Does socialism win elections in America?

The scoreboard question. The win column is real and growing; the disputes are about why wins happen and whether they travel beyond deep-blue cities.

Position A: Yes — and beyond New York

Held by: sympathetic media, DSA

• The roster: Mamdani (NYC mayor), Wilson (Seattle, label contested), a democratic socialist “all but certain” to be DC mayor, Kiros’s Colorado upset, three NY congressional primaries, results in Philadelphia and NJ suburbs (What Is Democratic Socialism?)
• Mechanism: door-knocking capacity and volunteer armies (Hasan Piker: This Is Why So Many Young People Are Turning to Socialism (Impact Theory w/ Tom Bilyeu))
• The named test: the 2026 Michigan Senate Democratic primary — a Trump-carried state

Position B: The wins are protest votes in Democratic monopolies, not socialist mandates

Held by: skeptics across lean — including Wolff

• Wolff: “The people of New York were not voting for socialism. They were voting against what they’re living through” (On Socialism: Exploring Theory & Practice with Richard Wolff)
• Front Page: the socialist label among the young is “a protest vote in search of a label” (Socialism Always Fails...)
• Wilson won by 0.73 points in one of America’s most progressive cities — a margin, not a wave (Socialism Is Destroying Seattle)
• All flagship wins so far are Democratic primaries or one-party-city generals; no purple-state general has been won

Rebuttals & Cross-fire

• A→B: primaries against establishment incumbents are the contested terrain when the general is safe; that’s how party realignment always looks
• B→A: historical base rate — Wolff himself notes ~100–150 US cities have had socialist mayors (Milwaukee, Bridgeport) without a national breakthrough following
• Governance risk (B): Seattle’s capital-flight narrative and NYC budget criticism preview the incumbency test that wins set up

Evidence Ledger

The August 2026 Update (resolved test, split verdict)

The discriminating event happened — twice — and delivered a split verdict:

• Michigan (Aug 5): Abdul El-Sayed won the purple-state Senate primary, beating $32M of AIPAC money and the entire establishment slate — but narrowly, after polls showed +13/+15 (Abdul El-Sayed wins Michigan Senate primary — a major victory for progressives). General vs. Rogers pending; earlier hypothetical polling had Rogers +10 over him.
• Wisconsin (Aug 11): Francesca Hong lost the gubernatorial primary by 0.5 points after 18–36 point polling leads (Wisconsin surprise raises big questions about polls and Democratic voters).
• The measurement problem: WaPo’s polling desk now hypothesizes progressive supporters are systematically overrepresented in polls (nonresponse bias) — meaning the movement’s polled strength, including possibly its youth-favorability numbers, is inflated. “Supporters of strong progressive candidates are a passionate minority of the Democratic Party.”
• The establishment found its play: late consolidation behind one candidate (Evers/Crowley) beat a divided-field frontrunner — the first documented establishment counter-strategy in the corpus.

Current State

Updated verdict: socialists win Democratic primaries when the opposition splits, and win big-city generals; they underperform their polls consistently, and a unified establishment can beat them even in progressive states. Scoreboard additions (Aug 2026 ingest): same-night Minnesota Senate primary went left (Flanagan beat a moderate); Detroit’s Donavan McKinney (DSA) ousted Rep. Thanedar while Cori Bush lost Missouri badly; projected seven DSA House members post-midterms; DSA-aligned mayoral wins claimed in DC, LA, and Seattle. The environment is now adversarial in a funded way — Third Way’s $15M discrediting program and the “DSA dodge” pressure are producing visible label-avoidance (El-Sayed: “not technically or practically DSA”; Will Lawrence’s lapsed membership) — so the label and the movement may decouple: candidates keep the DSA ground game and drop the name. Next discriminating events: Florida Aug 18 (Larkin vs. Moskowitz — first test after the counter-offensive was announced), then the Michigan general (Nov 2026). Record margins and turnout.

Hard numbers now in corpus (2026-08-13 fetch)

• NYC 2025 general: Mamdani 50.78% (1,114,184) vs. Cuomo 41.32% and Sliwa 7.01%; turnout 43.47%, highest since 1969 — an outright majority on record turnout, the strongest pro-side datum and the direct answer to “only wins low-turnout primaries” (2025 New York City mayoral election (results record)).
• NY-7: Valdez +20.3 on ~67k ballots (open, safe-blue — don’t generalize); NY-13: Avila Chevalier +3.3 over the CHC chair; NY-12 counter-datum: establishment held the open seat.
• The establishment’s sharpest testable claim, from Third Way’s own memo: “Since 2018, the far left has not flipped a single Republican-held congressional seat” — concedes primary wins, denies general-election expansion; still unfalsified after Michigan/Wisconsin (2028ers: Don't Get Fooled Again). Its 6%-socialist-primary-voter figure is advocacy polling without published wording — flagged.
• Wisconsin margin reconciled to 0.4 (supersedes 0.5) pending certification.

Update 2026-08-22 — Florida verdict and first Michigan general poll

• Florida Aug 18 resolved: establishment +28. Moskowitz beat DSA-endorsed Larkin ~64–36 in a Trump-won, quarter-Jewish, redrawn FL-25 (Moskowitz defeats anti-Zionist challenger in heavily Jewish South Florida district, Jared Moskowitz Defeats DSA Challenger Oliver Larkin In Florida Democratic Primary; see 2026 Florida 25th District Democratic Primary (Moskowitz vs. Larkin)). The scoreboard pattern sharpens: DSA wins deep-blue primaries, loses purple ones — now Wisconsin (statewide, −0.5) and Florida (House, −28) on one side, NYC/Denver/Detroit/Philadelphia on the other. Whether Third Way’s $15M did any of the work is unproven — no source documents its spending here; district lean and Israel are sufficient explanations.
• The label question got its first purple-seat data point: Larkin ran the open-rose campaign Bread & Roses demands (Hide the Rose Era is Over: Emphasizing Socialist Messaging for DSA's National Endorsements) and lost big; El-Sayed ran the “not technically DSA” dodge and won narrowly. One pair, confounded — but it is the pattern the dodge predicts.
• Michigan general, first poll: AARP/Fabrizio Ward–Impact Research (Aug 9–11, n=877, ±3.3): El-Sayed 51–50 (or 48–47 per The Hill’s write-up); Rogers +9 among 50+, +33 among older swing voters; Fox had Rogers +4 (AARP Poll: In Tight Michigan Senate Race, Rogers and El-Sayed Are Nearly Tied). Apply the progressive-overpolling discount from Wisconsin surprise raises big questions about polls and Democratic voters and a nominal tie reads as a modest Rogers edge. The Democratic pollster’s own read: low youth motivation is the GOP’s “silver lining.”
• Counter-signal: TIME reports membership >120k and “we’re proving it works” triumphalism from the Chicago summit — published two days after Florida and still listing it as a possible gain ('We're Proving It Works': Inside the DSA's Push to Remake the Democratic Party). Next discriminating events: Michigan general (Nov 3); LA mayoral runoff (Raman); DC/Seattle governance metrics.
• WaPo-Ipsos (July 8–13, 2026): exact wording now retrieved — “Would you consider voting for a candidate for president who identifies as a democratic socialist?” Overall 39% yes/55% no/6% skipped; 75% yes among Democrats. This measures consideration, not committed vote intention (Washington Post-Ipsos Poll July 2026 — Trend Topline). Cohan’s reply for the left: 2016 was the centrist’s disaster.

Update 2026-09-03 — Florida’s split verdict, and a margin

• FL-House margin finalized: Moskowitz 63.5–36.5 (27 points). The Hill frames it as the cycle’s “last test for a DSA congressional candidate” and as the second straight non-blue loss after Hong (Moskowitz wards off DSA-backed challenger in Florida primary for toss-up House seat, Moskowitz taunts Piker after primary win).
• Counter-datum, same night: Angie Nixon, a DSA member (not nationally endorsed), won Florida’s Democratic U.S. Senate primary over DSCC recruit Alex Vindman despite a roughly 16:1 fundraising deficit ($16M+ vs. <$1M). Margin and turnout not in corpus. Vindman never debated her and campaigned against the Republican (Democratic socialist Angie Nixon scores upset in Florida's Democratic Senate primary). The “0-for-2 in purple contests” line from Aug 22 needs restating: DSA-endorsed candidates are 0-for-2 outside deep-blue terrain (WI statewide, FL House); a DSA member is 1-for-1 in a red-state statewide primary. The Position B reply is ready-made: a primary against a non-engaging opponent is not a general, and Florida’s Democratic primary electorate is not Florida.
• Establishment behavior diverged by race: in the House race the incumbent ran the electability/AIPAC playbook and won big; in the Senate race the DSCC recruited, lost, and endorsed the socialist the same night (Schumer/Gillibrand statement). Republicans have pre-registered the general-election frame: NRSC, “Florida will never elect a socialist to the US Senate”; Moody, “staring down crazy and the socialists.”
• Ledger addition: Nixon vs. Moody, Nov 3, 2026, joins El-Sayed vs. Rogers as a purple/red-state general with the socialist label on the ballot. Third Way’s “no GOP seat flipped since 2018” claim gets two statewide tests.

Update 2026-09-04 — Florida’s official numbers, Michigan’s polling band, and a smaller nonresponse discount

• Florida Senate margin retrieved: Nixon 56.06–43.94 (705,835 to 553,138; 1,258,973 votes; +12.1), carrying every large county but Palm Beach; all-party turnout 25.3% (Florida Division of Elections — August 18, 2026 Primary Election Official Results (U.S. Senate, U.S. House District 25, turnout)). For Position A: a 12-point statewide win by an open DSA member over a $16M recruit is not a fluke margin. For Position B: 1.26M Democratic primary ballots against 1.66M Republican ones, in a state Republicans carry by double digits, describes the general-election problem before a single Moody ad. FL-25 confirmed: 63.48–36.52 on 44,631 ballots, district number officially 25; the House test was a very small electorate.
• Michigan general, four polls, four El-Sayed leads: AARP +1, MSU/YouGov +5 LV, Susquehanna (R) +7, EPIC-MRA +4 with leaners (MSU Poll Shows Democrats Lead, Republicans Coalesce Behind Rogers (State of the State Survey, Aug 10–20, 2026), Abdul el Sayed outperforms Mike Rogers in latest Michigan Senate poll (Susquehanna, Aug 11–17, 2026), EPIC-MRA Statewide Poll of Likely November 2026 Voters (Michigan, Aug 22–28, 2026): frequency report). No post-primary poll has Rogers ahead. The structural warning is consistent across houses: youth-heavy coalition, older voters and independents leaning Rogers, 32% very unfavorable on El-Sayed (EPIC), and a 37% “less likely to back a Muslim candidate” figure (MSU) that is a separate penalty from the socialism frame.
• The overpolling discount is now sized, and it is smaller than August implied: Morris’s microdata reanalysis attributes 5–10 points of the Wisconsin miss to primary-specific age targets, 6–10 to late movement, and only 2–5 to within-cell nonresponse (I reanalyzed the raw data from Wisconsin's primary polls. Here's what actually went wrong.). The first cause does not exist in a general-election poll with census benchmarks; the third does. So “apply the Wisconsin discount to Michigan” (Aug 22 update) overstated the correction: a fair haircut on a general-election lead is a few points, which still leaves El-Sayed’s +4/+5 within the error band and turns Susquehanna’s +7 into a modest edge, not a tie.
• Wrinkle for the measurement story: the Examiner relays that Median Strategies, one of the pollsters showing Hong far ahead, admitted fabricating polls (Abdul el Sayed outperforms Mike Rogers in latest Michigan Senate poll (Susquehanna, Aug 11–17, 2026)). Part of the “progressive overpolling” pattern may have been fraud rather than bias; the LA Times report is not in the corpus.
• Ledger: Florida Senate primary, +12.1, official; FL-25, −27.0 on 44,631 votes, official; Michigan general polling band, El-Sayed +1 to +7 (Aug 9–28); Wisconsin nonresponse component, 2–5 points (Morris).