Political scientist Matt Dickinson intended to spend his sabbatical writing the definitive work on the Obama White House staff. Instead, his leave year was dedicated to chronicling one of the most improbable candidacies for the American presidency.

should be clear right up front: I didn’t immediately scrap my sabbatical plans on that June 16 morning in 2015 when Donald J. Trump declared his candidacy for the presidency of the United States. That morning in Trump Tower, when he rode the down escalator to a flag-draped stage, Neil Young’s “Rockin’ in the Free World” reverberating around the atrium, I was as skeptical as anyone that this would be a “real” candidacy.

Instead, as I listened to Trump’s rambling justification for running (“We don’t have victories anymore”) interspersed with an equal mix of outlandish policy promises (“I will be the greatest jobs president that God ever created”) and crude attacks on ethnic groups like Mexican immigrants (“They are bringing drugs, they are bringing crime, they’re rapists”) my immediate concern was to extract the maximum entertainment value from his short-lived candidacy before its inevitable implosion.

Toward that end, I sat down after his announcement and wrote a fiendishly clever (to my mind) satirical piece on my Presidential Power blog explaining why I would break my long-standing policy of not voting in presidential elections so that I could support “The Donald.” After lauding, tongue firmly in cheek, his many stellar qualities—“He’s rich!”—I ended my piece by telling Trump, “You’re hired!”—a not-so-subtle play on his signature tagline from Celebrity Apprentice. And then I waited, smugly, for his political comeuppance. Instead, the comeuppance was mine. During the next six months—a period dubbed by political scientists “the invisible primary” because it tends to weed out weaker candidates—Trump continued issuing vague policy promises while insulting anyone who dared to oppose him, including war heroes, journalists, and media anchors, and, not least, his major political rivals, each of whom he tagged with a demeaning sobriquet: “Low-Energy Jeb” Bush, “Lying Ted” Cruz, and “Little Marco” Rubio. Remarkably, even as an increasingly worried Republican Party establishment began belatedly mobilizing against him, and pundits like Nate Silver continued to predict his imminent demise, Trump’s candidacy did not collapse. Instead, he quickly rose to the top of the public opinion polls and stayed there, despite raising and spending comparatively little money, running few ads, and generally seeming to defy what political scientists thought they understood about how to win a major party nomination.

By the end of 2015, Trump was up 20 points over Cruz, his nearest rival based on the aggregate polls. Most of his other opponents—several had already dropped out—were even further behind and polling only in single digits. How was Trump doing it? Belatedly, I decided to find out. With my wife, Alison (she took notes while I live-tweeted events), I had already begun attending candidate rallies in neighboring New Hampshire, home to the crucial first-in-the-nation primary and thus a magnet for all the presidential hopefuls. These were typically intimate, low-key events of perhaps 50–75 people, held at restaurants, schools, and Elks clubs, in which candidates provided brief statements mapping out their major policy positions, fielded some questions, and concluded by asking for people’s support.

Then, in January 2016, about a month before the New Hampshire primary, I attended my first Trump rally, at a high school in Claremont, New Hampshire. I soon realized I had committed the social scientist’s cardinal sin: opining on a topic I had never really studied, never mind understood. On the night of the rally, Alison and I arrived early, only to find a half-mile-long line to get into the high school gymnasium. The night was bitterly cold—the temperature hovering in the lower 20s—and many of those around us, particularly the high school students, were lightly dressed, so we expected some attrition. Instead, in the more than an hour that it took us to get inside, we saw not a single person leave the line. People passed the time buying the ubiquitous Trump campaign paraphernalia—buttons, hats and T-shirts—talking with a small, friendly group of protestors, and trying to stay warm. When we finally passed through the metal detectors at the building’s entrance, we discovered a packed gymnasium of more than 1,000 people. His was by far the largest campaign audience we had seen to date, save for a Bernie Sanders rally in Lebanon, New Hampshire. The crowd appeared to be a demographic cross-section of New Hampshire voters. Of course, it was hard to discern who the solid Trump supporters were, who might still be undecided, and who came just for the experience. But whatever their leanings, they were in a festive mood.

And then Trump hit the stage. In my blog post, published a couple of days later, I tried to capture the essence of my first Trump campaign speech: “Trump finally took the stage, entering while Twisted Sister’s ‘We’re Not Gonna Take It’ blared from the speakers, almost an hour after the scheduled start time. After referencing the size of the crowd—‘There’s a lot of people still outside . . . should we wait for them? No!’—he immediately launched into a discussion of the latest Reuters poll, which had him leading the race with 42 percent, with Ted Cruz second at 14 percent. He then proceeded to work his way through a series of poll results, all of which showed him ahead of his Republican rivals. ‘And we’re winning big in a place called New Hampshire!’ he exclaimed, to loud cheers. After doing a state-by-state rundown of the polls, he noted, ‘People always ask, “Why do you talk so much about the polls?” Because I’m winning! Believe me, if I’m not winning I don’t talk about ’em.’ He took pains to point out that he was leading among Hispanics in Nevada. Later he would add evangelicals and the Tea Party to his list of supporters—‘I’m winning with everything . . . I’m winning with the smart people. I’m winning with the not-so-smart people too!’”

And, in fact, Trump was winning—and would continue to do so when the actual voting began. After finishing second to Cruz in the Iowa caucuses—a strong performance by Trump in a caucus state where social conservatives dominate the Republican vote—he began methodically racking up victories and clearing the Republican field, beginning with his blowout win in New Hampshire a week later. Ten days after his New Hampshire victory, when the remaining Republican candidates moved south to the more racially diverse state of South Carolina, Alison and I followed. On our arrival we experienced a welcome dose of Southern hospitality when the locals pulled our rental truck from the ditch I had driven it into late at night trying to find our hotel. The next morning, after attending a Cruz rally, we returned to our hotel to find armed security blocking the entrance. It turned out the Trump entourage had moved in to our hotel the previous night, and he and his staff were now preparing to leave to get to his rally. (Later that day we asked the hotel manager what Trump was like. He effusively praised Trump as the most genial of guests.) We hurried ahead to the rally, and once again found ourselves in the midst of a campaign experience like no other.

The next day Trump scored a 10-point victory, ending the candidacy of the one-time front-runner Jeb Bush. On my Presidential Power site I took another stab at explaining Trump’s growing support: “If you want to know why Donald Trump won in South Carolina tonight, you need only have attended his rally yesterday at the Myrtle Beach Civic Center. . . . Inside, Elton John’s ‘Rocket Man’ blared so loud the floor shook. There was an air of expectation as the large crowd waited for The Donald to arrive. The floor of the Civic Center was packed—I estimated maybe 5,000 people pressing forward to the stage, trying to get a closer glimpse of the candidate. . . . As is typical for a Trump crowd, there was a healthy cross-section of demographic groups, but there was a definite segment of what appeared to be the working-class voter; for example, a group of bikers gathered next to me, with one of them wearing a leather jacket and clutching a Trump poster. . . . Trump ended [the rally] by asking the people to come out and vote for him. ‘We are going to start winning, winning, winning,’ he intoned, to rising applause. As we left the arena, people seemed in a festive mood, as if they had attended a great rock concert or sporting event. ‘You’ll remember this great meeting,’ Trump told them near the end of the speech. And he may very well be right.”

Less than 10 days later, on so-called Super Tuesday—the biggest day of the nominating race in terms of the number of delegates at stake—Trump won six more contests to cement his front-runner status. Two weeks after that we packed our bags once more, this time flying to Florida where we had intended to see whether Marco Rubio could resuscitate his flagging campaign in his home state’s March 16 primary. By the time we arrived, however, Rubio was on political life support. With our interest in him dwindling, and unable to get into any of the packed Trump rallies, we settled for a rather sedate Hillary Clinton event hosted by her husband, Bill. The next day, Trump crushed “Little Marco” by nearly 20 percent, ending Rubio’s candidacy. Although Trump’s remaining two opponents, Cruz and Ohio Governor John Kasich, hung on for another two months, the race for the Republican nomination was essentially over.

All that remained was Trump’s official coronation as the Republican standard-bearer. In late July, after attending one more Trump rally in upstate New York, Alison and I headed to Cleveland, host of the Republican National Convention, where, in my role as political blogger, I had secured media credentials. Could Trump unify the party? The talking heads on cable television thought not; they speculated endlessly about the chances for a brokered convention in which the Republican delegates would revolt against a Trump candidacy, and a white knight—Mitt Romney? Paul Ryan?—would ride in to rescue the party from certain electoral disaster. Alison and I spent much of our time staked out at a bar near the security entrance to the Quicken Loans arena, where we could watch media pundits like Chuck Todd, Wolf Blitzer, and Van Jones arrive to pontificate about the coming political bloodbath. Alas, the pundits were wrong again. After watching from the nosebleed seats high inside the arena as Trump gave a lengthy acceptance speech punctuated by frequent applause from a very supportive audience, I summarized my experience as follows: “I flew into Ohio expecting to see a very divided set of Republican delegates, and braced for major demonstrations in the streets. Neither expectation was met. In fact, despite the media’s tendency to focus attention on dissenting delegates and other controversies . . . this was a relatively tame event. Once Trump’s team, allied with the Republican Party leadership, beat back an early effort to amend the rules to allow the delegates to vote their conscience, the battle for the nomination was essentially over.”

And the battle for the general election race now began in earnest.

***

Trump had done the improbable in securing the nomination. Could he now do the impossible and win the presidency? Many of my colleagues thought not. They believed his appeal was limited to the Republican-leaning portion of the electorate, and that he would lose more moderate voters in the general election, and by a wide margin. Based on what Trump’s supporters were telling me at his rallies, I thought otherwise. As I wrote in my convention blog post, Trump’s message of economic populism was likely to find a broader audience: “I think his message will resonate with that portion of the electorate that has experienced years of stagnant wages, and who are worried about growing economic inequality and security issues.”

Eager to see how Trump’s candidacy would play with a general election audience, Alison and I attended one last Trump rally, in Windham, New Hampshire, during early August. Once again the event was packed, and this time we spent most of it outside, talking with Trump supporters barred by the fire marshal from entering the building. In my blog summary, I wrote: “It quickly became clear that two themes dominated the thinking of Trump supporters. The first, expressed—unprompted by me—by every person I talked to, was economic anxiety. Interestingly, that anxiety was not directed so much at their own situation but toward that of their children, or others close to them. . . . The second theme that emerged, again unprompted by me, was a deep antipathy toward Hillary Clinton. One man, in his early 40s, told me he wasn’t voting for Trump as much as he was against Clinton. Almost to a person those I talked to expressed a fundamental belief that she could not be trusted. At one point in our conversation, a woman whose daughter was having twins lowered her voice to tell me, ‘I’m a Roman Catholic and a good Christian, but I just have to say this: that woman [Hillary Clinton] is evil.’”

For much of the general election campaign, however, it appeared my colleagues might be right. Despite the political science forecast models issued in late summer that indicated the race for the popular vote would be very close, Clinton maintained a consistent lead of 2–5 percent in the national polls throughout the fall, and election forecasts by pundits like Silver using state polls gave her a very high probability of winning the Electoral College on November 8. Could the polls be wrong? In a lecture given in late October, I assured my audience that, historically, the polls were increasingly accurate as Election Day approached. If that pattern held, Clinton should win a narrow victory, I said. As it turned out, the polls were wrong. The political scientist forecast models were not. 

On Election Night my colleague Bert Johnson and I gathered in Crossroads Café to analyze the returns and provide commentary. A large crowd of perhaps several hundred expectant students had gathered. Tittering with excitement, they anticipated history being made by the election of the first woman president. That afternoon, in my final weekly politics lunch, I had laid out the progression of states based on poll closing times—first Florida, then North Carolina, and then Ohio—that Trump had to win if he were to pull off an upset. If he lost any one of them, I told my audience, Clinton would be our next president. That night, as the polls closed in the key states I had identified, beginning with Florida at 7, it gradually became clear as the votes were tallied that Trump was outperforming the polls. He had narrow leads in each of the benchmark states and, more surprisingly, he also led in states like Michigan, Pennsylvania, and Wisconsin, which almost no one predicted would be in play. As Bert and I took turns explaining what we were seeing in the returns, the festive mood dissipated. Students grew increasingly hushed. One fled in tears, whispering, “I can’t take this anymore.” By 11:30 p.m., when the staff closed the café down, it was apparent that we had witnessed history in the making—but not the history most had expected. By the time I arrived home later that night, most of the key states had been called in Trump’s favor, and it was all but certain that a billionaire businessman with no political experience was going to be our next president.

How did it happen? The next morning I began my U.S. Elections class by recounting what I had seen at the Trump rallies, and how listening to his supporters for the past year provided clues to his victory. Despite the caricatures of Trump voters often presented in partisan social media outlets, I explained, the vast majority were not racist, misogynistic “deplorables.” Indeed, in my conversations, Trump supporters rejected his more inflammatory talking points. Moreover, I pointed out that, although the state polls had systematically underestimated Trump’s support, the fundamentals-based political science forecasts that predicted a closer race proved remarkably accurate. Given those fundamentals—most notably the uneven progression of the economic recovery under Obama, and the difficulty parties had in holding the White House for three consecutive terms—I reminded them that Trump’s narrow victory was not a complete surprise. Nor did it represent a break with previous elections in terms of what motivated voters. It was still largely the economy.

For the most part, my students were not yet ready to hear this. Less than 24 hours earlier they had expected a Clinton victory. Now they were grieving. Several spoke with anger about what they believed Trump’s election really meant about voters in this country. Others broke down in tears. A week later I tried again. Speaking at a College-wide post-election panel, I said, “When you survey most Americans on a gamut of issues and you compare their response in ‘red states’ vs. ‘blue states,’ you find that there’s much more that binds us together, and that we share a common set of values. It’s only the case that when you’re forced to choose between two candidates who may, themselves, be polarized, it appears we are polarized as a nation. In fact, Trump supporters and Clinton supporters agree on much more than they disagree on, and the question, again, is ‘How do you move toward that common ground?’”

How indeed? Presidential elections are inherently divisive. Candidates and their partisans spend most campaigns telling us of the high stakes involved, and the disastrous consequences should their opponent win. In the increasingly balkanized world of cable news and social media, moreover, these polarizing tendencies get magnified. I understand why Trump’s unprecedented (and unpresidentially) divisive language alarmed people, and further exacerbated partisan differences. However, having spent the better part of my sabbatical year listening to Americans explain how they would vote, and why, I am convinced that most of those who backed Trump do not share his more incendiary views. We would do well to remember that our choices of candidates are often more polarized than we are. In reassuring our students that we reject the more demeaning portions of Trump’s message, we must be careful not to demonize the great portion of voters who backed him despite, and not because of, his more hurtful words. We are, in the end, greater together.