Episode 179 - New York attempts Ranked Choice Voting
New York City's democratic primary is in the midst of a vote-counting debacle that may go on for weeks. This is their first time using ranked choice voting, but criticism of RCV can be conflating with other problems in the electoral process. Max and Aaron break it down with inside knowledge of NYC politics, social choice theory, and potential solutions.
Links
New York Times
Chaos as 135K Test Ballots counted in Preliminary Results
Ranked Choice could give it to Garcia or Wiley over Adams
New York Post:
Affluent White People vote far left while Minorities vote more Moderate
Fat Overtime Pay for Election Staff
Daily News:
Katherine Garcia may Takeit from Eric Adams after Absentee Ballots are Counted
Wikipedia
Page contains preliminary results
(I used this in my calculations but they may change this as the real count comes in)
Related Episodes
Episode 92 about Ranked Choice Voting being Introduced in New York City
Transcript
Max Sklar: You're listening to The Local Maximum Episode 179.
Time to expand your perspective. Welcome to The Local Maximum. Now, here's your host, Max Sklar.
Max: Welcome everyone. You have reached another Local Maximum. Today, we've got Aaron on the show again. Aaron, how are you doing?
Aaron Bell: I'm doing well. Had a great Fourth of July. I hope everybody listening did as well.
Max: Happy Fourth of July to everybody listening in the United States.
Aaron: Or should I say, Independence Day.
Max: Yeah. Well, one of my international friends on Twitter was complaining that the Fourth of July is becoming an international holiday. Not everybody likes that. But...
Aaron: Yeah. Well, actually, this was also right around my seventh anniversary. We did our honeymoon in Ireland and we were there for the Fourth of July. I'm trying to remember which town it was. Maybe, it was Killkenny. We came through and there was a Fourth of July parade in an Irish town.
Max: Oh, wow. I didn't know that.
Aaron: We were not expecting that at all.
Max: Yeah, okay. Well, for those of you who feel excluded because we're talking about the United States, today, we're going to really exclude you because we're gonna go hyperlocal and talk about New York City politics, but—
Aaron: America's least favorite city.
Max: Oh, come on. Well…
Aaron: It's a city we all love to hate.
Max: I'm tired of defending it. But, why am I still talking about New York City? I'm gone. But no, we're talking about ranked-choice voting. So that has more broad implications in terms of voting systems, and social choice, and electorates in general, and democracy in general. So hopefully, there will be more general takeaways on this. This learning about New York is inherently interesting.
Aaron: This is one of the first… Regardless of how politically important New York City is at the national scale, this is the first truly large-scale test drive of ranked-choice voting that we've seen. Everywhere else, it's been smaller elections or smaller states, but this is a big deal. I think we can safely say at this point, results may be mixed.
Max: Right. Yeah, yeah. Okay. We'll talk about that in a second. Yeah. So I just got back last week from PorcFest, which is a liberty-oriented festival in northern New Hampshire. That was a lot of fun. I'll probably talk about that more on a solo show maybe next week. So a lot of interesting stuff that came out of that. A lot of former guests on The Local Maximum. But why don't we just get right into the New York City election today? So first of all, how many points did you win for betting against Yang on Metaculus for the prediction? So I remember six months ago, everyone thought Andrew Yang was going to be the next mayor of New York City.
Aaron: He was going strongly because he had all that name recognition coming out of the presidential election. And—
Max: Yeah.
Aaron: He was leading. That did not stay the case for long. So I did win points.
Max: I knew it.
Aaron: Completely befit, but I came away with a not so impressive 16 points because my final prediction was 45% odds for a Yang victory, which, given that they didn't lock that until, I guess, until Yang actually officially dropped out. I was really leaving some points on the table. I should have revised that down significantly in the last weeks leading up.
Max: Even I feel like I should have revised mine down. I had him at 20%. I got 141 points. It’s good. But I was just like… On Metaculus, they were saying 60%. I'm like, “No, I know how the Manhattan Democratic Party works. This is not going to happen, very unlikely.” But then I looked at the 60%. I was like, “Oh, maybe some people know more than me.” I don't know. But it turned out no, they—
Aaron: I haven't looked recently. Do you know if the predictive markets were in line with Metaculus here or we're seeing a split in the demographics of who participates there?
Max: They were until a few weeks before the election. Yeah, but I remember this last time. By last time, I mean in 2013, which is the last time de Blasio won in a crowded field. He basically started to run away with it only in the last couple weeks of the election when all the democrats start closing ranks. So that's how it works. It's not necessarily the person with name recognition going into it.
Aaron: Yeah, there's a reason for a—
Max: Especially for a local election.
Aaron:—in presidential politics. They talk about the October surprise because there's a lot that can happen in the last week or two leading up to the election.
Max: Okay. Yes. So the interesting thing about this election... Well, first of all, we don't have a winner yet, and it's been almost two weeks. So that's pretty crazy. So let's talk about that. This is the first ranked-choice vote among the Democratic candidates that took place on June 22nd, still no winner. Curtis Sliwa, won among the Republicans, founder of the Guardian Angels and talk radio host for many years. He won with 68% of the vote among Republicans. But again, that election had only 30,000 voters, whereas the Democratic election had 800,000 voters. 30... Maybe a little more actually. But it's still... It's under 100,000. It's like 10s of 1000s on the Republican side—
Aaron: It's on the high end being able to call it just an order of magnitude.
Max: Yeah, yeah. So how does ranked-choice voting work in New York City? I used to say, I still agree with it. They removed the worst voting system and replaced it with the second-worst with this ranked-choice voting.
Aaron: The worst being the traditional, first past the poll?
Max: Yeah, yeah. Then, of course, every election has some complications in it. This one is included in terms of who gets on the ballot and all that stuff. But the way ranked-choice works in New York City is there are 13 candidates, and you can rank them, and basically, whoever… It's your first-place choice that matters. So it does matter who gets your first-place slot because the person with the lowest number of first-place votes ends up getting eliminated. Then their votes get reallocated. So, basically, in the next rounds, whoever's eliminated if you still have someone who's uneliminated on your ballot, then whoever is highest up on your ballot gets counted as your vote in the next round.
So it's essentially first past the post, but you get to reallocate your vote if your person doesn't make it out on top. So, I don't know if I explained that correctly. Also, another aspect of the New York system is you could only rank five candidates, so you can't rank all 13. So you do want to make sure that on your ranking, some of the people on your ballot are actually people who make it to one of the last rounds so that you get a say in that. So I just wanna...
Aaron: There's still very much a concern about, “Oh, you're wasting your vote,” because you could conceivably fill out a ballot where you rank your five top choices, and none of them even get counted because they're so far below the threshold that they all get discarded.
Max: Right. I think that happened to a lot of people. The strategy for this makes a lot of sense if you are... Or is a lot simpler if you have, say, a two-candidate election with some very unpopular third party candidate, not unpopular, but not well known, who has an issue that you really care about. You can vote for that little-known candidate in the first slot. Then in the second slot, choose the candidate among the top two popular ones who you would prefer to win. So that's one thing that you can do if it's a simple two-party with some minor party's race, but in this case—
Aaron: The mechanic is still in play here. It's just that with so many people on the ballot, and with... Even if they hadn't capped it at, you can only rank five, who can reasonably research and rank, what do you say was, 13 different candidates that—
Max: Right.
Aaron: That it’s overloading the load-bearing capabilities of such a system.
Max: It's all the same party. So you really have to do research on each one. We'll get into a second on the demographics of who voted for what, according to the Times and the Post and all that. I didn't go into all the numbers myself. So another point is generally thought the winner of this Democratic Party wins the election in New York. For 20 years, New York voted for non-Democrats: Giuliani, Republican, and Bloomberg was a Republican and then became an independent. The general sense is that those days are over, it's all Democrats from here on out. There is at least some diversity of thought in the Democratic primary, probably not enough, but it makes me think, “Well, maybe they should just have this be one nonpartisan race if you're going to do it like that.” But I don't know if that's going to happen.
Aaron: Yeah, there are definitely a lot of cities across the country that do have nonpartisan races for a mayor, or select people, or whatever their particular system of government is but not so much for the large cities. Your New York City being, I think, the largest city in the United States.
Max: Right.
Aaron: Cities like that, or Boston or Los Angeles. You're not going to see a nonpartisan race anytime soon there without a major restructuring of the system.
Max: Yeah, also the Democratic Party in New York City, they get to decide who gets the jobs and the new administration and all that. So they want to be in control of the process, even if everybody in New York City joins the Democratic Party and votes in this, they would prefer that they are the gatekeepers.
Aaron: Yeah.
Max: Which is also—
Aaron: I suppose the city is the recipient of federal funds on any significant scale. There is a game to be played with being partisan is, I feel, too low to determine, but being politically aligned rather than being a nonpartisan officeholder.
Max: I thought Yang wasn't going to win because he didn't put in his time in the Democratic Party in New York City. He wouldn't play by their rules necessarily.
Aaron: Yeah, I was about to say you could say the same thing about Bloomberg because—
Max: Well, he didn't run as a Democrat.
Aaron: Right. That’s why I walked it back a little bit, but he definitely didn't pay his dues in either political party before he rose to the mayorship.
Max: No, he just had billions and billions of dollars. So that allowed him to do it. There's so many things to cover. I'm interested in the new election system, I want to get into the pros and cons. First, before we get into that, let's talk about the absolute crap show that is the counting of these ballots. So it looks like Eric Adams won the election on election night. Should I even go through these candidates?
So I'm not going to go through all 13 candidates, maybe I can go through the four candidates, the four or five candidates that made it out on top and what the… Just for those people who are not in the city. So there's Eric Adams, African American, he's generally support former police officer. Some people say he's a Democrat in name only. He's supported by minority communities, entrepreneurs, he's considered more pro-business, more pro-law enforcement. Hiort comes from the outer boroughs. So that's Manhattan—
Aaron: It’s definitely an interesting cross-section of the Democratic base in the current era to be pro-law enforcement and have significant minority support. But, maybe, that's just something about how New York is different than some other parts of the country.
Max: No, I mean, some people are saying, basically, it's the white socialist true believers that are different. Then you had Kathryn Garcia, also still in the running. I call her the New York Times candidate because she was endorsed by the New York Times. They say, wealthy upper-class Manhattanites.
Aaron: Sounds very establishment.
Max: Yeah. So Maya Wiley, she's the most closely associated with de Blasio, supported by socialists, AOC. They said, far-left hipsters who are now all of a sudden, far left. I just thought they wore funny clothes. I didn't know. Okay, but uh—
Aaron: They have too much time in Brooklyn.
Max: Yeah, so her voters tend to go to Garcia in the second round. We'll talk about that in a second. Mainly because her voters don't want Eric Adams. Then finally, the fourth one is Andrew Yang, who's... You guys probably know him from the presidential race. He's also less doctrinaire progressive, but still firmly Democrat. He actually asked his voters to go to Garcia in second place, as well. So it seems like everyone's ganging up on Adams here.
Aaron: No. Yang, I know, in the presidential election, his big talking point was UBI. Was that part of his platform for New York City as well? Or is it more a pro-business?
Max: No.
Aaron: But also, some progressive policies platform.
Max: Right. So that's the thing. He was saying some of the same things that Eric Adams was saying. He was pro-business, pro-law enforcement, but he was telling his voters to go to Garcia. If you look at the numbers, and I'll get into that in a minute. Did they listen to him? A slight majority listened to him and put Garcia down as second, but just almost as many put Eric Adams as second.
Aaron: How does it work? It's a majority but not a plurality?
Max: Yeah. So some voters were willing to put him in second place but they weren't willing to listen to him or willing to put him in first place but weren't willing to listen to him in terms of what's the rank afterwards. Or it could have been voters that don't particularly liked him, but ranked him third or fourth and then had one of the other candidates further down the list. Then some of the minor candidates, I guess, Scott Stringer had maybe 5%. He's very similar to Garcia, another Manhattan guy. Okay, so I just point that because I think I met him once. Alright.
So anyway, Eric Adams looked like he won on election night. Now, we're not sure. They keep releasing unofficial counts, but they're waiting for official counts because they've got to open up all of the mail-in ballots, they're like— Well, mail-in ballots take 10 days. We want to give it 10 days to get in. So it's been 10 days and they closed it. Now, they're still trying to count it. They're like, “It's gonna take us another couple weeks to count all these.” So that sounds a little bit crazy.
Maybe it doesn't have to be like that. Then secondly, they released some preliminary results. Then like Adam said, “Well, where's all these extra votes coming from? This doesn't make sense.” Then some people in the media were calling it, saying, “Well, that's Trumpian. He's questioning the outcome of the election.” Then the election board said, “Whoops are bad, we actually added 135,000 votes. These are just test ballots.” I don't know what they were testing. Were they doing a test count beforehand? We don't really know what that means. That's what I assume it means.
Aaron: The Steelman assessment: that's the most charitable possible determination.
Max: Right.
Aaron: It's easy to see why there might be some parallels to Trump here because you've got leading in the counts at the close of election night. Then it seems to have been reversed based on mail-in ballots and additional counting, much like we saw in November of last year. Although, I think labeling him as Trumpian may be… But perhaps only because it's clearly a pejorative there. It’s a little unfair, given that he was correct to call out, at least in that specific case, there was something fishy in the numbers.
Max: Oh, they rushed judgment on that. So now a lot of people are spinning this in a lot of different ways. I read the comments on The New York Times. Oh, people are getting all bothered by this, The New York Times readers. So some of the spin is that, “Well, this shows that our election system works because the error ended up getting corrected. So that means that, of course, questioning any other election on the national scale is crazy.”
Aaron: It does raise some alarms, though, about what kind of self-checks and what policies do they have in place to... If Eric Adam’s campaign hadn't raised an alarm about this, would it have gone on unchecked, uncorrected?
Max: Yeah, maybe? Maybe not? I don't know. There at least should be some investigation into that. Why do we assume that the system is self-correcting always?
Aaron: This on a long list of things.
Max: Yeah.
Aaron: The long time to resolution here in general in that specific case. You want to make sure that at least you don't fail the same way twice.
Max: Yeah. Also, if I'm testing software and I'm finding a lot of bugs, I wouldn't be like, “Well, look, I'm finding bugs. That means there's no bugs that I didn't find.” That would raise some red flags to me. Now, some people are blaming ranked-choice voting for this and it's being conflated. I think that's wrong. But that is certainly a point of view here because it does seem a little bit more complicated. Another group on The New York Times comment board is like, “Well, basically, their whole thing is, this is really bad that this is happening because now it gives Trump... Ah! Trump more ammo to question elections.”
Aaron: It’s certainly not good optics.
Max: No, no. But that's the one thing they're concerned about. Their main complaint is with ranked-choice voting. There's some people who are defending ranked-choice voting, but they think that all of the critics of what's going on in this election are critical of ranked-choice voting and not critical of the other problems with the election boards. They conflate it on the other side if that makes sense. They're right because some of the critics are overly focusing on ranked-choice voting, but then the people who are... Then the people are like, “Well, why are you complaining? Ranked-choice voting is fine.” When it's like, “Well, there are other things to complain about.”
Aaron: Yeah. From a very superficial point of view, they are correct in that this is the fault of ranked-choice voting because if they kept the old system, they know how to do it, we would certainly have a result by now. But I think that’s un—
Max: Well, not necessarily. They could have still added test ballots, and they still had to wait for the mail-in ballots to come in.
Aaron: I suppose that's true. But I think extensively, it would have been a non-news story. It would have just been another New York City election.
Max: Right.
Aaron: But that's not really a fair yardstick to measure this against, given that this is the new system. Saying that it's a bad system because we didn't do a good job at it is not a good argument to convince me we should go back to the old one.
Max: Right. I did make a comment on Twitter that I want to discuss, which is, I said, “There's a strong case to be made for ranked-choice voting, at least over plurality voting. But if you don't have timely, accurate, and trusted results; timely meaning, you don't take weeks and weeks to get a response; accurate, meaning, you're actually getting the real numbers of what people voted for, which is important; and trusted, meaning that people can verify it, people largely trust what you're doing, none of it matters. I prefer to have a plurality system with those three things versus a ranked-choice or any other system without those three things. So, if they don't get this right, the voting system, to me, doesn't matter nearly as much because it's not a sound democracy if you don't have trustable elections.
Aaron: Yeah, and those of you who are in the locals, in The Local Maximum, you would have seen that there was some discussion of that particular tweet earlier. Was that over the weekend?
Max: Yeah. Oh, did I say? I said tanked choice voting. I didn’t say ranked-choice voting. That was an accident.
Aaron: I thought that might have been a bit of a Freudian slip. Although, well, you mentioned the… What is it? The board of elections and their role in here. I don't think you're implying that they're in the tank for a particular candidate, but they certainly have some influence here.
Max: The Post and The Times are writing about the board elections, they seem to be in agreement that there's something really wrong there where... I've been reading, and I don't understand this as much as I should on this. But I've been reading, there's like, there's nepotism, and they're collecting all this overtime. I don't know, it just sounds like a big problem. So first, I'll defend ranked-choice, maybe a little bit before we get into the mathematics of what is happening here.
So Tucker Carlson, one of his takeaways is, he hates ranked-choice voting. The way he put it is, it's because the losers get to pick the winners. I think that's a direct quote. Actually, some of the people in The New York Times comment board have been saying the same thing. Well, it's not fair because the losers get to pick the winners. I just think that it’s such a bad take. That's not really what's going on here. So if your candidate gets ousted, you're not the loser. You still have a valid ballot. Even if it was plurality, you probably would have picked a different first choice if it was plurality.
Aaron: It's not a perfect metaphor or perfect analogy, but I think of it like you're on the playground in middle school when you're picking teams for softball or whatever.
Max: Okay.
Aaron: Maybe there's a particular player who would be my first choice, but they get picked by the other team. That doesn't mean that I just don't get to pick. I get to make my next best pick from the available candidates.
Max: Right. Right.
Aaron: That's not the losers getting to pick the winners. It's agreed that the metaphor doesn't doesn't hold up 100%. But there’s something to be said for that.
Max: It’s like if you go to the store, and you have to buy milk, and you say, “Well, if you can't get 2%, just get low fat. If you can't get low fat, just get a whole milk.” People can do... I feel like any competent person said—
Aaron: No, if there's not 2%, then you get no milk. Drink water, and you'll be happy.
Max: Yeah, okay. No, but a valid ballot is a valid ballot. Ranked-choice voting, it doesn't mean that losers get to pick the winners, just because your first choice is somebody who's out of the mainstream.
Aaron: There's, structural is maybe not the right term, but there are inherent downsides to ranked-choice voting, but I think what Carlson and you're saying some of the people in The New York Times comment boards have pointed out, that's not really one of them. That is actually more of a positive to the system. They're just a little salty about it for some reason.
Max: Right. Well, he gets into something where… So there's a lot of racial politics here, where some people are like, “Oh, ranked-choice voting is designed to suppress minority votes,” or some people are like, “Oh, ranked-choice voting is great.” Then there are other people who are saying, “Well, ranked-choice voting is great because it's going to give a stronger voice to minorities.” It's just, none of that is true.
Aaron: Like you alluded to in our earlier discussion, in a less crowded field scenario, and when it's not necessarily in a party primary, it can benefit minority parties. But we're not talking about—
Max: We’re not talking about… Yeah.
Aaron: —racial minorities there, which I think is the context of minorities that they're generally referring to.
Max: So then there's the argument that the average person doesn't know how to rank or doesn't understand what ranking is, which is also... There is a certain level of competence you should have to vote. People do understand ranking. I think the problem is that when you have 13 candidates, you shouldn't be expected to do research on each and every one of them. But I do think that the average person understands what ranking is. That's another thing. “Oh, it's so complicated. How do I know... Is the second place, do I want them more than my third place? I don't know.” It's just that's a lot of the arguments I get against it.
Aaron: Well, it's easy to draw that naive conclusion looking at some of the data, too. I guess, you're gonna get into talking about the actual numbers a little bit later. But there are a surprising number of ballots that we see that have no subsequent choice. I don't think that's necessarily because people are filling out ballots and only picking one or two or literally submitting a completely empty ballot. I think it has more to do with, at that point in the winnowing down there, subsequent selections have already been eliminated.
Max: Right, right. Then Carlson is gonna say, “Well, this helps the far left, but de Blasio won in 2013. He might not have won if it were ranked-choice voting. He got just, I think, under 40% of the vote, but everyone who preferred him was probably already on board at that point. His next best rival, Bill Thompson, maybe could have made a run for it if they had ranked-choice voting at the time. But I do want to point out: ranked-choice voting can't fix certain things. So if your entire electorate is made of complete idiots, I'm not saying that that's the case in New York, but it might be. It probably is. It doesn't fix that. It can't fix a corrupted and incompetent board of elections, which again, I'm not saying is definitely the case, but it looks like it might be. So there's a problem with that.
I guess, my last point is, I would prefer a Condorcet method which... We'll get into the practicality of doing that where basically, the winner is the candidate that beats everyone else head-to-head. So there's no elimination round because the elimination rounds does force you to think, “Okay, well, who's going to be in the last round. I want to make sure that one of the candidates that's in the last round, I have ranked on my ballot because, otherwise, I'm not going to have a say in that last round.” That makes RCV really... Then you have to be more of a political pundit. I don't think...
Aaron: Does that risk disappear if we shrink the field, and we no longer have the 5 out of 13 scenario? If it was, rank your top five, and there are only five in the race? Does that become functionally the same?
Max: Yes.
Aaron: Or is there more nuance there?
Max: So there's more nuance there. You're right, the problem that I just stated would not occur if there's five people in the race, and everybody's ranking all five. But you'd still have a problem where let's suppose… I'll give you an example. Let's suppose you wanted Yang, right? Let's suppose you didn't really want Yang. You wrote him low, maybe, you put them in second place.
Aaron: Okay.
Max: Okay. Essentially, he gets eliminated early, okay? Your vote didn't help him. So you could have voted, you could have put him at first place, and he wouldn't have gotten eliminated. But because he got eliminated, it could be that somebody worse came up, instead of him. I don't know why I picked Yang. I think I did. No, I did because of the numbers below, which we're going to get to because I think this actually happens to people. Whereas somebody worse than Yang could come up because you contributed to him getting eliminated by not putting him in the first round. Whereas if you count all the head-to-heads at once, then you don't have that problem anymore.
Aaron: Now, what do you mean when you say count all the head-to-heads just to be clear?
Max: Okay, so when you count all the heads-to-heads, everybody has their ranking, right?
Aaron: Yep. So I’ve got my one through five.
Max: Great. So then you make a five-by-five box and then just say, “Okay, let's look at two-party races of any two candidates against each other.” To get the count for that two-party race, you say, “Okay, people who voted Candidate A over Candidate B in their ranking, voted for A. Anyone who voted B over A voted for B.” Then you look at that, and then you'll get an answer as to who won in that head-to-head race between A and B. You can do that with any pair of candidates. Typically, what happens in Condorcet is that one candidate beats every other. That's your answer. I know every once in a while—
Aaron: No… The voting systems on a previous episode, is this the one where you can get into a cycle where it loops? There's not a clear winner?
Max: Yeah, so that would be pretty rare.
Aaron: Especially at the scale we're talking here, it seems unlikely.
Max: Yeah. But again, it's the best of the lot in terms of preventing these types of problems. Now, you have to have it—
Aaron: What is the downside… Since we did mention before that there are downsides to all these systems. Is the downside there that it's difficult to implement or is this more, it's difficult to sell to the public?
Max: Probably both. I don't think it's difficult to implement. But given the current board of elections in New York City, who can't figure out first-grade math, it might be difficult to implement. Yes, it's difficult to sell to the electorate. You have to have an electorate that's at least slightly more sophisticated. Finally, you do have to have some elimination round beforehand, where you want a field of candidates that's small enough that everybody could do research in each one.
But you want as much representation among the candidates as possible. So you don't want them to have all to be clones of each other. So, maybe, some proportional representation could happen for that first elimination round. Some primary, where maybe you vote for one person, and then the top vote-getters go on or something like that. I haven't really thought enough about that. But there has to be an open... There has to be some pre-selection where you're guaranteed to get a broad, diverse set of candidates but small enough for people, for the average person, to be able to research and rank.
Aaron: That sounds like a new level of hell's nightmare scenario, and that we need to have a primary to select the candidates to run in the primary.
Max: Yeah. Well, in this case, it wouldn't be the primary, it would be the final result. Now there are things you have to do to get on the primary ballot. So maybe if they make that a little tougher, it would be lower than 13. But there's some ridiculous candidates running.
Aaron: Without diving too much into your politics, is it simply a matter of collect X number of signatures? Or are there a lot more hoops to jump through in that process?
Max: I actually am not entirely sure. But there are certainly hoops to jump through. It could be pretty expensive as well. So here's another problem with what they're doing. They're actually doing multiple counts. So what they're doing is, they're counting all of the first-choice ballots, then they're eliminating the bottom one. After the first round, that's like two votes for whoever got the bottom one. Then they're counting the whole thing again with those two ballots reallocated. So, instead of doing that, they should actually just publish all of the ballot configurations.
So this is how I see it would work. So what is a ballot configuration? So if I rank five candidates, let's say I rank ABCDE. That's my configuration. It's ABCDE. Maybe, there are 30,000 people that had that exact same configuration. Your configuration might be BAEDC. So that's a different configuration. Now, could there be too many configurations? If everybody ranks 13 candidates and you had 13 slots, that's 13 factorial configurations. I calculate that's 6.2 billion. Theoretically, every ballot could be unique.
I think in reality, there's only 800,000 voters. So theoretically, all 800,000 ballots could be unique. I don't think that's the case. But if you want to rank all 13 candidates, that would be implausible. Now, if you want to rank 5 candidates, it's actually less crazy because it would be 13, choose 5, if everybody chose five. Then some people aren't gonna choose 5, some people are gonna choose 4, 3, 2, and 1. So that's a mathematical formula for that, right? How do you take 13 and you choose five, although it's actually you're not choosing five. It's permutations, not combinations, right? It's not just 13, choose 5, it's—
Aaron: Because orders matter.
Max: Right. Order matters. So basically 13 for your first choice, times 12 for your second choice, times 11 for your third choice, times 10 for your fourth choice, etc., etc., and times 9 for your fifth choice. So essentially, you would have only a set number of configurations possible. Basically, if you limit it enough, I imagine that… Again, certain configurations won't be selected. If you have a case where... Okay, there are a few 100 different ballot configurations that we see, you could basically put it all in a spreadsheet and type in the numbers for each one. Then everybody could figure out what the rankings ought to be based on those answers. Did I explain that right? Or maybe it was a little bit off.
Aaron: I think so, yeah. Then the only thing we have to argue about is whether they entered all that data into the spreadsheet, if they counted them correctly, not if they did the math on the analysis correctly because the raw data would be available.
Max: Once the raw data, once the raw information is available, then doing the whole system of rounds or if you do Condorcet every one-to-one matchup, is just a matter of math on the spreadsheet. It should be pretty easy.
Aaron: Right. The other thing that I think I neglected to ask you about, but we talked about a little bit before the show, if, let's just say that miracle of miracles, the board of electors or some ballot initiative next year changes New York City to a Condorcet system, it wouldn't actually change what the ballot looks like to the voter in any meaningful way, they pretty much go through the same process. It's just that how the math is done on the back end would be different. As you've asserted here, it would come up with a fair result, one that satisfies more voters.
Max: Right. But also more than that, rather than—
Aaron: It’s not like you have additional strategizing to understand. How does my vote work in Condorcet that you didn't have to do for the initial assessment with ranked-choice voting?
Max: Yeah, so it's my understanding with the ranked-choice voting, they're not actually giving us the ballot configurations. They're just giving us the first-place counts on each subsequent round. So we don't really know how the ballots are configured, which makes the election process a lot more opaque, less transparent, which I think is a problem.
Aaron: Maybe at some future date, some Ph.D. student will have the opportunity to go through all these ballot records and collect that information and then rerun it and say, “Okay, if given the same voting, what would the Condorcet result have been?” That would make for a very interesting bit of investigation, but we don't have the data to do that currently. We don't expect they're going to be releasing it in that kind of a format anytime. At least not—
Max: Once people are satisfied with them for this, I think they're going to be like, “Oh, we're done. Yeah, I'm not gonna do any more overtime”.
Aaron: I'm morbidly curious if this is a system where once the election is done, they have to preserve these ballots in a locked and sealed container for X number of years. Or if once the election is decided, they ceremonially burn them so that nobody can go back and relitigate the process.
Max: Yeah, exactly. So I do think that needs to be more transparent in that regard. I think that there should be a way to make sure that for someone to check that their vote actually got counted correctly. So let's get into that a little bit. I wrote some lessons here. Well, before we get into some lessons, let's actually talk about what happened in these elimination rounds. Because we're talking a lot in theory about ranked-choice voting. Let's actually see what happens in practice.
So once you eliminated all the minor candidates, you had four candidates left. You had Adams, Garcia, Wiley, and Yang. Among those four, the first one that got eliminated was Andrew Yang. At that point, he had 111,000 votes in his favor. So those got eliminated. 30,000 of those 111,000 did not rank any of the other three candidates left. So those got thrown out. So that's actually something to consider for those Yang voters, that they did not include any of the candidates who are higher in the polls.
Aaron: That doesn't mean they just left it blank, just, it may have. But it seemed more likely that their subsequent choices were somebody who had already been eliminated.
Max: Right. Again, we don't know this because we don't know the valid configurations. By the way, these numbers include the test data that is false, so we might have to change them later. But see, he asked his voters to go to Garcia. Of the 111,000 voters, only 36, 37,000 went to Garcia. Now it's the most but only 37,000 of 111,000. So that's not that much. Then 31,000 went to Adams. So yes, his voters helped Garcia a tiny bit, but not much. Then only 12,000 went to Wiley. So I guess they don't want socialism. Or maybe a little bit with Garcia. So that's what happened. Then the next round, Maya Wiley gets eliminated now. The next round, you had Adams, Wiley, and Garcia. At that point, Wiley and Garcia are almost neck and neck. So this is one of the things that could change on the recount. That's another problem with ranked-choice voting, as opposed to Condorcet, is that it's not very robust. A few votes and a little bit of different configuration can cause the candidates to get eliminated in different orders and then change the entire outcome of the election.
Aaron: Yeah, it cascades down a very different branching path.
Max: Exactly. So this is a problem. So Maya Wiley, actually, at that point, when you have three candidates left, she has 227,000 votes. Of those, 78,000, a little over a third, get thrown out. They did not rank between Garcia and Adams. Then hers were very one-sided between the two remaining candidates. 117,000 went to Garcia, and only 31,000 went to Adams, 117 to 31. So her vote tracks heavily to Garcia. Maybe if Garcia gets eliminated first, her second places don't go so heavily to Wiley. So that could help Adams if Garcia gets eliminated first. But that's just speculation on my part. But I think that's probably right.
Aaron: At this point, all the votes are in. They're not expecting or accepting any additional mail-in votes or anything like that. They've presumably counted them all, but maybe not. So we're still waiting on official final numbers. Never mind what the breakdown of rounds actually is.
Max: Yeah, so the mail-in ballots are closed. They came in, but they're apparently going to take a few weeks to count them for some reason.
Aaron: I was just looking it up while we were talking. It appears I'm not looking at an actual ballot, but a guide to how to fill out the ballot. It looks like it's a pretty simple Scantron. You've got the candidates listed down the left-hand side, and then you've got across the top, ranking first, second, third, fourth, fifth. Bubble in the appropriate ranking next to the candidate of your choice, that should be pretty fast. If they can run the SATs through that Scantron machine in a weekend for that many hundreds of thousands of high school students, you'd think they could do this.
Max: That's a good point. It seems like all of this should happen automatically. Why is it taking several weeks to—the balance doesn't make sense.
Aaron: That's it, I fully expect that there are people who filled this out incorrectly. They filled out more than one person for their first choice or things like that. So there's probably plenty of those that need some manual inspection for “Is this valid. If so, how do we count it?” We haven't done that.
Max: Same with the SATs.
Aaron: Because it's not a scenario where you have, “Oh, we've got the Republican and Democrat poll workers looking over each other's shoulders”. It's all within the party. So there's not necessarily that same controversy.
Max: I've read there is a little bit of that. There's poll watchers being like, “No, they marked this one, and then they mark this one, and there's a stray line here, and all that”. Some lessons, some takeaways for me from actually putting this voting system into practice. First, it's probably best to whittle down the field a little bit before doing ranked choice. I don't think it should just be collect more signatures because then you just end up with a bunch of insiders who know how to do that.
You don't end up with interesting candidates. So there should be some ways to ensure that all sectors are represented in the field, maybe through some proportional scheme. But then again it’s like you have a primary and an election. But yeah, maybe you could have, maybe if ranked-choice voting or Condorcet voting were the last rounds like the not the primary, but the main election, maybe you could have a primary that's some proportional representation.
Aaron: If this is coupled with a move to a nonpartisan race, so that the primary is not within the party that it's a city-wide primary that then goes to the top X number of candidates across the board, that could help mediate some of that need for having multiple rounds. Not rounds in the sense that we were talking with ranked-choice voting, but you're having multiple election days leading up to the final election.
Max: Yeah. So I don't know, I picked maybe limited to seven candidates. I don't know, I put seven. But I feel like a lot of the multiparty democracies have like seven big parties, seven reasonable parties. I feel like—
Aaron: It seems like a lot, but we definitely have a bias towards having had two meaningful parties for pretty much our whole life. So I think we could probably adjust.
Max: Now I think seven choices is doable in general, but that's based on human nature. That's a question that we'll have to figure out. I feel like 13 is probably too much. Three, you could probably do a little more. ‘Cause you don't want the number of ballot configurations to get unwieldy because that makes the process a lot less auditable. Then, there's too much research that you need to complete the ranking.
I think that the mail-in vote counting is a problem. I don't know why it's taking so long. It's also crazy that they need to wait 10 days for the vote, and then take another couple weeks to count it. So something's got to be done about that. Then finally, there should be some receipts, maybe this can be done cryptographically, where it's like, they don't keep how each individual voted, but they keep how each, they have a code number for each person, and only you know your code number, so you could actually look up your vote, something like that, to ensure that votes are counted correctly and that individuals can check because I feel like there's just a lot of questions about our voting system that it's not good enough to say, like, don't question them. They're the experts, and they're doing a good job. If we are smart enough, we should be able to come up with a system where people can verify that this is correct.
Aaron: Yeah. I don't like the idea of being able to change your vote after the fact. Especially—
Max: No, no, no. Not change it, but—
Aaron: I'm thinking of, what was it that they talked about curing ballots in the last election? Primarily, if you have a way of tracking to your individual vote, then, if yours is one of those that comes up with you filled it out wrong, instead of putting two candidates as your first and second choice, you accidentally mark them both as your first choice, if the system can't properly process it, they could reach out to you and see what was your actual intention? Can you clear this up for us? I don't feel great about that in the context of there will certainly be preliminary results, the first night. So now, I can update how I actually want my vote to be based on who's leading. There's probably some things you can do to mitigate that, but it raises my hackles a little bit being able to do that after the fact.
Max: No, I'd be against that.
Aaron: It raises a little bit of a question. Obviously, it's better to know that they counted your vote correctly than to not know. But what do you do then? If it turns out they counted your vote incorrectly, if you go up, and you look up cryptographically. It wasn't just that I filled out the form wrong. It's that something went wrong in the system, and you just counted it wrong. How do you rectify that? Maybe there's some paths there. But I almost get the impression that it's just going to tell you that they screwed up but not give you a path to fix it.
Max: I don't want to get into it. But actually there was an election in New York back in ’06, ’08, I think. I tested their system a little bit, and I know they counted my vote incorrectly. It wasn't a vote that mattered, but that's definitely happened to me. You just gotta believe me on this. I proved that they screwed up the counting. So that is interesting. But yeah, you're right, there has to be some system where if the votes are not there correctly, it has to be investigated and and corrected for the next time.
Aaron: Yeah, I mean, this is what people have been talking about for a number of elections with concerns about going to electronic voting. We need a paper trail and you should get some sort of receipt when you place your vote. That's not going to be solved certainly on a nationwide basis anytime soon, given the distributed nature of our elections. There are a lot of things that could stand to be improved and fixed in how we do elections across the country.
Max: Yeah, I just don't understand the attitude of, “Trust us, there are no bugs”. I mean, that's always my famous last words.
Aaron: You’ve worked on software. Of course, there are never any bugs. Everything works right the first time.
Max: Yep. Yep. For sure.
Aaron: You just got to work a little longer.
Max: It's not a bug. It's a feature. Any further takeaways on this? I think I got my thing out or any more questions. We’re good to go.
Aaron: So I think I know the answer to this. But I'm gonna ask it anyway to clarify. Well, I guess it wouldn't matter because come the actual election, not the primary in the fall, there are only two candidates. So ranked-choice voting would have no relevance there.
Max: Right and there is no ranked-choice voting for the fall. It's back to plurality, again. Sliva versus... could be Eric Adams, could be Garcia.
Aaron: Has New York City ever had a viable third-party candidate for mayor?
Max: Yes, yes. Actually—
Aaron: Did Bloomberg run as an independent at one point?
Max: Yes, but he was cross-endorsed by the Republicans.
Aaron: Okay. So that really doesn't count.
Max: There was I think in the 60s, there was a... Let's look this up because there was a candidate from the Conservative Party of New York that actually did pretty well. Let me see. Mayoral 19— I don't know.
Aaron: Was that a New York State-only party? Kind of like how, is it... Is it Minnesota that has the democrat farmers party?
Max: Yeah. William F. Buckley ran as a conservative. Now, he was not gonna win.
Aaron: The William F. Buckley of, was that the National Review?
Max: Yes, he ran for mayor in 1965 under the Conservative Party and got 13% of the vote. So that's actually quite a lot.
Aaron: For a third-party candidate, that's an unusually strong showing.
Max: Then in 1969, actually a third party candidate John Lindsay of the Liberal Party won versus a Democrat and a Republican. So yes, there was something called the Liberal Party that I believe in like the 90s cross-endorsed Giuliani, actually. So there's some moderate party at some point. Although these little parties in New York, they have a name, but they're always captured by the small group of people who decide to do whatever they want, right?
Aaron: You can’t assume that a party from 50 years ago actually has any of the same positions of a party with the same name today, necessarily.
Max: No. Well, it also might not be liberal. There's always the Workers Party. The, what was it? The Working Families Party? WFP. Then Cuomo created the Women's Equality Party, WEP.
Aaron: I'm sure there was no confusion at all there.
Max: No, no, people speculated that he did that on purpose to confuse people who wanted to vote for WFP into voting for WEP instead. So that's how screwed up some of this stuff is.
Aaron: You got to get a name change before you get on the ballot to leverage somebody else's name recognition.
Max: Yeah. Because it's literally one line of difference. It's insane. But that's the kind of shenanigans that go on with ballots and all that. So another way of your perfect mathematical model of how an election should work doesn't always pan out in the real world. But all right, I think this is an interesting test case, and I'm sure it will come up later. I'm sure we'll talk about it when the actual candidate gets selected. I know we also want to talk about the billionaires going to space. You want to go into hyper-local to like Interstellar or whatever interplanetary, but maybe we'll do that next time. So yeah, I think that's it for today. I think that's all we can handle. I think we're good. Anything less. Are we good to go?
Aaron: Yeah, we’ll be keeping an eye on these things as they develop. You heard it here first. We'll have more in the future.
Max: All right, have a great week, everyone.
That's the show. To support The Local Maximum, sign up for exclusive content and our online community at www.maximum.locals.com. The Local Maximum is available wherever podcasts are found. If you want to keep up, remember to subscribe on your podcast app. Also, check out the website with show notes and additional materials at www.localmaxradio.com. If you want to contact me, the host, send an email to localmaxradio@gmail.com. Have a great week.
