NYC Mayor/2021 Megathread

[Sorenroy]

It seems odd that all of these simulated ranked-ballot polls only show the last hypothetical face-off. Even the best polls have a margin of error of a few points: it would be nice to see how the simulated ballot might end up if, say, the last round is Adams-Wiley. Of the polls taken this month, Emerson showed a 41-30-29 race; Change showed a 37-33-30 race; Schoen Cooperman showed a 40-32-28; Public Opinion showed a 38-33-29 race; and Marist showed a 43-30-27 race. The common factor (besides Adams in first): Wiley was third place in all of them and never by more than a few percentage points. It doesn't seem like that much of a stretch to say that the polls might be a little off or that Wiley continues gaining enough steam to get that 1-4 percent.

[Sorenroy]

https://twitter.com/freedlander/status/1406352294023733257

Campaigning together is racist in the same way as Buttigieg, Klobuchar, et. all endorsing Biden after South Carolina was collusion: it's not. It's strategic, in both cases, but there is nothing wrong with trying to work together to make sure you or a candidate with your values is elected.

Actually, the language of Ashley Sharpton is super similar to some of those who were backing Sanders in 2020. Adams would have done the same thing as Garcia/Yang if he could do so, in the same way as Sanders would have loved a wave of left wing endorsements. It's not just a silly thing to say, it's hypocritical and dishonest and weakens legitimate political discussion.

[Sorenroy]

So I decided to look at the numbers for what Wiley/Garcia might need to overcome Adams in the "instant" runoff, and it's a very, very high bar. The New York Times and Politico seem to think there's another 150k+ ballots left to come in and, while their estimates of which boroughs are currently over/underrepresented does little to change the percentages of each candidate, it does expand Adams's lead by another 16,000 votes. All that considered, Wiley would need to net ~92,000 votes in reallocation to win, while Garcia would need ~118,000.

Depending on ballot exhaustion, that ends up being between 61-39 Wiley (hypothetical 0% exhaustion), up to 66-34 Wiley (with 35% exhaustion). In the case of Garcia, those numbers are 63-37 to 70-30. Wiley's path looks very steep in this regard: of the three runoff polls done where she was the final challenger to Adams, her highest percentage of transfers was 52%. However, Garcia might have a narrow shot. Of the seven polls conducted over the past month that had her as the final challenger, she's gotten anywhere from 47% (the only poll to see Adam's beat her in transfers) to 62% (in the most recent DfP poll). Her median has been 56%, which would leave her about 40,000 votes behind Adams even at 0% exhaustion.

[Sorenroy]

Who do you guys think will ultimately end up as the Dem nominee for Comptroller?

If outstanding ballots reflect the boroughs that the New York Times/Politico thinks they are missing from, Johnson will need to make up about 74,000 votes to win. Depending on ballot exhaustion, that means that, by the final round, Johnson will need to pick up between 59% (at a hypothetical 0% exhaustion) and 64% (if exhaustion is way up at 35%) of the transfer votes.

There's only been two polls of the race in the past month, and both have had undecideds around a quarter of the total, but the simulations of Data for Progress showed the candidates going just about even on reallocation, and they nearly nailed the percentages with Lander at 33% and Johnson at 26% in the first round (ignoring those massive undecided numbers). The other poll, out of Schoen Cooperman, showed Johnson getting 60% of the transfer votes, but they massively missed the first-round mark, where they had Johnson beating Lander 29 to 18.

All in all, it seems like a tall order for Johnson to pull ahead, but a lack of polling and outstanding ballots mean that there's still a chance. Probably about the same, if not slightly better, than Adams being beat out for the mayoralty.

[Sorenroy]

So, after all is said and done, how public will the ballots be? Will there be an enumeration of each set of rankings, or is that held private? Seems like it should be public information, but I can't find anything either way.

[Sorenroy]

I said before that it seems like it should be public information how many ballots of each ranking were, but thinking about that more, there's a reason for that that's more than academic. While some people might find the back-end mechanics of ranked choice voting confusing, releasing those numbers acts a bit like an audit in that it allows for third party's to check the work of the BoE. In a plurality-winner race, this is very easy: just look at the election returns and whichever candidate has the highest number is the winner. But in a ranked-choice race, those full ballots need to be released to make that calculation.

That's not to say that anyone should be calling fraud over the ballots being private, but it seems like a good government thing to do, just like paper trails for elections are good just in case. The only issue I can see is that it lowers the anonymity of voters slightly. If there was a hypothetical vote-buying scheme, or some type of machine politics counting votes, it's easier to identify people if you demand a very specific line of preferences. Any precinct is going to have a least a few dozen people who vote for a major candidate, but it's highly unlikely that they would vote for, lets say, two major candidates followed by a very specific non-entity.

Edit: When I was thinking about reasons for why releasing the full race tabulations would be a bad idea, the biggest concern I could come up with of was violating anonymity. I was coming from a stance that the results would be given precinct by precinct, just like elections in plurality system races. However, there really is no need for that if your main reason for releasing the tabulations is good government or audits on how the eliminations flowed. Just like a plurality race, the precinct by precinct breakdown does not actually matter for deciding who wins. There are still reasons for why having regional results would be a good idea - just like in a plurality system, doing so allows for people to see if their ballot was counted or for other organizations to check and make sure the elections were clean - but it's not required. That moves the issue of anonymity from a very granular level up to an election-wide level. Perhaps in some of the smallest races, like those in small towns of only a couple thousand people over a handful of precincts, you'd still be able to track the votes, but in larger elections, any specifically ranked ballot is that much more likely to blend in with all the other ballots cast. If someone really was trying to run an election scheme across multiple areas, the race-wide elections would mean that it would be impossible to track who followed the lead and who did not.