← Election Series · All house races

Our verdict

LiberalMind endorses Anthony Constantino — as unproven, and on watch

Election Day is Tuesday, November 3, 2026, and voting is already underway in New York. This is the whole race — the records, the claims against each candidate, and what the sources actually show. Read the research, and decide which flaw you can live with.

The setup

New York's 21st District — the North Country — is an open seat because Rep. Elise Stefanik (R) is not seeking re-election; she ran for governor and later withdrew (North Country Public Radio, Oct 9, 2026; Wikipedia, "2026 United States House of Representatives elections in New York").

Start with the district's fundamentals, because everything strange in this race is measured against them: Donald Trump won this district by roughly 20 to 21 points in 2024 (Bloomberg Government, Oct 14, 2026: 21 points; Washington Reporter, Oct 2026: 20 — printed as a range because the two reports differ by a point). Its PVI is R+10 (Wikipedia 2026 ratings table). This is deep-red paper.

The Democratic nominee is Blake Gendebien, a dairy farmer from Lisbon in St. Lawrence County, who won the June 23 primary with 64.7% (15,216 votes) (PollingSource primary table). The Republican nominee is Anthony Constantino, the CEO of Sticker Mule, who won his primary with 59.3% (26,578 votes) over Assemblyman Robert Smullen's 40.2% (PollingSource; Wikipedia, above) — a primary with its own story: Smullen held the state GOP's and the Conservative Party's endorsements; Constantino won with President Trump's endorsement instead. The certified State Board of Elections list was not directly retrieved in our research pass; additional qualifiers are unconfirmed, and Constantino's "Taxpayer Rights" line is part of his confirmed ballot presence (Wikipedia, above).

The ratings say Lean Republican — with one compilation conflict we print rather than resolve. Cook Political Report: Lean R (Oct 23, 2026). Inside Elections: Lean R (Oct 15, 2026). Sabato's Crystal Ball: PollingSource's expert block prints Likely R; Wikipedia's ratings table prints Lean R. Both compilations cannot be right about Sabato; we print both and resolve neither without Sabato's own page.

The polling says something the fundamentals say should be impossible, and because the polls conflict with each other less than they conflict with the district, we print all three: PollingSource's average is Constantino +0.3, with PollingSource's own model rating the race a Toss-up against the experts' Lean R. Siena Research Institute (Sep 30–Oct 6, ~500 likely voters, ±4.6) found Gendebien 46, Constantino 42 — and, the number that should make every Republican in the North Country sit up, 21% of Republicans saying they would vote for Gendebien, even as district voters preferred Republican control of Congress 50–45 (NCPR, Oct 9, 2026). Two Democratic-sponsored polls found Constantino ahead: Public Policy Polling (Oct 13): Constantino 46, Gendebien 44; Impact Research (Oct 4): Constantino 48, Gendebien 45. Constantino's campaign (senior advisor Alex DeGrasse) has dismissed the Siena poll, claiming Siena "included a write-in, conservative candidate, but did not include a far-left write-in candidate, intentionally skewering results" (NCPR, above). That is the campaign's characterization, printed as a campaign claim — Siena's published crosstabs are the check on it, not the quote.

The money is enormous on both sides and in conflict across compilations; we print one dated set and flag the rest. Gendebien has raised $7,655,529 with $364,012 in cash on hand (PollingSource compilation, read Oct 25, 2026); a second compilation puts him at just under $7.2 million raised and about $470,000 on hand through September 30 (Northeast Independent, citing PaidForPolitics). Constantino sits at roughly $14.4 million in total contributions and $4.1 million on hand through September 30 (Northeast Independent via PaidForPolitics) — a figure swollen by heavy self-funding, a fact stated plainly because it explains his campaign's scale: it is substantially his own money. (An earlier figure on Wikipedia's page — $7,453,187 raised — comes from a primary table that appears mislabeled by date; we do not print it beside the September figures.) Outside groups have spent $2.45 million total (PollingSource), with PaidForPolitics counting about $1.64 million in general-election outside spending supporting Gendebien or opposing Constantino through October 20, against about $40,000 opposing Gendebien (Northeast Independent, above).

Blake Gendebien (Democratic)

The candidate and his record: Gendebien is a dairy farmer from Lisbon, St. Lawrence County (NCPR, Oct 13, 2026). He has been running for this seat for nearly two years — he entered when Stefanik was nominated as UN ambassador, stayed in after that nomination was withdrawn, and stayed again after Stefanik announced her retirement (NCPR, above). Persistence is documented. Public office is not: no office record was located in our research, and he is unproven — a watch-closely posture, not a demerit.

His pitch is moderate and cost-focused in his own words — "fight to lower our costs… save our rural health care" (NCPR interview, Oct 13, 2026) — alongside an explicitly anti-administration rationale: "We need to pump the brakes on this [Trump] administration's really strange policies, and that is why I'm running" (quoted by the Washington Reporter, a conservative outlet — its "far-left stances" characterization of Gendebien is that outlet's framing, not an established fact; the direct quotes are the usable record). Asked about impeaching Trump or cabinet secretaries, he said he wants to "hold people accountable for mistakes and incompetence, but not at the expense of getting things done" (same source). And the Siena crosstab is real: one in five Republicans in this Trump+20 district told a nonpartisan pollster they'd vote for him.

Steelman him at full strength, because the file supports it: a farmer, a neighbor, two years of showing up, a cost-and-rural-health-care message aimed at voters his party usually lectures, and measurable crossover into the other party's electorate. That is the profile of a Democrat built for exactly this district.

Pros: A biography rooted in the district's actual economy; a moderate, cost-centered pitch; documented Republican crossover (21% in Siena) that no other Democrat in this batch can claim; $7.65 million raised — his campaign is fully real.

Cons: None of it is a break in service toward our values, because there is no service — and the breaks he describes run the other way: he is running, by his own statement, to brake the administration this publication supports. A candidacy-position moderation from a man who has never held office is a promise, and his party's program is what a Democratic vote organizes. The Manchin exception is earned in office. He has never been in one.

The flaw, plainly: Gendebien's flaw is that his moderation is unproven and his purpose is opposition. He asks a Trump+20 district to send a Democrat to Washington to slow down the president the district chose by twenty points — politely, moderately, and in the other party's caucus.

Anthony Constantino (Republican)

The candidate and his record: Constantino is the CEO of Sticker Mule, the printing company in Amsterdam, New York, and a former boxer (NCPR, Oct 9, 2026; Bloomberg Government, Oct 14, 2026). He has never held public office — unproven, watch-closely, not a demerit. He beat the state party's endorsed candidate for the nomination on the strength of President Trump's endorsement, which is a documented instance of the president's word outweighing the state organization in a Republican primary.

His campaign's scale is a function of his own wealth — roughly $14.4 million in total contributions through September, heavily self-funded (Northeast Independent via PaidForPolitics). That buys saturation advertising without traditional fundraising, and it also means his candidacy has never had to pass the market test of other people's money.

One incident, printed with its caveat at full strength, because printing it any other way would be a smear: a man, Josef Marris, was charged by the Essex County Sheriff's Office with two counts of criminal possession of stolen property after allegedly being caught on video taking Constantino campaign signs; Constantino's campaign alleged the thefts were carried out on Gendebien's behalf, and Constantino claims more than $100,000 in losses to sign theft and vandalism (American Almanac, relaying Fox News Digital, Oct 2026). The charge against Marris is a matter of record per the sheriff's office. The campaign's attribution of the thefts to Gendebien is unproven — the reporting itself states that "no charging document in the available reporting ties Gendebien or his campaign staff to the alleged thefts" — and this article prints the allegation only to label it that. It earns Constantino sympathy for vandalized signs. It earns Gendebien nothing worse than being accused by an opponent, which is not evidence.

Pros: The conservative direction in a district that votes conservative by twenty points; the president's endorsement, won against the state machine; resources that guarantee his message is heard; a district baseline (Trump +20–21, PVI R+10) that does the heavy lifting of the race for him.

Cons: Unproven in the complete sense — no office, no votes, no service record — with a campaign financed substantially from his own pocket and a nonpartisan poll showing him behind a Democratic farmer in a Trump+20 district. His campaign's public response to that poll was to accuse the pollster of rigging it, a claim we print as a claim. If the Siena number is even roughly right, he is underperforming his district's fundamentals by a historic margin, and the file does not contain the record that would explain why voters should trust the fundamentals over the poll.

The flaw, plainly: Constantino's flaw is the empty file at scale: a self-funded candidacy with no service behind it, running behind in the only independent poll in a district his party carries by twenty. Voters are buying a direction and a balance sheet.

OUR CALL

LiberalMind endorses Anthony Constantino — as unproven, and on watch.

The district is the first fact: Trump by 20 to 21 points, PVI R+10. The second fact is the Democrat's own mission statement: Gendebien is a genuinely appealing moderate-profile candidate who is running, in his own words, to pump the brakes on the administration — which makes him, under our doctrine, an opposition candidate with a farmer's biography, not an exception candidate with a record. The exception is earned by breaks in service toward our values; he has no service and offers a break in the opposite direction. His 21% Republican crossover is a tribute to his candidacy and a warning to ours, and it does not change what his vote would organize.

Constantino is unproven, self-funded, and — if Siena is right — running the weakest Republican campaign this district's fundamentals have seen in a generation. All of that is printed above. What he offers is the direction, the president's endorsement earned in an open fight with his own state party, and a seat that must not be handed to the opposing caucus in a House divided this finely. In the House, where the organizing vote weighs lighter than in the Senate, we still do not give a Trump+20 district to the brakes.

We flag the live risk in our own call: if Siena's crossover is real and holds, this district is about to teach its party a lesson about nominating wealth over record, and Constantino's empty file will be the reason. He is endorsed as what he is — unproven, and watched.

This race is one of hundreds in LiberalMind's Election 2026 coverage. For the complete breakdown — every race, claim by claim — go to liberalmindhq.com.

Vote early. Research your candidates. Choose the best moderates, whatever their party — no extremism from either side. And encourage your friends and family to do the same.

Sources

Election result — November 7, 2026

Winner: Anthony Constantino (Republican) — called by the Associated Press. This result was appended on November 7, 2026 (results wave 5); the original article above is unchanged, exactly as published before Election Day.

AI-assisted article — but the sources are human, and the facts are very real.

← Back to the Election Series