House · 2026

New York’s 21st District

Blake Gendebien (D)v.Anthony Constantino (R)LIKELY R

Open seat, currently Republican.

Projected margin

R+6.2

Nine in ten outcomes: D+2 to R+15

Chance of winning

88%

Anthony Constantino

Polls alone

R+6.2

5 polls from 4 pollsters; latest Sep 10. Weight in projection: 84%.

Fundamentals alone

R+5.8

Weight in projection: 17%.

Polls

The average over time

Each poll is first shifted by its sponsor's and its pollster's measured lean, then weighted by recency and track record. Sample size gets no weight: in 2,000-plus past races it predicted nothing about accuracy. Why.

adjusted weighted average shading: which side the average favors each poll, as published party-aligned pollster or sponsor

Before any poll

What the fundamentals say

A regression on past House races, fit without looking at this year. The national environment term uses the generic ballot after its historical correction (D+5.3).

What each term adds to the estimate

BaselineD+0.8

Partisan leanR+16.3

National environmentD+3.6

Incumbency0.0

Incumbent's last over/under-performance0.0

Fundraising gapD+6.1

Fundamentals estimateR+5.8

Input: partisan lean

R+19.2

Presidential vote relative to the nation, 2024 on this year's lines.

Input: raised from itemized individual donors, through June 30

$2.5M D · — R

FEC filings; the model uses the log of the ratio.

Handicappers, for comparison

Cook: Safe R · Sabato: Likely R · Inside Elections: Safe R

Not an input to the projection.

Table view

All 5 polls in the average

“Share” is how much of today’s average each poll supplies. It depends on recency, the pollster’s record, whether the poll screens for likely voters, and how many other polls the same firm has in the race. It does not depend on sample size.

Dates	Pollster	Sample	Gendebien	Constantino	As published	Adjusted	Share of average
Sep 9–10	Public Policy PollingDEM.-ALIGNED for The Bench	557 RV	43	48	R+5	R+8.5 moved 3.5 toward R	21%	source
Aug 30 – Sep 2	Impact ResearchDEM.-ALIGNED for Gendebien’s campaign	500 LV	44	47	R+3	R+6.7 moved 3.7 toward R	30%	source
Aug 15–17	J.L. Partners	500 LV	40	53	R+13	R+12.5 moved 0.5 toward D	23%	source
Aug 4–8	SurveyUSA for WNYT-TV Albany	612 LV	43	39	D+4	D+3.1 moved 0.9 toward R	21%	source
May 26–31	Impact ResearchDEM.-ALIGNED for Gendebien’s campaign	500 LV	44	45	R+1	R+4.7 moved 3.7 toward R	4.8%	source

LV: likely voters. RV: registered voters. A: all adults. “Adjusted” is the published margin after removing the sponsor shift and the pollster’s measured house effect; the small print says how far that moved the poll. Dem.-aligned and Rep.-aligned mark polls paid for by a party, a campaign, or an allied group, or taken by a firm whose published work is mostly for one side. They are shifted by the average lean of such polls, then counted at full weight.