I strongly question the Silver Bulletin assessment that Kamala Harris currently is a 58.9 to 41.1 percent favorite in Pennsylvania.
The above chart is courtesy of Nate Silver’s Silver Bulletin.
On the surface the chart smacks of bias. I think the odds are wrong but bias has nothing to do with it.
There are 11 pages of polls that go into the forecast. They date all the way back to a YouGov poll on 2024-08-23.
Personally I would throw that away. Silver nearly does because that poll has an influence of 0.012. Recent polls have more influence.
38 polls make up the current projection. The above chart shows influence tapering off and it continues down to the above YouGov poll before falling off.
Where to Cut the Line?
I don’t know the answer but the cutoff point influences the odds.
For example. A Siena NYT poll 9/11 – 9/16 has Harris up by four percentage point with an influence of 0.66.
A new poll from Siena will reduce the influence of that poll. And if it tilts to Trump the odds will change significantly.
Siena is a top tier pollster as is Fox News. Silver comments on Pollster Bias.
Fox News’s polling department has long played it straight down the fairway, even if their editorial coverage hasn’t.
If we get an R +1 Rasmussen poll for example, the model thinks about it more like a D +2 poll due to Rasmussen’s large house effect. But an R +1 NYT/Siena poll is treated pretty much like an R +1 result.
Among our two highest-rated pollsters, for instance, NYT/Siena has had some bad data for Harris recently, while Selzer just published some excellent numbers for her in Iowa (only down 4 to Trump).
There were at least 10 consecutive polls with an influence above 0.5 in mid-September where Trump never had a lead. The result is shown in the lead chart.
Now four of the most recent polls have Trump ahead with one even.
I suspect the influence of the Harris-Trump debate is finally wearing off and Silver has not properly factored that in.
You can see the same thing but to a smaller degree in Michigan and Wisconsin.
The Best News for Trump
The best news for Trump is the Beacon Research/Shaw & Co. Research (Fox News) poll which has a historical tiny Democrat bias. That recent poll has Trump ahead in Pennsylvania by 0.5 percentage points.
Patriot Polling only has a Republican bias of 0.3 so that was another solid poll for Trump.
Trafalgar is best read as Harris plus 0.5 percent not Trump +2.2 percent.
Unfortunately, Silver did not list the bias of AtlasIntel, but unless that pollster has a huge Republican bias that also looks good for Trump.
I consistently see people mocking NYT and Fox news polls as biased but they are both very high quality and minimally biased.
What Are the Odds?
I am a big fan of momentum and also more recent polls, more so than Silver. A solid poll from NYT in favor of Trump with no deterioration elsewhere would have me thinking Trump is ahead.
Right now it looks more like a tossup than a 58.9 to 41.1 edge for Harris.