The DSA's big city takeover

...and why only the very rich and very poor are immune from it. Plus: prospects for the DSA going national, Ossoff might be too scripted, quants vs. handicappers, the red-blue fertility gap

No. 406 | June 26, 2026

🇦🇴 The DSA is now the big city default

The Democratic Socialists of America’s victories in Congressional primaries feel seismic because they come on the heels of a string of major victories in big cities, including a DSA-backed candidate winning the Washington, D.C. mayoral primary, a DSA candidate making it to the runoff in Los Angeles, and Zohran Mamdani’s victory in New York City. By the end of the year, the mayoralty of the country’s three largest cities may be in the hands of hard-left progressives.

If you’re thinking this is a workers’ revolution — not quite yet. Support for these candidates came from wealthier, better educated — and whiter — precincts. Even the race the DSA was not supposed to win — Darializa Avila Chevalier’s win against Adriano Espaillat, came in a Harlem-based precinct that’s seen a lot of gentrification in its Manhattan-based parts. Meanwhile, moderate Ritchie Torres saw no real challenge in a South Bronx district that’s still the poorest in the nation.

Another area where you did not see a comparable DSA effort was the actually-wealthy 12th district, where Upper West Side liberal Micah Lasher prevailed. According to figures compiled by Kyle Tharp, the results here were inversely proportional to the fame (as measured by social media followings) of each candidate:

The sweet spot for socialism are downwardly mobile highly educated young voters — the junior copyeditor living in a studio apartment whose parents were corporate lawyers living in the suburbs. And largely missing from the party are the urban working class, who are dwindling in number.

Play this out, and it’s to see mainstream Democrats hanging on in cities anywhere, at least those which are magnets for college-educated young people. The actually rich voters like those in the 12th are limited in number while the working class are increasingly priced out of their old neighborhoods and moving to the suburbs. Who else is left? Basically the DSA’s target voter.

Charles Franklin also breaks down DSA favorability, which is at 20-48 nationally, but 40-20 among Democrats. But it’s not limited to big cities: all sorts of Democrats, including those who live in rural areas, view the DSA favorably.

Could far-left politics go national in 2028? The Michigan Senate primary will be a big test of this, as I wrote on Wednesday. You also saw a progressive win in a Trump district, Maine’s 2nd, where it would have behooved Democrats to nominate a more mainstream alternative. Substantively, most Democratic primary voters appear to want socialism.

🔵 Checking in on the normie liberals

Meanwhile, normie liberals like Gavin Newsom appear to be walking a tighter line. Once surging in polls and prediction markets, he has since fallen back. His comeback strategy? Tethering himself to Joe Biden and an unpopular Democratic establishment. He’s also tried to negotiate with the unions to get a state billionaire tax off of California’s ballot this fall. As it’s become apparent that’ll be a hard sell for this primary electorate, today’s he’s come out for a national tax.

One Democrat who does seem to be making a name for himself is Jon Ossoff, whose viral rally clips have catapulted him to the top of the 2028 buzz. But something about it feels off-key, writes Kyle Tharp, a perfectly stage-managed 2012 production out of place in 2028:

To play devil’s advocate, however, the problem with this strategy is that the Perfectly-Optimized, Almost-Robotic Ossoff hasn’t proven he can turn the switch off. He almost never seems like he’s speaking off the cuff, and he rarely displays any casual human banter on camera. Like Martin noted, he “speaks in paragraphs,” and every clip seemingly lands because his team engineered it to land, which is exactly the issue. When a politician’s best moments all arrive pre-planned, voters and the reporters covering them will start to wonder what he sounds like when the advance team isn’t carefully picking the camera angle. Eventually, people will get bored.

Democrats seem to care about ~authenticity~ far more in 2026 than they did during the Obama years or even when Ossoff was first elected. The politicians often breaking through in the second Trump era do so, in part, by seeming unfiltered. Trump joked and rambled for hours on Rogan, Flagrant, and Bussin’ with the Boys and his supporters loved him more for it. Zohran Mamdani built a following on selfie-cam monologues and creator collabs, many of which look like he filmed them between subway stops. Gavin Newsom is hosting his own hour-long podcast every week, and AOC has long been the master of the art of informal Instagram live streams. Younger Democrats may now spend their time watching people like Hasan Piker reacting live for six hours straight, and older Democrats are watching candidates shoot the shit with Tim Miller on the Bulwark (When Ossoff recently appeared with Miller, he seemed like The Bulwark host was giving him a root canal).

That’s a perfect example of the “Joe Rogan CEO” test from a16z highlighted in Nathan Brand’s must-read media newsletter. Are you authentic and interesting enough to spend three hours shooting the breeze with Rogan? Kamala wasn’t—and probably knew it. Which 2028 Democrats will be?

🗳️ Quants vs. handicappers

The Argument’s Lakshya Jain has been critical of the Cook Political Report, whose ratings currently forecast a jump ball for the House despite evidence of a pro-Democratic environment.

That’s a criticism I’m sure I’ve also made in Republican wave years, one that speaks to a fundamental difference between the job of the quantitative forecaster and the traditional handicapper. Cook’s method is rooted in study of individual districts—including interviews with candidates, reviewing internal polling, and an overall assessment of how the district might behave differently compared to national trends. These bespoke reviews mean that they only tend to move a handful of districts at once, when at least in theory a shift in the national environment should cause all districts to move at once. It’s only late in the cycle in a wave election when Cook and the traditional handicappers “catch up” to where national polls imply the election should be at, something documented recently in a recent Ethan Chen Substack post.

📰 Trust in media has declined

In their annual report on trust in media, YouGov finds a landscape of mostly declining trust — with the biggest shift coming among recently rebranded or overhauled outliers like CBS and MSNOW/MSNBC.

👫 The red-blue fertility gap

New data confirms a meaningful fertility gap between red and blue counties — with a hypothetical Trump +100 county at replacement level and a hypothetical Harris +100 county at 1.2.