March 01, 2004

Does Australia exist?

Posted by John Quiggin

Eric Maskin and Partha Dasgupta are smart guys, and its hard to believe they are totally ignorant of what happens in the Southern Hemisphere. So how can they justify writing a piece promoting a system of rank-order voting as superior to the existing American (plurality) and French (top-two runoff) systems, without mentioning that Australia has had this system (in a range of variants) for many decades.1

A minor side point is that, in addition to having the world’s most complicated voting systems, Australia also has compulsory voting.2 Typically more than 95 per cent of votes are formal, that is, list all candidates in order of preference, with no missing numbers or repetitions. In Dennis Mueller’s generally excellent book on Public Choice, he discusses the single transferable vote and suggests that, while attractive in theory, it’s too complicated to work in practice. Either Australians are a lot smarter than everybody else, or public choice theorists aren’t as smart as they think they are.

1 To be precise, Maskin and Dasgupta advocate the Borda weighted vote, whereas Australia has the single transferable vote (called preferential voting in Australia), but nothing in their argument distingushes the two.

2 More precisely, compulsory registration and attendance at the polling station - there’s nothing to stop you casting a blank ballot.

Posted on March 1, 2004 12:54 PM UTC
Comments

Well called John. They’re as dumb as dogshit. OK, we’re only 20 million folks, but surely they must know that our PM is known throughout the contemporary world as The Man of Steel, and that we’re fully paid up members of the Coalition of the Willing. What if Dolly Downer hears of this? There’ll be hell to pay. Inexcusable on every count.

Posted by cs · March 1, 2004 01:23 PM

I hate to break this to you, but the relationship between the rest of the world and Australia is pretty similar to the relationship between Australia and New Zealand.

Posted by dsquared · March 1, 2004 01:33 PM

Has anyone noticed a certain desperation to be noticed in the voices of Australians ever since the mighty England flayed them on the rugby field?

Posted by jdsm · March 1, 2004 01:52 PM

Overuse of the word “noticed” - sorry.

Posted by jdsm · March 1, 2004 01:53 PM

What? Being a member of the Coalition counts for nothing? I’m aghast! Wait, there’s more. Have I told you about our deputy sherrif job …

Posted by cs · March 1, 2004 01:58 PM

Oh? We forgot about that world cup thing as soon as we won the Davis Cup the next weekend.

Posted by Graham · March 1, 2004 02:33 PM

As I recall, casting a blank or otherwise ‘spoiled’ ballot is illegal. Utterly unenforceable, of course— but you’ve then got a criminal act which others can be, and sometimes are, punished for advocating.

Posted by Jacob T. Levy · March 1, 2004 02:49 PM

So how does this compulsory vote thing work, anyway? Is there an election mutawa that hunts down non-voters and drags them to the polls?

Posted by Andrew Case · March 1, 2004 02:53 PM

Jacob, the law is a little unclear on whether it is illegal to cast an informal vote, but it doesn’t appear to be: sect. 245 - Compulsory Voting.

What is illegal is to distribute untrue information about voting, which is what Albert Langer was (quite improperly) jailed for back in 1996.

Andrew, no one drags you to the polls, you just get a $20 fine if you don’t turn up and can’t come up with a good excuse why you didn’t.

Posted by Brian Weatherson · March 1, 2004 03:18 PM

And what about Ireland? Since the foundation of the state we’ve been using the STV.

Posted by Keith Gaughan · March 1, 2004 03:21 PM

I am probably not as well versed in public choice theory as I should be. Could someone point out why STV would be preferable to actually using the Condorcet method? The ballots will look the same; is there some criteria against which Condorcet pairwise counting fails miserably? (I don’t think cyclic ambiguities count, since STV has to deal with them somehow too.)

Posted by Jack Lecou · March 1, 2004 06:00 PM

They’re smart guys, but they’re also high-powered theorists, so I’m surprised their paper even mentions that one system is used in America and another in France. Not that there’s anything wrong with that, just you shouldn’t expect Eric Maskin (whom I respect greatly — he was an economics prof of mine about five years ago) to be the one to tie his recommendations to the real world and know what goes on in what country.

Posted by Sasha Volokh · March 1, 2004 06:18 PM

Keith is right. Not only do we share Ireland’s voting system (not sure who came first on this one), but we copied Ireland’s version of the VAT/GST, with fresh food exempted.

And of course the only international football competition that matters is Australia v Ireland (mixture of Australian and Gaelic rules).

Posted by John Quiggin · March 1, 2004 07:54 PM

Yeah, don’t forget Ireland. And of course we are notoriously stupid, so if the public choice people think STV is too complicated, and Irish people do not, well …

Posted by Kieran Healy · March 1, 2004 08:47 PM

jack lecou: thanks for the comment on Condorcet methods. I’ve been wondering that myself.

For more information on voting systems, see

http://en.wikipedia.org/wiki/Voting_system

http://en.wikipedia.org/wiki/Condorcet_method

IMO, STV is functionally the equivalent of the caucus system, without the hassle of showing up at an inconvenient time. But it takes away the only part of the caucus system that makes it fun, the art of persuading voters for losing candidates to shift their preferences to your alternative.

The other aspect of alternative voting systems that nobody talks about is how to winnow down candidate lists to a number (5 or 6, say?) that is manageable for the voter. In Washington State, the whole open primary system is in question because individual parties refuse to honor the result of an open primary election, based on the right of free association.

If the object is to preserve the best-choice centrist compromise candidate into the final election, some solution must be found that is palatable to both mainstream and alternative political parties.

Posted by shooting_star · March 1, 2004 09:08 PM

This site convinced me that STV is a poor method.

Posted by novalis · March 1, 2004 09:11 PM

Jack, to be implementable, Condorcet must be modified to give a winner in every case, and the comparison depends on which version you want to use.

The main divergence between Condorcet and STV arises because to win in STV you have to avoid early elimination, which normally means that you need to have a significant number of first-preference votes. In a system with two large voting blocs, this tends to work against centrist candidates. Iin this respect STV is closer to plurality voting in its properties than is Condorcet.

The merits or otherwise of this are debatable. Arguably, when STV is used in a constituency system, it will come closer to proportional representation than will Condorcet, which will generally over-represent centrists.

Posted by John Quiggin · March 1, 2004 09:13 PM

Daniel, if the relationship between the rest of the world and Oz was the same as between Oz and NZ, Dasgupta and Maskin would have claimed the credit for the whole idea of the STV.

It’s a well-known rule of Australian journalism that any creditable accomplishment by a New Zealander becomes attributable to Australia on the completion of one day’s Australian residence by said Kiwi.

BTW, didn’t we do well at the Oscars?

Posted by John Quiggin · March 1, 2004 10:44 PM

“surely they must know that our PM is known throughout the contemporary world as The Man of Steel”

Y’all elected Stalin Prime Minister?

Posted by rea · March 1, 2004 11:09 PM

Donald Saari has a convincing (to me) argument for the Borda count, including an argument that even a unique Condercet winner should not necessarily be elected.

He uses the 3-candidate example that Condercet himself exhibited:

30 votes A>B>C
10 votes B>C>A
10 votes C>A>B
1 vote C>B>A
29 votes B>A>C
1 vote A>C>B

A is the unique Condercet winner, while B wins the Borda count. Now consider this subset of ballots:

10 votes A>B>C
10 votes B>C>A
10 votes C>A>B
1 vote C>B>A
1 vote B>A>C
1 vote A>C>B

These ballots are completely symmetrical with respect to the three candidates and so clearly define a three-way tie: in a very natural sense the ballots in this subset “cancel out” each other, and what you are left with is

20 votes A>B>C
28 votes B>A>C

So B should win, as the Borda count suggests. Saari proves some theorems that say, roughly, that the Borda count is the only system that always respects such natural cancellations among ballots defining ties.

Posted by Daniel Lam · March 2, 2004 12:04 AM

A big problem with Borda is that the addition of a fourth candidate, D, ranked last by all voters, will change the relative value of higher preference votes.

Posted by John Quiggin · March 2, 2004 12:31 AM

I retract the comment immediately above. An extra candidate of the type described will just add one to the Borda value of all votes for the others.

Posted by John Quiggin · March 2, 2004 12:58 AM

But if D is ranked dead last by all voters, then the points awarded to A, B, and C under Borda will all go up by the same amount, namely the total number of voters. So the winner remains the same.

On the other hand, an additional candidate D can change the winner without winning itself. For example, if all the A>B>C voters all changed to A>D>B>C while everyone else puts D last, A will become the Borda winner instead of B. So Borda doesn’t satisfy the Independence of Irrelevant Alternatives (IIA).

Saari proves that if the notion of IIA is modified to take into account the number of other candidates ranked between a given pair of candidates, rather than just the binary question of which one is preferred to the other, then Borda satisfies this modified IIA, together with all of Arrow’s other conditions.

Posted by Daniel Lam · March 2, 2004 01:11 AM

It’s a well-known rule of Australian journalism that any creditable accomplishment by a New Zealander becomes attributable to Australia on the completion of one day’s Australian residence by said Kiwi.

Residence? They don’t even have to be conceived there!
http://www.theaustralian.news.com.au/common/story_page/0,5744,8842284%255E601,00.html

“Australia may well lay claim to an acting Oscar, too, as Renee Zellweger, who won best supporting actress for her role in Cold Mountain, revealed afterward that her parents met in Sydney. Her father, Emil, was born in Switzerland but was sent to Australia by his family during World War II, and became a surf lifesaver in Cronulla. He met Zellweger’s mother, a Norwegian nurse, Kjellfrid, in Sydney and later followed her to Texas when she found a job there.”

Posted by Errol · March 2, 2004 03:28 AM

lol, Errol! You couldn’t make it up if you tried, could you?

How about “Although Charlize Theron has never actually been to this country, her childhood liking for Captain Kangaroo qualifies her as yet another Australian winner”.

Posted by John Quiggin · March 2, 2004 04:08 AM

20 bucks!!??? A few months ago i got a fine for $150 in the mail for an election i didn’t even think i’d missed. But still, compulsory voting rules. When else would you get to see all the halls of our country? Be they church, school, scout, norwegian club or boy’s brigade?

Posted by Ads · March 4, 2004 01:21 AM
Followups

→ The War for the War II.
Excerpt: But this is really immature! First: What serious intellect (see the consensus here - you can ignore Christine Chinkin, who has always been a simple-minded literalist in international law, as for that Bournemouth chap . . . ) takes international...Read more at Full Context
→ The War for the War II.
Excerpt: But this is really immature! First: What serious intellect (see the consensus here - you can ignore Christine Chinkin, who has always been a simple-minded literalist in international law, as for that Bournemouth chap . . . ) takes international...Read more at Full Context
→ The War for the War II.
Excerpt: But this is really immature! First: What serious intellect (see the consensus here - you can ignore Christine Chinkin, who has always been a literalist in international law, as for that Bournemouth chap . . . ) takes international law...Read more at Full Context

This discussion has been closed. Thanks to everyone who contributed.