The past two election cycles have ushered in a more empirically grounded style of political analysis. A new generation of data-minded analysts now compete with the old-school pundits who predict election results based on what their gut tell them about a race, or deduce voter trends based from remarks they overheard in a bar in Iowa or a cab in Chicago. And this new, more empirical approach has brought some key insights from political science into the national conversation.

So far, the 2016 cycle has left the pundits flummoxed. It’s a running theme that Donald Trump and Ted Cruz’s presence at the top of the Republican ticket defies easy explanation. They’ve been similarly hard-pressed to explain how a 74-year-old Brooklyn Jew who identifies himself as a democratic socialist is giving a candidate with as much establishment support as Hillary Clinton a real challenge. But what about the political scientists? Have they been equally surprised at the way the 2016 race has played out?

Last week, I invited David Karol, a professor of government and politics at the University of Maryland, College Park, onto my show, Politics and Reality Radio, to get his take. Karol co-authored The Party Decides: Presidential Nominations Before and After Reform, an influential book that argues that, while the two major parties ended the days of deals cut in smoke-filled rooms when they adopted a transparent primary process in the 1970s, “party actors” still hold an enormous amount of influence over the nomination process. Rather than analyzing how candidates woo voters, Karol and his colleagues looked at how party activists vet candidates during an “invisible primary,” long before most people start tuning into a race, and then help shape primary voters’ views of those vying to lead their party.

Donald Trump’s commanding lead in the GOP contest seems to undercut their thesis, and Karol was quick to acknowledge that, “in real time, there’s certainly been some startling developments.… Donald Trump is really a unique candidate. It just has to be said: He doesn’t fit a lot of…paradigms and models. In some ways every election is different. We try to generalize, and find patterns, and I think when it’s all said and done, even this year, there will be a lot of resemblances to past years. But I certainly think Trump has been a disruptive force in the Republican nomination process this year and has kept people off balance.”

“So the dynamics that we identified in the book haven’t been evident in this race,” Karol admitted. “Then the question becomes, Is it because the world has changed? The book was written in 2008. It’s possible. The world evolves. In the case of Trump, were he to become the nominee, I would say he really is a unique figure. He was better known than any of the other Republican candidates before he even got in the race. He’s better known than Jeb Bush—better known than anyone running except Hillary Clinton, and he’s actually been famous for even longer than she has.”

But he didn’t think that Trump’s success should necessarily change the way we look at this process. Karol and his co-authors define “party actors” broadly—the term isn’t limited to people working within the party infrastructure. It includes everyone who has some influence, both inside and outside the parties, and who are heavily invested in the outcome and highly engaged in the process. Unions are a good example on the Democratic side, as are leaders of evangelical churches for the GOP.

I asked Karol what he thought about the fact that, in this cycle, so many voters on both sides of the aisle have been railing against their parties’ respective “establishments.”

He said, “I think the term ‘establishment’ is kind of unfortunate. When people talk about the Republican establishment, they seem to mean the business community, and elected officials and lobbyists. That’s part of it, but that’s not the whole story. So I think the term is being used in a kind of misleading way. On the Democratic side, it’s interesting. When liberal interest groups like Planned Parenthood and the Human Rights Campaign—the LGBT Rights Group—endorsed Hillary Clinton, Bernie said, ‘Well, they’re part of the establishment.’ I think that’s partially misleading, but in a way it’s right. ‘The establishment’ is a pejorative term, and people got very upset about that. But of course they are part of the Democratic Party establishment. They’re aligned with the party. They have activists, they have money, their endorsement carries weight, and they are players, so I think that’s true. And I think that’s how parties work.

“I think on the public level there’s kind of this idea that there really shouldn’t be parties, or the idea of parties, because everyone doesn’t have a completely equal voice. But you could be somebody who’s not even registered with the party. If you show up one day, should you have an equal say to someone who’s spent their life working for this party? Working, organizing, raising money—that’s part of being a party actor. But there’s this idea that the establishment is somehow sinister.”

Some Sanders supporters are concerned that the Democratic “super-delegates”—many of whom are party officials and officeholders who are free to choose which candidate they’ll back at the Democratic National Convention—will tilt the balance toward Hillary Clinton if Bernie Sanders holds a small delegate lead at the end of the primaries. There are 712 super-delegates, and they make up around 30 percent of the total number of delegates a candidate needs to secure the party’s nomination. According to a survey conducted last November by the Associated Press, 359 of them planned to support Clinton, eight planned to back Sanders and 210 remained on the fence.

Karol thinks those concerns are, at best, overblown.”There is no historical evidence that super-delegates have the backbone to go against a candidate who is leading the primary and caucus voting,” he said. “The mere suggestion this might happen in 2008 was turned into a scandal by the Obama people, including the ‘neutral’ Nancy Pelosi. As a party scholar, I am all for super-delegates as an institution, but it is really unclear whether they retain the legitimacy to do much of anything. The only time they played a meaningful role was in 1984 when they put Mondale over the top, but he was well ahead of his rivals and that was a long time ago.”

Some Sanders fans take heart from Obama’s 2008 primary victory over Clinton. In that race, as in 2016, Clinton began the contest as the heavy favorite. If Obama could rise from underdog status to take the win, they say, why shouldn’t Sanders be seen as having similar chances?

There are some obvious differences between the races: In 2008, Democrats had the wind at their backs after two disastrous terms of George W. Bush, it was a three-way contest until John Edwards dropped out after the Florida primary and the campaign finance landscape was markedly different. I asked Karol how he saw this race compared to 2008.

He said, “I don’t think it’s nearly as close as it was in 2008. Hillary Clinton was never the overwhelming party choice the way she is this time. She had, let’s say, a preponderance of support at that time, but it wasn’t close to the level we see today. There were more unions supporting Obama, even though Obama was maybe less union-oriented than Sanders. He had more elected officials, and more major fundraisers. He was seen from the beginning as a credible candidate, where with Sanders a lot of people said, ‘Well, he’s running to raise his issues.’”

Turning to the general election, I asked Karol whether he thought a possible Trump or Sanders win in the primaries would shake up the way political scientists tend to handicap a race. For the past few cycles, we’ve heard a lot about how the quality of candidates and their campaigns, while not meaningless, are far less important than most voters—and pundits—believe. In this view, the “fundamentals”—how the economy is doing in an election year, and how favorably or unfavorably an incumbent or outgoing president is—play a major role in determining which party is favored to win. This is central to Nate Silver’s 538 election prediction model, but we haven’t heard as much from the political science fundamentalists this year as we have in the past.

Karol: “I don’t think it’s inconsistent to say that in general, when we’re talking about candidates within the normal scope of politics, like a Mitt Romney, or a John McCain, or a John Kerry, then these structural considerations could be very important. But within the range of cases that we’ve seen—the past elections that these models are built on, post–World War II elections—who is remotely like Trump in that relatively sample of cases? Nobody. With Cruz, you could say, ‘Well, maybe he’s kind of like Barry Goldwater or like George McGovern.’ He appeals only to the hardcore of his party, and those candidates did very badly.

“As a generalization, there’s some evidence in political science that candidates who are seen as extremists pay some penalty, and I think that would happen to Cruz and that would happen to Sanders. Trump is just in a category of his own, I think.”

And what happens if you have two candidates who are perceived as being outside of the mainstream?

”Well, in presidential politics we’d probably be in uncharted territory,” he replied. “We haven’t had the Goldwater-versus-McGovern election. Maybe you could say, ‘Well, if they’re both perceived as extremists than the structural fundamental conditions matter—the economy, and so on.’”

Hillary Clinton is well within the mainstream, but she’d be the first woman to secure a major party’s nomination. And unless a more traditional candidate catches fire in the Republican race, it looks likely at this point that the GOP will choose Donald Trump or Ted Cruz. And that means that 2016 will be, at least to some degree, “uncharted territory.”

David Karol’s responses have been condensed and edited for clarity. You can listen to the entire 20-minute interview below.

You can also download the whole show at iTunes. It also featured Salon’s Gary Legum reporting from New Hampshire and Nazita Lajevardi, a researcher at the University of San Diego who discussed a new study she co-authored that found that strict voter-ID laws are far more effective in suppressing the Democratic vote than previously believed.