RCP Polling Averages Flipped MI, WI to Harris; Here’s Why It Doesn’t Matter

RCP Polling Averages Flipped MI, WI to Harris; Here’s Why It Doesn’t Matter

After becoming very comfortable with former President Donald Trump’s leads in the RealClearPolitics polling averages of all seven battleground states, I was alarmed to see that Michigan and Wisconsin had flipped into the Harris column on Tuesday.

The reversal in Wisconsin was caused by the release of a CNN/SSRS poll that found Vice President Kamala Harris up 6 points in the Badger State. This result, most definitely an outlier, flies in the face of surveys taken by all other pollsters since Labor Day. Of the nine additional polls included in the current RCP average, four show the race to be tied, three show Trump up by 1 point, and he is up 2 points in another. In the remaining poll, Harris leads by 1 point.

New polls in Wisconsin:

Marquette: Harris +1 (last poll was Harris +4)
CNN: Harris +6
Atlas Intel: TIED (last poll Harris +1)

Harris leads by 0.2% in the RCP Avg. https://t.co/nPh4lZbRcx

— Tom Bevan (@TomBevanRCP) October 30, 2024

It should be noted that this was the first time during this election cycle that CNN had commissioned a poll in the state. So what, if anything, did CNN see that nine other pollsters had missed? Hot Air’s Duane Patterson, who follows polling trends closely, provided some insight into this anomaly on Thursday morning.

First, he looked at the accuracy rate, or more precisely the inaccuracy rate, of CNN/SSRS polls in 2020. Patterson found, “They missed in the six states they polled, including Wisconsin, by an average of 7.3%. They were as bad as Quinnipiac in that cycle, and only Monmouth was worse.”

Next, he focused on the poll’s “methodology.” In other words, what was the breakdown by party affiliation of the participants?

He notes that CNN does not provide this information in the print story that accompanied the poll. “If a poll at this stage of the game will not give you the methodology behind it,” he said, “it’s in all likelihood a voter suppression poll. Polls have to have transparency behind them, especially when they’re released in the last week. There is none of that in the CNN poll.”

Here’s what CNN told readers about the new polls:

Interviews were conducted October 23-28, 2024, online and by telephone with registered voters, including 726 voters in Michigan, 819 in Pennsylvania and 736 in Wisconsin. Likely voters include all registered voters in the poll weighted for their predicted likelihood of voting in this year’s election. Results among likely voters in Michigan and Pennsylvania have a margin of sampling error of plus or minus 4.7 percentage points; it is 4.8 points among likely voters in Wisconsin.

Patterson zeroed in on the judgment call made by those who conducted the CNN poll and how it may have impacted the results.

So the sample size of Wisconsin is 736 registered voters. And of that subset of registered voters, an unspecified smaller number of people were given weight based upon what CNN/SSRS believes is their predicted likelihood of showing up next Tuesday. We don’t know out of which orifice they’re pulling that prediction out of, but the sample size is somewhere around what, 400? 450?

We are left wondering how many voters CNN/SSRS pollsters believe will ultimately cast a ballot. Judging from the result, they likely think more Democrats will show up to vote compared to their Republican or independent counterparts.

Exaggerating the point to emphasis the poll’s higher than average margin of error, Patterson asks, “Why don’t they just go ahead and say she’s up 10 with a MOE of 9 while they’re at it?”

Similarly, he looked at the Susquehanna poll that catapulted Harris into a 0.4 point lead in the RCP average of polls in Michigan on Tuesday and reached the conclusion. The poll, with a likely voter sample of 400 and a 4.9%, showed Harris up 5 points in the state. Patterson described both surveys as “garbage polls.”

I’d also like to highlight a Quinnipiac poll of Michigan voters released last week that showed Harris up 4 points. A previous iteration of this poll, released two weeks earlier, showed Trump up 4 points, a swing of 8 points. Weird, huh?

At any rate, rather than relying on polls, which can say whatever the pollster wants them to say, Patterson called upon readers to ask themselves the following questions:

When’s the last time Kamala Harris, at least politically, won the day against Donald Trump? When was her last honestly good day? You’d be hard-pressed to name one positive day for her on the campaign trail since her debate with Donald Trump. And that didn’t even move the needle for her very much, and not for very long.

Has the right track/wrong track number moved at all towards parity? By a margin of 64-27%, Americans believe the country is moving in the wrong direction, a spread of 37 points. That’s electoral death for the incumbent party. That number has not moved significantly in the last two years, and Kamala Harris has offered nothing rhetorically to persuade people she’ll be an improvement.

Interestingly, Fox News chief political analyst Brit Hume, who at 81 has covered presidential elections for a very long time, made similar points on Monday night. Hume described how reporters used to assess political candidates in the “old days” when there were far fewer polls available. He said:

We didn’t have the kind of polls and the number we have today. You’d go in to cover a race and polls would be nonexistent or they’d be old. So you’d rely on other things. You’d rely on how the candidate seemed. You relied on their events, how the events seemed to go, how well-organized they seemed to be. You looked at the response of the audience at these events, and how they seemed to be doing. You watched for other signs to pick up a sense of the race. And you could pretty well do it. … There are upsets in every election cycle, but you could get a sense of it.

If I were covering it the same way we used to cover it, I would look at this and say, Trump appears to be in the ascendancy. His campaign seems to have momentum. His events seem to be more exciting. They seem to be going better.

She seems to be struggling. She struggles to answer questions. She’s not doing well in interviews. And so on.

But the polls don’t reflect that. The polls say this is absolutely neck and neck. Judged the old-fashioned way, it wouldn’t appear to be. Judged the way we judge them now because we’re just surrounded by polls, that’s where we’re getting the idea that this race is tight.

I’ll leave you with the following assessment of the race from CNN senior political data reporter Harry Enten:

If Trump wins, the signs were there all along.

No incumbent party has won another term with so few voters saying the country is on the right track (28%) or when the president's net approval rating is so low (Biden's at -15 pts).

Also, big GOP registration gains in key states. pic.twitter.com/knDQ2HOFtJ

— (((Harry Enten))) (@ForecasterEnten) October 30, 2024

Just a little something for us to think about as we await the election results!

Elizabeth writes commentary for The Washington Examiner. She is an academy fellow at The Heritage Foundation and a member of the Editorial Board at The Sixteenth Council, a London think tank. Please follow Elizabeth on X or LinkedIn.

Donations tax deductible
to the full extent allowed by law.