The Civic Pulse — Issue #6

If you voted in last year's NYC mayoral primary, you ranked.

You walked into the booth, picked up a ballot, and got asked to list your top five candidates in order. You put Zohran Mamdani at #1 (or Andrew Cuomo, or Brad Lander, or Adrienne Adams, or the guy you'd never heard of who was running on a horse-and-buggy platform).

Then you went home, and a few days later you found out Zohran Mamdani won — even though no one got a majority on the first count.

How? That's the part most explainers never actually explain. Let's fix that.

This is a Civic Pulse explainer — part of our Government, explained. No spin. education series.

Here's the full breakdown →

Ranked choice voting (RCV, sometimes called instant runoff voting) lets you rank candidates instead of picking just one. If nobody gets more than 50% of first-place votes, the lowest-finishing candidate is dropped, and their voters' next-choice votes get counted instead. The process repeats until someone has a majority.

The point: the winner has support from a majority of voters, not just a plurality. No more "Candidate X won with 32% in a crowded field."

Still confused, let me go deeper…

How NYC's Version Actually Works

Here's the play-by-play, using the 2025 Democratic mayoral primary as the example.

Step 1 — You rank up to 5 candidates. NYC ballots let you fill in 1st through 5th choice. You don't have to use all five. You can rank one. You can rank three. The rest of the bubbles stay empty.

Step 2 — They count first-place votes. This is just like any normal election. Whoever you put at #1 gets your vote in Round 1.

Step 3 — If someone hits 50%+1, it's over. Done. They win. RCV doesn't kick in.

Step 4 — If nobody hits 50%, the math starts. The candidate in last place gets eliminated. The voters who ranked that person #1 don't lose their vote — their #2 choice gets counted instead.

Step 5 — Repeat until someone clears 50%. Each round, the bottom candidate drops out, and their voters' next-ranked choice gets transferred to whoever's still in the race. Eventually, someone has a majority.

Step 6 — Exhausted ballots. If all five of your ranked candidates get eliminated, your ballot is "exhausted" — it stops counting. (We'll come back to this in the Cons section. It matters more than people think.)

That's it. That's the whole system.

In November 2019, NYC voters approved Ballot Question 1 by 73.6%, amending the city charter to use RCV starting in 2021.

It applies to primary elections and special elections for five offices: Mayor, Public Advocate, Comptroller, Borough President, and City Council.

It does not apply to general elections. So in November 2025, when Mamdani faced Cuomo (running as an independent after losing the primary) and Republican Curtis Sliwa, that was a regular plurality race. Whoever got the most votes won. Mamdani took it with 50.78% — a clear majority anyway, so the system didn't matter much. But if it had been close, the candidate with even 35% could have won outright.

It also doesn't apply to state or federal races. Your governor, your congressmember, your state senator — those are still all plurality elections.

This is the part everyone has an opinion on. Let's run the actual numbers.

Round 1 (the raw count, June 24, 2025):

Per the Wikipedia primary results, Mamdani led the first round outright by about 7 points and 70,000 votes. But nobody had a majority, so RCV started.

The DREAM Team and the cross-endorsement. Two weeks before primary day, on June 13, Mamdani and Lander did something unusual: they cross-endorsed each other. They told their supporters to rank each other #2 and to leave Cuomo off the ballot entirely. A separate independent campaign called DREAM — "Don't Rank Evil Andrew for Mayor" — pushed the same message: don't put Cuomo anywhere on your ballot.

This is the move that RCV makes possible. Under a normal election, two progressives splitting the anti-Cuomo lane would have been a disaster — classic spoiler. Under RCV, they could collaborate without canceling each other out.

Round 2 and 3 (the transfers). After Adrienne Adams and the rest were eliminated, Lander dropped next. When his roughly 130,000 first-place votes were redistributed, more than 85,000 flowed to Mamdani. About a third of them flowed to Cuomo. The rest are exhausted.

Final result, after three rounds:

Total transfers Mamdani picked up: ~99,000. Total transfers Cuomo picked up: ~53,000.

The cross-endorsement worked. The DREAM strategy worked.

Short answer: No.

Here's why this is the question worth running the math on: it's the one everyone gets wrong in both directions.

Under a normal first-past-the-post (FPTP) primary — whoever gets the most first-place votes wins — Mamdani would have won anyway. He had 432,305 first-place votes. Cuomo had 361,840. That's a 7-point, 70,000-vote gap. Mamdani didn't squeak past. He led the field outright.

What RCV did was take a 7-point margin and turn it into a 13-point landslide (56.4% to 43.6%). That's not nothing — a majority winner has political legitimacy that a plurality winner doesn't. But it didn't pick the winner. It confirmed the winner.

The bigger RCV effect was upstream of the ballot: it let Mamdani and Lander collaborate openly without spoiling each other. Under FPTP, that cross-endorsement would have been impossible. Lander's 12.9% would have stayed siloed, Mamdani would still have led, and Cuomo's path to victory would have required overtaking an opponent who was already 70K votes ahead.

The honest read:

Both can be true.

1. The winner has a real majority. Plurality elections in crowded fields routinely produce winners with 25–35% support. RCV forces a majority outcome, which most political scientists treat as more democratically legitimate.

2. Less negative campaigning. Multiple studies have found that candidates run more positively under RCV because they need to ask for #2 and #3 votes from rivals' supporters. Trash-talking a candidate makes their voters less likely to rank you at all. The Mamdani-Lander cross-endorsement is a textbook example.

3. More candidate diversity. Research from FairVote and 19th News shows women and candidates of color win RCV elections at higher rates, partly because the "vote-splitting" excuse for not running disappears.

4. No spoiler problem. You can rank a long-shot candidate first without "wasting your vote." If they lose, your second choice still counts.

5. Cheaper than runoffs. RCV replaces runoff elections, which cost cities millions and produce dismal turnout.

1. Exhausted ballots. A study of more than 600,000 ballots found exhaustion rates ranging from 9.6% to 27.1%. That means in a hot race, between 1-in-10 and 1-in-4 ballots stop counting before the final round. The "majority winner" is a majority of remaining ballots, not of all voters.

2. Disparate impact concerns. Harvard's Ash Center found exhaustion rates are higher in districts with concentrations of Asian, Hispanic, and other voters of color. The mechanism: less translated voter education, less familiarity, more people ranking only one or two candidates instead of all five. This is the criticism RCV defenders take most seriously.

3. Complexity. A first-time RCV voter has to learn an unfamiliar ballot. Critics argue this depresses turnout among older voters, immigrants, and first-time voters. Defenders point to Alaska's 2022 election, where 85% of voters said RCV was simple and 99.8% of ballots were valid.

4. Slower results. Final tabulations can take a week or longer. NYC's 2025 primary first-round count was June 24; the final RCV result was July 1. In our news cycle, that's an eternity.

5. The "monotonicity" problem. This is a wonk objection but it's real: in rare scenarios, ranking a candidate higher can actually hurt them, because of how transfers cascade through eliminations. It almost never matters in practice. But it's mathematically true, which gives RCV critics a reliable talking point.

NYC just had a special election yesterday in Council District 3 — the Manhattan seat covering Hell's Kitchen, Chelsea, and the West Village. Carl Wilson won the first round with 43.08% against Boylan's 25.66%. Sound familiar?

Because nobody hit 50%, RCV is now playing out in the background. Final results may slip to Tuesday, May 5.

If you've been wondering why your news app is saying "Wilson wins" while the official results are still pending — that's RCV. The first-round leader almost always becomes the final winner (Mamdani did, Wilson likely will), but the certified result waits for the rounds to finish.