We support our Publishers and Content Creators. You can view this story on their website by CLICKING HERE.

I will not believe in single polls. I will not believe in single polls. I will not believe in single polls.

But that doesn’t mean we can’t look at them in the final days of the election cycle. As everyone tries to get a read on voters in the Blue Wall battleground states, each new survey gives us a little more clarity. Or so we hope, but the new USA Today/Suffolk poll delivers more ambiguity about Pennsylvania instead. In their sample of just 500 likely voters (!), they find the presidential race tied at 49/49.

Advertisement

However, some of the crosstab data may offer clues:

A poll of 300 likely voters in Erie County, which could indicate which way the state trends, was also tied 48% to 48%. Northampton County, another Pennsylvania bellwether, leaned slightly towards Trump, with 50% saying they supported him, to Harris’ 48%. The results of the county polls are within the margin of error of 5.65 percentage points. … 

Trump is up by 20 points among men in Pennsylvania, 57% to 37%, while Harris has an 18% hold on women over Trump, 57% to 39%. That’s compared to Trump’s 16-point advantage among men nationally and Harris 17-point advantage among women.

Trump has the gender-gap edge here, slightly. However, women usually turn out slightly more than men do, too, so this may be a wash. A wash would parallel what happened in 2020 among PA voters too, when Trump and Biden each got 55/44 in their respective male and female demos. 

This looks more on point, although the way in which this is parsed seems odd:

Unsurprisingly, more than 70% of people who viewed current economic conditions as poor said they supported Trump. Harris outperformed Trump with those who believed the economy was in fair, good or excellent shape. 

Normally, pollsters would group “poor” and “fair” together so as to allow for balanced response. Taking a look at the crosstabs, it turns out that 70% of respondents rate the current economy as poor (39%) or fair (31%). Later in the crosstabs, we discover that Harris does outperform Trump among those who chose “fair,” but not by much — 25 respondents, or about 5% of the sample. Among the 70% who rate the economy as poor or fair (350), Trump wins 234-104, or 46.8/20.8. 

Advertisement

What about the ethnic demos? That’s more straightforward. Harris wins the black vote 78/17 and the Hispanic vote 52/48. Those numbers are a huge problem for Harris when compared to the 2020 exit polls. Joe Biden won the black 92/7 and Latinos 69/27. In this poll, the sample respondents say their 2020 votes split 88/10 and 52/41, respectively. If those splits are valid, then this isn’t going to be a tie. 

The same is true for independents. This poll has Trump leading among unaffiliated PA voters 51/43, with the same voters saying they went 48/34 for Biden. The 2020 exit poll showed Biden winning independents 52/44 four years ago. If Trump has flipped the gap with indies by sixteen points, I’d expect that he’s doing better than just a one-point pickup over 2020. 

With all this movement toward Trump in these demos, how can this race be a dead heat at this point, when he only lost PA in 2020 by 1.2 points? It’s possible, I suppose, but it seems rather improbable. 

The same issue appears in a new Marist poll that shows Harris up by two points, 50/49. This is the only PA poll in a week showing her with a lead at all, which makes it a mild outlier. Marist also has Trump getting 16% of the black vote, more than double from four years ago. That seems too significant to support a topline that exceeds Biden’s performance against Trump, but YMMV. 

Overall, Trump continues to lead the RCP aggregate in PA by +0.5 points. That’s a dead-heat result too, but it’s still better than trailing. Trump has had a lead now since October 6 in the RCP aggregate, and the current gap has held pretty steadily since October 15th. On this date eight years ago, the polls had narrowed a bit, but Hillary Clinton still led the aggregate by three full points; four years ago, Biden led by 3.6 points. If the pollsters haven’t corrected for those misses in this cycle, we could see Trump sweep the Blue Wall states, and perhaps especially in PA. 

Advertisement

But that’s still an if, so … don’t get cocky, kid.