Earlier this month, the media burst with joy when historian Allan Lichtman offered the latest update to his election prediction model, the “Keys to the White House.” According to Lichtman, this model, which has correctly forecasted every election outcome since 1984, showed that Kamala Harris will win the 2024 election.
The problem with his model was that many of the keys are subjective, and when I reassessed them, it took no effort to show that his model actually predicts a Trump victory.
On Thursday, another election forecaster said Kamala Harris would win in a landslide.
Northwestern University data scientist Thomas Miller said that Harris will knock Trump in a landslide election, telling Fortune Magazine: 'It's gone from a drastic landslide in Trump's direction to a drastic landslide for Harris'.
Miller first gained attention after he accurately predicted the 2020 presidential election by unconventionally using betting markets rather than traditional polls to forecast the outcome.
He achieved this by developing a model, which is based on 16 presidential elections, that converted betting prices into the popular vote and Electoral College projections.
This model shows a narrow correlation between betting odds and the anticipated popular vote.
Currently, Miller's model projects that Kamala will win 55% of the popular vote.
If that doesn’t raise serious alarm bells, you’re not paying attention.
Let's put this 55% number into historical context. Barack Obama didn't even win 55% of the national popular vote in 2008. Despite being such an "inspirational" figure who got 95% of the black vote and drove up turnout, he only won with 52.9% of the national popular vote. I don't doubt that Kamala Harris could potentially win the presidential election, but does anyone actually believe that she will outperform Barack Obama's national popular vote share? Heck no. In 1980, Ronald Reagan won 44 states, defeating Jimmy Carter in the Electoral College 489 to 49, and still only got 50.7% of the national popular vote. A 55% national popular vote victory would even be better than George H.W. Bush's landslide victory in 1988 when he won with 53.4% of the national popular vote.
In other words, according to Thomas Miller, Kamala Harris is on track to win a blowout landslide victory akin to Ronald Reagan's 1984 victory, when he won 49 states and got 58.8% of the national popular vote.
What makes Miller's prediction even more absurd is that polls have consistently shown that Donald Trump has gained significant ground with traditionally Democrat constituencies, like black and Hispanic voters—a trend that has continued even after Kamala replaced Biden on top of the Democratic ticket. Meanwhile, according to Gallup, Harris has a net favorability among independents of -25 points, with 60% viewing her unfavorably. Meanwhile, she couldn't even secure the endorsement of the Teamsters, and polls show her margins in blue states are far behind Joe Biden's 2020 margins. Even Minnesota is too close for comfort for the Harris-Walz campaign.
This isn’t a blueprint for a landslide victory; it’s a recipe for defeat. Now, to be clear, I’m not saying Kamala Harris can’t win—I’m saying there’s no way she’s getting 55% of the national popular vote. To believe that, she'd have to be seen as a once-in-a-generation transformative candidate. And let’s be honest. While the media is trying hard to pump her up as one, even her most die-hard supporters don’t honestly see her that way.
Election forecasters clearly think we're stupid.
Join the conversation as a VIP Member