I think it's fair to say that for Trump to win the election, he really needs to do better in battleground states than nationally, since he seems likely to lose the overall popular vote. Harry Enten has looked carefully at recent data and raises the possibility that Trump is losing that relative advantage in swing states.
Trump won the presidency in 2016 by winning the Electoral College while losing the popular vote. Previous polling has suggested that phenomenon could repeat itself in November: A lot of data suggested that Trump was, indeed, stronger in the battleground states than he was nationally.
But polling over the past month indicates his standing in those battleground states could be fading, bringing those numbers more in line with his national polling...
Florida: Presumptive Democratic nominee Joe Biden led Trump by six points in a University of North Florida poll and four points in a Quinnipiac University poll. The previous polling average of probability polls had Biden up 2 points.
Michigan: Biden was up by eight points in a Fox News poll compared to the prior polling average of probability polls that had him up five points.
Pennsylvania: Biden held an eight-point advantage in a Fox News poll, while he was up just three points in the longer term probability polling average.
Wisconsin: Biden was ahead by three points in a Marquette University Law School poll. The longer term probability polling average had the race tied.
Now, we're just looking at four polls here, so I don't want to make too much of it. It is interesting, however, that Biden seems to be doing better than his longer term averages in this limited state polling data, while he is not doing the same in the national polling.
Biden's lead nationally in those polls has been consistently around six to or seven points, as it is now in those.
Could it be that Biden is eliminating Trump's relative advantage in the swing states compared to his national standing?
Take a look at recent national polls from ABC News/Washington Post, CNN/SSRS and two from Monmouth University. Thanks to the Roper Center archives we can drill down to the state level in these polls. Specifically, we can look at the 15 closest states in the 2016 election.
Over the last month, Biden was up by about two points more in these 15 states than he was nationally in these same polls. That's quite different from 2016 when Hillary Clinton did more than 3 points worse in these battlegrounds than nationally."
A very interesting pattern. Frankly, I'd like to see a lot more data along these lines before I'd be confident of this trend's staying power. But if it did stay, that would imply that even if Biden's national lead compresses he would be less likely to lose because of an electoral college-popular vote mismatch.