The Washington PostDemocracy Dies in Darkness

The real reason Donald Trump is winning: No one thought it was possible

Political science said Trump would lose. Did GOP elites trust the theory so much they forgot to do anything to make it come true?

Donald Trump at a rally in Nevada just before he won the caucuses there. Did GOP elites fail to stop him because they thought they didn’t have to? (EPA/MIKE NELSON)

Donald Trump is now favored to win the GOP nomination for president, and it’s not close. All the polling suggests that Trump will win almost every Super Tuesday state this coming week. Prediction markets have also swung behind the billionaire. Trump skeptics remain, and given his polarizing effect on Republicans, some skepticism is probably warranted. But the idea that a winnowing field will help a non-Trump alternative surge to victory seems less plausible by the day.

If Washington has a comparative advantage in anything, it is in postmortems. Should Trump continue Making America Great Again all the way to the nomination, let alone the presidency, there will be plenty of autopsies about how and why it happened. Some will point to the frustration that many Republican-leaning voters feel toward the GOP establishment over illegal immigration, trade policy and myriad other issues. Others will note Trump’s symbiotic relationship with the media, which lets him exploit coverage of his bluster to drown out any other candidate’s perceived momentum. And it’s quite possible that Republican Party leaders actually prefer Trump to a candidate like Ted Cruz.

Donald Trump cruised to victory in Nevada, building a broad coalition that left his two rivals trailing far behind heading into Super Tuesday. The Washington Post’s David Weigel talks about what this means for Trump ’s chances of securing the GOP nomination. (Video: Alice Li/The Washington Post)

These explanations cover a lot of ground, but I want to propose an additional, somewhat unorthodox hypothesis. Trump is winning because no significant Republican coalition seriously tried to oppose him when there was still time for it to work. And the reason no powerful Republican coalition emerged to stop him is that the GOP believed all the analysts who said Trump had no chance. In other words, the political science theories predicting that someone like Trump was highly unlikely to win a major-party nomination were so widely believed that they turned out to refute themselves.

Let’s start with the part of my hypothesis that’s easy to prove: how little serious opposition Trump faced within the Republican Party. The New York Times said his popularity last summer evoked “barely concealed delight” from Jeb Bush’s campaign, which thought it could position Bush as the sensible alternative to the businessman’s bombast. One top GOP operative wrote in the Wall Street Journal in September that there was no reason for Trump’s rivals to attack him at a debate, because “wasting time attacking somebody who won’t get the nomination is just that: a waste of time.” According to Politico, only $9.5 million in super PAC advertising — 4 percent of the $238 million spent by outside groups this election cycle — has been aimed at opposing Trump. Since GOP strategist Liz Mair created the anti-Trump Make America Awesome super PAC in December, she’s raised a whopping $10,351.77 for her efforts.

No one took Silvio Berlusconi seriously, either

Yes, a few Republican elites have pushed back. Both the New Hampshire and South Carolina GOP chairs blasted Trump after he proposed banning Muslims from entering the United States. National Review organized an entire issuewith much fanfare, dedicated to opposing Trump. And Fox News’s moderators grilled him aggressively during the first GOP debate.

But as important as conservative media stalwarts might be, mainstream political science suggests that elected party officials matter more. Very few have endorsed Trump, but neither they nor their super PAC benefactors have spoken out against him much, either.

While the field of candidates was crowded, Trump posed what economists call a collective action problem. It was in everyone’s general interest for someone to attack Trump — but it wasn’t in anyone’s specific interest to do it or to draw Trump’s fury in response (ask Bush, whom Trump mocked mercilessly until he finally quit). So everyone hoped to benefit when a candidate went after Trump — only no major candidate ever bothered with a sustained attack. Even this past week, as the likelihood that anyone could stop Trump dwindled further, the remaining contenders seemed more intent on attacking each other, issuing snarky news releases about their rivals’ inability to win Nevada and ignoring the guy who did, though Sen. Marco Rubio (Fla.) did spend most of Friday making fun of Trump in a rather Trumpian manner.

Five myths Donald Trump tells about Donald Trump

As a political scientist, I have to at least consider the possibility that all this has happened because GOP leaders were reading too much political science.

I am not saying that Mitch McConnell and Paul Ryan have been combing through back issues of the American Political Science Review or even “The Party Decides,” the 2008 book that argued that party insiders continue to exercise significant control over the nomination process, despite reforms in the 1960s and 1970s. But there is now a whole cottage industry of political scientists and poli-sci-curious columnists who write about presidential campaigns for the kinds of serious outlets that politicians read. And what have these people been saying?

When Trump entered the race, FiveThirtyEight’s Harry Enten wrote, “Trump has a better chance of cameoing in another ‘Home Alone’ movie with Macaulay Culkin — or playing in the NBA Finals — than winning the Republican nomination.” In July, John Sides argued in The Washington Post that Trump’s surge was a product of media attention and that media scrutiny would be the effective antidote. In August, political forecaster Nate Silver declared that “Trump’s campaign will fail by one means or another.” In September, the New York Times’s Nate Cohn labeled Trump “an extreme long shot.” Bloomberg View’s Jonathan Bernstein said in September that a flailing Rick Perry still had a better chance of winning than Trump. After Perry dropped out, Bernstein re-upped his theory in October.

By November, even as Trump continued to poll well, auxiliary hypotheses emerged about how his support in surveys was suspect, his ground game was weak and his unfavorable ratings would lead to a low ceiling once voting started. In December, Cohn wrote again that Trump was unlikely to win. And Silver said, “The most difficult hurdles between Donald Trump and the Republican presidential nomination are still to come.” (I, too, made this argument again and again on The Post’s website.)

Armed with that analysis, GOP insiders issued very similar pronouncements. In August, Mitt Romney’s onetime campaign manager, Stuart Stevens, dismissed Trump, arguing in the Daily Beast that “all that we know about politics has not evaporated because Donald Trump says he’d like to be president.” Politico’s surveys of GOP operatives and activists repeatedly showed them writing off Trump’s chances. In August, one New Hampshire insider said, “Trump is generating a lot [of] controversy, but he is not taken seriously as a potential president.” At the beginning of 2016, numerous Republican elites told Politico they did not think Trump would be the nominee. Barely two weeks ago, an update was headlined, “Insiders: Hard road ahead for Trump.”

Dear Donald Trump supporters: Hear me out before you vote. Yours, Kareem Abdul-Jabbar

It is easy to mock these assessments now, but these are smart people who grounded their analyses in previous campaigns and in how political scientists analyzed those elections. It was not difficult, using this logic, to think of Donald Trump as a more obnoxious and racist version of Herman Cain.

The theories about how major-party candidates secure a nomination employed quality political science — but they have some flaws that Trump is illuminating. The first is that there are not a lot of data points on modern nomination fights with primaries, caucuses and superdelegates. Even if the theories were largely correct, they are based primarily on post-Watergate campaigns with no incumbent. That’s only 13 cases.

The second problem is that social science theories are, by their nature, reflexive. They try to explain human behavior, but humans can, in turn, read about these theories and adapt to them. We know from reporting on how both parties are attempting to better use data to reach voters that Republicans have paid close attention to political science theories about campaigns. If GOP decision-makers read these analyses about Trump, they might have concluded that they did not need to do anything to stop him, because he would inevitably fail anyway.

That conclusion would have been wrong, as voters in 12 states are poised to demonstrate Tuesday. Most of the political analyses concluding that Trump would not win were based on variations of “The Party Decides.” But that book argues that the reason insurgents are rarely successful is that party elites coalesce around an establishment front-runner, giving that person a commanding edge in resources, news coverage and endorsements over any alternatives before the voting starts and before most of the public is really paying attention.

In other words, party leaders actually have to do something to stop an insurgent. The whole reason smart analysts believed that Trump had no chance was because they thought GOP leaders would eventually strike against him. But that didn’t happen.

Looking backward, political victories always seem inevitable. As the political class moves from denial to anger to bargaining to depression and acceptance, there will be plenty of autopsies claiming that Trump’s political genius was unbeatable. But the rise of Trump was very resistible. That should haunt Republicans who fear what a Trump nomination would do to their party and their country. And if he goes on to win in November, it could haunt all of us.

This article is an expanded version of an earlier one.