Hacker News Reader

Comment by Animats | Hacker News Reader

Not that it matters this time, there being no third party candidates of note in Maine. If Florida had ranked voting in 2000, Al Gore would be President.

Florida 2000:

    Bush      2,912,790  
    Gore      2,912,253  
    Nader        97,488  
    Buchanan     17,484  
    Browne       16,415
    Others      < 6,000

View on HN

s/If Florida had ranked voting in 2000/& and the candidates, campaigns, and voters all behaved as though it didn't/

You're technically correct, but how do you think people would behave differently with ranked-choice voting?

I’m Australian and I consciously vote very differently. My first votes are for small parties that are very unlikely to get elected but who’s campaign largely rests on an issue I support.

This encourages the other, more popular parties to adopt that issue if they see these parties are popular.

Otherwise I’d just vote for the “best” and most popular.

I'd like to point out that Australia is still dominated by two parties even with the benefit of the parliamentary system (giving proportionate seats to other parties). This forces two major coalitions which isn't extremely different.

This is true in the lower house, but not in the upper house (the senate) where there are 5 significant parties holding seats (counting the LNP coalition as one party and the Centre Alliance as a significant party) plus some independents, and no natural majority across any combination[1].

[1] https://www.aph.gov.au/Senators_and_Members/Senators/Senate_...

I said "dominate" not "absolute control." I mean you're talking 14 candidates between 5 parties, with 9 from Green (5 between 4). By "dominate" I mean that the 2 parties (Coalition + Labor) control ~82%. Even America isn't 100% controlled by Dems + Repubs, but it is the proportion that is the problem. Really, that the parties are not representative and there is a lack of mechanism for a ore representative party to start and gain power. I'd argue that the same is true for Australia in that respect.

I should also point out that I mentioned a specific advantage.

Well the non-Labor/LNP candidates hold the balance of power, so there is that.

I do agree that the dominance of two major parties is mostly bad though.

Yeah, it’s not ideal. I’m no political scientist though, I’m just sharing my own motivations and voting patterns under a different system.

I just want to point it out because I strongly advocate for cardinal systems like Approval or STAR (I also like proportionate representation). The big reason is that we have a 100+ year experiment that disprove Fair Vote's (the one pushing RCV/IRV) largest claim: "solving the spoiler effect." The math doesn't side with the claim either, but Australia is a fantastic experiment. So if we're going to do something new, why not better? Especially if better is also simpler (approval is near trivial).

They can vote for candidates other than the top 2 without wasting their vote. Because when any candidate is eliminated, their votes flow to each voter’s next preference. That encourages co-operation between coalitions of candidates instead of polarization of only 2.

i don't know or (in this case) care what would be different. (and i say that as a fan of RCV and similar).

i'm just tired of people positing counterfactual worlds, and assuming whatever tiny little change they make will have no ramifications except for the one that they want. it is sloppy thinking.

Gotcha. The parent comment should have said that it "might have" changed the outcome.

And if there’d been ranked choice voting in 92, Bush Sr. would likely have won over Bill Clinton.

You can't get a winner for ranking (or even ordinal) when you made people vote singular. These numbers give us no insight into how much or how little voters liked/approved of the other candidates. It is likely that Nader voters would have also approved Gore, but it also wouldn't be surprising if Gore voters approved Nader and he came out on top. It gets much trickier than just assigning votes over.

It's also possible that some Nader voters wouldn't have entered a second choice, or that the balance of 1st and 2nd choices between Nader and Gore would have resulted in a non-monotonic outcome (i.e. some group of voters cause, for instance, Gore to lose to Bush because too many of them entered Gore as their first choice; that can actually happen with RCV, which is why I am uncomfortable having the integrity of our democracy depend on that particular voting system).

> Gore to lose to Bush because too many of them entered Gore as their first choice; that can actually happen with RCV

How does this happen?

The only scenario I can see where this happens is if Nader received more votes than Gore, and those Gore voters didn't put Nader second. But then the problem isn't too many voters putting Gore as their first choice, it's too few!

Wikipedia has an example of a non-monotonic result.

https://en.wikipedia.org/wiki/Monotonicity_criterion#Instant...

Yes, despite Fair Vote's claim IRV doesn't solve the issue of spoilers. Despite the 100+ years of experimentation and despite the math. I'm always confused why there isn't a bigger push for methods like Approval and STAR. They are dead simple and solve the issues. The conspiracy theorist in me (joking) says its because it would actually change things where IRV won't.

I think part of it is that people tend to be laser-focused on the problem they're having right now and not the problem they're going to have.

For instance, RCV (aka IRV) is great solution to the problem that a minor candidate siphoned a few votes away from a major candidate and swung the results of a national presidential election several times in recent memory. If voters who really wanted to vote for Nader could have also voted for Gore, maybe Bush wouldn't have won, and that election would have better reflected the will of voters.

The problem that RCV is not prepared to deal with is what happens if people actually start voting for Green or Libertarian or Democratic Socialist or whatever in large enough numbers that they actually have a chance of winning elections. At that point, you run into problems and RCV starts producing nonsensical results. The voters learn from this that it's not actually safe to vote for third party candidates, and the two party system remains in place. It's a little bit better because you don't have minor candidates swinging elections, but it's still a duopoly.

I think that is a fair assessment. My big fear is that we'll implement IRV and then when we run into the next problem (let's be real, the next election) people will say "well that didn't work" and discount voting systems as being part of the solution.

I mean we've literally done this in America (though not nationally). We've had IRV before. People turned against it. So I don't think my fear is unjustified since there's historical precedent.

Australia has IRV (has for 100 years) and is still dominated by two parties even though they have the advantage of proportionate representatives through their parliamentary system. And in reality they have their two coalitions.

What worries me is that CGP and Hasan have large audiences but look to have done surprisingly little background research into the topics and alternatives. And let's be real, voting math is hard and unintuitive. It is overly constrained and there is no global optima/easy answer. I don't expect the average person to look more than surface level, but I do expect more from groups like HN.

Yeah, that's also my take: if you introduce a weird new voting system and voters don't like it, it's not just one step in a progression of incremental reform that will eventually lead us to a smoothly-functioning democracy. Rather, voters will just demand we revert back to the old system because (as far as they can see) that fancy voting nonsense is all junk. Then we'll be stuck with FPTP until either democracy fails outright or the previous attempt at reform has been forgotten, perhaps in a hundred years or so, and someone tries again. When you're changing the basic mechanisms of voting, you don't get any do-overs.

I am optimistic that within the last couple of years, it seems that any online conversation about RCV among even moderately technical/nerdy groups of people seems to almost always include someone bringing up the downsides of RCV and the better alternatives. I used to kind of dread RCV discussion threads because I felt like I ought to explain the problems, but it's very technical and abstract. Now I usually find someone has already written what I would have said, often in a more concise, clear way.

It has already mattered downticket; it is, I think? part of the reason Jared Golden now represents ME-02.

20 years is quite the tenure for a president

And Hillary Clinton as well if I recall correctly.

I don't think that is obviously true. Stein and the Green Party finished behind the Libertarian Party by a signification margin.

What states would have switched? In Wisconsin Clinton + Stein did outpoll Trump alone (by ~10K votes), but the Libertarian party had another 106K votes that would have been allocated and it's not hard to imagine Trump getting more than Clinton of them.

It was a similar story in Pennsylvania and Michigan.

Reply on news.ycombinator.com