I'm co-leading this initiative. It's democracy, not politics: a simple, non-partisan, candidate-neutral change that leads to far more representative[1] winners.
If anyone would like to invest time or money to make this happen, please contact us or donate here: https://SeattleApproves.org/
The first link there does actually compare against ranked choice. It is just called IRV. There's a larger comparison that includes more types of ranked choice (also called "ordinal") here[0]. Though I particularly like this animation[1]. I'll also add that there is a lot of nuance to voting and VSE isn't everything. There's a lot of factors to consider including: how do strategies work? Is it scalable? Can it easily be counted and/or verified statistically? How easy is it for voters? How effective are (strong and weak) spoiler candidates? And many more questions. For the most part cardinal methods (like Approval) fair better on all these accounts. There is a slight tradeoff to VSE. Note STAR in [0] has a slightly lower VSE than RP/Schulz (ranked), but a tighter grouping (better vs strategies). But also STAR is more complex for voters and counting than Approval is (I'd still argue STAR is easier for voters and counters than ordinal systems). So these are the types of tradeoffs you're making
i'd actually love to see a voting system developed that's closer to a conjoint, which is a statistical method designed to reveal preferences, rather than relying on stated preferences (which are notoriously unreliable). it's more complicated, but we have plenty of technology nowadays to make it practical.
one method of implementing this (from thinking about it for all of 2 minutes), is developing a list of the top 10-20 issues that voters care about in a given election and providing tradeoff choices for each, then matching them to the candidate closest to their revealed preference. of course, the biggest issue with method is that it moves the problem up one level, from voting to choice design.
I honestly think it is very difficult to determine what voters' actual preferences are. I mean think of your standard political compass tests. Or even Meyer Briggs. They can change day to day or mood to mood and lots of questions tend to be leading. While I love the idea, I think logistically it would be a nightmare and greatly reduce transparency. It would also be far more vulnerable to so called experts manipulating the system if they aren't acting in good faith.
it's exceedingly difficult to fairly aggregate voter preferences, but all current voting systems seem to reduce all that complexity down to a singular choice in the overwhelming majority of cases. that's a shitload of information discarded, to put it crassly.
the neat thing about conjoint is that it multiplexes the choice-making (when designed well), so that the choice burden scales reasonably while also better representing collective choice. nothing is going to be perfect, but conjoint could potentially be better than all current systems, probably by an order of magnitude or more.
it's relatively complicated, so yes, 'experts' (i disdain this term, frankly) could in theory manipulate the system, but transparency of voting data (just like now) and survey design (the more likely bias vector) would make that difficult. transparency would allow competing independent journalists, academics/researchers, political analysts, etc. to weigh in and balance out biases (not perfectly of course), as well as verify results.
Referendums exist to get a state-wide policy change. In the last election, historically conservative states voted for liberal policies like legalized marijuana, $15 min wage, etc.
For politicians, for me, it's more about (perceived) character vs. their stated policies.
you could certainly include character traits among choice decisions, but that's more fraught by subjectivity. that said, that's the magic of conjoint done well: turning subjective hidden preferences into (more) objective choice revelation.
currently, there is no option to choose a candidate supporting marijuana legalization, $15 minimum wage, pro-life, and pro-gun policies (as an example), since we practically only have two relatively fixed packages of pre-ordained, all-or-nothing policy positions to choose from. revealed preferences would pressure parties into fielding candidates that better match the non-partisan preferences of constituents. or better yet, it would lead to the waning of the two-party system altogether, toward a more direct representative democracy.
That would work if all candidates are interchangeable if they assert the same plank. In practice, however, candidates X and Y may both have identical planks but wildly different levels of efficacy. People aren't machines.
Sorry, but I think that's a terrible idea. Unless we're talking about ballot initiatives, I would hate for us to focus more on policy-based voting. What we should be doing as citizens is voting for someone based on our belief that they will represent our interests in the long-term. If you vote for someone solely because they agree with you on a couple of issues, you could be voting in someone truly reprehensible. I would prefer to vote in someone with integrity that I disagree with than someone untrustworthy that says the things I want to hear.
Additionally, all of this ignores the possibility that a representative could change their stance on an issue after receiving more information on an issue. Presidents and Supreme Court justices have changes their stances once their terms have begun or events have taken place during their time in office. I would hate a system that relies on tightly binding the representatives to narrow definitions of hot-button issue stances.
That's all academic, though. A great real example is the 2021 Seattle City Attorney primary, which had only 3 candidates. The votes were split: 36%, 32%, 30%. These 2 candidates emerged from the primary: https://www.nytimes.com/2021/10/30/us/seattle-city-attorney-.... The candidate that most accurately represented the opinion of the electorate, Holmes, was eliminated. That would have still happened in a ranked ballot. The voters who voted for Thomas-Kennedy or Davison in our current system would have still put Thomas-Kennedy or Davison #1 in a ranked ballot.
I tried to distill this in https://twitter.com/troyd/status/1461389190206156802. And a 3-way race like that is actually the simplest possible case. Effects like that are much more likely with more candidates and they happen everywhere in the elimination cycle, they’re just less obvious to casual observers. That's what you see in https://www.youtube.com/watch?v=7btAd1HYvjU&t=1329 (note that IRV == RCV in that video).
(Obligatory disclaimer: I don't care who wins. I do care how accurately the winners reflect the inputs, namely, voter opinion.)
Separately, Washington law requires cities hold a primary that produces 2 candidates for the general election, so cities can't legally eliminate primaries. That said, whether primaries are a net positive or negative is a really complicated discussion. Beyond election mechanics, the disadvantage of primaries is that only a subset of voters in the general will turn out in the primary. The advantage is that voters don't spend much time on ballots and it focuses that time. Seattle had 15 candidates for Mayor (many races are 10+). If someone spends 10 minutes on their ballot and the ballot has 10 races, that's 1 minute per race - 6 seconds per candidate. No easy answer here. However, for practical purposes we don't need to answer that hard question; this one change delivers an incredibly representative election.
I've seen approval voting up close, approval and top 3 make it to council seats.
One issue seems to be domination by majority?
For example, let's say 55% of electorate is hard left, 45% moderate to conservative.
Under approval those 55% can vote for all 3 lets say hard left candidates, no moderate's make it in?
RCV, once someone has gotten their candidate in, their vote is exhausted. Seems to come out much better. Still might end up with 2 hard left, one moderate, but that seems fairer to me.
Multi-winner elections are different from single-winner elections. Approval voting doesn't work there, because as you note, people with fewer acceptable preferences get fewer votes.
Voters voting strategically are known to adjust their thresholds in approval voting to have the largest effect they can on the output, assuming they have some idea of the support base of candidates. That said, Range Voting is sort of a solution to this, and it's expected there that candidates will scale their top and bottom choices to the top and bottom scores respectively.
if X is a fixed number, then it's not approval voting. In approval voting, there is no limit, either minimum or maximum, on the number of candidates you can vote for. If X is limited, then you get something a bit more like plurality voting, since the spoiler effect is in play.
As others have said, if they're limiting the number of approvals you can make on your ballot, then that's not how approval voting is supposed to be administered.
Did you have three votes, or more than three? If you had the same number of votes as there were positions to be filled, that's a plurality block vote. Winning slate takes all.
> RCV, once someone has gotten their candidate in, their vote is exhausted.
Hi, Australian here. What happens with multi-winner single transferable vote is that there's a _quota_ and then ballots for candidates who have reached quota get _reweighted_ (not quite the same thing).
This quota-and-reweighting system can (and must!) be applied to multi-winner Approval too - jpfed linked the wiki article.
What blocks us from attempting a similar approach state-wide? Seattle's top-two approach has already made a big difference in making the general election relevant and competitive.
Where I see trouble these days is with Federal elections. Being able to vote for anyone but X or Z would be a wonderful way to send a signal as a voter. I suspect that both the east and west sides of the state would feel similarly (East: "Anyone but X!", West: "Anyone but Y!". Result: M gets elected, who is thankfully neither X nor Y.).
> What blocks us from attempting a similar approach state-wide?
Same thing that makes it difficult by attempting to do it country wide. That it is larger and more difficult. It would be nice if we could just wave a hand and change everything at once, but that also wouldn't be very democratic. What is also nice is that we can experiment, refine, and test at smaller levels before we advance to larger elections.
Sorry for the late response. I wasn't on HN yesterday.
Someone else answered this well, but: money. A Seattle initiative is roughly 10%-20% as expensive as a statewide one, and it's still an incredible challenge to raise enough to do this. Ask me where I spend my time.
Qualifying for the ballot in Seattle requires about 26,200 valid signatures (10% of votes cast in the last Mayoral election). They must be collected during a 180-day period that, due to deadlines to make the November ballot, must is basically January-June. Since January-March is wet and cold, people aren't outside as much and many initiatives just do April-June.
Collecting 26,000 signatures in 90-180 days is all but impossible with volunteers, which means the core challenge for an initiative is fundraising to pay signature-gathering costs. You can divide 26,000 by any number of valid signatures collected per hour - 3, 6, 12, whatever - and it's still a lot of time and thus really expensive.
So, that's for the Seattle ballot. Getting on the WA state ballot requires 324,516 (!). Yes, 12x more. And that cost scales roughly linearly; the few efficiencies of scale are offset by new inefficiencies like travel. There's some efficiencies with overhead, but that's a small % of the cost.
So, that's why. It takes 5-10x more money.
Of course, voters would need to support it statewide too. In Seattle, this will pass; our job is to get it on the ballot. A statewide campaign may also need to do at least an education/awareness campaign (roughly the last paragraph of your comment) and potentially even try to change public opinion. These can cost even more than collecting signatures.
(If you're reading this and thinking "Wow, Troy knows this stuff - and it's interesting and there's a plan to do this!" - it is and there is! I'd love your help as a volunteer and/or your donations.)
Despite the claim that this isn't political, it's very difficult not to see this in light of the fact that Seattle's city council has the only elected Marxist in the U.S.
Kshama Sawant fought for the $15 minimum wage, the Amazon tax, numerous renters' rights protections, a ban on police use of chemical weapons against protestors, and is now continuing the fight for rent control. She's currently facing a right-wing recall campaign because big business was unable to defeat her in the 2019 election, despite Amazon dumping $1.5 million into various races for citywide office.
A few here have remarked that approval voting tends to select more "moderate" candidates over "extremes." This isn't a neutral preference, it's obviously ideological, and targeted at one specific legislator in Seattle. Voters should reject this measure as a transparent attack on the most effective and sincere fighter for working people in the country.
Full disclosure, Kshama and I are both members of Socialist Alternative. I may be limited in my ability to respond today, as I'm rushing right after work to a volunteer shift to make sure she stays in office. I'll try my best to check in occasionally!
Sawant and her supporters have had a pretty loose relationship with the truth in general, so I wouldn't put too much stock in that. It seems far more likely that if it's specifically in response to recent events that it's actually the fact that "conservative" (by Seattle standards, where we just elected the first republican in 30 years to a city office) candidates did quite well in the Nov elections where they were running against far left candidates and had squeezed out any more moderate voices in the primaries.
That's a reasonable point, Ann Davison's victory could definitely be a factor as well. And I don't know much about the background of the petitioners here; I think it's likely they're sincere in their belief in the same way that the original recall petitioner probably is.
That doesn't change the fact that big business in Seattle would love to see Kshama out of office, and if she survives her recall and this initiative gains traction, I'm confident they'd throw their resources behind it to get another bite at the apple. With the exception of Sawant, they have a city government that works for them, and I think they'd prefer the middle path to rocking the boat in either direction.
As we all say, correlation is not causality, but more practically, it's impossible to announce an initiative that addresses a real problem without it seeming related to something recent :-) Until the prior election is done, no one cares, and yet signature-gathering deadlines basically require starting as soon as that election ends.
(AMA about running an initiative, particularly one that doesn't have an enemy - that is, we're not trying to "Move the goalposts" towards or against any viewpoint. That's incredibly valuable (IMO!) and quite rare, but also means the typical you're-with-us-or-against-us fundraising and voter pitches don't apply. "You're with us if you support representative democracy" requires a different campaign plan targeted at people and organizations who support systems and incentives, not only candidates and positions, and who put their own preferences second to the will of the voters.)
> $15 minimum wage, the Amazon tax, numerous renters' rights protections, a ban on police use of chemical weapons against protestors, and is now continuing the fight for rent control.
Sorry for the late response. I wasn't on HN yesterday.
Regarding:
> This isn't a neutral preference, it's obviously ideological, and targeted at one specific legislator in Seattle. Voters should reject this measure as a transparent attack on the most effective and sincere fighter for working people in the country.
This isn't ideological in any way, and it's not targeted at anyone. When I said it was non-partisan and candidate-neutral, it's really that simple. It's non-partisan and candidate-neutral.
Consistent with HN guidelines, I'd ask first for an assumption of good faith. Beyond that, I can give some practical comments:
1. Our group of volunteers includes a huge range of political beliefs. There is no "one" opinion that the group could try to target even if we wanted to. The shared belief is that leaders reflect public opinion as accurately as possible.
2. This doesn't help or harm anyone. Rather, think of the current system like a grocery store where, each time you walk in, the prices on every item have doubled or halved. No one wins today: one entity (sellers/candidates) has no idea what their real support is and the other entity (buyers/voters) has to cobble together ingredients for dinner.
That wouldn't be good for the dairy people and bad for the fish people, it's bad for all sellers/candidates and all buyers/voters. Our flawed system doesn't help anyone.
Imagine a system where every voter could express their whole opinion, without feeling like they were throwing away their vote. Every election becomes a simple measure of how many voters approve of each candidate.
3. Just practically, no one's job - politician or otherwise - is worth a year or two of my life. It violates the "The sun doesn't revolve around any of us" theorem :-) Heck, even if I got to personally pick every Seattle legislator, I wouldn't put in this much effort. I don't care anywhere near enough. The thing I care about, and the only thing that's worth this much effort, is a system that delivers whatever voters want.
PS: I would welcome your help making this happen! That's how much I mean it when I say this is candidate-neutral and we welcome everyone. As volunteers, our political preferences vary widely. Our commitment is to elect winners which best reflect voters' opinions (democracy), whether or not they match any of ours.
Hey Troy, I appreciate the response! I buy that you aren't motivated by this one city council position in particular. But we can't escape the fact that every policy decision exists in our particular reality and has material effects. There's no such thing as an ideologically-neutral electoral system, of all things, because a change in system will select different candidates from the actual pool of existing candidates or types of candidates.
Wealthy elites don't care what policy choices are optimal in terms of some abstract desiderata[1], they push for the policies that will be best for them materially. Working people need to understand that and behave the same way.
To take a different example: Republicans in the U.S. push for voter ID laws because they favor Republican election outcomes. This is pretty well understood. A voter ID law appears on its face ideologically neutral. It still is meant to bring about a certain outcome, and should be fought on the basis of that outcome. It doesn't even matter if somebody pushing the legislation has a different intention, even a noble one!
---
[1] Incidentally, I'm not really sure how well AV even meets those desiderata. For example, does a non-Condorcet system really reflect voters' preferences most accurately? Later-no-harm is another criteria that comes to mind. I'm skeptical of the claim that AV is really "more reflective" in some objective sense than FPTP.
You know, when I lived in Seattle, my biggest problem when voting was never, "I approve of too many of these candidates and it's hard for me to choose just one".
Well, having lived and voted in SF and Portland, I would say that I can think of races where I would have been fine with several of the contenders, but would have really not wanted an outlier to win, so AV seems like it could work. In my undergrad days I got to experience a ranked system, which is interesting, but seems sort of inefficient.
Inefficient in what way? Surely the small amount of computing power needed to process the ballots and determine the winner is not significant when compared to the increased democratic power of the system.
Or do you mean inefficient in the sense of "more effort for voters"? You could always just rank your favorite candidate as 1 and leave the rest blank, which should give you the same amount of voting power you had in a single-vote system.
I'd argue that cardinal systems are much more efficient, on two accounts. First, you just need to do reduced sums to determine the winners, whereas ranked systems you have blocks at each round (preventing it from being asynchronous). Second, you have a better embedding space for cardinal systems than ranked. In ranked you're saying that your preference is equidistant. This is not a requirement in cardinal systems (e.g. score or star). Thus you can express that you ''really,, like candidate A over B rather than just A is better than B.
I live in Seattle and would benefit from being able to protest-vote for goodspaceguy or whoever the lunatic of the month is, at the same time I hold my nose and vote for the least awful "likely" candidate.
I don't live in the US, but I can't remember a single election where I wasn't ok with more than one candidate for majority positions. And it is really enraging to see random extremists win because the people vote strategically.
Are you sure your situation isn't due to too few candidates?
Is this a case where "one person one vote" going away may allow more candidates to actually appear?
If you're a fringe single-issue candidate, you can more easily push your issue by running on that issue (while adopting most other parts of the most similar mainstream candidate).
In our current system you would be a spoiler. With a better voting system like Approval, you would not. I prefer RCV to Approval (so the single-issue candidate can signal issue strength) but either are a big improvement to the status quo.
Voting systems is one of the few political things I'm very passionate about (those of you that know me will know I won't shut up about this). I really do think this is one of the most important things we can do to fix our democracy. I've always been saddened that Ordinal systems have captured the public eye but Cardinal systems haven't. Everything I have read suggests that Cardinal is leagues better: scalability, ease, verifiability, resistance to strategies, and vse. I'm really happy to see more and more Approval initiatives (I'm a bigger fan of STAR, but Approval is good enough that I'd shut up about voting).
I also suggest other users look for Clay[0] in these threads. He's the co-inventor of STAR and writes a lot about voting. Also a HN user.
Just getting rid of first-past-the-post would do wonders for our country and by the looks of it approval voting is inferior to run-off voting by a mile.
> by the looks of it approval voting is inferior to run-off voting by a mile.
I'm curious why you think this? I think exactly the opposite. RCV/IRV has major problems, especially for strong spoilers. They are also more difficult to tabulate and with the recent events of the last election the importance of ease to tabulate went up in my priority list (making me even question my preference of score over approval). In fact, IRV also favors extremists, which is something I really don't want.
For more context, this post has some cool interactive features to explain the voting types: https://ncase.me/ballot/
Another problem with RCV is that it’s difficult to understand, at least when you get into the runoffs. That makes it harder for people to understand and trust the election system, which is also very important.
I'm not too big of a fan of this site but it is nice. One thing I wanted to point out is the spoiler effect. Often we talk about it but similarly we generally only talk about a weak spoiler. So I wanted to show an example of a strong spoiler[0]. In this example Hexagon splits Triangles votes just by a little, but the sum of Triangle and Hex are larger than Square. But here IRV has Square winning but Approval has Hex winning.
This is one big issue for me with IRV vs Cardinal systems. Both handle _weak_ spoilers but IRV doesn't handle ''strong,, spoilers. Understanding that these are two different types of spoilers is really important. It is "Bernie vs Hillary" and "Cruz vs Trump." Even with IRV we'd still see the parties trying to prevent strong primary candidates from running independently because they could spoil the election. With cardinal methods (approval, score, star) this is far less of a concern and thus I think more democratic of an election. I honestly think we should be much more concerned with strong spoilers than weak spoilers (weak spoilers don't affect elections much anyways).
Considering our present division, I have a big "spider-sense" tingle that approval voting will simply result in everyone just voting all blue or all red, that sounds like a bad deal to me. Run off voting would give moderates a good chance of putting the country right side up again.
FWIW, this already happens. That will also happen under any system. I'm not sure why you think IRV encourages moderates. It actually encourages extremist candidates which is why I'm particularly against it.
"Everyone" doesn't vote all-red or all-blue. 30% vote all red, 30% all blue, and 40% stay home because there is never a sensible centrist candidate. Each of the 30% active voters then get mistaken for 50% of citizens.
All your assertions are actually complaints about our existing plurality voting system.
IRV/Approval/Borda/Condorcet voting systems all improve those factors. That they have some edge conditions that could be theoretically exploited belies that fact that plurality voting (ie, FPTP / one person one vote) is a complete train wreck for spoilers and favoring extreme voting.
No voting system is perfect (Arrow's impossibility theorem) but some are harder to exploit than others. Staying with a broken system because the proposed replacement isn't perfect is a poor justification.
Down below I actually give an example of IRV/RCV not solving a strong spoiler case. Clay has also written extensively on the subject matter so I'll defer to him. I'll especially defer to him on explaining how IRV actually encourages extremists.
I think the problem here is that people get caught up in VSE. I get it, that's what I was caught up in when I first started learning about this stuff. Then I had to really dig through strategic voting and spoilage. The nuance of the systems actually do make a very big difference in the outcomes.
I'd love to hear how our current electoral system is in any way preferable to RCV (or Star Voting for example).
Because for many municipalities, it's not whether they move to Approval vs. RCV or IRV or Star, etc - it's moving away from the dumpster fire of plurality voting.
It seems likely that a given municipality will only have the political energy to change voting systems once. Given that, it would be preferable to move to something that isn't RCV/IRV in the first place.
The non-monotonicity shown there for instant runoff voting seems to be more undesirable than any downsides of approval voting.
IMO the point of elected representation is to be representative, not to elect specific people, so the possibility of a lesser known candidate winning over a major party's favorite candidate isn't a problem.
> No voting system is perfect (Arrow's impossibility theorem)
Arrow's impossibility theorem only says that no _ranked_ voting system is perfect, so it doesn't apply to cardinal methods such as Approval or Score.
> That they have some edge conditions that could be theoretically exploited belies that fact that plurality voting (ie, FPTP / one person one vote) is a complete train wreck for spoilers and favoring extreme voting.
I agree, but I think the nature of the exploitation should still be considered. For IRV, we would be introducing a dependence on unauditable voting machines to tabulate votes, which would make it harder to verify the integrity of future elections.
I'd argue that instant-runoff is _worse than plurality_. It complicates the voting system, makes the wrong tradeoffs (lack of monotonicity is a very serious one), and I'd really like to hear more about why you think approval voting is worse than runoff (instant or otherwise).
IRV is also very hard to explain to average voters. Approval voting is easy to explain: you get to give a thumbs up or down to every candidate.
In a local mayoral election that used IRV, I've seen some candidates tell their supporters to vote for them in every slot. Either those candidates didn't understand IRV themselves, or they didn't want to try to explain IRV to their supporters.
Though with a candidate modification: add 'none of them / no confidence' as a virtual candidate that can never be removed by the runoff processes. If 'NoT' is the winner then the entire election is thrown out and all of the candidates are banned from the re-run.
Condorcet methods are also hard to explain to voters. The tabulation is rather complicated. At least in comparison to something like cardinal methods. In light of recent recounts I think this is a really important factor to consider. While Condorcet methods have the best VSE, they aren't winning by a huge margin. Besides, VSE isn't everything.
Condorcet methods perform, at best, only a little better than approval voting in computer simulations. And they have only ever been used in one US city (Marquette, MI) which shortly thereafter repealed it. And they are radically more complicated and unlikely to ever see the light of day.
> Condorcet methods perform, at best, only a little better than approval voting in computer simulations
Computer simulations pretty universally paper over the main problem with approval compared to any ranked-ballot (whether forced or unforced preference) methods by pretending that “approval” has a consistent, cross-voter meaning. (This is an even bigger problem with score-based methods than approval.)
Personally, for simplicity and good enough single winner methods, I prefer Bucklin which is both better than and simpler to explain than IRV, but more critically I prefer minimizing single-winner elections. 5-member legislative districts using a candidate-centric proportional method like STV or the method that has the same relationship to it as Bucklin as to IRV, which AFAIK isn't a named method, would be far superior to any possible system with single-member districts, and basically neutralize districting as a lever for partisan advantage.
Paramount executive elections with a parallel successor (e.g., US Presidential tickets, if they were directly elected, or state Gov/LtGov pairs) can even be made into phased single-winner systems to get some of the advantages of multiwinner systems (not proportionality, but you do get some ability to separate partisan and personal preference in the general election, and to thereby weaken incumbency advantage: basically, use any ranked ballots [0] single winner election for the top spot, eliminate the winner, and rerun the election on the same ballots with the same method for the constitutional successor.)
[0] while I am not partial to them, approval or score systems could be used to; simple FPTP fails because it mean those who voted for the top winner have no voice in the designated successor.
People are actually very familiar with cardinal systems. You have your 5 star ratings (score) for restaurants, apps, etc. You have your thumbs up/down on Netflix. "On a scale from 1 to 10 how do you feel about x" etc. I always see people say that cardinal systems are hard to explain but the fact is that we use systems similar to them on almost a daily basis. (At least the front end is the same)
I don't support every candidate on the ballot equally because I believe every candidate's abilities to do the job vary. First choice, second choice, third choice is not hard to explain and I do not understand why some people think Americans have been lobotimized. NYC now has run off voting and it turned out marvelously, with Eric Adams being elected.
I give an example from this year's Seattle primaries where RCV/IRV would have eliminated the candidate that most accurately reflects the electorate but Approval Voting would not have. That's not a trivial or rare situation, either, it's just such a simple race that it's easy to spot.
I have long liked and advocated for approval voting (or cardinal methods more generally), or barring that, basically anything but FPTP.
But I'm disillusioned. Many non-plurality methods (including approval) can effectively solve the spoiler problem (i.e. satisfy IIA)... but I no longer think that spoilers are the absolute most important problem to solve.
The most important problem is single-seat districts creating anti-majoritarian, disproportional legislatures. Because of how people geographically sort themselves among the likeminded, even independent districting commissions will not be able to effectively prevent anti-majoritarian legislatures.
For this reason, I believe that we would greatly benefit from explicitly proportional representation. This can be accomplished through cardinal or ordinal means - I no longer care about that dimension. That doesn't mean that we wouldn't enjoy some incremental benefit from approval or really any non-FPTP system. We would. Seattle should go ahead and do its thing. But the fight for better forms of democracy can't stop with any system that retains single-seat districts.
> Everything I have read suggests that Cardinal is leagues better
Speaking as a fellow kin who also won't shut up about this, right now we need to just acknowledge that ANYTHING is better than first-past-the-post. At this point I honestly don't care which alternative it is, none of them are worse than the current system. In every jurisdiction where any form of alternative voting system is on the ballot, I will throw my support behind whatever effort has the most momentum, regardless of whether it's approval, star, RCV, whatever. First-past-the-post is killing us with its inevitable descent into bipolarization, and we can figure out a perfect solution once we've stopped the bleeding (keeping in mind that voting system reform is merely a necessary first step, and not sufficient on its own).
I disagree with this stance. Mostly because how these methods are presented. RCV has gained such popularity because its claims of preventing spoiled elections. But people think that's Bernie spoiling Hillary and not Stein spoiling no one. RCV only prevents weak spoilers where cardinal systems prevent strong spoilers.
The reason I'm so concerned with this is because this is also why RCV got repealed in the past. IN AMERICA. This is my big fear. We all get excited about it, implement a system that doesn't solve the problems we are seeking to solve, and then go back and look for other solutions. We have a solution now, and a wide breadth of them. They just aren't the popular ones.
I think you misunderstood. Bernie made an agreement with the DNC to not run as an independent candidate. The reason being that he would have spoiled the election.
The US was stable for the entire 20th century with FPTP. If it's so bad, why was the US stable for so long with literally the exact same system we have now? Canada's used FPTP for its entire democratic existence, it's one of the most stable democracies on Earth. Britain not only uses FPTP, but uses the most extreme version possible in the Westminster system where the losing party has far fewer rights than they would in the US. Again- another perfectly stable 1st world democracy.
Meanwhile, lots of far more consensus-oriented political systems have utterly collapsed. Weimar Germany used PR- so did Argentina when Peron took over. Rwanda was using PR leading up to its famous genocide in the 90s! The number of parliamentary countries using PR that have collapsed probably exceeds the total number of FPTP countries globally.
Voting or electoral systems just don't matter that much- culture does. If you're a basically stable society, you can use any voting method and be fine- if Japan or the Scandinavian nations adopted FPTP tomorrow, nothing fundamental would change. If you're unstable, you'll be unstable no matter what type of voting system you use. Governments are not computers where a different operating system leads to different results, I'm sorry
You spent two paragraphs arguing that FPTP voting systems are responsible for the stability of the US, Canada and Britain, ("If [FPTP is] so bad, why was the US stable for so long with literally the exact same system we have now?") whereas countries that use consensus-oriented systems tend to collapse into autocracies ("The number of parliamentary countries using PR that have collapsed probably exceeds the total number of FPTP countries globally.") ... and then your final paragraph argues "Voting or electoral systems just don't matter that much- culture does. If you're a basically stable society, you can use any voting method and be fine."
So which is it? Is FPTP the lynchpin of stable democracies, or otherwise irrelevant? Are you making an explicit statement about politics or an implicit statement about culture, here?
>You spent two paragraphs arguing that FPTP voting systems are responsible for the stability of the US, Canada and Britain
I did not say that, I'm sorry that you got that impression. I am saying that political systems overall are simply much less important than many people, including anti-FPTP campaigners, believe. These examples were simply meant to illustrate the non-correlation between government structure & stability- I could just as easily point to how stable most European PR countries are, or how unstable FPTP users like Jamaica, Pakistan and Yemen can be. It's my belief that the US, Canada and Britain would be equally stable under PR. I am very much making an implicit statement about culture.
Fun fact, with a little more research I learned that both South Korea & Taiwan use FPTP for at least some parliamentary offices. Again, if the voting system makes such a huge difference....
Lee Drutman talks a lot about this in his book, "Breaking the Two-Party Doom Loop." [0]
Short version: there used to be much less "party discipline" in US politics. There were lots of sub-factions within the two big parties that would vote differently, depending on the issue. So, it was common for bills to pass with (say) 50% of Democrats voting for them, and 60% of Republicans.
In the 90's (according to Drutman), the parties started being much more "coherent" -- like, all Democrats would favor a bill, and all Republicans would oppose it. This makes it much harder to pass legislation.
I read it, I thought it was pretty handwavey. Anyways, doesn't explain how Canada, Britain, South Korea & Taiwan all manage FPTP successfully as stable democracies
> doesn't explain how Canada, Britain, South Korea & Taiwan all manage FPTP successfully as stable democracies
There's a lot of other factors in there besides the method of voting. People often compare the US to other countries over voting but ignore that the systems are substantially different. For example, the UK parliament (and many parliaments) has proportional representation. This is something we don't see in the US congress. And it is important to distinguish that presidential elections are vastly different than congressional/parliamentary elections. Then again, maybe this is part of the problem, that US is hyper focused on the presidential elections but nowhere near similarly engaged in the congressional ones.
The UK does not have proportional representation, that's not correct. FPTP & PR are opposites- you have to choose one system or the other. I agree that presidential systems are probably suboptimal, but it is worth noting that South Korea has a president in addition to FPTP, single rep districts, etc. Again my point- if you're a stable culture, you can make any system work!
On Korea: The National Assembly has 300 members elected for a four-year term, 253 in single-seat constituencies and 47 members by proportional representation.
Yes, and I believe Taiwan has started doing something similar. That's still under 16% of all seats, and- they just started doing this in the last couple of years. They were 100% similar to the US system until very recently
As a general rule, it is true that plurality voting absolutely sucks. The vast majority of highly rated democracies use something different. The UK and US are ranked 16th and 25th respectively.
Taiwan and South Korea both use a significant amount of proportional representation.
But Drutman's biggest fallacy is thinking that you have to have proportional representation. Simple approval voting would plausibly have as big or greater impact.
You missed Canada at #5. And Taiwan (#11) & SK allocating around 15% of their seats to PR- recently, as in the last three years- I would not call a 'significant amount' of PR
> Instead of reading “Vote for one,” the ballot instructs voters to “Vote for as many as you approve of.”
So they've changed it from vote for one to vote for zero. Thanks but it's been a long time since a politician I approved of made it onto my ballot. Better wording would be "Vote for as many of the lesser evils as you want." Otherwise I have to lie in order to vote.
There's a significant difference between "Not Voting" and "Voting for 0". Other than Nevada, most states don't allow "Voting for 0". The closest thing we get is writing in Mickey Mouse where writeins are allowed.
Changing the voting system changes the incentives for different kinds of candidates to run.
Suppose right now you have a Democrat, a Republican and a moderate. In first past the post, the moderate splits the vote with whichever major party candidate is most like them, causing the one less like them to win. Also, everybody knows this and then only votes for one of the major parties. So then the moderate doesn't bother to run and the only thing on the ballot is the one evil or the other.
With approval voting, if the moderate runs, most of the Republicans approve of the Republican and the moderate (because better the moderate than the Democrat) and most of the Democrats approve of the Democrat and the moderate (because better the moderate than the Republican). So then the moderate runs, and wins. And you get a candidate you might actually want to vote for.
Gentle reminder that approval voting has been tried & discarded by a number of organizations over the decades (it's literally centuries old!) The main issue is that consistently 80+% of voters 'bullet vote', or simply select one candidate despite their options. Is it a terrible flaw or something that makes AV unusable? Of course not- but it should cool some of the more heated claims.
Moreover, it's inferior to a simple runoff with a large number of candidates. As Dartmouth found, when you have 6-8+ candidates, even with AV the plurality winner will be a pretty small chunk of total votes cast (like less than 40 or even 30%). Pretty ironic result for a method that claims that to help find the consensus solution!
Unpopular opinion, FPTP with a runoff is superior to AV or RCV just for sheer workability & finding the consensus candidate. Even more unpopular opinion, voting systems just don't matter that much- healthy democracies have been plenty stable with FPTP (the UK has like the most extreme version possible), unhealthy democracies will be unstable no matter which voting system is used
> "...with AV the plurality winner will be a pretty small chunk of total votes cast (like less than 40 or even 30%). Pretty ironic result for a method that claims that to help find the consensus solution!"
that's not ironic. that's closer to the complex reality of heterogeneous interests and perspectives. what's unnatural is the idea that 50+% of a group would naturally, in an unbiased setting, pick only 2 viable candidates out of millions, and then vote consistently for one of them.
Many people find selecting a candidate for political office with 20-30% support to be deeply suboptimal. You're flirting with your government being perceived as illegitimate by regular citizens. At the national level these are incredibly powerful offices- imagine a divisive President ruling the US having won under 30% of votes cast. This is how civil wars start. Yes there are a few countries whose systems work out that way (the UK at times), and it's been criticized for decades.
>what's unnatural is the idea that 50+% of a group
This is literally how a runoff works- a pragmatic system that's been used for decades by 50+ countries globally. It also enhances democratic legitimacy, as the majority of people feel invested in having voted for the eventual winner.
(To be fair, I will say that using AV to select the top two candidates who then go to a runoff is pretty interesting)
what you're literally suggesting is that manufacturing consent is the righteous course. democracy shouldn't hinge on getting the vote 'right' or on the 'legitimacy' of a single elected representative. we're at a point in history, where we can instead have dozens/hundreds of representatives without much administrative burden via technology.
plurality of perspective and expertise is what makes a society stronger, not a singular hegemony. the legitimacy of government emerges from the totality of its actions (and non-actions), not from cult of personality.
The most important objective of a democracy should be the peaceful transition of power, facilitated by a perception of legitimacy and fairness. Anything else is secondary to that perception. Peaceful transitions of power are important foremost because political violence is abhorrent. Secondly, systems with established track records for peaceful transitions of power encourage people to feel confident about the future, which in turn encourages long term investment and robust economic growth.
In the US we do have something like a runoff. We have primary elections and then general elections. It's not quite the same thing, but neither is it the same as plain FPTP.
The American system acknowledges the fact that there are long-term stable coalitions. Even if you went to a single election with a runoff, you're going to end up with similar dynamics. Most of the candidates will say, "If I'm your first choice, your second choice should be that other guy who is in my coalition and believes most of the same things that I do."
There may well be other candidates on the ballot, but they'll usually find themselves unable to win without a coalition to present a slate of candidates. They're in exactly the same position as "third party" candidates already are.
The systems aren't identical and don't produce identical outcomes, but they're a lot closer than I think people make it out to be. A simple procedural change isn't going to fix what's wrong with the American body politic.
There is no perfect tiny party out there with all the right answers that everybody would agree with if only the two main parties would just get out of the way. That is what each of hundreds of tiny parties tell themselves. And that's why there's really no such thing as a "third party". Really, there are hundreds of "hundredth parties" who disagree with each other as much as they disagree with the two main parties, because they haven't put forth the effort of hammering out and nurturing the uneasy coalitions that go into making a dominant party.
That's a weirdness for the Presidential election, the only elected national office holder. That one is weird a lot of different ways.
But for the various state and district offices, that doesn't matter. There, you really are voting at the same time as all of the other people with an interest in it.
I'd say the bigger issue is that for a number of places, only one party really has a chance, and the general election is a foregone conclusion. Voters really only get one bite at the apple, during the primaries. (The voters for the other party never really get a bite at the apple, but they are so far outnumbered that no tweak to the voting system will fix that. They have to hope that the elected representative will keep their interests in mind.)
This is nonsense. For instance, in the last approval voting election for the board of trustees at Dartmouth, there was an average of 1.8 approvals per ballot for four candidates. The average voter voted for almost half of the candidates.
Approval voting is beloved by game theory experts specifically for its immense resistance to strategy. It was adopted by a 64% landslide majority in Fargo and a 68% landslide majority in St Louis.
Yes, approval voting is resistant to strategy. Here are Bayesian regret and VSE figures showing it performs excellently, generally better than IRV/RCV in all kinds of strategic scenarios.
Here's a mathematical proof that, even in the worst case scenario where 100% of voters are maximally strategic, approval voting elects the Condorcet winner (candidate who beats every rival head-to-head).
Moreover, in 422 of the 440 ranked choice voting elections since 2004, the winner was the candidate with the most first-place votes. So even if it was true that everyone would just bullet vote with approval voting, approval voting would still elect the same winner as RCV 96% of the time. At radically lower cost and complexity.
Gentle reminder that instant runoff voting has been tried & discarded by a number of organizations over the decades (it's literally centuries old!) It's been tried in America and discarded in the past several times in many different places. It is also how they vote in Australia (which is similarly divided as the US).
I don't disagree that Approval has issues. I think we can improve upon these with score or STAR. But I think it is very naive to start a post with such a criticism and then ignore that it applies even more strongly to the alternative method being presented (either Exhaustive Voting or Two-Round).
I'm not for IRV, you're confusing that with a 'normal' or two round runoff. In a normal runoff there are two rounds, the second round is a few weeks or a month after the first one- the French, most famously, do this for their President. I agree that IRV is pretty terrible
> Unpopular opinion, FPTP with a runoff is superior to AV or RCV just for sheer workability & finding the consensus candidate
You had me up until here, but there's no way this claim is true. FPTP with a runoff is exactly how San Francisco elections are held, and it has made the elections there essentially a one-party system. Voters get to choose between two establishment wings of the Democratic Party at each runoff. It's even worse than plain old FPTP.
San Francisco uses ranked choice voting or IRV, not a two-round runoff. I should've used a different term to avoid confusion. Here's what I mean by a two-round system https://en.wikipedia.org/wiki/Two-round_system
I don't think that the voting system has made SF 'essentially a one-party system', as tons & tons of cities across the US just use normal FPTP and are still very Democratic. Just as tons of rural states just use normal FPTP and are still very Republican. This is what I mean by, voting systems are overrated and just don't have the large effects that enthusiasts want them to have. It's a seductive argument for left-brained engineering types who are systems thinkers, but it's just not empirically true
California has a top-two primary system, not IRV, which as far as I can tell is the same as the "two-round system" you link to. So the primary is a free-for-all and the actual election has the top two polling candidates from the primaries, regardless of party affiliation.
This was sold as a way of breaking the two-party duopoly, which it did, but not in the expected way. Instead of getting third parties or independents into that second spot, in heavily blue areas like SF they instead typically get the democratic vote split among two candidates, who go on to compete against each other in the general election. So now instead of third parties being locked by the mechanics of FPTP voting, they get third parties and republicans locked out! Went from a duopoly to a monopoly.
(E.g. the last senate election in California was between Dianne Feinstein and Kevin de León, both establishment dems. Nancy Pelosi also faced a democrat in 2020.)
The key feature of this proposal is that it only applies in our open primaries. The top two vote getters advance to the general election for a normal FPTP with only two candidates. Personally, I think this is an elegant way to get third parties into the process without the complexity of the various IRV concepts.
This makes it sound very much like California's system. As garmaine described elsewhere in this discussion[1], you won't get third parties, you'll get one party.
I'm confused by how "all you approve of" should work. In an election with Center-Left, Center-Right and Fascisf, if my preferences are in that order, do I approve of the Center-Right candidate? I suppose they are better than the Fascist, but I wouldn't really approve of them except if it was a choice between them and the Fascist. Is "approves of" dependent on the candidate set?
I think that's part of the point -- you get to draw that line and decide if you're more in favor of promoting your favorite over all others vs opposing your least favorite. Where you decide to put your approval mark favors one vs the other.
Voting schemes aren’t Seattle’s problem. It’s things like district based elections for council members, maneuvers to set up low income housing in particular areas to skew district voting patterns (see Magnuson Park), sitting council members (see Kshama Sawant) abusing city resources for political gain, a rioting extremist left that is willing to use violence for political gain, suspicious late vote swings in one direction (left), ballot harvesting of the homeless, and so on. These things undermine the city and its constituency, and unfortunately it may be too late to reverse the shift in political power that has resulted.
134 comments
[ 3.6 ms ] story [ 178 ms ] threadIf anyone would like to invest time or money to make this happen, please contact us or donate here: https://SeattleApproves.org/
[1]: https://electionscience.github.io/vse-sim/VSEbasic/ , https://www.youtube.com/watch?v=7btAd1HYvjU&t=1329
[0] https://electionscience.github.io/vse-sim/VSE/
[1] https://www.youtube.com/watch?v=-4FXLQoLDBA
one method of implementing this (from thinking about it for all of 2 minutes), is developing a list of the top 10-20 issues that voters care about in a given election and providing tradeoff choices for each, then matching them to the candidate closest to their revealed preference. of course, the biggest issue with method is that it moves the problem up one level, from voting to choice design.
[0]: https://wikipedia.org/wiki/Conjoint_analysis
the neat thing about conjoint is that it multiplexes the choice-making (when designed well), so that the choice burden scales reasonably while also better representing collective choice. nothing is going to be perfect, but conjoint could potentially be better than all current systems, probably by an order of magnitude or more.
it's relatively complicated, so yes, 'experts' (i disdain this term, frankly) could in theory manipulate the system, but transparency of voting data (just like now) and survey design (the more likely bias vector) would make that difficult. transparency would allow competing independent journalists, academics/researchers, political analysts, etc. to weigh in and balance out biases (not perfectly of course), as well as verify results.
For politicians, for me, it's more about (perceived) character vs. their stated policies.
currently, there is no option to choose a candidate supporting marijuana legalization, $15 minimum wage, pro-life, and pro-gun policies (as an example), since we practically only have two relatively fixed packages of pre-ordained, all-or-nothing policy positions to choose from. revealed preferences would pressure parties into fielding candidates that better match the non-partisan preferences of constituents. or better yet, it would lead to the waning of the two-party system altogether, toward a more direct representative democracy.
Additionally, all of this ignores the possibility that a representative could change their stance on an issue after receiving more information on an issue. Presidents and Supreme Court justices have changes their stances once their terms have begun or events have taken place during their time in office. I would hate a system that relies on tightly binding the representatives to narrow definitions of hot-button issue stances.
That's all academic, though. A great real example is the 2021 Seattle City Attorney primary, which had only 3 candidates. The votes were split: 36%, 32%, 30%. These 2 candidates emerged from the primary: https://www.nytimes.com/2021/10/30/us/seattle-city-attorney-.... The candidate that most accurately represented the opinion of the electorate, Holmes, was eliminated. That would have still happened in a ranked ballot. The voters who voted for Thomas-Kennedy or Davison in our current system would have still put Thomas-Kennedy or Davison #1 in a ranked ballot.
I tried to distill this in https://twitter.com/troyd/status/1461389190206156802. And a 3-way race like that is actually the simplest possible case. Effects like that are much more likely with more candidates and they happen everywhere in the elimination cycle, they’re just less obvious to casual observers. That's what you see in https://www.youtube.com/watch?v=7btAd1HYvjU&t=1329 (note that IRV == RCV in that video).
(Obligatory disclaimer: I don't care who wins. I do care how accurately the winners reflect the inputs, namely, voter opinion.)
Separately, Washington law requires cities hold a primary that produces 2 candidates for the general election, so cities can't legally eliminate primaries. That said, whether primaries are a net positive or negative is a really complicated discussion. Beyond election mechanics, the disadvantage of primaries is that only a subset of voters in the general will turn out in the primary. The advantage is that voters don't spend much time on ballots and it focuses that time. Seattle had 15 candidates for Mayor (many races are 10+). If someone spends 10 minutes on their ballot and the ballot has 10 races, that's 1 minute per race - 6 seconds per candidate. No easy answer here. However, for practical purposes we don't need to answer that hard question; this one change delivers an incredibly representative election.
[1] https://electionscience.org/library/approval-voting-versus-i...
One issue seems to be domination by majority?
For example, let's say 55% of electorate is hard left, 45% moderate to conservative.
Under approval those 55% can vote for all 3 lets say hard left candidates, no moderate's make it in?
RCV, once someone has gotten their candidate in, their vote is exhausted. Seems to come out much better. Still might end up with 2 hard left, one moderate, but that seems fairer to me.
I think the appeal of the simple ballot is what drove adoption (ie, vote for as many as you like, top X win).
In California the trick is they treat all open seats as one office.
https://m.youtube.com/watch?v=-4FXLQoLDBA
The ballot is approval style voting, you can vote for X of Y candidates.
Approval voting is a single winner system.
If they're just adding up each candidate's votes, that's not how approval is supposed to be tabulated in a multi-winner context (which is more complicated); see https://en.wikipedia.org/wiki/Proportional_approval_voting
https://en.wikipedia.org/wiki/Multiple_non-transferable_vote
> RCV, once someone has gotten their candidate in, their vote is exhausted.
Hi, Australian here. What happens with multi-winner single transferable vote is that there's a _quota_ and then ballots for candidates who have reached quota get _reweighted_ (not quite the same thing).
This quota-and-reweighting system can (and must!) be applied to multi-winner Approval too - jpfed linked the wiki article.
Where I see trouble these days is with Federal elections. Being able to vote for anyone but X or Z would be a wonderful way to send a signal as a voter. I suspect that both the east and west sides of the state would feel similarly (East: "Anyone but X!", West: "Anyone but Y!". Result: M gets elected, who is thankfully neither X nor Y.).
Same thing that makes it difficult by attempting to do it country wide. That it is larger and more difficult. It would be nice if we could just wave a hand and change everything at once, but that also wouldn't be very democratic. What is also nice is that we can experiment, refine, and test at smaller levels before we advance to larger elections.
Someone else answered this well, but: money. A Seattle initiative is roughly 10%-20% as expensive as a statewide one, and it's still an incredible challenge to raise enough to do this. Ask me where I spend my time.
Qualifying for the ballot in Seattle requires about 26,200 valid signatures (10% of votes cast in the last Mayoral election). They must be collected during a 180-day period that, due to deadlines to make the November ballot, must is basically January-June. Since January-March is wet and cold, people aren't outside as much and many initiatives just do April-June.
Collecting 26,000 signatures in 90-180 days is all but impossible with volunteers, which means the core challenge for an initiative is fundraising to pay signature-gathering costs. You can divide 26,000 by any number of valid signatures collected per hour - 3, 6, 12, whatever - and it's still a lot of time and thus really expensive.
So, that's for the Seattle ballot. Getting on the WA state ballot requires 324,516 (!). Yes, 12x more. And that cost scales roughly linearly; the few efficiencies of scale are offset by new inefficiencies like travel. There's some efficiencies with overhead, but that's a small % of the cost.
So, that's why. It takes 5-10x more money.
Of course, voters would need to support it statewide too. In Seattle, this will pass; our job is to get it on the ballot. A statewide campaign may also need to do at least an education/awareness campaign (roughly the last paragraph of your comment) and potentially even try to change public opinion. These can cost even more than collecting signatures.
(If you're reading this and thinking "Wow, Troy knows this stuff - and it's interesting and there's a plan to do this!" - it is and there is! I'd love your help as a volunteer and/or your donations.)
Kshama Sawant fought for the $15 minimum wage, the Amazon tax, numerous renters' rights protections, a ban on police use of chemical weapons against protestors, and is now continuing the fight for rent control. She's currently facing a right-wing recall campaign because big business was unable to defeat her in the 2019 election, despite Amazon dumping $1.5 million into various races for citywide office.
A few here have remarked that approval voting tends to select more "moderate" candidates over "extremes." This isn't a neutral preference, it's obviously ideological, and targeted at one specific legislator in Seattle. Voters should reject this measure as a transparent attack on the most effective and sincere fighter for working people in the country.
Full disclosure, Kshama and I are both members of Socialist Alternative. I may be limited in my ability to respond today, as I'm rushing right after work to a volunteer shift to make sure she stays in office. I'll try my best to check in occasionally!
That doesn't change the fact that big business in Seattle would love to see Kshama out of office, and if she survives her recall and this initiative gains traction, I'm confident they'd throw their resources behind it to get another bite at the apple. With the exception of Sawant, they have a city government that works for them, and I think they'd prefer the middle path to rocking the boat in either direction.
I can answer this conclusively: it's not a response to any single election or event. We've been working on this for almost 2 years. Here's a tweet from June 2020: https://twitter.com/SeattleApproves/status/12776993817564528...
As we all say, correlation is not causality, but more practically, it's impossible to announce an initiative that addresses a real problem without it seeming related to something recent :-) Until the prior election is done, no one cares, and yet signature-gathering deadlines basically require starting as soon as that election ends.
(AMA about running an initiative, particularly one that doesn't have an enemy - that is, we're not trying to "Move the goalposts" towards or against any viewpoint. That's incredibly valuable (IMO!) and quite rare, but also means the typical you're-with-us-or-against-us fundraising and voter pitches don't apply. "You're with us if you support representative democracy" requires a different campaign plan targeted at people and organizations who support systems and incentives, not only candidates and positions, and who put their own preferences second to the will of the voters.)
Those are not adjectives that I think the majority of people would ascribe to Sawant.
As a moderate, are these extremist ideals now?
Regarding:
> This isn't a neutral preference, it's obviously ideological, and targeted at one specific legislator in Seattle. Voters should reject this measure as a transparent attack on the most effective and sincere fighter for working people in the country.
This isn't ideological in any way, and it's not targeted at anyone. When I said it was non-partisan and candidate-neutral, it's really that simple. It's non-partisan and candidate-neutral.
Consistent with HN guidelines, I'd ask first for an assumption of good faith. Beyond that, I can give some practical comments:
1. Our group of volunteers includes a huge range of political beliefs. There is no "one" opinion that the group could try to target even if we wanted to. The shared belief is that leaders reflect public opinion as accurately as possible.
2. This doesn't help or harm anyone. Rather, think of the current system like a grocery store where, each time you walk in, the prices on every item have doubled or halved. No one wins today: one entity (sellers/candidates) has no idea what their real support is and the other entity (buyers/voters) has to cobble together ingredients for dinner.
That wouldn't be good for the dairy people and bad for the fish people, it's bad for all sellers/candidates and all buyers/voters. Our flawed system doesn't help anyone.
Imagine a system where every voter could express their whole opinion, without feeling like they were throwing away their vote. Every election becomes a simple measure of how many voters approve of each candidate.
3. Just practically, no one's job - politician or otherwise - is worth a year or two of my life. It violates the "The sun doesn't revolve around any of us" theorem :-) Heck, even if I got to personally pick every Seattle legislator, I wouldn't put in this much effort. I don't care anywhere near enough. The thing I care about, and the only thing that's worth this much effort, is a system that delivers whatever voters want.
PS: I would welcome your help making this happen! That's how much I mean it when I say this is candidate-neutral and we welcome everyone. As volunteers, our political preferences vary widely. Our commitment is to elect winners which best reflect voters' opinions (democracy), whether or not they match any of ours.
Wealthy elites don't care what policy choices are optimal in terms of some abstract desiderata[1], they push for the policies that will be best for them materially. Working people need to understand that and behave the same way.
To take a different example: Republicans in the U.S. push for voter ID laws because they favor Republican election outcomes. This is pretty well understood. A voter ID law appears on its face ideologically neutral. It still is meant to bring about a certain outcome, and should be fought on the basis of that outcome. It doesn't even matter if somebody pushing the legislation has a different intention, even a noble one!
---
[1] Incidentally, I'm not really sure how well AV even meets those desiderata. For example, does a non-Condorcet system really reflect voters' preferences most accurately? Later-no-harm is another criteria that comes to mind. I'm skeptical of the claim that AV is really "more reflective" in some objective sense than FPTP.
Or do you mean inefficient in the sense of "more effort for voters"? You could always just rank your favorite candidate as 1 and leave the rest blank, which should give you the same amount of voting power you had in a single-vote system.
Are you sure your situation isn't due to too few candidates?
If you're a fringe single-issue candidate, you can more easily push your issue by running on that issue (while adopting most other parts of the most similar mainstream candidate).
In our current system you would be a spoiler. With a better voting system like Approval, you would not. I prefer RCV to Approval (so the single-issue candidate can signal issue strength) but either are a big improvement to the status quo.
I also suggest other users look for Clay[0] in these threads. He's the co-inventor of STAR and writes a lot about voting. Also a HN user.
[0] https://twitter.com/ClayShentrup
I'm curious why you think this? I think exactly the opposite. RCV/IRV has major problems, especially for strong spoilers. They are also more difficult to tabulate and with the recent events of the last election the importance of ease to tabulate went up in my priority list (making me even question my preference of score over approval). In fact, IRV also favors extremists, which is something I really don't want.
Another problem with RCV is that it’s difficult to understand, at least when you get into the runoffs. That makes it harder for people to understand and trust the election system, which is also very important.
This is one big issue for me with IRV vs Cardinal systems. Both handle _weak_ spoilers but IRV doesn't handle ''strong,, spoilers. Understanding that these are two different types of spoilers is really important. It is "Bernie vs Hillary" and "Cruz vs Trump." Even with IRV we'd still see the parties trying to prevent strong primary candidates from running independently because they could spoil the election. With cardinal methods (approval, score, star) this is far less of a concern and thus I think more democratic of an election. I honestly think we should be much more concerned with strong spoilers than weak spoilers (weak spoilers don't affect elections much anyways).
[0] https://imgur.com/a/vV15A7Z
STAR: 5, Score 4, approval 4, IRV 1, FPTP 1
It would be ranked in this order too, but there's more expressivity here than an ordinal system can provide.
FWIW, this already happens. That will also happen under any system. I'm not sure why you think IRV encourages moderates. It actually encourages extremist candidates which is why I'm particularly against it.
IRV/Approval/Borda/Condorcet voting systems all improve those factors. That they have some edge conditions that could be theoretically exploited belies that fact that plurality voting (ie, FPTP / one person one vote) is a complete train wreck for spoilers and favoring extreme voting.
No voting system is perfect (Arrow's impossibility theorem) but some are harder to exploit than others. Staying with a broken system because the proposed replacement isn't perfect is a poor justification.
I think the problem here is that people get caught up in VSE. I get it, that's what I was caught up in when I first started learning about this stuff. Then I had to really dig through strategic voting and spoilage. The nuance of the systems actually do make a very big difference in the outcomes.
Because for many municipalities, it's not whether they move to Approval vs. RCV or IRV or Star, etc - it's moving away from the dumpster fire of plurality voting.
The non-monotonicity shown there for instant runoff voting seems to be more undesirable than any downsides of approval voting.
IMO the point of elected representation is to be representative, not to elect specific people, so the possibility of a lesser known candidate winning over a major party's favorite candidate isn't a problem.
Arrow's impossibility theorem only says that no _ranked_ voting system is perfect, so it doesn't apply to cardinal methods such as Approval or Score.
> That they have some edge conditions that could be theoretically exploited belies that fact that plurality voting (ie, FPTP / one person one vote) is a complete train wreck for spoilers and favoring extreme voting.
I agree, but I think the nature of the exploitation should still be considered. For IRV, we would be introducing a dependence on unauditable voting machines to tabulate votes, which would make it harder to verify the integrity of future elections.
I tend to not get into this fight because 1) it extends to cardinal systems with Gibbard's and 2) Arrow's is more about social choice than voting.
In a local mayoral election that used IRV, I've seen some candidates tell their supporters to vote for them in every slot. Either those candidates didn't understand IRV themselves, or they didn't want to try to explain IRV to their supporters.
Though with a candidate modification: add 'none of them / no confidence' as a virtual candidate that can never be removed by the runoff processes. If 'NoT' is the winner then the entire election is thrown out and all of the candidates are banned from the re-run.
https://electionscience.github.io/vse-sim/vse.html
Computer simulations pretty universally paper over the main problem with approval compared to any ranked-ballot (whether forced or unforced preference) methods by pretending that “approval” has a consistent, cross-voter meaning. (This is an even bigger problem with score-based methods than approval.)
Personally, for simplicity and good enough single winner methods, I prefer Bucklin which is both better than and simpler to explain than IRV, but more critically I prefer minimizing single-winner elections. 5-member legislative districts using a candidate-centric proportional method like STV or the method that has the same relationship to it as Bucklin as to IRV, which AFAIK isn't a named method, would be far superior to any possible system with single-member districts, and basically neutralize districting as a lever for partisan advantage.
Paramount executive elections with a parallel successor (e.g., US Presidential tickets, if they were directly elected, or state Gov/LtGov pairs) can even be made into phased single-winner systems to get some of the advantages of multiwinner systems (not proportionality, but you do get some ability to separate partisan and personal preference in the general election, and to thereby weaken incumbency advantage: basically, use any ranked ballots [0] single winner election for the top spot, eliminate the winner, and rerun the election on the same ballots with the same method for the constitutional successor.)
[0] while I am not partial to them, approval or score systems could be used to; simple FPTP fails because it mean those who voted for the top winner have no voice in the designated successor.
https://vote.nyc/page/ranked-choice-voting
I give an example from this year's Seattle primaries where RCV/IRV would have eliminated the candidate that most accurately reflects the electorate but Approval Voting would not have. That's not a trivial or rare situation, either, it's just such a simple race that it's easy to spot.
Always good to see you Clay. Hope you are doing well.
But I'm disillusioned. Many non-plurality methods (including approval) can effectively solve the spoiler problem (i.e. satisfy IIA)... but I no longer think that spoilers are the absolute most important problem to solve.
The most important problem is single-seat districts creating anti-majoritarian, disproportional legislatures. Because of how people geographically sort themselves among the likeminded, even independent districting commissions will not be able to effectively prevent anti-majoritarian legislatures.
For this reason, I believe that we would greatly benefit from explicitly proportional representation. This can be accomplished through cardinal or ordinal means - I no longer care about that dimension. That doesn't mean that we wouldn't enjoy some incremental benefit from approval or really any non-FPTP system. We would. Seattle should go ahead and do its thing. But the fight for better forms of democracy can't stop with any system that retains single-seat districts.
Speaking as a fellow kin who also won't shut up about this, right now we need to just acknowledge that ANYTHING is better than first-past-the-post. At this point I honestly don't care which alternative it is, none of them are worse than the current system. In every jurisdiction where any form of alternative voting system is on the ballot, I will throw my support behind whatever effort has the most momentum, regardless of whether it's approval, star, RCV, whatever. First-past-the-post is killing us with its inevitable descent into bipolarization, and we can figure out a perfect solution once we've stopped the bleeding (keeping in mind that voting system reform is merely a necessary first step, and not sufficient on its own).
I disagree with this stance. Mostly because how these methods are presented. RCV has gained such popularity because its claims of preventing spoiled elections. But people think that's Bernie spoiling Hillary and not Stein spoiling no one. RCV only prevents weak spoilers where cardinal systems prevent strong spoilers.
The reason I'm so concerned with this is because this is also why RCV got repealed in the past. IN AMERICA. This is my big fear. We all get excited about it, implement a system that doesn't solve the problems we are seeking to solve, and then go back and look for other solutions. We have a solution now, and a wide breadth of them. They just aren't the popular ones.
Meanwhile, lots of far more consensus-oriented political systems have utterly collapsed. Weimar Germany used PR- so did Argentina when Peron took over. Rwanda was using PR leading up to its famous genocide in the 90s! The number of parliamentary countries using PR that have collapsed probably exceeds the total number of FPTP countries globally.
Voting or electoral systems just don't matter that much- culture does. If you're a basically stable society, you can use any voting method and be fine- if Japan or the Scandinavian nations adopted FPTP tomorrow, nothing fundamental would change. If you're unstable, you'll be unstable no matter what type of voting system you use. Governments are not computers where a different operating system leads to different results, I'm sorry
So which is it? Is FPTP the lynchpin of stable democracies, or otherwise irrelevant? Are you making an explicit statement about politics or an implicit statement about culture, here?
I did not say that, I'm sorry that you got that impression. I am saying that political systems overall are simply much less important than many people, including anti-FPTP campaigners, believe. These examples were simply meant to illustrate the non-correlation between government structure & stability- I could just as easily point to how stable most European PR countries are, or how unstable FPTP users like Jamaica, Pakistan and Yemen can be. It's my belief that the US, Canada and Britain would be equally stable under PR. I am very much making an implicit statement about culture.
Fun fact, with a little more research I learned that both South Korea & Taiwan use FPTP for at least some parliamentary offices. Again, if the voting system makes such a huge difference....
Short version: there used to be much less "party discipline" in US politics. There were lots of sub-factions within the two big parties that would vote differently, depending on the issue. So, it was common for bills to pass with (say) 50% of Democrats voting for them, and 60% of Republicans.
In the 90's (according to Drutman), the parties started being much more "coherent" -- like, all Democrats would favor a bill, and all Republicans would oppose it. This makes it much harder to pass legislation.
[0] https://www.amazon.com/Breaking-Two-Party-Doom-Loop-Multipar...
There's a lot of other factors in there besides the method of voting. People often compare the US to other countries over voting but ignore that the systems are substantially different. For example, the UK parliament (and many parliaments) has proportional representation. This is something we don't see in the US congress. And it is important to distinguish that presidential elections are vastly different than congressional/parliamentary elections. Then again, maybe this is part of the problem, that US is hyper focused on the presidential elections but nowhere near similarly engaged in the congressional ones.
https://thefulcrum.us/amp/actually-approval-voting-beats-rcv...
As a general rule, it is true that plurality voting absolutely sucks. The vast majority of highly rated democracies use something different. The UK and US are ranked 16th and 25th respectively.
https://en.m.wikipedia.org/wiki/Democracy_Index
Taiwan and South Korea both use a significant amount of proportional representation.
But Drutman's biggest fallacy is thinking that you have to have proportional representation. Simple approval voting would plausibly have as big or greater impact.
So they've changed it from vote for one to vote for zero. Thanks but it's been a long time since a politician I approved of made it onto my ballot. Better wording would be "Vote for as many of the lesser evils as you want." Otherwise I have to lie in order to vote.
[1]: https://www.sos.wa.gov/elections/faq_vote_by_mail.aspx
Suppose right now you have a Democrat, a Republican and a moderate. In first past the post, the moderate splits the vote with whichever major party candidate is most like them, causing the one less like them to win. Also, everybody knows this and then only votes for one of the major parties. So then the moderate doesn't bother to run and the only thing on the ballot is the one evil or the other.
With approval voting, if the moderate runs, most of the Republicans approve of the Republican and the moderate (because better the moderate than the Democrat) and most of the Democrats approve of the Democrat and the moderate (because better the moderate than the Republican). So then the moderate runs, and wins. And you get a candidate you might actually want to vote for.
The IEEE, Mathematical Association of America, and the Dartmouth Alumni Association have all tried & discarded AV- as they consistently found that the vast majority of voters bullet vote https://en.wikipedia.org/wiki/Approval_voting#Other_organiza...
Moreover, it's inferior to a simple runoff with a large number of candidates. As Dartmouth found, when you have 6-8+ candidates, even with AV the plurality winner will be a pretty small chunk of total votes cast (like less than 40 or even 30%). Pretty ironic result for a method that claims that to help find the consensus solution!
Unpopular opinion, FPTP with a runoff is superior to AV or RCV just for sheer workability & finding the consensus candidate. Even more unpopular opinion, voting systems just don't matter that much- healthy democracies have been plenty stable with FPTP (the UK has like the most extreme version possible), unhealthy democracies will be unstable no matter which voting system is used
that's not ironic. that's closer to the complex reality of heterogeneous interests and perspectives. what's unnatural is the idea that 50+% of a group would naturally, in an unbiased setting, pick only 2 viable candidates out of millions, and then vote consistently for one of them.
>what's unnatural is the idea that 50+% of a group
This is literally how a runoff works- a pragmatic system that's been used for decades by 50+ countries globally. It also enhances democratic legitimacy, as the majority of people feel invested in having voted for the eventual winner.
(To be fair, I will say that using AV to select the top two candidates who then go to a runoff is pretty interesting)
plurality of perspective and expertise is what makes a society stronger, not a singular hegemony. the legitimacy of government emerges from the totality of its actions (and non-actions), not from cult of personality.
https://web.archive.org/web/20190219005032/https://sites.goo...
The American system acknowledges the fact that there are long-term stable coalitions. Even if you went to a single election with a runoff, you're going to end up with similar dynamics. Most of the candidates will say, "If I'm your first choice, your second choice should be that other guy who is in my coalition and believes most of the same things that I do."
There may well be other candidates on the ballot, but they'll usually find themselves unable to win without a coalition to present a slate of candidates. They're in exactly the same position as "third party" candidates already are.
The systems aren't identical and don't produce identical outcomes, but they're a lot closer than I think people make it out to be. A simple procedural change isn't going to fix what's wrong with the American body politic.
There is no perfect tiny party out there with all the right answers that everybody would agree with if only the two main parties would just get out of the way. That is what each of hundreds of tiny parties tell themselves. And that's why there's really no such thing as a "third party". Really, there are hundreds of "hundredth parties" who disagree with each other as much as they disagree with the two main parties, because they haven't put forth the effort of hammering out and nurturing the uneasy coalitions that go into making a dominant party.
But for the various state and district offices, that doesn't matter. There, you really are voting at the same time as all of the other people with an interest in it.
I'd say the bigger issue is that for a number of places, only one party really has a chance, and the general election is a foregone conclusion. Voters really only get one bite at the apple, during the primaries. (The voters for the other party never really get a bite at the apple, but they are so far outnumbered that no tweak to the voting system will fix that. They have to hope that the elected representative will keep their interests in mind.)
http://scorevoting.net/BulletBugaboo
Approval voting is beloved by game theory experts specifically for its immense resistance to strategy. It was adopted by a 64% landslide majority in Fargo and a 68% landslide majority in St Louis.
https://www.rangevoting.org/BayRegsFig
https://rpubs.com/Jameson-Quinn/VSE5key
Here's a mathematical proof that, even in the worst case scenario where 100% of voters are maximally strategic, approval voting elects the Condorcet winner (candidate who beats every rival head-to-head).
https://www.rangevoting.org/AppCW
Moreover, in 422 of the 440 ranked choice voting elections since 2004, the winner was the candidate with the most first-place votes. So even if it was true that everyone would just bullet vote with approval voting, approval voting would still elect the same winner as RCV 96% of the time. At radically lower cost and complexity.
https://clayshentrup.medium.com/bullet-voting-is-fine-52e6f2...
I don't disagree that Approval has issues. I think we can improve upon these with score or STAR. But I think it is very naive to start a post with such a criticism and then ignore that it applies even more strongly to the alternative method being presented (either Exhaustive Voting or Two-Round).
https://en.wikipedia.org/wiki/Two-round_system
You had me up until here, but there's no way this claim is true. FPTP with a runoff is exactly how San Francisco elections are held, and it has made the elections there essentially a one-party system. Voters get to choose between two establishment wings of the Democratic Party at each runoff. It's even worse than plain old FPTP.
I don't think that the voting system has made SF 'essentially a one-party system', as tons & tons of cities across the US just use normal FPTP and are still very Democratic. Just as tons of rural states just use normal FPTP and are still very Republican. This is what I mean by, voting systems are overrated and just don't have the large effects that enthusiasts want them to have. It's a seductive argument for left-brained engineering types who are systems thinkers, but it's just not empirically true
This was sold as a way of breaking the two-party duopoly, which it did, but not in the expected way. Instead of getting third parties or independents into that second spot, in heavily blue areas like SF they instead typically get the democratic vote split among two candidates, who go on to compete against each other in the general election. So now instead of third parties being locked by the mechanics of FPTP voting, they get third parties and republicans locked out! Went from a duopoly to a monopoly.
(E.g. the last senate election in California was between Dianne Feinstein and Kevin de León, both establishment dems. Nancy Pelosi also faced a democrat in 2020.)
> consistently 80+% of voters 'bullet vote', or simply select one candidate
St. Louis's election with Approval Voting in March 2021 seems to dispel that. In a Mayoral race with 4 candidates, voters approved an average of 1.56 candidates (https://stlouis-mo.gov/government/departments/board-election...).
Voters understood and used the system, and they didn't bullet vote. 1.56 is a ton of additional expression.
As I wrote in https://twitter.com/SeattleApproves/status/14589200428425748..., with 15 Mayoral candidates in 2021, Seattle's average may be much higher. It's quite possibly the average will be above 2.
https://80000hours.org/podcast/episodes/aaron-hamlin-voting-...
It's an amazing in-depth interview about Approval and other types of voting systems.
[1] https://news.ycombinator.com/item?id=29271948
This is something everyone can get behind. Literally Democrat or Republican, you would be happier with more parties and more options.
There really shouldn't be anyone on the other side of this issue except the politicians who use the two party system to keep power among the few.