It only takes one sentence to describe: "Vote for everyone you approve of."
It also gracefully handles incomplete information—if you don't know about someone, just don't vote for them.
But I can already tell you by the complexity of what that article is trying to show that most people won't understand it and won't care.
When I vote, it's in part to declare what I want and what I believe in. I generally don't like Democrats and I don't like Republicans. If I want to vote for the Green Party candidate, then I want to vote for the Green Party candidate and make it known that that's who I want to win. I don't want to be stating that I equally approve of both the Green Party candidate and the Democratic candidate since that doesn't accurately represent my opinion.
Given that the Green Party candidate is not going to win, I generally vastly prefer the Democratic candidate to the Republican candidate, even if I don't like the Democratic candidate one bit.
I don't want to vote in a way in which my statement is that I equally prefer the Democratic candidate to Green Party candidate. I want to vote in a way that states, "I want the Green Party candidate to win. And if I can't have that, I'll reluctantly settle for the Democratic candidate over the Republican candidate."
Voting in a way in which one's desires are clearly indicated is what most people would want, I imagine, even if it has some counterintuitive results at times.
Also, at least on a cursory reading, it seems that the author of the article has a bias towards "moderate" candidates winning. Who says that I want a moderate candidate to win? I want some serious progressive change!
The posit is that the spoiler effect is bad: Fielding an additional extreme candidate on one side should not be helpful to the other side.
So with Instant Runoff elections, the majority gets their wish. The Green Candidate wins.
With Approval Voting, 60% of the people vote the Green Party Candidate, 40% of the people vote for the Republican Party Candidate, and 100% of the people vote for the Democratic Candidate. The Democratic candidate wins.
I don't see how this is a win for Approval Voting. The candidate that wins is a candidate that no one really wanted.
Well, clearly my example is rather contrived, but I suspect (perhaps out of ignorance) that any situation in which Instant Runoff would cause counter-intuitive results is also rather unlikely.
I'll definitely have to look at your reference, though. I do find such issues very interesting
https://www.youtube.com/watch?v=-4FXLQoLDBA
The video was created by the Equal Vote Coalition. They also have some good information on their site.
Preferential voting. If your sincere choice (ie Greens) didn't win, your vote goes toward the second choice until a candidate gets clear majority.
Proportional voting, people can vote for their sincere choice and there's a pool of 10 representatives. That means each candidate needs at least 10% of vote. If over 10% people in the state for Greens but rest vote for Democrats/Republicans. Thats 1 Greens and 9 Democrats/Republicans.
In USA, it is not possible to have a sincere choice. Also known as the illusion of choice.
Elected officials are initially largely irrelevant, changes to the voting system will start with state-level voter initiatives.
Pretty much everything worth doing to voting systems aside from (1) replacing or reforming the electoral college, and (2) moving away from mandatory single-member districts in the House of Representatives, can be done entirely there in many states without elected officials being involved except in their ministerial role in certifying the results and implementing the dictated policies.
Keep in mind that our system has changed a great deal already (e.g. women and non-whites being allowed to vote), the politicians at the time benefitted greatly from how things were set up before, but those changes were still accomplished. We shouldn't give up on the idea of reform just because politicians would resist it.
I'd probably check off both, personally, because I don't care. See, the system works!
This is a pervasive logical fallacy, we call the "expressiveness fallacy". https://sites.google.com/a/electology.org/www/expressiveness
In short, you could allow partial approvals (1/4, 1/3, etc.), or mathematically equivalent ratings on a 0-N scale. But Approval Voting behaves almost as well in aggregate. http://scorevoting.net/BayRegsFig.html
And IRV/RCV will maintain two-party duopoly. https://asitoughttobe.com/2010/07/18/score-voting/
Voting theory is highly counterintuitive.
It is not exactly counterintuitive that aggregated orderings have problems - the problems relate to the fact that ordering cannot be averaged and ties are not allowed. The latter problem also happens in the oldest score system called Bucklin voting.
Approval voting is essentially 0-1 score voting system with ties.
I don't particularly like to argue over semantics, but this is not a "logical fallacy". It may be some sort of fallacy, but it's not a logical one.
Approval voting would encourage people to NOT vote for people they would accept, for fear that it would hurt their favorite.
You wind up with an arguably more benign form of strategic voting: preference deals.
The point here is that approval voting, range voting etc work better than IRV under a number of ideal conditions.
But political parties have no incentive to cooperate with the ideal conditions. They want to beat the other party.
Requiring exhaustive (or even second-level) ranking has been ruled unconstitutional (at the state level) in a couple of US states when forms of Bucklin voting were used; which makes any method that relies on compulsory multiple ranking (rather than accepting truncated rankings) more difficult to implement.
Which isn't a fundamental problem, but it's something to be tactically aware of.
(I have thought this before about other things I have come across. Universal healthcare, and taxes to pay for it, come to mind)
It doesn't matter. Voters will do what's in THEIR best interest, regardless of what their favorite party wants. E.g. Ralph Nader wanted his supporters to vote for him, but 90% told pollsters they voted for someone else, mostly Gore.
I repeat: voters do what's in their own best interest.
With Approval Voting, that often means voting for your favorite PLUS one or more other candidates who are more electable.
In a large 2014 exit poll including Approval Voting, the result of the Maine gubernatorial race was completely REVERSED. http://scorevoting.net/Maine2014Exit
Sure, the Democrat didn't want his supporters also voting for the independent. But they DID IT ANYWAY, because they didn't want the Republican.
There is a reason that experts in game theory generally favor Approval Voting and/or Score Voting, such as outlined in the 2008 book "Gaming the Vote" by William Poundstone.
I disagree with this somewhat, and the reason is because some ballots have an option to vote for one party across the board, and people do that. Parties choose their candidates, therefore parties are telling people, who have the option of voting on their favorite color, who to vote for.
If you align yourself with one of the two big parties in the US, you're almost always going to vote for whatever they tell you to (of course there are exceptions, but I'm speaking generally). Generally people believe the party they like is acting in their best interest (because why else would they like it), and now we've come full circle. Ralph was not a member of one of the two major political parties.
That said, I agree with the rest of your comment.
That naturally leads to the same ability to discern support for compromise candidates as ranked choice. Each individual voter can't express that but the population as a whole will.
Because IRV (“ranked choice voting”) may be the worst non-malicious use known of ranked choice ballots, but approval allows even less information about preferences to be expressed and does so in a way where the honest marking of ballots isn't even clear.
Plus, once you get people using ranked ballots, debate over better counting procedure is possible without disrupting how people cast ballots. Approval is a worse method from which any improvement is more disruptive.
1. You need to support transitional back-ends for certain users (e.g. instant runoffs)
2. You strongly want to avoid a "new system" that cannot be replaced and has no room for expansion.
___
Any voting system able to handle ranked-preference voting can trivially be "dumbed down" for a certain set of customers who want approval-voting -- it's almost as simple as mapping integers to booleans.
In contrast, if everything is built assuming inputs are boolean values for each item in a set, your system is crippled. You either have to hope that everybody loves approval-voting, or admit defeat and switch it back to the old (i.e. current) stupid system of "pick one".
You can convert ordinal ranks to scores directly meaning the set exactly same as number of candidates. By that, each median score represents median ordering in a ranked ballot. You may lose certain properties that are relevant in case of direct ties but nobody really cares about that in big constituencies.
Heck, you can also keep counts of each score in case the dominant is needed for a tie breaker.
Approval voting is a score system with a binary score.
Rules are simple: in case of rank scores, candidate with lowest median rank wins, if there is a tie candidate with higher count of low scores wins. This converts IRV or other ranked system to Bucklin vote.
In case of approval scores, replace lowest with highest.
Under ranking voting systems (such as AV) is is obvous how to vote. You do it in this in this order: Candidate-A, Candidate-C, Candidate-B.
Under scoring systems, where you give each candidate a score (say 10 for the best to 0 for the worst), it is also reasonably simple for to vote: Candidate-A 10, Candidate-B 0, Candidate-C somewhere in between. But here again there are problems? If I give C a high score, he is more likely to beat B (which I want) but also more likely to beat A (which I don't want). So that's an incentive for dishonest voting.
Either way it doesn't incentivize dishonesty, and it doesn't cause weird squeeze-outs of the middle candidate the way IRV does.
I can see how that's less clear cut to a voter, though.
This would allow voters more freedom of expression than the current system, while being simpler than ranked voting. I understand that it could conceivably (and probably frequently would) result in the winner being the person with the smaller negative number of votes. But at least then they couldn't claim a "mandate for change" based on a broad support of the populace.
Obviously I vote for the last one and don’t vote for Satan, but do I vote for the middle one?
* Get a 5 sided die.
* Rank each candidate from 0-5.
* For each candidate roll the die and vote for them if your roll is less than or equal to your rank.
I very much dislike approval voting because the way to make people the happiest is to turn it into stochastic STAR.
http://scorevoting.net/RVstrat6.html
The evidence overwhelmingly shows that Approval Voting works well, arriving at outcomes that represent the will of the people.
People tend to approve of candidates they know of. Likewise they tend to not approve candidates they do not know of or they dislike for any number of reasons.
So these marks are not drawn from random distribution at all.
- Ballot expressiveness
- Strategic voting (which distorts sincere information)
- Tabulation efficiency (e.g. IRV has more data on the ballot, but destroys a lot of it due to inefficient tabulation)
You are making a common fallacy outlined here. https://sites.google.com/a/electology.org/www/expressiveness
When you look at total combined information throughput, Approval Voting does exceptionally well. http://scorevoting.net/BayRegsFig.html
Thus using the standard strategy I mentioned is the most accurate way to convey your preferences. If you want them to appear as honest as possible, you just treat all candidates as equally likely to win, such that the formula simplifies down to your approval threshold being "average utility of all candidates".
There's literally no difference. It's up to you to make your choice.
It's much simpler in a plurality system. First, I decide if I want to make a statement or a difference. If I want to make a statement, I vote for the candidate I like best and that's it. If I want to make a difference, I choose the candidate I prefer from the two major parties. In the unlikely event that I like them both equally, I can either keep looking until I find a reason to prefer one, or I can abstain knowing that either one is equally good (or bad). Note, I'm not saying this is good, just that it's simple.
The problem with instant runoff voting is that you can actually be harming your preferred candidate by ranking them ahead of the middle candidate.
In my example, the decision of whether to “approve” of the horrible candidate has enormous stakes.
The whole point of alternative vote systems is to be able to both designate your preferred candidate while also being able to specify your tertiary preferences so you can both vote for who you want to win while also not sabotaging your ability to vote against those you don't want to win.
Approval only lets you avoid the spoiler effect.
You can do it, but what does it get you?
Instant runoff is not monotonic: there are times when the best strategy is to vote your favored candidate second and your second choice first so that the second choice can push out your least favored candidate.
Edit: I wasn't asking you to explain it to me; I was asking how you break it down for the average person so that they don't need to keep depending on the election officials to disambiguate every case.
You recon “approve” is exclusive? Write down the one you would have voted for in FPTP.
Had trouble deciding which of three to vote for? List all three.
Someone’s borderline? Go with your gut. Or your head, whichever.
Do you want to increase some candidates' total tallies? Then "approve" of them on your ballot.
(Seriously, if we get anything at all other than first-past-the-post in my lifetime, I'll be happy.)
VERY Roughly: Eliminate the weakest candidates until there's a winner.
Only this method is a formal mathematical twist on that to make it computationally easier for computers to process while yielding the same result.
In previous posts I've wanted two additions to this method, but I'm modifying the second after thinking about it.
* None of the Above / No Confidence (same, this wins == new vote all old candidates banned)
* Allow for a single preference list but don't require all candidates ranked.
I can’t see this majority criterion as an absolute good.
This is an unavoidable vulnerability in any scheme that doesn't involve an exhaustive and exlusive ranking.
I do like approval voting as well, for the same mentioned reasons, but I worry that after a few elections people would start voting strategically, and then it could take another few decades to change the voting system again.
The more important part isn't ranked choice or approval voting anyway - both are single-winner voting systems which are far from ideal for elections involving Congress/Parliaments and local legislature. It's much preferable to have a proportional representation voting system for those.
This isn't a hypothetical. Queensland introduced "optional" preferential voting and one of the majors promptly campaigned on "Just Vote 1" in order to wedge the opposing coalition.
What Queensland uses, (which they call) Preferential voting, is IRV (instant-runoff voting). That's not approval voting.
It might initially seem like IRV's ranking (A>B>C>D>E) would be a better voting system than Approval Voting (where voters simply say A-yes, B-yes, others-no). Relative rankings seem at first glance to express more information. But it turns out that the way you have to count relative rankings either leads to pathological consequences for 3rd parties (see IRV), just like winner-takes-all voting, or the counting method becomes complex and byzantine in corner cases (see Condorcet). It's all explained on rangevoting.org
Most real and proposed balloting methods (other than range/score voting) are preferential, they differ in the restrictions on he expressed preferences; IRV (and lots of other systems), as usually defined, relies on a fully-ranked forced-preference (no ties) ballot. Approval has an implicitly fully-ranked two-preference-rank ballot with unlimited ties, FPTP (whether plurality or majority/runoff) has an implicitly fully-ranked two-preference-rank ballot with ties allowed only in the second rank.
> Relative rankings seem to express more information
They do, there is no “seem". They also do so more consistently. Forced preferences can have the reverse problem of forced ties, though (which is abstractly a point in favor of unforced ranking ballots, but there's less work done on resolution procedures for them.)
> but it turns out that the way you have to count relative rankings either leads to pathological consequences for 3rd parties (see IRV), just like winner-takes-all voting, or the counting method becomes complex and byzantine in corner cases (see Condorcet).
Bucklin is simpler to count than IRV and doesn't have it's pathological effects; the main concrete criticism I've seen is that if truncation is allowed rather than forced-preference, and if the no-majority fallback is highest-vote-total-on-last-round, it may encourage truncation and devolve to plurality. (But a similar charge of incentive to bullet-vote is made against approval.)
What does “approve of multiple candidates” even mean?
Preferring one candidate over another has a clear meaning. Approval does not; as well as meaning honest, uninfluenced ballots are going to vary from voter to voter even with the same actual assessment of candidates, that makes it a lot easier for voters to be manipulated by influencers as to the right way to map preferences to ballot markings.
Approval is good for open-ballot voting on, e.g, where to go to lunch (where “approve” means “commit to attend if the option is chosen”.) It's a poor choice for public, secret-ballot elections.
If it doesn't, then your ballots have no consistent meaning and any process for aggregating them produces garbage output in terms of aggregating social preference.
Consistency of meaning of ballot markings and the resolution process being sensible in terms of that meaning are the most critical minimal standards for a voting system. Otherwise, voting is a farce and distraction, not a serious way of guaranteeing that government is accountable.
That isn't a reason to choose a system that maximizes the degree to which that problem is preserved.
It's still consistent in that in any case, the candidates you're approving of are all preferable (for you) to all of the candidates you're not.
There's no ambiguity.
I'll be an anecdote here, from 2016. I voted for Sanders in the Democratic primary. I voted for Johnson in the general election, but had I been in a battleground state where Clinton had a chance I'd have voted for her.
Did how she was doing in the pre-election polling and exit polls determine whether or not she was my favorite candidate? Of course not. Would my single vote for her to try to avoid electing Trump have looked any different from a devout partisan Democrat's vote for the same office if I had voted for her? No, not at all. What FPTP almost convinced me to do is to VOTE FOR SOMEONE WHO WAS NOT MY FIRST CHOICE.
That's the problem, you see. If I'm tempted to vote for someone I really didn't care for rather than who I prefer because of the mechanics of the counting, that dilutes my voting power. I wanted to vote for Johnson. If I could have ranked the general, it would have been Johnson, Clinton, Stein, Trump. If I could have voted Approve/Disapprove, it would have been Johnson and Clinton approve and the others disapprove. If it was Approve, Neutral, Disapprove for each I'd have either voted J+C, S, T or J, C+S, T.
The last thing I want is to have to choose between my candidate of preference and some tolerable major party candidate based upon who else is running.
Perhaps, but the question was why people seeking to replace FPTP advocate IRV instead of Approval, so pointing out FPTP has the same problem being pointed to in Approval when compared to IRV is missing the point.
The last few ranks in ranked choice voting go unused or are randomly filled, except that making an explicit ranking of a least favorite candidate requires ranking them all. The first few ranks also may well be random, as the first two or three candidates may all be equally acceptable to a voter. On a ballot with city council for a ward, council members at large, mayor, county board, county chair/county judge, school board, sheriff, constable, state House, state Senate, governor, lieutenant governor, comptroller, secretary of state, land office commissioner, agriculture commissioner, several judges, school board members, US House, US Senate, President, and more that's a lot of ranking and a lot of time to do so. It may no better support the voter's will than Approval after all the effort by the voter and effort to update the counting process. However, some people may well have a strong opinion on where each candidate in a seven-way race belongs in a list.
Approval/Disapproval or Net Approval voting, especially with a neutral option, solves that problem of having to support candidates you don't care about higher or lower relative to other candidates that neither RCV / IRV nor FPTP solves. In FPTP, you can choose just one and either Approval or RCV solves that. In Approval you can denote clearly that you approve of any number of candidates, without your first four choices being arbitrarily ranked. You can disapprove of any number as well, without your last choice of disapproval getting ranked right behind your choice for winning the office.
I genuinely can’t comprehend the problem you’re raising. Even if it can’t solve Arrows Impossibly Theorem, and even though I’m leaning towards technocracy rather than democracy anyway, it doesn’t seem worse than the current systems.
http://scorevoting.net/BayRegsFig.html
And this is true regardless of whether there is any objective "meaning" to the votes.
In mathematical terms, approving the candidate means you prefer that candidate to the expected utility of the winner.
In approval voting, the strategy to maximize the influence of your vote is to vote for half of the candidates, or at least half of the candidates who are likely to win. (Well, except that arguably in nearly all election schemes, the instrumentally rational strategy is to stay home and do something productive, like baking a pie, having sex, or writing some software; this is why voting is obligatory in nearly all functioning democracies.)