Anyone who remembers (shudder) the 2016 presidential contest is acutely aware that national and state polls often tell a different story (though contrary to what many people seem to believe, the former were quite a bit more accurate than the latter). In part because Trump won while losing the popular vote by 2.1 percent in 2016, there’s more battleground-state polling being conducted this year, and, at a fraught moment just before the candidate debates begin, the story they tell is confusing.
First, though, it should be observed that Biden’s lead over Trump in national polling averages remains very steady: He’s ahead by 6.8 percent (with 50.4 percent of the vote) at FiveThirtyEight and by 6.5 percent (with 49.5 percent of the vote) at RealClearPolitics. Biden has led Trump in national polls for well over a year now, and no, Hillary Clinton’s lead over Trump was never that consistent (at this point four years ago, she was up by 2.6 percent in the RCP averages).
State polls, though, are another matter. In the past few days, respectable surveys came out showing Biden tied with Trump in Iowa (Selzer) and Georgia (University of Georgia) and down by only two points in Texas (CBS News/YouGov). Trump carried these states by nine, nine, and five points, respectively, in 2016. There’s no fresh polling in Arizona, a state Trump won by 3.5 percent, but Biden is leading in the RCP averages for the state by 5 percent. And by the way, while there hasn’t been much polling in Ohio, because nobody figured the state would be competitive, Biden leads in the RCP averages there by 2.4 percent (Trump carried the state by eight points in 2016).
At the same time, a new poll in Florida (CBS YouGov) shows Biden up by just two points in that critical state (his lead in the RCP averages is even smaller there, at 1.6 percent), and the last two polls in Pennsylvania (Ipsos and Trafalgar) have Biden up there by three points and two points, respectively. Nate Cohn of the New York Times sums up the implications of recent state polling:
For those of us who are worried by the “red mirage” scenario of a temporary Trump lead on Election Night due to mailed ballots, which he baselessly attacks as illegitimate, Cohn notes this possible trajectory:
So yeah, there are reasons for both sides to be jittery, though you have to like Biden’s strategic flexibility in being ahead or close in so many states Trump must win.
Some new Senate polling shows how close the battle for control of the upper chamber remains, with some of the same dynamics as the presidential race: Democrats are doing well in a lot of red-state Senate races but not enough to survive any kind of late shift toward polarization (caused, say, by a Supreme Court confirmation fight). The University of Georgia has incumbent Republican David Perdue leading Jon Ossoff by just two points and a close three-way race involving Republicans Kelly Loeffler and Doug Collins and Democrat Raphael Warnock for Georgia’s other Senate seat (the top two will go on to a January special-election runoff). Selzer has Democrat Theresa Greenfield leading incumbent Joni Ernst by three points in Iowa. A New York Times/Siena poll shows Democrat Steve Bullock just one point behind incumbent Steve Daines in Montana. And in the biggest shocker, Quinnipiac has the exceptionally well-funded Democrat Jaime Harrison tied with Lindsey Graham in South Carolina.
At the same time, though, a new MRG Research survey in Michigan shows Republican John James within two points of incumbent Democrat Gary Peters and trending upward. Yes, it’s a typically GOP-leaning pollster showing an unexpectedly tight race, but it wouldn’t be a complete shock if the presidential contest tightened and the perhaps excessively low-key Peters got in trouble.
With races this tight, the presidential debates and the dynamics of the Supreme Court confirmation fight could matter a great deal, very soon.