Allan Lichtman vs. Nate Silver: Whose US Presidential Election Model Can Predict Correctly? Experts weigh in

The rivalry between top US election forecasters Allan Lichtman and Nate Silver will soon be put to the test as voters cast their ballots on November 5 and political pundits assess who could win the presidency.

Professor Lichtman of American University, who has accurately predicted nine of the previous presidential elections, has predicted that Vice President Kamala Harris will win the 2024 White House race.

What are Lichtman and Silver's forecasting methods?

In a recent New York Times article, Silver, the pollster and statistician who developed FiveThirtyEight, said that while the race was virtually even, he had a “gut feeling” that former President Donald Trump would win.

Both election observers argue about the merits of their different methods.

Silver turned to Lichtman and asked him if he accurately evaluated the “13 keys” he used to predict the September election results, claiming that the professor's method really worked to Trump's advantage. Lichtman clapped back, saying that Silver, who has a degree in economics, is “not a historian or a political scientist” and has made incorrect predictions in the past.

“At least seven of the keys, perhaps as many as eight, are clearly pro-Trump. Sorry, bro, but that's what it says on the keys. “Unless you admit they are completely arbitrary?” Silver wrote on social media.

According to Lichtman, Harris is favored by at least eight key people in 2024. However, Silver looks at the election results with a very different approach and data set.

Also read: US election Nostradamus reveals vulnerability that could be a “huge problem” for Kamala Harris

Allan Lichtman vs. Nate Silver: A look at their prediction records

In the ten U.S. elections since 1984, Lichtman has accurately predicted the results of nine of them. What he predicted wrong, however, was the fight between Al Gore and George W. Bush. Al Gore lost the 2000 presidential election to George W. Bush.

On the other hand, Silver became known across the country in 2008 after his statistical model accurately predicted the results of presidential elections in 49 out of 50 states. Since then, the results of the 2012 and 2020 presidential elections have been anticipated by his model. In the 2016 election, Silver's model predicted that Hillary Clinton would win, but gave Trump about a 30% chance, which was far greater than most other forecasters.

HT in the USA: Special coverage of the American elections by Prashant Jha

Here's what experts say about the accurate prediction

According to Thomas Miller, director of Northwestern University's Data Science School, both Silver's and Lichtman's approaches are “wrong in different ways.” Miller developed his own election prediction algorithm using data from the Predict It betting market and 60 years of historical analysis.

He claimed Lichtman's model doesn't take into account the way campaign messages and significant events shift public opinion in the final months of an election, USA TODAY reported.

Miller also recognized flaws in Silver's strategy, including an over-reliance on survey data that was inconsistent and error-prone. Silver will make false predictions if the polls are wrong.

Despite the divergence, David Wasserman, election analyst at Cook Political Report, said he believes Silver's strategy is “methodologically more rigorous.”