BOSTON — As editor and founder of FiveThirtyEight, Nate Silver makes a living by predicting the outcomes of sporting events and political elections. But even his exhaustive statistical forecasting models could not accurately project how far Donald Trump would get in the Republican presidential nomination race.
Silver discussed Trump’s unprecedented rise at the MIT Sloan Sports Analytics Conference, spending a half-hour on a non-sports-related topic that has dominated headlines over the past several months.
“There are a lot of complicated lessons on the rise of Trump,” he said. “A lot of things have to come together for something like this to happen, and they have.”
Silver said there were three arguments his team made for why Trump was unlikely to be the Republican Party nominee.
The first was that Trump was in a media bubble. Silver noted that there were moments in his campaign where Trump was getting 70 or 80 percent of network news coverage in a field of 17 candidates.
“There was some notion that when he got more coverage, the polls went up, and there was a positive feedback loop,” Silver said.
He said that argument is an important part of the story, but not sufficient enough to explain why exit polls show that Trump’s supporters have been behind the candidate for so long.
“The media bubble theory is somewhat wrong, but you can’t talk about Trump without talking about the role media played,” Silver added.
The second theory was that Trump was a “factional” candidate” who would win support in the way former candidates like Pat Buchanan, Mike Huckabee, and Ron Paul did in the past.
“I don’t mean to put them together, but they got 20, 30 percent of the vote,” Silver said. “That theory might actually be right — Trump has not won a majority in any of the states.”
The third theory revolved around the fact that Republicans, many who dislike Trump, would “stop him somehow.” But Silver noted that the party was too disorganized and dysfunctional early on in the race.
“You’re not supposed to have 17 candidates running,” he said. “There weren’t very many endorsements and people weren’t getting off the sidelines in the GOP until very late in the game.”
Silver said there are some historical examples similar to what’s going on with Trump, dating back elections during the 1960s, 1930s, or 1890s — “periods of social upheaval,” he noted.
But the Trump situation is certainly unique.
“It’s not that he just violates some academic theories — he violates precedent in so many ways,” Silver said.
