There are many kinds of voting which involve the voters ranking their choices. The difference is in how those choices are resolved. You can get very different election results depending on whether you do instant runoff voting, or use a condorcet method, or use borda count. They're all ranked voting, but they have very different properties.
Interesting, thanks for the clarification. I didn't realize how nuanced it can be. For example, when using the Condorcet method there's a chance of not having a winner. From wikipedia[1]:
> This is sometimes called a Condorcet cycle or just cycle and can be thought of as Candidate Rock beating Candidate Scissors, Candidate Scissors beating Candidate Paper, and Candidate Paper beating Candidate Rock.
The first time I heard about it was called ranked voting, but instant run off is the more popular term. There are multiple rounds, where the least voted candidate is removed, so it makes sense.
It's a kind of optimised variant of the 'runoff voting', where the election is rerun with one fewer candidate if noone gets a majority. That's where the name comes from.
In these parts (the UK) the same system is known as the Alternative Vote (AV) which is just as meaningless a name. But in defence of my fellow political scientists, there are a lot of different voting systems, many of which have alternatives, rankings, and runoffs, so unambiguous and useful names are hard to find.
> This makes a lot of sense as a system, but who came up with "instant runoff"? should be called "ranked voting"
“Ranked Choice Voting” is one of th names for the system, but since their are dozens (maybe hundreds) of voting systems which use identical ranked choice ballots with different resolution procedure, naming it “instant runoff” (since it uses ranked choice ballots to implement a variation of majority/runoff) is much more specific of name.
Because runoffs already exist in a lot of places around the US, and IRV does accurately describe what happens if you are comparing it to a runoff election. I agree that "ranked voting" is a better term in a vacuum (and probably a better term regardless), but since there are more people in the US who have been exposed to runoff elections than ranked voting, I understand the choice.
In Australia it's called "preferential voting", since you are voting according to your order of preference. Each round of the count is typically described as "distributing preferences".
Places with IRV generally still have two major parties, even if weaker, like I said in the original comment. I'm just saying that how this normally plays out is you still end up with 2 major parties (though weaker).
We really don't, historically (for the last century or so) it's been pretty cut and dry between either Labour or Conservatives with the occasional coalition (always involving one of those two parties).
Well, if San Francisco's IRV races are to be used as an example, you end up with a one-party system. But more seriously, I think "clone-proof" electoral systems tend to better allow for more than two parties, or at least more than two candidates at a time.
In Australia (which has 2 chambers - 1 under this system and the other proportional):
It is very difficult for a 3rd voice to gain traction. A party which has attracted over 10% nationwide for the past 2 decades, Greens, has exactly 1 seat out of 150 under this system.
We've had this method of voting for a century but voters generally see it as a binary choice - Roughly 35-40% will vote for A and 35-40% will vote for B. Typically there will be up to 5-10 candidates on the ballot paper. So that 20-30% of the "neither of major 2" could be split 8 ways across a wide political spectrum. Inevitably, most outcomes fall back to the 2 majors.
This is a feature, not a bug. In a Westminster system the government is formed on the floor of the House of Representatives by being able to pass motions of confidence and supply. Australia provides an unusually generous opening for minor parties through the Senate, which is a proportional system.
Your statistic about the Greens, in particular, is true only for the House. The Greens hold 9 Senate seats out of 76, or approximately 11% of the Senate. Given that their national Senate vote was 8.7%, I think this is a pretty reasonable outcome.
Just the other day, I discovered a group that is pushing for this very thing. If you're interested in making this system the norm, start pushing for a change in your local area. https://www.fairvote.org/
Or you could learn more about other options and critiques of IRV in order to realize that our effort at reform would be better spent on STAR voting (or maybe approval voting), even though IRV is better than the status quo.
IRV has the fundamental flaw in that it counts some voters' preferences and ignores others. It's an overly-complex and mediocre unequal system.
Check out https://www.equal.vote to understand what it means to have voting equality and how we can pursue systems that honor that.
If you really wanted to encourage 3rd party candidates, then Proportional Representation would work even better.
Both are better than first past the post, but agreeing on an alternative is HARD.
The situation in Canada is instructive: the governing left-central Liberals promised to replace first past the post, but reneged on their promise when they realized that the committees studying the issue were going to come back with a Mixed Member Proportional recommendation which would hurt them and help the left-wing New Democratic Party, rather than the ranked ballot mechanism that would help them.
(That's speculation, of course. They gave some other BS reason for breaking the promise)
Proportional voting is a vastly superior system for arranging many-people blocks (senate and congress in the US), and also completely breaks gerrymandering.
It doesn’t work for individual candidate positions like president, because you can’t have a non whole number percentage of a single candidate
Electing the president separately, when the president is a position of power and not just ceremonial, doesn't make sense anyway. With a multiparty system, you are better off waiting for a coalition to form in parliament and letting that select the president. Otherwise, either the president or the parliament can act as a roadblock.
Compare with how the system of the Netherlands works: a single country-wide electorate, and no threshold except for the number of votes needed for a seat. The outcome is numerous parties in parliament (i.e., far more democratic than the US) and coalition governments which must be formed or at least supported by parties that got a majority of the votes between them.
I'd argue that if the president is intended to be a position of power, then a separate election is good - otherwise, one party (presumably the largest) in the ruling coalition has power far beyond what their vote count would justify, and the person gets power without real input from the electorate.
Especially with many parties, if parties A and B are reasonably similar, it's quite plausible to like party A's platform while preferring party B's presidential candidate. A separate presidential election accounts for that, allowing you to elect A for the house, and B for president.
Note that parties A and B are similar - I'd argue that they'd have to be, and that it's extraordinarily unlikely to get a president the ruling coalition is incapable of compromising with. This works out because overall, the vote for president should be ideologically similar to the votes for congress/parliament. (That working, of course, is contingent on some form of ranked voting for the president, rather than the FPTP system the US has now.)
All that said, I'm from Canada, and I don't really agree with the idea of a single person with large amounts of power, but that's my take if there must be one.
Well, in Canada you have a Prime Minister who is elected by the House of Commons? Would it be so much better if the Prime Minister was elected separately? How would a Prime Minister run a government if his/her party was in opposition?
The US constitution and founding documents make it very clear that the president is meant to be a separate entity - the fact that it’s devolved to the point that one party will rubber stamp pretty much anything the president tells them to is a failing of that party.
Specifically the president represents the executive branch, then you have the two houses, one that represents the people approximately proportionally, and one that gives 20x multipliers to the voting power of people from mostly empty states.
Seems like it pushes everything to a lukewarm center, change already difficult to enact would be even harder, it goes to "who sorta kinda appeals to the most". I imagine turnout would decline.
I know I've "voted defensively" before, and I really don't like having to do it. I want to vote for those that I agree with, I don't want to vote against those who I disagree with.
It may seem like a small distinction, but if someone loves candidate A, but they are extremely unlikely to win, most will end up voting for B because they REALLY don't like C.
That leads to A never getting in office until they become more like B.
But with instant runoff, I'm free to know that I can vote for A first, and B second, and not worry that my support for A is implicit support for C.
It not a perfect system, but IMO it's a lot better than what we have now in the US.
Edit: another piece that I forgot to mention is that instant runoff can give more signals beyond the winners or losers. Currently you have no way outside of polling to understand what voters want, it leads to a vocal minority being able to control the narrative.
With instant runoff, the runners can easily see that even though candidate A didn't win, they did get a sizeable number of votes, and maybe next time B runs, they will incorporate some of A's ideas into their campaign.
You end up with the candidate who most voters at least like while offering more choices. Seems like it would wipe out the ‘rile up the base’ strategy in favor of trying not to offend anyone.
Right, I'm saying it's a stable but stale form where candidates optimize being tepid. May seem fine but I see how it would lead to further entrenchment of a ruling class
I must be misunderstanding something, because it seems to me that this basically gives the top candidate their votes, and the second top candidate every other vote. What am I missing?
EDIT: Oh, on one of the rounds the top candidate might become the bottom candidate, I see.
Let's say Clinton, Trump, and Sanders on on the ballot. Some voters could rank their choice, from most to least preferable, Sanders -> Clinton -> Trump, while others could rank Sanders -> Trump -> Clinton. If Sanders received the least amount of votes, voters with the first ranking would have their vote transferred to Clinton while the voters with the second ranking would have their votes transferred to Trump. It wouldn't necessarily reallocate votes to third and lower ranking candidates to the second ranking candidate.
Ahh, I see, okay, I got it completely wrong then. I thought you voted for a single person and those votes were reassigned, I didn't realize you gave a strict ranking.
What happens if the candidate that gets removed is in the middle? I'm guessing nothing, since it "stops" at your first preference that's still in the race, ie your vote goes to the first person?
The way I understand it, you're correct. If a candidate that you ranked higher is still in the running, your vote goes to them. Only when your highest ranked choice is eliminated does your vote get transferred.
That seems managable when you have 3 candidates. What happens when there are 10 or 20 candidates? Would voters have to list their top 10 or 20 choices?
No, you would only have to specify/rank as many choices as you want your vote to carry over. If you only like 5 candidates and hate the other 15, your vote will simply be thrown away and not counted if you don't specify any of the other candidates.
If all your choices are eliminated, then for future rounds it counts as abstaining. If enough people do so that no one crosses the threshold, the process repeats until the last candidate remaining is (normally) deemed elected without reaching the quota (though it could also be treated as cause for a rerun if preferred).
In practice here, even in such large contests the top 5 candidates normally end up with 90% of the vote and there's enough overlap between the parties (the further left parties transferring to centre left, the two centre right parties transferring to each other) that there's normally a clear leader by then, it's rare that e.g. a candidate with 48% gets eliminated in favour of a candidate with 49% in the last round.
I'm continually upset and confused by Instant runoff voting's popularity. IRV is, by my reckoning, a much inferior (and more complicated) system than other systems, like range voting or ranked pairs. And it has some bizarre properties that make me wonder if it is even worse than the traditional first-past-the-post.
One property of IRV that makes it so questionable is its non-monotonicity. This means that you can help a candidate by ranking them lower, and hurt a candidate by ranking them higher. It is extremely easy to visualize how ill-behaved IRV is in the charts on this website:
Meanwhile, ranked pairs (a "condorcet" method) has the property that, if candidate A is more popular than all other candidates pairwise, then candidate A will always win. This is _not_ true in IRV.
If you're interested in falling down the rabbit hole of electoral systems, I recommend looking at this chart that compares their properties:
To me it also seems IRV is complicated. That said, doesn't the non-monotonicity property of IRV support outcomes where a compromise candidate might win when they otherwise would lose? That doesn't seem to be a bad thing. If people vote strategically to make a centrist their 1st vote to avoid what they consider an extreme, that seems like the essence of political compromise and an improvement over 1st past post.
Perhaps the most important important property of a voting system is that it is easy to grok, so that people trust it. In particular, aggregates at the local level should combine simply with other districts to tally overall winners.You can tally IRV incrementally, it's just that ranking permutations are the units of aggregation. That's non-trivial. For this reason, I prefer approval voting.
Voting systems like range voting seem to me, based on simulation, to have this property but much more well-behaved. Range voting is also much simpler than IRV. Approval voting is just a special case of range voting. I'm afraid people want to take one step forward, three steps back with IRV.
Range voting is challenging because people often have difficulty putting an exact number on how much they want something, and also scales differ between different people. On a 0-10 scale, some people will be putting every candidate as either 0 or 10; others will always put in 3 and 7; and this depends a lot more on the voter than the candidate.
Approval voting is simple enough, but is vulnerable to bullet voting: approving an additional candidate can make your first choice lose. This makes it unsuitable if your goal is to make third-party voting more viable, which seems to be the premise of the article.
> Range voting is challenging because people often have difficulty putting an exact number on how much they want something, and also scales differ between different people. On a 0-10 scale, some people will be putting every candidate as either 0 or 10; others will always put in 3 and 7; and this depends a lot more on the voter than the candidate.
And, particularly, there is very notable cultural variation, which means (ab initio at least; over time, one expects behavior modification efforts would be deployed around this), support from different cultural groups in a diverse polity will have different electoral value even when turnout is the same.
> Approval voting is simple enough,
But also has the same problem: concretely what “approval” means is unclear and will be subject to voter-to-voter variation even when the actual preferences are identical.
> a compromise candidate might win when they otherwise would lose? That doesn't seem to be a bad thing.
Only if you assume that compromising is fundamentally good. Do you believe the propensity is to compromise to get the best of the both worlds or the worst of the both worlds? I would point out, for example, that during the bush (II) administration, the only truly bipartisan, compromise initiatives were war in Iraq, No Child Left Behind, and Corporate Bailouts.
They have IRV in San Francisco for the mayoral race. What ends up happening is that a candidate who would not otherwise be in the majority can end up as mayor because he/she was everybody’s second choice. Additionally, candidates who think they might lose team up and start promoting each other. It’s bizarre and very difficult for voters to understand the implications of their actions.
"What ends up happening is that a candidate who would not otherwise be in the majority can end up as mayor because he/she was everybody’s second choice."
But, isn't that the whole point? ... that a compromise, perhaps centrist, candidate can emerge victorious?
But the question is, is this a good thing? If 51% of the population want an extremist candidate, but also like a centrist candidate, then IRV incentivizes them to lie on their ballot and promote a candidate they don't like over the candidate they do like.
Compared to the common “first past the post” voting, yes it is a good thing. This is because the common voting system encourages a “two extremist parties” system where everyone is effectively voting “greater horrible evil vs lesser horrible evil”.
Here's the problem though; IRV doesn't really do the thing you want it to. It can actually make an extreme candidate win even when voter opinion is centered around a moderate candidate. It's very obvious in this visualization (from the site linked a few comments ago): http://zesty.ca/voting/sim/54,47_77,64_13,10_irv.png
Just a friendly note: this visualization means nothing without context, and is far from "obvious". Unfortunately, I'd place this in the category of "things that are obvious to people who already believe it". Is there a better way to go about explaining your idea?
To paraphrase, this is a simulation which places candidates and voters on a two-parameter preference space. This is useful because two dimensions are easy to visualize, but the same intuitions apply to higher-dimensional spaces.
Each point is colored to answer "which candidate wins when the popular opinion is normally distributed about this point?". That means the fractured visualization shows IRV's non-monotonicity, and how even when popular opinion is close to a centrist candidate (here in red) it might be that the voting system results in the election of an extreme candidate (green).
I don’t think the problem with having only two viable parties is that they will both tend to be extreme (and that doesn’t strike me as empirically true at the US federal level). Rather, I think the problem is that parties have aggregate power, which means that individual policy preferences get bundled together in ways that may not match an individual voter’s preference. What if I prefer one policy that’s only part of the party X platform, and another that is only part of the party Y platform? Or, another way to think about this party aggregation: what if I prefer the candidate from party X for my local congressional district, but I can’t stand the thought of the likely Speaker of the House of party X?
What I often see in my local election is that 3rd-party candidates are often (not always, but often) very inexperienced, and might have platforms that don't make sense or aren't coherent. I've frequently got serious doubts about their ability to function competently - let alone effectively - in the position they're applying for.
With IRV, though, I could see a situation where they are much more likely to win simply because, nowadays, Republicans' and Democrats' general level of mutual disdain is such that I wouldn't be surprised if it ends up being common for people who identify with one major party to rank the candidate from the other major party below every other candidate more-or-less out of spite.
>What I often see in my local election is that 3rd-party candidates are often (not always, but often) very inexperienced, and might have platforms that don't make sense or aren't coherent
This is a symptom of our current system. Anybody who has any ambition in politics will modify some of their positions to fit one of the major two parties. If running for another party wasn't a death sentence, you would see more competent candidates in those parties.
Yep. Just trying to think through plausible practical scenarios where IRV might produce a bad outcome.
Which is worth doing because, regardless of what the ideal situation is, if it's associated with a spectacularly bad outcome in one of the earlier elections that uses it, that's going to be a big impediment to wider adoption.
Even in a sharply polarized contest, it's usually the case that most of the "third party" candidates receive fewer first-preference votes than those from the major parties.
So the minor parties (and independents) are eliminated first, and their preferences are distributed. Which usually means than one of the major party candidates scrapes over 50%, reflecting the minor party voters' basic preference for one of the major parties over the other.
In this (common) situation, I think IRV actually works really well.
The major parties still end up in government, but they see the first-preference percentages for the minor parties, and understand what policy positions are important to the electorate.
... while still being everyone's second choice out of several?
Edit: As much as everyone's trashing IRV as "horrible" I don't see a simple, worrisome case where it violates monotonicity; those seem limited to really pathological electorates where nothing will make them happy anyway.
Yeah, I'm not sure where the qualms about monotonicity are coming from. In an IRV the lower-ranked preferences do not count until the higher-ranked preferences have already been eliminated for not reaching majority
> Edit: As much as everyone's trashing IRV as "horrible" I don't see a simple, worrisome case where it violates monotonicity;
It's terrible because the loser-elimination step that makes non-monotonicity an issue also makes it less straightforward to tabulate, not because non-monotonicity is an issue outside of edge cases.
Yeah I guess I’m just not really invigorated by fears of “difficulty of tabulation” and “bad behavior on edge cases”. Maybe I react differently from the typical voting system enthusiast?
>Difficulty in tabulation affects speed of, certainty of, and confidence in results, as well as cost of implementation.
Which doesn't rise to the level of "that is so bad it must be stopped".
>> Maybe I react differently from the typical voting system enthusiast?
>Nope,
Kind of strange, since the entire reason for my confusion and joining this thread is because of enthusiasts' doom-and-gloom about the quality of the results themselves, and how they worry about the edge cases and I don't.
Are you invigorated by the cost fears? Because they seem like a sliver compared to the cost of Gore vs Bush.
I think that the Burlington example that has been cited a lot in this thread shows that non monotonicity is a problem for IRV in not just pathological edge cases.
Basically, imagine you have three candidates: A, B, and C, and there are 50 voters, and the votes come in like this:
100 - A > C > B
100 - B > C > A
40 - C > A > B
10 - C > B > A
After round 1 of IRV, candidates A and B each have 100 votes, C has 50. C gets knocked out, 40 votes move to A, 10 votes move to B. A now has 140 votes, and B has 110, and A is declared the winner.
The problem is that the B > C > A don't get their second choice considered. Why does this matter? Look at it when you put the votes into pairs.
A vs B: 140 people think A > B. 110 think B > A.
A vs C: 100 people think A > C. 150 think C > A.
B vs C: 110 people think B > C. 140 think C > B.
C is actually the more popular candidate, but C LOSES using IRV!
If you and your dining partner rank the kind of restaurants you'd like to go to tonight.
You rank 1) Indian 2) Chinese and 3) Italian
Your partner ranks 1) Gastropub 2) Chinese and 3) Steakhouse
you're going to get Chinese food, not because it exists on a line between each of your first choices, but because it is acceptable to each of you, if not preferred. It is a compromise by definition.
That's a compromise by the numbers, not by the nature of the food.
Chinese food isn't a compromise, a blend between Italian, Indian, Gatropub and Steakhouse food.
In Chinese food there isn't a little bit from every other choices. A candidate being chosen on virtue of a compromise by the numbers doesn't mean s/he aggregates some ideas from other candidates.
Hence my remark about that candidate being on a different spectrum from the others.
I am not saying he shouldn't be chosen or that people wouldn't be happy about it, I am saying that his ideas aren't a compromise from the other candidates' ideas.
Say in the US 2016 election Sanders had run as a 3rd party. He was a liberal populist. He might well have been the second choice of both the Trump and Clinton voters, despite not being positioned between the two. He'd still have been the compromise candidate.
Not a great example, because neither you nor your friend probably despise any of those 5 choices, and even if you do, it’s no big deal to just decide to get dinner separately tonight. Remember, political elections are about choosing the candidate or policy that exerts power over everyone in the jurisdiction.
I don't see how that works. If we have say 3 candidates, a rightist, a leftist, and a centrist, and the centrist is almost nobody's first choice but almost everybody's second choice, the centrist is eliminated first, and either the rightist or leftist will win. I could see someone arguing the centrist should win, but given the rules as described he won't.
This scenario happened not in San Francisco, but in Oakland with Jean Quan's election.
Since then the second-tier candidates have done what you say about promoting each other: "vote for me and for X". If one loses, the other could win in the next round. IRV has a moderating influence on how candidates campaign. By making a bid for those later rounds viable, now everyone has an incentive to build strong coalitions before they even get into office.
Would another system work better? Maybe, but I still count it as an improvement over FPTP.
It happened in the SF 2018 mayoral election too. Mark Leno and Jane Kim (the two progressive candidates) teamed up against London Breed (a moderate liberal) and suggested each of them mark each other 1 & 2. As a result, after the first round of voting Mark Leno was in first place despite London Breed getting 50% more votes, and it was only after a recount and full tabulation of absentee ballots that London Breed was declared the winner.
Now, arguably this is IRV functioning as designed, where the city's progressive wing could vote as a block without worrying about 2 candidates splitting the vote. It was very much non-intuitive, though, and a lot of London Breed supporters were quite pissed off. IRV allows you to "vote for a position" rather than "vote for a person", but many voters are still stuck in the "but....they got fewer votes!" mindset.
> It happened in the SF 2018 mayoral election too. Mark Leno and Jane Kim (the two progressive candidates) teamed up against London Breed (a moderate liberal) and suggested each of them mark each other 1 & 2. As a result, after the first round of voting Mark Leno was in first place despite London Breed getting 50% more votes
You meant after completing all the rounds of IRV, Mark Leno had more votes despite London Breed having more first place votes.
Which, yeah, is the whole point of IRV (and, with slight variations, most ranked ballots methods): a plurality winner that doesn't have broader support but ends up at the bottom of ballots not ranking them first shouldn't generally win over someone who has broader appeal just because the broad appeal splits first place support with another candidate.
> IRV allows you to "vote for a position" rather than "vote for a person", but many voters are still stuck in the "but....they got fewer votes!" mindset.
No, it lets you vote for a person, and also specify, “if not Alice, then my vote goes to Bob”. The winner in IRV is always the candidate with the most votes at the end of the process; reporting only first place votes as votes for the candidate is just wrong reporting of IRV counts.
As a contrasting anecdote, Australia uses IRV for many of its elections, and I think it's fair to say that most people understand how it works and like it.
Absolutely, it's more complex than a simple majority. But I think "bizarre" is unfair.
In practice, I think it helps to provide a diversity of representatives that more closely reflect the diversity of views in the electorate.
Ah yes, people complaining why we can't have X++ when we can't even have X. We should take what we can get and move the goal post from there. Any progress is good.
Look at the results linked at zesty.ca: it’s not clear that IRV is progress. Meanwhile, approval voting is simple, easy to understand and to switch to, and a clearer improvement on the status quo. This is why the frustration.
I mean, if we're going to finally be able to switch it, why not switch to the best option instead of starting with the second worst and switching again later?
> I mean, if we're going to finally be able to switch it, why not switch to the best option instead of starting with the second worst and switching again later?
Because that kind of lesser-of-two-evils incrementalist attitude is inherent to the first-past-the-post system we have now.
But these things aren't degrees of one another, they're totally different.
Saying "we only ever eat meat, let's try eating an orange and see if we like it before we eat any of this lettuce" doesn't follow any actual logic. You might hate lettuce, you might hate oranges, but there's no specific factor that binds them together.
Also what sort of things would you factor into what you prefer? How important is each of them?
Because X++ is just as easy (and difficult) to implement as X.
If my choice is between 1% of what I want, 2% of what I want or 99% of what I want, saying "why don't you just be happy with 2% since it's twice as good as the 1% you could have gotten" isn't the most logical point to make.
I care much more about getting a good voting system than I do improving our voting system, if that makes sense.
I would make the stronger claim that IRV is an even worse voting system than the one we already have. At least with FPTP voters can easily figure out the effect of their vote. With IRV it can be ridiculously hard to interpret, and relies on a lot of strategic thinking like "well, I think candidate A isn't great and so will not have enough support to make it to the later rounds, so I might as well rank A first so that maybe they can eliminate someone else I don't like before they get eliminated and then my vote gets transferred to my actual top choice... oops too many people did this and now candidate A won"
This could also be viewed as a feature rather than a bug. Make it complex enough and maybe people will fall back to "Vote for the candidates you want to see win, in order of preference, and quit worrying about how everybody else is voting", which is how elections are supposed to work.
There's an impossibility theorem (Gibbard's) that states that for any non-dictatorial process for choosing among > 2 outcomes, there is no way to avoid tactical voting potentially swaying the results. The next-best alternative might be to make tactical voting so complex and error-prone that nobody bothers.
I worry that this provides a very strong tool for people to make arguments that "ranked voting systems don't work, since you're always confused by their results" and have a strong case to bring everyone back to FPTP.
I've gone down that rabbit hole, repeatedly. It's just too fascinating. Likewise, I feel obligated to bring up the Voter Satisfaction Efficiency (or index) metric that seems to be favored by the approval/range voting crowd: basically estimating how utilitarian a given voting system can be. There's a fascinating write-up of their modeling results on github, with code (python).
I like pointing people to https://modernballots.com/ for all those "where should we eat lunch?" type questions. People usually seem quite surprised how well the Schulze method generally works.
I implemented the Schulze method in a hobby project once, and the math is a rather simple digraph algorithm, which also gave me a lot of confidence that it's a well thought out voting method.
Runoffs are something people understand, IRV is conceptually similar and makes intuitive sense to people, it's very easy to explain to people, 'if nobody gets a majority, eliminate last place and redistribute their votes, repeat'.
Compare that to to some of the other systems which often have complex mathematical descriptions which are hard for people to wrap their heads around.
Another thing is how big a deal is 'non-monotonicity' is it actually going to be something that will affect elections all the time, or is it something that's more minor than say turnout which honestly has no solution and on one side you have USA where turnout is low so the game is all about making sure your base turns out, or on the other side you have Australia where voting is mandatory so making sure people who don't care, don't vote at random is a big part of the game.
IRV is more complex in every way than STAR voting.
STAR is just score with a single automatic runoff. It gets everything you are talking about in terms of intuition, explanation… everything you like about IRV is present in STAR and basically everything wrong with IRV is fixed by STAR.
I think the idea of rating candidates vs voting for them could be less intuitive and feels more complex, though there isn't any reason you couldn't just change it to 'rank candidates in order of preference 1-5, candidates may tie' or something.
What it's missing compared to IRV is that you can't go, it's just like what people already do (runoffs), just in a single round.
It's true that ratings can feel a little harder to grasp, but just any concrete experience with the system, it's easy to get. Check out http://star.vote/ and try for yourself.
In STAR, 5 points is simply "most points" i.e. most support. It doesn't mean something semantically. You can give up to 5 points, like if there's a race, you can push each candidate forward as much as 5 steps. The only thing it means is that you pushed them forward more than another candidate.
That's not that hard to get intuitively.
IRV is not what people already do but in a single round, it's much more complex than that. And whoever's 1st choice in IRV loses in the final round, they (and that could be as much as 49% of the voters!) never get any of their other preferences counted at all, unlike either STAR or our current two-round system.
> In STAR, 5 points is simply "most points" i.e. most support. It doesn't mean something semantically.
Yes, so you import the well-known cultural variation in numeric rating systems without concrete semantics for rating levels into your voting system. That's a great idea.
More data on ballots is good only as long as it has a consistent meaning and is used in a way that is consistent with that meaning.
STAR (and other score voting systems) gather noise and pretend it is signal.
STAR gathers scores that voters choose when looking at the pool of candidates. It uses those in tabulating the outcome. They have an effect. There's no sense in which this is noise. I don't even understand the claim.
STAR Voting allows voters to rate all the candidates similar to how we rate books on Amazon
Ah yes, Amazon, that site with famously reliable and unbiased user reviews.
This is a terrible idea. It’s a bad voting system made worse by appeal to similarity with popular apps and websites, which themselves use a really bad voting system.
It's not a terrible idea, it's a bad metaphor from imperfect people trying to explain a perfectly fine idea.
STAR voting is NOT like rating books on Amazon, it's just superficially similar.
At Amazon, if you don't like any of a bunch of books, none of them get 5 stars. With STAR voting, you award points not to mean anything, but only to decide how much ahead in the race you want to push each candidate versus the others. Whoever you want to win most, you push them ahead with all 5 available points. That's it.
Your reaction and misunderstanding does further my annoyance at the people explaining STAR using the Amazon analogy, and I think that bad comparison is harming support for what is truly an excellent voting system.
I’ll admit I didn’t read much further than that, yeah. It’s possible it’s a good system and I’m just put off by the unfortunate comparison.
My issue with all of these point- and range-based voting systems is that when you see the results, you might want to tweak your scores.
Let’s say I’m range voting out of 10 and I give four candidates 1, 3, 8 and 10 points respectively. But the candidate I rated 3 wins, with the candidate I rated 8 a close second. Damn it! If I’d known that was going to happen, I would have voted 1, 1, 10, 10.
So with extra information, range voting boils down to approval voting, and I think it all ultimately becomes ranked voting. It’s hard to imagine any scenario where I would want to change my rankings, even if I have information about the results. If I like both candidates A and B, but A slightly more, I’ll always put them at the top of my list and in that order.
Now, it’s true that in IRV there are scenarios where I can get a better result for myself by voting for candidate ranks in a different order; but those are rare (I think) and definitely unintuitive. In the example above, if B were a real long shot candidate, it might be worth ranking them above A in case B gets eliminated early. But it’s unlikely to make a big difference because, you know, B is a long shot. And that’s just IRV; other ranked voting systems are much more robust.
> Let’s say I’m range voting out of 10 and I give four candidates 1, 3, 8 and 10 points respectively. But the candidate I rated 3 wins, with the candidate I rated 8 a close second. Damn it! If I’d known that was going to happen, I would have voted 1, 1, 10, 10.
This is the EXACT reason STAR voting was invented. Unlike plain score, STAR has a runoff where it compares your preference of the top two and puts your full vote to your preference.
So, this is like an exact case study of why STAR is worthwhile. STAR proposes 0-5, but we can stick with 1-10 for this case. You vote 1, 3, 8, 10 and the top two candidates are the ones you voted 3 and 8. With STAR, it doesn't matter that your 3 candidate got a total score higher than your 8 candidate, your vote in the runoff goes to your 8 because that's rated higher. If more voters rated that candidate higher, they win even if their plain score total was lower.
Thus, you don't have to exaggerate all your scores. In fact, you have an incentive not to, because if you just give ties, you risk abstaining in the runoff instead of having a say there.
I mean literally you just spelled out the most common critique of plain score voting which is itself THE argument for why we should use STAR instead.
It sounds like STAR is equivalent (or at least similar) to the Condorcet-with-score-based tiebreaker that somebody proposed elsewhere (edit to add: https://news.ycombinator.com/item?id=18395358 ) in this thread. I understood that one but not STAR. It needs much better explanation and marketing! Edit: doh, wait, STAR is in fact the exact opposite -- scores first, ranks as a tiebreaker.
I’m reading https://www.equal.vote/starvoting and I still don’t entirely follow. It’s lacking some detail I’d like to understand. For example, it looks like it is not a Condorcet method. That seems bad to me -- if you’re not electing the Condorcet winner, you need a very good explanation why; and I don’t trust an appeal to numerical ratings that each voter will interpret and use in different ways.
I do agree that it looks probably better than IRV, but I’m not yet convinced that it’s a really good system and is worth pushing as a good approach for real elections.
Thanks for being open-minded. I mentioned in my other reply (to your other reply) why Condorcet isn't strictly ideal. Discussing the edge-cases where STAR does not support the Condorcet candidate is too long a discussion, I can't get into it here.
One short point: the primary problem with not electing Condorcet is that coordinated strategic majority could, in hindsight and for future elections, force the Condorcet by refusing to support later preferences. In STAR, that strategy is risky and unpredictable. If the Condorcet is a strong case, it will win STAR anyway. If it's an edge case, nobody knows in advance if it will be the true majority or just a bit less, so strategically zeroing all by the favorite is so risky in that edge case, most people won't do it (it could backfire for them to refuse to support their 2nd favorites if their hail-mary hope that they just might have a Condorcet turns out to be wrong). No strategies with STAR are reliable. The best outcomes come from just being honest.
Note: STAR doesn't use rank as a tie-breaker only. It can flat-out overrule a score difference with the 2nd-highest scoring candidate winning. It's not just for ties.
> I do agree that it looks probably better than IRV, but I’m not yet convinced that it’s a really good system
Well, since IRV gets pushed for real elections and STAR is better on a large number of factors, and both are better than FPTP…
Here is my less-hot take: STAR is exactly back-to-front.
You’re combining ranked voting with range voting.
Ranked voting is objective and clear (the Condorcet criterion is very compelling) but unfortunately incomplete (Arrow’s theorem, which in this case boils down to unresolvable three-way cycles.)
Range voting is more tractable (counting is easy, no awkward cycles) but subjective (no-one knows what the scores actually mean) and prey to tactical voting.
STAR starts with range voting (so you lose Condorcet) and adds cut-down ranked voting as a final step (only after some rank information has been discarded; and ranking isn’t a sound way to resolve ties).
It would be better to use ranked voting first, then range voting to resolve tie-breaks. That would make better use of the unique advantages of each voting style.
Yes, STAR mixes rank and range to create a balance.
I still don't buy your assertion that no one knows what scores mean. That's no stronger of an argument than the opposing critique of ranking: nobody knows in a rank ballot whether it's a slight preference between two supported candidates or a huge difference between a liked and a hated (but maybe some other candidate is even worse) situation.
In both ranking and scoring, the ballot does not give a transparent view of the voters' beliefs. They only mean what they mean on the ballot, that there are relative scores or a ranked order. We can't and shouldn't assert that we can know more deeply what's behind those marks.
Condorcet is not a bad principle per se, but it's not the absolutely most desirable. If it were the only factor, nobody would accept IRV. And Condorcet means that a candidate who is 2nd choice of 100% of the voters, maybe even a strong 2nd choice (rankings don't carry that info) will always lose even in a case where all the other candidates are grossly polarizing and will lead to civil war. There are at least cases where a consensus is superior to Condorcet.
> It would be better to use ranked voting first, then range voting to resolve tie-breaks.
How does this proposed idea work? What actual system are you suggesting that would do this? EDIT: I see this referenced in your other post.
Condorcet means that a candidate who is 2nd choice of 100% of the voters, maybe even a strong 2nd choice (rankings don't carry that info) will always lose even in a case where all the other candidates are grossly polarizing
That's not correct, if there are 4 or more candidates.
If there were exactly 3 candidates the split would have to be something like 51-49, and then the candidate with 100% second place would come second. But it's arguable whether that's a bad result, and it seems like a very unlikely scenario anyway. Even 10% of people voting that candidate in first place would be enough for them to win it.
Yes, you're totally right. I was writing too quickly and mixed up that sentence's point.
What I meant to express was that the 100% 2nd choice candidate will lose to Condorcet winner if there's 51% support for the Condorcet winner. I.e. majority support. That's true also for non-Condorcet systems. My point was that Condorcet systems are always also strict majoritarian, and there are reasonable arguments that majoritarianism can be undesireable (such as cases of polarizing slim-majority winner.
I think this improperly confuses an axiom with an argument. What I mean is, some of this comes down to a question that asks what your values are:
If, in a two-candidate election, one candidate gets more votes, should that election be the winner, even if the losing candidate has more passionate supporters?
In democratic elections that respect one-person/one-vote, the answer to that question is simply yes, because that is the definition of one-person/one-vote democracy. Even if there are 100 voters, and it's 51-49, and the 51 voters are lukewarm while the 49 voters are passionate, the candidate with 51 votes wins. No matter what.
The reason that matters is because it is actually fairly slippery to measure passion. If one person submits a 10 rating, and another submits a 7 rating, what does that mean? It could mean that one person is passionate and the other person is lukewarm. But it could also mean that one person is emphatic while the other person is meek. Or that one person is authoritarian and the other person is abused. The entire reason we have one-person/one-vote is to make sure that voters are treated equally, even if society is otherwise telling certain people that they are "less than".
And for those sorts of elections, Condorcet voting is basically perfect. All the flaws people attribute to "Condorcet methods" are only attributable to when there are multi-member Smith Sets.
Now there are other kinds of elections; elections where we don't follow the one-person/one-vote principle. Smaller electorates, non-governmental, self-governing organizations. In those cases, "consensus" is more possible without risking systemic abuse. And in those cases, it might be entirely appropriate to award the winner to the 49% candidate with more passionate support. In which case you would perhaps choose something other than Condorcet.
What is this "push ahead" crap? I have no idea how that's different from the Amazon analogy.
You should come to terms with the fact that whatever you mean is completely unintuitive to many/most of us, and that almost everyone will just this like they use Amazon. Whether or not the page says "it's like Amazon" or "it's not like Amazon".
Calling something crap doesn't encourage constructive discourse if you're looking to actually gain something from a discussion. But I can ignore that.
The ballots are proposed as labeled "no support" to "most support".
It isn't entirely unlike Amazon, it's like Amazon if your standard for what's a 5, like "the best book" is "the best of THESE candidates". All people need to do to get the system is care about their impact on the outcome of the election in order to realize that 5 doesn't mean "love" or some other semantic thing, it just means the top of the options.
And if people just use it like Amazon and some give no candidate a 5, the system doesn't fall apart, it works pretty well anyway.
STAR has never been used in a public election. At some point, we can't both argue about lack of evidence and against getting the evidence.
http://star.vote is not a controlled scientific survey or anything, but people overall seem to have no trouble using the full range.
In fact, in my anecdotal experience, even over years in the past when I told people about plain score voting and tried to get them not to use the full range (because I misunderstood that issue myself), many people insisted that they would use the full range anyway. A lot of people find it intuitive to give their favorite all the points they can.
But the real answer is that, yes, we need more studies and trials of all these things.
IRV has problems, but complexity, intuition, and explanation are not three of them.
Intuiting what "giving one star more or less to a candidate" means is really difficult to understand as a voter. "If more voters had given Kodos four stars instead of three, he could have won" is a headache inducing explanation.
Approval voting is the only alternative to IRV or plurality voting that seems to match the level of intuitiveness of both the voting process and the tabulation process.
The way to turn ballots into results is just a simple division, numerator (sum of scores) over denominator (number of voters). Try to explain to me the function from ballots to results in IRV with 5 candidates in as simple of terms? It's absolutely much harder to analyze IRV and predict what is going to happen. In every range voting scenario, I can show you a bunch of ballots and you could quickly figure out the smallest change required to make a different person win. With IRV, I imagine you'd first start by drawing a big tree of options, and then simulating every way each could play out...
I seriously doubt that anyone would get a headache over "if people had given that losing candidate higher scores, they would have won". I don't even see how you can honestly claim that to be anything but completely intuitive to everyone.
If IRV complexity and explanations were not a problem, we wouldn't see today's reality where IRV proponents themselves constantly make incorrect claims about it.
Not always (and not in the linked article here), but often people get the intuition that their 2nd choice will get counted when their 1st choice is eliminated, and then try to get them to understand why their 2nd choice was not counted (because their 2nd choice had already been eliminated) while someone else's 2nd choice was counted…
What happened in Burlington before they repealled IRV shows not only people getting intuitions about IRV wrong but people often don't even get what happened in Burlington because it seems so counter-intuitive. https://www.equal.vote/Burlington
And yet, that scenario is common enough. Burlington is just the case where all the ballot stats were released so we can study it.
Thanks for pointing to the Burlington election, that's a remarkable data set showing the theoretical worst case of IRV actually occurring in real life!
The problem, though, is one of correctness, not complexity -- I would maintain that almost anyone could look at that election summary and understand why the conclusion that was reached was reached. And if they had had a traditional non-instant runoff election, the result would likely have been the same. It's hard to even explain why the outcome was "wrong" because the tabulation of the results is so intuitive.
> I seriously doubt that anyone would get a headache over "if people had given that losing candidate higher scores, they would have won". I don't even see how you can honestly claim that to be anything but completely intuitive to everyone.
The question for the intuition is "how many stars to I give this candidate". What does it mean to give them one more star? One person's four-star vote carries exactly as much weight as four people's one-star vote, which seems incredibly odd. An Amazon product with a hundred one star reviews is intuitively worse than one with 10 five star reviews.
> if they had had a traditional non-instant runoff election, the result would likely have been the same
Certainly not! In Burlington, people know the Republican can't win. So, like everyone in the current system, people would (and have both before and since IRV) vote strategically. So, the less-hardcore Republicans would vote Democrat in order to stop the Progressive.
In fact, if IRV had not been repealed, a good portion of the Republicans would betray their favorite and vote Dem as 1st choice (dishonestly) in the future in order to stop the Progressive from ever winning again.
Furthermore, when vote splitting happens in a current system, everyone readily gets why it was "wrong" and they adapt their strategies accordingly.
> What does it mean to give them one more star?
It means they are one more point ahead of the others.
> One person's four-star vote carries exactly as much weight as four people's one-star vote, which seems incredibly odd
The weight is relative, that's all that matters. As long as people understand the system, they can realize that they are throwing away part of their vote if they only put a 1-star and nothing else. Four people who put 1-star for candidate A might have put 5 stars for candidate B. They are influencing the race just as much as anyone else.
Here's what it means to be equal: Any voter can score the opposite of another voter and they will cancel out. Our votes can have equal weight if that's possible. That's not always possible in our current system or with IRV.
> An Amazon product with a hundred one star reviews is intuitively worse than one with 10 five star reviews.
If there were 110 voters, then every item is relative to that vote. There's 100 1-star votes for candidate A and 10 5-star votes for candidate B? That means candidate B got zeros (blanks are no points) from 100 candidates. A candidate with 100 zero-star votes and only 10 5-star votes is a worse option.
That Burlington result just doesn't look clearly wrong to me.
With FPTP voting, C would have won of the three candidates, and A would have placed last. If you removed A from the race, B would have won - and did!
Alternate voting systems are supposed to produce different results from plurality voting, and that's often unsettling the first time it happens.
That page makes the claim that IRV produced a bad result and that's why it was replaced. But at least the candidate with the 2nd-most 1st place votes won. If instead the Condorcet winner, A, had won, do they really think it would have been better? A got the _least_ 1st place votes of the three, who's to say there wouldn't have been an even larger backlash?
If you don't see what's wrong with the Burlington result, it just proves how hard it is for people to understand IRV.
FPTP is even worse than IRV, except that people know how it is bad so they adapt accordingly.
Of course, it's speculation about the specific backlash, but the core issue in IRV is this (and it happens in the Burlington case):
ALL the voters whose 1st choice loses in the final round never get any of their other preferences counted and they lose their 1st choice. They get NOTHING, no say, totally screwed. They could be as much as 49% of voters. Now, WITHIN IRV, they can later learn (as people have learned about strategy in FPTP) that if they betray their favorite and vote for their lesser-evil choice as 1st, then they WILL swing the election to their lesser-evil instead of the greater evil.
By ignoring the preferences of some voters and counting others, the weighting is unequal and people will feel disenfranchised.
The core point is that voters in IRV can get a preferable outcome via favorite betrayal (a dishonest strategy). Either they do use that strategy and we're back to the lesser-evil problems we have now, or they don't use that strategy and we're back to vote-splitting like we have now, where a candidate choosing to run can both lose and cause a worse outcome for their supporters.
> ALL the voters whose 1st choice loses in the final round never get any of their other preferences counted and they lose their 1st choice. They get NOTHING, no say, totally screwed. They could be as much as 49% of voters. Now, WITHIN IRV, they can later learn (as people have learned about strategy in FPTP) that if they betray their favorite and vote for their lesser-evil choice as 1st, then they WILL swing the election to their lesser-evil instead of the greater evil.
Another way of seeing this could be that such voters had their 1st choice considered and given value all the way until the last round, whereas many other voters had their lesser choices considered in their previous rounds. The 1rst-choicers-all-the-way-till-the-end could be considered more favored than the latter voters I mentioned, who lost their favorite candidates earlier on in the process.
I'm not saying that this is a better interpretation of the situation you posit. I'm just using it to illustrate that it's hard, maybe impossible, to get away from subjective criteria when we consider the virtues of various voting systems.
If electing the Condorcet winner in Burlington would have looked wrong, then that just underscores the problem of using a ranked voting method. You want a rated voting system like STAR, Score, or approval voting.
by the way, Fargo North Dakota just adopted approval voting for their City elections tonight.
It sounds like you’re assuming that all the voters would have chosen their top IRV candidate as their plurality candidate, but that’s almost certainly not the case.
Burlington may be an argument against IRV (as not superior to majority/runoff in that case), but its the whole description of why the intuition is wrong (loser elimination prevents seocnd-place votes from being considered if the candidate is the first-round loser) is an argument for Bucklin, not STAR.
Bucklin has its own different issues. If anyone were promoting Bucklin, we could get into why it's not a good option.
Within a strict ranked-ballot approach, the overall best tabulation is probably Ranked Pairs.
STAR has some similarities with Bucklin but does not have its problems. STAR's advantages are several and aren't directly just a forgone conclusion merely by showing what's wrong with IRV. STAR is much simpler to tabulate than Ranked Pairs and doesn't need to be tabulated at the highest level, it can be counted at precincts like FPTP.
All score methods are bad (especially in a large multicultural polity) because mapping from subjective opinion and preference to numeric scores is inconsistent and, particularly, highly culturally variable.
If there were an objective way to measure actual utilities, score-based methods would be great (and economics and policy assessment would be much simpler, and...)
There is no such mapping, especially with STAR. So, the critique of scoring based on mapping being variable is fundamentally a misunderstanding.
In STAR voting, you have up to 5 points to give to each candidate. You effectively get to push each candidate ahead in the race relative to the others by as many as 5 steps forward. There is NOTHING subjective about the meaning of points. They are just points. More points wins. They have no cultural signficance, no opinion, nothing subjective.
Yes, there's a mapping to a (typically 0-5) numerical rating like those popular in online review systems, which is a motivation for the forced acronym, and exactly the similarity that is viewed as making it particularly intuitive and accessible.
Cultural differences in how such rating systems are used for similar preferences have been studied considerably.
It's true that you can also view STAR as a limited bucket (and so forced-tie for any but the smallest candidate pool) ranked-ballot method, in which you force voters to arbitrarily select which intercandidate ranking preferences to suppress, and then do a completely wacky (from a ranked preference perspective) method of tallying the ballots.
It's unfortunate that the branding emphasizes the bad-to-good popular ratings metaphor. I agree that is a problem.
It is indeed a form of ranking that allows ties and has a limited resolution. But the limited resolution does help with simplicity.
From a UI perspective, presenting a ranked ballot that allows ties is unfortunately complicated. That's what STAR actually could/should be though.
And the tallying method is fine, not wacky. In a sense, it's like Borda Count to get to two finalists and then IRV for the finalists. That avoids all the worst aspects of IRV (which are all symptoms of the way its multi-round elimination ignores the preferences of many voters while counting the preferences of others).
> But the limited resolution does help with simplicity.
I disagree. Generally, unforced rankings (allowing full expression of preferences with optional ties) are simpler than forced rankings (with no ties) which in turn are simpler than limited bucket arrangements that force ties; “Between X and Y, do you prefer X, Y, or neither?” is the simplest question about voting preferences, and unforced preferences can be answered by answering that question alone; forced rankings may, as the name suggests, force the invention of an artificial preference in some cases, but limited buckets methods force deciding which preferences are most important and which to suppress. And limited buckets methods which apply something more than ordinal meanings to buckets,as STAR does, are most complicated.
> From a UI perspective, presenting a ranked ballot that allows ties is unfortunately complicated
Tallying unforced preference ballots is complicated, “number your preferences starting with ‘1’ for most preferred, ’2’ for next most, and so on, with equal preferences getting the same number” with a ballot with appropriate entry blanks isn't hard UI either in paper or machines.
Even if you want paper optical scan ballots, the main problem is the size of the ballot (which is a logistical problem when you've got multiple simultaneous races) more than the UI.
> And the tallying method is fine, not wacky.
Viewed as a preference ballot, giving non-ordinal meaning to rankings is wacky, since it is inventing information. The wacky description was in that specific “viewed as” context.
> In a sense, it's like Borda Count
I don't disagree with this characterization, though I view it as more of a condemnation than a defense.
> That avoids all the worst aspects of IRV (which are all symptoms of the way its multi-round elimination ignores the preferences of many voters while counting the preferences of others).
If you want to eliminate all the worst aspects of IRV, which you view as all due to ignoring preferences via loser elimination, the simple solution is just to drop loser elimination (giving Bucklin or a form of Majority Judgement, depending on where you use forced preferences or allow ties; classic Majority Judgement even uses limited buckets, though you can use a full unforced preference ballot, so if you really think the preference compression is justified by UI benefits, you can use exactly the same ballot as STAR, with a much more straightforward resolution procedure.)
This is certainly far simpler than STAR, whether or not you use the same ballot.
I think your points are reasonable and thoughtful, thanks for the constructive engagement.
You are right in noticing that it's not exclusively an unforced ranking (which I agree is vastly superior to forced rankings!). So, in STAR, there's a different between 3 candidates scored as 5,1,0 versus 5,4,0. They count identically in the runoff stage (which is effectively the rank part of the system). But they count different in the score.
I'm not convinced that difference is necessary, valuable, or intuitive to people, but I'm not convinced it isn't. I think these things need more real-world case studies. I definitely think the distinction isn't fatal or anything. I see how it could have a positive impact on outcomes.
I don't agree that multi-round eliminations are simpler than STAR per se.
I'm certainly inclined to see 3-2-1 voting as a competitor with STAR for best overall system.
There are many examples of voters in real-world IRV elections _do_ rank their favorite candidates last, to help them win (the same problem plurality voting has, but worse, because there are multiple plurality runoff votes).
Mathematical models also agree – situations in which IRV produces bad results are extremely common. Something like 70% of possible situations where IRV and plurality voting produce different results, the IRV result is worse than the plurality result.
Approval voting and range voting are really the only voting systems I know of that are strictly better than plurality. They have their flaws, but per Arrow's Impossibility Theorem, there aren't any voting systems without flaws.
Huh? There isn't a clear-cut way of determining the best voting system, so how were they compare the two? That matters quite a lot.
There are situations where IRV get so complicated it is unpredictable, but living in Australia that does sometimes feel like a feature - if the polity expresses no strong consensus preference, random selection amongst candidates is not a bad idea.
> Approval voting and range voting are ... strictly better
Yeah, totally [0]. Approval voting looks like the one to use just in terms of doing what a voting system does.
Any system that lets neutrals and moderates in occasionally is better than plurality. And plurality is streaks ahead of having no choice at all.
The resource is link is a very weak critique of IRV. consider their criticism of Aus 2007 general election.
https://rangevoting.org/Aus07.html
The key point which is laboured a couple of times is that some seat results are different than plurality outcomes. But that is a feature of IRV, not a bug.
Had the election been run under plurality rules voters would of voted differently and this fact seems to of escaped the author of this website.
Voters who preferenced greens first, recognising a minor parties zero chance of winning under plurality would of instead cast votes according to two party preferred lineup, who mostly would ideologically align with labour. But under IRV they can vote for anyone they feel more closely represents them but still have meaningful input into a 2PP outcome.
Show me a critique of an IRV outcome where 2PP outcome violates Monotonicity and you'll have a real critique of IRV.
Agreed. The lived experience in Australia (which has now used "IRV" in national elections for 100 years) is that in almost every case, the IRV winner was also the likely Condorcet winner (which certainly can not be said of first-past-the-post in practice).
However, in close three-cornered races, tactical voting does sometimes come into play. The recent Wentworth by-election was an example of this: Liberal voters could have assisted their party to win the election by switching their first preference to the Labor candidate, because if the Labor candidate had beaten the eventual winner (Phelps, Independent) into second place, then Phelps' preferences would have been distributed and almost certainly given the win to the Liberal. Of course, this is a risky strategy, because if too many Liberal voters did this then they would have instead engineered a Labor win. And in reality, despite the possibility of tactical voting, the Condorcet winner Phelps won anyway.
An example where the Condorcet winner may not have won is the seat of Rockhampton in the 2017 Queensland State election. The primary votes had Labor in first, independent Margaret Strelow in second, One Nation in third and the LNP in fourth. However preferences from the LNP candidate pushed One Nation into second, which gave Labor the win on Strelow's preferences. It seems likely that Strelow would have beaten any of the other three head-to-head, though. And as previously mentioned, this sort of situation is notable for its rarity.
Tactical voting is only a concern if people don't fully preference, which I always do (and encourage others to do). Surely if you can be bothered to vote "tactically" you can be bothered to fully preference
The given example (Wentworth) was a House of Reps by-election, in which full preferences are required for a vote to be formal anyway.
That doesn't affect the tactical voting possibility there - the possibility arose because the Liberal candidate's chances were greatly affected by which of the other two major contenders made it into the final runoff.
>Compare that to to some of the other systems which often have complex mathematical descriptions which are hard for people to wrap their heads around.
Similar issue for end-to-end auditable homomorphically encrypted voting systems. Yes, anyone can check the math in theory ... but not the typical voter.
But does the typical voter have anybody they trust who _can_ check the math? It's analogous to people trusting FOSS software despite lacking the skills or time to check the code -- they trust it because they trust people who are able to check the code.
>Runoffs are something people understand, IRV is conceptually similar and makes intuitive sense to people, it's very easy to explain to people, 'if nobody gets a majority, eliminate last place and redistribute their votes, repeat'.
> Compare that to to some of the other systems which often have complex mathematical descriptions which are hard for people to wrap their heads around.
> Another thing is how big a deal is 'non-monotonicity' is it actually going to be something that will affect elections all the time, or is it something that's more minor than say turnout which honestly has no solution and on one side you have USA where turnout is low so the game is all about making sure your base turns out, or on the other side you have Australia where voting is mandatory so making sure people who don't care, don't vote at random is a big part of the game.
> Runoffs are something people understand, IRV is conceptually similar and makes intuitive sense to people, it's very easy to explain to people, 'if nobody gets a majority, eliminate last place and redistribute their votes, repeat'.
Compare that to to some of the other systems which often have complex mathematical descriptions which are hard for people to wrap their heads around.
It's fascinating how a similar constraint arises in industry software engineer; more than fascinating
I've never understood the criticism that Condorcet is hard to understand. If one candidate would beat all others head-to-head, that candidate wins. That's it.
The weirdness comes in the strange tiebreaker methods if there's a "loop", but when you have a lot of voters, loops are rare.
(Condorcet proponents would probably get more mileage if they had a better conceptual split between methods that identify the Smith Set, and methods that seek to award single winners of multi-member Smith Sets - those "tiebreakers" are where all the problems happen, and none of those flaws are relevant in elections where there is a Condorcet Winner.)
> One property of IRV that makes it so questionable is its non-monotonicity. This means that you can help a candidate by ranking them lower, and hurt a candidate by ranking them higher.
Could anyone take a stab at explaining how this could happen? I looked at the linked site but it's over my head. The visualizations aren't doing anything for me.
The simple explanation (I hope) is that by ranking a candidate lower, you can promote a second candidate so that they survive elimination instead eliminating a third candidate, which would otherwise have defeated your first candidate in a later round where the second candidate will not.
It is a consequence of the loser-elimination step. You can avoid it by eliminating loser elimination, which gives the (unlimited ranks variant of) Bucklin (with ties and truncated ballots allowed, essentially Majority Judgement) voting system, which has its own pros and cons. I kind of lean towards unlimited ranks, ties-allowed Bucklin—count first-place votes, if you have a candidate with a majority of ballots cast, they win, if not add all candidates second place votes and see if you have a majority (of ballots, not counted votes which will now exceed ballots), etc., with a tie-breaking procedure if you get two majorities in the same round (bigger majority, either in current or subsequent rounds, random selection of the tie persists when ballot preferences are exhausted)—is my preferred simple ranked-choice system. Condorcet methods arguably choose a better winner in the case where they disagree, but are less simple.
The charts on this website are eye opening. Even in 4 candidates we can see very strange behaviour from IRV. I initially didn't understand why IRV would be non monotonic, but reading some examples it can be seen that the results of IRV can drastically change depending on who got the lowest votes. The graphs show several weird behaviors for IRV and I would recommend anyone interested in voting systems to have a look.
Agreed. For me it seemed like IRV would be fairly stable, and I was surprised by how strange it was.
My updated intuition now says that IRV means that as long as your top candidate is in first place, your other preferences are disregarded by the system. "My #2 was eliminated too early -- if I had known that my #1 would be eliminated in the end I would have ranked #2 higher".
Agreed. That said, if you're interested in pushing something that is better than IRV and that people will actually use, I'd suggest approval voting (or possibly range voting or STAR). I don't really expect the Scholze method to ever see wide adoption.
Basically, I suggest pushing approval voting over other methods for realistic cases.
My favorite for single winner elections is range voting with top-2 runoff, actually. But I at least find all of these to be acceptable. Whether or not the trade-offs of range / approval voting vs condorcet methods are better seems to depend on whether you think of utility as an ordinal or a cardinal thing.
Your zesty link detracts from the conversation. The zesty link makes a simplifying assumption, that under first past the post (called Plurality) voters vote for the nearest candidate according to some distance.
For example, if 40% believe in A and 60% believe in B, and there one candidate, X, standing on platform A, and two candidates, Y and Z, standing on minor variants of platform B, the votes split 30% for Y, 30% for Z, and X gets elected with 40%.
It is easy to see how the 2016 presidential election would have worked out if this assumption was realistic. Hillary 52%, Donald(Republican) 24%, Never-Trump(Republican) 24%. Hillary wins by a landslide, despite playing a sub-optimal electoral college game.
The assumption is very far from realistic. I next turn to whether the lack of realism matters.
Imagine a three way race, with two old established parties, A and B, and a new third party, C. Under first past the post voters who prefer C > B > A, vote B, to avoid letting A in. Meanwhile voters who prefer C > A > B vote A to avoid letting B in. Under first past the post (FPTP) it is very hard to start a new party.
We might frame the issue this way: who decides the fate of the third party candidate? Under FPTP the third party candidate loses and this decision was taken long ago, by men now dead, when they opted for FPTP. Under IRV the fate of the third party depends on today's voters. If they are voting C > B > A and C > A > B then the third party candidate stages an upset victory. If they are voting B > A > C and A > B > C then the third party candidate is humiliated and doesn't bother running again.
The argument for IRV over FPTP is that we prefer that the fate of the third party candidate be in the hands of the voters, rather than seeing the third party candidate lose automatically, by the game theory of FPTP.
The zesty link simplifies away the problem that IRV is intended to solve. But it doesn't argue that the problem is small; it cannot do that because every-one can see that the spoiler effect under FPTP is a big issue. Instead it refuses to mention the problem, simplifies it away, and hopes that nobody notices. That is derailing, not contributing.
I don't have time to jump down that rabbit-hole, but it seems to me, the best way would be for each voter to have two votes available to use for each candidate - one positive, as normal, the other negative, meaning, basically, "anyone but this idiot"... Voters could use one, both, or neither.
It's funny we're using this exact voting method to rank the comments in this conversation. Imagine promoting IRV for this: "Everyone just sort all of the comments into your preferred order, and then we'll do these rounds of selection based on that data to pick the top comment. Simple." Sorry, I'm being very unfair -- politics has different tradeoffs -- but the idea amuses me.
Restricting the number of votes each voter can cast causes the same problems as FPTP, the so called "spoiler effect". Ideally a voting system tries to get information from each voter about their opinion of each candidate. Why prematurely throw away most of that information?
However, a system similar enough to your example has been used for years by the Quakers, where each voter gives a score of +1, 0, or -1 to each option/candidate. The difference here is that you can give +1 to multiple candidates, and -1 to multiple candidates. Giving your opinion on one candidate in no way limits your ability to give opinion on other candidates.
You know, reading about Ranked Pair voting you have just made a convert of me. This is a fantastic method.
You'll never be able to sell it to the general public, of course, but if you could do it for just one election I think you'd have everyone on board just from the results of it.
I wish that the simulator worked in reverse, i.e. instead of showing for which center-of-opinions a given candidate would win, it would show given a distribution-of-opinion, how should a candidate position themselves to win?
This is a much more interesting question to me, first because the candidates' positions are more mutable than the voters' opinions and second because as someone whose political opinions are often non-aligned with major parties, I really wish that there was an incentive for viable 3rd-party candidates to give a voice to my views. The ncase.me simulator linked elsewhere on this thread seems more useful:
Almost all ranked methods use the same kind of ranked ballots with differences in tabulation, differing occasionally in whether ties or truncated ballots are allowed.
Range/Score (and STAR, which is a variation) use numeric score rankings on a preset range, and some methods use a limited set of ranks rather than (allowing voters to use up to) as many as candidates (Approval can be viewed as two-rank system.)
Condorcet is definitely an important property to have. But you also need a solution when there isn’t an outright Condorcet winner, and there is no good solution to that. The tie-breaking mechanism of every Condorcet method is basically impossible to explain in simple terms, ranked pairs included.
I often wonder if we should just use Condorcet and if there isn’t a clear winner, declare the result a tie. (That would likely be vulnerable to a kind of DoS attack, though.)
> But you also need a solution when there isn’t an outright Condorcet winner, and there is no good solution to that. The tie-breaking mechanism of every Condorcet method is basically impossible to explain in simple terms, ranked pairs included.
The Smith set is as easy to explain as the Condorcet winner, and random selection is trivial, so random selection from the Smith set seems to be a simple Condorcet tiebreaker.
Limited to the Smith set, but sure. Same as lots of places do with FPTP and a tie; no Condorcet winner is effectively a “Condorcet tie” between the members of the Smith set, each of whom beats every candidate outside the set, but where victory within the set is undeterminable.
Yes, UK elections use coin tosses on ties. It’s pretty rare to get an exact tie when thousands of people are voting, though, and it’s also hard to argue against the coin toss when literally nothing separates the two candidates.
The Smith set tie would occur more often, and in complex scenarios where it really isn’t clear that it actually is a dead heat.
You can do other tiebreakers in the Smith set, like plurality first-rank winner in the Smith set (proceeding down preference ranks, with random selection only if you fall off the bottom without resolving the tie.)
Indeed, if you allow randomness, you can get all the other desired attributes listed in Arrow’s impossibility theorem. Random ballot voting, for instance, has nearly every desirable property. There’s no reason to vote strategically (dishonestly). The race between candidates A and B can’t be affected by the presence or absence of any third candidate (there are no spoiler candidates). At least one voter is guaranteed to be happy with the election results (ok, that one is sort of a joke).
But of course, random ballot has just one pesky undesirable property: the fact that it’s random.
I kinda see range voting as the solution. To explain my intuition... Technically it's not condorcet, because a pair-wise winner doesn't always win, but if you relax the definition of "pair-wise winner" to allow for each voter to scale down the strength (weight) of their vote, then range voting does in fact always select the candidate which wins all weighted-pairwise competitions. I think whether you buy this intuition has to do with whether you think preference/utility is fundamentally ordinal or cardinal.
Your intuition is in fact correct. Balanski and Laraki have shown that a median based rating system, in the limit where all voters are rational and their utilities are maximized by electing their desired candidate, that the winner is always a Condorcet winner. This is one of the reasons that Majority Judgement (their name for this system) has so much going for it.
Yeah, I've thought similarly. I actually consider multi-candidate Smith Sets a feature though, not a bug. Because I believe it means the electorate is honestly confused and not ready to make a decision yet.
So one option is to hold a runoff between the candidates in the Smith Set.
Another is crazier, but I haven't found a flaw yet. Basically introduce range/score voting on the ballot, but only use the scores if there is a multi-member Smith Set. That way you guarantee Condorcet compliance. And since the voters don't know if the ranges/scores will be used (and they probably won't), they don't have incentive to strategically score a less-desired candidate ahead of a more-desired candidate.
I like this idea the more I think about it, and I think it compares very well to methods being discussed elsewhere in this thread such as STAR (see https://news.ycombinator.com/item?id=18397839)
I just thought of a way to possibly make it more intuitive. Use ranked voting, with the addition of a “nobody below this line” item that you can add to your ranking anywhere you like. Then at the runoff stage, you use approval voting, with “nobody below here” determining which candidates you approve.
That gives you a nice easy-to-understand runoff, and is maybe even more understandable than normal ranked voting because you get to explicitly separate out the candidates you really can’t stomach. I suspect some people don’t like ranked voting because you’re expected to stack-rank lots of candidates you dislike.
I'm convinced that a "None of the Others" mark is pretty essential to making any ranked system good.
But we're talking a lot of complexity to have both ranks and scores as separate marks. Super long ballot, will turn away most voters and seem really hard to process.
It wouldn't have to be both ranks and scores. It could just be scores. But you then order the scores numerically to infer the rank. And you use the rank to find the Condorcet Winner. If there is no CW, then you use the scores from the candidates in the Smith Set to identify the winner from that Smith Set.
(The point of the above was to respect the Condorcet Criterion, and 3-2-1 voting does not meet that criterion.)
It's really all the mistakes that even proponents of IRV are making about how it behaves that really keeps me up at night. I'm not convinced voters understand the system well enough to vote optimally. This effect and how it results in bizarre strategic voting is enough for me to consider IRV to be worse than FPTP.
Yes, that is a property of IRV, but in a real life situation involving hundreds of millions of voters, does this situation have an even slight chance of happening?
IMO the biggest problem with IRV is that it bypasses the root problem, which is that we assume we are electing a single person at all.
It's entirely possible (even within the confines of the US Constitution!) to virtually never do this. Elect the House of Representatives by statewide party list. Senate is hard because each state has two senators elected non-simultaneously; otherwise statewide STV would work. Since the President is already elected by the Electoral Congress, skip the popular Presidential election and select the Electoral Congress by sortition among all eligible registered voters in a state (similar to jury selection except without being able to arbitrarily send people home) and have them personally interview the candidates before voting their personal consciences.
Maybe IRV got an early start as a poster child of alternative voting systems, thus it's popularity. In any case, I would argue along with other commenters that IRV's run-offs would probably be far easier to understand for the general electorate than other, more complex vote-processing systems. I would very much disagree with your (xvedejas's) characterization of ranked pairs as being simpler than IRV, though range voting would indeed seem easier to explain and implement than IRV. A math major (or at least a minor) seems like a requirement for someone to understand the ranked pair system.
I was unfamiliar with the ranked pairs system, and will study it more to understand it better. It looks promising as far as technical excellence, though I'm afraid it'd still be too hard to understand for the general populace in most (or all) countries.
One thing I suspect about range voting is that it seems to require strategic voting. A voter would still need to give a very favorable score to a majority party that's kind of aligned with their views, as opposed to giving them a lower score than a minority/third party he or she would prefer. One of the strongest appeals to me from IRV is the opportunity it provides for a minority party to unseat majority parties, which is appealing given the political stagnation experienced by many countries. I had also thought that it allowed voters to avoid strategic voting and express their true, individual voting preferences.
After reading IRV's description in http://zesty.ca/voting/sim/ , I realized that I had misunderstood one important detail. I had thought that the least popular candidates were eliminated first until only the top two candidates remained, and the one with most votes among them remained. I dislike that the "real" IRV (not the one that was in my head) first checks for a 50%-plus-one majority before eliminating the bottom candidates. That feels too much like the first-past-the-post system for my liking. I wonder if having successive rounds where the least-favorite first place candidate is eliminated would make most people's favorite candidates "bubble up" to the top, producing a final match between two highly-favored candidates. I wonder if such a system is closer to a Condorcet method than the real IRV...
> I dislike that the "real" IRV first checks for a 50%-plus-one majority before eliminating the bottom candidates.
Suppose that a candidate has a majority at the beginning. Then this continues to be the case as other candidates are eliminated. Thus, this majority candidate will be elected at the end. So this extra "if-clause" only speeds up computation, it does not change the voting rule.
I think this post is really just arguing for some kind of ranked choice voting system to better support third party candidates. You’re criticizing a specific implementation of ranked choice, but I think the author was just using IRV to represent the general concept. It’s possible they are not aware of the technical differences.
Could you elaborate on why you think plurality is better than IRV?
Every time I see something along these lines, I think of a discussion between people debating which base OS to use to build a platform to power billions of mobile phones: Plan 9, or BeOS?
It doesn’t matter. It ain’t gonna happen. Not because IRV or condorcet is or isn’t better, but simply because of the installed base. Getting literally everyone to a) understand the subtle differences and then b) agree to switch and c) agree to switch to the same system is impossible because literally every step of the process is itself impossible.
This is not me being pessimistic, this is me being realistic.
Humorously, a top down dictatorial approach would be the only way to “improve” (by some estimations) the current democratic implementation. You’d have to not care about democracy to upgrade democracy.
Yes, it's impossible to convince a large number of people to switch, which perfectly explains why Maine is still using FPtP despite the large push for RCV a few years ago.
/s
Proof by contradiction, mate. It's not impossible, just hard. There's plenty of things that once we're said to be impossible, and the data agreed... until they happened. Everything had to have had a first time.
BTW if you want to play with the model of IRV voting shown on that website (so you can get an idea of when the non-monotonic results hit, etc), I threw together a jsfiddle:
Straight IRV is an improvement in that it removes the third-party candidate spoiler effect and the incentive for negative campaigning, yes, but it can actually make results less proportional than FPTP.
What's needed isn't just IRV but STV - think IRV but districts grouped together and multiple winners. That way, you can actually get results that are proportional to the votes cast because the difference between a narrow win and a landslide changes who gets elected down the ticket.
The multi-winner system that appeals the most to me is either range (score) or approval voting, with seats allocated proportionally. Unlike STV, it would be possible to support or oppose as many candidates/parties as the voter wants simultaneously. And so it avoids some of the nasty properties of IRV and STV.
There's no problem with opposing candidates in IRV / STV - you just rank them at the back of the list. I can't claim to be an expert on all the finer points of STV versus range or approval voting, but it's significant to me that the Electoral Reform Society (https://en.wikipedia.org/wiki/Electoral_Reform_Society) continue to back STV.
It's not obvious to me that all the information about your ranking gets percolated up to the final decision, especially since your middle rankings can disappear without warning during intermediate stages. If you can't do anything to prevent your #2 from getting eliminated, might as well rank them last as far as the system matters.
One very important part of getting STV to work well is the quota - if a candidate has enough votes to guarantee that they would be elected (so, for a 3-seat district, 25%+1 of votes) then additional votes should be transferred BEFORE eliminating any candidates.
Some systems do this randomly (randomly choose ballots and take the next preference), and some do this fractionally (e.g. transferred votes are worth whatever fraction wasn't needed to elect their previous choices). Overall, though, the point is that if your middle choice is eliminated, then either your vote wasn't enough, or a previous choice needed your support already.
That said, most systems 'skip' over previously elected candidates as well. Strictly speaking, doing that can mean ballots with a losing first candidate are more powerful than ballots where the first candidate is elected by a narrow margin[1]. That said, as per the wiki link, there's a solution (recalculate all instead of skipping previously elected) and tactical voting around that issue isn't seen in practice anyway.
There's a big unaddressed obstacle to IRV which is that it's a little confusing and complicated to explain. This is a much bigger problem than it might first sound like. It is an absolute requirement of any voting system that voters understand how it works and feel confident in it.
This is deeply unfortunate if it's really true, because the prime factor in favor of IRV is that it is the second-easiest system to explain after FPTP.
But it's harder to understand consequences, which change over time. So the explanation is deceptively simple. Like the game of life. Random Ballot is simple to explain and is a lottery pool. Simpler than IRV.
Random ballot works just as well for Senate elections, and votes for Presidential electors (or even votes by Presidential electors) as it does for House votes.
Which is, to be sure, not very well, but there's nothing special about the House that makes it work there better.
IRV is nowhere near the second easiest to explain.
Approval voting: you can vote for any of the candidates you like, not just one.
Way simpler than IRV.
Score voting: give any or all candidates scores.
Easy explanation.
STAR: score the candidates, there's an auto-runoff where the two highest-score candidates get checked to see which is preferred by more voters.
At least as easy to explain as IRV, and much easier to actually understand overall, particularly in making sense of the results and following what happened.
Your vote is NOT applied to each candidate in turn. That happens to SOME voters sometimes. But you can also have your 1st choice eliminated after your 2nd choice, and then your 2nd choice is NEVER counted.
IRV is strictly less simple (both to explain and to implement) than Bucklin (or Majority Judgement), since Bucklin is IRV without loser elimination, and MJ is (or can be implemented as) Bucklin with ties allowed.
IRV (and any preferential voting system) is quite close to the voting system for any of the weekly elimination talent shows. You rank people in order, the last one gets knocked out, you try again - with the difference that this all happens at once.
Are we really saying that people are too dim to rank their choices in order of preference, or understand the voting system behind some of the most popular TV formats worldwide for the past decade?
Maine has put a whole lot of effort into making our RCV based ballot understandable, including providing this sample that clearly outlines how it works and several options:
ncase.me is very good, but it neglects what I consider to be the elephant in the room - dictatorship of the majority.
If you have a country with 60% A, 30% B and 10% C, then you might like to a parliament roughly representative of that. (A, B, and C could be anything - attitudes to same sex marriage, daylight saving - there are always many ABC sets.) If parliament consists people elected in geographic regions, and A, B and C are uniformly distributed over the regions than all voting systems here will give you a parliament consisting only of A's.
People here seem to enjoy pointing out some perverse outcomes for IRV, but the Arrow's impossibility theorem proves (and ncase.me beautifully demonstrates) all voting systems have perverse outcomes. Yet none are as bad as electing all A's to represent a diverse country and every one of them will do that if used the wrong way, which also happens to be the most common way they are used.
Back to the topic - IRV does have its downsides, but it's not perverse outcomes. Australia tries to use IRV everywhere. The Australia Federal Senate has a voting system that does fix the dictatorship of the majority and naturally they try to use IRV there too. If you ever have the unfortunate experience of voting in an Australian Senate election you'll see why IRV isn't so good in that situation. In fact it's so bad not even the Australian's can use it in its pure form. If you are wondering why, it leads asking the public to number 100 odd boxes in strictly consecutive order, on a 1m wide ballot paper that can't lie flat in the ballot box with writing so small they had to issue magnifying glasses so you could read the candidate names. https://www.abc.net.au/news/6870332
Sadly, voting terminology is nasty. I've long advocated for a better voting system than first-past-the-post (or "pick one", as I call it for non-voting-system nerds). IRV has issues and is not my first choice, but it is better than FPTP. That said, the major objection I get from people is "anything else is too hard".
So I say "can you list your choices in order of preference?" They almost always say that's acceptable. You can run IRV from that, but you can also run most any Condorcet variant. The only remaining part is to show the results in a fashion people understand, which is non-trivial but is also not what the "anything else is too hard" group was complaining about.
So don't say "ranked pair" or "STV"...say "order of preference", and you'll find people more willing to consider ideas. It's not EASY, but it's definitely EASIER.
We use FPTP in the UK because it directly correlates an MP with a constituency. Even if the candidate you didn't prefer wins, they are still your MP and have a duty to represent you.
We had a referendum on PR a few years ago (which was rejected) and its proponents kept banging on about it, then went very very quiet when UKIP got 4 million votes (more than the SNP and the Greens combined).
FPTP is not the only system to do that - if the vote is to decide who represents _you_, any system can do that. If the vote is instead to give the party a certain amount of power, it won't do as you want regardless of the voting system.
The people vote for their representatives directly, and the party with the most has the opportunity but not the obligation to form the government. The alternative, PR, breaks the link between constituency and representative. That's why very few wanted it - generally those who thought it would favour their party (Lib Dems) who are unable to muster much support at the local level.
I assume that PR means proportional representation. Instant runoff voting does not necessitate proportional representation. You can have an instant-runoff vote to elect a single MP for your constituency. All you do is rank the candidates in order of preference rather than picking just one of them. There will be one winner, and that's your MP.
I suppose it is fair to call FPTP "pick one" in some contexts, with people who aren't familiar with different voting systems, but it is possible for a proportional voting system to be implemented with "pick one" ballots.
A relatively unknown voting system is Direct Party and Representative Voting:
which basically involves giving voters two ballot papers at each election: one a simple "pick one" for the local candidate (counted like a normal FPTP election), and the other a "pick one" for the party that the voter prefers at the national level (if any). The proportions for these second ballots are used to adjust the "voting power" of the MPs or Representatives that were elected for each constituency / district by the first ballots. For example, each politician from an under-represented party could be given effectively two votes in the parliament / congress.
An even simpler approach (which is a variant of Mixed-member Proportional Representation) would be to use the same ballot papers (and constituencies) as now, and produce the same set of winners as a FPTP counting process produces, but add an extra 20% of representatives to parliament / congress to readjust the composition of that house to reflect the national vote totals. In particular, these extra representatives should come not from a party list (since those give parties the ability to give seats to unpopular candidates) but from the "best near-winners", that is, the candidates who did best in their elections without winning the plurality vote. Such a system is already implemented in Baden-Württemberg.
Both of these systems have a clear benefit over Ranked, Score, and even Approval voting, which is that the vote counting process involves putting each piece of paper into one of a few piles, once. There are no extra rounds, and the only information that needs to be communicated for collating is the number of votes for each candidate (which correlates well with the size of the respective piles).
Many people talk about proportional representation; my favorite PR system is the "Random Ballot." It is a very simple voting system and is the only voting system I know of that never suggests strategic voting (when voting in a way different from your preferences is required to get your preferred outcome).
Basically if you have e.g. 100 seats in your house, you choose 100 ballots at random and whomever is voted for on those ballots are the representatives. It averages out to proportional representation, while still allowing geographically distributed candidates.
It also satisfies all of Arrow's criteria[1], which no deterministic system can.
> It also satisfies all of Arrow's criteria, which no deterministic system can.
Approval voting is deterministic and satisfies all of Arrow's criteria. Arrow's theorem doesn't apply to voting systems, such as approval voting, that are not ranked choice systems.
I used to be in favour of this one, and talk about it all the time, but I've kinda changed my view.
The core problem with most voting systems could be solved with more representatives and just straight proportional representation. In the range of 1 to 10k–100k. Then you do a parliamentary system with one primary chamber that does most of the law making and one secondary chamber that does some checks but is ultimately toothless. A supreme court that can make law, but tends to refer legislation back to the parliament. A clause to allow the primary legislator to override all but the most basic freedoms required for democracy to work like voting and a free press.
Essentially Canada, but with proportional representation and maybe double the number of MPs.
The problem with random ballots is that unserious people will will or vote for themselves and they're harder to do securely with straight paper ballots that is understandable to laypersons.
My ideal voting system would be similar to this but without the actual voting part. Just take a sample of your population until you're some agreed-upon percentage confident you know the preference breakdown of your population and just go with that.
If you're going to randomly sample ballots just randomly sample voters and skip the line.
> Some publications even blame third party voters for Hillary Clinton's loss in 2016.
I don't think there is any plausible reason to believe Clinton would have won in 2016 if not for third-party candidates running. The math just doesn't add up.
The raw numbers are there for third party candidate to make a difference. It would only have taken Michigan, Pennsylvania, and Wisconsin going to Clinton instead of Trump to have changed the result.
In Michigan, Trump won by around 11k votes, and there were around 223k third party votes (172k Johnson, 51k Stein).
In Wisconsin, Trump won by 22k votes. Johnson got 107k there and Stein got 31k.
In Pennsylvania, Trump won by 46k. 147k went to Johnson and 50k to Stein.
I'd guess that almost all Stein voters would pick Clinton over Trump, so take Stein out of the race and in all three of those states Clinton wins that statewide vote unless the Stein voters decide to stay home. I'd guess that those who would vote for Stein would be sufficiently "anyone but Trump" to not stay home.
Those states are all winner-takes-all for the Electoral College, so that would do it. (Maine and Nebraska are the only states that split their electoral votes).
What would have happened without Johnson is a harder call. Trump and Clinton are both terrible from a Libertarian point of view. I'd expect Libertarian voters to be more likely to go Republican than Democrat if there is no Libertarian on the ballot most years, but Trump is so different from what had been mainstream Republicans before that I have no idea who the majority of Libertarians would have picked.
Also, I suspect that many of the Johnson voters were "never Trump" Republicans rather than Libertarians, who would agree with conservative writer and research fellow at the Cato Institute P.J. O'Rourke who said that Clinton lies about everything and is wrong on every issue, but he was endorsing her over Trump because she was wrong "within normal parameters".
The option of "just stay home" is an option that's not being considered by the OP. Plenty of people cast a protest vote for Stein against both Clinton and Trump and wouldn't have voted for either if those were the only choices.
Stein was the Green Party candidate, and Trump was pretty much the anti-Green candidate. Clinton ranges from agreeing with the Green party to indifferent on most of their major issues.
Even where Trump and Stein agree, such as on getting rid of Obamacare, they are massively incompatible (Stein wants to get rid of it for a single-payer system, Trump for just letting the market work it out).
Back in the day, the Georgia Tech Linux Users' Group had a very simple two-round system. The first round was approval voting; the second round was a run-off between the top two candidates. We voted like this every week to decide where to eat dinner. This extremely simple system satisfies participation, monotonicity and a weak form of Condorcet, where the winner always won pairwise against other candidates in terms of expressed preferences (but not all preferences are expressed). The participation criterion and Condorcet criterion are incompatible in the full sense, so this is as close as you can get. The run-off acts as a correction to AV's tendency to select underqualified moderates.
However, it is a two-round election. A run-off is always necessary. But it can be made one-round via what I call rank-and-star: voters rank candidates by preference, and put a star next to the top N that they like. The two candidates with the most stars are then compared on rankings. It's pretty easy to prove this is mathematically equivalent to the two-round system, but it's not a ranked-choice method because the ballots (1-2-3) and (1-2-3) are different.
Interestingly, it might still be possible to satisfy both participation and monotonicity if you expand the run-off to three candidates and use minimax Condorcet for the final round. For three candidates, minimax = Schulze = ranked-pairs and they satisfy participation; for four or more candidates, no Condorcet method satisfies participation. However, doing this costs you the attractive simplicity of the method (but minimax Condorcet is simple to understand, relative to IRV or ranked-pairs).
Anyway, this puts me in the odd position of advocating a voting system nobody has ever heard of =p
Maine voted on using this system for some elections in 2016.
It passed, but -- perhaps not surprisingly -- many current, established politicians weren't happy. Maine's governor called it "the most horrific thing in the world" and the state legislature refused to fund implementation and voted to delay it until 2022.
So this year the the citizens of Maine voted again to confirm instant runoff voting, overruling the state legislature and vetoing their delay. It passed by an even bigger margin this time.
Maine's gov, Paul le page, was elected when 2 vanity candidates split the moderate & left vote. He wouldn't have been elected under an IRV system. No wonder he doesn't like it.
In any case, he's leaving office and leaving Maine, so who cares what he thinks?
If you want to see how it is complicated, review https://www.rangevoting.org/Burlington.html and (A) see if you even understand (I've seen people literally deny the basic facts or miss the point if they happen to feel politically happy with the outcome in that one example), (B) see if you can understand the situation enough to explain to someone else what happened and why people see it as a problem.
Simply running IRV is easy enough. Understanding it and its ramifications is complex enough that even the majority of IRV advocates don't understand it.
Well of course, Arrow's theorem shows that you can have pathologies in every system. This is also true of every apportionment system as well. As we live in the real world, the point is to choose voting systems that have the best outcome. What are the criteria?
- people need to believe it is fair.
- it should avoid lop sided results (one party always wins regardless of vote) yet should avoid excessive diffusion (i.e. you don't want to have only fragile coalitions of many small parties).
- others, all with the aim of clarity.
As there is no magic bullet people use combinations of different systems:
- Australian ballot (secret ballot) which most systems use these days; hard to enforce in USA because of first amendment.
- quasi-independent apportionment processes/committees (as used in CA, AZ et al)
- party lists (not used in USA) or party-plus-candidate (e.g. Germany)
- non-party systems (e.g. CA)
- multi-candidate districts (not used in the US)
- proportional representation (most democracies) or first-past-the-post (most dependents of Westminster). I'm a fan of the latter BTW.
- IRV
Mix, match, iterate until an overwhelming majority of people are satisfied.
Arrow's theorem only applies to ranked ballots. But there's another concept that still shows how nothing can be perfect.
But really, the question isn't whether pathologies are possible ever, the question is how likely they are. What systems have the best balance to be the most robust while still being practical? STAR Voting and 3-2-1 voting. Definitely not IRV. IRV's pathologies have high enough probability that they do and will continue to show up significantly.
Good starting reference from one of the leading voting theorists:
STAR voting is good, but it makes the (IMO) mistake of taking the average rating as each candidate's score. I think it's absolutely crucial in a voting system to make it resilient to manipulation, and that means that taking the median rating (as in Majority Judgement) is the way to go.
STAR does NOT make that mistake. I mean, it literally doesn't. It uses the SUM of scores, NOT the mean or the median. But sum is effectively the same result as taking the median.
Well, the main concern most people have with mean is if it ignores abstentions. Sum does not ignore abstentions, they are the same as a marked 0. But you're right that sum is like mean if mean counts all the abstentions as 0's.
I'm open to the idea that actually using median is superior. Could you expound on that?
You're just picking one element. People have to understand and believe the system. A system in which you weigh the full set of a:b, a:c, b:c would be too tedious for anything but the board of the mathematical society.
If we could only get those pesky humans out of the loop...but we don't.
I don't understand why this follows my comment. I agree with you about that weighing challenge, about the humanity…
But I don't know what you mean about "picking one element" or understanding or believing. Both STAR and 3-2-1 are quite understandable, believable, and don't require weighing full sets of pairings. That weighing you describe sounds like how Ranked Pairs version of ranked-ballots is done.
But I chose that link in this case, despite its problems, because it more strongly emphasizes the complexities.
FairVote has engaged in really intellectually dishonest arguing, but the level of over-the-top reaction that the rangevoting site does in response comes across poorly and is regretful.
Having voters rank all the candidates from most to least desirable is a great idea. The IRV method of tabulating their choices is pretty questionable. There has to be a better way.
There are better ways, but they are generally more complicated and thus harder to explain to voters. IRV already gets a lot of flack for being too hard to understand, but it's the second-easiest system to explain after FPTP.
I think this should be implemented at the Party primary level. It's too hard to change the Constitution and all 50 states to some kind of run off system. But at the party level, I think it makes sense and would be much easier to roll out and implement.
The purpose of the change is to move away from a two-party system, where people are forced to vote red or blue, because green and grey are too small to have a real chance. Only making the change within parties wouldn't fix that at all.
I understand that, but from a pragmatic point of view, you have little chance of achieving that. As long as Congressional districts are winner-take-all. And Senate seats are winner-take-all (there's only one seat up for each state at a time, usually). Nothing will change at the Federal level. And as for the Presidency, each state gets to decide how they allocate their electoral votes. Game theory currently suggests it's in a state's best interests to allocate all of their votes to the winner of the popular vote in that state.
And in the meantime, you're going to hit roadblock upon roadblock in trying to get a major disruptive system installed at the Federal election level.
You would have a much easier time getting the system to change from within the Democrat and Republican parties for the primary system, since those rules are set by the parties themselves and not by the states.
So long as the parties are open and accessible to anyone, I don't see a big problem. The parties themselves are composed of many smaller factions (greens, libertarians, evangelicals, etc, etc).
There are already Congressional districts that use runoff elections (MS-special, for one that's being voted on today). States can choose the way they elect their representatives so long as it isn't contra the Constitution, which only requires direct election at the moment.
> You would have a much easier time getting the system to change from within the Democrat and Republican parties for the primary system, since those rules are set by the parties themselves and not by the states.
In any state with citizen initiatives, it involves less wrangling with esconced elites to change statewide rules than party internal rules. That's where the real big changes are likely to happen.
You don't have to change the Constitution to implement instant runoff (at least at the state and below level). Plenty of states use runoff elections already, and local governments are already using IRV.
While I think IRV is a better system than we have, I think it concentrates on the wrong problem. In practice it only really works on high profile elections where all of the candidates get plenty of public scrutiny. The lower houses like the Senate and House, especially at the state level, usually receive very little scrutiny compared to governors and presidents. But the real power lies more in these lower houses than with the heads of state. They really only have power because their political party already has a significant control in the lower houses. The lower houses usually have the power to override anything the head of state does, and the vast majority of the power the head of state has are just things the lower houses has designated to him/her, which they can just as easily take away.
So I think the real issue is to open up the lower chambers to third parties, and that's best done through the use of Proportional Representation. IRV will not work at that level, because very few races are scrutinized by the public very much. Proportional Representation also eliminates Gerrymandering (or significantly reduces it depending on how it's implemented).
Once you've got a significant number of third parties in the lower chambers, whoever is head of state really doesn't matter that much. Since they'll have to work with pretty much everyone in the lower chambers to get anything done.
Approval voting for executive positions, proportional representation for assemblies.
The upper house (eg Senate) in bicameral systems is a special case. I'm undecided if they should be abolished (7 states have more Senators than Representatives) or our state boundaries redrawn.
The Senate, and the changes to Senate balance caused by granting statehood, already was a contributing factor to one civil war. It also disenfranchises many Americans: DC has large populated areas that aren't part of any state, but Democrats insist on turning DC into a state in its own right (which would give them 2 extra Senators) while Republicans don't want the Democrats to have 2 extra senators, so the most they would agree to is retrocession to Maryland (similar to how Alexandria was retroceded to Virginia), which the Democrats won't accept because then they get 0 extra senators. (Most DC residents would prefer statehood, but then, most DC residents would also vote Democrat, and it's hard to tease out the conflicting motivations of "we want to have representation" and "we want our side to be more powerful in Congress".)
Puerto Rico also seems to want statehood, but there's no progress on getting it for them for various reasons I don't fully understand.
Redrawing state boundaries would either be politically impossible, or could only happen with a large enough national one-party supermajority that it would basically amount to gerrymandering anyway.
Why not approval voting? It's much easier to explain to people and is way more compatible with the current system. You simply approve of any number of candidates, and the candidate with the highest number of approvals wins. Maybe in practice people would only ever approve one candidate and the situation wouldn't change, but I think it's a least a good small step forward.
#1 - I switched from advocating IRV to approval voting once I better understood the election integrity aspects. Tabulating IRV is much harder, necessitating more software, which obfuscates the process.
#2 - Election administrators hate IRV. They'll oppose, sabotage it at every turn. We had a local trial run. As an observer, it'd be hard to argue the admins made an honest effort to make it succeed.
#3 - IRV is confusing to voters. Try explaining it to your parents.
Any changes to our voting systems should be incremental, deliberate. First with local races and then progressively up ballot. Give people time to adapt.
I'll suggest Australia as a counter-example to all three numbered points, albeit on a relatively small scale (population ~25m).
But I agree 100% with your final point: people need time to adapt, and be willing to do so. Changing the system to something people trust even less would be bad.
> Tabulating IRV is much harder, necessitating more software, which obfuscates the process.
This is a feature, not a bug. The count should be done by hand. The multiple rounds increase the probability that shennanigans will be detected, whether by sheer weight of watching eyes or by numerical irregularities between rounds.
> Election administrators hate IRV. They'll oppose, sabotage it at every turn.
I don't see how to predict which one they will like.
> IRV is confusing to voters. Try explaining it to your parents.
My parents are in their 70s and have voted this way since before I was born. My paternal grandparents were voting this way before my parents were born. My maternal grandparents voted this way when they moved to Australia.
Tens of millions of Australians vote this way and have done so for about a century.
It does not seem statistically likely that somehow Australia has had an uninterrupted run of particular genius at a multi-decadal/multi-million-person scale that nobody else in the world can replicate ever again.
How many races (contests) does a typical urban Australian election have? How many candidates for hotly contested races? How many ballots are procssed?
My jurisdiction often has 30+ contests. We had 22 belligerents for mayor 2 years ago. I think we had 12 for senator in our most recent primary. I dimly recall my county counts 450k ballots.
> How many races (contests) does a typical urban Australian election have?
Commonwealth, state and local elections are held separately.
How many candidates for hotly contested races?
For Senate elections it is not uncommon for there to be literally hundreds of names on the ballot, which is why "above the line voting" is an option.
How many ballots are procssed?
For a Commonwealth and State elections, millions, with an indicative count (ie, whether there a clear winner on first preferences) made on the same night by tallying across hundreds or thousands of polling stations.
More to the point: it scales. The indicative count is easily parallelised to any level, since summation is a commutative operation.
I think your implication is that Australia doesn't do it at a large enough scale and that therefore, our experiences don't count. That isn't true. Commonwealth elections are comparable in scale and complexity to the operations of those run by the largest states in the USA to handle federal elections.
Of course your experience counts. I'm just making sure we're comparing apples to apples. It appears my very shallow understanding of Australian election administration is WAY off. (I'll need to find some sample ballots.)
If you're handling large scale, complicated IRV elections with ease, than that should inform the efforts of our local proponents (eg fairvote.org).
No worries. You and I have already crossed swords a few times before. And my writing style has been described as "aggressive". I do appreciate you setting me straight on this. I hate being wrong.
Yeah, I agree. I do research on things very closely related to voting systems, and my first thought is approval voting. It's basically rating with two options, approve and disapprove. Has good properties and is relatively easy to understand and implement, especially in the age of social media.
People seem to get fixed on ranked choice voting versus plurality voting (the current system), and what we really need is a movement to change the voting method to something more accurate.
But it's puzzling to me why the discussion seems to highlight ranking rather than approval voting. The only argument I've really heard against it is that people would get confused and not realize they can vote for more than one person. But it seems to me in that case you've just re-implemented plurality voting, and people would figure it out well enough. Just put out a big public ad campaign. People would learn.
I think Approval voting would solve a lot of the first-past-the-post voting while also being much easier to explain to non-math nerd voters than IRV or Condorcet voting: just give every candidate a thumbs up or down. :)
Oakland switched their mayoral election to IRV (maybe ten years ago?) and it was so misunderstood that even one of the candidates recommended that his supporters vote for him in every IRV rank position! (Or perhaps he just didn't want to recommend his supporters even think about the other candidates names?)
Approval voting has the same problem as any other scheme which doesn't force an exhaustive enumeration of preferences: political parties can and will turn them into de facto first-past-the-post systems.
How?
"Just vote one". Or "just mark one". Or "100 points for us and 0 for anyone else". Etc etc. Enough voters will comply -- especially in voluntary systems which magnify the power of fringe voters -- that any advantages of non-ranking voting schemes will be neutralised.
Voting systems like approval voting, range voting and so on work if the options or candidates are prepared to behave uprightly. Outside of situations like a local club or professional organisation, they won't. They will meta-game. It happened in Queensland when they introduced "optional preferential" voting.
Because there is no consistent meaning to “approval”; this is generally a problem with limited-buckets systems in that they force artificially creating ties with no consistent meaning; if you have to have a single-winner system, it should be a ranked-ballot system without limited buckets that does something sensible with the rankings; IRV is about the worst, IMO, that meets that minimum standard.
The main point of IRV is to allow me to say "I'd actually prefer third party C, but if it comes down to main parties A or B (which it probably will), I guess I prefer B". Under approval, I have two options:
* Vote for B and C. My vote for C is guarenteed to fail because I'm also giving a vote to B, which was likely to be better than them in the first place. There's no way C can win if everyone who votes for them also votes for B.
* Vote for C alone. This increases C's chance of beating B, but it means that if C doesn't win I now had no say in whether A or B won.
Politics isn't black and white. There isn't a set of candidates I'm ok with vs a set I am not. My preferences are a ranked list, and any system that doesn't let me express those preferences is, in my opinion, worse than IRV, even if it does produce some theoretically sub-optimal results in some weird edge cases.
EDIT:
> Maybe in practice people would only ever approve one candidate and the situation wouldn't change, but I think it's a least a good small step forward.
I didn't see that bit before. Yes sorry, if we're talking about in comparison to FPTP, I'm all for approval voting.
I like Majority Judgment, which I essentially view as approval voting on steroids. It allows you to effectively rank candidates, for one thing.
Really, we just need something better than first past the post.
FPTP < IRV < Kemeny–Young < Majority Judgment
Really, if anyone here hasn't looked at cardinal voting systems before, I strongly recommend reading some of the academic work in this area. There are a lot of strong considerations in favor that surprised me.
STAR is on the ballot in Lane County, OR. We just might find out tonight or tomorrow that we'll see the first ever implementation of a more equal voting system.
Except for people who actually want a healthy society and see the bigger picture.
If my candidate wins with a plurality and is stymied and sabotaged by a divided population with many people feeling totally unrepresented, that does not create a world I prefer. I want a society where as many people as possible feel included and trust the system. I don't want a situation where losers feel disenfranchised and interested in undermining democracy.
If my views are actually democratically popular, I will be okay with a fairer voting system. If I like that the current system is unfairly weighted in my favor, then it's still very short-sighted for me to defend that system.
Incidentally, I don't advocate for IRV, it's a flawed system. Much better options are STAR, 3-2-1, or approval voting.
336 comments
[ 2.5 ms ] story [ 297 ms ] thread> This is sometimes called a Condorcet cycle or just cycle and can be thought of as Candidate Rock beating Candidate Scissors, Candidate Scissors beating Candidate Paper, and Candidate Paper beating Candidate Rock.
[1] https://en.wikipedia.org/wiki/Condorcet_method
In these parts (the UK) the same system is known as the Alternative Vote (AV) which is just as meaningless a name. But in defence of my fellow political scientists, there are a lot of different voting systems, many of which have alternatives, rankings, and runoffs, so unambiguous and useful names are hard to find.
“Ranked Choice Voting” is one of th names for the system, but since their are dozens (maybe hundreds) of voting systems which use identical ranked choice ballots with different resolution procedure, naming it “instant runoff” (since it uses ranked choice ballots to implement a variation of majority/runoff) is much more specific of name.
Can you explain why you think that is the case?
We really don't, historically (for the last century or so) it's been pretty cut and dry between either Labour or Conservatives with the occasional coalition (always involving one of those two parties).
https://en.wikipedia.org/wiki/Independence_of_clones_criteri...
It is very difficult for a 3rd voice to gain traction. A party which has attracted over 10% nationwide for the past 2 decades, Greens, has exactly 1 seat out of 150 under this system.
We've had this method of voting for a century but voters generally see it as a binary choice - Roughly 35-40% will vote for A and 35-40% will vote for B. Typically there will be up to 5-10 candidates on the ballot paper. So that 20-30% of the "neither of major 2" could be split 8 ways across a wide political spectrum. Inevitably, most outcomes fall back to the 2 majors.
Your statistic about the Greens, in particular, is true only for the House. The Greens hold 9 Senate seats out of 76, or approximately 11% of the Senate. Given that their national Senate vote was 8.7%, I think this is a pretty reasonable outcome.
Parliamentary system is merely the opposite of a presidential system and has nothing to do with voting methods.
[1] https://en.wikipedia.org/wiki/Parliamentary_system
IRV has the fundamental flaw in that it counts some voters' preferences and ignores others. It's an overly-complex and mediocre unequal system.
Check out https://www.equal.vote to understand what it means to have voting equality and how we can pursue systems that honor that.
Both are better than first past the post, but agreeing on an alternative is HARD.
The situation in Canada is instructive: the governing left-central Liberals promised to replace first past the post, but reneged on their promise when they realized that the committees studying the issue were going to come back with a Mixed Member Proportional recommendation which would hurt them and help the left-wing New Democratic Party, rather than the ranked ballot mechanism that would help them.
(That's speculation, of course. They gave some other BS reason for breaking the promise)
It doesn’t work for individual candidate positions like president, because you can’t have a non whole number percentage of a single candidate
Compare with how the system of the Netherlands works: a single country-wide electorate, and no threshold except for the number of votes needed for a seat. The outcome is numerous parties in parliament (i.e., far more democratic than the US) and coalition governments which must be formed or at least supported by parties that got a majority of the votes between them.
Especially with many parties, if parties A and B are reasonably similar, it's quite plausible to like party A's platform while preferring party B's presidential candidate. A separate presidential election accounts for that, allowing you to elect A for the house, and B for president.
Note that parties A and B are similar - I'd argue that they'd have to be, and that it's extraordinarily unlikely to get a president the ruling coalition is incapable of compromising with. This works out because overall, the vote for president should be ideologically similar to the votes for congress/parliament. (That working, of course, is contingent on some form of ranked voting for the president, rather than the FPTP system the US has now.)
All that said, I'm from Canada, and I don't really agree with the idea of a single person with large amounts of power, but that's my take if there must be one.
Specifically the president represents the executive branch, then you have the two houses, one that represents the people approximately proportionally, and one that gives 20x multipliers to the voting power of people from mostly empty states.
If not, do they compensate PEI for the senatorial clause?
I know I've "voted defensively" before, and I really don't like having to do it. I want to vote for those that I agree with, I don't want to vote against those who I disagree with.
It may seem like a small distinction, but if someone loves candidate A, but they are extremely unlikely to win, most will end up voting for B because they REALLY don't like C.
That leads to A never getting in office until they become more like B.
But with instant runoff, I'm free to know that I can vote for A first, and B second, and not worry that my support for A is implicit support for C.
It not a perfect system, but IMO it's a lot better than what we have now in the US.
Edit: another piece that I forgot to mention is that instant runoff can give more signals beyond the winners or losers. Currently you have no way outside of polling to understand what voters want, it leads to a vocal minority being able to control the narrative.
With instant runoff, the runners can easily see that even though candidate A didn't win, they did get a sizeable number of votes, and maybe next time B runs, they will incorporate some of A's ideas into their campaign.
EDIT: Oh, on one of the rounds the top candidate might become the bottom candidate, I see.
What happens if the candidate that gets removed is in the middle? I'm guessing nothing, since it "stops" at your first preference that's still in the race, ie your vote goes to the first person?
In practice here, even in such large contests the top 5 candidates normally end up with 90% of the vote and there's enough overlap between the parties (the further left parties transferring to centre left, the two centre right parties transferring to each other) that there's normally a clear leader by then, it's rare that e.g. a candidate with 48% gets eliminated in favour of a candidate with 49% in the last round.
One property of IRV that makes it so questionable is its non-monotonicity. This means that you can help a candidate by ranking them lower, and hurt a candidate by ranking them higher. It is extremely easy to visualize how ill-behaved IRV is in the charts on this website:
http://zesty.ca/voting/sim/
Meanwhile, ranked pairs (a "condorcet" method) has the property that, if candidate A is more popular than all other candidates pairwise, then candidate A will always win. This is _not_ true in IRV.
If you're interested in falling down the rabbit hole of electoral systems, I recommend looking at this chart that compares their properties:
https://en.wikipedia.org/wiki/Comparison_of_electoral_system...
Perhaps the most important important property of a voting system is that it is easy to grok, so that people trust it. In particular, aggregates at the local level should combine simply with other districts to tally overall winners.You can tally IRV incrementally, it's just that ranking permutations are the units of aggregation. That's non-trivial. For this reason, I prefer approval voting.
Approval voting is simple enough, but is vulnerable to bullet voting: approving an additional candidate can make your first choice lose. This makes it unsuitable if your goal is to make third-party voting more viable, which seems to be the premise of the article.
And, particularly, there is very notable cultural variation, which means (ab initio at least; over time, one expects behavior modification efforts would be deployed around this), support from different cultural groups in a diverse polity will have different electoral value even when turnout is the same.
> Approval voting is simple enough,
But also has the same problem: concretely what “approval” means is unclear and will be subject to voter-to-voter variation even when the actual preferences are identical.
http://scorevoting.net/Beaumont.html
Only if you assume that compromising is fundamentally good. Do you believe the propensity is to compromise to get the best of the both worlds or the worst of the both worlds? I would point out, for example, that during the bush (II) administration, the only truly bipartisan, compromise initiatives were war in Iraq, No Child Left Behind, and Corporate Bailouts.
But, isn't that the whole point? ... that a compromise, perhaps centrist, candidate can emerge victorious?
https://youtu.be/s7tWHJfhiyo
Just a friendly note: this visualization means nothing without context, and is far from "obvious". Unfortunately, I'd place this in the category of "things that are obvious to people who already believe it". Is there a better way to go about explaining your idea?
To paraphrase, this is a simulation which places candidates and voters on a two-parameter preference space. This is useful because two dimensions are easy to visualize, but the same intuitions apply to higher-dimensional spaces.
Each point is colored to answer "which candidate wins when the popular opinion is normally distributed about this point?". That means the fractured visualization shows IRV's non-monotonicity, and how even when popular opinion is close to a centrist candidate (here in red) it might be that the voting system results in the election of an extreme candidate (green).
With IRV, though, I could see a situation where they are much more likely to win simply because, nowadays, Republicans' and Democrats' general level of mutual disdain is such that I wouldn't be surprised if it ends up being common for people who identify with one major party to rank the candidate from the other major party below every other candidate more-or-less out of spite.
This is a symptom of our current system. Anybody who has any ambition in politics will modify some of their positions to fit one of the major two parties. If running for another party wasn't a death sentence, you would see more competent candidates in those parties.
Which is worth doing because, regardless of what the ideal situation is, if it's associated with a spectacularly bad outcome in one of the earlier elections that uses it, that's going to be a big impediment to wider adoption.
So the minor parties (and independents) are eliminated first, and their preferences are distributed. Which usually means than one of the major party candidates scrapes over 50%, reflecting the minor party voters' basic preference for one of the major parties over the other.
In this (common) situation, I think IRV actually works really well.
The major parties still end up in government, but they see the first-preference percentages for the minor parties, and understand what policy positions are important to the electorate.
Edit: As much as everyone's trashing IRV as "horrible" I don't see a simple, worrisome case where it violates monotonicity; those seem limited to really pathological electorates where nothing will make them happy anyway.
It's terrible because the loser-elimination step that makes non-monotonicity an issue also makes it less straightforward to tabulate, not because non-monotonicity is an issue outside of edge cases.
Difficulty in tabulation affects speed of, certainty of, and confidence in results, as well as cost of implementation.
> Maybe I react differently from the typical voting system enthusiast?
Nope, especially on neglecting pragmatics like tabulation; that's practically a hallmark of voting system enthusiasts.
Which doesn't rise to the level of "that is so bad it must be stopped".
>> Maybe I react differently from the typical voting system enthusiast?
>Nope,
Kind of strange, since the entire reason for my confusion and joining this thread is because of enthusiasts' doom-and-gloom about the quality of the results themselves, and how they worry about the edge cases and I don't.
Are you invigorated by the cost fears? Because they seem like a sliver compared to the cost of Gore vs Bush.
If I can hurt my preferred candidates chances of winning by ranking them higher on my ballot then the voting system is broken.
Non-monotonicity solely on a few pathological edge cases where voter preferences are a nigh-intransitive clusterf—- is not a problem.
The reason there are not more examples like this one is that most governments don't release enough information about ballots in IRV elections.
But ... is there maybe a truncated explanation of the general dynamic there?
The problem is that the B > C > A don't get their second choice considered. Why does this matter? Look at it when you put the votes into pairs.
C is actually the more popular candidate, but C LOSES using IRV!The theoretical candidate was everyone's second preference. That sounds like quite a good compromise.
You rank 1) Indian 2) Chinese and 3) Italian
Your partner ranks 1) Gastropub 2) Chinese and 3) Steakhouse
you're going to get Chinese food, not because it exists on a line between each of your first choices, but because it is acceptable to each of you, if not preferred. It is a compromise by definition.
Chinese food isn't a compromise, a blend between Italian, Indian, Gatropub and Steakhouse food.
In Chinese food there isn't a little bit from every other choices. A candidate being chosen on virtue of a compromise by the numbers doesn't mean s/he aggregates some ideas from other candidates.
Hence my remark about that candidate being on a different spectrum from the others.
I am not saying he shouldn't be chosen or that people wouldn't be happy about it, I am saying that his ideas aren't a compromise from the other candidates' ideas.
How many axes can you chart a poltician's positions on?
Authoritarian<>Libertarian
Traditional Values <> Social Relational Values
Survival Values<>Self Expression Values
Interventionalist <> isolationist
That's just a few from https://en.wikipedia.org/wiki/Political_spectrum
Say in the US 2016 election Sanders had run as a 3rd party. He was a liberal populist. He might well have been the second choice of both the Trump and Clinton voters, despite not being positioned between the two. He'd still have been the compromise candidate.
So let's talk 2016 presidential election
I'd imagine that a Trump Voter's ballot would look like the following if Sanders had run as an independent.
1) Trump 2) Sanders 3) Gary Johnson (Maybe with a flip on 2/3)
Clinton voters would be
1) Clinton 2) Sanders
Neither primary choice would be acceptable to the other. There would be some voters who preferred Sanders first.
Since then the second-tier candidates have done what you say about promoting each other: "vote for me and for X". If one loses, the other could win in the next round. IRV has a moderating influence on how candidates campaign. By making a bid for those later rounds viable, now everyone has an incentive to build strong coalitions before they even get into office.
Would another system work better? Maybe, but I still count it as an improvement over FPTP.
Now, arguably this is IRV functioning as designed, where the city's progressive wing could vote as a block without worrying about 2 candidates splitting the vote. It was very much non-intuitive, though, and a lot of London Breed supporters were quite pissed off. IRV allows you to "vote for a position" rather than "vote for a person", but many voters are still stuck in the "but....they got fewer votes!" mindset.
You meant after completing all the rounds of IRV, Mark Leno had more votes despite London Breed having more first place votes.
Which, yeah, is the whole point of IRV (and, with slight variations, most ranked ballots methods): a plurality winner that doesn't have broader support but ends up at the bottom of ballots not ranking them first shouldn't generally win over someone who has broader appeal just because the broad appeal splits first place support with another candidate.
> IRV allows you to "vote for a position" rather than "vote for a person", but many voters are still stuck in the "but....they got fewer votes!" mindset.
No, it lets you vote for a person, and also specify, “if not Alice, then my vote goes to Bob”. The winner in IRV is always the candidate with the most votes at the end of the process; reporting only first place votes as votes for the candidate is just wrong reporting of IRV counts.
Absolutely, it's more complex than a simple majority. But I think "bizarre" is unfair.
In practice, I think it helps to provide a diversity of representatives that more closely reflect the diversity of views in the electorate.
Because that kind of lesser-of-two-evils incrementalist attitude is inherent to the first-past-the-post system we have now.
Saying "we only ever eat meat, let's try eating an orange and see if we like it before we eat any of this lettuce" doesn't follow any actual logic. You might hate lettuce, you might hate oranges, but there's no specific factor that binds them together.
Also what sort of things would you factor into what you prefer? How important is each of them?
If my choice is between 1% of what I want, 2% of what I want or 99% of what I want, saying "why don't you just be happy with 2% since it's twice as good as the 1% you could have gotten" isn't the most logical point to make.
I care much more about getting a good voting system than I do improving our voting system, if that makes sense.
There's an impossibility theorem (Gibbard's) that states that for any non-dictatorial process for choosing among > 2 outcomes, there is no way to avoid tactical voting potentially swaying the results. The next-best alternative might be to make tactical voting so complex and error-prone that nobody bothers.
http://electology.github.io/vse-sim/VSE/ https://rangevoting.org/vsi.html
When it comes up in conversation (or I can work it into conversation), I push people to this - it has some great interactive visualizations
https://ncase.me/ballot/
(AKA Schwartz Sequential dropping (SSD), cloneproof Schwartz sequential dropping (CSSD), the beatpath method, beatpath winner, path voting, and path winner)
The part that is most appealing to me is how this method eliminates the least performing options first, then proceeds to better pairs.
The only modifications necessary to "fix" it are to add in a "None of the Above" / "No Confidence" slot AND to require fully filled out rank lists.
If NotA wins then the entire election is re-run with all prior candidates eliminated as possibilities.
I implemented the Schulze method in a hobby project once, and the math is a rather simple digraph algorithm, which also gave me a lot of confidence that it's a well thought out voting method.
Compare that to to some of the other systems which often have complex mathematical descriptions which are hard for people to wrap their heads around.
Another thing is how big a deal is 'non-monotonicity' is it actually going to be something that will affect elections all the time, or is it something that's more minor than say turnout which honestly has no solution and on one side you have USA where turnout is low so the game is all about making sure your base turns out, or on the other side you have Australia where voting is mandatory so making sure people who don't care, don't vote at random is a big part of the game.
STAR is just score with a single automatic runoff. It gets everything you are talking about in terms of intuition, explanation… everything you like about IRV is present in STAR and basically everything wrong with IRV is fixed by STAR.
What it's missing compared to IRV is that you can't go, it's just like what people already do (runoffs), just in a single round.
In STAR, 5 points is simply "most points" i.e. most support. It doesn't mean something semantically. You can give up to 5 points, like if there's a race, you can push each candidate forward as much as 5 steps. The only thing it means is that you pushed them forward more than another candidate.
That's not that hard to get intuitively.
IRV is not what people already do but in a single round, it's much more complex than that. And whoever's 1st choice in IRV loses in the final round, they (and that could be as much as 49% of the voters!) never get any of their other preferences counted at all, unlike either STAR or our current two-round system.
Yes, so you import the well-known cultural variation in numeric rating systems without concrete semantics for rating levels into your voting system. That's a great idea.
More data on ballots is good only as long as it has a consistent meaning and is used in a way that is consistent with that meaning.
STAR (and other score voting systems) gather noise and pretend it is signal.
STAR Voting allows voters to rate all the candidates similar to how we rate books on Amazon
Ah yes, Amazon, that site with famously reliable and unbiased user reviews.
This is a terrible idea. It’s a bad voting system made worse by appeal to similarity with popular apps and websites, which themselves use a really bad voting system.
STAR voting is NOT like rating books on Amazon, it's just superficially similar.
At Amazon, if you don't like any of a bunch of books, none of them get 5 stars. With STAR voting, you award points not to mean anything, but only to decide how much ahead in the race you want to push each candidate versus the others. Whoever you want to win most, you push them ahead with all 5 available points. That's it.
Your reaction and misunderstanding does further my annoyance at the people explaining STAR using the Amazon analogy, and I think that bad comparison is harming support for what is truly an excellent voting system.
My issue with all of these point- and range-based voting systems is that when you see the results, you might want to tweak your scores.
Let’s say I’m range voting out of 10 and I give four candidates 1, 3, 8 and 10 points respectively. But the candidate I rated 3 wins, with the candidate I rated 8 a close second. Damn it! If I’d known that was going to happen, I would have voted 1, 1, 10, 10.
So with extra information, range voting boils down to approval voting, and I think it all ultimately becomes ranked voting. It’s hard to imagine any scenario where I would want to change my rankings, even if I have information about the results. If I like both candidates A and B, but A slightly more, I’ll always put them at the top of my list and in that order.
Now, it’s true that in IRV there are scenarios where I can get a better result for myself by voting for candidate ranks in a different order; but those are rare (I think) and definitely unintuitive. In the example above, if B were a real long shot candidate, it might be worth ranking them above A in case B gets eliminated early. But it’s unlikely to make a big difference because, you know, B is a long shot. And that’s just IRV; other ranked voting systems are much more robust.
This is the EXACT reason STAR voting was invented. Unlike plain score, STAR has a runoff where it compares your preference of the top two and puts your full vote to your preference.
So, this is like an exact case study of why STAR is worthwhile. STAR proposes 0-5, but we can stick with 1-10 for this case. You vote 1, 3, 8, 10 and the top two candidates are the ones you voted 3 and 8. With STAR, it doesn't matter that your 3 candidate got a total score higher than your 8 candidate, your vote in the runoff goes to your 8 because that's rated higher. If more voters rated that candidate higher, they win even if their plain score total was lower.
Thus, you don't have to exaggerate all your scores. In fact, you have an incentive not to, because if you just give ties, you risk abstaining in the runoff instead of having a say there.
I mean literally you just spelled out the most common critique of plain score voting which is itself THE argument for why we should use STAR instead.
I’m reading https://www.equal.vote/starvoting and I still don’t entirely follow. It’s lacking some detail I’d like to understand. For example, it looks like it is not a Condorcet method. That seems bad to me -- if you’re not electing the Condorcet winner, you need a very good explanation why; and I don’t trust an appeal to numerical ratings that each voter will interpret and use in different ways.
Aha, yes, Wikipedia has a clearer explanation and confirms that it is not Condorcet: https://en.m.wikipedia.org/wiki/STAR_voting
I do agree that it looks probably better than IRV, but I’m not yet convinced that it’s a really good system and is worth pushing as a good approach for real elections.
One short point: the primary problem with not electing Condorcet is that coordinated strategic majority could, in hindsight and for future elections, force the Condorcet by refusing to support later preferences. In STAR, that strategy is risky and unpredictable. If the Condorcet is a strong case, it will win STAR anyway. If it's an edge case, nobody knows in advance if it will be the true majority or just a bit less, so strategically zeroing all by the favorite is so risky in that edge case, most people won't do it (it could backfire for them to refuse to support their 2nd favorites if their hail-mary hope that they just might have a Condorcet turns out to be wrong). No strategies with STAR are reliable. The best outcomes come from just being honest.
Note: STAR doesn't use rank as a tie-breaker only. It can flat-out overrule a score difference with the 2nd-highest scoring candidate winning. It's not just for ties.
> I do agree that it looks probably better than IRV, but I’m not yet convinced that it’s a really good system
Well, since IRV gets pushed for real elections and STAR is better on a large number of factors, and both are better than FPTP…
You otherwise want to promote 3-2-1 https://wiki.electorama.com/wiki/3-2-1_voting or approval or maybe some good PR approach like PLACE voting https://medium.com/@jameson.quinn/place-voting-the-elevator-...
If you want to discuss this stuff more with other folks, https://forum.electionscience.org is the place to go, btw
You’re combining ranked voting with range voting.
Ranked voting is objective and clear (the Condorcet criterion is very compelling) but unfortunately incomplete (Arrow’s theorem, which in this case boils down to unresolvable three-way cycles.)
Range voting is more tractable (counting is easy, no awkward cycles) but subjective (no-one knows what the scores actually mean) and prey to tactical voting.
STAR starts with range voting (so you lose Condorcet) and adds cut-down ranked voting as a final step (only after some rank information has been discarded; and ranking isn’t a sound way to resolve ties).
It would be better to use ranked voting first, then range voting to resolve tie-breaks. That would make better use of the unique advantages of each voting style.
I still don't buy your assertion that no one knows what scores mean. That's no stronger of an argument than the opposing critique of ranking: nobody knows in a rank ballot whether it's a slight preference between two supported candidates or a huge difference between a liked and a hated (but maybe some other candidate is even worse) situation.
In both ranking and scoring, the ballot does not give a transparent view of the voters' beliefs. They only mean what they mean on the ballot, that there are relative scores or a ranked order. We can't and shouldn't assert that we can know more deeply what's behind those marks.
Condorcet is not a bad principle per se, but it's not the absolutely most desirable. If it were the only factor, nobody would accept IRV. And Condorcet means that a candidate who is 2nd choice of 100% of the voters, maybe even a strong 2nd choice (rankings don't carry that info) will always lose even in a case where all the other candidates are grossly polarizing and will lead to civil war. There are at least cases where a consensus is superior to Condorcet.
> It would be better to use ranked voting first, then range voting to resolve tie-breaks.
How does this proposed idea work? What actual system are you suggesting that would do this? EDIT: I see this referenced in your other post.
That's not correct, if there are 4 or more candidates.
If there were exactly 3 candidates the split would have to be something like 51-49, and then the candidate with 100% second place would come second. But it's arguable whether that's a bad result, and it seems like a very unlikely scenario anyway. Even 10% of people voting that candidate in first place would be enough for them to win it.
What I meant to express was that the 100% 2nd choice candidate will lose to Condorcet winner if there's 51% support for the Condorcet winner. I.e. majority support. That's true also for non-Condorcet systems. My point was that Condorcet systems are always also strict majoritarian, and there are reasonable arguments that majoritarianism can be undesireable (such as cases of polarizing slim-majority winner.
If, in a two-candidate election, one candidate gets more votes, should that election be the winner, even if the losing candidate has more passionate supporters?
In democratic elections that respect one-person/one-vote, the answer to that question is simply yes, because that is the definition of one-person/one-vote democracy. Even if there are 100 voters, and it's 51-49, and the 51 voters are lukewarm while the 49 voters are passionate, the candidate with 51 votes wins. No matter what.
The reason that matters is because it is actually fairly slippery to measure passion. If one person submits a 10 rating, and another submits a 7 rating, what does that mean? It could mean that one person is passionate and the other person is lukewarm. But it could also mean that one person is emphatic while the other person is meek. Or that one person is authoritarian and the other person is abused. The entire reason we have one-person/one-vote is to make sure that voters are treated equally, even if society is otherwise telling certain people that they are "less than".
And for those sorts of elections, Condorcet voting is basically perfect. All the flaws people attribute to "Condorcet methods" are only attributable to when there are multi-member Smith Sets.
Now there are other kinds of elections; elections where we don't follow the one-person/one-vote principle. Smaller electorates, non-governmental, self-governing organizations. In those cases, "consensus" is more possible without risking systemic abuse. And in those cases, it might be entirely appropriate to award the winner to the 49% candidate with more passionate support. In which case you would perhaps choose something other than Condorcet.
You should come to terms with the fact that whatever you mean is completely unintuitive to many/most of us, and that almost everyone will just this like they use Amazon. Whether or not the page says "it's like Amazon" or "it's not like Amazon".
The ballots are proposed as labeled "no support" to "most support".
It isn't entirely unlike Amazon, it's like Amazon if your standard for what's a 5, like "the best book" is "the best of THESE candidates". All people need to do to get the system is care about their impact on the outcome of the election in order to realize that 5 doesn't mean "love" or some other semantic thing, it just means the top of the options.
And if people just use it like Amazon and some give no candidate a 5, the system doesn't fall apart, it works pretty well anyway.
http://star.vote is not a controlled scientific survey or anything, but people overall seem to have no trouble using the full range.
In fact, in my anecdotal experience, even over years in the past when I told people about plain score voting and tried to get them not to use the full range (because I misunderstood that issue myself), many people insisted that they would use the full range anyway. A lot of people find it intuitive to give their favorite all the points they can.
But the real answer is that, yes, we need more studies and trials of all these things.
Intuiting what "giving one star more or less to a candidate" means is really difficult to understand as a voter. "If more voters had given Kodos four stars instead of three, he could have won" is a headache inducing explanation.
Approval voting is the only alternative to IRV or plurality voting that seems to match the level of intuitiveness of both the voting process and the tabulation process.
If IRV complexity and explanations were not a problem, we wouldn't see today's reality where IRV proponents themselves constantly make incorrect claims about it.
Not always (and not in the linked article here), but often people get the intuition that their 2nd choice will get counted when their 1st choice is eliminated, and then try to get them to understand why their 2nd choice was not counted (because their 2nd choice had already been eliminated) while someone else's 2nd choice was counted…
What happened in Burlington before they repealled IRV shows not only people getting intuitions about IRV wrong but people often don't even get what happened in Burlington because it seems so counter-intuitive. https://www.equal.vote/Burlington
And yet, that scenario is common enough. Burlington is just the case where all the ballot stats were released so we can study it.
The problem, though, is one of correctness, not complexity -- I would maintain that almost anyone could look at that election summary and understand why the conclusion that was reached was reached. And if they had had a traditional non-instant runoff election, the result would likely have been the same. It's hard to even explain why the outcome was "wrong" because the tabulation of the results is so intuitive.
> I seriously doubt that anyone would get a headache over "if people had given that losing candidate higher scores, they would have won". I don't even see how you can honestly claim that to be anything but completely intuitive to everyone.
The question for the intuition is "how many stars to I give this candidate". What does it mean to give them one more star? One person's four-star vote carries exactly as much weight as four people's one-star vote, which seems incredibly odd. An Amazon product with a hundred one star reviews is intuitively worse than one with 10 five star reviews.
Certainly not! In Burlington, people know the Republican can't win. So, like everyone in the current system, people would (and have both before and since IRV) vote strategically. So, the less-hardcore Republicans would vote Democrat in order to stop the Progressive.
In fact, if IRV had not been repealed, a good portion of the Republicans would betray their favorite and vote Dem as 1st choice (dishonestly) in the future in order to stop the Progressive from ever winning again.
Furthermore, when vote splitting happens in a current system, everyone readily gets why it was "wrong" and they adapt their strategies accordingly.
> What does it mean to give them one more star?
It means they are one more point ahead of the others.
> One person's four-star vote carries exactly as much weight as four people's one-star vote, which seems incredibly odd
The weight is relative, that's all that matters. As long as people understand the system, they can realize that they are throwing away part of their vote if they only put a 1-star and nothing else. Four people who put 1-star for candidate A might have put 5 stars for candidate B. They are influencing the race just as much as anyone else.
Here's what it means to be equal: Any voter can score the opposite of another voter and they will cancel out. Our votes can have equal weight if that's possible. That's not always possible in our current system or with IRV.
> An Amazon product with a hundred one star reviews is intuitively worse than one with 10 five star reviews.
If there were 110 voters, then every item is relative to that vote. There's 100 1-star votes for candidate A and 10 5-star votes for candidate B? That means candidate B got zeros (blanks are no points) from 100 candidates. A candidate with 100 zero-star votes and only 10 5-star votes is a worse option.
With FPTP voting, C would have won of the three candidates, and A would have placed last. If you removed A from the race, B would have won - and did!
Alternate voting systems are supposed to produce different results from plurality voting, and that's often unsettling the first time it happens.
That page makes the claim that IRV produced a bad result and that's why it was replaced. But at least the candidate with the 2nd-most 1st place votes won. If instead the Condorcet winner, A, had won, do they really think it would have been better? A got the _least_ 1st place votes of the three, who's to say there wouldn't have been an even larger backlash?
FPTP is even worse than IRV, except that people know how it is bad so they adapt accordingly.
Of course, it's speculation about the specific backlash, but the core issue in IRV is this (and it happens in the Burlington case):
ALL the voters whose 1st choice loses in the final round never get any of their other preferences counted and they lose their 1st choice. They get NOTHING, no say, totally screwed. They could be as much as 49% of voters. Now, WITHIN IRV, they can later learn (as people have learned about strategy in FPTP) that if they betray their favorite and vote for their lesser-evil choice as 1st, then they WILL swing the election to their lesser-evil instead of the greater evil.
By ignoring the preferences of some voters and counting others, the weighting is unequal and people will feel disenfranchised.
Maybe this older more thorough discussion of the Burlington case will help you understand: https://www.rangevoting.org/Burlington.html
The core point is that voters in IRV can get a preferable outcome via favorite betrayal (a dishonest strategy). Either they do use that strategy and we're back to the lesser-evil problems we have now, or they don't use that strategy and we're back to vote-splitting like we have now, where a candidate choosing to run can both lose and cause a worse outcome for their supporters.
Another way of seeing this could be that such voters had their 1st choice considered and given value all the way until the last round, whereas many other voters had their lesser choices considered in their previous rounds. The 1rst-choicers-all-the-way-till-the-end could be considered more favored than the latter voters I mentioned, who lost their favorite candidates earlier on in the process.
I'm not saying that this is a better interpretation of the situation you posit. I'm just using it to illustrate that it's hard, maybe impossible, to get away from subjective criteria when we consider the virtues of various voting systems.
by the way, Fargo North Dakota just adopted approval voting for their City elections tonight.
Within a strict ranked-ballot approach, the overall best tabulation is probably Ranked Pairs.
STAR has some similarities with Bucklin but does not have its problems. STAR's advantages are several and aren't directly just a forgone conclusion merely by showing what's wrong with IRV. STAR is much simpler to tabulate than Ranked Pairs and doesn't need to be tabulated at the highest level, it can be counted at precincts like FPTP.
If there were an objective way to measure actual utilities, score-based methods would be great (and economics and policy assessment would be much simpler, and...)
In STAR voting, you have up to 5 points to give to each candidate. You effectively get to push each candidate ahead in the race relative to the others by as many as 5 steps forward. There is NOTHING subjective about the meaning of points. They are just points. More points wins. They have no cultural signficance, no opinion, nothing subjective.
Yes, there's a mapping to a (typically 0-5) numerical rating like those popular in online review systems, which is a motivation for the forced acronym, and exactly the similarity that is viewed as making it particularly intuitive and accessible.
Cultural differences in how such rating systems are used for similar preferences have been studied considerably.
It's true that you can also view STAR as a limited bucket (and so forced-tie for any but the smallest candidate pool) ranked-ballot method, in which you force voters to arbitrarily select which intercandidate ranking preferences to suppress, and then do a completely wacky (from a ranked preference perspective) method of tallying the ballots.
It is indeed a form of ranking that allows ties and has a limited resolution. But the limited resolution does help with simplicity.
From a UI perspective, presenting a ranked ballot that allows ties is unfortunately complicated. That's what STAR actually could/should be though.
And the tallying method is fine, not wacky. In a sense, it's like Borda Count to get to two finalists and then IRV for the finalists. That avoids all the worst aspects of IRV (which are all symptoms of the way its multi-round elimination ignores the preferences of many voters while counting the preferences of others).
I disagree. Generally, unforced rankings (allowing full expression of preferences with optional ties) are simpler than forced rankings (with no ties) which in turn are simpler than limited bucket arrangements that force ties; “Between X and Y, do you prefer X, Y, or neither?” is the simplest question about voting preferences, and unforced preferences can be answered by answering that question alone; forced rankings may, as the name suggests, force the invention of an artificial preference in some cases, but limited buckets methods force deciding which preferences are most important and which to suppress. And limited buckets methods which apply something more than ordinal meanings to buckets,as STAR does, are most complicated.
> From a UI perspective, presenting a ranked ballot that allows ties is unfortunately complicated
Tallying unforced preference ballots is complicated, “number your preferences starting with ‘1’ for most preferred, ’2’ for next most, and so on, with equal preferences getting the same number” with a ballot with appropriate entry blanks isn't hard UI either in paper or machines.
Even if you want paper optical scan ballots, the main problem is the size of the ballot (which is a logistical problem when you've got multiple simultaneous races) more than the UI.
> And the tallying method is fine, not wacky.
Viewed as a preference ballot, giving non-ordinal meaning to rankings is wacky, since it is inventing information. The wacky description was in that specific “viewed as” context.
> In a sense, it's like Borda Count
I don't disagree with this characterization, though I view it as more of a condemnation than a defense.
> That avoids all the worst aspects of IRV (which are all symptoms of the way its multi-round elimination ignores the preferences of many voters while counting the preferences of others).
If you want to eliminate all the worst aspects of IRV, which you view as all due to ignoring preferences via loser elimination, the simple solution is just to drop loser elimination (giving Bucklin or a form of Majority Judgement, depending on where you use forced preferences or allow ties; classic Majority Judgement even uses limited buckets, though you can use a full unforced preference ballot, so if you really think the preference compression is justified by UI benefits, you can use exactly the same ballot as STAR, with a much more straightforward resolution procedure.)
This is certainly far simpler than STAR, whether or not you use the same ballot.
You are right in noticing that it's not exclusively an unforced ranking (which I agree is vastly superior to forced rankings!). So, in STAR, there's a different between 3 candidates scored as 5,1,0 versus 5,4,0. They count identically in the runoff stage (which is effectively the rank part of the system). But they count different in the score.
I'm not convinced that difference is necessary, valuable, or intuitive to people, but I'm not convinced it isn't. I think these things need more real-world case studies. I definitely think the distinction isn't fatal or anything. I see how it could have a positive impact on outcomes.
I don't agree that multi-round eliminations are simpler than STAR per se.
I'm certainly inclined to see 3-2-1 voting as a competitor with STAR for best overall system.
https://www.youtube.com/watch?v=2q_eMUGCU5U&t=487
https://rangevoting.org/IrvExec.html
There are many examples of voters in real-world IRV elections _do_ rank their favorite candidates last, to help them win (the same problem plurality voting has, but worse, because there are multiple plurality runoff votes).
Mathematical models also agree – situations in which IRV produces bad results are extremely common. Something like 70% of possible situations where IRV and plurality voting produce different results, the IRV result is worse than the plurality result.
Approval voting and range voting are really the only voting systems I know of that are strictly better than plurality. They have their flaws, but per Arrow's Impossibility Theorem, there aren't any voting systems without flaws.
Huh? There isn't a clear-cut way of determining the best voting system, so how were they compare the two? That matters quite a lot.
There are situations where IRV get so complicated it is unpredictable, but living in Australia that does sometimes feel like a feature - if the polity expresses no strong consensus preference, random selection amongst candidates is not a bad idea.
> Approval voting and range voting are ... strictly better
Yeah, totally [0]. Approval voting looks like the one to use just in terms of doing what a voting system does.
Any system that lets neutrals and moderates in occasionally is better than plurality. And plurality is streaks ahead of having no choice at all.
[0] http://zesty.ca/voting/sim/
The key point which is laboured a couple of times is that some seat results are different than plurality outcomes. But that is a feature of IRV, not a bug.
Had the election been run under plurality rules voters would of voted differently and this fact seems to of escaped the author of this website.
Voters who preferenced greens first, recognising a minor parties zero chance of winning under plurality would of instead cast votes according to two party preferred lineup, who mostly would ideologically align with labour. But under IRV they can vote for anyone they feel more closely represents them but still have meaningful input into a 2PP outcome.
Show me a critique of an IRV outcome where 2PP outcome violates Monotonicity and you'll have a real critique of IRV.
However, in close three-cornered races, tactical voting does sometimes come into play. The recent Wentworth by-election was an example of this: Liberal voters could have assisted their party to win the election by switching their first preference to the Labor candidate, because if the Labor candidate had beaten the eventual winner (Phelps, Independent) into second place, then Phelps' preferences would have been distributed and almost certainly given the win to the Liberal. Of course, this is a risky strategy, because if too many Liberal voters did this then they would have instead engineered a Labor win. And in reality, despite the possibility of tactical voting, the Condorcet winner Phelps won anyway.
An example where the Condorcet winner may not have won is the seat of Rockhampton in the 2017 Queensland State election. The primary votes had Labor in first, independent Margaret Strelow in second, One Nation in third and the LNP in fourth. However preferences from the LNP candidate pushed One Nation into second, which gave Labor the win on Strelow's preferences. It seems likely that Strelow would have beaten any of the other three head-to-head, though. And as previously mentioned, this sort of situation is notable for its rarity.
That doesn't affect the tactical voting possibility there - the possibility arose because the Liberal candidate's chances were greatly affected by which of the other two major contenders made it into the final runoff.
Similar issue for end-to-end auditable homomorphically encrypted voting systems. Yes, anyone can check the math in theory ... but not the typical voter.
> Compare that to to some of the other systems which often have complex mathematical descriptions which are hard for people to wrap their heads around.
> Another thing is how big a deal is 'non-monotonicity' is it actually going to be something that will affect elections all the time, or is it something that's more minor than say turnout which honestly has no solution and on one side you have USA where turnout is low so the game is all about making sure your base turns out, or on the other side you have Australia where voting is mandatory so making sure people who don't care, don't vote at random is a big part of the game.
> Runoffs are something people understand, IRV is conceptually similar and makes intuitive sense to people, it's very easy to explain to people, 'if nobody gets a majority, eliminate last place and redistribute their votes, repeat'.
Compare that to to some of the other systems which often have complex mathematical descriptions which are hard for people to wrap their heads around.
It's fascinating how a similar constraint arises in industry software engineer; more than fascinating
The weirdness comes in the strange tiebreaker methods if there's a "loop", but when you have a lot of voters, loops are rare.
(Condorcet proponents would probably get more mileage if they had a better conceptual split between methods that identify the Smith Set, and methods that seek to award single winners of multi-member Smith Sets - those "tiebreakers" are where all the problems happen, and none of those flaws are relevant in elections where there is a Condorcet Winner.)
Could anyone take a stab at explaining how this could happen? I looked at the linked site but it's over my head. The visualizations aren't doing anything for me.
It is a consequence of the loser-elimination step. You can avoid it by eliminating loser elimination, which gives the (unlimited ranks variant of) Bucklin (with ties and truncated ballots allowed, essentially Majority Judgement) voting system, which has its own pros and cons. I kind of lean towards unlimited ranks, ties-allowed Bucklin—count first-place votes, if you have a candidate with a majority of ballots cast, they win, if not add all candidates second place votes and see if you have a majority (of ballots, not counted votes which will now exceed ballots), etc., with a tie-breaking procedure if you get two majorities in the same round (bigger majority, either in current or subsequent rounds, random selection of the tie persists when ballot preferences are exhausted)—is my preferred simple ranked-choice system. Condorcet methods arguably choose a better winner in the case where they disagree, but are less simple.
My updated intuition now says that IRV means that as long as your top candidate is in first place, your other preferences are disregarded by the system. "My #2 was eliminated too early -- if I had known that my #1 would be eliminated in the end I would have ranked #2 higher".
Basically, I suggest pushing approval voting over other methods for realistic cases.
For example, if 40% believe in A and 60% believe in B, and there one candidate, X, standing on platform A, and two candidates, Y and Z, standing on minor variants of platform B, the votes split 30% for Y, 30% for Z, and X gets elected with 40%.
It is easy to see how the 2016 presidential election would have worked out if this assumption was realistic. Hillary 52%, Donald(Republican) 24%, Never-Trump(Republican) 24%. Hillary wins by a landslide, despite playing a sub-optimal electoral college game.
The assumption is very far from realistic. I next turn to whether the lack of realism matters.
Imagine a three way race, with two old established parties, A and B, and a new third party, C. Under first past the post voters who prefer C > B > A, vote B, to avoid letting A in. Meanwhile voters who prefer C > A > B vote A to avoid letting B in. Under first past the post (FPTP) it is very hard to start a new party.
We might frame the issue this way: who decides the fate of the third party candidate? Under FPTP the third party candidate loses and this decision was taken long ago, by men now dead, when they opted for FPTP. Under IRV the fate of the third party depends on today's voters. If they are voting C > B > A and C > A > B then the third party candidate stages an upset victory. If they are voting B > A > C and A > B > C then the third party candidate is humiliated and doesn't bother running again.
The argument for IRV over FPTP is that we prefer that the fate of the third party candidate be in the hands of the voters, rather than seeing the third party candidate lose automatically, by the game theory of FPTP.
The zesty link simplifies away the problem that IRV is intended to solve. But it doesn't argue that the problem is small; it cannot do that because every-one can see that the spoiler effect under FPTP is a big issue. Instead it refuses to mention the problem, simplifies it away, and hopes that nobody notices. That is derailing, not contributing.
However, a system similar enough to your example has been used for years by the Quakers, where each voter gives a score of +1, 0, or -1 to each option/candidate. The difference here is that you can give +1 to multiple candidates, and -1 to multiple candidates. Giving your opinion on one candidate in no way limits your ability to give opinion on other candidates.
You'll never be able to sell it to the general public, of course, but if you could do it for just one election I think you'd have everyone on board just from the results of it.
This is a much more interesting question to me, first because the candidates' positions are more mutable than the voters' opinions and second because as someone whose political opinions are often non-aligned with major parties, I really wish that there was an incentive for viable 3rd-party candidates to give a voice to my views. The ncase.me simulator linked elsewhere on this thread seems more useful:
https://ncase.me/ballot/
Range/Score (and STAR, which is a variation) use numeric score rankings on a preset range, and some methods use a limited set of ranks rather than (allowing voters to use up to) as many as candidates (Approval can be viewed as two-rank system.)
I often wonder if we should just use Condorcet and if there isn’t a clear winner, declare the result a tie. (That would likely be vulnerable to a kind of DoS attack, though.)
The Smith set is as easy to explain as the Condorcet winner, and random selection is trivial, so random selection from the Smith set seems to be a simple Condorcet tiebreaker.
Hmm, there might be something in that...
The Smith set tie would occur more often, and in complex scenarios where it really isn’t clear that it actually is a dead heat.
But of course, random ballot has just one pesky undesirable property: the fact that it’s random.
So one option is to hold a runoff between the candidates in the Smith Set.
Another is crazier, but I haven't found a flaw yet. Basically introduce range/score voting on the ballot, but only use the scores if there is a multi-member Smith Set. That way you guarantee Condorcet compliance. And since the voters don't know if the ranges/scores will be used (and they probably won't), they don't have incentive to strategically score a less-desired candidate ahead of a more-desired candidate.
I just thought of a way to possibly make it more intuitive. Use ranked voting, with the addition of a “nobody below this line” item that you can add to your ranking anywhere you like. Then at the runoff stage, you use approval voting, with “nobody below here” determining which candidates you approve.
That gives you a nice easy-to-understand runoff, and is maybe even more understandable than normal ranked voting because you get to explicitly separate out the candidates you really can’t stomach. I suspect some people don’t like ranked voting because you’re expected to stack-rank lots of candidates you dislike.
But we're talking a lot of complexity to have both ranks and scores as separate marks. Super long ballot, will turn away most voters and seem really hard to process.
It sounds like you might appreciate 3-2-1 voting. https://wiki.electorama.com/wiki/3-2-1_voting
(The point of the above was to respect the Condorcet Criterion, and 3-2-1 voting does not meet that criterion.)
IRV has its flaws but they are minor in comparison to the voting system we have.
It's entirely possible (even within the confines of the US Constitution!) to virtually never do this. Elect the House of Representatives by statewide party list. Senate is hard because each state has two senators elected non-simultaneously; otherwise statewide STV would work. Since the President is already elected by the Electoral Congress, skip the popular Presidential election and select the Electoral Congress by sortition among all eligible registered voters in a state (similar to jury selection except without being able to arbitrarily send people home) and have them personally interview the candidates before voting their personal consciences.
I was unfamiliar with the ranked pairs system, and will study it more to understand it better. It looks promising as far as technical excellence, though I'm afraid it'd still be too hard to understand for the general populace in most (or all) countries.
One thing I suspect about range voting is that it seems to require strategic voting. A voter would still need to give a very favorable score to a majority party that's kind of aligned with their views, as opposed to giving them a lower score than a minority/third party he or she would prefer. One of the strongest appeals to me from IRV is the opportunity it provides for a minority party to unseat majority parties, which is appealing given the political stagnation experienced by many countries. I had also thought that it allowed voters to avoid strategic voting and express their true, individual voting preferences.
After reading IRV's description in http://zesty.ca/voting/sim/ , I realized that I had misunderstood one important detail. I had thought that the least popular candidates were eliminated first until only the top two candidates remained, and the one with most votes among them remained. I dislike that the "real" IRV (not the one that was in my head) first checks for a 50%-plus-one majority before eliminating the bottom candidates. That feels too much like the first-past-the-post system for my liking. I wonder if having successive rounds where the least-favorite first place candidate is eliminated would make most people's favorite candidates "bubble up" to the top, producing a final match between two highly-favored candidates. I wonder if such a system is closer to a Condorcet method than the real IRV...
Suppose that a candidate has a majority at the beginning. Then this continues to be the case as other candidates are eliminated. Thus, this majority candidate will be elected at the end. So this extra "if-clause" only speeds up computation, it does not change the voting rule.
Could you elaborate on why you think plurality is better than IRV?
It doesn’t matter. It ain’t gonna happen. Not because IRV or condorcet is or isn’t better, but simply because of the installed base. Getting literally everyone to a) understand the subtle differences and then b) agree to switch and c) agree to switch to the same system is impossible because literally every step of the process is itself impossible.
This is not me being pessimistic, this is me being realistic.
Humorously, a top down dictatorial approach would be the only way to “improve” (by some estimations) the current democratic implementation. You’d have to not care about democracy to upgrade democracy.
/s
Proof by contradiction, mate. It's not impossible, just hard. There's plenty of things that once we're said to be impossible, and the data agreed... until they happened. Everything had to have had a first time.
http://jsfiddle.net/p4L6do1m/1/
What's needed isn't just IRV but STV - think IRV but districts grouped together and multiple winners. That way, you can actually get results that are proportional to the votes cast because the difference between a narrow win and a landslide changes who gets elected down the ticket.
Some systems do this randomly (randomly choose ballots and take the next preference), and some do this fractionally (e.g. transferred votes are worth whatever fraction wasn't needed to elect their previous choices). Overall, though, the point is that if your middle choice is eliminated, then either your vote wasn't enough, or a previous choice needed your support already.
That said, most systems 'skip' over previously elected candidates as well. Strictly speaking, doing that can mean ballots with a losing first candidate are more powerful than ballots where the first candidate is elected by a narrow margin[1]. That said, as per the wiki link, there's a solution (recalculate all instead of skipping previously elected) and tactical voting around that issue isn't seen in practice anyway.
[1] https://en.wikipedia.org/wiki/Issues_affecting_the_single_tr...
STV = Single Transferable Vote
Which is, to be sure, not very well, but there's nothing special about the House that makes it work there better.
Approval voting: you can vote for any of the candidates you like, not just one.
Way simpler than IRV.
Score voting: give any or all candidates scores.
Easy explanation.
STAR: score the candidates, there's an auto-runoff where the two highest-score candidates get checked to see which is preferred by more voters.
At least as easy to explain as IRV, and much easier to actually understand overall, particularly in making sense of the results and following what happened.
There -- simple!
Way too complicated for 50% + 1 of the population
IRV (and any preferential voting system) is quite close to the voting system for any of the weekly elimination talent shows. You rank people in order, the last one gets knocked out, you try again - with the difference that this all happens at once.
Are we really saying that people are too dim to rank their choices in order of preference, or understand the voting system behind some of the most popular TV formats worldwide for the past decade?
https://www.maine.gov/sos/cec/elec/upcoming/MarkedBallotExam...
If you have a country with 60% A, 30% B and 10% C, then you might like to a parliament roughly representative of that. (A, B, and C could be anything - attitudes to same sex marriage, daylight saving - there are always many ABC sets.) If parliament consists people elected in geographic regions, and A, B and C are uniformly distributed over the regions than all voting systems here will give you a parliament consisting only of A's.
People here seem to enjoy pointing out some perverse outcomes for IRV, but the Arrow's impossibility theorem proves (and ncase.me beautifully demonstrates) all voting systems have perverse outcomes. Yet none are as bad as electing all A's to represent a diverse country and every one of them will do that if used the wrong way, which also happens to be the most common way they are used.
Back to the topic - IRV does have its downsides, but it's not perverse outcomes. Australia tries to use IRV everywhere. The Australia Federal Senate has a voting system that does fix the dictatorship of the majority and naturally they try to use IRV there too. If you ever have the unfortunate experience of voting in an Australian Senate election you'll see why IRV isn't so good in that situation. In fact it's so bad not even the Australian's can use it in its pure form. If you are wondering why, it leads asking the public to number 100 odd boxes in strictly consecutive order, on a 1m wide ballot paper that can't lie flat in the ballot box with writing so small they had to issue magnifying glasses so you could read the candidate names. https://www.abc.net.au/news/6870332
So I say "can you list your choices in order of preference?" They almost always say that's acceptable. You can run IRV from that, but you can also run most any Condorcet variant. The only remaining part is to show the results in a fashion people understand, which is non-trivial but is also not what the "anything else is too hard" group was complaining about.
So don't say "ranked pair" or "STV"...say "order of preference", and you'll find people more willing to consider ideas. It's not EASY, but it's definitely EASIER.
We use FPTP in the UK because it directly correlates an MP with a constituency. Even if the candidate you didn't prefer wins, they are still your MP and have a duty to represent you.
We had a referendum on PR a few years ago (which was rejected) and its proponents kept banging on about it, then went very very quiet when UKIP got 4 million votes (more than the SNP and the Greens combined).
You had a vote on IRV, not for proportional representation. The Australian lower house is based on the same Westminster system but we use IRV.
A relatively unknown voting system is Direct Party and Representative Voting:
http://www.dprvoting.org/
which basically involves giving voters two ballot papers at each election: one a simple "pick one" for the local candidate (counted like a normal FPTP election), and the other a "pick one" for the party that the voter prefers at the national level (if any). The proportions for these second ballots are used to adjust the "voting power" of the MPs or Representatives that were elected for each constituency / district by the first ballots. For example, each politician from an under-represented party could be given effectively two votes in the parliament / congress.
An even simpler approach (which is a variant of Mixed-member Proportional Representation) would be to use the same ballot papers (and constituencies) as now, and produce the same set of winners as a FPTP counting process produces, but add an extra 20% of representatives to parliament / congress to readjust the composition of that house to reflect the national vote totals. In particular, these extra representatives should come not from a party list (since those give parties the ability to give seats to unpopular candidates) but from the "best near-winners", that is, the candidates who did best in their elections without winning the plurality vote. Such a system is already implemented in Baden-Württemberg.
Both of these systems have a clear benefit over Ranked, Score, and even Approval voting, which is that the vote counting process involves putting each piece of paper into one of a few piles, once. There are no extra rounds, and the only information that needs to be communicated for collating is the number of votes for each candidate (which correlates well with the size of the respective piles).
Basically if you have e.g. 100 seats in your house, you choose 100 ballots at random and whomever is voted for on those ballots are the representatives. It averages out to proportional representation, while still allowing geographically distributed candidates.
It also satisfies all of Arrow's criteria[1], which no deterministic system can.
1: https://en.wikipedia.org/wiki/Arrow%27s_impossibility_theore...
Approval voting is deterministic and satisfies all of Arrow's criteria. Arrow's theorem doesn't apply to voting systems, such as approval voting, that are not ranked choice systems.
The core problem with most voting systems could be solved with more representatives and just straight proportional representation. In the range of 1 to 10k–100k. Then you do a parliamentary system with one primary chamber that does most of the law making and one secondary chamber that does some checks but is ultimately toothless. A supreme court that can make law, but tends to refer legislation back to the parliament. A clause to allow the primary legislator to override all but the most basic freedoms required for democracy to work like voting and a free press.
Essentially Canada, but with proportional representation and maybe double the number of MPs.
The problem with random ballots is that unserious people will will or vote for themselves and they're harder to do securely with straight paper ballots that is understandable to laypersons.
If you're going to randomly sample ballots just randomly sample voters and skip the line.
> Some publications even blame third party voters for Hillary Clinton's loss in 2016.
I don't think there is any plausible reason to believe Clinton would have won in 2016 if not for third-party candidates running. The math just doesn't add up.
In Michigan, Trump won by around 11k votes, and there were around 223k third party votes (172k Johnson, 51k Stein).
In Wisconsin, Trump won by 22k votes. Johnson got 107k there and Stein got 31k.
In Pennsylvania, Trump won by 46k. 147k went to Johnson and 50k to Stein.
I'd guess that almost all Stein voters would pick Clinton over Trump, so take Stein out of the race and in all three of those states Clinton wins that statewide vote unless the Stein voters decide to stay home. I'd guess that those who would vote for Stein would be sufficiently "anyone but Trump" to not stay home.
Those states are all winner-takes-all for the Electoral College, so that would do it. (Maine and Nebraska are the only states that split their electoral votes).
What would have happened without Johnson is a harder call. Trump and Clinton are both terrible from a Libertarian point of view. I'd expect Libertarian voters to be more likely to go Republican than Democrat if there is no Libertarian on the ballot most years, but Trump is so different from what had been mainstream Republicans before that I have no idea who the majority of Libertarians would have picked.
Also, I suspect that many of the Johnson voters were "never Trump" Republicans rather than Libertarians, who would agree with conservative writer and research fellow at the Cato Institute P.J. O'Rourke who said that Clinton lies about everything and is wrong on every issue, but he was endorsing her over Trump because she was wrong "within normal parameters".
I see absolutely no reason to believe that.
https://www.isidewith.com/candidate-guide/jill-stein-vs-hill...
https://www.isidewith.com/candidate-guide/jill-stein-vs-dona...
Also check them all out at ontheissues.org.
Stein was the Green Party candidate, and Trump was pretty much the anti-Green candidate. Clinton ranges from agreeing with the Green party to indifferent on most of their major issues.
Even where Trump and Stein agree, such as on getting rid of Obamacare, they are massively incompatible (Stein wants to get rid of it for a single-payer system, Trump for just letting the market work it out).
However, it is a two-round election. A run-off is always necessary. But it can be made one-round via what I call rank-and-star: voters rank candidates by preference, and put a star next to the top N that they like. The two candidates with the most stars are then compared on rankings. It's pretty easy to prove this is mathematically equivalent to the two-round system, but it's not a ranked-choice method because the ballots (1-2-3) and (1-2-3) are different.
Interestingly, it might still be possible to satisfy both participation and monotonicity if you expand the run-off to three candidates and use minimax Condorcet for the final round. For three candidates, minimax = Schulze = ranked-pairs and they satisfy participation; for four or more candidates, no Condorcet method satisfies participation. However, doing this costs you the attractive simplicity of the method (but minimax Condorcet is simple to understand, relative to IRV or ranked-pairs).
Anyway, this puts me in the odd position of advocating a voting system nobody has ever heard of =p
It passed, but -- perhaps not surprisingly -- many current, established politicians weren't happy. Maine's governor called it "the most horrific thing in the world" and the state legislature refused to fund implementation and voted to delay it until 2022.
So this year the the citizens of Maine voted again to confirm instant runoff voting, overruling the state legislature and vetoing their delay. It passed by an even bigger margin this time.
https://www.theatlantic.com/politics/archive/2018/06/maine-l...
https://en.wikipedia.org/wiki/Paul_LePage#Drug-dealing_comme...
In any case, he's leaving office and leaving Maine, so who cares what he thinks?
Simply running IRV is easy enough. Understanding it and its ramifications is complex enough that even the majority of IRV advocates don't understand it.
- people need to believe it is fair.
- it should avoid lop sided results (one party always wins regardless of vote) yet should avoid excessive diffusion (i.e. you don't want to have only fragile coalitions of many small parties).
- others, all with the aim of clarity.
As there is no magic bullet people use combinations of different systems:
- Australian ballot (secret ballot) which most systems use these days; hard to enforce in USA because of first amendment.
- quasi-independent apportionment processes/committees (as used in CA, AZ et al)
- party lists (not used in USA) or party-plus-candidate (e.g. Germany)
- non-party systems (e.g. CA)
- multi-candidate districts (not used in the US)
- proportional representation (most democracies) or first-past-the-post (most dependents of Westminster). I'm a fan of the latter BTW.
- IRV
Mix, match, iterate until an overwhelming majority of people are satisfied.
But really, the question isn't whether pathologies are possible ever, the question is how likely they are. What systems have the best balance to be the most robust while still being practical? STAR Voting and 3-2-1 voting. Definitely not IRV. IRV's pathologies have high enough probability that they do and will continue to show up significantly.
Good starting reference from one of the leading voting theorists:
https://www.quora.com/What-is-the-best-single-winner-voting-...
I'm open to the idea that actually using median is superior. Could you expound on that?
If we could only get those pesky humans out of the loop...but we don't.
But I don't know what you mean about "picking one element" or understanding or believing. Both STAR and 3-2-1 are quite understandable, believable, and don't require weighing full sets of pairings. That weighing you describe sounds like how Ranked Pairs version of ranked-ballots is done.
But I chose that link in this case, despite its problems, because it more strongly emphasizes the complexities.
FairVote has engaged in really intellectually dishonest arguing, but the level of over-the-top reaction that the rangevoting site does in response comes across poorly and is regretful.
And in the meantime, you're going to hit roadblock upon roadblock in trying to get a major disruptive system installed at the Federal election level.
You would have a much easier time getting the system to change from within the Democrat and Republican parties for the primary system, since those rules are set by the parties themselves and not by the states.
So long as the parties are open and accessible to anyone, I don't see a big problem. The parties themselves are composed of many smaller factions (greens, libertarians, evangelicals, etc, etc).
In any state with citizen initiatives, it involves less wrangling with esconced elites to change statewide rules than party internal rules. That's where the real big changes are likely to happen.
So I think the real issue is to open up the lower chambers to third parties, and that's best done through the use of Proportional Representation. IRV will not work at that level, because very few races are scrutinized by the public very much. Proportional Representation also eliminates Gerrymandering (or significantly reduces it depending on how it's implemented).
Once you've got a significant number of third parties in the lower chambers, whoever is head of state really doesn't matter that much. Since they'll have to work with pretty much everyone in the lower chambers to get anything done.
The upper house (eg Senate) in bicameral systems is a special case. I'm undecided if they should be abolished (7 states have more Senators than Representatives) or our state boundaries redrawn.
Puerto Rico also seems to want statehood, but there's no progress on getting it for them for various reasons I don't fully understand.
Redrawing state boundaries would either be politically impossible, or could only happen with a large enough national one-party supermajority that it would basically amount to gerrymandering anyway.
#2 - Election administrators hate IRV. They'll oppose, sabotage it at every turn. We had a local trial run. As an observer, it'd be hard to argue the admins made an honest effort to make it succeed.
#3 - IRV is confusing to voters. Try explaining it to your parents.
Any changes to our voting systems should be incremental, deliberate. First with local races and then progressively up ballot. Give people time to adapt.
But I agree 100% with your final point: people need time to adapt, and be willing to do so. Changing the system to something people trust even less would be bad.
This is a feature, not a bug. The count should be done by hand. The multiple rounds increase the probability that shennanigans will be detected, whether by sheer weight of watching eyes or by numerical irregularities between rounds.
> Election administrators hate IRV. They'll oppose, sabotage it at every turn.
I don't see how to predict which one they will like.
> IRV is confusing to voters. Try explaining it to your parents.
My parents are in their 70s and have voted this way since before I was born. My paternal grandparents were voting this way before my parents were born. My maternal grandparents voted this way when they moved to Australia.
Tens of millions of Australians vote this way and have done so for about a century.
It does not seem statistically likely that somehow Australia has had an uninterrupted run of particular genius at a multi-decadal/multi-million-person scale that nobody else in the world can replicate ever again.
My jurisdiction often has 30+ contests. We had 22 belligerents for mayor 2 years ago. I think we had 12 for senator in our most recent primary. I dimly recall my county counts 450k ballots.
Commonwealth, state and local elections are held separately.
How many candidates for hotly contested races?
For Senate elections it is not uncommon for there to be literally hundreds of names on the ballot, which is why "above the line voting" is an option.
How many ballots are procssed?
For a Commonwealth and State elections, millions, with an indicative count (ie, whether there a clear winner on first preferences) made on the same night by tallying across hundreds or thousands of polling stations.
More to the point: it scales. The indicative count is easily parallelised to any level, since summation is a commutative operation.
I think your implication is that Australia doesn't do it at a large enough scale and that therefore, our experiences don't count. That isn't true. Commonwealth elections are comparable in scale and complexity to the operations of those run by the largest states in the USA to handle federal elections.
If you're handling large scale, complicated IRV elections with ease, than that should inform the efforts of our local proponents (eg fairvote.org).
People seem to get fixed on ranked choice voting versus plurality voting (the current system), and what we really need is a movement to change the voting method to something more accurate.
But it's puzzling to me why the discussion seems to highlight ranking rather than approval voting. The only argument I've really heard against it is that people would get confused and not realize they can vote for more than one person. But it seems to me in that case you've just re-implemented plurality voting, and people would figure it out well enough. Just put out a big public ad campaign. People would learn.
Oakland switched their mayoral election to IRV (maybe ten years ago?) and it was so misunderstood that even one of the candidates recommended that his supporters vote for him in every IRV rank position! (Or perhaps he just didn't want to recommend his supporters even think about the other candidates names?)
How?
"Just vote one". Or "just mark one". Or "100 points for us and 0 for anyone else". Etc etc. Enough voters will comply -- especially in voluntary systems which magnify the power of fringe voters -- that any advantages of non-ranking voting schemes will be neutralised.
Voting systems like approval voting, range voting and so on work if the options or candidates are prepared to behave uprightly. Outside of situations like a local club or professional organisation, they won't. They will meta-game. It happened in Queensland when they introduced "optional preferential" voting.
Because there is no consistent meaning to “approval”; this is generally a problem with limited-buckets systems in that they force artificially creating ties with no consistent meaning; if you have to have a single-winner system, it should be a ranked-ballot system without limited buckets that does something sensible with the rankings; IRV is about the worst, IMO, that meets that minimum standard.
* Vote for B and C. My vote for C is guarenteed to fail because I'm also giving a vote to B, which was likely to be better than them in the first place. There's no way C can win if everyone who votes for them also votes for B.
* Vote for C alone. This increases C's chance of beating B, but it means that if C doesn't win I now had no say in whether A or B won.
Politics isn't black and white. There isn't a set of candidates I'm ok with vs a set I am not. My preferences are a ranked list, and any system that doesn't let me express those preferences is, in my opinion, worse than IRV, even if it does produce some theoretically sub-optimal results in some weird edge cases.
EDIT:
> Maybe in practice people would only ever approve one candidate and the situation wouldn't change, but I think it's a least a good small step forward.
I didn't see that bit before. Yes sorry, if we're talking about in comparison to FPTP, I'm all for approval voting.
Really, we just need something better than first past the post.
FPTP < IRV < Kemeny–Young < Majority Judgment
Really, if anyone here hasn't looked at cardinal voting systems before, I strongly recommend reading some of the academic work in this area. There are a lot of strong considerations in favor that surprised me.
https://www.quora.com/What-is-the-best-single-winner-voting-... lays it out clearly.
STAR is on the ballot in Lane County, OR. We just might find out tonight or tomorrow that we'll see the first ever implementation of a more equal voting system.
IRV is not an equal system, it counts some people's preferences and discards others. https://www.equal.vote/star-vs-irv
If my candidate wins with a plurality and is stymied and sabotaged by a divided population with many people feeling totally unrepresented, that does not create a world I prefer. I want a society where as many people as possible feel included and trust the system. I don't want a situation where losers feel disenfranchised and interested in undermining democracy.
If my views are actually democratically popular, I will be okay with a fairer voting system. If I like that the current system is unfairly weighted in my favor, then it's still very short-sighted for me to defend that system.
Incidentally, I don't advocate for IRV, it's a flawed system. Much better options are STAR, 3-2-1, or approval voting.