The election of Donald Trump signifies a lot of things—is one of them a rebuke to popular culture’s political influence? Hillary Clinton lined up A-list entertainers for fundraising, endorsements, and performances, from Katy Perry to Lena Dunham to Bruce Springsteen to Beyoncé and Jay Z. Donald Trump, a TV star himself, boasted fewer well-known entertainers in his camp, and in the campaign’s final days made fun of Clinton for relying on celebrities.

David J. Jackson, a political science professor at Bowling Green State University, studies the influence of celebrities on elections. In a 2015 survey of 804 likely general-election voters in Ohio, he asked people whether particular celebrity endorsements would make people more or less likely to support a candidate. In almost all the cases, the net effect of any particular endorsement on a sample of the general electorate was negative—voters were less likely to support the endorsed candidate. But the effect often switched to positive when you just focused on demographics already favorable to any given celebrity.

I spoke with him on Thursday for a post-mortem on celebrity’s role in the election. This conversation has been edited.


Spencer Kornhaber: You’ve studied the intersection of politics and celebrity and popular culture for a long time. What were the things to know about the subject going into this election?

David Jackson: Initially my research focused on looking at the effect of celebrity endorsements on beliefs about particular topics. What that research had shown was that celebrities taking political positions that were popular could make those positions more popular, or celebrities taking positions that were unpopular could make those positions less unpopular. But there wasn’t a tremendous amount of data showing moving from agree to disagree or disagree to agree. In other words, a reinforcement.

Bono said some very positive things about George W Bush’s policies on AIDS in Africa. I’d survey young people, and young people weren’t a fan of Bush, but when Bono said something about him, it didn’t make them more likely to agree with it, it just made them less likely to disagree with it. So that’s something.

More recently, I’ve begun looking at endorsements of particular candidates. In a survey we did on likely voters in Ohio in 2015, we asked about a bunch of traditional endorsers—The Cleveland Plain Dealer, The New York Times, UAW, NRA—but a bunch of celebrities as well. The greatest percentage of people said that celebrity endorsements would have no effect. But there were some who said it would make them more or less likely to support a candidate. In every case, if you subtract the less likely from the more, then that effect is negative [on voters’ likelihood to support the endorsed candidate]. The smallest effect is George Clooney, who was only negative by around 9/10ths of a percent. And the two least effective, or more negative, were Ted Nugent and Beyoncé, who became fairly significant towards the end of the campaign.

But in Ted Nugent, if you only look at people who are sympathetic to the Tea Party, [his endorsement] goes from being a negative to a positive. Trace Adkins, a country star who was the winner of The Celebrity Apprentice, if you look at just country and western fans, he becomes positive, whereas overall he’s negative. Oprah Winfrey, among African Americans only, positive effect; over the electorate, overall negative effect.

The suggestion in terms of strategy is that candidates have to deploy their celebrities to the selective, appropriate audience. A lot of the thinking comes from marketing research about celebrity endorsements as well: You want someone who might be influential amongst your target audience if you’re trying to sell ‘em cars or coffee. Same thing is true of if you’re trying to persuade them to vote for a candidate.

Now, towards the end of the campaign, I don’t think Ted Nugent was out there trying to persuade people in Michigan for Trump. Or that Beyoncé, Bon Jovi, Springsteen, Jay Z, Katy Perry, that late in the game were trying to persuade people. They were trying to energize people to go out and vote. And that was quite an uphill battle for the Clinton campaign to begin with, because the energy and enthusiasm gap was in favor of Trump.

Analysis from less sophisticated places is [going to be] what they said in 2004: Lots of celebrities endorsed Kerry, Kerry lost, therefore celebrities either caused it—which was the worst analysis—or had no effect, which was equally lazy analysis. It’s illogical to say that celebrities cost her the election, though they didn’t win her the election.

Kornhaber: From what you’re saying, though, it is fair to think that celebrities could’ve had a somewhat negative effect overall for Clinton. Especially given that Trump was going around trash-talking the ones who endorsed her.

Jackson: Sure. Lately there’s been some folks throwing this hypothesis out there saying celebrities now may be perceived not just as wealthy pampered people [to whom] a lot of people say, “Fuck you, I don’t wanna listen to what you have to say.” Now they’ve actually been involved in so many campaigns—since 1992, when [Bill] Clinton went on Arsenio Hall—perhaps celebrities are perceived as part of “the establishment.” Their endorsement of Clinton may have had a detrimental effect in that respect. Trump was perceived and promoted as an anti-establishment candidate.

I argued initially when that hypothesis was presented that Trump was critiquing her use of celebrities because he didn’t have any, that it was just a reaction to [only] having Scott Baio, the soap opera actress selling avocados, the Duck Dynasty guy, and Kirstie Alley but then she changed her mind. I don’t know to what extent that was a strategic choice because they were believing the rhetoric that celebrities were representative of the establishment. That hypothesis remains untested but certainly has some face validity.

Kornhaber: There are a couple different ways celebrities can support a candidate: They can raise money, they can speak out, they can do what Jay Z and Beyoncé did in those final days and perform. What are your thoughts on the different forms of celebrity activism?

Jackson: The classic example of the most effective celebrity endorsement was Oprah Winfrey’s endorsement of Barack Obama during the primary of ‘07-‘08. There were some economists who estimate that that was worth a million votes in the primaries, and that’s pretty meaningful. She also hosted a fundraiser which raised a few million bucks. She wasn’t performing for him per se, but her not having endorsed anybody beforehand probably added to the value of that endorsement. Because once you’ve endorsed a hundred candidates you’re just that guy who endorses.

The difference between performance and endorsement, it mostly goes back to persuasion and energizing the base and trying to get out the vote. At the end, who knows what effect it had? I don’t know if anybody measured it. Jay Z and Beyoncé’s performance was in Cleveland: Ohio went 52-44 against Hillary, African American turnout was down. It’s really hard to say, there aren’t a lot of people in political science still who will actually ask these questions. I guess I should’ve been outside the arena in Cleveland giving out surveys.

Kornhaber: I think some people will be surprised that in the survey you did of likely voters in Ohio, Beyoncé was the most polarizing celebrity, having the greatest negative effect on a general sample. What determines a celebrity’s effect on people politically?

Jackson: I can speak in general terms, I don’t necessarily want to say anything about Beyoncé. Certain factors matter in terms of celebrity endorsements. The first threshold is if the person is recognized and well-known. You would think that that’s a given because they wouldn’t be a celebrity otherwise, but there are plenty of people who are celebrities within specific communities that aren’t elsewhere.

The second seems to be, is the person well liked? If not, the endorsement could be very negative. There are plenty of people who are famous, but people don't have positive feelings towards them. The third threshold that seems to matter sometimes is credibility. Is the celebrity perceived of as actually knowing something about the issue they’re talking about? But credibility doesn't seem to be essential, although a guy like Bono has been perceived as knowing what he’s talking about.

In the survey data, we took an actual statement that Kim Kardashian said, a critique of Obama for not referring to the events of 1915 as a genocide against the Armenian people. Her saying that makes people [in general] less likely to agree with the statement. But if we only look at people who know she is, who also have positive feelings towards her, it switches the direction. We look at a bunch of other celebrities and for almost every one of them it works the same way. Which is to say, overall, either no effect or negative effect, but once you add in recognition and positive feelings, the correlation becomes positive.

I’ve been doing this for 20 years now, and we’re finally getting it down to that granularity. It’s hardly rocket science, but somebody’s gotta prove it the first time.

Kornhaber: What’s the impact on a celebrity when they speak up for a politician?

Jackson: It has an effect. We’re a 50/50 country when it comes to presidential candidates. You pick one and you’ve inherently offended the other 50 percent. I think this time, with Donald Trump’s controversial comments throughout the campaign, the same sort of self-preservation concerns that drove guys like Paul Ryan to vote for the nominee of his party but not endorse him, motivated celebrities too.

Kirstie Allie, not exactly an A-lister anymore, initially supported Trump. And then after the comments on the bus, took it back. So our understanding of the celebrity endorsement process is becoming more sophisticated from a scholarly perspective, and I also think it’s becoming more sophisticated from the actual celebrities’ perspective.

Kornhaber: Does the political split in the electorate naturally also mirror the divides in who regards which celebrities negatively and positively?

Jackson: Probably. What I’ve been hearing a lot of intelligent thought about has been this general sorting going on in the country. Some people refer to it as people putting themselves in bubbles, just a critical way of saying people tend to connect with and be with people who are more like them. I think that comes down to partisanship and ideology, and that then is connected with cultural and entertainment preferences.

When I was 15 years old—who gets attracted to Walter Mondale as an inspirational figure? I did. And I worked on that campaign. And I also loved Neil Young. And when Neil Young seemed to endorse Reagan, it broke my heart. I’m like, “I just assumed we’d all agree on everything.” But increasingly that [agreeing on everything] is the case. People encase themselves in echo chambers of preference-and-bias-affirming friends and culture. Going to a Dixie Chicks concert kind of became a political statement. Republicans don’t go see Michael Moore movies. And maybe that’s why MSBC is so much like Fox News sounded in 2012. Like, “How could this have happened, I don’t know anyone who voted for him!” Exactly.

Kornhaber: Trump is a celebrity. What role do you think that played?

Jackson: [After the election] I was thinking about how people would say, “Celebrity endorsements don’t work.” Well, that’s only if you look at the Clinton side and do so simplistically. That would require ignoring the fact that the winner of the electoral college vote has no public-policy experience, has never held elective office—and probably has the highest name-recognition of any challenger ever because of his experience with New York City celebrity tabloid media and his popular television show. We just elected a celebrity.

It seems like it’s a one-off concept, because how many of him are there? This guy has been in the American mind since the ’80s and ’90s right? To build that level of name recognition and brand recognition, I don’t know who else there is like that. Nonetheless it proves that celebrity status matters.

Kornhaber: But some celebrities are seen as establishment and some aren’t.

Jackson: The guys who voted for Trump aren’t going to listen to Bruce Springsteen. What I mean is, they might listen to Bruce Springsteen’s music but if you tell them who to vote for, it’s just another rich guy.  I mean, Chris Christie has seen him a hundred times in concert. He doesn’t agree with him, obviously, but he probably tears up when he hears “Born to Run.” So somebody is simultaneously able to receive a celebrity’s creation positively and still not be influenced by their political positions.

Kornhaber: It’s interesting because Springsteen did perform for Hillary, but only on the very last day of the campaign, whereas he performed for Obama and Kerry with numerous concerts.

Jackson: You’re right. It kind of personifies the way a lot of people in the Democratic or liberal side of things felt about Hillary. They waited, and then when it became too late people started coming home.

I don’t know if this means anything, but he chose to do “Thunder Road” and I thought he would actually change the last line of that song—“it’s a town full of losers and we’re pulling out of here to win.” That’s a bleak little number. And he also did “Dancing in the Dark” and that’s a bleak little song too. I wonder about the song choice. There’s no way to suggest it had any effect.

Kornhaber: Yeah, it was kind of a solemn performance.

Jackson: It didn’t seem like he had the same glee he had performing for Obama. He was clearly reading the speech he gave in the middle. It was somber. And then one time he turned around and made some extra-loud noise on the guitar to rile up the crowd, and that was it.