Uri Gneezy: The Art of Incentivizing People

26:01Right. So$500 is a lot of money. But imagine that you go to Office Depot to buy a mouse and it costs$50 and the guy tells you, if you walk 10 minutes, you can get it for 50 % off. You can get it for$25. Many of us will do that. Now, imagine that you're buying a laptop and you want to buy also a mouse and he will tell you this. The laptop costs$20 ,000. Sorry,$2 ,000. That would be a very expensive laptop. Yeah,$2 ,000. and he'll tell you, well, you can get$25 off. If you'll go 10 minutes, he'll say, I'll stick with it. So you look at the$25 as part of the deal. And the same is true for this discount.

26:45So I worked with Edmond, the car online company that does this, and they wanted people to go to the dealership through their website. So they gave, say,$500 for a$20 ,000 car. Now,$500 discount is a lot of money, but not relative to the$20 ,000. When we gave them, instead of$500, we gave them$500 to fill up their car with gas. Now, that's a lot of money because you can imagine yourself at the pump for a few weeks, maybe even a few months if you don't drive that much, just using this. And that would be a great feeling. So this kind of mental accounting, that's what we call it in behavioral economics, could really make the incentives much more powerful.

27:27The same amount of incentives more powerful if you just frame them as gas cards. And in this real world situation with Edmonds, let's say the way they give you that$500 is they give you a MasterCard debit card that has$500 stored on it. There's also going to be breakage, right? So people are going to lose the card, not use the card, whatever. So that's even better, right? Well, that's a cynical view of it, but that's definitely what you see, for example, with rebates. So people go and they buy something and get, say that you buy this laptop that you talked about before and they get$200 rebate.

28:07For the rebate, they have to go home, go online, do something, maybe even mail something to you. You see that a very small fraction of the people actually do it. When you buy the laptop, you look at it, oh, it's$1 ,800. Well, I'll pay$2 ,000, but I'll get the$200, everything is okay. Then you go home and you forget about it. So a large part of that business is actually based on that. But I think that it's powerful beyond this point that people really care about some things more than others. Money is not fungible. In the section about entrepreneurship and innovation, you talk about the positive effects of try a lot of things, try them fast, rewarding people for risky innovation.

28:50But my interpretation of what you read is, so what happens if you have this unintended consequence of fostering innovation and quick failures and all that? What if people in the company start saying, let's just try shitty ideas? Because the more shitty ideas we try, the more we break, the more we get rewarded. That's clearly not what you want to do. So as a background, when I'm talking about this mixed signal, basically everyone will tell you, be creative. But then if you fail, they punish you. And that's a problem because creativity means that you increase the variance, basically. There is a higher chance of success in something really interesting, but also a chance of failing.

29:29And when I'm saying that you should not punish people if they failed, I don't mean that people will come up with an idea, then go home, drink beer and do nothing. And then they'll come, oh, we failed. Right? So the three things you need to look, what happened? Why did it happen? And how can you do better in the future? If you do the why did it happen because it was a stupid idea to begin with, then you have a problem. Then you don't want to reward it. But as a behavioral economist, I run experiments very often with incentives. I look at incentives for many, many years now, and very often I predict wrongly.

30:03My intuition is wrong about what's going to happen. So I take it to the lab. I test it. I see that it didn't work. Maybe what I found is more interesting because it's counterintuitive. or maybe I just need to redesign my experiment and so on. So failing is not bad if what you did made sense to begin with. And that takes me, you asked me before what you need to do. I talked about the common sense thing that you need to do. The second thing that you really need to do with incentives is test them. So something like A-B testing that you do for the color of the background of the screen, you should also do with incentives.

30:37Try incentives, see how they work. If it works, great. If not, change them. Adjust them until they do. And don't stop over there. Once you're happy with it, use it. And then go back and test them very frequently. Because people are really creative in finding ways to gain incentives. So even something that might work now, maybe in two months, people will find a way to gain it. So make sure that you constantly test it. Because our intuition can take us only part of the way. I think many men who are married to women have a built-in common sense officer, which is ask your wife, because your wife is going to penetrate the bullshit and tell you, have you freaking lost your mind?

31:19That's your incentive idea? Do you think women have better judgment? And no, I think that spouses have better judgment than you. I would put it this way. So I think that it works both ways. That someone else, ask someone else that really cares about you, but has a different perspective. And they think that will work. Getting a different perspective of what you're saying can really work. So that's true in general. I don't think that it's a gender thing. I really think that it's about just getting somebody else's perspective that helps you. I drive a Tesla, but I can't be politically correct. Well, I'm not worried about being politically correct.

31:56I would also make the case that based on my relationships with Israelis, if your special friend who cares about you is Israeli, that's even better. Israelis really know how to cut through the bullshit. So that's actually a very important point, because when I talk about incentives and the signal that you send, it also very much depends on the culture from which you're coming. So what you said, you know, in Israel, if I give a talk, a presentation, and my friend over there will come and say, yeah, it was okay. That's equivalent to an American coming to me and telling me that was fantastic, right?

32:31So you need to adjust because the Israeli will give you a much harsher criticism. So you need to understand when you design the incentives and the signals and you think about it, you really need to understand the culture. And the culture doesn't have to be Israeli versus Japanese versus American. because even within San Diego, where I live, you can think about taxi drivers and teachers and lawyers and whatever. Each one of them will have a different culture. So the work culture, finance people might have different than healthcare providers. You need to understand the culture in which you work.

33:02You need to understand that the incentives that you design for people that might buy the Prius is very different than the incentives that will work on someone who drives a pickup truck. So they are not your target. Don't think about them because very few people will switch from a large pickup truck to driving a Prius. You have to think about people who care about the environment to begin with, and you can convince them. As you can tell, I'm very fascinated with incentives and signaling. And I majored in psychology. If I could do it all over again, I would major in behavioral economics. You and John List, such interesting stuff.

33:39Based on my affection and just how much I love this topic, I want to do a speed round. Okay, this is a Guy Kawasaki speed round. I'll try. Let's just say that I want your advice in a few different areas just as fast as you can. Just give me the gist. So I want you to tell me your idea for the best incentive programs, and I'm going to give you a list of things, all right? So number one is kids. What's the incentive program for kids? That's a very hard thing because they understand the incentives better than you. So they know how to trick you faster. So don't try it on your kids. Don't try it at home.

34:18So here's one. So think about you want to get your child to stop using diapers. You can bribe them with candies. And the nice thing about it is that they're not going to go back to diapers once they're done. If you stop giving them candies, they're not going to go back. So there are things that are really important that you want to change with kids. And it's once and done. Then you can use incentives. If your annoying teenager doesn't want to study for the SAT and you know how important the SAT is, if you can bribe them with something, that's a good idea. Don't try to bribe them to enjoy reading books.

34:53That's not going to work. Okay. How about how do you incentivize the U.S. Supreme Court? I think that the fact that they're elected for life is a problem. I think that they're so old is a problem. So you care about tech. You get people that don't know the difference between Google and Facebook over there because I'm only 55 and I'm completely out of it. I've never used TikTok. I'm not on social media. My son is like 20 times. You need younger people over there. They need to have term limits and that way they'll be better incentivized maybe to do well. I think that in general there is too much power given to old people.

35:33okay how about the incentives for ceos like per share price growth in sales what do you do for ceo so if i would be the owner of a company and i'll choose a ceo i will promise that person that their job is guaranteed for a while so i'm not looking for the quarterly earnings they i don't care about that i really want them to invest in the long run that's one of the mixed signals that you send. Every owner tells that to the CEO, but then reward them or fire them based on quarterly earnings or annual earnings. And if you need to invest in infrastructure, you need to change the network. And that's going to cost a lot of money today.

36:14So the next quarter is not going to do well. The company is not going to do well, but in the long run, that's the right thing to do. So I would just tell the CEO, look, you're on your own for maybe two years. I'm going to watch. I'm not going to talk with you about it. Maybe Silicon Valley Bank should have tried that, but I digress. How about programmers? Was it lines of code per day? What's the incentive? No, definitely not lines for codes per day, because we know that then they're going to write more lines for the same codes. My daughter is actually, that's what she does. She writes code. She's a software engineer.

36:50What kind of incentives can you give them to do? It's a good question. I Do you have a suggestion? What do they care about? Well, I mean, what's so hard about a programmer is fundamentally what you want them to do is in conflict. You want them to create a great program really fast with no bugs. And those three things, it's like you want something good, fast and cheap at a restaurant. You cannot get all three. Pick one. Right. No. So I think that the cheap part, we know that is problematic. So we know that one good programmer can be better than 100 mediocre programmers, right? So I would definitely try to find the best one and reward them for being, if I can measure what does it mean to be fast and effective, then I would clearly try to encourage them.

37:38And so here's one thing that could work, increase their salary very fast. So because many of them, you come in and it takes you time to learn the job. And then very often by the time you know the job, you leave. So just start. It's okay to start with the low salary, but then increase their salary by a lot if they're performing well very fast. So they will be more committed to stay with you. How about to use another of your theories? You say to the programmer or the team of programmers, we're going to give you each$100 ,000 bonus if you ship by this date. If you don't, you lose it. And their aversion to loss will kick in.

38:13Right. So that's one of the things, by the way, by Kahneman that we mentioned before. It talks about loss aversion. Imagine that I tell you, look, you need to write this code. If you'll do it within three months, I'll give you$100 ,000 bonus. Or you tell them you need to do exactly the same story. I'm going to give you the bonus now. But if you will not do it within three months, you'll have to pay it back. Turns out that the second one, the one in which you can lose the$100 ,000, that's much more painful. And people are more motivated when they have something like this, when they have fear of loss.

38:46because now they feel that the$100 ,000 is theirs already to lose. But in the real world, how do you take back money from a programmer or a teacher who failed? No, you don't want to put it in escrow. You want them to actually have it, but you write a contract just like any other contract. It's enforceable by the courts, if not in any way else, any other way. So you write the right contract, and you can do that. Okay. So just as hard as programmers, how do we properly incentivize teachers? So that's a very hard thing. So my life is very calm. You know, I'm a professor. I don't get much conflict in my life, apart from when I talk about incentives with educators.

39:32You go into educators and they're, you talk about incentives, they want to kill you, right? Because of, like we said, you shouldn't bribe kids. You need to get them to intrinsically be motivated in learning. Now, the question is about teachers. If you incentivize them based on some kind of standardized test, we know what's going to happen. They're just going to teach the kids how to do well on the standardized test. They're not going to teach them to be a better citizen, to enjoy learning or anything like that. But one thing that you can have, for example, is give them much more money. Just pay them a much higher salary.

40:04Because now, imagine that I can write code or I can go and teach. Some people should write code and not teach. Some people should teach and not write code, but some people can do both. And if you write code, you earn, say,$200 ,000 a year. If you teach, you get$60 ,000 or$70 ,000 a year. So you lose many people that could have been great teachers and don't do it. Just give them enough money to have the selection to begin with of people that are more motivated. And I think that would be one way to really improve the people. It's not that the teachers that we have are not good. If I can think about one very hard job that you can have is to be a teacher to these kids, to stand in front of these kids.

40:45Some of them are annoying. Their parents are very annoying usually. And you need to do this day in, day out. It's a very hard job. I think they should be compensated more.

41:00as an aside i read this and you're a college professor so literally do parents of kids call you up complaining about the difficulty of the subject and grades no no so that doesn't reach college fortunately not yet at least so i never got a call or an email from a parent And I think that's reserved for the younger kids. That's good to know. John List did this study in Chicago where they paid parents to keep their kids in school. And as I recall, he told me, all things considered, that was good. That people complained that that's buying off people and buying off education and all that. But to get them into the habit of going to school, it was worth the money.

41:48Do you agree with that? Right. So we write about it in our joint book, The Y-Axis. And the story is, imagine, he's talking about very rough neighborhoods in Chicago that the kids go out in the afternoon and they join gangs and they shoot each other. Then if you can pay them to just stay in school, even like we talked about Harvard, even if they learn nothing during this time, but they stay in school, they don't join a gang, they don't go out and get killed. That's already a huge success over there. So we're talking about a broken system, a system that has lots of problems over there. And in a perfect world, you wouldn't need to do that.

42:26But it's so far from perfect, the system that we have over there, that something like that can really be useful. You mentioned a very interesting real estate example. And my interpretation was that, let's say you're the seller. So you want to price it, as you say, just this side of crazy to signal quality. And then if you set it just this side of crazy, people are not necessarily going to try to literally lowball you for social reasons, et cetera, et cetera. So I understand all that. But where I live in Santa Cruz, at least the theory that seems to be working here is you establish the selling price as reasonable, if not low.

43:14And what that happens is it attracts multiple bids that makes the price go higher than what you really wanted. And so I'd like to hear what you say about these two contrary theories. but the bigger question for me is I understand intellectually the theory of price high signal quality put a stakeholder in the ground so nobody tries to lowball you I also understand the theory of put a low price make people go into a buying frenzy and jack up the actual price so these are two conflicting theories how do you decide that's great I think that's what makes my life interesting right? Because that's a great example where you need to test it.

43:57My guess is that in a market like we had a couple of years ago, the choice of getting in low, I know about San Diego at least, getting in low and getting 20 offers and then start playing them one against the other can work. Now the market is much colder. So it's not anymore a seller's market. And in this market, such a system won't necessarily work. So you might come up with a low number and get one person that will still try to negotiate you even lower. So you need to be careful. So I think that which system exactly works best depends on the market to a great extent. In the current market, I would say, try to be above what you think you're going to get.

44:39Make it reasonable such that the other side will not say, no, no way, I'm not going to pay that much and won't even counter. But if you get the other side to play with you, to counter your offer, then you did well. So if you are pushing it high and you don't get the offer or the counteroffer, then you push too much. I can give an example. A few years ago, I was in Singapore and I wanted to buy a lens to my camera. There is an area in town in which there are like 50 stores, all of them by Indians, that sell exactly the same cameras over there. So I went to the first place. They asked me for$700.

45:09I said$300 and leave the store. I knew that I'm not going to get it for$300. Next one, I tried$400. and they didn't want it. That way, when I said 450, he said no in the third store. And when I started walking out, he said, wait, wait, wait. So then I knew that I got to the price. So what I'm trying to say is that you need to find, and once you got into a very low offer that is still acceptable to the other side, the other side is willing to talk with you. So when you want to buy, when you want to sell, if you give a very high asking price and you get the other side to bite, then you're in good shape.

45:48If I can push back on you, I can see how that could work if you know exactly you want a Sony E-serie, E-mount, 50mm 1.8 prime, and all of those lenses are created equal. But if you go in the Grand Bazaar of Turkey and you say, okay, I want to negotiate for this carpet, all carpets are different. What do you do there? Clearly, you're in better shape if you're talking about the product, something that I can go on Amazon, check how much it costs. But even in the bazaar, you can think about different ways of doing it. So you mentioned Israel a lot. So if we'll go in the market in Israel and I'll speak with them Hebrew, I will get a different price than if you'll go and try to speak English with them, right?

46:32So even there, there are different things that you can do that are like that. So if you have a house in a condo that there are three other apartments for sale, you can't really play this trick. If you have some kind of unique house that Zillow cannot give you good comps to it, then you're in a different situation. So if you sell a commodity, then negotiation and using the price, high price as a tactic, won't work as well as if you're selling, like you said, a rug that only you have. For the record, whether I'm in Turkey or Israel, I would never go shopping without a native with me. Just hold a sign on your head that says, I am a goy.

47:17Yes. I would just like you to summarize, you know, sort of best practices about incentives. And then give us a pitch for your book. I think people should read your book. It's so interesting, and I think it'll open your eyes and help you avoid many mistakes. So give us your best shot and maybe a good example, the Yuri Hall of Fame. This company really understands incentives. This is what you should do, and this is why you should buy my book. So a few rules to remember. First of all, your incentive sent a message. Don't forget it. And use common sense. Don't just, oh, you know, all incentives are created equal.

47:56So very often I go to companies and they tell you, oh, we tried incentives and they didn't work. And my analogy to this is that you went to a bad Japanese restaurant and your conclusion was that Japanese food is bad. No, you were unlucky. You went to a bad Japanese restaurant. You should try something else. Same is true about incentive. You tried, but you didn't think about it enough or you were unlucky. But that incentives work. There is no question about that. But they don't always work the way we think about. So think about it a lot. And then after you come to something that you think tells the story that you want to tell, test it.

48:32Right? Like the A-B testing that we discussed, just go out and test it, see that it works. If it doesn't, improve it until you get to something that actually works. Why should you buy the book? Because the book will tell you how to do it in a smart and funny way, such that you will understand how incentives work. So incentive work, there is no question about it. How they work, that's very complicated and somewhat simple if you understand what you're doing. So understand the signal that you're sending. Don't send mixed signals and just think about it. And who's in the URI Hall of Fame as, yeah, these guys really know what they're doing?

49:07So one example is the airlines that came with a frequent flyer. So I fly a lot and someone else is paying for the flight ticket. I'm lucky in that respect that people invite me. And imagine that I fly to San Francisco to visit you. There'll be two flights, one southwest that will cost$300 or one by United that depart 10 minutes later and cost$500. I'm a United guy. I have frequent flyer miles. I will buy the United one and you will pay the$500. I don't care how much it costs. So the airlines were smart enough to create incentives to individuals that they don't pay the cost, but they get the benefit of it.

49:48And I think that was a very smart program to create. So if I'm the company bringing you in, I should tell you that I'm going to pay you X and that includes your travel. And then you have to make the decision whether the difference is enough so that you fly Southwest versus United, even though you're not getting United miles? There is no question that that's the right way of reimbursing people. So we are, you know, I work in a public university and we fly in people and we give them all these restrictions. And then we need 70 people to sign off on this. It costs much more to reimburse you. Instead of this, there are some places that you just tell them, OK, or you come to San Francisco for two days, we'll give you whatever X amount of dollars.

50:32You can buy the ticket, the hotel, do whatever you want. take Uber or you can walk to the airport. We don't care. Then I'll be much more careful with the way I spend the money. That's, by the way, with my flights, you know, that's peanuts. The healthcare really suffers from this. When I first went to a dentist in the US, it was in Chicago, the clinic had an amazing view to the lake. It was in a good place. The offices, the clinic was amazing. And I've never seen it in Israel. In Israel, you know, dentists are in the basement. And the reason is that in the US, the insurance is paying for it. You can charge and so they have to compete.

51:09They don't compete on being cheaper. So I don't care how much it costs, because it's going to cost me the same whether I go to this beautiful clinic or to a less good clinic. And I want to live in a world in which I can decide, okay, I can afford myself to go to the nice clinic and pay three times more. Or I can go to a basement and get just as good care, but it's going to be a bit less pleasant the view is going to be less pleasant right okay thank you very much and um i hope your stomach feels better and as the japanese say zygazunt zygazunt yeah very japanese yiddish word thank you very much it was very interesting it's very different than other podcasts that i've been on so it was much more fun why what what happens in the other podcasts They ask more structured questions, right?

51:59So they read the passage and they ask me to describe it or something like that. And this one was more exciting. So thank you. You made my day, Yuri. I hope this conversation with Yuri has provided you with valuable insights into the world of incentives and behavioral economics. Yuri's new book, Mixed Signals, How Incentives Really Work, is a must read for anyone interested in understanding how incentives shape our behavior and decision-making process. Again, my thanks to Katie Milkman for the directive to have Yuri on this show. And my thanks to the Remarkable People team. That would be Peg Fitzpatrick, Jeff C., Luis Magana, Shannon Hernandez, Alexis Nishimura, and the drop-in queen of Madison Nismar, who just moved into a new place.

52:54I'm Guy Kawasaki. This is Remarkable People. Until next time, aloha. Listeners of the Remarkable People podcast will learn from some of the most successful people in the world. They provide practical tips and inspiring stories that will help you be more remarkable. If you live in the U.S. or Canada, text 831-609-0628 to get notified of each new episode. This is Remarkable People.