Minnesota Senate poll has the race at 43-42 as our forecast drops to 78 percent

A co/efficient survey of 833 likely voters in Minnesota, fielded September 14-15, put the Democrat at 43 percent and the Republican at 42 percent, with 2 percent for an independent and 2 percent for other candidates. That one-point gap is too small to call a lead. The race, on this poll, is close.

Our model now gives the Democrat a 78 in 100 chance of winning the open Minnesota seat, with a mean margin of D+4.6 and a 5th-to-95th percentile range running from R+5.2 to D+14.5. That is a move. Our previous run, earlier the same day, had the Democrat at 92 percent with a mean margin of D+8.5 and a range of R+1.4 to D+18.3. The new poll narrowed the central estimate by about four points and widened the downside end of the distribution to include a mid-single-digit Republican win.

This is the fourth straight reaction on this race in which our Democratic win probability has come down, and it is the largest single step so far. Tina Smith, a Democrat, holds the seat and is not on the ballot.

Nationally, our model gives Democrats a 48 in 100 chance of winning Senate control, against 51 percent for Republicans, with a mean of 50.1 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 them. Our generic-ballot average stands at D+6.8 across 24 polls over the past 45 days. In the House, where 218 seats are needed, our model gives Democrats a 95 percent chance of a majority and a mean of 241 seats; that figure is marginally firmer than it should be, because our simulation correlates district error nationally, by census division and by state, but has no data to correlate demographically similar districts in different places.