A Graveyard of Bad NYC Mayoral Election Narratives
The electoral outcome and its drivers were far more banal than partisans across the board seem willing to recognize.
According to his detractors, Zohran Mamdani won the mayoral race primarily because of highly-educated, relatively affluent, urban and suburban whites (including and especially symbolic capitalists). By implication, “normies” across racial and ethnic lines sided with other candidates at disproportionately high levels.
Meanwhile, the mayor-elect’s supporters argue that Mamdani built an extraordinary coalition, pulling in lots of people who would not otherwise have voted for Democrats, and ultimately pulled off an historic win for an overtly socialist candidate.
In reality, both of these narratives (and many other popular culture war claims about the 2025 NYC mayoral election) are demonstrably false.
Historic? Not So Much…
The Democratic nominee won the 2025 NYC mayoral race. This is hardly an unprecedented or surprising turn of events.
Democrats control the NY city council by 9:1. They dominate the NY state legislature by 2:1. The governor of NY is a Democrat. Looking at New York’s federal delegation, the state’s representatives in the House skew Democrat nearly 3:1. Both of the state’s senators are Democrats. Kamala Harris went down in flames in the 2024 U.S. presidential election. Yet, despite losing every single swing state, she notched more than twice as many votes as Donald Trump in NYC.
Put simply, New York City is very “blue.” Democrats winning municipal elections in the Big Apple is the most banal outcome imaginable. Between 1932 and 2024, there were 26 NYC mayoral elections. Republicans won 7 of them. Independents won 3. The rest were won by the Democratic Party. Democrats won the last three consecutive mayoral races, each of them by 2:1 margins. Even the last Republican mayor of NYC was functionally a Democrat.1
Zohran Mamdani is not the reason Democratic Party succeeded in the 2025 mayoral race. If anything, Democrats won despite him. In a world where Mamdani was not on the ballot, the outcome would have almost certainly been the same: the Democratic candidate (be it Adams, Cuomo or someone else) would have won. In Mamdani’s absence, the Democratic nominee likely would not have faced a well-funded intraparty challenger in the general election. Consequently, they probably would have won the election by far larger margins than Mamdani did (albeit with significantly lower turnout). In other words, the main “Mamdani effect” on the topline outcome of this cycle was that he made the race much closer for Democrats than it otherwise would have been.
Granting this point, one might think that what made Mamdani’s win impressive was not that the Democrat won, but that a Democrat with his specific politics managed to secure the Democratic nomination and win the general election. But here, too, the outcome is not so extraordinary.
Every single mayoral election from 1913 through 1933 featured a socialist candidate. Following the demise of the Socialist Party, Democrats nominated multiple socialists to the top of their ticket. Occasionally socialist candidates won the general election. Beyond the mayoral race, one of the most prominent Democratic Socialists in the country (and a likely 2028 U.S. presidential contender), Alexandria Ocasio-Cortez, represents New York’s 14th Congressional District. Socialists are hardly unheard of in New York politics.
In many respects, Zohran Mamdani’s election is a replay of Bill DeBlasio’s 2013 victory. DeBlasio also launched from obscurity and had limited relevant experience and some troubling political positioning in college. He won through a grassroots campaign focused tightly on economic issues. His election led to similar overwrought narratives from friends and foes alike about what was likely to happen under his administration. Here’s how the NY Times characterized the 2013 race (quoted at length, emphasis mine):
“Bill de Blasio, who transformed himself from a little-known occupant of an obscure office into the fiery voice of New York’s disillusionment with a new gilded age, was elected the city’s 109th mayor on Tuesday. His landslide victory… amounted to a forceful rejection of the hard-nosed, business-minded style of governance that reigned at City Hall for the past two decades and a sharp leftward turn for the nation’s largest metropolis…
It was the most sweeping victory in a mayor’s race since 1985… In Manhattan, Mr. Lhota, a 59-year-old Republican, quieted boos from his disappointed supporters as he conceded the race from behind a wooden lectern at a hotel in Murray Hill…
The lopsided outcome represented the triumph of a populist message over a formidable résumé in a campaign that became a referendum on an entire era… Throughout the race, Mr. de Blasio overshadowed his opponent by channeling New Yorkers’ rising frustrations with income inequality, aggressive policing tactics and lack of affordable housing…
On Tuesday, Mr. Lhota put on a brave face… Like many New Yorkers, he was taken aback by Mr. de Blasio’s improbable rise. Raised a Boston Red Sox fan in Massachusetts, Mr. de Blasio embraced the cause of leftist Sandinistas in Nicaragua as a young man, married a woman who once identified as lesbian, and has never managed an organization larger than 300 people.
But Mr. de Blasio… oversaw a highly disciplined political machine that committed few errors and took little for granted, in stark contrast with Mr. Lhota. On Election Day, Mr. de Blasio had amassed around 10,000 volunteers at 40 locations to turn out voters; Mr. Lhota recruited about 500 workers at nine locations. The coordinated outreach paid off, with Mr. de Blasio capturing majority support from voters of all races, genders, ages, religions, incomes and education levels.”
Does any of this sound a little familiar? Perhaps the best way to understand Zohran Mamdani is as the second coming of Bill DeBlasio. Indeed, Mamdani has declared Bill DeBlasio as the best mayor New York City has had in his lifetime and DeBlasio, in turn, has been a strong supporter of his erstwhile successor.
One might think that what separates Mamdani from DeBlasio or overt socialist predecessors was turnout: Mamdani won more than a million votes. The widely-reported topline number is that Zohran is the first candidate to crack this threshold since 1969. But here, basic context is necessary.
Several previous mayoral candidates have won more than a million votes. From 1937 through 1969, literally every single race was won by someone who pulled in more than a million votes. Even the losing candidates in those cycles occasionally got more than a million votes!
Critically, the overall population of the city was significantly smaller in those previous races. Their candidates exceeded a million votes by mobilizing a much higher share of the electorate than turned out in 2025.
The overall turnout in this cycle was high relative to any race since 2001. Nonetheless, participation was significantly lower than most races before 1997, and much lower than races prior to the 1970s (when mayoral elections routinely pulled in more than double the vote share Mamdani inspired).
Of all the races we have electoral turnout data for, 2025 currently ranks 13th out of 19. Not exactly gangbusters.
Mamdani’s vote totals and the overall 2025 participation rate only seem impressive from our contemporary vantage point because New Yorkers have largely tuned out during the most recent mayoral election cycles (each of the previous five elections had record-low turnout relative to their predecessors).
In a similar vein, comparing Mamdani’s vote share (50.4%) to other races over this period, he’s tied for 12 out of 19. He secured the lowest vote share of any winning candidate since 2001.
Some quick math reveals additional context: by multiplying the turnout rate in each cycle by each winning candidate’s vote share, we can see that, not only did most successful candidates since the 1960s receive a higher vote share than Mamdani among those who turned out, but several won a larger share of the total electorate (which includes eligible voters who went to the polls that year alongside those who abstained) despite netting fewer voters than Mamdani in terms of sheer number. Among them: Beame (1973), Koch (1981), Dinkins (1989) and Giuliani (1993).
The only reason Mamdani ended up with more than a million votes (while Beame, Koch, Dinkins and Giuliani did not) is because the population of New York City is larger than it used to be. In reality, the mayor-elect reached his seven-figure threshold despite middling turnout and vote share thanks to changing city demographics that have nothing to do with Mamdani himself (or his campaign).
Moreover, as it relates to turnout, it is important to note that a huge share of the increase between 2021 and 2025 consisted of voters headed to the polls to vote against Zohran Mamdani.
Comparing 2025 to 2021, an additional 924,373 New Yorkers came out to vote. Sounds inspiring, right? Yet, Zohran only got 282,250 more votes than Eric Adams. This means that the vast majority (more than two-thirds) of net additional votes this cycle compared to last were against the Democratic nominee, not for him (which is why Mamdani’s vote share is markedly lower than Adams’).
An analysis by the New York Times found that if you zoom in on precincts where more than 10 percent of all voters registered after the Democratic primary results became known (i.e. places where tons of people signed up to vote because Mamdani became the Democratic nominee), Andrew Cuomo these areas by 52 percentage points. It seems as though Mamdani emerging victorious in the Democratic primary did inspires lots of New Yorkers to register to vote… so they could cast ballots against him in the general election.2
Put simply, the turnout numbers do not suggest Mamdani was especially popular, they show he was highly polarizing. In terms of raw totals, Mamdani may have won more votes than any candidate since 1969, but he also had more votes against him (aggregating the totals of his opponents) than any candidate since 1969. Andrew Cuomo got more votes than the winners of all other NYC mayoral contests since 1993.
The mayor-elect’s supporters selectively look at votes for Mamdani in order to pretend he enjoys spectacular support while conveniently overlooking the high levels of opposition he also generated among NYC residents. They trumpet the turnout percentages while ignoring that the bulk of the additional turnout compared to last cycle were demonstrably anti-Mamdani votes.
In conservative and anti-woke spaces, meanwhile, many have derided Curtis Sliwa as a “spoiler” candidate. This is absurd for many reasons.
First, as I’ve detailed elsewhere, appealing to “spoilers” is an impoverished way to understand any electoral outcome. Sliwa didn’t stop anyone from voting for Cuomo. If people didn’t vote for Cuomo, it’s because he failed to earn their vote. Many of those who declined to vote for Cuomo in a three-person race would have also declined to vote for him in a two-person race (they’d have stayed home or perhaps even voted for the other guy instead). But bracketing this point, as Sliwa himself emphasized, “Even if every person who voted for me, voted for him - the prince of darkness still wouldn’t have won.”
A more fundamental problem with this narrative is that Curtis Sliwa was the GOP nominee appointed by the usual process. If anyone could be reasonably understood as a spoiler in this race, it’d be Andrew Cuomo. In terms of intent, this was literally his purpose: he launched a third party bid to deny Mamdani victory because he was dissatisfied with how the Democratic primary shook out. And, although he failed to keep Mamdani out of office, he succeeded in muddying his opponent’s win.
Many point to Mamdani’s margin of victory and (correctly) observe that it’s the lowest in decades – conveniently ignoring the fact that there hasn’t been a comparable three-person race (featuring an intraparty Democratic challenger) in nearly a half-century.
Ironically, the last time we saw a race like this was in 1977 when Mario Cuomo lost the Democratic primary and then launched an unsuccessful third party bid to sandbag Ed Koch. That race ended almost exactly the same way as this one: the Democratic nominee won just over 50 percent of the vote and Cuomo won 41 percent (with the Republican and other candidates taking the remainder).
Critically, the 1977 Democratic nominee didn’t just survive Cuomo’s intraparty campaign, he went on to serve two more terms (in each cycle facing other unsuccessful primary challengers who went on to run against him in the general election too – with these intraparty rivals securing smaller and smaller vote shares each cycle). The lesson: Mamdani may not have won by a huge margin in his initial outing (thanks to Cuomo), but this doesn’t signal anything about his eventual reelection prospects.
In any event, anyone trying to throw cold water on Mamdani’s victory by noting that he won by much smaller margins than Eric Adams or Bill DeBlasio is comparing apples to oranges. In a world where Mamdani squared off in a two-party race against Curtis Sliwa, he’d have likely won in a landslide just like his immediate predecessors did (although turnout would likely have been significantly reduced in a two-person race precisely because its outcome would not be in doubt, so Mamdani’s overall vote totals may have ended up the same or smaller despite Cuomo’s absence).
That said, when we directly compare 1977 to 2025, a true apples-to-apples comparison, it doesn’t exactly reflect positively on Mamdani either. Unlike Koch, Mamdani was operating in a period of Democratic hegemony over New York city and state. He was running against a much weaker intraparty candidate (Mario Cuomo was a far better politician than his son ever was). Yet he couldn’t manage to exceed Ed Koch’s vote share, and his race inspired a significantly smaller share of the electorate to the polls. That is, even looking at equivalent cases, Mamdani’s performance was far from extraordinary.
Ultimately, the outcome of the 2025 mayoral election was extremely banal: the Democrat won (surprise!). What’s more, the winner’s electoral coalition was super ordinary for a Democrat in New York City too. As I’ll demonstrate below, Mamdani was propelled to his decisive (albeit, far from “historic”) victory exclusively through constituent blocs that typically vote Democrat – and they supported Mamdani at levels consistent with other Democratic nominees.
Race and the Race for City Hall
Starting in the primaries, a narrative emerged that Mamdani’s primary base of support consisted of highly-educated and relatively-affluent whites. By implication, normie voters of most other ethnicities were aligned behind Cuomo. Even during the primaries, this narrative was an exaggeration. With respect to the general election, it’s straightforwardly false.
When you look at how the race changed between June and November, Mamdani saw significant gains in black and Hispanic majority precincts while Cuomo’s support came overwhelmingly from white-majority precincts – including and especially white-dominated precincts that typically swing Republican.
When you get more granular, looking at neighborhoods rather than precincts, the results are basically the same. Cuomo dominated white-majority neighborhoods, while Mamdani swept places with mixed populations and areas where most residents are black or Hispanic.
That said, at the neighborhood level, we see that although Mamdani won most of the vote in Asian-majority precincts, more Asian-majority neighborhoods swung to Cuomo. This probably has something to do with the role of class in shaping vote preferences this election – more on that soon.
Exit polls tell a similar story to the precinct and neighborhood data. Mamdani won a clear majority of votes among all non-white ethnic populations. Among white voters, however, it was a statistical tie (the difference between Mamdani and Cuomo falls within the 2.2 percentage point margin of error for the poll):
Put simply, there isn’t racial culture war story to be had here. Mamdani won with a broad ethnic coalition. Cuomo lost with everyone except whites… but he didn’t decisively win whites either.
This might sound like an impressive victory for Mamdani, but it really isn’t. When Bill DeBlasio won in 2013, he scored significantly higher margins with all of these ethnic groups.
When Democrat William Thompson Jr. lost to Michael Bloomberg in 2009, he secured roughly the same percentage of White and Hispanic voters as Mamdani did. He also scored a much higher share of the black vote than Mamdani did (although this may be, in part, because Thompson stood to be New York City’s second-ever black mayor). Asian voters weren’t captured in exit polls at the time, so we can’t compare Mamdani’s performance to Thompson’s among this voter bloc.
All said, there is nothing unusual about the Democratic nominee winning most Asian, black or Hispanic New Yorkers. Mamdani didn’t win any of these groups by huge margins (relative to other Democratic NYC mayoral candidates).
If you compare the NYC precincts Kamala Harris won in 2024 to the precincts Mamdani won in 2025, the maps are not very different. Mamdani performed a bit stronger in Asian (especially South Asian) precincts, but these modest gains were more than offset by underperformance with African Americans and whites. Defections among these latter groups (to Cuomo), especially among whites, are why Kamala Harris was able to win 67.7% of the NYC vote while Mamdani only secured half.
But, again, despite underperforming other Democrats with black voters, Mamdani did, in fact, win this constituency decisively. His numbers with Hispanic voters were typical for a Democrat. His numbers with Asian voters were strong. He tied Cuomo with white voters. All said, Mamdani secured his victory through a broad ethnic coalition that was not particularly different from those of other Democratic nominees.
Keeping it Classy
According to exit polls, Zohran Mamdani did well with voters who possess college degrees and performed poorly among those who don’t. A graph visualizing the relevant data was circulated far and wide in order to suggest that Mamdani’s core base was either
Privileged elites, or
Downwardly-mobile frustrated erstwhile elites.
An implication of these narratives — often made explicit by folks sharing the chart — is that Andrew Cuomo is the candidate of normie New Yorkers. In fact, the data do not show this at all.
To understand what the chart actually shows, the first thing to note is that New York City has a significantly higher concentration of college graduates than America as a whole. Moreover, constituents who have completed college turn out to vote at significantly higher rates than those who have not – especially for elections where the presidency is not on the ballot (such as the 2025 NYC municipal contests).3 Consequently, the “normal” or “modal” New York voter is someone who has, in fact, obtained a college degree of some type. Folks who have not obtained a postsecondary credential are the outliers in the NYC mayoral electorate – not the other way around.
The exit poll actually makes this clear in the left sidebar, which includes SSRS projections of how much each educational bloc contributed to the electorate overall. They estimate that 59 percent of 2025 NYC mayoral voters had a BA or higher while an additional 8 percent had an AA degree. Only a third lacked any form of postsecondary credentials (and many of those are likely currently enrolled in college). To hold up this extreme minority of voters as the “real” voice of New York is a strange move.
In any event, talking heads are wildly exaggerating the degree of educational polarization in this election.
According to the exit polls, Cuomo won 5 in 10 New Yorkers who had no postsecondary credential. Guess what? Mamdani won 4 in 10 voters from this same demographic. These voters did not uniformly rally behind Cuomo or against Zohran – they were pretty split. Voters with associate degrees were exactly evenly split.
We see similar patterns in the precinct data: according to New York Times estimates, Mamdani won precincts where a majority of residents have a BA or higher by 15 percentage points. However, he also won precincts where most voters do not have BAs (albeit, by a smaller margin). He did well with both groups.
On the flip side, not only does “the narrative” oversell Cuomo’s support among voters without postsecondary degrees, it also undersells his support among the “privileged” degree holders.
College graduates broke for Mamdani, but far from unanimously. Cuomo won 4 in 10 voters with BAs or graduate degrees. In fact, when we attend to base rates in the sample, the exit poll data actually suggests that the vast majority (63%) of Cuomo’s voters had some form of postsecondary credential.4 Another way of saying this is that both candidates drew most of their support from voters with college degrees.
In fact, Democrats almost certainly performed worse with college voters this cycle compared to last. Again, exit polls suggest Mamdani won just 60 percent of voters with a BA or higher. Adams certainly beat that. There’s no way that the Democratic nominee (who stood to be just the second black mayor in NYC history) performed worse than Mamdani did with college voters in a two-person race against Republican Curtis Sliwa. In fact, when we look at how different voting precincts skewed in the last election, and attend to the fact that Adams won all voters 2:1 — and he certainly overperformed with degree holders compared to the electorate as a whole (as Democrats typically do) — we can reasonably estimate that Adams may have won 75 to 80 percent of the college vote. Again, educational sorting was likely less pronounced in this cycle than last.
Likewise, if you compare the 2025 mayoral race to the 2024 U.S. Presidential Election you can see that Zohran Mamdani underperformed Kamala Harris with both the least and most educated precincts while outperforming her in the middle.
Put another way, Cuomo significantly outperformed Trump with college educated voters. He performed particularly well (relative to Trump) in enclaves with especially high concentrations of voters with college degrees. The education polarization narrative is clearly a bust for the NYC mayoral race.
But, of course, education is only way to measure class. There’s also the small matter of income and wealth. Here, the exit polls tell an interesting story:
Mamdani was decisively favored by working class, middle class, and upper-middle class voters (i.e. constituents whose total family income fell anywhere between $30,000 and $200,000). This income range includes the vast majority (67 percent) of voters. However, Mamdani underperformed with the very poorest New Yorkers; they slightly favored Cuomo. The mayor-elect performed even worse with NYC’s richest constituents (from households that rake in $300,000 per year or more). Those folks favored Andrew Cuomo by 2:1.
When we examine Cuomo’s support in each income bracket relative to that income bloc’s share of the total electorate, we can see that Cuomo received far more votes from New Yorkers from households bringing in least $200k per year than from those who make $30,000 or less. Not exactly a working class hero.
Precinct data paint a similar picture: Mamdani won precincts where most residents use public transportation to get around. Cuomo won areas where most people drive private cars. Cuomo ran up huge margins in precincts where most residents own the home they live in (exit polls also show Cuomo winning strongly with homeowners). Meanwhile, the (vast majority) of precincts where most residents live in rentals went decisively for Mamdani. Among renters, however, Cuomo did win the precincts with the very lowest rent.
When you break out precincts by income, Mamdani decisively won precincts with lower and middle income residents, while Cuomo ended up on top in higher-income precincts (see the first chart in the previous section).
Should you compare the 2025 NYC mayoral race to the last U.S. presidential election, you can see that Cuomo significantly outperformed Harris in affluent enclaves and traditionally Republican voting areas, while significantly underperforming in working class areas and traditional “blue” precincts.
Looking at individual neighborhoods, many of Mamdani’s critics have circulated a New York Times graphic showing that Zohran performed strongest in Brooklyn. Typically, these partisans conveniently edit out the part of the chart showing that most of Cuomo’s best neighborhoods were also in Brooklyn, alongside a few affluent neighborhoods in other boroughs.
When we look at campaign donations and expenditures, we see a similar story as the other data.
Open Secrets finds that two-thirds of donations to Mamdani’s campaign were from folks who contributed $100 or less. Andrew Cuomo’s campaign received far fewer donations overall, but the lion’s share of his contributions were for $250 or more.
With respect to expenditures, Drop Site found that both campaigns spent similar amounts of their own money – most of it from public funds. However, for “independent expenditures” (i.e. billionaire-funded PACs aimed at undermining one candidate and/or bolstering another) there was no contest. Cuomo backers outspent Mamdani supporters by 7:1.
Andrew Cuomo wasn’t just the candidate overwhelmingly supported by the rich at the ballot box (as reflected in exit polling and precinct voting patterns), he was also the candidate they threw their money behind.
In the 2024 U.S. presidential election, the affluent lined up behind Kamala Harris. Buoyed by billionaire support, her campaign and aligned PACs outspent Trumpworld by 2:1 but she ultimately lost every swing state. In a similar vein, looking at total campaign spending across the entire 2025 NYC mayoral cycle, Cuomo & co. outspent Mamdani 2:1 and still lost decisively. The lesson in both cases: it’s hard to buy elections… especially if the billionaires’ preferred candidate is objectively terrible.
With respect to class, all the data tell a consistent story: Mamdani won with a broad coalition that cut across most education and class levels. Cuomo’s support, however, was heavily polarized by income: the very poorest and least educated voters seemed averse to Mamdani and chose to back Cuomo instead. Simultaneously, the most privileged New Yorkers rallied even more fiercely behind the former governor. In fact, with respect to overall vote totals, campaign contributions, or any other metric, Cuomo’s support was skewed heavily towards the very top of the wealth and income distribution.
To be clear: symbolic capitalists were clearly an important part of Zohran Mamdani’s coalition. However, there’s no evidence he performed particularly strongly with these voters (relative to other NYC Democrats) and he enjoyed plenty of support from “normies” as well. The mayor-elect performed solidly in minority-majority and working-class areas in addition to Brooklyn’s “commie corridor” or Manhattan’s Upper West Side.
Meanwhile, Cuomo did especially well in areas like West Chelsea and Tribeca, where New York’s wealthiest are now congregating in droves. Many of Cuomo’s most prominent billionaire backers (such as Bill Ackman to Michael Bloomberg) derived their wealth from the symbolic industries. And, as noted above, exit poll data suggest that nearly two-thirds of Cuomo’s voters possess postsecondary credentials.
Partisans like to tell selective culture war stories. However, the big trend in the data is that there actually wasn’t a clear divide between symbolic capitalists and everyone else in this election. The symbolic capitalists were, themselves, deeply divided (with the most privileged among them tilting towards Cuomo). This is something they shared in common with non symbolic capitalists.
Sex and the City
On social media, “woke” and anti-woke culture warriors alike have fetishized one particular image from the exit polls:
Among Democrats, this image was held up as proof that Cuomo’s history of sexual misconduct was coming back to haunt him. Others viewed this graph as proof that right-aligned and anti-woke hostility towards feminism (paired with tolerance of misogyny) would consistently lead the “bad guys” to electoral defeat. It’s hard to win races without female support, after all…
On the right, this same chart fed into a conversation rekindled by a recent viral essay by Helen Andrews that blames women, feminism and institutional feminization for most adverse political and cultural trends of the last decade. People who buy into that thesis have been pointing to this chart as evidence that Mamdani won because of extraordinary support from women (and both this outcome and its alleged drivers are viewed as “bad” in this telling).
There are many silly elements to this gender polarization narrative. For one, the spread between Cuomo and Mamdani was 9 percentage points. Those 18-29 year old women who supported Mamdani over Cuomo at a rate of 7:1? They comprised only 6 percent of the estimated electorate. Consequently, every single one of these women could have voted for Cuomo instead of Mamdani and it would not have changed the outcome of the race at all. To blame or credit these voters for Mamdani’s victory is show that one has not paid close attention to the data that’s right there on the graph.
Moreover, within the 18-29 age range, there is not actually much gender polarization. Those who fetishize the fact that young women were more likely to support Mamdani than male peers seem to have lost track of the fact that roughly 7 in 10 men under 30 also voted for the Democratic nominee. Young people across gender lines overwhelmingly cast ballots for Mamdani (more on this soon).
Even as gender differences in young folks’ vote preferences have been wildly exaggerated by culture warriors, there has been chronic neglect of the poll’s finding that outside of the 18-29 age bracket, there is no gender divide at all: two thirds of women and men between 30 and 44 supported Mamdani. For voters aged 45-64 or 65+, the difference between men and women also falls within the poll’s margin of error. If anything, for respondents 30 and up, Mamdani may have been marginally more popular with men — a finding that doesn’t suit the preferred narratives of either party.
When we aggregate the data across age groups, we can see that there is, in fact, no gender difference at all with respect to Mamdani support. According to SRSS estimates, women turned out to vote in this race at significantly higher levels than men (typical of U.S. elections over the last 50 years). However, given that men and women supported the Democratic nominee at statistically indistinguishable levels, these turnout differences didn’t matter at all for shaping the ultimate outcome of the race.
In fact, the main gender difference in candidate support seems to lie in the anti-Mamdani camp, where men seem to have been slightly more likely to support Sliwa while women opposed to Zohran lined up more uniformly behind Cuomo. This, of course, sits awkwardly with “woke” narratives about why the election shook out how it did (according to those narratives, one would expect that women registering a protest vote against Mamdani would choose the alternative candidate who wasn’t a confirmed sexual predator. But that’s not what we see).
In truth, Mamdani did not perform especially strongly with women. In fact, he significantly underperformed most other recent Democratic mayoral nominees (not to mention Kamala Harris in 2024). On the flip side, despite his baggage, Andrew Cuomo didn’t perform poorly with women. If anything, he was slightly more favored by anti-Mamdani women than anti-Mamdani men. These are tough patterns for culture warriors on any side of the election.
When you intersect gender and race, you see an interesting pattern where white women were more supportive of Mamdani than white men. Yet. for non-whites, this pattern is reversed: black and Hispanic women were less supportive of Mamdani than black or Hispanic men.
It’d be perhaps tempting for some explain this pattern in terms of some nonsense about “white women” and “wokeness.” The only problem is, in fact, Mamdani’s support is significantly weaker among white women than with black or Hispanic women. Black and Hispanic women decisively favored Mamdani over Cuomo overall – it wasn’t even close. But among white women, the difference between the two candidates fell within the margin of error.
Again, there isn’t a clean culture war story to be told here.
The only remotely interesting gender-related pattern I could see in the data was not between men and women, but rather, among women…. based on whether or not they were parents.
Among men, parenthood made no difference in support for Zohran Mamdani (although anti-Mamdani men without kids are marginally more likely to support Sliwa over Cuomo). Among women, the picture is very different. Women who do have children are significantly more likely to support Cuomo, while women who don’t have children are significantly more likely to support Mamdani.
But again, at the aggregate, all of this washes out, such that there is literally no difference at all between male and female support for Mamdani. He decisively won both gender blocs with roughly identical vote shares.
Put another way, the big story of the race, which comes through crystal clear in the data, is that there was no electorally-relevant gender polarization this cycle.
Distinct from the question of gender polarization in the electorate, many anti-woke and right-wing culture warriors have also mischaracterized the sexuality break down of the exit polls. According to these narratives, Mamdani won City Hall primarily because of votes by LGBTQ New Yorkers (who gave 82 percent of their votes to the mayor-elect).
But, in fact, LGBTQ New Yorkers did not support Mamdani at unusual levels. They voted the same as they typically do. Consider 2021: it was a two person race between Eric Adams and Republican Curtis Sliwa. In this race, which was a landslide, we can safely conclude that LGBTQ Americans supported Eric Adams at similar rates that they supported Mamdani in 2025. Curtis Sliwa did not pull in more than a fifth of the queer vote.
Alternatively, consider the last race we have good exit poll data for: DeBlasio’s win in 2013. In that race, the Democratic nominee performed a bit better than Mamdani with LGBTQ voters.
In 2024, Kamala Harris performed poorly overall. Yet she still won 86 percent of LGBTQ voters.
Put simply, there was nothing extraordinary about Mamdani’s performance with queer and trans voters. He put up the same numbers Democrats typically secure with these constituents. With respect to both gender and sexuality, there is very little that was unusual in this race.
Kids These Days?
There is a culture war consensus that any outcome favoring the left should be laid at the feet of young people.
Reactionary conservatives and anti-wokes like to blame any unwelcome developments on “kids these days” — even when it was clearly older people driving the outcomes in question — while telling inaccurate nostalgic stories about how things used to be “back in the day.” People on the left don’t reject these distortions; they largely embrace them. Depicting stuff they’re personally championing as having originated with young people (or as otherwise representing the will and interests of the youth) affirms progressives’ sense that they’re forward thinking, “with it,” and on the “right side of history.”
And when you look at exit poll data on Mamdani’s victory, they superficially support “the narrative.” Young people are far more supportive of Mamdani than older New Yorkers.
However, Zohran’s vote share among young people was hardly extraordinary. Consider the 2021 electoral contest: Curtis Sliwa racked up only 28 percent of the overall vote, and he certainly performed better with older people than younger voters. With this in mind, we can reasonably estimate that Eric Adams’ vote share among young people was likely the same or higher than Mamdani’s.
Given that young voters likely had the same vote allegiance this cycle as last, the focus on the youth seems misplaced. The folks whose alignments actually changed between this cycle and last (who, therefore, had the most potential to change the electoral outcome) were older voters.
Given Adams’ landslide win, and the heavy turnout skew towards older voters in the last cycle, we can confidently conclude that the Democratic nominee did far better with voters over 40 in 2021 compared to 2025. Here, there is an actual change in voting patterns. Cuomo’s votes were primarily achieved by poaching older voters who cast ballots for main-party candidates in previous cycles. But rather than focusing on the “olds,” and trying to figure out what’s going on with middle-aged and senior voters to explain their change in voting behavior, there’s been an intense focus on “kids these days” instead.
When we look at precinct voting data, it largely confirms the picture from the exit polls. There Is a linear relationship between candidate vote share and median precinct voter age. However, the skew is not as dramatic as many seem to assume. Among precincts where the median age is 50-55, half of voters supported Cuomo while 4 in 10 support Mamdani. In precincts where the median age was 40- 49, these ratios were identical but flipped. In both cases, the differences were real, but not huge.
We can reinforce this point by aggregating precinct data: Mamdani did really well in the minority of precincts where the median age of residents was under 45. But then again, this would be true of the Democratic nominee in most other cycles as well. More interesting is Cuomo’s razor-thin lead in precincts where most residents are over 45. The narrowness of his lead in these areas suggests that huge numbers of older voters also cast ballots for Mamdani. and the “age polarization” narrative is likely overstated.
We can put the final nail in the coffin of the “kids these days” rhetoric by attending to the SSRS estimates (provided in the left-hand column of the exit poll charts), modeling how much each bloc of voters contributed to the overall electoral outcome. These data also allow us to estimate each age bloc’s contribution to each candidate’s coalition.
Making use of these data we can see that young people were not Mamdani’s core base. Instead, the majority of his voters would have been 40 or older. Making use of the SSRS estimates of the electorate included in the exit poll, the single biggest numerical block of Mamdani voters seem to fall in the 30-39 age range (i.e. the age of people settling down, forming families, hitting stride in their careers — not “kids these days”). Voters under 30, who’ve been the primary object of media discussion about age and this race, comprised less than a quarter of all Mamdani voters. In short, the widely-circulated chart does not insinuate that Mamdani won primarily because of “kids these days.” When you make use of all the data presented to viewers in that image, we can see that the “narrative” gets the facts exactly backwards.
That said, there is some evidence that the SSRS estimates – modeled based on previous cycles and early voting patterns – sold the youth vote a little short. Because voters must include their birthdays when they register to vote, the Board of Elections can produce exact estimates of turnout by age. Those provisional data suggest that younger voters turned out at significantly higher rates this cycle than they typically do… which means they hit the polls at the same rates as older voters did in 2021 (but still significantly less than older voters did in 2025). Voters over 40 also turned out at higher levels, but the increases among younger voters were far more dramatic, and so younger ended up comprising a significantly higher share of the electorate this cycle compared to last (and a slightly higher share of the electorate than SSRS predicted).
If we use these data as the baselines for how the electorate broke down by age but assume that the SSRS exit poll accurately captures vote preferences within each age bloc, the picture of Mamdani’s base shifts a little – but not by much!
On this model, voters under 30 would still comprise only a quarter of Mamdani’s vote total. It’d still be the case that roughly half of all votes cast for Zohran Mamdani came from voters 40 years or older (see the final column of the chart below).5 The broad picture remains: the age polarization narrative is oversold.
In fact, Andrew Cuomo would have still received less votes than Mamdani in a world where turnout by age looked like it did in 2021 (skewing much more heavily towards older voters, holding constant registrations and vote preferences within each age bloc). Under that scenario, the mayor-elect would have ended up with 46 percent of the vote to Cuomo’s 45, with Sliwa doing a little better, at 9 percent.6
Put simply, youth turnout this cycle is not the reason Mamdani won (you get the same outcome if you reverse it back to 2021 levels — a year with record low turnout overall). In terms of party preference, there is no reason to believe youth vote choice this cycle was meaningfully different from the last. There’s simply not much to see by looking at the votes of young people.
If you want to understand what made this cycle different from the last, and why it ultimately shook out how it did, you need to look at voters over 40 instead.
The Culture War That Wasn’t
The 2025 NYC municipal elections actually were historic in the sense that they led to the election of New York’s first Muslim mayor (who is also the city’s first Asian mayor). Andrew Cuomo and aligned PACs and politicians leaned heavily into Mamdani’s faith and immigrant background. They portrayed his views, policies and values as foreign to “real” New Yorkers. They ran ads darkening Mamdani’s features. They drew spurious associations between Mamdani and 9/11. They implied that Mamdani would pursue “Sharia Law” in New York. They sought to conflate Mamdani’s criticism of Israel with antisemitism. They argued that his election would make Jewish voters and other New Yorkers unsafe, and so on.
But for all that, did Mamdani’s background actually matter in shaping the outcome of the race? Were culture war issues actually central to voters’ decision-making this cycle? Probably not.
Despite aggressive emphasis on Middle East politics by Cuomo et al., Mamdani largely declined to cooperate in making it an “issue.” This was sensible, of course, because the mayor of New York does not set American Middle East policy, so a mayoral candidate’s position on these issues isn’t particularly germane to the role they’re trying to fill. That’s how voters seemed to see it too.
According to exit polls, the overwhelming majority of voters saw Middle East politics as a trivial or completely irrelevant factor in their decisions about who to vote for. Only 38 percent of respondents viewed this as “major” issue – and those folks’ votes were evenly split between Mamdani and Cuomo, cancelling each other out. As a consequence, the election was ultimately shaped by people who didn’t care much or at all about this issue (for the purpose of deciding who should by NYC’s mayor). And the less people cared about this issue, the more they supported Mamdani (and the less they supported Cuomo).
In short: Mamdani’s Muslim identity and his positions on foreign policy were ultimately tangential to the electoral outcome. Cuomo and his allies made their best attempt to stir up a culture war. They failed.
Since the election, many on social media have tried to write a culture war into the race anyway by appealing to graphs like this:
Right-wing and anti-woke culture warriors interpret this graph as showing that Mamdani won because of immigrants (with foreign values and politics) paired with rootless and transient “anywheres” (who are pulled in and out of NYC for work but have no real connection to the city and its culture, and who aren’t invested in its long-term future). Mamdani’s alleged base is then contrasted with “real” New Yorkers — on this account, people who were born in the Big Apple – a constituency that favored Cuomo.
There are many obvious problems with this read of the chart.
According to this framework, a 50 year old who moved to New York as a kid because of his parents’ job, but who has lived here for 40 years, raised a family here, and plans to die here – that person would be a rootless “anywhere” and not a “real” New Yorker. That seems strange. Meanwhile, a 19 year old who plans on moving somewhere else once he’s able would be a “real” New Yorker, despite having lived in the city for half as long as the first person and despite explicitly planning to leave. This also seems strange.
A second obvious problem is that most of the New York City electorate, in fact, was not born in the city. It seems odd to say the decisive majority group does not reflect what the city is actually “about” while simultaneously asserting that an unrepresentative minority population does.
But for the sake of argument, let’s just bracket all of this. The biggest problem with the culture war narrative is that folks circulating the graph above clearly didn’t pay attention to the population base rates that are included in the image (on the left hand column). When you look at how much of the electorate each group represents in addition to looking at vote share, it’s evident that folks who lived in NYC for less than five years were not decisive for this election. You could give 100 percent of their votes to Cuomo and he’d still lose.
With some basic math, you can also see that, in fact, significantly more Mamdani voters were people who lived in New York their whole life (34.8 percent of all Mamdani voters) than folks who lived in the city for 10 years or less (who collectively comprised 26.2 percent of Mamdani voters). All said, people who lived here for at least a decade (to include folks who lived here their whole lives) delivered 73 percent of all Mamdani’s votes.
The mayor-elect did enjoy strong support from new arrivals (who are, themselves, a key part of NYC’s dynamism and history), but their support was not dispositive for the race. In fact, we can see from the exit poll data (by attending to vote share and electoral base rates) that even if NYC mayoral elections were restricted exclusively to folks who lived in the city for a decade or longer (both 10+ year residents and lifelong New Yorkers), Andrew Cuomo would still have lost the popular vote! In a world where we only counted folks who lived here for 10+ years and/or their entire lives, Sliwa would end up with 8 percent of the vote, Cuomo with 45 percent of the vote, and Mamdani would have 47 percent of the vote.
Another way of saying this is that, in fact, Mamdani’s core base of support were folks who have lived in the city for a long time. His votes came primarily from folks who’ve built a life here and are, in fact, invested in the city and its future. On top of this, he did enjoy strong support from new arrivals – but they were icing on the cake. When we pay attention to all the data provided on these graphics, the widely-circulated exit polls actually show that Mamdani would have won even with the pool restricted to folks who’ve lived here for a long time.
With respect to religion, a similar story holds: popular culture war narratives are unsupported by basic facts.
For instance, many focused on the “other” and “none” categories of exit poll charts to suggest that Mamdani won primarily because of New Yorkers who fall outside the “Judeo-Christian” tradition.
But in fact, there is zero reason to believe that Mamdani’s margins with these voters are atypical. Again, in 2021, only 28 percent of voters cast ballots for Republican Curtis Sliwa. Were atheists, agnostics and non-Christian faiths overrepresented among Sliwa supporters? Likely not. Instead, these voters likely supported the Democratic nominee in 2021 at roughly the same level they supported Mamdani in 2025.
For the last mayoral race we have exit poll data for, Bill DeBlasio got a higher share of non-Christian and non-Jewish voters than Zohran did, as the graph below will illustrate.
Put simply, Mamdani’s vote shares with “others” and “nones” were completely typical for Democratic mayoral candidates. It’s his support with Jews and Christians that deserves additional scrutiny. So, let’s do that:
At the national level, Jewish voters are a reliable Democratic voting bloc. In New York City, however, they’re a swing constituency. In 2013, when DeBlasio delivered landslide wins with all other religious blocs (and in the election overall), he won only a slim majority of Jewish voters.
In 2009, Middle East politics, antisemitism and related topics were not a big part of the mayoral discourse. Democratic nominee Bill Thompson was not a Muslim. Thompson enjoyed support from Chuck Schumer, Andrew Cuomo, and lots of other New York political mainstays. Nonetheless, he ended up with less than a quarter of the Jewish vote. He performed significantly worse than Mamdani with Jewish voters.
There were no exit polls for the 2021 mayoral race, but reporting suggests that Eric Adams didn’t rack up huge support from Jewish voters either. As the Times of Israel observed at the time, “while Adams had been endorsed by the Satmar Hasidic group in June, which may have helped him come out ahead in the summer’s crowded mayoral primary, he was not the favorite in Borough Park on election night. In the 48th Assembly District, a district represented by an Orthodox assemblyman and which encompasses much of Borough Park, 54.9% of voters chose Republican Curtis Sliwa, with only 39.2% voting for Adams.”
You’d be hard pressed to find a more aggressive cheerleader for Israel than Eric Adams. It didn’t necessarily translate into a strong showing at the ballot box among Jewish voters. He seems to have outperformed Mamdani – but not by as much as most seem to be assuming. And Zohran’s performance with Jewish voters was hardly unprecedented. Again, other Democratic candidates have fared significantly worse with these constituents when none of this culture war stuff was even part of the conversation.
One thing many outside observers fail to understand is that there are lots of conservative and Republican Jewish New Yorkers. And at the national level, before October 7, Jewish voters writ large had been shifting towards the GOP. Many seem to be overinterpreting the 2025 exit poll data because they misunderstand how Jewish New Yorkers typically vote.
In fact, the most interesting story in the chart on religious vote share — the lines that are genuinely unusual relative to other cycles — are the Christian votes.
Nationally, Democrats don’t do well among Protestants and Catholics. However, in New York City, they typically do. This is because huge shares of NYC Protestants are black or Hispanic (and the black church is, in fact, a key driver of Democratic support). Likewise, a huge share of NYC Catholics are Hispanics and immigrants from other parts of the world. Consequently, Democrats tend to do significantly better with Catholics and Protestants in New York City than elsewhere. This was not the case in the 2025 mayoral race.
Here are Kamala Harris’ religion vote share stats:
You can see that the Democratic mayoral nominee did outperform his party’s 2024 presidential nominee among Protestants – largely owing to the heavy concentration of black and Hispanic Christians in the city. However, he severely underperformed Bill DeBlasio and Bill Thompson with these constituents. And among Catholics, there is no overperformance at all: Mamdani didn’t just underperform other Democratic mayoral nominees, he even managed to underperform Kamala Harris.
However, the trend lines we see among Catholic voters may have more to do with his opposition than with Mamdani himself. Andrew Cuomo is famously Catholic. Although he’s had occasional tensions with the church over some gender and sexuality issues and New York’s COVID-19 policy, many Catholic voters are inclined to support one of their own over non-Catholics. Notice in the exit polls that Curtis Sliwa, who is also Catholic, performs much better with his co-religionists than with any other faith bloc.
In sum: with respect to religion, there isn’t anything particularly unusual in the Jewish, “other,” or “none” vote. The genuinely unusual line, the voters who behaved differently than they typically do, were Christians. Mamdani’s weakness with Catholic voters can be partially chalked up to the fact that he was running against two Catholics. His weakness with Protestants is more unique among NYC mayoral candidates, although he did outperform Harris with Protestant voters (as Democratic NYC mayoral nominees typically do).
The big story looking at all the culture war angles is that there was not, in fact, a major culture war evident in the political coalitions or final outcome of the race.
What did happen, then?
Over the course of this essay, I’ve shown that prominent culture war narratives about the role of class, age, gender, sexuality, race and religion are bunk. Most prominent culture war claims rely on cherry picking data and interpreting charts in lazy or statistically-illiterate ways.
Here a reader may wonder: if Mamdani didn’t win the race because of any unusual culture war dynamics, then why did he win? The most direct answer is: he won because he was the Democratic nominee.7 Mamdani’s win was decisive but hardly extraordinary. The coalition that propelled him to victory was broadly consistent with other Democratic candidates. There is far less here than meets the eye.
Of course, part of why the race was so normal, in the end, was because Mamdani wisely refused to let it turn into a circus. He ran a popularist campaign that was focused tightly on bread-and-butter issues like housing, public transportation and affordability.8 He did not let activist groups or staffers pull him into taking unpopular stands on niche issues. He did not let his opponents make culture war issues central to the election either. Many tried to derail the campaign, to make it about something other than Mamdani’s preferred issues, but he largely refused to take the bait.
On his topics of interest, he was happy to engage with stakeholders wherever they were. This included multiple appearances on Fox News and online manosphere shows (which Democrats typically avoid) and tons of face time with regular New Yorkers across all five boroughs. He didn’t write anyone off or take anyone for granted. This is in stark contrast with Democrats like Joe Biden and Kamala Harris who were highly risk averse and generally declined to engage with folks who weren’t already firmly in their camp.
By following these distressingly-uncommon political best-practices, Mamdani was able to ultimately deliver a conventional result with a conventional Democratic coalition despite facing an intraparty challenger and in the face of tepid support from New York’s Democratic establishment. In a two-person race against Sliwa, he likely would have put up a landslide victory comparable to Adams in 2021. He may have even been able to approach DeBlasio’s 2013 performance.
Of course, in addition to the fact that Mamdani ran a good campaign, it also worked to his favor that this third-party challenger was… not great. Elections are always “compared to what?” In this case, from the primaries through the general election, the main alternatives to Mamdani were:
A deeply unpopular current mayor who engaged in quid pro quo with Donald Trump in order to avoid accountability for corruption charges (Eric Adams).
A deeply unpopular former governor who resigned in disgrace after getting #MeToo’d, mismanaging COVID-19, and in the wake of corruption scandals and who, on top of all this, was endorsed in this race by Donald Trump (Andrew Cuomo).
A likeable but off-beat Republican who garnered less than a third of the vote in the previous election (Curtis Sliwa)
In the two-candidate race that functionally emerged between Cuomo and Mamdani, the former governor and his allies tried their best to make Zohran’s plans sound unrealistic or undesirable (and some of them, in my personal view, are not ideal9 ). Yet, they overlooked the basic reality that you can’t beat something with nothing. Cuomo himself had absolutely no plans. He had no concrete alternative proposals to mitigate practical challenges New Yorkers face; he was solely focused on the culture wars and demonizing his opponent. He had no aspirational vision or positive energy. Instead, everything about his campaign was deeply unpleasant and geared towards evoking negative emotions – fear, resentment, hatred. He seemed to have no particular goal other than blocking Mamdani and taking the office for himself. It was hard to discern any motive for his run other than a hunger for more power, status and money. It’s not hard to figure out why New Yorkers decided to vote for the other guy instead.
If Mamdani had faced an intraparty challenger who had less baggage, something to actually offer voters, and/or who wasn’t in bed with Donald Trump, he could have been in a really tough spot. But as things stand, although Cuomo was able to poach some Democratic voters at the margins, his primary base ended up being Sliwa voters. As a result, Mamdani was able to ultimately deliver a pretty conventional win with a pretty conventional coalition.
Given how idiosyncratic New York’s demographics and politics are, there isn’t much to take from the electoral outcome, in my view, except that popularism works. Focus on issues that voters care about. Take popular positions on those issues. Convey your position on those issues in a concise, accessible and compelling way. Avoid taking unpopular stands on divisive issues. Be disciplined in your messaging (don’t be pulled off-course by allies or opponents). Pound the pavement to engage with a broad swath of prospective voters (you can only make the shots you take). Have a positive vision rather than relying purely on negative partisanship. Be likeable.
The secret to Mamdani’s success is that there was no secret. He won because he well-executed the obvious stuff that politicians across the spectrum regularly decline to implement.
Trump Meets Mamdani (11/22/2025 Update)
Mamdani’s opponents tried hard to make the mayoral race a culture war during the primaries and general election. They failed to get New Yorkers to buy into that narrative. And now, it looks like they’re going to fail nationally as well.
According to polling, Mamdani is one of the most popular (if polarizing) figures in American politics today. Today, New York’s mayor-elect Mamdani and President Trump met at the White House. The press conference afterwards was… genuinely remarkable.
When Trump was asked whether he agreed with Elise Stefanik that Mamdani was a jihadist, the president declared, “No, I don’t. I met with a man who’s a very rational person.” He then went on to defend Mamdani in the face of a “defund the police” accusation offered by a right-leaning reporter.
When asked if, as a billionaire and former New Yorker, would Trump feel comfortable living in NYC during Mamdani’s administration, Trump answered “Yeah. I would. I really would, ESPECIALLY after the meeting. We agree on a lot more than I would’ve thought. I want him to do a good job - and we’ll help him do a good job… I would feel very, very comfortable being in New York.”
Trump continued, “I feel confident he can do a good job. I think he’s going to surprise some conservative people actually. And the very liberal people, he won’t surprise them, because they already like him.”
The president stressed that both he and Mamdani agree on a bunch of stuff — from building more houses, to lowering the price of goods and utilities, to prioritizing American interests over foreign powers, and more. He stressed how many Bernie Sanders voters cast ballots for him in previous elections after Sanders was pushed out of the race. He emphasized that both he and Mamdani share a commitment to seeing NYC thrive. He even went so far as to declare, “The better [Zohran] does, the happier I am.”
Critically, Mamdani didn’t soften any of his positions. When asked if he believed Trump was a fascist, Mamdani didn’t walk back that sentiment. He didn’t offer false flattery. He didn’t present himself as a sycophant. He didn’t dodge difficult questions.
For instance, when asked about Israel, Mamdani described its actions in Palestine as genocide and emphasized that the U.S. should not be funding it, and should enforce international laws, rules and norms. Not only did these answers fail to alienate Trump, the president actually offered to have Mamdani come back to the White House when he hosts the Lebanese president in order to help broker peace in the Middle East.
The lesson: getting cooperation and building bridges across difference doesn’t require anyone to be a squishy moderate, to be spineless, or to be fake. It requires pragmatism. It requires an openness to talk, and to look for zones of agreement, and to deal with tradeoffs and occasionally compromise. But you absolutely don’t have to do the pathetic nonsense that Democrats typically do.
Having failed to get this memo yet, even as Mamdani headed to Washington to successfully build a line of cooperation between his administration and the federal government, the House of Representatives passed a resolution condemning socialism — endorsed by most Democrats, to include Hakeem Jeffries. Ladies and gentlemen, the “opposition” party. The kind of politics represented in this vote — rejected by Trump and Mamdani alike in this visit — is precisely why the political establishment is so widely detested.
In anticipation of today’s meeting, many in left and right aligned media alike were gearing up for a culture war. That is not what they got. This is a testament to both Trump and Mamdani’s pragmatism, and a full-circle moment for for the complete failure of the culture wars to define Mamdani up to now.
Appendix: Picking Bones, Not Cherries
In order to get a good understanding of what happened in the 2025 NYC mayoral race (or the 2024 U.S. Presidential Election, or any other big social phenomenon) it’s important to look at different types of evidence – to contextualize findings from one type of evidence in terms of other evidence – creating a holistic picture of the outcome and its drivers. I’m going to walk through this at length because it’s important and not commonly done.
One tool for understanding this race is to look at precinct voting data: which areas voted for whom. We can then triangulate how precincts voted with the demographics of those precincts (leveraging the American Community Survey, etc.) to get a sense of which voter blocs supported which candidates.
But, of course, relying purely on precinct data can lead to a version of the “ecological fallacy.” For instance, just because a precinct with lots of African Americans voted for a candidate doesn’t mean that candidate has strong support with black people. It could be that whites who live in communities with lots of black people are especially likely to vote, and to vote for a particular candidate, while black voters in those communities are more ambivalent about voting and vote choice. This would lead to a situation where a candidate could have strong support in black precincts, but the effect is driven primarily by white voters. To get around this problem, we can look to exit polls, and see if the picture they paint matches up with the image we see from the precinct data.
Exit polls, on their own, are also a very limited tool. All said, roughly 2 million people voted across multiple modalities (early voting, in-person voting), spread across 4,338 electoral precincts. SSRS (the company that produced the exit poll used by most mainstream media organizations) collected data from 4744 respondents. While this is a large sample, it’s also equivalent to talking to only one person in each precinct and having them stand in for everyone else who voted there. In order to try to make the data representative, they have to do a bunch of statistical weighting based on previous elections, early voting data, other polling, and other factors. The modeling decisions that end up getting made can, itself, produce distortions.
In the NYC mayoral race, the SSRS estimates a margin of error of at least 2.2 percentage points. What this means is that any difference of 1-3 percentage points may just be statistical noise. If one candidate got 46 percent support from a given group, and another got 43 percent support, the two of them were statistically tied. Critically, this is the margin of error looking at the NYC electorate as a whole. The more fine-grained you get – i.e. looking at particular subsets of ethnic groups, or looking at how black women voted compared to white men – the noisier the data becomes.
The lesson: no one should take the exact percentages of exit polls too seriously. They mainly indicate broad differences between groups. The bigger those differences are, the more likely they can be trusted. Conversely, the more specific the group is that you’re looking at (i.e. non-BA Asian men) and the smaller the difference in preference between the candidates among these voters, the more one should think that observed differences are just noise… unless and until any observed patterns are corroborated by other data.
Critically, exit polls do not just provide the share of voters estimated to support each candidate, they also regularly include the pollsters’ estimates of how much each sampled group contributed to the final electoral outcome. We can make use of this data to get a clearer sense of how much different subgroups contributed to a candidate’s coalition. If a politician gets 90 percent support from one group who comprise 2 percent of the electorate and 55 percent support from another group that comprises 25 percent of the electorate — the latter group represents a much higher share of the candidates overall support. He has far more voters from the latter group than the former, despite the other population supporting him at far higher levels. This information is important to attend to when interpreting exit polls.
In addition to looking at exit polling and precinct data from this cycle, we can also compare how the 2025 electoral results compare to other races.
For instance, we can look at how Mamdani performed in the 2025 NYC mayoral race relative to Kamala’s performance in the 2024 U.S. presidential election. This is helpful because Kamala was a deeply unpopular candidate who lost most swing voters, so her numbers likely reflect what the absolute core of the Democratic base looks like. Therefore, by comparing Mamdani to Kamala, we can see which subsets of the national base defected to Cuomo, and where Mamdani has particular strength.
However, it’s also important to recognize that U.S. presidential elections tend to draw much larger and more heterogeneous voters than mayoral elections do. About 600k more New Yorkers showed up to vote in the 2024 presidential election as compared to the 2025 mayoral race. This is a large difference. And the distribution of who showed up and who stayed at home in each race did not very in a random or even way. Folks who are less affluent, less educated, and less white are especially likely to sit out non-presidential races. Part of the difference between Mamdani and Harris could come down to different electorates showing up to the polls. This needs to be accounted for when analyzing the race.
We can get additional perspective on this cycle by comparing the (three-person) 2025 electoral outcomes to the (2 person) NYC mayoral race of 2021. Here, though, we have the opposite problem. Turnout in the 2021 NYC mayoral race was the lowest on record, so the results would reflect only the highest-propensity voters (who are, themselves, not representative of the broader electorate).
It’s also possible to compare exit poll results from this race to data from previous cycles to help contextualize Mamdani’s victory. However, as a result of the transition to ranked-choice voting, lingering COVID disruptions, and other factors – there were no publicly-available exit polls conducted last cycle. Other cycles did have exit polling, but the results were typically not posted as comprehensive data tables.
The only two races I could find clean data for were the races in 2013 and 2009. Incidentally, however, these are good cycles to illuminate 2025:
As noted above, Bill DeBlasio was a similar candidate to Mamdani who ran a very similar campaign in 2013. However, he didn’t face an intraparty challenger. Looking at DeBlasio’s performance can therefore help us see what Mamdani’s numbers might have looked like if Cuomo wasn’t in the race.
The 2009 cycle between Thompson and Bloomberg is also an interesting comparison point. Bloomberg began his career as a Republican but governed largely as a Democrat. By 2009, he had abandoned the GOP and ran for his third term as an Independent. There was no GOP challenger, so it ended up being a two person race between an Independent and a Democratic nominee, both of whom were Democrats at bottom. The race ultimately resolved itself with the Independent candidate winning. Looking to Bloomberg’s winning coalition (built by a pseudo-Democrat in a 2-person race against the Democratic nominee) can therefore help us see where Cuomo fell short and what a winning coalition might have looked like for him.
Finally, in addition to precinct data and polling records, we can look at behavioral data like donations to understand who supported whom (and why).
Looking at these and other indicators collectively – putting them into conversation with each other -- a picture becomes clear: Mamdani mobilized a conventional Democratic coalition to achieve a conventional electoral outcome. Folks across the spectrum seem to be wildly over interpreting the race (and its implications) in order to write culture war narratives into the outcome that are simply unsupported by the evidence.
Michael Bloomberg began his career as a Republican. However, he built an empire out of the symbolic industries and largely shared the values of mainstream symbolic capitalists. Some of his most notable mayoral achievements include significantly raising property taxes in order to fund expanded social safety nets, imposing policies to discourage consumption of high-fat or high-sugar foods, banning smoking indoors, taxing the use of plastic bags, advocating for gay marriage, and pushing for environmental sustainability in NYC buildings and processes. Increasingly out of step with his own party, Bloomberg ran as an Independent in 2009. He then endorsed Barack Obama for president in 2012 and ran for the Democratic presidential nomination himself in 2020. He has subsequently emerged as one of the nation’s top funders of universities, scientific research, and other symbolic institutions. This was New York’s last Republican mayor!
The same chart also shows that Mamdani decisively won districts where more than 10 percent of voters registered this year before the primaries closed. These would be voters who signed up to vote in order to weigh in on Mamdani’s primary campaign. Given that he won these precincts by 19 percentage points in the general, and ultimately won the primary in a convincing fashion, we can infer that registrations in these areas were led by voters who wanted to support (rather than block) Zohran’s nomination.
It looks like Mamdani inspired new voters for him at the primary stage and against him in the general. To see both sets of precincts, click here, go to “Filter precincts by category” and scroll down to “Voter registration.”
Precisely because low-propensity voters are especially likely to avoid non-presidential races, there was a campaign to move New York’s mayoral elections to align with the U.S. presidential cycles. However, because low-propensity voters have been increasingly drifting towards the GOP, Mamdani recently came out against this rescheduling. For all the talk about Mamdani mobilizing or maximizing the voting base, in fact, he’s shrewdly trying to curate the base to maximize Democrats’ electoral advantage. He’s not trying to “rock the vote.”
For each education category, you can multiply the share of the electorate that falls into that bloc by the share of the electorate that Cuomo won. This will give you the total share of the electorate that were, for instance, Cuomo voters who didn’t graduate high school (1.4 percent in that case: 0.03 x 0.48). Next, folks can sum these products for all categories of voters who have no postsecondary credentials (you end up with 0.1504, meaning 15 percent of the total electorate were Cuomo voters who possess no college degrees of any kind). You can do the same with the categories of voters with an AA or higher (which add up to 0.2605, meaning 26 percent of the total electorate were Cuomo voters who have some form of postsecondary degree).
As a validity check, when you add these two numbers together (which would comprise Cuomo’s support at all education levels). You get 0.4109, or 41 percent of the vote, which is exactly what Cuomo is estimated to have won in the general election.
You can see how much of Cuomo’s total votes came from voters with some form of college degree by dividing the latter by the former (in this case, that would look like 0.2605 / 0.4109, which comes out to 0.633 — indicating that 63 percent of Cuomo’s voters had a postsecondary degree. Only 37 percent had no form of postsecondary degree).
You get the totals for column 4 my multiplying the data from each row of columns 2 and 3 (converted into decimals rather than whole number percentages), and then dividing the product by 0.504 (Mamdani’s share of the total vote). You get the totals for column 6 in a similar way: multiply columns 2 and 5, and divide the product by Mamdani’s overall vote share.
We can do this, back of envelope, by multiplying the 2021 turnout rate for each age bloc against the total number of registered voters with each age bloc. This will allow us to see that the distribution of votes by age would look like if turnout rates in 2025 were the same as last cycle. On that model, we can see that overall turnout would have been much lower – 1.13 million votes rather than over two million.
With these pieces of information in hand, we can then divide the total number of estimated votes from each age bloc in our counterfactual world by the total number of estimated votes overall to determine what percentage of the total electorate each age bloc would represent. You end up with this:
Then, you simply multiply each candidates vote share in each age bloc (from the 2025 exit polls) against the share of the electorate that that age bloc would have represented. Next, sum the totals for each candidate across age blocs, and you can see the total vote share that each candidate would have received in this hypothetical world.
In a world where Cuomo won the nomination and Mamdani launched a third-party bid, Cuomo would almost certainly be the mayor-elect instead.
In developing and testing their popularist messaging, Mamdani’s campaign consulted with Blue Rose — the firm created by David Shor, one of the leading lights of popularism (who also happens to be a socialist). Popularism and leftism are often discussed as incompatible. However, the approach was developed by a socialist to help other socialists win elections (and, subsequently, use the apparati of the state to improve people’s lives and life prospects).
Consider bussing. A recent MTA survey asked New Yorkers what they wanted. Reduced or free fares ranked near the bottom of their concerns.
The fact that Mamdani is focused on public transportation is great. However, on my read, his plans are not well-calibrated to the primary concerns that riders have.
In a similar way, I suspect Mamdani’s public housing scheme will ultimately redound to the benefit of symbolic capitalists (who are great at disguising their income in order to evade taxes or even make themselves seem “poor”) while benefiting ordinary New Yorkers far less than expected. This is a recurring problem with NYC affordable housing schemes: although intended to help the genuinely marginalized and disadvantaged, it’s often white-collar professionals who end up as the primary beneficiaries (see pp. 252-253 or pp. 258-260 of my book for examples of this dynamic in action for other spheres).
Creating a convoluted housing system that requires a lot of attention, effort, discipline, paperwork, institutional knowledge, and connections in order to secure one of these new affordable units – this is a scenario that typically works to the benefit of symbolic capitalists and other elites, largely to the exclusion (or at the expense) of less-advantaged people. The goal, in my view, should be to make the housing system in New York simpler rather than creating new carve outs and administered programs, etc. Again, it’s good that he’s focused on affordability and housing, but I strongly suspect that, if implemented, the primary beneficiaries of many of his proposed housing policies will not be the folks in whose name those policies are justified. The target is good IMO. The specific policies he’s advocated for, however, I’m skeptical of.
I wonder if Rob Henderson would walk back some of his recent comments if he read your piece.