This new “53% of Republicans” poll doesn’t mean what you think it does
The latest Economist/YouGov poll says that 53% of Republicans now think Donald Trump is a better president than Abraham Lincoln. This number is as astonishing as it is disconcerting as it is cringe-worthy. And there’s every reason to believe that it’s accurate. But this number does not at all mean what it sounds like it means.
While I like to see multiple new polls say the same thing before presuming it’s true, there’s no reason to presume that this particular new poll is false. At the least, there’s no other polling data that conflicts with it. So let’s give this poll the benefit of the doubt. What does it tell us about the overall American voting populace?
There are roughly 32 million registered Republicans in the United States. 53% of that is roughly 17 million. So we’re talking 17 million people, out of the 250 million Americans who are of voting age. In other words, this poll tells us that not much more than 5% of Americans think Trump is a better president than Lincoln. Even if you want to quibble with registration numbers, you’re still going to end up with a single digit number.
In other words, this group of people is electorally irrelevant. Donald Trump didn’t get elected because of the single digit number of people who think he’s the messiah. He won’t get reelected because of a single digit number of people. They’re already locked into voting for Trump, and they’re already factored into his overall very weak poll numbers. So these folks will have no impact on the election whatsoever.
If Donald Trump does win reelection, it’ll be because of his quasi-supporters outside of his base, who have never loved him, and still aren’t sure what to think of him, and could end up doing a variety of things in 2020 (unenthusiastically voting for him, unenthusiastically voting against him, staying home). It certainly won’t be because a statistical handful of deeply deranged people think he’s the greatest president of all time, as disturbing as they may be.
