The stars are aligned against a unified front to stop a radical takeover of America’s greatest city.

N ew York City seems to be careening toward electing as its mayor Zohran Mamdani — a young, inexperienced, unaccomplished, red-diaper socialist born with a silver spoon in his mouth who has built his political identity around Israel-bashing, pie-in-the-sky economic ideas debunked by the whole history of humanity, and law enforcement ideas very specifically debunked by the in-living-memory history of New York City. Frustratingly, Mamdani faces a fractured opposition: incumbent Mayor Eric Adams, who fled the Democratic Party after his federal indictment for garish corruption made renomination impossible; Republican candidate, radio host, and Guardian Angels founder Curtis Sliwa, who was trounced by Adams in 2021; and former Governor Andrew Cuomo, who resigned in a state of disgrace (specifically, New York) and lost to Mamdani in the primary. To top it off, there’s a wealthy independent vanity candidate, Jim Walden.

I’m glad I moved from Queens to Long Island nine years ago, after two decades living inside city limits. But what happens in New York City still matters quite a lot.

If there’s to be any hope of beating Mamdani, who inspires particular enthusiasm among voters who did not live in the city before Bill de Blasio became its mayor in 2014, there must be some winnowing of that field. In the RealClearPolitics average, comprising five polls taken over July, August, and early September, Mamdani has a double-digit lead but averages 38.6 percent of the vote. That suggests that he is still a ways from the majority support needed to win a two-way race — but in the two most recent polls (by AARP in early August and Siena in early September), he was at 42 and 46 percent, respectively — inching closer to sealing the city’s fate.

The latest of those, from the New York Times/Siena, came out this morning. It shows Mamdani with almost double the support of his nearest rival among likely voters, with Cuomo at 24 percent, Sliwa at 15 percent, and Adams at 9 percent. In a hypothetical head-to-head matchup with Cuomo, Mamdani leads 48 percent to 44 percent; against Adams, he leads 55 percent to 36 percent. Siena did not poll Mamdani head-to-head against Sliwa.

There are problems with persuading any of these people to drop out of the race. Sliwa is an ornery character and can quite reasonably argue that he is, unlike Adams or Cuomo, the choice of one of the two major parties. Adams and Cuomo are quite aware that this is their last chance to win elective office and change the narrative of having ended their careers in scandal. Neither man has much of a record of knowing when to take no for an answer. If Cuomo didn’t give up after a humiliating primary loss for governor in 2002, and didn’t quit this race after the primary, when will he?

The Trump political operation has zeroed in on the well-deserved perception that Adams can be bought. Running fourth in most polls and increasingly struggling to get out of single digits, he has to know that he isn’t getting reelected and accomplishes nothing for himself by continuing to try. So, he held a press conference on Friday to announce that he’s . . . staying in the race, more or less for the open purpose of holding out for a better offer. Oh, and “I want to be clear: Andrew Cuomo is a snake and a liar.”

There are a lot of egos and grievances involved here. Also many eccentricities, such as Sliwa running not only as a law-and-order guy but also as an animal-rights cat lover.

The fundamental dilemma is not just that you can’t beat a single candidate with a crowd of alternatives, that you can’t beat somebody with nobody. We’ve been through this dynamic in Republican primaries before, trying to unify the opposition to Donald Trump or Mitt Romney or statewide candidates such as Mehmet Oz. Cobbling together their opponents assumes that everybody is equally willing to back one alternative — even when none of them excite everyone. The Siena poll found that 60 percent of voters not supporting Mamdani would never consider supporting him — but 19 percent said that they could consider supporting him under some circumstances, and another 15 percent were either strongly or not so strongly considering him.

There are a lot of problems with the incentives here, and many problems with stitching together a coalition from the extremely disparate elements of opposition to Mamdani. Adams draws a lot of his support from appeals to his status as the only black candidate in the race. Many Sliwa supporters are Republicans, conservatives, or otherwise people who have hated Andrew Cuomo for so many years that they are unlikely ever to be persuaded to vote for him. Many Cuomo supporters are lifelong Democrats with no intention of ever voting Republican. While I take seriously Jim Geraghty’s view that Sliwa has a stronger base of support than Cuomo or Adams, the likelihood is that many current Cuomo supporters would see Mamdani as the lesser evil compared with Sliwa. Sliwa’s best showing is likely to come in a three-way race against Mamdani and Cuomo, but that does no good for the anti-Mamdani cause.

Voters who hate Trump — and there are vast numbers of these, despite the great inroads he made on the margins in New York in 2024; even Adams has a higher approval rating in the Siena poll — are ill-disposed towards Sliwa for being a proto-MAGA Republican, and toward Adams after he effectively bent the knee to Trump in order to get Trump to call the DOJ off him. They are apt to be similarly skeptical of Cuomo if Trump’s muscle clears the field for Cuomo, no matter how loudly Cuomo denounces and threatens the president.

Then there are the national GOP’s incentives. Mamdani may be a disaster for the nation’s most visible city, and this may be enough to worry Trump, who remains a lifelong New Yorker in his bones. But Republican ad makers, consultants, and direct-mail fundraisers salivate at the prospect of Mamdani’s radicalism as a foil. That puts a cap on how far the White House is likely to be willing to go in order to stop him.