A Good Reason NOT To Trust Early Polls You Should Know

I have written repeatedly that as accurate as polls may be overall, especially closer to election day when people are more engaged, polls this early are largely useless. As pundits and strategists sager than I like to say, polls this early, no matter how accurate, are simply a snapshot in time.

But this time there’s an even better reason to take these early polls with a cowlick size grain of salt. It’s not a slur on the polls. They may have great and well proportioned sample sizes, and the questions may be fair and equitable. But what’s fallible in these polls is what’s missin g from the cross tabs.

Sarah Longwell is a political analyst, as well as the publisher of the website The Bulwark, which she obtained from GOP strategist Charlie Sykes. She has been doing research on 2024 GOP voters who voted for Trump in 2016 and 2020, and none of it is good news for the GOP, and may indicate their desperation at Trump’s guilty verdicts.

When she appeared on MSNBC today, she was fresh out of a focus group. The focus group consisted of nine registered GOP voters who had voted twice for Trump, but were not planning on voting for him in November. And the two major takeaways are a nightmare scenario for the GOP.

In multiple focus groups of similar sized, when asked what was the deciding factor for them being unwilling to vote for Trump again in 2024, the majority stated that it was the events of January 6th that was the break-glass moment for them. They could stomach a lot, but not a literal insurrection to keep Trump in power. When today’s group of nine was asked if Trump’s convictions made them more or less likely to vote for him, bad news again. Five said it made it less likely, one said more likely, and three said it didn’t matter.

Full Disclosure. The majority of focus group members are rock solid white suburban voters, the kind of voters Trump is going to have to harvest like summer wheat to have any chance at all in November. The fact that January 6th was the tipping point for the majority of them, and that more than 50% find his convictions disqualifying, presents Trump and the GOP with an almost insoluble problem.

Then Longwell dug into the polls, but she started to dissect them with a different view in mind. Most polls show Trump with a lead in most battleground states, but within the margin of error. She split the polls into two categories, Trump voters and Biden voters, and then went on to cross tab their backgrounds and most importantly, their reliability.

She found that Trump was surging with lower income, lower education, largely disengaged FUX News voters. Meanwhile Biden was surging with better educated, higher income, more engaged voters who got their information mostly from mainstream media, newspapers and magazines.

See the Balrog in the woodpile there? It was the words I highlighted for you. Trump is knocking it out in polling with voters who don’t traditionally show up to vote. And Biden is killing it with a smaller number, but his supporters do tend to show up on election day. And that’s not showing up in the cross tabs, and makes all the difference in the world.

Let’s say that you have 1000 Trump voters, and only 850 Biden voters. That should give Trump a close but comfy 150 vote margin of victory. But if on election day only 500 Trump voters show up, and 700 Biden voters do, then the result is flipped. And you’d never know it from looking at the polls, because they weren’t cross tabbed for high intensity and low intensity voters. The polls weren’t skewed, just incomplete.

All that being said, I’m still going to tell you to watch the polls over the next 2-3 weeks. Not so much the national polls, but the battleground states, since that’s where this election will be won or lost. For this simple reason. Previous polling before and during the trial showed that as much as 24% of GOP voters said that a Trump conviction would be the final straw for them. And we know from Longwell’s focus groups that the slippage rate may actually be higher than that, closer to 45%.

So when you look at battleground polls over the next 2-3 weeks, after the convictions were announced, if you see Trump slipping by 2-3 points, from where I’m sitting, those are conviction defections. And once they’re gone, they ain’t coming back. And as for the rest? Just keep repeating high intensity vs low intensity. These polls will still be a snapshot in time, but they may be a snapshot of rats tunning down the tow ropes of a garbage scow shipping water.

I thank you for the privilege of your time.

Support the site with a subscription today and see no more ads!

Go Ad-free Now!