Georeactor Blog

Zohran Mamdani's surprise win in the NYC mayoral primary is getting framed as a victory for RCV. But little changed about a system which was seen as troubled in 2021.

The voters understood this to be a Cuomo-Mamdani race, and 80% voted accordingly in the first round.

We don't know how many voters did any ranking yet! NYC has some of the most flexible elections for mail-in and affidavit ballots, so we won't see counts until July 1. An election system with these two components can be frustrating and difficult to read on Election Day. In 2021 it took until July 6 for the AP to call Adams the winner.

If being the winner of a plurality in the first round again means you're the mayor, to me, it means if you tell me RCV changed the election, you must mean voter and candidate behavior outside of the actual counting.

I know there is research about RCV increasing turnout or diversifying the pool of candidates - this is part of the reason why people wanted to replace the old voting system.
But if you look at modern politics, elections are in trouble. In 1991, Adam Przeworski wrote the cynical minimalist definition of democracy as "a system in which parties lose elections". This means that there is some market or reputational benefit to recognizing a loss. When the 2020 presidential election was so contentious, it launched many thinkpieces about "loser's consent" as a component of democracy, which I've heard as elections must convince the losers that they've lost.
RCV leaves room for confusion and doubt, most visibly in Alaska in 2022, where Donald Trump demanded "NO ranked choice!" among familiar grievances against dropboxes, and Sarah Palin vowed to end it.

"It's bizarre, it's convoluted, it's confusing and it results in voter suppression," Palin told the CPAC crowd. "It results in a lack of voter enthusiasm because it's so weird." https://www.motherjones.com/politics/2022/08/sarah-palin-is-taking-a-few-pages-from-trumps-election-lies-playbook/

In Alaska, RCV narrowly passed in 2020 and survived by 737 votes in the 2024 election. RCV lost in other state referenda, including a new ban in Missouri: https://ballotpedia.org/Results_for_ranked-choice_voting_(RCV)_and_electoral_system_ballot_measures,_2024

This is not a red state problem. According to a study of 'come from behind victories' (CFBV) systems which adopt RCV are at their greatest risk during the first two elections. New York narrowly avoided its own repeal in the 2021 race https://onlinelibrary.wiley.com/doi/full/10.1111/polp.12544

The number of exhausted ballots was large (more than 14% of the total) (New York Times, 2021). In this context, "the Adams campaign had suggested that it would amount to voter suppression if the candidate who had the most first-choice votes did not become mayor" (Weigel & Yuan, 2021). Commented the Executive Director of Common Cause New York: "If Adams had lost, I shudder to think what would have happened to RCV."

Older, minority voters are most likely to make errors and report difficulty on RCV ballots. Here's what that looked like in NYC:

The top three council districts with the most fatal overvote errors in any race were all located in the Bronx. Council District 14 had a fatal error rate of 3.8%, District 17 had a rate of 3.6%, and District 16 had a rate of 3.2%. https://www.nyccfb.info/pdf/2021-2022_VoterAnalysisReport.pdf

In neighborhoods where young people are coming and going, it may also remain difficult to establish familiarity and trust in RCV.

If anywhere near as many ballots get eliminated this year by overvoting or exhaustion before the final Cuomo-Mamdani round, that would be pretty nuts (suppose 80% ranked them first, and 10% are eliminated, leaving only 10% of the vote to actually use later rankings).

Deep blue City Council District 2, where Mamdani has a 37-point lead over Cuomo, got extra attention this cycle because Anthony Weiner and SNL fave Harvey Epstein were on the ballot. After bike advocate x congestion pricing opponent (yeah idk) Allie Ryan received fewer first-round votes than Weiner, her husband posted that she "got ZOHRAN'd" elaborating

you can go for the young underdogs THINKING your vote will go to second choices eventually, but in REALITY, unless it's a tight race, never does

This reminds me of the Alaska criticism where an RCV loss is often seen as weird and wrong. Theories emerge that someone else ought to have won in a proper election. In reality, Epstein has 39% of the first round vote and the closest competitor Batchu has 21%, meaning she would need to gain votes 2:1 in future RCV rounds (or >2:1, considering exhausted ballots). This is generally seen as unrealistic. I believe the Ryan math is either adding all Batchu and Gordillo votes to reach 50%, or supposing that they could win 40%-39% in a winner-take-all race. In 2021 and 2023 Ryan lost two-person races.

The election online

As early voting began, local subreddits were full of people who were uncomfortable or uncertain. Two common issues are:

• Strategy (ranking someone later and they get eliminated before the vote gets counted). This might be mathematically interesting, but as we saw in the mayoral race and District 2, it's not flipping many elections.
• Ranking last (falsely thinking that if you like candidate A and hate candidate B, put A first and B last). This is why the message "Don't Rank Cuomo" had to be emphasized.

Having Lander and Mamdani cross-endorse (and appear together on The Late Show) was probably effective, but more in a media/messaging sense. If this is interpreted as an RCV strategy, are we saying that a significant number of Lander-Cuomo or Lander-nobody voters were converted? Or that Lander voters were motivated to put Mamdani first and discredit Cuomo in the first round?

I could see an argument that Mamdani benefited either because (a) it seemed more likely he could win through some RCV math or (b) when he was a longshot candidate everyone was friendlier. There was some research in 2016 about RCV campaigns being more civil, or perceived as more civil https://www.sciencedirect.com/science/article/abs/pii/S0261379416000299

Cuomo originally planned to reappear as a third-party candidate in the November election, which would drag out this same Mamdani-Cuomo contest and make the primary a bit redundant. It now looks like Cuomo might be convinced by statistics or by funders to stay out? Mamdani received some high-profile endorsements and disses, but if there's any factors where he gets pushed aside in the November election, I think that will include undermining credibility of RCV.

Note

An anti-RCV site, ProtectMyBallot.com, hasn't been updated since 2021 but collected a variety of statements and arguments.