New Vermont polling leaves VT-1 where our model already had it

Two University of New Hampshire surveys of 954 likely voters in Vermont's at-large district, each fielded July 15-20, put the Democrat at 63 percent and 25 percent for the Republican in one, and 64 to 25 in the other. That is a gap of 38 to 39 points.

Our forecast did not need much revision. The model gives the Democrat a 100 in 100 chance in VT-1, unchanged from our previous run on August 17, and a mean margin of D+41.9, against D+41.8 before. The 5th-to-95th percentile range now runs from D+24.7 to D+59.0, a shift of about a tenth of a point at each end. The new numbers sit inside that range and confirm what was already there rather than changing it.

At the national level, our average of 26 generic-ballot polls over the last 45 days stands at D+6.7. Our model gives Democrats a 93 percent chance of winning a House majority, with a mean of 242.1 seats against the 218 needed for control. One caveat on that figure is ours to own: our simulation draws error nationally, by census division, by state and by district, but it has no term linking demographically similar districts across state lines, because we have no cluster data for it. That absence makes the House control number marginally firmer than a fuller model would report.

The Senate picture is closer. Our model gives Democrats a 45 percent chance of winning Senate control against 55 percent for Republicans, with a mean of 50.0 Democratic seats. Democrats need 51 seats for control; Republicans need 50, because the vice president is a Republican and a 50-50 tie goes to their side. Nothing in the Vermont polling speaks to that. With 77 days to the November 3 election, VT-1 remains among the least competitive races on our board.