Impact Research puts NY-21 within three points; our forecast stays at R+13.7

An Impact Research survey of 500 likely voters in New York's 21st District, fielded August 30 to September 2, put Republican Anthony Constantino at 47 percent and Democrat Blake Gendebien at 44 percent. On a sample that size, a three-point gap is close enough that the poll does not establish a leader.

Our forecast did not follow it. The model's mean margin in NY-21 is R+13.7, unchanged from the run at 8:32 AM ET on September 8, with a 5th-to-95th percentile range of R+29.4 to D+2.0. It gives Gendebien a 7 in 100 chance, against 8 percent in that earlier run — a change smaller than the precision the number is worth. The last time we wrote about this district, on August 20, a poll had the Republican up 10 and the forecast barely moved. A tighter number has now arrived and the forecast has again barely moved.

That gap between one poll and our average is the thing to watch rather than the one-point drift in the probability. A single 500-person survey carries limited weight in the model, and our range still runs well past R+20 on the Republican side while stopping at D+2.0 on the Democratic side. It would take more polling in this direction, not one more result, to move the district's expected margin.

Nationally, our generic-ballot average sits at D+6.2 across 24 polls over the last 45 days, and our model gives Democrats a 93 percent chance of winning a House majority, with a mean of 239.0 seats against the 218 needed. One note on that figure: our simulation draws error nationally, by census division, by state and by district, but it has no term linking demographically similar districts in different places. We lack the cluster data to add one, so our House control number is somewhat firmer than a fuller treatment would leave it.