Every presidential election is just like all the other ones until it’s not. The many people who predicted Donald Trump would flame out in the Republican primary aren’t idiots; mostly their predictions were based on past primaries. Many people are now predicting Trump will be destroyed in the general election. This is based on reams of voter surveys and demographic data from past elections. But what if everyone is repeating the same mistake?

In August, Nate Silver predicted Donald Trump was doomed to be a blip, just like Herman Cain and Michelle Bachmann in 2012, or perhaps Mike Huckabee in 2008. Almost exactly four years earlier, Republicans were passing around a slideshow noting that no president had been reelected when the consumer confidence index was below 75, and it was at 55.7. (It rose to a mere 72.2 a few days before President Barack Obama was reelected.) Until Obama, no Democrat from outside the South had been elected president since 1960. There are countless examples of rock-solid election laws that suddenly crumbled.

Since just after the 2012 election, political punditry has been almost Calvinist in its certainty about 2016: The electorate was predestined to have a particular demographic makeup, and each demographic group would turn out and vote a particular way, and the math added up to Republicans being toast. It wouldn’t matter much whether the GOP candidate was Chris Christie or Scott Walker or Jeb Bush, because voters’ views of the parties were set. 

But what if the candidacy of Donald Trump is so weird and new and different that it can actually change all that? Last summer, one of the few people who bet big on Trump winning the GOP nomination and the election was a British Scientologist. Of course a Scientologist would understand the power of celebrity in America—the church has long used movie stars to get people to do something even harder than switch parties: switch religions. John Mappin, who bet almost $10,000 on Trump and could win $100,000 come November, explained, His rise marks a paradigm shift in media, really ... What I could see was that the pundits in America hadn’t woken up to that.” Mappin continued, “People say, ‘Well, this is ridiculous, he’s a reality TV show star.’ But hang on, he’s an icon. He can use his power as an icon to attract attention. We’ve moved into a new media age, which is the age of icon control. And somebody with a Twitter account has effectively more power, perhaps unjustly, than somebody who perhaps has been sincerely working their whole life doing something really, really good.” 

What’s the evidence Democratic voters will be more immune to the power of celebrity than Republican voters? Perhaps the strongest such evidence is the phenomenon of negative partisanship—increasingly, people vote against the party they hate instead of for the party they like. Last summer, political scientists at Emory University published a paper showing that while people are more likely to call themselves independent, they are also more likely to vote straight ticket. In a June post at Larry Sabato’s Crystal Ball, Emory’s Alan I. Abramowitz and Steven Webster explained:

“Between 2000 and 2012 the proportion of positive partisans—voters who liked their own party more than they disliked the opposing party—fell from 61 percent to 38 percent while the proportion of negative partisans—voters who disliked the opposing party more than they liked their own party—rose from 20 percent to 42 percent.” 

It would only get worse in 2016, Abramowitz and Webster wrote, “given the deep divisions between Democrats and Republicans over issues ranging from immigration and health care to climate change and same-sex marriage.” Further, they added, “We can predict with a high degree of confidence that party loyalty will be very strong in next year’s elections.”

Nine months ago, in other words, hate-voting seemed like a recipe for stability. But what if it’s the opposite? What if a candidate came along and did something crazy, like change some of his party’s positions? There was a curious result in a recent Wall Street Journal pollOnly 8 percent of all voters think Trump “represents the values and positions of the Republican Party.” For Democratic-leaning voters, maybe that’s not more to love, but at least less to hate.

In the weeks before the Massachusetts primary, almost 20,000 Democrats changed their registration to Republican or independent, which a state official called a product of “the Trump effect.” In Pennsylvania, 46,000 Democrats switched to Republicans in 2016. Only half that number of Republicans have become Democrats (though more independents became Democrats than became Republicans). NBC Miami reports that in early 2016, “Florida saw Republican Party registration increase by 67,065 while Democratic Party registration increased by 34,943 and independent voters decreased 27,721.”

Mark Munroe, chairman of the Republican Party in Mahoning County, Ohio, said last week that he got calls every day from Democrats wanting to vote in the GOP primary. “And nine times out of 10, or 19 out of 20, you get the sense they are doing so because they want to vote for Trump,” Munroe said. Trump, he said, is “getting Democrats to cross over ... but he’s also getting a large number of unaffiliated voters—people who don’t participate in primaries.” (Munroe supports John Kasich.)

According to Pew Research Center, 2016 primary turnout rivals that of 2008. For Republicans, it’s the highest since at least 1980, a projected 17.3 million voters, and more than five million more than projected turnout on the Democratic side. As a column in The Hill recently noted, there’s no historic correlation between high turnout in the primaries and high turnout in the general. But the column also noted that people who pay less attention to the election are likely to sit out the primary and vote in the general. Isn’t it possible those less engaged voters might find Trump’s celebrity appealing? In July 2008, John McCain attacked Obama for being a “celebrity”—then he picked Sarah Palin to be his VP.


Last week, Politico reported that Republican donors expect Trump to lose the general election, and so contributors are “forking over piles of money to contain down-ballot collateral damage from a potential Trump nomination.” On a Politico podcast released Monday, Hillary Clinton pollster Joel Benenson said that if Trump is the Republican nominee, he could potentially expand the swing state map for Democrats, putting places like Arizona in play.

Benenson dismissed the idea that Trump could have a path to the White House through the Rust Belt. “What’s the evidence of it? The evidence of it, they’ve turned out a lot of people,” Benenson said. Romney won Rust Belt primaries too, and then lost them in the general. Trump, he said, “doesn’t have a message that appeals to these folks. It’s not real.”

Primary results thus far suggest that Trump appeals to at least some of the folks Benenson is referring to. His strongest constituency is whites without a college degree. If you look at FiveThirtyEight’s great interactive on what demographic changes would flip states from blue to red, you can see it takes a significant but not unfathomable increase in non-college educated white Republican votes to swing the election.

FiveThirtyEight

For a long time, liberals have been asking what’s the matter with Kansas—why do so many working-class whites vote against their economic self-interest? Obama infamously suggested they were clinging to white identity politics. It now seems crazy it took so long for a guy to come along and combine the two. Donald Trump trashes immigrants but says we shouldn’t let people die in the streets without health care.

A recent Atlantic Media/Pearson Opportunity poll finds that Americans think the biggest obstacles to “achieving your personal goals in life so far” are slow wage growth, few local jobs, and their lack of education. A large majority thinks they could make more money if they had more education, but they can’t afford to invest in it. The “central conundrum” of this poll, The Atlantic’s Ronald Brownstein writes, is that “those who might benefit the most from more education and training often feel least equipped, for cost and other reasons, to obtain it.”

Trump offers them someone to blame for the barriers to their economic success—immigrants, Wall Street, companies that move factories overseas. And even though he’s a billionaire who got his start with the help of his millionaire dad, and who wears different colors of hats to signal his mood to his butler, he talks like them. He makes jokes like them, he’s not polished and prim. He brags about being a winner and makes them feel like winners, too. Trump can give them something to vote for, instead of pointing to the other party as a monster to vote against.

To be sure, Trump has historically high unfavorable ratings for a general election candidate. He could end up turning out high numbers of Democrats who find him disgusting. Clinton has held a lead over Trump in national polling averages for six months. But his “ceiling” of support has slowly inched upward, and he hit 50 percent support among Republicans for the first time in one poll this week. While Clinton admits she’s not a natural politician, Trump delivers his stump speech like a comedian, playing to the mood and rhythm of the crowd. And unlike most Republicans, Trump can attack Clinton for her ties to Wall Street, the source of so much anger across party lines.

Will Donald Trump crush Hillary Clinton in the general election? Probably not. But instead of having the worst chance among all Republicans to beat her, maybe he has the best.