You may want to keep an eye on a new Monmouth University poll expected out tomorrow morning.
The school’s last one, issued Sept. 2, showed some interesting regional trends because of a different approach.
Plenty of polls since have the lead wider, one reason Monmouth’s new poll is worth watching.
More interestingly, Monmouth used a regional breakdown that differs from the standard geographic one. They grouped counties by which ones Trump won by more than 10 percentage points in 2016, which counties Hillary Clinton won by more than 10 points and which both won by less than 10 points, referred to as swing counties.
So, in 2016, Hillary won the cumulative vote in the swing counties by a 48.6% to 47.4% margin. The swing counties are Berks, Bucks, Centre, Chester, Dauphin, Erie, Lackawanna, Lehigh, Monroe, Northampton. They accounted for 26% of the vote.
The poll showed Trump leading there by 46% to 44%.
So that’s good for the president. He’s doing better on turf Clinton won, albeit narrowly.
In the counties Clinton won by more than10 points, she won the overall vote 66.3% to 30.7%. Biden led in these counties by 65% to 30. These counties, which accounted for 34% of the vote, are Allegheny, Delaware, Montgomery, Philadelphia.
So that’s about the same as four years ago and good for Biden.
Trump won the other 53 counties, including Luzerne, by 64.8% to 31.4%. They accounted for 40% of the vote.
The poll showed Trump leading in those 58% to 37%.
So that’s good for Biden, too, because he’s doing better in the counties where Clinton got clobbered.
This is essentially what U.S. Sen. Bob Casey expected would happen two years ago. Casey said Biden won’t lose as big as Clinton in the counties Trump dominated, but will do about the same as Clinton where she did well.
One word of caution against taking the last poll or probably the new one too seriously. The last poll questioned 400 registered voters between Aug. 28 and 31, starting the day after the Republican National Convention.
It had an overall margin of error of plus or minus 4.9 points, but the margins of error are larger for the regions.