According to recent polls, Americans are supposedly suffering a severe case of buyer's remorse over Donald Trump — so severe, in fact, that if the 2024 election were held again today, Kamala Harris would win by eight points. Call me crazy, but I’m just not buying it.
CNN’s chief data analyst, Harry Enten, walked viewers through the numbers earlier this week. He pointed out that Trump's approval rating has dipped, and the polling industrial complex has sprung into action. Enten laid out what he called a dramatic shift. "The actual was Donald Trump winning by about a point and a half," he said, before explaining that a new NBC News survey — weighted, he was careful to note, to reflect the actual 2024 result — showed something very different. "Kamala Harris wins in a redo… she wins by, get this, eight points. A massive shift from what we saw back in November of 2024, when Donald Trump won by a point."
Eight points. In a sample weighted to show Trump winning by one. That is, if you take it seriously, a genuinely big number.
Host John Berman couldn't help himself. "So, the good news is, Kamala Harris, for her, wins the 2024 election. The bad news is, it's actually 2026, so it doesn't count for her."
Regrets: Harris now clearly leads Trump in a 2024 redo.
— (((Harry Enten))) (@ForecasterEnten) February 17, 2026
Big implications for 2026: Dems for Congress are holding onto Harris voters more than the GOP is holding onto Trump voters.
Further: non-2024 voters are heavily favoring Dems.
Puts not just the House, but Senate in-play. pic.twitter.com/qioTChA2Zh
I’m sure many would like to dismiss the polling because hypothetical election-redo polls are, at their core, fantasy. They measure sentiment and frustration without the weight of consequences. But look, we saw polls like this under Biden, showing that Trump would win a redo election, and we all felt validated by them.
Does that mean we should take these polls at face value? I wouldn’t.
For one thing, we’ve been treated to stories of Trump voter regret pretty much since the moment he took office again. Stories claimed a mass exodus of Trump voters was taking place, even as polls showed record-high approval of Trump from Republican voters.
ICYMI: Gutfeld Nukes AOC After Her Munich Humiliation, and It Was Epic!
Polling at this stage of a midterm cycle should be consumed with significant skepticism. We are in February 2026, months away from any actual votes being cast or from voters even being all that engaged.
The 2024 cycle demonstrated, again, that polls can miss badly — and that the media's preferred narrative has a way of running ahead of reality. Countless polls claimed that Kamala was beating Trump, even as Trump had record support from minority voters and women. The polling made no sense, unless you considered the fact that the pollsters found a way to get the results they wanted, even when the data clearly conflicted with it.
In fact, it wasn’t until after the election that we learned that the Harris-Walz campaign’s internal polling never had her leading Trump, even as public polls showed her ahead through much of the race.
A poll showing Harris winning a do-over election is a fun data point for Democrats to gush over. But the only thing it accomplishes is giving Kamala Harris a reason to think she should run again in 2028, and to believe she might actually win.
