The Three Hardest Words in the English Language

Hey, it’s Stephen Dubner. Our new book is out, Think Like a Freak, and we want to talk about it with you. So we are forming the Think Like a Freak Book Club. How’s it work? You send in your questions about the book and we’ll answer them on this podcast. There are nine chapters in the book, so why don’t we do the first installment of our book club with chapters 1, 2, and 3. Those are “What Does It Mean to Think Like a Freak?,” “The Three Hardest Words in the English Language” and “What’s Your Problem?” Just send your questions to radio@freakonomics.com with the subject line “Book Club.” If we choose your question for the podcast, we’ll send you a signed copy of Think Like a Freak or a limited edition Think Like a Freak t-shirt. Thanks!

Stephen J. DUBNER: Hey, Levitt, can I read you a story?

Steve LEVITT: Sure.

REPORTER: OK, so I am going to tell you a story.

DUBNER: A little girl named Mary goes to the beach with her mother and brother.

REPORTER: A little girl named Mary goes to the beach with her mother and brother.

REPORTER: They drive there in a red car.

DUBNER: They drive there in a red car.

REPORTER: At the beach.

DUBNER: They swim.

REPORTER: Eat some ice cream.

DUBNER: Play in the sand.

REPORTER: And have some sandwiches for lunch. So that’s the end of the short story.

DUBNER: Now, Levitt, let me ask you a few questions about it.

REPORTER: So now I’m going to ask you the questions.

KIDS: [Giggle]

REPORTER: Ready?

DUBNER: Number one. What color was the car?

LEVITT: I think it might have been red.

REPORTER: What color was the car?

KID: The car was red.

REPORTER: Fantastic!

DUBNER: Number two: Did they have fish and chips for lunch?

LEVITT: No, they had sandwiches.

KID: Uh… yes they did have fish and chips for lunch?

KID: No. They do not have fish and chips for lunch.

REPORTER: What do they have?

KID: Sandwiches.

KID: They had sandwiches.

DUBNER: Number three: Did they listen to music in the car?

KID: No they didn’t.

LEVITT: I’m sure they did. Doesn’t everyone listen to music?

REPORTER: Did they listen to music in the car?

KID: No. They didn’t listen to music in the car?

DUBNER: And number four: Did they drink lemonade with their lunch?

KID: No, they didn’t drink lemonade with their lunch.

LEVITT: Probably not. Most people drink soda.

DUBNER: [Laughs]

REPORTER: What flavor ice cream did Mary have?

LEVITT: I don’t know.

KID: Mary had chocolate flavored ice cream.

REPORTER: And what flavor ice cream did Mary have?

KID: Mary had vanilla ice cream.

REPORTER: And what flavor ice cream did Mary have?

KID: (long pause)

REPORTER: You can say, “I don’t know.”

KID: Strawberry?

ANNOUNCER: From WNYC: This is Freakonomics Radio, the podcast that explores the hidden side of everything. Here’s your host, Stephen Dubner.

Steve Levitt and I have just published a new book, called Think Like a Freak. One chapter is called “The Three Hardest Words in the English Language.” Now, what do you think those words are? In the past, some may have argued that people have a hard time saying “I love you.” I don’t think so. People running around saying, “I love you all” the time – to friends, to pets, to strangers. So what are those three most difficult words?

DUBNER: So Levitt, one piece of business advice that people give all over the Web… LinkedIn, and just read all of the regular kind of job search and career advice places, one piece of advice they give is that when you are in a job interview there is one thing you should never say to the person interviewing you, which is “I don’t know.” You should never admit that you don’t know the answer to a question. Do you think that is pretty good advice?

LEVITT: That actually probably is good advice for an inner Jew —

DUBNER: Did you just say inner Jew?

LEVITT: [Laughs.] That probably is good advice for a job interview. But it is terrible advice for life…

DUBNER: And why is that such terrible advice for life?

LEVITT: Let me try to answer that by example. So we wrote Freakonomics. And because it was deemed a business book and we sold a lot of copies that made us business experts. And since we wrote that book I’ve been asked a ton, you as well, to go talk to companies and give them advice. And what’s amazing to me, and I’ve kept very close attention over the last decade almost, is I could count on one hand the number of occasions in which someone in a company, in front of their boss, on a question that they might possibly have ever been expected to know the answer, has said, “I don’t know.” Within the business world there’s a general view that your job is to be an expert. And no matter how much you have to fake, or how much you are making it up that you just should give an answer and hope for the best afterwards. And I have seen it teaching the business school students, that they are incredibly good, the MBAs, at faking like they know the answer when they have no idea. But that is so counterproductive. I mean think about it. It might keep your job for another week or another month, it might make people think you are good, but that’s not the point. I mean what fun is life if all you do is go through life trying to fake that you are something that you are not when really the goal is to be good, and to improve, and to learn, and to make things better? And the only way to do that is to start by saying,”I don’t know.” And when I don’t know, I don’t know.

All right, so it’s easy to ridicule the fakers and bluffers of the world – but we can understand their motivation, can’t we? Everyone wants to look smart. Look competent. Look like we belong. But wouldn’t it be even better if – as Levitt says – to improve, to learn, to make things better? If you think so, listen on. Let’s try to figure out where this reluctance to say, “I don’t know” comes from…

DUBNER: Hi, Amanda.

Amanda WATERMAN: Hey Stephen…

Amanda Waterman teaches developmental psychology at the University of Leeds. A lot of her research is about people’s willingness – especially children’s willingness — to say, “I don’t know.” Her early research looked at whether kids would say, “I don’t know” when asked a patently ridiculous question.

REPORTER: What do feet have for breakfast?

KID: Toenails if you chop them off.

KID: Smelly socks.

REPORTER: You can say, “I don’t know.”

KID: They have your… they have your skin.

KID: They have a walk for breakfast.

KID: Shoes and socks for breakfast.

REPORTER: Ok, next question: Is red heavier than yellow?

KIDS: [Giggle]

REPORTER: Is red heavier than yellow?

KID: No, because yellow is my favorite color.

REPORTER: Bella, what do you think?

BELLA: I think it is heavier than yellow, because… because, because, because… red is in hell, and then yellow is in the sky, isn’t it? Yeah, so it is down below. Down there.

REPORTER: Is a jumper angrier than a tree?

KID: No.

REPORTER: Is a jumper angrier than a tree?

KID: No, because trees look like they have faces.

KID: Yes, a jumper is angrier than a tree because you can pick jumpers up and that’s how they move. And trees just go like this… they sway from side to side in the wind… whoo.

DUBNER: So, Amanda, is a sweater—what you there call a ‘jumper’—angrier than a tree?

WATERMAN: OK, well, that’s perhaps where this research started, where we asked children very silly questions about whether a jumper or indeed, a ‘sweater,’ is angrier than a tree, to see whether they would indicate in any way if they were confused by the question, or whether they would just go along with the fun spirit of the question. And we found that often children did want to decide whether a jumper was angrier than a tree. But we moved on after that. I was more interested in looking at whether —

DUBNER: You got bored by that stuff, didn’t you?

WATERMAN: I got bored. Jumpers are obviously more angry than trees, right?

DUBNER: Oh, really? I thought trees.

WATERMAN: No no… Jumpers are definitely angrier, because I think they are kind of used by people. They have to be put on and off all the time. But we were more interested in looking at perhaps something that would be a little more applicable to real life, because generally adults don’t ask you very silly questions all the time. But you do come across questions to which you do not know the answer. So it might be a perfectly sensible question, but you just happen not to know the answer to it. So the research went on from “silly questions” to what we’ve called “unanswerable questions.”

DUBNER: Which brings us back to the story you heard earlier. About the little girl named Mary…

REPORTER: A little girl named Mary goes to the beach with her mother and brother…

This story was part of a study that Amanda Waterman ran to see how children will answer a question that is essentially unanswerable. A question to which the most sensible answer is, “I don’t know.” In the story about Mary and her family going to the seashore, we learn several facts: The car was red… they ate sandwiches for lunch. But other facts aren’t given to us. Did they listen to music in the car? We aren’t told that. Did they have lemonade with their sandwiches? We don’t know.

REPORTER: Did they drink lemonade with their lunch?

KID: No, they didn’t drink lemonade with their lunch.

REPORTER: Did they listen to music in the car? You can say I don’t know.

KID: [Long pause] No. They didn’t listen to music in the car?

DUBNER: The vast majority of children pretend they know the answer to unanswerable questions.

WATERMAN: It has varied between studies…but you’re looking at maybe two-thirds to three-quarters of children—and we’re talking in the age range here of about five to eight years old—would say, “Yes” or “No” to a yes/no question that we know they don’t know the answer to.

So these kids are preparing themselves very well for careers in business… or politics… maybe as a pundit of some sort… or wherever else in modern society people pretend to know more than they do. Why is this impulse so strong among children?

WATERMAN: Personally what I think what’s going on is that the children are very used to a classroom situation. Obviously, you do get asked questions by your parents, as a child. But a lot of question/answer exchanges happen in school. Teachers ask kids questions all the time to ascertain what they’re picking up in the lesson, et cetera. And generally, in that situation, it’s sensible to have a go. You don’t generally earn many points for saying, “I don’t know” when a teacher asks a question. So, having a go and actually feeling good that maybe you have that piece of knowledge and you can show that you are clever. So I think that when it seems like a test situation, they fall back on what they’ve already started to learn from the context that they spent a lot of time in.

DUBNER: I guess when you describe that as an academic setting, even an academic setting among children, it reminds me certainly of academic settings around older people—university and so on. But it also reminds me of settings in the workplace, where very often when a question is raised, someone is expected to know the answer. And maybe the asker of the question, if they don’t necessarily know the answer, they certainly have strong opinions about it. So I’m curious about how you would connect what you see in young children with what you see among the rest of us, adults, whether it’s in the work world or politics or policy, things like that.

WATERMAN: Well, I think that ties back into this whole social aspect of any kind of conversation or dialogue or question/answer exchange, where a lot of how you respond is affected by how you perceive the expectations of the person asking the question. So it’s not just about giving a straight factual answer as to the state of your knowledge. It’s actually trying to work out all the other contexts that surround why that question’s been asked, what you think the interviewer wants from you. And I do think children do start to do that from a very young age. I think one of the things that happens with kids, but you can actually, definitely apply this to adult conversations, is that there’s quite a power differential between a child and an adult who’s asking them a question. And the child will feel that the adult is the person that has more of the power, and therefore perhaps they feel slightly disadvantaged. They feel like they want to show what they can do, and they don’t feel as comfortable admitting when they don’t know something. That clearly applies across, into the work context with adults, where you have a boss with an employee. Or where you have somebody who’s higher up the hierarchy than you are—and you feel that expectation that somebody should know the answer to this.

As it turns out, Waterman also tested some adults with unanswerable questions. How’d they do?

WATERMAN: We found that about a quarter of the adults tried to answer the yes/no unanswerable questions.

Well, at least that’s better than the three-quarters of the kids who couldn’t bring themselves to say, “I don’t know.” But still: not very encouraging. Think about it for a minute. When’s the last time someone you know – maybe even yourself – gave a firm answer to a question that you really didn’t know. A week ago? An hour ago? Maybe… just now? Coming up on Freakonomics Radio: the consequences of not being able to admit what you don’t know:

LEVITT: It was easier to spend a billion dollars pretending they knew than to actually find out the answer.

DUBNER: And what we can do about it:

LEVITT: There’s only one way to learn, and that’s through feedback.

ANNOUNCER: From WNYC: This is Freakonomics Radio. Here’s your host, Stephen Dubner.

One of the first steps in learning to think like a freak is learning to say, “I don’t know.” Why? Because until you can admit what you don’t know, it’s virtually impossible to learn what you need to. Because if you think you already have all the answers, you won’t go looking for them. But let’s be clear: Simply saying “I don’t know” isn’t a solution. It’s just a first step. You have to figure out what you don’t know – and then work like a dog to learn. And how do you learn?

LEVITT: There’s only one way to learn, and that’s through feedback

That’s Steve Levitt again.

LEVITT: The thing about always faking is that if you fake like you know the answer, you don’t have the freedom to explore other possibilities. But if you actually care about the outcome and the truth, saying, “I don’t know” is critical. One thing we’ve learned is that the only way to learn is through feedback. That whether you’re a human being, an animal or an organization, the way that you learn is by trying different things and seeing the outcome when you try different approaches, and comparing those outcomes.

Feedback comes in many flavors. It might mean going out and asking a lot of questions. It might mean gathering a lot of data. If the data don’t happen to be just lying around, waiting to be plucked, you may have to generate data yourself. How? One great way is to run some kind of experiment. If you want to think like a freak, you need to fall in love with experimentation. Now scientists have been doing it for centuries. The rest of us? Not so much. But why should scientists have all the fun?

LEVITT: So, something that puzzled me for the longest time when I started working with firms is that I would preach to them the gospel of experimentation, of either randomized experiments or accidental experiments, just all the things that I do. And always it was met with great enthusiasm. Of course, yeah, you know, experiments — that’s the way to go. And then when it actually came down to it we could almost never convince anyone to do things experimentally.

DUBNER: Why don’t they want to do it? What are they reluctant about? Are they afraid it is going to be expensive? That they won’t know how to do it? That it is going to embarrass them? What’s the problem?

LEVITT: For years, I struggled with trying to understand what it was. And I thought about, was it expensive? Well, no, I could design it in ways that it really wouldn’t cost them anything. “Would it take too long? Too many resources?” I mean, it was none of that. And it finally hit me what was totally obvious. It is that in order to run an experiment you have to start from the premise that you don’t know what the answer is. So I don’t know if I should do the red version or the blue version, should I price high, should I price low, should I give people a money back guarantee, or shouldn’t I? So, all of those are questions that might be perfect for trying to answer with an experiment. But in order to be willing to run the experiment you have to say, “Well, geez, I don’t know the answer.” Wait, but you are the expert. You are in charge of marketing. You are in charge of pricing. You are supposed to know the answer. And so the people in the firms we were working with felt so much pressure to be the expert that they wouldn’t run the experiments because that was an admission that they weren’t really as expert as they maybe had been pretending to be or as other people wanted them to pretend to be.

Levitt, in addition to teaching at the University of Chicago, helps run a consulting firm.

LEVITT: One of my favorite early stories… this might have been the first company that I went to visit.

This was a company whose unwillingness to experiment, to admit what they didn’t know, had pretty big consequences. It was a huge multinational retailer. They spent many millions of dollars in advertising in the U.S. alone. But they weren’t sure how effective the advertising was. So far, they had come to one concrete conclusion: TV ads were about four times more effective, dollar for dollar, than print ads. How’d they know this? That’s what their sales data told them. They advertised every week in newspapers. But because TV advertising is so expensive, they only bought ads three times a year: Black Friday, Christmas, and Father’s Day.

LEVITT: So then I said, “I think I might understand why your TV ads seem to have a return on investment that is four times as great as your newspaper ads.” And they said, “Why?” And I said, “Well, every single time you go and do a big advertising campaign, your sales go through the roof right afterwards.” And they were like, “Yeah, Father’s Day, Thanksgiving and Christmas.” And I said, “Don’t you think maybe if you didn’t do the ads you would sell stuff around Christmas and on Black Friday?” And they said, “Yeah!”

DUBNER: So basically this company is spending hundreds of millions of dollars on TV advertising to persuade people to go to their stores to buy stuff on the three times of the year when people were already going to be going to stores to buy stuff.

LEVITT: Exactly right.

DUBNER: So what about the print ads? This retailer had run newspaper ads pretty much every Sunday for 20 years in more than 200 markets. How effective were they? It was hard to know.

LEVITT: So I go to them and I say I think you cannot learn from the existing data the answer to your question. And to their credit they said, well, geez this is a really important question for us. How could we learn the truth. And I said, well, TV ads are tougher, I don’t know how we will ever know that, but the newspaper inserts, that’s an easy one. So you said you have a couple of hundred markets and I did some calculations about how much power we would have if we did some randomized experiments and I think I would need you to give me about 40 of your newspaper markets, and I would randomly assign 20 of those to a control group where we wouldn’t do any newspaper inserts for about three months and we will compare the sales in the markets where you kept on doing the inserts to the ones where you didn’t do inserts, and by doing that comparison we would learn with a great degree of confidence what the impact of newspaper inserts was on sales.

DUBNER: And did you expect them to say, “Hey, that’s a fantastic idea, and just think of all the money we will save by not buying the ads, that’s even cheaper,” did you expect them to be enthusiastic about it?

LEVITT: They were spending hundreds of millions of dollars a year on these newspaper inserts. They had no idea if they worked. And they even had this belief, which I am almost certain was erroneous anyway, but they had reason to believe that the newspaper inserts weren’t that good relative to TV. So I thought for sure this is a no-brainer. This is easy. Of course. Let’s get rolling. How fast can we get it started?

DUBNER: And what was their response?

LEVITT: So we went to them with the proposal and their reaction surprised me. They said, are you crazy? We can’t not advertise in 20 markets. We’ll all get fired.

LEVITT: And they went on… one of the other guys chimed in and said, yeah, we once had this summer intern, an MBA that we had hired, and one of his jobs was to do the ordering of the media in different cities, and they guy was so incompetent that for reasons we never really figured out, he just forgot to order the newspaper inserts for a big chunk of Pittsburgh for the entire summer.

LEVITT: And just kind of joking around with them I said, “Well I hope you hired that guy.” And they said, “Hire him? He was terrible. We would never hire him. Why should we hire him?” And I said, “At least he was terrible in just the right way that you are all too chicken to be terrible.” Because, I said, “What happened to the sales in your stores in Pittsburgh when this guy forgot to do the newspaper inserts?”

DUBNER: So…what did happen?

LEVITT: Well, to their credit they went and they looked at the data and they called me with a spring of excitement in their voice and they said to me, “You are not going to believe what we found. We found no impact whatsoever on the sales in Pittsburgh when this guy didn’t put in the ads.” And I said, “Wow, that’s amazing, alright, so when can we start the experiment. Again, I just need 40 markets and I’ll randomly assign them and we’ll stop the ads for three months.” And they said, after a pause they said, “Are you crazy? We can’t not advertise in 20 markets, the CEO will kill us.”

LEVITT: And it is truly amazing that the only piece of real learning they had, the only piece of real feedback that they were given is that the newspaper inserts didn’t work. And yet, it’s now five years later, and they’ve probably spent close to another billion dollars putting ads in the newspaper and the reason was because if they tried to do an experiment to see whether newspaper inserts worked or not, it would be an admission that they did not know whether newspaper inserts worked. It was easier to spend a billion dollars pretending they knew than to actually find out the answer.

DUBNER: It’s easy to criticize people who get locked into pretending to know answers that they don’t. But as we said at the beginning of this program – those three little words – “I don’t know” — can be hard to say.

LEVITT: First, it takes a lot of confidence to be willing to say I don’t know, and I think actually it is easy for someone like me to say I don’t know, because, you know, I’ve had some success, we’ve sold some books, I’ve done well in academics, so if I say I don’t know, people say, “Oh, isn’t that cute, the Midwestern, self-deprecating, blah blah blah.” But if you are just, you know, a middle manager at a firm and you say, “I don’t know,” people look around and are like, this guy doesn’t know? What’s wrong with him?

So the incentives to fake it – to never admit you don’t know an answer – are pretty strong. But maybe there’s a way to flip those incentives. To encourage people to say “I don’t know” when they really don’t know. Remember Amanda Waterman, the British psychologist who studies kids? We went back to her with this.

DUBNER: So, in our new book, we make the argument that people should be much more willing to admit what they don’t know. Especially in settings, whether it is business, or academia or politics, where you are dealing not just with a simple set of facts, or even a simple set of beliefs, but with a process or, you know, when you are trying to tease out cause and effect which can be very complicated, not only… certainly in the future, what policy will work, but even trying to understand what has worked in the past, and we make the argument that it would be better if more people admitted what they don’t know because that is kind of the first step on the road to learning and discovery and getting feedback and running some experiments, perhaps, to find out what’s real. I’m just curious whether you would endorse that argument of ours or whether you think it is not really a principal obstacle.

WATERMAN: No, I absolutely endorse it. And I would go back a little to saying that I guess at least part of the responsibility lies with the people who are the top of whatever hierarchy you happen to be talking about. If you create a culture where it’s OK to say,”You know, I don’t know the answer to that, but wouldn’t it be good to find out?” then that then perpetuates through the organization. But I think the problem is, I guess, that the people that’ve come up to the top are maybe used to having got to the top possibly by not admitting when they didn’t have the answer. So you get that cycle. But no, I absolutely agree with you in that saying you “don’t know” appropriately is a good thing.

So let’s break the cycle, shall we. Let’s start with the generation that really matters…

REPORTER: You can say I don’t know.

KID: Uh….

KID: I don’t know.

KID: Uh, I don’t know.

KID: I don’t know.

KID: I don’t know.

KID: I don’t know, miss.

Kid: I don’t know.

REPORTER: That’s perfectly fine.