With the first general-election presidential debate set for Tuesday night, state polls are coming thick and fast. And perhaps none of them speak to how much President Trump has his work cut out for him than in Pennsylvania.
That’s notable for a few reasons. Among them: Early this month, Pennsylvania was arguably the one pivotal state in which Trump appeared to be closing the gap. Biden’s lead in his birth state had declined to as little as 3.2 points in the RealClearPolitics average of polls. Monmouth University had Biden’s double-digit lead dropping to a virtual tie among likely voters.
Something similar happened in 2016, and we all remember what happened then, when Trump edged past Democratic nominee Hillary Clinton by less than a percentage point. It wound up being one of three Rust Belt/Midwestern states to go for Trump by less than a point — along with Michigan and Wisconsin — giving him his margin of victory in the electoral college. Clinton had reasserted a lead in October, but then things closed again, to the point where it was close enough for Trump to pull the upset.
At this point, though, the Pennsylvania of 2020 looks very little like it did in 2016, and a deep dive into the new polls shows why.
One vital area is the Philadelphia suburbs, a region that was huge for Democrats in their 2018 midterm victories and comprises more than 1 in 5 Pennsylvania voters. The NYT-Siena poll shows Biden leading in them by 24 points (54 to 30), while the Post-ABC poll has him up an astounding 38 points (69 to 31). Either margin would be significantly bigger than in any recent presidential election, with Democrats having carried them by seven to 17 points in every race since 2004. Trump in 2016 lost them by just 13 points (55 to 42).
Trump’s erosion is also particularly pronounced among White Americans with college degrees. The Post’s Philip Bump wrote Monday about how that looked on the national level, and Pennsylvania is a case in point. Although this demographic split 48 to 48 in 2016, Biden now leads by 23 points in the Post-ABC poll (61 to 38) and by 26 points in the NYT-Siena poll (59 to 33).
But it’s not just among college-educated Whites; it’s also among noncollege-educated Whites. Although Trump carried them by 32 points in 2016 (64 to 32), he currently leads by 13 in the NYT-Siena poll (52 to 39) and 17 in the Post-ABC poll (58 to 41).
This is the point at which Trump supporters begin throwing things at their screens. But the polls were wrong in 2016! Fair enough, but they were off by only a few points. Clinton led by two points in the final RealClearPolitics average, so they wound up being off by about three points. The media’s failure was, to a large extent, not recognizing margins of error and how much undecided voters would break for Trump. That was clearly the case in Pennsylvania, where Trump won voters who decided in the final week by 17 points.
And there’s another crosstab in the poll that speaks to just how much his base has shrunk. The Post-ABC poll asked voters whom they voted for in 2016. And although just 1 percent of Clinton supporters say they’re now backing Trump, 8 percent of 2016 Trump supporters now say they’ll cross over and back Biden.
That’s a small number, and it’s subject to the margin of error, but it reinforces the base erosion that we see in several other parts of the poll. Given his narrow 2016 margin, Trump can’t afford any of these shifts, much less all of them. His task beginning Tuesday night is to try to reassemble what he was able to piece together in 2016. Because without Pennsylvania, his path to victory becomes significantly more prohibitive.
