New CBS News/YouGov polls from Pennsylvania and Wisconsin seem to tell a familiar story. Former Vice President Joe Biden leads President Donald Trump by six points among likely voters in both states (49% to 43% in Pennsylvania and 48% to 42% in Wisconsin).
Biden's advantages in those states are nothing new and matches what the average shows.
Yet, a look at the CBS News/YouGov polling over the last month suggests a potentially important split between how pollsters are seeing the presidential election break down.
Specifically, CBS News/YouGov polling is suggesting that the potential for a popular vote and electoral college split exists as strongly in 2020 as it did in 2016. Meanwhile, traditional live interview polls indicate that Biden may be able to avert that fate in a tight election by performing disproportionately stronger in the Great Lake battleground states than Hillary Clinton did.
CBS News/YouGov does its polling online and, since the beginning of July, has surveyed the nine closest states Trump won in 2016. As first pointed out by the New York Times' Nate Cohn on Twitter, CBS News/YouGov has shown that Biden is doing better than Clinton's margins by a very consistent margin.
Biden's improvement over Clinton has been within a point of seven points (i.e. between six points and eight points) in every state they've polled except Arizona.
Across the four Great Lake battleground states (Michigan, Ohio, Pennsylvania and Wisconsin), Biden's doing about 6.7 points better than Clinton.
Across the five Sunbelt battleground states (Arizona, Florida, Georgia, North Carolina and Texas), Biden's doing about 6.5 points better than Clinton.
Now compare that to the traditional live interview polls, like CNN does with SSRS, in those states since June (i.e. a point since the polls have been fairly consistent).
Biden's doing about 10.4 points better than Clinton in those same Great Lake battleground states that CBS News/YouGov polled. In the same Sunbelt states, he's doing only 8.1 points better.
You'll note two things here.
First, Biden's doing better in live interview polling overall than online polling, which is true nationally as well.
Second, Biden's doing disproportionately better than Clinton in live interview polls in the Great Lakes. The same cannot be said of CBS News/YouGov.
The regional breakdown of the live interview polling makes a lot of sense given the live interview national polling. Biden's been doing much better than Clinton among White voters. And, as I pointed out last week, the Great Lakes have a lot more White voters as a proportion of the electorate than the nation as a whole.
The regional breakdown difference between CBS News/YouGov and the live interview polls may seem trivial, but it could make all the difference in the world if the race were to tighten.
Right now, CBS News/YouGov polling indicates that Biden would be wise to devote as much energy and time to the Sunbelt as the Great Lakes. The live interview polls recommend that Biden might be better off putting more energy into winning in the Great Lakes.
Perhaps more importantly, a disproportionate overperformance by Biden in the Great Lakes could help close the electoral college and popular vote split from 2016. Since June, Biden's doing a little less than nine points better in live interview polls nationally than Clinton. That's less than the 10.4 shift we're seeing in the Great Lake battlegrounds.
If the live interviews polls were exactly correct, the difference between the tipping state in the electoral college (i.e. state containing the median electoral vote plus one) and the national vote would be under two points and perhaps closer to one point. It was nearly three points in 2016. A slightly under two-point gap is much more in-line with what has happened historically.
Put another way, Biden would likely win in the electoral college if the national margin were the same as it was in 2016.
Meanwhile, CBS/YouGov polling suggests a different fate could occur. Biden's margin over Trump is no greater than six points in any of the states polled by CBS News/YouGov. This includes all the Great Lake battleground states. In other words, the tipping state in the electoral college (i.e. state containing the median electoral vote plus one) gives Biden a six-point margin.
At the same time, the one national poll CBS News/YouGov has done had Biden up 10 points. (This mostly matches other YouGov polls done over the same period.)
If the CBS News/YouGov were exactly correct, the difference in margin between the tipping point state and the electoral college would be at least as large as it was in 2016 and conceivably greater.
In the CBS News/YouGov universe, Biden would likely lose if the national margin were the same as it was in 2016.
Again, this doesn't really matter if Biden continues to lead by a fairly wide margin nationally. If Biden's national margin shrinks, this would turn into potentially a very big deal.
Now which group of polls is correct? The honest answer is I don't know. Live interview polls do tend to be slightly more accurate over the long term. Not to the extent, however, that we can count out the possibility of a large electoral vote and popular vote split like what happened in 2016.