Comments

1

Hey, just thought I'd point out how RCV and Approval would affect the City Attorney's race:

Under RCV, we'd get the same outcome we had: Pete Holmes would have had the fewest 1st place votes and be eliminated, and Davison would go on to beat NTK in the general election.

Under Approval Voting, voters could have registered a vote for two candidates. It's highly likely that Holmes would have advanced due to crossover voters that selected Holmes-Davison or Holmes-NTK.

Who knows if it would have been an NTK-Holmes or a Davison-Holmes general, but it almost certainly would have been one of the two, and that would have been better than the outcome we got.

2

@1 I agree. I think that is a much better example. I also think the District 9 race would have been different. The consensus candidate (Thomas) would have made it through to the general election, and probably would have won.

I don't think the mayoral election would have been any different under approval voting. Gonzalez was popular, it is just that Harrell was more popular.

I should point out that RCV systems aren't all the same. Under some systems, Holmes would have made it to the general. It actually gets very complicated -- there are a bunch of different systems; each have their strengths and weaknesses. https://en.wikipedia.org/wiki/Ranked_voting, https://en.wikipedia.org/wiki/Comparison_of_electoral_systems.

3

@2 Yeah, there are a TON of systems. We like to look at Voter Satisfaction Efficiency to compare them. All the RCV systems in use today would have had the City Attorney race problem, and the hypothetical ones that wouldn't are so complicated to tabulate I don't think the average voter would have any idea how the election actually works.

Approval Voting strikes the nice balance of being near the top in terms of selection representative winners, be very easy to use, and is legal today.

But if you want to get your nerd on, check out STAR voting. That's another great option, it's just too complicated, IMHO. In the current world where election integrity is constantly questioned, I think it's really important to have simple tabulation and rules.

4

Would approval voting get the overall percentage of eligible voters to increase?
Ranked choice type voting makes more sense, in that the voter tells you their preference. Non-preferential voting means a Mike the Mover vote equals a vote for anyone else.

5

Wouldn't this just automatically rule out Stranger endorsed candidates? Anyone not brainwashed knows that their ideas are extremely unpopular and why they lose year after year after year.

7

Approval voting will pick the same winner as ranked choice voting 95% of the time, so you might as well go with the simpler more transparent option, which is approval voting.

How do we know? As of September 2021, there have been 440 single-winner RCV elections in modern history (since 2004). In 422 of them, the winner was the person with the most first-place votes. As long the that same candidate had been in the top two most approved, then the same outcome would have occurred under the approval+top-two system being proposed.

Another analyst put it more starkly:

In the 1362 IRV elections I’ve personally looked into:
- 1263 went to the person who was in first on the first round of counting
- 98 went to the person who was in second on the first round of counting
- 4 went to the person who was in third on the first round of counting
- none went to anyone who was worse than third on the first round of counting

And in those very rare cases where the approval voting results would have differed, it seems to have the upper hand. Approval voting likely would have sent Pete Holmes to the general, where he quite plausibly would have beaten Davidson.

Kamau Chege calls this compromise result "cheese pizza", but as a vegan, I think such a compromise outcome is precisely what we want in this age of polarization. This needn't imply boring or "milquetoast" candidates however. In the March 2021 mayoral primary in St Louis, approval voting advanced two progressive women to the general, resulting in their first black female mayor.

https://electionscience.org/press-releases/st-louis-voters-use-new-approval-voting-system-in-march-primary-election/

And a 2019 national approval voting poll of Democratic voters had Elizabeth Warren 1st and Bernie Sanders 2nd, with Pete Buttigieg 3rd.

https://electionscience.org/press-releases/new-poll-74-of-democratic-primary-voters-would-support-warren-for-president/

As for those RCV proponents who wish to eliminate the primary altogether, be careful what you wish for. In the 2021 St Louis elections, which used approval voting in the primary, two of the 16 races featured come-from-behind victories for the (generally more progressive) underdog. Apparently winnowing the field down to just two candidates gave the public an opportunity to learn more about their options, and that helped those underdogs get their message out and compete against the "machine" candidates.

9

@1: Seattle's City Attorney's elections went exactly the way they should have, given how much voters hated the status quo in Seattle.

First, voters in the primary election opposed the status quo, so they eliminated the incumbent. Second, in the general election, the voters still opposed the status quo, so they eliminated the candidate (NTK) who promised even more of it. That left Davison the winner.

The Stranger has a problem with this outcome, because the Stranger is very, very happy with encampments and petty crime everywhere, and didn't want voters changing that. Hence the Stranger's attempt to ice out Davison, which promptly exploded the the Stranger's face. The Stranger won't admit it is itself the problem here, and so goes looking for ways to tinker with election mechanisms in the hopes of preventing voters from getting what voters want.

10

@8 "More reason to never adopt RCV. If any of those three had been at the top of the ticket in November 2020 we’d be finishing up the first year of Trump’s second term."

Unless...unless...Trump and the other Republican candidates were all running alongside Bernie, Warren, et al in one big approval voting election, and the top two most broadly appealing were squaring off in the general. The proposal isn't to have partisan primaries, of course.

11

Approval Voting worked great in St Louis. They implemented it in 4 months and fixed their vote splitting problems. Their winners went from having less than 33% support to 57%. Despite the FUD from the RCV folks, Approval Voting is easy to implement and effective.

Compare the complexity of this: https://www.youtube.com/watch?v=FeMg30rec58
with this: https://www.youtube.com/watch?v=_eXYsjnVhBk

13

@3, @7 -- Agreed. Simplicity is one of the key factors, and approval voting has that.

@6 -- You need to read the first example to see why this is a problem. Most of the city does not like our new attorney general. Most. Your solution (which you shared for a problem you believe doesn't exist) wouldn't help. Activist candidates (like NTK and Nelson) have no trouble getting enough people to sign petitions. The more extreme you are, the more likely you are to cross that hurdle. It is the boring candidate that struggles.

@9 -- It is quite likely that just about all of the people who supported NTK in the primary would be OK with Holmes. At the same time, there were people who were only OK with Holmes (for obvious reasons). That would have been enough to propel Holmes above NTK. Whether he faced off against NTK or Davison is irrelevant -- he would have easily won the run-off.

The endorsements get simpler as well. It is quite likely that The Stranger would have endorsed both Holmes and NKT in the primary. If they both advanced, then they would support NKT. Likewise, they would have endorsed both Oliver and Thomas. In both cases, it is quite likely that the moderate candidate would have advanced, and won the run-off.

17

@14 That 51.51% is Davidson vs. Thomas-Kennedy, not vs Holmes. This is the whole point. With choose-only-one ballots, you get vote splitting and often don't see the true overall support for candidates. It's entirely possible Holmes would have beaten Davidson head-to-head.

18

@9 "voters in the primary election opposed the status quo, so they eliminated the incumbent."

We don't know whether they really preferred Davidson to Holmes. It could have just been the classic "center squeeze" pathology. Vote splitting. The "spoiler" effect. Approval voting would fix this.

https://web.archive.org/web/20190219005032/https://sites.google.com/a/electology.org/www/utilitarian-majoritarian

19

@16 "The top two primaries are too easy for unqualified and joke candidates to both get on the ballot and to use the crowded field to squeak into the General Election.
Making candidates petition onto the ballot solves that problem."

Petitioning doesn't solve vote splitting. If you have more than two candidates, vote splitting can still occur, and that's where approval voting works wonders. There's no logical reason to restrict primary voters to only one choice.

20

@12 Approval voting passed by 64% landslide in Fargo and a 68% landslide in St Louis, and will almost certainly pass in Seattle. Everyone intuitively knows there's no logical reason to restrict primary voters to a single choice when a large number of candidates are running.

23

So, if I understand this right a voter can just "approve" every candidate if they wanted? For what purpose? To prevent vote splitting? Pick one. It's not hard. Voters shouldn't be able to say "I really like CPLOGO and I really don't want Harrell so I'm also going to approve Farrell and Echohawk in the hopes they'll pass Harrell."

That's not voting. It's hedging. I'd like to hear a compelling argument for why any voter should be able to get two votes per race per election, which is exactly what happens if someone approves the top two. At least in RCV it's still only one vote.

The whole thing sounds messy and inaccurate and ripe for even more extreme niche candidates. Not to mention the potential for voter nullification. What's to stop supporters from approving their top choice and GoodSpaceGuy? That would be some general election.

The voter turnout in Seattle is low. Maybe candidates should work on enticing more voters to support them.

25

Council President Lorena Gonzalez

28

Logan Bowers is no different than Republican candidates who lose the election and then look to change the system because they lost.

29

Obviously, the solution is a weighted approval voting. If there are 10 candidates for mayor, then every voter gets 10 points to allocate. The typical voter would vote 0 votes for the 6 unacceptable candidates, 6 points for their favorite, 2 points for their second favorite, and 1 point to the remaining acceptable candidates. Top 2 point getters advance to the general election.

In the city attorney race, NTK supporters could have voted 2 NTK and 1 for Holmes. Davison supporters also would vote 1 point for Holmes. Holmes supporters would probably have voted 3 points for Holmes and he would have advanced to the general.

30

@29 You're proposing cumulative voting which is strategically identical to the status quo.

https://www.rangevoting.org/rangeVcumulative.html

The system as proposed is the result of decades of social choice theory research. It is intentionally designed to satisfy mathematical properties like independence of irrelevant alternatives.

31

@23 There's no logical / mathematical reason to limit people to one vote. It's just a weird accident of history. Your example perfectly illustrates the necessity of approval voting. If someone votes for Harrell, the only way for me to have an equal but opposite vote that cancels that vote out (to essentially cast a vote against Harrell) is to be able to vote for all of Harrell's opponents—Farrell and Echohawk in your example.

https://www.equal.vote/equalvote

Under the status quo, if a policy platform is popular causing a lot of candidates to run on it, then it is paradoxically less likely to win. This is problematic to say the least.

32

@22 "Vote splitting is only a problem in the mind of losing campaigns."

Well no, it's just an objective mathematical reality, known in social choice theory as a failure of "Independence of your relevant alternatives".

https://en.m.wikipedia.org/wiki/Independence_of_irrelevant_alternatives

If voters prefer X to Y, then it makes no sense for the winner to switch from X to Y just because Z joins the race.

If a large number of candidates embrace popular policies, and two unique wingnuts embrace the extremes, they can easily advance simply due to the fact that they aren't splitting their votes with any other similar candidates.

Everyone intuitively knows that candidates should advance based on how popular they are, not based on how many other similar candidates run.

33

@28 The advantages of approval voting are real regardless of whether you like Logan Bowers. It did a fantastic job when it was used for the first time in St Louis this year, after passing by 68% landslide last November 2020.

https://electionscience.org/press-releases/st-louis-voters-use-new-approval-voting-system-in-march-primary-election/

One thing you'll notice is that no commenter can articulate a clear logical reason that we should limit people to one candidate. Why not two? Why not three? What makes one the ideal number? This is apparently just bias for the status quo.

35

@34 "The logical practical reason to allow people only 1 vote for mayor is that only 1 person can be mayor."

That's a non sequitur. How does the fact that there is one winner make it advantageous to limit people to only one candidate? Your logic is like if I said, the sky is blue therefore everybody should get free cupcakes. It doesn't make any sense.

Indeed this can be mathematically proven. If there are three candidates and you vote for candidate number one, the only way for me to have an equal but opposite vote against candidate one would be to vote for both candidate two and candidate three.

If we can't do that, then candidates get an advantage simply by having more opponents. Which literally means that the more popular a policy platform is, the less likely it is to win. Which is obviously completely unpathetical to the whole point of having elections.

37

@34

There are multiple candidates for 1 job. You can’t hire them all so pick one.

That often elects the wrong candidate. For instance, in the 2016 WA State Treasurer race, a 51.6% majority voted for a Democrat, yet the two finalists were both Republicans. Why? Because there were 3 Democrats running and only 2 Republicans running.

https://en.wikipedia.org/wiki/2016_Washington_elections#State_Treasurer

Election results should be about voters' support for the candidates, not about weird vote splitting anomalies caused by having too many candidates running on the most popular platform.

And how presumptuous of you to think that because I voted for a mayoral candidate you didn’t like you deserve multiple votes to counter my one.

It is not a presumption to say that all voters should have equal weight. Your vote should not count for more than mine just because more candidates support my political ideas. You know something is broken in our system when political opinions become less likely to succeed when they grow more popular.

39

@38 Trump would have lost his primary in 2016 under Approval Voting.

See this presentation with polls from the Wall Street Journal: https://www.youtube.com/watch?v=6KqjT21fbT8&t=672s

See also this video on how the current system leads to minority rule and polarization:
https://www.youtube.com/watch?v=s7tWHJfhiyo


Please wait...

Comments are closed.

Commenting on this item is available only to members of the site. You can sign in here or create an account here.


Add a comment
Preview

By posting this comment, you are agreeing to our Terms of Use.