you to make an informed choice as to who you place 43rd and who you place 44th. Another advantage is that you can support your favourite party while snubbing that annoying, out-spoken senator and ranking her or him way down the list.

The advantage of voting above and below the line is obvious - it is just the same as voting below with a backup in case you mess the form up. Of course you can always just ask for another form if you spot the error yourself, or do what I do and take in an A4 copy of the draft white ballot already filled in and tripple checked. Draft white ballots will be available on this blog, or just on the other side of a link, later this year.

Compulsory Voting, O How I Loathe Thee.

Compulsory voting does several good things. It forces the AEC to make sure ballots are available to all Australians, even if they are outside of their seat, their state or even the country on polling day. In the U.S., by contrast, many people did not get the opportunity to vote in the 2012 Presidential election because they were outside their states helping to deal with the Hurricane that smashed much of the East Coast, and there was no real system to deal with people from some states being outside of that state.

Compulsory voting also ensures no one is prevented from voting by work commitments, pressure groups or the like since voting is required and secret. Our electoral commission does a fantastic job setting up booths, maintaining the checks and balances and giving us a level of secrecy almost unparalleled in elections anywhere else in the world. It also forces people like me to get out of the house one Saturday every few years. These are all positives, although the extent of Australians denied their vote while travelling, or the rate of election-day kidnappings without these measures might not be all that high in a voluntary voting system anyway.

The main gripe I have with compulsory voting is that it encourages the politically uninterested to sway the election results. In the U.K. the average voter turnout since World War Two is just under 74%. In Canada it is just over 65%. Voter turnout in the U.S. only exceeded 60% five times in the last century. Click here to check up on many other countries.

Even if we assume only a quarter of Australians don't care about voting, it is important to remember that in Australian psephology "very safe" refers to seats with a margin of 20% or more. Only a few seats are ever won by more than 25% of the voting public. In theory, the "I don't care" vote could make safe seats for one party become safe for another.

In reality the votes are more evenly divided, but not so evenly as to make no difference. A "Donkey Vote" is probably the most simple means of filling in your ballot, normally filling the boxes 1, 2, 3, 4... from top to bottom. Because this is a valid vote it is difficult to determine how prevalent this is (although I hope to try and get some estimates on this in a later post), but it does give the first name on the ballot a boost.

Also, because of the redistribution system, a small change in a close run fourth-place contest could drastically affect the flow of preferences and determine the outcome. A minor candidate might, for example, advise her or his supporters to preference the Nationals, then the Labor party, then the Liberals. In some cases this could be enough to overtake the Labor Party, result in a Lib vs Nat contest and give the seat to the Coalition. If the donkey voters and "I don't care" participants manage to nudge the minor candidate above the Nationals, the Nats will drop out and the majority of their votes would presumably flow to the Libs or (to a lesser extent) the ALP. The minor candidate would fall next, and the majority of those votes would go to the Labor party too, possibly taking the seat out of Coalition hands.

Of course many people forced to vote still exhibit some independent preference between parites which, because of the enduring Labor vs Liberal dichotomy, normally sees votes for the two major parties swell while minor parties and independents stay relatively unaided.

Now I am not suggesting there should be some kind of minimum standard of research imposed on voters; anyone who wants to vote should be allowed (excluding, of course, those too young to understand the issues). And I am not advocating that anyone submit an informal ballot, partly because this just clogs up the system and partly because I once read that it was illegal to advocate invalid voting though I cannot find anything to back that up now.

There is an argument that a government elected by everyone is somehow more democratic. Ignoring the fact that compulsory voting drowning out the votes of people who actually give an expletive arguably makes such a government less democratic, it also gives less credibility to any government's mandate to rule. In voluntary voting systems there is a simple rule of thumb - if you didn't vote then it is inconsistent to complain about the government. You had your chance to cast an opinion and didn't. In Australia it is easy to shrug off any responsibility for your vote since you had to vote for someone. Peter Singer, being the controversial thinker he is, argues that this rejection of responsibility for the government could lead to a rejection of the law altogether. I don't fully accept that myself, but I do think that forcing people to vote leads to many people feeling disconnected from the government. When people are forced to support some party or another there is no choice to actively participate in elections and less sense of getting behind a government.

There is another argument that compulsory voting leads governments to consider the entire voting public in their policies, since being too harsh on any one  sector will backfire when they inevitably vote. On the contrary, I would argue that voluntary voting will lead to wider consideration of the public, since any policy that negatively affects a group is liable to turn any non-voters in that group into voters for an opposing party - effectively recruiting for the opposition

It is probably true that less time and money will be spent by candidates, since they simply need to convert wavering voters to support them over any alternative, without enthusing them to actually turn up in the first place. Of course, if they had worthwhile policies the public would normally be enthused anyhow, except for people who just don't care. Trying to get these people to vote would probably be a waste of resources anyhow, and they would only skew the results away from the will of the people who actually care.

The AEC also suggests that "[v]oting is a civic duty comparable to other duties citizens perform e.g. taxation, compulsory education, jury duty".

I'm going to skip right on past the bit where things like education are listed as "duties", and focus on the core claim that we somehow owe our societies our vote. Beyond my legal obligations I don't consider my vote to be "required". If I genuinely did not care one way or the other (or the other (or the other)) this election, I wouldn't want my randomly allocated vote to skew the result. I don't see how that benefits my local or national community in the least. If anything, I'm doing a disservice.

Then, of course, there is that incredibly patronising line of argument that says that compulsory voting is educational and teaches voters about the political system. Its as though anyone who would opt not to vote must be as ill informed about our political system than a third-grade student. Don't forget to make sure your parent(s)/guardian(s) sign the consent form before you go on the field trip to the polling booth, children!

It is actually possible to be intelligent and politically aware and still not care to support a political party (or independent) in a given election. In fact some would probably say that those who follow politics the closest are the most likely to be fed up with the whole business.

All I am really advocating is a box on the ballot that says "none of the above", or "frankly, my dear, I don't give a damn."

In The Land of Danistan

I think we know that many things would be better if I ruled the world. Equal rights, free tertiary education and compulsory beards would be well established. And, of course, we wouldn't need elections, since I would be the obvious choice for supreme and glorious overlord. But if we were going to hold elections, just for the fun of it and to give me something to blog about, this would be my system:

Within the month between the issuing of the writs and the election anyone would be able to attend a post office or other registered place with valid photographic identification and either enter an early vote or announce their wish to not vote at all.

Anyone who has not announced their wish to not vote in advance will still be expected to turn up at a polling booth on election day. There is another little gripe I have with our system where you can dislike two parties equally, yet be forced to prefer one over the other. A 1,2 vote for the X-Wing and Chicken Wing parties respectively could indicate you like both parties a lot, or that you hate both but not as much as the others, or that the X-Wing party is your idea of perfect government while the Chicken Wing party lags well behind.

My ballots would look something like this:

The voter places a mark next to some or all of the candidates, in boxes marked from 3 (strongly approve) to -3 (strongly disapprove). The totals for each candidate are totalled and the highest-scoring candidate wins. There would also be an option for a little perforated circle in the corner. If you fail to punch it out the vote is invalid. This means it is possible to submit a ballot that you have marked but still choose not to vote. It also means that incorrectly filled out ballots can be returned and destroyed while still being able to prove they were never valid votes. Because the vote is validated by the removal of the circle, rather than invalidated, it is not possible to invalidate valid votes.

In this system there are, then, three ways to not vote: declare that you do not wish to vote, take a card and do not fill it in (or only mark the 0 circles for each candidate), or (optionally, if the system is deemed to require it,) do not punch out the validation circle.

But until the glorious revolution installs me as a benevolent yet totalitarian dictator, we have compulsory voting. So, I guess, enjoy your legally enforced right to choose.


  1. Remind me that I could never be a citizen of Danistan, I look terrible in a beard, and so do most women.
    I'm not convinced by your voting system, oh overlord, for the following reasons.

    1. Assume that the RWP and the LWP get exactly the same number of positive votes, and vote -3 for the opposing candidates. This would cancel each vote out and essentially elect, to grossly oversimplify, the party that people hate the least, which would most likely be just because we don't know anything about them, rather than because we approve of them. The Shooters and Fishers and the Sex party would rule the roost for a term, until they were replaced by the DLP and the Socialists, and we end up with a parliament of inoffensive, generally inept politicians, like...Tasmania.

    2. I'd make it so that you had to punch out the bottom to NOT vote, rather than the opposite. It's hard enough to vote properly for a large percentage of people as it is, so why not include people's vote by default and allow those that care (or can read) the ability to opt out?

    1. To reply to your points in reverse order, the advantage of the press-out circle is that it cannot be reversed, making it much harder for officials to manipulate.

      This would not be a great concern in Australia; I cannot overstate what a fantastic job the AEC does in setting up, maintaining and organising supervision for the voting booths around the country, let alone keeping the entire system honest and transparent. Since Danistan is an example separated from any context it is harder to comment on the need for anti-corruption measures. My gut feeling is that if I were in power, you would want as many checks against manipulation as I allowed you to keep. I will say, however, that if such an invalidation mark were to be added to Australian ballots, a simple check box would suffice and validation should be opt out as you propose.

      I personally think the addition of such a box to our ballots would be a positive step. Casting an invalid vote currently has an air of underhandedness to it. Further, depending on how you wish to invalidate the ballot, it can be as complex as voting correctly and is certainly less clearly advocated. There are many 'how to vote' cards but no 'how to not vote' cards. An opt-out checkbox would legitimise invalid voting as the alternative that pro-compulsory voting advocates insist it is, and allow people to easily and confidently not-vote.

      I would expect this to have greater implication for early voting, since it is more conceivable that people would take the option of not-voting early. I doubt very many people currently arrange early voting with the intention of not-voting - they would be more likely to not-vote in person or not attend at all.

    2. As to your first point, you are quite right that minor parties would take a larger portion of the vote. However, any diehard supporter of the RWP or LWP would down-vote all opposition so heavily that it would presumably rule the really low profile candidates out of contention (i.e. the ones who get a few hundred votes from their local community and not much else). The parties that thrive will be those that get the most supporters and least opponents. This does not mean ineffective, non-committing saccharine in place of policies, merely that the policies must be broadly supported, and that support must out-weigh opposition. For example, polling over the last few years has repeatedly resulted in support for the recognition of same-sex marriages at 60% to 70%, although the Coalition still openly opposes the concept, and the ALP only began moving in this direction in 2011 by allowing the matter to be resolved by a conscience vote.

      Without wishing to sum up the opinions of voters this far out from the election, let us suppose that the electorate is by-and-large fed up with both Labor and Liberal. It will come as no surprise under the current system that the next Prime Minister will be either Labor or Liberal, regardless. In this scenario under the proposed system, however, both parties would end up in the negatives. I don't necessarily agree that parties will end up positive simply by being less well known, since this would still require the people who have heard of them and support them to outnumber those who oppose them AND those who down-vote all but one other party.

      However, minor parties could certainly end up less negative through keeping a low profile. That is, however, what the public would be supporting by ranking them at '0' and above the major parties. However I believe that before we reached the scenario you suggested we would see a major shift in rhetoric away from decisive, race-to-the-bottom politics towards painting positive visions fir the future. Parties would no longer be fighting to win votes from each other, but rather to keep their own approvals high independent of other parties' ratings.

      It is hard to judge which parties would do well, since current polling is based on our present electoral system, and I think minor parties would get more coverage - or at least be more broadly criticized by major parties and others - under a different system.

      An alternative system could rank all candidates from -1 to 2, thus ensuring major parties are more likely to get positive momentum in the long run, whilst still enabling voters to quantify their support for, or opposition to, each candidate.

      The precise outcome of either system can only really be judged by a trial, which I would be interested to conduct if I had the resources. I doubt it will be happening any time soon, though.