Hidden Brain XX
[0] This is Hidden Brain.
[1] I'm Shankar Vedantam.
[2] A few years after the communist revolution brought Mao Zedong to power, China's leader ordered his people to exterminate four pests.
[3] Rats, flies, mosquitoes, and sparrows.
[4] The birds were eating China's grain.
[5] Chairman Mao had recently collectivized agriculture, and he wanted to kill off the birds consuming his country's food.
[6] During the Great Sparrow campaign, as the effort was called, millions of birds were killed.
[7] People shot sparrows or banged on pots and pans until the birds fell to the ground in exhaustion.
[8] But the eradication of the sparrows produced an unintended consequence.
[9] Insect population soared because one of their natural predators had been eliminated.
[10] Locusts swept across China's fields, decimating crops, and contributing to a great famine.
[11] Official figures put the death toll at 15 million people.
[12] Unofficial numbers were two to five times higher.
[13] While the scale of the disaster was unimaginable, China is hardly alone when it comes to wrong -headed policies.
[14] From nations and governments to companies and families, humans regularly fail to foresee how initiatives can backfire.
[15] This week, we continue our series success 2 .0 with a look at one of the principal sources of failure in our lives.
[16] Unintended consequences.
[17] How to expect the unexpected when we plan for the future.
[18] This week on Hidden Brain.
[19] Much of life is about getting others to behave in the way we want.
[20] We come up with carrots and sticks to persuade children to do their homework, prompt partners to pick up their socks, and motivate co -workers to do their best.
[21] But often these inducements are ineffective.
[22] Sometimes they even backfire.
[23] Uri Gnizi is an economist at the University of California, San Diego.
[24] He studies how we craft incentives and smart ways to do it better.
[25] Uri Gnizzi, welcome to Hidden Brain.
[26] Thank you.
[27] Great to be here.
[28] Uri, when you were a child growing up in Tel Aviv, in Israel, the heroes of Israeli society were the fighter pilots who served in Israel's Air Force.
[29] When you were 11 or 12 years old, you heard a story about these fighter pilots that made quite an impression on you.
[30] Tell me that story.
[31] So in Israel, being a fighter pilot was the highest social status you can have.
[32] And everyone wanted to be one, including me. So think about top gun on steroids.
[33] And my hobby was to read all these World War II books and other books about fighter pilots going out and shooting down enemy airplanes and becoming heroes.
[34] And so I remember this story that I read about the Six -Day War.
[35] It was a Mirage pilot that told the story.
[36] He was there with his three bodies.
[37] The four planes were looking for Egyptian migs over the Suez Canal.
[38] And suddenly he saw two of them in the horizon and he turned immediately over there.
[39] His friends didn't see where he went.
[40] And they asked him and he said that he just is chasing the migs to the west, when he actually chased them to the east.
[41] And basically what he did was trying to get these migs by himself, to shoot them down by himself.
[42] Now, if he worried about the country and safety and the chance of actually doing this, he would have called his buddies to join him because that's much better.
[43] But no, he really wanted to shoot them down by himself.
[44] So he lied to them, and after that he was very proud of it.
[45] So that shows that the incentives are really working in this case in, it might be a funny way, but I remember, like you said, when I was 11 or 12, thinking about, okay, so that's the real incentive.
[46] That's what motivates him.
[47] That's what gets this guy to go and risk his life, because he really wanted the social status that came up with shooting down the mix.
[48] So it's interesting, you know, the very high status awarded to pilots, you could say that that's a good thing for a country to be doing, because of course, by recognizing, you know, soldiers, and pilots, the country basically incentivizes people to want to join the armed forces and to protect the country.
[49] But what you're showing here is that that incentive now can actually change people's behavior in ways that actually don't serve the best interests of the country.
[50] Right.
[51] You get the best people to actually want to go out and shoot enemy planes risk in their life.
[52] But in some cases, it can really make you do stupid things, because now your motivation is not completely aligned with the organization.
[53] The organization, in this case, the state of Israel, wanted the MiGs to be shot down and you to be safe.
[54] So sometime later, when you were a teenager, you and your classmates knocked on doors each year to raise money for charity.
[55] And one year, your teachers came up with an incentive to get kids to do a better job.
[56] Tell us how the charity program worked, what kind of activities you did each year with your friends, and what the incentive was that teachers came up with that one year.
[57] So when I grew up in Israel, we used to go out a few times a year to collect money for charities like the Cancer Association or kids in danger.
[58] And together with another 15, 16 year old kid, you went around knocked on doors and tried to raise money.
[59] Now, it was really, you didn't have to persuade the donors because they knew exactly what it is.
[60] And the more doors you knocked on, the more success you had.
[61] And I remember as a kid that I was very happy when I got to the age when I could participate in this.
[62] And my friend and I went out and tried to knock on as many doors.
[63] We got receipts for 500 shekels, which was about $170, $180 at the time.
[64] But then there were also some kids who didn't do anything.
[65] They just didn't go.
[66] And the teacher came up with the brilliant incentive scheme.
[67] She said, if you raise at least 100 shekels, then you won't have to do homework over the weekend.
[68] And it worked.
[69] So it was a brilliant incentive in the sense that now everyone went out and collected at least 100 shekels.
[70] But actually, many of us, including me, stopped now at 100 shekels.
[71] And as a result, what happened is more kids actually went out, but the overall collection was much smaller than before.
[72] So in some ways, the incentive was trying to change the behavior of the weakest performers.
[73] And in fact, it successfully did so, but at the cost of undermining the performance of the strongest performers.
[74] Exactly, and the overall performance as well.
[75] One last example I want to touch on.
[76] Years later, after you became a parent yourself, you once took your son to Disney World and you spotted a sign setting out the ticket prices for admission to the park.
[77] What did the sign say, Uri?
[78] So the sign said, if you're under three -year -old, it's free.
[79] Over three -year -old, it's $117.
[80] Now, my son just turned three a couple of months earlier, and I loved that period because he started communicating.
[81] You can talk with them.
[82] It's a lot of fun.
[83] But they also start lying when they start talking.
[84] That's also very, very human, right?
[85] And, of course, as parents, we told him that he shouldn't lie.
[86] Only bad kids lie.
[87] So we got to the counter, and then I saw that if the kid is under three, it's free.
[88] And the cashier asked me, you know, how old is your kid?
[89] I looked at him, and I said, well, he's almost three.
[90] Now, it wasn't a lie completely because it was almost three, but from the wrong direction.
[91] But, you know, I lied and got to pay, whatever.
[92] And then we started walking, and about half an hour later, my son pulled my shirt and said, Daddy, Daddy, you told me that only bad people lie, and now you just lied.
[93] And, of course, I turned pale, and my reaction was something of this sort of do as I say and not as I do.
[94] A few minutes later, we got to another attraction, and this one was, my son, Ron, really wanted to do it, but it turns out that you needed to be at least four in order to be able to go on it.
[95] And Ron, without hesitation, said, yeah, I'm four.
[96] And, you know, turns out that what he actually learned was not to do as I say, but to do as I do.
[97] So you saved $117 on Ron's ticket, but the price.
[98] was that you sent your three -year -old kid the message that lying is okay.
[99] Exactly.
[100] And I guess the question to ask then is, was that the right price to pay for the message that you were sending your kid?
[101] I think that that's exactly the question.
[102] So when you do something like that, you need to think, is it worth $517 in order to send this message?
[103] In some cases, you might say yes.
[104] But in retrospect, no, it wasn't.
[105] I wouldn't have done it again.
[106] Classical economics presents us with a very simple understanding of incentives.
[107] Incentives can get people to do things they might not otherwise do.
[108] Once an incentive is introduced, people's motivation will increase as you increase the size of the reward.
[109] But the way incentives play out in real life is considerably more complicated.
[110] When we come back, we peel back the layers of the incentives onion to understand why it's so hard to get other people to do what we want.
[111] You're listening to Hidden Brain.
[112] I'm Shankar Vedantam.
[113] This is Hidden Brain.
[114] I'm Shankar Vedantam.
[115] Uri Gnizzi works at the University of California, San Diego.
[116] When he was being trained as an economist, he was taught that incentives were simple and straightforward.
[117] Pay someone to do something, they'll do it.
[118] The more you pay them, the harder they'll work.
[119] But over many years of experience and research, Uri has come to see that reality is a lot more complicated.
[120] Uri, I'm wondering if we could start with the basics.
[121] What is an incentive and what kind of signal is it designed to send?
[122] So incentive is something that will motivate you to do an activity or produce something or whatever that you wouldn't do without it.
[123] And one of the problems with setting incentives is that many people don't understand that these incentives are not just giving you some kind of material payoff.
[124] They also send you a signal about what's important for me. And this signal is something that can really impact behavior.
[125] So poorly crafted incentives can backfire in that they can actually discourage people from doing things you want them to do.
[126] But it's sometimes hard to tell when an incentive is poorly crafted.
[127] For example, we all know that it's nice to say thank you when someone does something nice for us.
[128] But the social scientist George Newman and Jeremy Shen once looked at the effect of thank you gifts on charitable contributions.
[129] Tell me what they found, Uri.
[130] So they found that if you give people a gift after a thank you gift, say a coffee mug, it can actually backfire and get people to donate less.
[131] And I think that the reason is that it basically changes the meaning of what you're doing, turning the situation into some kind of exchange.
[132] Now I'm not doing it because I'm a nice guy.
[133] I'm doing it because I want this bag or pen or whatever the gift is.
[134] And then I say, well, I'm not going to spend $100 on this.
[135] pen or this bag, it's not worth it.
[136] It completely changes the way I perceive the interaction.
[137] And because of that, I want to do the activity even less than before.
[138] So sending unclear or contradictory signals can change the way people think about a particular situation and in some ways it can lead them to engage in behavior.
[139] That is the opposite of what you're trying to encourage.
[140] There's a somewhat related story involving the Welsh government.
[141] Officials wanted to crack down on parents who pull kids out of school for family vacations.
[142] Presumably the policymakers wanted to keep kids in class so they could learn more, right?
[143] Right.
[144] So anyone who had kids and wanted to go on vacation knows that that's the worst time to go.
[145] So when you have breaks from school, that's where everything is more crowded, more expensive, less pleasant.
[146] And some parents decided, you know what, we'll take our kids off school, say a week before or week after the spring break.
[147] That way we'll save money and we'll have a nice vacation because it's going to be less crowded.
[148] The government didn't want to do it and it introduced a small fine.
[149] I believe it was 60 pounds if you did that.
[150] Now the situation that the parents faced was very different.
[151] Before that it was look, the teachers are not going to like it.
[152] The principal is not going to like it.
[153] Good parents won't do that.
[154] Now the story is, oh, if we'll do that, it's 60 pounds.
[155] You put a price on this activity, and now instead of being immoral, now it's like, well, for 60 pounds, I can save much more money by going a week earlier.
[156] And it turns out that many parents decided to do it.
[157] And some travel agencies even offered to pay this fine for people.
[158] Oh, my gosh.
[159] Because it was so much better to do it this way, right?
[160] Everyone is happier, apart from the schools that were less happy.
[161] One of the things that's tricky about incentives is that in any situation, there is often more than one motivator at play.
[162] So an incentive can amplify one source of motivation, but sometimes at the expense of something else, and that something else might actually be more important.
[163] Uri once saw this play out in British professional soccer, where a team was awarding bonuses to individual players for every goal they scored.
[164] As a soccer player, there are different ways to motivate you, to incentivize you.
[165] One of them is just pay your salary, another one pay you for victory.
[166] Another one is per scoring a goal or for assist if you send the ball to someone else.
[167] So there was a situation in which if you scored the goal, you got 75 ,000 pounds.
[168] If you sent an assist that ended up in a goal, you got 20 ,000 pounds.
[169] Now, imagine that you're facing two options.
[170] One option is to try and run yourself and score yourself, say that you have 50 % chance of succeeding.
[171] Then you have 50 % chance of getting 75 ,000.
[172] pounds.
[173] Or you can send it to your friend and say that your friend has 75 % chance of scoring it.
[174] So now you'll get lower than if you try yourself.
[175] So in other words, passing the ball might end up getting you less, will get you less money than trying to score yourself, even though the odds of you're scoring are actually lower than if you pass the ball.
[176] Exactly, exactly.
[177] So now the team clearly cares about scoring the goal.
[178] They don't care who is going to do it.
[179] They want to win.
[180] That's what the coach and management try to incentivize.
[181] And I'm sure that the coach told everyone, all the players, look, we are a team, we want you to be a team player and blah, blah, all the good stuff.
[182] But the signal that the incentive sent was very different than the signal that the coach told them before.
[183] Now, it's a complicated situation because in some cases, if you want to attract the best players in the world, if you want to attract the messy of the world, you really need to pay them a lot per success, right?
[184] But at least don't send a message telling them that you care about team game and incentivize individual success.
[185] That's not going to work.
[186] This is a little bit like the story of the Israeli pilot, right, which is the Israeli pilot cared more about his own glory than about what was actually best for the country.
[187] Exactly.
[188] And still, I'm not saying that that's necessarily a bad incentive because these incentives really motivate the fighter pilot and the soccer player to do their best.
[189] But in some cases, it could create a problem.
[190] So it's also the case that in group settings, different people often have competing incentives.
[191] Uri, I want to play you a clip from the TV show Friends.
[192] This scene is unfolding at a restaurant.
[193] And for the gentlemen.
[194] Yeah, I'll have the Thai chicken pizza.
[195] Miss?
[196] Okay.
[197] I will have the side salad.
[198] And for you?
[199] Who?
[200] I'm going to have a cup of the cucumber soup.
[201] I will have the Cajun catfish.
[202] Everyone enjoys their meal.
[203] Then the bill arrives.
[204] Okay, everyone owes $28 bucks.
[205] No, no way.
[206] Sorry, not going to happen.
[207] So, Uri, you've talked about a problem known as the unscrupulous diners dilemma.
[208] What is this dilemma?
[209] So I love this clip from France.
[210] Basically, what happens is that if you know that you're going to split the bill, then what the economist called, your marginal cost goes down.
[211] So imagine that there are five people in the restaurant.
[212] And we're going to split the bill.
[213] So if I spend another $10 on my main dish, I will actually pay only $2 out of that because the other people are going to pay the rest.
[214] So then my calculation becomes very different.
[215] If before that I would say, well, the lobster costs $50.
[216] I don't want to spend that much.
[217] Now I say, well, the lobster will actually cost me only $10.
[218] I should go for it, right?
[219] That's my chance.
[220] And that's indeed what we found in our experiment.
[221] We brought students to actual restaurant.
[222] And we saw what happened when we told them that we're going to split the bill versus when we told them that each one will pay for themselves.
[223] So they ordered more.
[224] They ordered the drinks suddenly.
[225] They were real festive.
[226] And by the way, it was true for men and women.
[227] The entire bill was so much higher when they split the bill that it was funny to see.
[228] In many cases, restaurants will say if you're a group bigger than X, we want you to actually split the bill, and that could be one of the motivation.
[229] They know that people order more when they split the bill.
[230] Now, we've looked at how incentives are designed to send signals, but of course the signals we're sending are not being heard by computers or robots.
[231] They're being heard by human beings, and human beings don't just respond to the signal.
[232] They ask, what is this signal really about?
[233] What is it telling me?
[234] During the COVID pandemic, Uri, some states like Ohio decided to encourage their citizens to get vaccinated using a rather unusual incentive.
[235] Tell me what they did.
[236] So they offered a lottery that will give you, say, a million dollar.
[237] Some of them were even higher than that.
[238] So they're going to go to all the people that got vaccinated this week, and one of them is going to get a million dollar in a lottery, which is great.
[239] you know, it's nice, what can go wrong?
[240] So here's a thought experiment.
[241] Imagine that someone invites you to join a test for a new drug.
[242] They tell you that's, you know, a headache, a pill that you can take, and your headache will go away.
[243] And there are no side effects that we know, and we'll pay you $50 for participating.
[244] That's basically what the FDA does in many cases, and I think that it makes sense to do.
[245] And I would be willing to take part in such a study.
[246] Now imagine that it's the same, but they tell you, will pay you $50 ,000 for participating.
[247] Now I'll be worried, right?
[248] If there are no side effects, why are they offering me so much money to take this one pill?
[249] And I think that this signal that if you're paying me a million dollars to get vaccinated, it must be really bad for me to get vaccinated, right?
[250] Because otherwise, why do you have to pay me to do it?
[251] Yeah.
[252] Remember that we talked about completing the story.
[253] This completes the story in a different way than you intended.
[254] You wanted to tell him, look, I care about you so much, that I'm willing to pay you because it's so important and the bad things that can happen are so bad that I want to encourage you.
[255] That was your story, but the story that the people may hear is that this is actually bad for you.
[256] It's so bad that we are willing to pay you a million dollars to do it.
[257] Another dimension of incentives that makes them tricky has to do with the fact that the people we're trying to influence perceive not just a carrot but also a stick.
[258] So when people don't achieve an objective, they start to ask, What's going to happen if I fail?
[259] Uri, tell me what happened at Wells Fargo Bank in the early 2000s when they introduced an incentive to increase performance.
[260] So the CEO had a great idea.
[261] I want every customer that we have to have at least eight products from the bank.
[262] So if the customer has, say, a checking account, I wanted to also have a credit card and saving account and whatever.
[263] And they gave very strong incentives for the people working in the bank to actually do that.
[264] And actually, if you did not meet the goals that they gave, it's not only that you didn't get the incentives, you were also very likely to be fired because your colleagues were able to do it.
[265] Now, it turns out that it's very hard to do it, and the way the colleagues were able to do it is by making up accounts.
[266] So, you know, think about the image of a worker coming to the bank in the morning, drinks his or her coffee, sit down at the computer, and start making up fake accounts.
[267] According to authorities, inside Wells Fargo, employees called it gaming, forging customer signatures, creating pins to activate unauthorized debit cards, and moving money to unauthorized accounts.
[268] I had 15 accounts at once.
[269] It was just very frustrating.
[270] And these are accounts I never opened.
[271] And it wasn't one or two people.
[272] We're talking about over 3 million accounts, and they fired at the end over 5 ,000 people for doing this.
[273] So basically, the CEO wanted to have more.
[274] products per account, got exactly what he wanted.
[275] The stock of Wells Fargo went up.
[276] Everything was fine, apart from the fact that it was all a balloon that was clearly going to explode at one point because it was all fake.
[277] I'm wondering if part of the problem here, Uri, is that the leaders were trying to incentivize an explicit behavior, which is to aggressively sell the company's products, but they ignored how this incentive might affect other behaviors, including the implicit norm, to be honest.
[278] They probably weren't thinking at the point they set up the incentive program.
[279] Well, what happens when people don't meet this target?
[280] What are they going to do?
[281] So giving people incentives to create X -ray account is not bad in itself.
[282] It will work.
[283] People will try harder.
[284] But at the same time, they have to put some safety measures that will make sure that people don't lie.
[285] So when you design incentives like this, think about how can people try and game the incentive, because that's what we do.
[286] We game the incentives all the time.
[287] We are very good at it.
[288] We are very creative at it.
[289] So, for example, if there was enough auditing done, and once you catch someone making up an account, you immediately fired that person, that would have worked.
[290] But instead, what happened is that even whistleblower that came to management said, this is happening, they were actually fired.
[291] So the incentive in itself wasn't bad.
[292] It just lacked the safety measure that was needed.
[293] You cite a revealing incident from colonial era Vietnam.
[294] French authorities were confronting a problem and they came up with a solution.
[295] It involved incentivizing ordinary people to go out and kill rats.
[296] But that incentive ended up producing a problem of its own.
[297] Tell me what happened, Uri.
[298] So the French wanted to introduce toilets to Hanoi, which is very French of them to do.
[299] And when the toilets were introduced, you also need sewage.
[300] And with sewage come the rats.
[301] So there was a big problem with rats.
[302] And the authorities were very smart.
[303] He said, let's pay, say, 10 cents per rat tail that you bring.
[304] So there was a poor guy in the city hall that was counting rats' tails.
[305] And for each one of them, he gave 10 cents.
[306] That sounds, again, like a very good incentive.
[307] What can go wrong?
[308] That's, you give the people, you let the people go after the rats.
[309] Everything's going to be well.
[310] Well, it turns out, for example, that people just cut the tail of the rat and let the rat go free because the rat can have more babies that way.
[311] Some of them started their rat farms.
[312] So they started to raise their own rats in order to harvest their tails.
[313] Some of them brought rats from neighboring cities in order to cut their tails.
[314] So again, people are really good at gaming the incentives.
[315] And you need to find a way and incentives that actually is robust to such manipulation.
[316] You know, it's interesting because I think this draws attention to the fact that when you're leading a company or a city or a country, you know, it's very easy to focus on something that is not going well and come up with a simple plan to fix that thing.
[317] But the problematic thing usually doesn't exist in a vacuum.
[318] It's connected to other things or it can influence things one layer beneath the surface.
[319] And in some ways, I think what all these examples are pointing to is the complexity of human behavior and perhaps how we need to exercise a certain amount of.
[320] out of humility when it comes to trying to engineer that behavior.
[321] So when I work with companies, very often I hear some version of, we tried incentives and they didn't work.
[322] And when I hear this, I think about, you know, someone went to a bad Japanese restaurant and his or her conclusion is Japanese food is bad.
[323] No, that's not the right conclusion.
[324] You went to a bad restaurant, and the same is true about incentives.
[325] It's not that incentives don't work.
[326] It's just that you didn't understand the complexity of incentives.
[327] incentives are powerful tools.
[328] They can make people act in ways they would not have left to their own devices.
[329] But misunderstood or misapplied, incentives can backfire and boomerang.
[330] When we come back, how to craft incentives that get yourself and others moving in the direction of success.
[331] You're listening to Hidden Brain.
[332] I'm Shankar Vedantam.
[333] This is Hidden Brain.
[334] I'm Shankar Vedantam.
[335] Economist Uri Gnizzi is the author of Mixed Signals, How Incentives Really Were.
[336] work.
[337] He studies the science of how we use carrots and sticks to get others to behave the way we want.
[338] Uri, when someone crafts an incentive, the people being incentivized use the signal of the incentive to draw conclusions about the sender, so poorly designed or poorly presented incentives can backfire by making the creator of the incentive look callous or greedy.
[339] You recently observed this in an initiative launched by a movie theater chain.
[340] Can you tell me what happen?
[341] So recently AMC decided to have different prices for different seats.
[342] They understood that people don't like to sit up front or on the sides.
[343] And their solution was, let's raise the price for the middle seats, for the premium seat.
[344] And that's what airlines are doing.
[345] That's what hotels are doing.
[346] But the way they did it really created a pushback.
[347] Many people were upset.
[348] You know, the prices are already so high.
[349] And now you're charging us more for the premium.
[350] seats, that's really a nasty behavior.
[351] That's the story that people heard.
[352] But instead of that, AMC could have said, look, we're sorry, prices are so high, the inflation, whatever, we need to raise our prices, we're sorry, but here's the good news.
[353] We're going to give you discount if you're willing to sit in the first row or on the sides.
[354] So exactly the same incentives, but frame them, tell the story, fill up the story in a way that instead of I'm greedy and want to take even more money from you is the story that you're telling is, look, we're trying to help you and we're offering this discount on less desirable seats.
[355] So in other words, let's say, for example, the ticket prices were $10 and you're charging $15 for the premium seats.
[356] What you're saying is, you know, raise the ticket prices to $15 for everyone and then basically make the less attractive seats discounted to $10.
[357] It comes to the same thing, but the story becomes completely different.
[358] Exactly.
[359] Instead of, I'm taking advantage of you whenever I can, I'm trying to help you whenever I can.
[360] I had to raise prices because everything costs more.
[361] We have to pay more for our workforce.
[362] We have to pay more for rent.
[363] So we had to do this, but we really care about the people that have hard time paying, and we really want them to actually keep coming.
[364] So this points, of course, to the importance of paying attention, not just to the incentive, but the story that the incentive is communicating.
[365] Some companies have, found a way to use incentives to tell a consistent story about their values.
[366] And you say that one such company is the sandwich chain, Predamagee.
[367] How did they design incentives, Zuri?
[368] So they wanted to give some kind of bonus like employee of the week or something like that and give you a bonus.
[369] Now, in every case that you're doing something like this, you're actually doing something else as well.
[370] It's not just the people that you invite to the party, it's also the people that you don't invite that you should care about.
[371] So the people that in this case, did not get the bonus.
[372] In this case, the bonus was $50.
[373] And the way they thought to reduce the problem was that if you were the employee of the week or of the month, you got the $50, but you were asked to spend it on someone else, on one of your friends.
[374] So the incentive on the surface is designed to reward people for reaching a milestone, but really the way the company is using it, the incentive is designed to enhance team building.
[375] Yes.
[376] if it's a subjective measure, for example, how well you did on something that we can't measure in dollars, then I'm going to be upset if you'll get the reward because I'm clearly better than you.
[377] This way, it reduced by a lot the tension between the employees about who's going to win this.
[378] So winning it was, you still got you the status and it was nice, you felt good about yourself.
[379] But the money you gave to other people.
[380] So other people, the workers became more friendly with each other and there was less tension over there.
[381] One problem with incentives is that people will change their behavior for as long as they are being incentivized to do so, but will then revert to their previous behavior once the incentive is withdrawn.
[382] Your own research, however, has shown that incentives can create long -lasting change if they are used in a particular way.
[383] Tell us about your study about building exercise habits, Uri.
[384] So creating habits is maybe the holy grail of incentives.
[385] How can we change your habits?
[386] So can we get you to exercise more, stop smoking or diet better, whatever is the activity that you want to do?
[387] It turns out to be very difficult because it's an ongoing behavior which is very hard to change.
[388] And many of us, those of us that struggle with this, including me, remember periods of their life in which they exercise.
[389] So there were periods in my life in which I exercised I was very happy about is going to the gym or surfing or biking, whatever it is.
[390] And then I have periods in my life in which I can't get myself off the couch.
[391] The only physical activity is watching Netflix.
[392] What's the difference between these two?
[393] And what we thought about, and there is some literature suggesting this, that once you get into the activity, so say for a month, after that it's much easier.
[394] So the first time you go to the gym, it smells bad, you don't know where to park, you don't know what to do, it's kind of embarrassing.
[395] After a month, suddenly you don't bother the smell that much.
[396] And you know exactly what you're going to do.
[397] And say to put it in your schedule that Monday and Thursday at night.
[398] 9 a .m. I'm going to do yoga classes, whatever it is that you're doing, suddenly it becomes much easier for you.
[399] And the idea was that if you can build this, if you can invest in this month, even if you're going to suffer and you think that the suffering is bigger than the benefit, just invest in this month.
[400] And then after that, it might become really easier.
[401] So we did it with incentives.
[402] It was Gary Charnes and I. So we pay students $100 to go to the gym eight times for a month.
[403] And then after that, we stopped paying them.
[404] And we saw that the people that got paid to go to the gym were much more likely to go to the gym than the control group.
[405] So we did create some kind of habit.
[406] Turns out that this habit dies quite fast.
[407] So, you know, winter break, they come back, the habit is dead.
[408] So how to maintain this habit is still a question.
[409] But it is interesting that if I can get you over this hurdle of actually going to the gym for a few times, it's going to be easier.
[410] Now, you can also think about it for yourself.
[411] You can incentivize yourself.
[412] The first time you'll go, like we said, you're going to suffer, but you should commit.
[413] I'm going to force myself to go for a month, and only then I'm going to evaluate whether I want to continue or not.
[414] And you might be surprised by the outcome.
[415] A related idea is that incentives can also be used to help people get past what you call switching costs.
[416] What are these, Yuri?
[417] So switching costs is something that we do all the time.
[418] So we have this, imagine going to the supermarket.
[419] You know which kind of shampoo you're going to buy, or toilet paper.
[420] Maybe at the beginning you searched a bit, but then you chose one, brand of toilet papers, and that's what you're taking.
[421] Switching to another one will require standing over there, thinking about it, comparing prices, is it soft enough or not?
[422] You're not going to do it.
[423] You're going to stay with the toilet paper that you know.
[424] Now, imagine that you go, you reach to your regular toilet paper, and you see that the competitor is as a discount, you know, half price, say.
[425] Now you're saying, wow, that's a big discount.
[426] Let me try the other one.
[427] I'll try it.
[428] You'll take it, try it, and if you like it, you might switch to it, and, you might switch to go back, when you go back, you'll actually search for it because you like it, even if there won't be a discount on it.
[429] So there is a switching cost, which is, it could be a mental switching cost like that, that you just don't want to think about it.
[430] There could be some other switching costs.
[431] Imagine now smoking, for example.
[432] Imagine that I can actually pay you to stop smoking for a month.
[433] Maybe after a month, it's going to be less attractive for you to go back.
[434] And that's the idea, right?
[435] So switching from one activity to another is costly if I'll give you at this count.
[436] If I'll give you incentives, I might get you to change your behavior in a way that you wouldn't do after that.
[437] And if I'm lucky, you'll stick with the new behavior later on.
[438] In many parts of our lives, we don't see the consequences of our actions until much later.
[439] So if you start smoking, for example, it can take years or even decades before you pay the price in terms of health effects.
[440] The same thing goes for schooling, but in the opposite direction.
[441] It often takes years before you can see the benefits of studying hard and doing well at school.
[442] You and others have done a lot of thinking about this problem and how incentives might be used to try and improve the performance of students in school.
[443] Tell me about this work, Uri.
[444] So as a university professor, my life is quite boring.
[445] I don't get that many people mad at me unless I go and talk about incentives with educators.
[446] That's, you know, they want to kill me. They want to chalk me very fast.
[447] And I think that they are right in many cases because you want the students to have intrinsic motivation to be successful.
[448] You don't want to pay the money in order to read books.
[449] And we know that I can pay you money to read a book.
[450] I can't pay you to start enjoying reading the books.
[451] So it's much harder.
[452] And they sympathize with the teachers and the educator that say this.
[453] However, there are things that you can actually use in education that will be useful.
[454] So imagine my example is I, you know, I write.
[455] a lot for my living, and yet I write, you know, one finger at the time.
[456] I don't know how to type.
[457] If someone would have paid me enough money or convinced me to invest a week in learning touch -typing, that would be amazing for me, right?
[458] So once you learn a skill, you're not going to go back.
[459] And in education, you can think about some landmarks like this that are important.
[460] Imagine your annoying teenager and the SAT.
[461] You understand how important it is.
[462] You really want your kid to be successful, but they have other things in mind.
[463] If you can find incentives to make sure that they actually prepare well for the SAT, that could still work.
[464] The bigger problem in which I agree with the teachers and the educator is that you cannot use it in order to get kids, students, to enjoy what they're doing.
[465] That we don't know how to do, and that's much harder.
[466] At the same time, you know, it seems like I think the picture you're painting here is that incentives are perhaps potentially useful in sort of these narrow circumstances.
[467] where, you know, you, the student basically is unable to really comprehend how, you know, applying himself or herself in school can pay off 20 or 30 years down the road, but a carefully targeted incentive can prompt the student to do better on a standardized test, for example, you know, three weeks print, for example, and that might in fact have long -term consequences that are beneficial.
[468] That's true, but I wouldn't be that pessimistic about the first part, because if you use the right incentives, you can make you actually studying more.
[469] enjoyable.
[470] We have a recent paper in which we looked at the effect of physical activity on success, academic success.
[471] So imagine that all you care about is the math success, success in math for the kids.
[472] And currently they get 10 hours of math and two hours of physical education every week.
[473] Now you have to cut two hours.
[474] Which one should you cut?
[475] Should you cut down the two PE classes like most schools do or maybe two hours of the math?
[476] Even if all you care about is math, performance.
[477] It could be, and that's what we find, we find that students that are engaged in physical activity are actually doing much better academically.
[478] So it might be better to take the two hours for math, let them run around and go back.
[479] And this could be used also as incentives, right?
[480] So if I tell you, look, finish up the work and you can go out and play some basketball, that could be used as incentive.
[481] So the kids will enjoy coming to school more and everything around it will become nicer.
[482] So I wouldn't be as pessimistic as we can't do anything, but we need to be very careful.
[483] Uri, when it comes to designing incentives, it's really important to pay attention to the size of the incentives that you're designing.
[484] You've run studies looking at the risks of running incentives that are too small.
[485] Tell me about that work.
[486] So we had the same setup with kids collecting money in these special donation days.
[487] And what we wanted to do is compare kids that are just motivated by that's the right thing to do.
[488] you should go out and get it, with kids that get a small kickback.
[489] So we offered them 1 % of what they collected.
[490] So if you collected $200, you got $2 as a reward, which is very little.
[491] And what we found was that when we offered this reward, these incentives, kids were much more likely to stay home and not go out to collect money because now they said, well, it's only $2, it's not worth it, right?
[492] I'm not going to put all this effort.
[493] So Aldo Rastikini and I ran an experiment like that.
[494] And the title, I think, is really revealing.
[495] We called it pay enough or don't pay at all.
[496] So the point is that if you would have offered these kids $100, they would have collected all the money they can.
[497] But when you effort too little, you change the story.
[498] Before that the story was, I'm doing it because that's the right thing to do.
[499] Now, when you offer incentives, I'm doing it because I'm getting paid.
[500] Well, if I'm getting paid, pay me enough.
[501] Otherwise, don't pay me. You've also done some research that suggests that making an incentive too large can cause people to choke, especially on tasks involving cognitive capacities such as attention and memory.
[502] Can you tell us about the study you conducted looking at the effect of very large incentives on performance?
[503] So think about public speaking or imagine that I ask you to do, say, free shot in basketball.
[504] I don't know how good you are.
[505] Maybe you'll do two out of ten.
[506] Now imagine that I'll pay you a million dollars if you'll be successful.
[507] Oh, boy.
[508] Right, exactly.
[509] So things that you would have done automatically before that, now you started thinking about it.
[510] And it turns out that in many cases, thinking about it is much less good than just doing it automatically.
[511] So just don't tell me that you'll give me a million dollars.
[512] I have a 20 % chance of doing it.
[513] If you'll offer me a million dollars, I'd probably be so far off because I'll be too excited.
[514] And that's basically what we found.
[515] We found it in cases in which thinking is not what you need to do.
[516] you really need to follow your instinct and do the automatic things?
[517] People were less successful when we did this.
[518] They basically choked under pressure.
[519] You said that incentives can also be a powerful truth -seeking device.
[520] They can reveal preferences that would otherwise remain hidden.
[521] I understand that some companies offer to pay employees to quit their jobs.
[522] Why would they do this, Uri?
[523] I love this.
[524] So I teach negotiation and I tell me students that the only time that they should lie is when they're interviewed for a job, they need to show excitement.
[525] If you're not excited about the job, why would I hire you?
[526] But because of that, if I'm the employer and they want to keep only the people that really want to be there, because they're much more productive.
[527] But if I'll ask you, are you excited to work for me?
[528] You'll say, yes, of course.
[529] How can I actually use incentives to find out who really cares about me?
[530] Instead of asking you, are you excited to work for me?
[531] I'll tell you, look, Shankar, here's $10 ,000 if you want to leave.
[532] You're welcome to leave.
[533] Thank you for working with us.
[534] We are happy and good luck.
[535] Now, if you're happy to work for me, you'll say, no, no, no, thank you.
[536] I don't want the $10 ,000.
[537] I'm happy over here.
[538] And if you are like on the margin and you're not sure, you'll just take the $10 ,000.
[539] We'll depart as friends, which is also important because otherwise you can bedmouth me, you can sabotage.
[540] I don't want to fire you.
[541] So I can really select on the people that want to be with me in the company.
[542] And by the way, it could have other effects.
[543] So now, because I gave up $10 ,000, I really have to justify this.
[544] If I'll go to work on Monday morning, now I really need to justify it.
[545] I really need to work hard in order to show that I was not stupid to give up the $10 ,000.
[546] So these people are going to be surprisingly even more motivated than they were before.
[547] It's interesting.
[548] So in some ways, it's revealing what might actually be in people's hearts, including things that they might not know themselves.
[549] Once you're actually faced with an incentive, you might come to a different decision, and that could actually reveal something, not just to the company, but even to yourself.
[550] It's really amazing how little we understand our preferences in many cases.
[551] So very often we get people that get a job offer and then they decline it, and not because they needed a job offer in order to negotiate.
[552] But really because they thought, you know, they come to San Diego, say, in my case, they come from a cold place now.
[553] It's very nice outside.
[554] They enjoy it.
[555] It's a vacation.
[556] They're really happy about this.
[557] They say, okay, we want to live in San Diego.
[558] Diego.
[559] Then they have the offer and they start thinking, oh, we have to take our three little kids from school.
[560] That's going to be hard.
[561] Where are we going to leave?
[562] They start thinking about the details and they discovered that they don't want it.
[563] So in many cases, we don't really know our preferences until we face the decision that we need to make.
[564] And with the pay to quit, until you are offered $10 ,000, you might not even understand that you are not happy where you are.
[565] You will not be very motivated.
[566] You will not be a good worker, but you won't understand why.
[567] You know, one of the things I've noticed as we've talked to, Rhee, is that when we've seen all the ways that incentives work and how they don't work, it seems to me that it's really important to run lots and lots of experiments because what works in one context might not work in another.
[568] What works at one point in time might not work at another.
[569] I'm really struck by how little such thinking shapes our efforts, whether that's at the level of companies or the level of countries.
[570] So often managers and leaders tell themselves, I know what needs to be done.
[571] My political loyalty is.
[572] tell me what's going to work.
[573] And especially I think when you feel like you're doing something important, something you care about, it's so easy to forget to question yourself, to doubt yourself.
[574] It's really, really hard to understand the importance of what you said.
[575] Think about culture.
[576] So I live in the U .S. for 22 years now, and I'm still surprised very often because they don't get the signal.
[577] So I learned that when American tells me, let's do lunch, basically they mean we don't want to see you again unless they say let's do lunch on Monday right in Israel everyone is much more blunt so you need to calibrate and the same is true about incentives and it doesn't have to be between different cultures it could be cultures could be also taxi drivers versus teachers versus bankers in San Diego and different people have different things that motivate them I talk in the book for example about incentives to buy Prius to buy hybrid cars this will not work on the guide that is driving the big pickup truck they don't care about the environment.
[578] They don't care about signaling that they care about the environment.
[579] You need to find the group that is really relevant for your incentives and find what works on them.
[580] So it's not that one size fits all.
[581] Different situations, different people, different cultures will require different things.
[582] And the way to do it is first to think about it, right?
[583] But then we try to see whether it works, and then maybe you need to tweak it a bit.
[584] And then maybe after you have it for a while, you see that people found the way to game it.
[585] so solve this problem.
[586] So it's a dynamic process that needs a lot of data in order to make sure that you're right.
[587] Yeah, and especially I think in politics, I don't know if you'd agree with me, but I think in politics, this is often so not the case, right?
[588] If you think about so many of the political debates that we see around us, so many of them are basically saying, I'm right because of this ideology or I'm wrong, you know, you're wrong because of that ideology, as opposed to saying we can actually run an experiment to figure this out.
[589] Let me ask you a question.
[590] Imagine that you have a big company and you invite two consultant.
[591] One of them tells you, I know the solution to your problem.
[592] Give me, you know, a million dollar and they solve your problem immediately, no problem, everything is good.
[593] And the other consultant tells you, I have no clue what's the right thing, but I can design a study that will tell you.
[594] Which one will you take?
[595] Many people will take the first one.
[596] And I think you want your politician to know what's right.
[597] You don't want them to run experiments in order to find out.
[598] You elect them because you think that they are good at it.
[599] That's what they have to present.
[600] And that's why the entire culture of politics is not, look, I'm good because they know how to run experiments and find out what's the right thing to do.
[601] No, I want a politician that knows what's the right thing to do.
[602] And I think that that's a mistake.
[603] Uri Gnizzi is an economist at the University of California, San Diego.
[604] Uri, thank you for joining me today on Hidden Brain.
[605] It was a great pleasure.
[606] Thank you.
[607] Hidden Brain is produced by Hidden Brain Media.
[608] Our audio production team includes Bridget McCarthy, Annie Murphy Paul, Kristen Wong, Laura Querell, Ryan Katz, Autumn Barnes, and Andrew Chadwick.
[609] Tara Boyle is our executive producer.
[610] I'm Hidden Brain's executive editor.
[611] For today's Unsung Hero, we bring you a story from our sister show, My Unsung Hero.
[612] The story comes from Sarah Feldman.
[613] In 2001, my hometown of Houston was hit by a tropical storm, and I was on vacation in Connecticut.
[614] We couldn't go home that day, of course, so my father decided to take me to the Bainiki Rare Book and Manuskip Library in Yale University.
[615] I was around 13 or 14 years old, and I was asking this gentleman named Bill Carver a bunch of questions, like, how do you keep all these rare books okay?
[616] And what's the oldest book you have?
[617] And then I told him about how my grandparents went into my house that was flooded and all of my books, lots of them, because I was a big reader as a child, were destroyed.
[618] And he said to me, Sarah, I'm going to send you a book in the mail so you can start your collection again.
[619] And he did.
[620] And I got a really nice book called The Medieval Book by Barbara A. Shaler.
[621] And I also received a letter in the mail, and I'd like to read it to you.
[622] December 8, 2001, dear Sarah, I have often thought of you and your family, and that terrible flood of 2000.
[623] He meant 2001.
[624] As promised, enclosed is a rather scholarly book, which you may be slightly mature for your age.
[625] But I thought it might help you shape your new library collection.
[626] And as the years progressed, this point.
[627] book may grow in stature and value to you.
[628] Have a happy holiday season.
[629] Wish you and loved ones all the best.
[630] Cordially, Bill Carver.
[631] I want to tell him that I'm 34 now and my life is great.
[632] I'm married.
[633] I have a dog.
[634] And my partner and I like to visit cathedrals in Europe and go to the Cloisters Museum in New York City from time to time.
[635] and I'm obsessed with medieval art. I actually sent the library a letter during the pandemic, but I never heard back.
[636] I'm not sure if he's alive still because he seemed old to me when I was 14 at the time.
[637] But I'd just love to tell Bill Carver that he changed my life.
[638] Sarah Feldman lives in New York City where she has her own jewelry business.
[639] If you liked today's episode, please be sure to share it with a few friends.
[640] I'm Shankar Vedantam.
[641] See you soon.