Your Brand’s Spokesperson Just Got Arrested — Now What?

Imagine for a second that you work for a big consumer brand — maybe it’s sneakers, or fast food, or high-end wristwatches. How do you persuade people that your product is the one worth buying? No matter how wonderful your sneakers or fast food or watches may be, they are also inanimate objects; they don’t have faces. (Well, a watch has a face but, c’mon, you know what I mean.) So eventually you may ask yourself: what if I hired a well-known actor or comedian or athlete to endorse my brand, to put a face on it? And now, potential customers who may not have noticed your brand are going to be like, Ooh, if they like it, maybe I will too! The practice of celebrity endorsement has been around for a long time. In the 1760s, the English pottery entrepreneur Josiah Wedgewood created one of the first luxury brands after receiving an endorsement from the queen. In ancient Greece, some of civilization’s earliest coins had images of gods and goddesses, like Athena. Who better to endorse a new product like money, which you might otherwise be suspicious of? But what happens if you attach a celebrity to your product and that celebrity messes up?

Larry KING: Police believe that — that O.J. Simpson is in that car. We’ve received reports of a gun in the car. The vehicle is registered to Al Cowling, a former teammate, close friend of O.J. Simpson’s, who has been a fugitive of justice now almost 12 hours.  

O.J. Simpson is just one example of many. Today on Freakonomics Radio: a case study of another endorsement deal that went terribly wrong.

Zab JOHNSON: I mean, this is not just a mistake This is a crime, and this is something horrible.

So what happens then? We’ll try to answer that question.

*      *      *

John Cawley is a professor at Cornell University.

CAWLEY: So I’m a health economist, and I’m really interested in the economics of risky health behaviors — in particular diet, physical activity, and obesity. 

Cawley is especially interested in how health behaviors can be affected by government policy and other top-down solutions. The short answer is not very much. In recent decades, we’ve heard a lot about nudges and tweaks and incentives that are designed to help people eat better, exercise more, etc. But most of the policies that John Cawley and his colleagues have analyzed just don’t move the needle. There was one mild exception.

CAWLEY: One thing that we did find worked is, we conducted an R.C.T., a randomized experiment, of putting calorie labels on restaurant menus. And what we found is that getting the menu with the calories did lead people to ordering three percent fewer calories. So, it’s not gigantic. It’s not single-handedly going to reverse the obesity epidemic. But it’s a cheap policy that had a demonstrable result.  

And this got Cawley to wondering what else he could learn by looking at data from restaurant chains.

CAWLEY: And so starting from that, and thinking about like, well, when would there be cases where people would be potentially attracted to or deterred from patronizing specific restaurant chains? The realization of, oh my gosh, that Jared Fogle scandal that occurred at Subway was definitely something to look at. 

DUBNER: Do you remember where you were, and/or what you felt when you first heard the news about the Jared Fogle scandal? 

CAWLEY: Definitely don’t remember where I was, but, um — I mean, revulsion, definitely. And kind of horror that someone who’d been going to schools and holding up his big pair of pants, trying to be like a role model, would use his money and fame to do such horrible things. This jumped out immediately as something that had the potential to really move demand. 

Okay, let’s back up and fill in some gaps, in case you don’t remember, or never even heard about, the scandal with Jared Fogle and Subway. Let’s start with: what is Subway?

CAWLEY: So Subway’s a fast food chain. It makes sandwiches. They make it to order.

Subway started in 1965, as a single sandwich shop in Bridgeport, Connecticut. It was owned by two families, the DeLucas and the Bucks. Under this same family ownership, Subway grew and grew and grew! — to more than 35,000 stores across 100 countries. In 2023, the original owners sold Subway to a private equity firm called Roark Capital, for about $9.6 billion. Roark also owns chains like Dunkin’, Arby’s, Jimmy John’s, and more. So, that’s Subway. And who is Jared Fogel?

CAWLEY: The Jared Fogle story begins in the late 1990s, when he’s an undergraduate in college. 

This was at Indiana University, in Bloomington.

CAWLEY: He weighed 425 pounds early in his college career, and wanted to make a big change. So he began eating at Subway, twice a day. So what he said is, “I’ll have a six-inch turkey sub for lunch and a foot-long veggie sub for dinner. No condiments, no cheese on either one, baked chips, and a diet soda with each. And I’m just going to walk a lot.” He lost 245 pounds. Someone who knew him as a freshman ran into him when he was an upperclassman, and was just shocked and asked, well, how did this happen? Jared told the story and the guy said, you know, I’m going to tell somebody at the school newspaper. So the school newspaper did a story on Jared Fogle. It then got picked up by Men’s Health magazine for a special they were doing on stupid diets that actually work. And so, both the school newspaper clipping and the Men’s Health article were seen by franchisees of Subway, who clipped it and sent it to headquarters and said, this seems like something we could use. They reached out to Jared and on January 1, 2000, they aired the first Jared Fogle commercial, where he explained that Subway is the healthy fast food, and that he lost 245 pounds by eating Subway twice a day and walking. 

AD: “Here is Jared Fogle. You may have seen him on the news or a talk show. He was inspired by Subway’s great-tasting sandwiches.”

CAWLEY: Later commercials would often show him with his pair of pants from when he weighed his heaviest to illustrate the dramatic change in weight loss.

AD: “Remember Jared from Subway? He’s inspired a lot of people. He’s looking good, to show you the way, his name is Jared, and he’ll lead you to Subway.” 

CAWLEY: They shot about 300 television spots with him, and he also made a large number of personal appearances.

And so it was that Jared Fogle, over the course of a decade, became one of the most prominent celebrity endorsers of his generation — although his case was special: it was the endorsements that made him a celebrity. There was something radically appealing, even sweet, about this combination of authenticity and fame. Jared seemed to unlock something wild in the culture of sandwich-buying. Between 1998 and 2011, Subway went on a tremendous growth spurt; its overall revenues tripled. Subway’s chief marketing officer, a man named Tony Pace, said that between one-third and one-half of Subway’s growth was due to the Jared Fogle campaign. Pace — who later died in a snowmobile accident — said that Jared Fogle was, quote, “woven into the fabric of the brand.” The influential trade publication Nation’s Restaurant News agreed.

CAWLEY: They said the Jared campaign is one of the most successful restaurant campaigns ever.

DUBNER: Now, you told us his story, which was that he was eating a six-inch sub for lunch every day, a foot-long vegetable sub for dinner. You know a little bit about calories, and about caloric intake versus exercise as well. When you heard that story for the first time, that he weighed 425 pounds, and lost 245 in a year primarily attributed to what he was eating, what was your first thought of that claim?

CAWLEY: My first thought is he’s probably walking a lot. But the other assumption is probably previously he was consuming an awful lot of calories. Sticking to this menu very likely could represent a major reduction in calorie intake. 

DUBNER: Do we know what he was eating, and how much he was eating, before the Subway diet? 

CAWLEY: I’ve never heard that.

DUBNER: I mean, it could have been 18 cheeseburgers a day. So, Subway is booming. Subway believes, and the restaurant industry believes, that it’s booming in some substantial part because of this guy named Jared Fogle. And then what happens? 

CAWLEY: Yeah, so in July of 2015, the F.B.I. raided Jared Fogle’s house. 

CNN: Investigators descend before dawn on the Indiana home of longtime Subway spokesman Jared Fogle. 

INSIDE EDITION: What makes the raid on Jared’s home particularly disturbing is that it comes nearly two months after the former executive director of his nonprofit foundation was arrested on federal child pornography charges.

CAWLEY: Jared, when he became relatively wealthy from his income from Subway, he set up a charitable foundation, and he hired someone to run it. The charitable foundation wasn’t really doing anything. It was a way to pay the guy who was heading it up, and what the guy was doing in exchange for the money was — and I apologize, this is gruesome — he was installing hidden cameras in his own house and filming his own children naked. And then also children of his relatives, children of his friends who were over. And then he was sending them to Jared, and Jared would be emailing back and commenting and asking for more and, you know, sharing what he’d like to do. So the F.B.I. got wise to the foundation head, raided his house, seized his hard drive, found the images and found the communications with Jared. They raided Jared’s house. And that was when it just was all over the headlines that Jared Fogle had been arrested for child porn. It came out too that he had crossed state lines to have sex with underage girls. He had actually solicited prostitutes saying, you know, “I want you to get me a young girl, the younger the better.” And so it was really reprehensible stuff. The sentencing judge said that it was “extreme in its perversion.” So it’s just really, really horrible events. And industry observers said this is really bad news for Subway because, they said things like, from now on, when you see a Subway ad, you’re going to be thinking of Jared and what he did to those girls. 

So how did Subway respond to the scandal? Think about how you might respond if your company’s superstar spokesperson was found to have used the money you paid him to set up a fake foundation that fed him child pornography? Subway had long been famous for being publicity-shy; their top executives rarely gave interviews or even appeared in public. And their response to this scandal was predictably muted; on the day that Fogle was arrested, they published a post that said: “Subway and Jared Fogle have mutually agreed to suspend their relationship due to the current investigation.” They also scrubbed all mention of Fogle from their social media accounts. In August, 2015, Fogle pleaded guilty to child pornography trafficking and sexual conduct with minors; he’s now serving a 15-year sentence. The scandal happened under the old, family ownership of Subway. We reached out to the current owners, Roark Capital; they sent a statement that reads, “Our thoughts continue to go out to all of those who were victimized by Mr. Fogle. When we learned of Mr. Fogle’s behavior, we took immediate action, and he has not been associated with the company in any way since his arrest.” But can a company just say, “Hey, stop paying attention to this person we’ve been begging you to pay attention to”? That’s what John Cawley wanted to know. As he considered how to answer this question, he turned to some influential economic research on the notion of repugnance.

CAWLEY: So, Al Roth, who’s a Nobel Prize-winning economist, wrote an article about how repugnance, or visceral disgust, can be something that affects market transactions just as much as prices. He gives examples of how there’s things are outlawed, like the consumption of horse meat or, selling a kidney, or paid sex work. And people just find them offensive, and want them to not happen. And so it really wouldn’t be surprising if people, when they see a Subway, remember the headline they just saw, the CNN story they just saw, about these reprehensible crimes.

Al ROTH: When I think of repugnance, I think of transactions that some people want to engage in, and other people don’t think they should be allowed to, normally for moral reasons.

And that is Al Roth himself. He teaches at Stanford. He points out that repugnance doesn’t always affect markets the way you might think.

ROTH: I find it very hard to predict. And one reason it’s hard to predict is, it’s different in different places. Disgust is easy to predict. If someone spits into your coffee, you won’t finish your coffee no matter where in the world you are. But kidney exchange, surrogacy, prostitution — those are things that are repugnant in some places and not in others. Horse meat — you can order it in some places and not in others.

It’s also worth pointing out that repugnance is a moving target — it shifts over time. Consider slaveholding: today, that is repugnant in most places; in the past, it was the norm. Or, here’s an example of things changing in the opposite direction: life insurance. Not so long ago, it was considered repugnant for someone to profit from the death of a loved one; today, life insurance is seen as something you’d be foolish to do without — or at least that’s the story we’ve been told by life insurance firms. So how did Al Roth think the public would react to the repugnant news about Jared Fogle?

ROTH: It seems like a plausible hypothesis that if the spokesperson were really important, and then he turns out to be a terrible guy, then you might change your mind — in much the same way, but not with exactly the same consequences, that you might associate your feelings about Elon Musk and Tesla and about Donald Trump and Trump hotels. If you don’t like Elon Musk, not buying a Tesla avoids giving him some of your money. But not buying a Subway sandwich, which you may have discovered you liked because of this criminal guy, doesn’t harm him at all. It wasn’t that Subway sandwiches approved of molesting children. They were as much a victim as the general public was. 

And John Cawley again:

CAWLEY: So this seemed to be a great opportunity to test for, to what extent do people — even though Jared had nothing to do with the quality of the food or the management — people might just see that Subway sign and remember him and just turn away and go somewhere else.

So, what did Cawley find? That’s coming up.

*      *      *

The Cornell health economist John Cawley recently teamed up with five coauthors on a research paper. The coauthors are Julia Eddelbuettel, Scott Cunningham, Matt Eisenberg, Alan Mathios and Rosemary Avery. The paper is called “The Role of Repugnance in Markets: How the Jared Fogle Scandal Affected Patronage of Subway.” And what did they want to do with this paper? 

CAWLEY: What we want to do is estimate what was the impact of the Jared Fogle scandal on patronage of Subway. You can’t just go look at a stock price to see how investors responded, because it’s a privately held company, there is no publicly traded stock. You can’t look at quarterly earnings reports, because there aren’t any. And also, you don’t want to just rely on whatever executives of the company say, because you can’t independently verify it. 

DUBNER: So what did you do instead? 

CAWLEY: So what we did is we went to a data set called the Simmons National Consumer Survey. They survey a large number of people, nationwide, and they ask them really detailed questions about the stores, the companies they patronize. And in particular, they ask about their patronage of 58 different restaurants in the past 30 days. They conduct it four times a year. So that allows us to look at changes in relatively short periods of time in the restaurants that people say they’re visiting. 

DUBNER: And these data go back how far? 

CAWLEY: We’re looking at from January 1st, 2014, until December 31st of 2016. So we’re looking over a three-year period, the middle of which is the information shock.  

“Information shock” is economist-speak for, in this case, the burst of news that accompanied the revelation of Jared Fogle’s child-pornography crimes.

CAWLEY: One thing that makes the event we’re studying suitable for study is that nobody knew the bad stuff about Jared. Nobody knew that until the F.B.I. raided his house. And as soon as they raided his house, it was front-page news all over America, and it was confirmed shortly thereafter through a guilty plea. Most of the time, with an ad campaign, it’s hard to measure how much people have been exposed to, but this is a case where really nobody knew anything until one day, like, everybody knew everything. 

DUBNER: John, I have a question that is personal and perhaps a bit intrusive, so feel free to take a pass if you’d prefer. But I once heard you give an interview where you discussed a troubling incident in high school with your debate coach. And I was really taken with your response to it, which was to get through it but then come back and get justice — not only at your own school, but then at the next school where the guy went to. And I was really curious if that experience at all informed the way that you approached this topic with Jared Fogle and Subway. 

CAWLEY: Yeah. So the background is, I had a high school debate coach who molested me and he molested other people. It wasn’t really until college that I was reckoning with it, and then wrote a letter to the school principal and got him fired. He moved to another high school in town. I went to the people who knew the administration of that high school and got him fired from there. And I went to the police and explained what happened, but he never went to jail. It does make me think that we — it’s — this kind of issue is worthy of study, and there’s lots of different ways you can contribute to that. And this is a small way that I’m contributing to better understand what people think, and what they do when they learn information like what Jared Fogle did. 

DUBNER: It did make me wonder if your personal experience contributed to your assessment of what you thought would happen with Subway after the Fogle scandal. 

CAWLEY: Another thing that you’re making me think about is the distinction between being an objective researcher and letting the data speak for themselves and fairly and honestly reporting what you find, the difference between that and the way you wish things worked in the world. I would hope and I think there was, like, universal disgust and anger at this person who betrayed so many people’s trust. That doesn’t mean Subway has to experience lower sales because of it, though.

So did Subway experience lower sales? Cawley and his colleagues did have data from the Simmons marketing survey. To get at the Jared Fogle effect, Cawley would need to measure the Simmons data against a control variable — a different fast food restaurant. This required the use of what researchers call synthetic control.

CAWLEY: Rather than us saying, well, here’s what we think is a good comparison or control firm for Subway and guessing like, DiBella’s or Firehouse Subs because they’re sandwich companies, we can instead use the method of synthetic control and that will go and find the optimal firms that best resemble Subway prior to this shock. In this case, interestingly, it picked just three. It was McDonald’s, Whataburger, and Jack in the Box. So all three coincidentally just turned out to be burger chains. It didn’t pick any sandwich chains. And —

DUBNER: I’m sorry, back up. The control or the metrics on which they aligned with Subway were what, though? Is it population metrics?  

CAWLEY: What we want is a synthetic, a fake Subway that’s as similar to Subway before the shock. But then the question is, similar in terms of what? The standard answer is lagged values of the dependent variable, or, in our case, patronage, that it has similar trends in patronage prior to the shock. But we also matched it on demographics of people who patronize it — so, the percent who are women, low-income, lower education.

DUBNER: So these are what you might call observationally equivalent customer populations, yes? 

CAWLEY: Mhm, yep. Chains with similar demographics among their patrons, and similar levels in trends and patronage. And the funny thing is, like, that’s the feature, not the bug. It doesn’t depend on us using deduction or logic to guess what’s the best control. It lets the data speak for itself. 

DUBNER: Okay, so you start working with the data. Tell me what happens next. 

CAWLEY: Yep. We now have a synthetic Subway that we can compare to regular Subway. And what we look to see is, does the real Subway become significantly different from the synthetic Subway after the shock. 

DUBNER: And what were you expecting, John? 

CAWLEY: I mean, I expected, yes, that people would not want to go to Subway as often or as much in the wake of this information. 

DUBNER: And what magnitude of a drop were you thinking?

CAWLEY: I wouldn’t have been surprised by 5 to 10 percent drop in patronage that eventually went away. Like, maybe it goes away in six months. And we really find that patronage remained flat. There’s no significant change in patronage. 

DUBNER: Can you just summarize this in lay terms? Like, a sandwich chain has an everyday endorser who turns out to be very, very popular. He becomes almost the face of the franchise. He pleads guilty, ultimately, to sex crimes and child abuse. And nothing happens to the product? I mean, it just sounds hard to believe. Did you have a hard time believing it? 

CAWLEY: Well, I mean, there are these two decision-making systems. So there’s system one, which is fast and emotional and may involve repugnance. And then there’s system two, which is more rational and slower to react. I’m certain when people saw the news that they were revolted and they were disgusted and experienced repugnance, but when it came time to choose where to go to eat, they apparently used system two and thought like, well, it’s got nothing to do with the food. It’s got nothing to do with the company, because we’re not finding any effect of this scandal on people’s probability of going.  

DUBNER: Is it possible that this evidence just shows that our personal preferences, including, like, what I want to put in my mouth today, are just much stronger a driver than our sense of — I don’t want to say “morality” or “guilt,” but, you know plainly, a person who was affiliated with this brand did something horrible. I would think that that would outweigh my decision about what I want to eat for lunch today — but it doesn’t. So what’s the takeaway in that regard for you? 

CAWLEY: So you’re right, like, people’s habits around food are strong. It’s really hard to change people’s behavior. We know that from a lot of different studies. But there’s a lot of other research that does document that people do respond to negative information about firms when it’s relevant to the product. For example, when there was adulterated infant formula coming over from China that had killed several children and hospitalized many others, there was a significant decline in exports of all dairy products from China for quite a while. Or when British Petroleum had the Deepwater Horizon oil spill. And it turned out that it was due to their lack of safety precautions, in the U.S. people decreased their purchases of B.P. gasoline. 

DUBNER: Another example you write about is sexual abuse in the Catholic Church. And you write that it reduced membership, at least in the measured area, by three percent — which is, you know, a large number when you’re dealing with something as big as the Catholic Church. But you were thinking that the Jared Fogle effect would be two or three times larger than that. Again, is this all because of the disconnect between who the person was and what the product was?

CAWLEY: So this paper concerned the Boston Globe’s Spotlight series that exposed not just sexual abuse by Catholic priests, but institutional protection of these priests, where the church administration knew what these priests were doing and hid it, and moved them around to help them evade accountability. And so it implicated not just numerous employees of the firm, you could say, but also the management in a very systematic way. And so that may explain why there was a much more dramatic reaction.

DUBNER: So in the paper, you write that “the absence of a detectable impact of the Jared scandal on patronage of Subway raises the question of whether Subway may have been previously overestimating the extent to which Jared was responsible for their increased sales.” Say a little bit more about this, and I’d love to hear you talk as much as you want about how firms will often tell a story that sounds believable, sounds appealing, but is kind of free of empiricism. 

CAWLEY: Yeah. So the fact that we can’t reject the null hypothesis, that this had no impact, like there’s no detectable impact of this information on people’s purchases or patronage of Subway, does make you question — well, wait a second: How could this be true when we were told all along by insiders to the company, by the restaurant industry, that this was one of the most successful ad campaigns ever, and he was the face of it? It is possible that maybe he was incredibly influential and these ads were really powerful and moved demand early in the ad campaign. But maybe by this point, it had run its course. And even though Subway was continuing to pay him, it wasn’t really adding much. But another possibility is that all along, the firm was overestimating how much the ad campaign was contributing to their bottom line.

ROTH: Well, I think that’s right. I think that there’s an incentive for marketing directors to tell their companies that they — the marketing directors — are having a great effect on sales. 

That, again, is Al Roth.

ROTH: And if what you’ve done is hired a spokesperson, then you say the spokesperson has had a great effect on sales. 

We did a two-part series of Freakonomics Radio a while back, called “Does Advertising Actually Work?” The short answer: not nearly as much as advertising and marketing departments say. One paper we cited in those episodes is by the Berkeley economist Steve Tadelis and two co-authors; they used data from eBay. Here’s Al Roth again:

ROTH: Apparently eBay used to buy ads on Google search on the name eBay. So if you searched for eBay, you saw not just eBay’s website but you also saw an ad from eBay. And what the marketing department at eBay told the executive suite at eBay was, “We’re really effective at driving sales because lots and lots of people who click on our ad proceed to buy something on eBay.” And what Tadelis and his colleagues did as an experiment was, they said, “You know, a lot of people who click on the ad, they were searching for eBay. They were intending to buy something on eBay. They just wanted to find the web address so they could click on it. And if we don’t advertise, we’ll get just as much revenue, because instead of clicking on the ad, they’ll click on the organic search result, which will come up first.” So that was a case where there was a claim but no evidence that the ads were leading to the purchases. And that could well be the case with the spokesperson. 

It may be that academic researchers are suspicious of the power of advertising, and of spokespeople — but most companies don’t seem to be suspicious at all. Take a look at advertising during the Super Bowl, which is easily the biggest annual TV event in the U.S. — and, therefore, the biggest annual advertising event. Over the last four years, between 60 and 75 percent of Super Bowl ads featured at least one celebrity, and usually more than one. So, if you believe that celebrity endorsements do work, the question is why?

*      *      *

Even though academic researchers are skeptical of the power of celebrity endorsement, the practice rolls on, despite the significant cost to the firms. LeBron James signed a lifetime deal with Nike that could pay out an estimated $1 billion. Taylor Swift has endorsed Diet Coke, Apple Music, Keds, AT&T, CoverGirl, and she reportedly almost reached an agreement to endorse the cryptocurrency exchange FTX, for a reported $100 million. And for nearly 20 years, starting in 1975, Hertz Rental Cars ran what was considered an exceptionally successful marketing campaign featuring O.J. Simpson, dashing through airports, and wowing the spectators.

KIDS: “Go, O.J., Go!” 

SINGERS: Getting you in, getting you out with super speed. Hertz: the superstar in rent-a-car.

Hertz took Simpson off the air once he was charged with murdering his ex-wife and her friend; at the time, he was being paid more than half a million dollars a year.

Zab JOHNSON: Celebrity endorsements are extremely expensive.

And that is:

JOHNSON: Elizabeth Zab Johnson. I am the executive director and senior fellow with the Wharton Neuroscience Initiative at the University of Pennsylvania.

DUBNER: Where does Zab come from? 

JOHNSON: The middle three letters of Eli-zab-eth. There are many Elizabeth Johnsons in the world, but there is only one Zab Johnson. 

DUBNER: As far as you know. 

JOHNSON: As far as I know. Actually, I know that there’s a second Zab Johnson, and she’s a developmental psychologist, and I’m pretty sure she Googled her name and found my nickname and liked it. 

DUBNER: Oh no.

JOHNSON: But I’ve never called her out on it.

DUBNER: You’re doing it right now.

JOHNSON: I am.

Johnson teaches in the marketing department at Penn, and she has studied celebrity endorsements. As she was saying, they are expensive.

JOHNSON: And they are inherently risky because you’re associating your brand with a person that may have behaviors that don’t all match up with your values and your idea of what the brand represents. 

DUBNER: So, “expensive” and “risky.” Those are two pretty strong negative words. And yet there is a multibillion-dollar industry in celebrity endorsement. So who is not buying what you are selling, and why not? Why do so many people proceed with it, despite those risks?

JOHNSON: Social hierarchy and social status matter to us as humans, and as primates. And so we want the things that high-status individuals have. We’re in a really fractured, noisy advertising environment. So you want to create memorable content, you want to create things that people will go back to and view again and again. 

DUBNER: But there are other ways to do all that without a celebrity. Why is the celebrity the first choice? 

JOHNSON: I think there’s some expectations around it. You know, you’re not just watching the game. You’re coming to watch the commercial. Why are you watching the commercials? Because you also want to see the celebrities and see what — you know, again, that sort of driving that fundamental biological need that we have.

DUBNER: So I don’t know how closely if at all you followed the breakup of Kanye West and Adidas. They had this hugely successful partnership — his Yeezy sneakers brought in something like $2 billion a year, roughly 10 percent of their revenue. But he’d always been a complicated partner, let’s say, and his behavior got more erratic, more inappropriate, including some very public anti-semitism. So Adidas finally broke things off with him, but they had, I guess, millions of pairs of sneakers already manufactured when they broke up. I guess they could have just burned them or something, but instead they decided to sell them and donate the proceeds to charities. What’d you think of that solution?  

JOHNSON: Their decision to release and sell but not make profits or be obligated to give Kanye any of the revenue — is probably a positive one. Adidas realizes that right now consumers really need them to espouse social values that resonate with younger people, and with the consumer at large. So I think the idea of burning shoes and creating environmental pollution from that process, as well as waste, I think was probably too big of a burden for them. So this was the right solution. I think it’s a great solution, actually. 

DUBNER: On the other hand, you’re wearing, or at least owning, the product of someone who’s been disgraced, at least in the eyes of many, right? 

JOHNSON: I mean, with every product that we consume, there are always those things at play. With shoewear in particular, we know that that those are mostly being manufactured in environments that I don’t think we think are the right manufacturing environments for modern-day consumers. But at the same time, we still will pay hundreds of dollars for them. 

DUBNER: I don’t know if you’ve ever visited Auschwitz, but there are these collections of things that were taken from the people who were sent there, and then usually killed there, most of them Jews, but not all. These gigantic piles of eyeglasses, gigantic piles of suitcases, and gigantic piles of shoes. I wonder, do you think there might have been something fruitful to do with those, whatever, million or however many Yeezys there were? Some kind of public sculpture that might have been a better solution than going ahead and selling the thing that was made by the person that you are supposed to be distancing yourself from? 

JOHNSON: That’s an interesting idea. I mean, I think Adidas had to be really careful because the founder had been part of the Nazi Party, and it was a point of shame for the company. There could have been a huge backlash to something like that. 

DUBNER: A little bit too on the nose.

JOHNSON: Too on the nose, right? And, like, not at all the same, right? 

DUBNER: Not at all the same.

JOHNSON: So I think that this is probably the better move. 

DUBNER: I guess this is why I’m not in corporate marketing.

JOHNSON: And maybe better than sending Yeezys to underprivileged children in developing nations, that could also be seen as being bad. I think in a sense, maybe this was the right call. 

And yet despite the risk of celebrity endorsements, and despite the expense, there are significant upsides. That is the conclusion that Zab Johnson and three co-authors reached in a recent study called “How Celebrity Status and Gaze Direction in Ads Drive Visual Attention to Shape Consumer Decisions.”

JOHNSON: In this particular study, we used Snoop, Scarlett Johansson, Michael Jordan, David Beckham.  

They also used ads that didn’t have celebrities, in order to track the difference. Some of the ads would show the celebrity looking toward a snack food they were endorsing.

JOHNSON: Cheez-Its, Nutri bars, Snickers, Oreos.  

By tracking eye movements, Johnson could tell whether research subjects followed the celebrity’s gaze and looked at the product — or whether they were so fixated on the celebrity that they ignored the product.

JOHNSON: In the psychology literature, that’s called the Vampire Effect — the celebrity overshadows the product. If that’s true, then it’s not going to work, right? That could be a really big disaster.

DUBNER: Okay, so what effect did you find when consumers look at a celebrity who’s looking at a product? 

JOHNSON: So we found that even though people spend less time looking at the product, it builds confidence and surety around the product just by being paired with a celebrity. And that was enough to nudge people to make choices of snack foods that they had felt on the fence about towards the one that had been presented with a celebrity. 

DUBNER: And could I take that finding, and extrapolate from it the idea that celebrity endorsement, quote, “works”? 

JOHNSON: I mean, this is a single intervention. This was the single presentation of an advertisement for four seconds. So if you think about how many times where we’re actually visually presented with this kind of information and that little bit of confidence, if you can think about that as an aggregation over time, it is probably going to be pretty successful. That’s my take.

Now, this was one small study by one neuroscientist. The economist John Cawley walked us through some of the other literature on celebrity endorsements.

CAWLEY: Craig Garthwaite at Northwestern has done a neat paper on the effect of an endorsement of your book by Oprah’s Book Club, and found that it led to a jump in sales not just of the recommended book, but also of other books written by that same author. And interestingly, it didn’t increase book sales. People didn’t buy any more books. They just switched to buying the one Oprah recommended. And then — I guess Craig really is a fan of Oprah — he wrote another paper about Oprah’s endorsement of Senator Obama for president, and estimates the impact that that had on his votes and contributions and voter turnout. He estimates it got Senator Obama an additional million votes.

DUBNER: To be fair, Oprah is an outlier as well because she’s so huge. Additionally, with her book recommendations, there’d never been anything like that in the history of book recommendations, at least in modern history. And it wasn’t just an endorsement. It was a call to action, to everybody, “go buy this book.” What do we know about the more general state of endorsement? Let’s say LeBron James is endorsing Nike. What do we know about the power of that type of endorsement where there’s some affiliation but not necessarily a call to action?  

CAWLEY: It’s a great question. I don’t have a good answer for you. You know, the discussion that you had in your earlier episodes about, does advertising really work — the work of Anna Tuchman finds very little increase in sales of e-cigarettes due to T.V. advertising. The work of Steve Tadelis finds a negligible return to paid search ads for eBay. Brett Gordon has found that firms’ rule of thumb for guesstimating the payoff to ads often vastly overstates what the true payoff actually is. 

DUBNER: There were explanations in that series that we did, one of which came from Steve Levitt, which is the people responsible for soliciting the budget for marketing and advertising are the same people responsible for measuring and making claims about the efficacy. Is that the best explanation for why there can be a gap between the two?  

CAWLEY: I think it’s one possible explanation. I also think of public health. So, a lot of the work I do is in health economics. And very often people in public health will be very well-intentioned and will propose some intervention, and they claim we’re seeing amazing results in pilot studies. There’s a sense of like, aren’t good intentions enough? And I could imagine something being similar with marketing of like, how couldn’t this work? This is what generations of ad people have done. But again, this is how science progresses. This is how we make sure we’re not wasting resources, is we need to rigorously evaluate everything we’re doing. 

DUBNER: So is Jared Fogle a good story, or example, about why it makes sense for firms like Geico to use an animatronic or whatever, clay gecko for a spokesperson? I mean, anybody you hire, potentially, is risky.  

CAWLEY: That’s a really good point. I mean, another example is like, universities have gotten very careful about naming any building after a living person. So when someone is deceased and you know everything that they’ve done, pretty much, then you can make that decision. But you don’t want to be on the hook if — you know, there’s a lot of Sackler wings of museums out there that since the OxyContin opioid scandal that are probably causing a lot of regret. 

DUBNER: I’m curious, when you look back, were there other firms in the wake of the Jared Fogle scandal who cut loose endorsers that perhaps they may have been quietly worried about? 

CAWLEY: Oh, wow, that’s a good question. I don’t know of any because of Jared specifically, but like, Kanye West, when he was making really hateful comments, he lost a lot of sponsors. I mean, Tiger Woods is another example of where, when he crashed his car under the influence and — 

DUBNER: What about Lance Armstrong? How quickly was he cut loose? 

CAWLEY: Oh, that’s a great example.  

DUBNER: Michael Vick, Oscar Pistorius. I mean, it’s a pretty long list, really, of celebrities who do terrible things. Have you looked at other endorsers gone bad? 

CAWLEY: No. I think if we can find another example of where a person had such an incredible turnabout in reputation and is so closely identified with one single brand in the public mind, then we’ll jump on it. But a lot of these cases, there’s just celebrities famous for something else who’ve endorsed 100 different things. 

DUBNER: So what about you, John? If you had to be a brand spokesperson for any company, any product, what would it be? 

CAWLEY: Oh my gosh. Well, you know, I’d be happy to be the face of the Freakonomics podcast and the Freakonomics book empire. So, keep me on speed dial.

Huh. Is that a risk I’m willing to take? John Cawley seems like a good guy but, I dunno. After what I learned today about the science of celebrity endorsement, I think it’s a pass. But I do appreciate the offer, John. I would also like to thank John for the good conversation we had today, along with Zab Johnson and Al Roth. I’m curious to hear your views about celebrity endorsement; send us an email: we’re at radio@freakonomics.com.

*      *      *

Freakonomics Radio is produced by Stitcher and Renbud Radio. This episode was produced by Zack Lapinski. Our staff also includes Alina Kulman, Augusta Chapman, Dalvin Aboagye, Eleanor Osborne, Elsa Hernandez, Gabriel Roth, Greg Rippin, Jasmin Klinger, Jeremy Johnston, Julie Kanfer, Lyric Bowditch, Morgan Levey, Neal Carruth, Rebecca Lee Douglas, Sarah Lilley, and Theo Jacobs. Our theme song is “Mr. Fortune,” by the Hitchhikers; our composer is Luis Guerra.