A second South Carolina poll ties the race; our model still has it R+11

An Impact Research survey of 700 likely voters in South Carolina, fielded August 18-24, put Democrat Annie Andrews at 41 percent and the Republican at 41 percent, with 5 percent for other candidates. No margin of error is given in the poll as we have it.

Our forecast did not follow the poll. Our model gives Andrews a 16 in 100 chance against Republican incumbent Darline Graham, down from 17 percent in our previous run earlier the same evening. The mean margin moved to R+11.3 from R+11.1. The 5th-to-95th percentile range now runs from R+30.5 to D+7.6, a shade wider on the Republican side than the R+30.1 to D+8.0 we had before.

That gap between a tied poll and a double-digit projected margin is the story of this race as our model sees it, and this survey did not narrow it. The forecast blends state polling with the broader environment and with how South Carolina has voted; a single 700-voter sample showing a tie moves the estimate by roughly a point, in this case in the Republican direction, as the other inputs update alongside it.

Nationally, our average of 23 generic-ballot polls over the past 45 days stands at D+6.6. In the Senate, our model gives Democrats a 44 percent chance of winning control and Republicans 56 percent, with a mean of 49.9 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. South Carolina is one of the seats that would have to break Democratic for that to happen, and at 16 percent it is not one our model expects to. Election Day is November 3, 67 days out.