Are The “Warning Signs” Real?

The New York Times came out with a piece last week about how the warning signs are showing up again for the Democrats. The “warning signs” they were talking about were the poll numbers that weren’t very accurate the last few election cycles, and there’s a good reason for that. But are they actual warning signs for the Dems this time around?

The first thing that most (not all) of these polls do is over-sample Democrats. I’ve seen polls where as high as 46% were registered Democrats. In actuality, it’s nowhere near that much. According to statistics from states, only 24% of voters identify as Republican. 28% identify as Democrats, and 46% identify as independents.

Second, you have to take a really hard look at who is doing the poll, and what has their track record been in the past. While there were some that were accurate prior to 2016 (before Trump beat Clinton), they have fallen off quite a bit in the last five or six years. Zogby is one of those polls. Emerson and Harris are also pretty liberal.

Third, take a look at what happened in 2016, 2018, and 2020 versus what the polls said would happen. Wisconsin, Pennsylvania, Nevada, and Michigan all over-played a Biden win and predicted Democrat wins in 2016 and 2018, by several points. In some states it was a percentage point or less, and they had Biden winning by as many as eight points! The same polls had Florida and North Carolina going to Biden right up until the election in 2020, but he ended up losing both states to Trump. And Texas, Iowa, and Ohio were all close enough to be called “toss-up” states in the polls, but were comfortably in Trump’s corner in 2020.

We’re seeing some huge Democrat leads in a lot of the polling right now, especially in the Senate races. That would concern most Republicans, and even has led to Mitch McConnell calling the field of candidates “weak”. And in a lot of cases, he’s probably right. But the races in pretty much all of the Senate races are tightening to the point where maybe the Dems won’t expand their lead in the Senate. Maybe they won’t end up with a lead in the Senate at all. Here in Arizona, Blake Masters has closed the gap on Mark Kelly to within the margin of error in the last two weeks after being down by 11 points earlier.

In Pennsylvania, what was once a John Fetterman 13 point lead over Mehmet Oz is down to 4 points. In Florida, Val Demings had a slim lead over Marco Rubio, but that’s disappeared, and now Rubio leads by 3. In Georgia, Herschel Walker was down by 4-6 points to Raphael Warnock by now has been leading for the last two weeks. Catherine Cortez Masto once led Republican Adam Laxalt by 7-10 points. Now it’s down to 1 to 2 points, well within the margin of error. North Carolina has Ted Budd now leading Cheri Beasley by 3 points. He was behind by the same amount.

The generic ballot over in the House is another indicator that polls aren’t working. Between 1996 and 2020, the generic ballot (which Rasmussen now gives to the GOP by 4 points) was off in all but two of the 13 elections. In 2008, Democrats did 2 points better than the polls indicated, and in 2018, they did exactly what the polls indicated. Other than that the polling heavily favored the Democrats and Republicans out-performed the polls significantly. So, a GOP by 4 point lead tells me that it’s going to be a huge GOP win. The last time it was even more than one or two points for the GOP, was 2010 when the GOP took back the House during the Obama administration.

So, overall, I would have to agree the New York Times election analyst, Nate Cohn that states that are showing large Democrat wins right now, like Wisconsin, Ohio, and Pennsylvania, may just be a mirage. We can only hope!

Carry on world…you’re dismissed!