Hacker News Reader

Comment by bedobi | Hacker News Reader

Australians love to hype their voting system, not knowing or caring to know it too is fraught with problems, just like this proposed one probably is.

That's not to say the US or UK or what have you systems are better or worse. All voting systems kind of suck and suffer from mathematical and political paradoxes.

View on HN

As a Brit that now lives in Australia, from what I can tell the voting is just better here. With preferential voting you (hopefully) end up with more of a consensus for the winner, rather than purely a representative of the largest minority as with FPTP.

It also makes it easier for people to vote for trash parties like One Nation. In other countries people would think twice before voting for small, extremist parties like that lest they waste their vote if the party doesn't make it past the threshold for getting a seat. (this is just one of the many problems I see with the Australian system)

"Stop voting for the wrong people!" is your problem?

Yes! I don't think liberal democracies should tolerate parties and policies that dismantle them. In many European countries people like Pauline Hanson would have been put jail (where she belongs) for hate speech (and probably any number of other things) a long time ago.

The irony being, that FPTP has resulted in Trump which is arguably not much better than One Nation.

Trump has far more to do with the electoral college and other quirks of the US system than FPTP

Don't we want voters to be able to just honestly write that they like these small extremist parties? Are you suggesting that we want all the voters to vote for one of the top two parties, out of a fear of wasting their vote?[0] Besides, those small parties are obviously eliminated almost immediately, so the only thing those votes are worth is a bit of money from the Australian Electoral Commission.[1]

[0]: https://en.wikipedia.org/wiki/Duverger's_law

[1]: https://en.wikipedia.org/wiki/Australian_Electoral_Commissio...

I think the cost for voting for minority extremist parties and policies should be higher, not lower, yes.

That's government by the majority, not the whole. Democracy is “a system of government by the whole population or all the eligible members of a state” (OED; emphasis added).

Just because you think the cost of voting for small extremist parties should be high doesn't mean you're in favor of absolutist majority rule?

+ liberal democracies have (at least in theory) unalienable basic rights that can't be taken away, even by majority. Eg just because a 90+% majority votes to revoke citizenship from all muslims that doesn't mean that it will go through.

The effect of eg. FPTP is to effectively disenfranchise large portions of the population. A country with a voting system like that makes a mockery of the very term democracy.

As for these "inalienable" right, I'm not aware of any system without a mechanism for altering their constitutional laws.

To your specific example, the UK just had a government representing a minority of voters vote through a law that allows them to strip citizenship from millions as they please.

Who determines what is extremist and what is small? It’s not like the dominant parties dropped from the sky with their existing “bigness”. They too started out as small parties.

> Who determines what is extremist

we're not talking about some ambiguous grey zone stuff here

* proposed a drastic reduction in immigration with particular reference to immigrants from Asia.

* stated that African immigrants were bringing diseases into Australia and were of "no benefit to this country whatsoever"

* called for a ban on Muslim immigration to Australia

* announced policies including installing CCTV cameras in all existing mosques

* stated "all terrorist attacks in this country have been by Muslims"

it's safe to call the above extremist and xenophobic

But you're not solving that problem by pretending they have less support than they do.

A huge danger of systems that downplay the support for extremist parties is that they create an incentive for infiltrating the bigger parties instead.

In a single-winner election, why would you think that? The "minority extremist parties" are going to lose anyway, so why not try to let people vote more honestly, so that they don't need to strategize? For example, I don't want people who naively vote Green to end up wasting their vote just because they don't know how the voting system works.

Personally I think that makes it a better system.

I may not agree with the politics of One Nation (in fact I'm almost certain I don't, without even going to read what they are) but having a system where I can place a vote for what I really want and then put an acceptable compromise candidate as second choice seems to me a much better way of allowing newer parties and newer politics in than the more or less institutionalised stagnation of something like the US system.

From what I can tell the Australian system rarely lets such smaller parties gain very much anyway.

For all that we have instant-runoff preferential voting, we still have single-member districts for our House of Representatives (and 150 members of said House). We also have compulsory voting and (at the federal level, and in most states), a full-preferential ballot.

When you have 150 single-member districts you're probably going to get two major parties regardless of whether you have FPTP or IRV or even Approval voting. What compulsory voting and preferencing add on top of that is both of Australia's majors have to win the centre, rather than energising their base per the US.

What our system does allow for is for smaller parties in our Senate. It's generally elected 6 people at a time, allowing for e.g. 1 Green, 2 Labor, 2 Liberal-National and 1 One Nation (this is a very plausible result for Queensland next year).

So what does the US system allow then? Well, every system has to let its steam out somewhere. The US has its primaries, which often has a level of access and openness that greatly surpasses any Australian preselection (generally still getting out of the smoke-filled-room era).

The primaries still only really allow people who have climbed the ranks in an existing party to get a shot, AFAICT. But I'm no expert.

I like the idea that the main parties have to capture the centre, and I really like that I can vote for something a bit out there, without the fear that my vote is effectively a vote for my least favourite big incumbent. The "You must vote Labour or the Tories win!" thing has always rubbed me up the wrong way, as someone who isn't especially fond of either of those options.

> That's not to say the US or UK or what have you systems are better or worse. All voting systems kind of suck and suffer from mathematical and political paradoxes.

The argument that everything has >0 flaws, therefore everything is the same, is a bit bizarre to find on HN. Is every solution in every category the same?

> just like this proposed one probably is

Even when we don't know what it is?

the point is, Australians in particular are very very quick to hype their system

and they have a fairly good understanding of the alleged benefits of their system

and they get told this over and over throughout their schooling and by the electoral commission

but not once have I ever seen any Australian ask themselves if the alleged benefits of their systems are really true in the first place, and even if they are, what the inescapable downsides of their system are

it doesn't even occur to them, the question doesn't even make sense to them

as for this quadratic one,

try the online demo they made at

https://www.economist.com/interactives/2021/12/18/quadratic-...

it's cool but tbh I would rather have infinite points to spend, at which point you might as well have binary endorse / oppose votes (maybe with an added option abstain)

point being that I don't think the intensity of my endorsement of policies to combat climate change should cost anything for my equally intense endorsement of paying reparations for colonialism - I consider both incredibly important and essential policies

Correct link: https://www.economist.com/interactives/2021/12/18/quadratic-...

> All voting systems kind of suck and suffer from mathematical and political paradoxes.

Yes. Arrow's Impossibility Theorem.

https://en.wikipedia.org/wiki/Arrow%27s_impossibility_theore...

That said, some systems are clearly worse than others.

> Arrow's impossibility theorem

Ah, that's it, thanks, I was looking for it but forgot what it was called. There's also https://en.wikipedia.org/wiki/Condorcet_paradox (and any number of other ones that may or may not even have names)

Arrows impossibility theorem only proves things about voting systems where the method of voting is simply ranking the candidates in an order, so it doesn't actually apply to this quadratic voting system.

Does anyone happen to know if an analogous theorem does apply to this quadratic voting system?

Range voting and approval voting aren't subject to anything similar, according to William Poundstone's book Gaming the Vote. If that's correct then I'd think quadratic voting is the same.

ATW, score voting (which some call range voting) and approval voting both suffer from the defect that a choice that is favored by an absolute majority (which would thus win in a "normal" election) can wind up losing.

That wouldn't have to happen very often to lead to a lot of very angry voters.

Australia’s system (IRV) doesn’t necessarily elect the Condorcet winner, either, and it hasn’t led to mass revolts.

Well, you have to be aware of the potential quirks in the system, and Australians are not, they uncritically know their system is infallible and the best.

How can approval voting suffer from this defect? Is the idea here that we factor in people who don’t vote for things they actually favor because they’re hoping to avoid tipping the result away from something they favor even more? Or do you mean an election could end up selecting a result that is favored by the majority even though there is a different result that is more favored by a smaller majority?

https://en.wikipedia.org/wiki/Approval_voting#Effect_on_elec...

This isn't a defect, it's a benefit. It's mathematically proven that an electorate can prefer X even if a majority of its members prefer Y.

Commonly proposed "majoritarian" methods using rankings can, for instance, elect X even though Y is preferred to X by a huge majority of voters and Y has twice as many first-place votes as X. http://scorevoting.net/CoreSuppPocket

And there's not even a clear mathematical definition of "majority winner". https://web.archive.org/web/20190219005032/https://sites.goo...

The goal of voting methods is to produce the most satisfying results possible (highest expected utility). And score/approval voting are excellent at that, and also radically simpler than ranked methods.

https://electionscience.org/library/approval-voting-versus-i...

https://rpubs.com/Jameson-Quinn/VSE5key

https://medium.com/election-science/star-voting-is-simpler-t...

etc.

In practice, approval voting and score voting will almost always elect the Condorcet winner. Possible even more often than real Condorcet methods.

http://www.rangevoting.org/AppCW

Try the online demo they made at

https://www.economist.com/interactives/2021/12/18/quadratic-...

it's cool but tbh I would rather have infinite points to spend, at which point you might as well have binary endorse / oppose votes (maybe with an added option abstain)

Quadratic voting is an extremely irrational idea.

https://www.rangevoting.org/MonetizedRV

putting science in political science.

Arrow's theorem only really applies to social welfare functions not voting methods. You use a utilitarian social welfare function when you calculate Bayesian regret for a voting method.

https://www.rangevoting.org/ArrowThm

There’s no need for fatalism, especially when you admit you haven’t analyzed the protocol’s failure modes.

Reply on news.ycombinator.com