looking at all kinds of behaviors in which we might have the same kind predictable, srrxsñonal tendencies and we it something more systematic? that is actually important question because if there were three bad apples he would want to eliminate the bad apples but it is something more general maybe we need to think about a different solution.

so like everything we do i decided to do a small experiment and the basic structure of my experiment with the following. a tecate sheet of paper with 20 simple math problems. onesy founded to numbers, you solve the problem. everybody can add two numbers 210, it just takes time to figure out which of the two numbers to add to ten. i gave 20 questions and i said you have five minutes and when the five minutes are over i will give you a dollar for each question you get correct. you would wert five-- card for five minutes into account how many questions to saltz correctly and i would pay you on average $4. that is what people's alben five minutes. in the second condition, the cheating condition when the five minutes are over we ask people to shred the piece of paper. put the pieces in their pocket or their back packin tell us how

many questions they saw it correctly. presumably people are not smart and because they shred the piece of paper but now they claim to sell seven. it is not its if we had a few bad apples that skewed the distribution. instead we have lots of people who cheated just a little bit. why do people cheat? the bekka model of cheating is every day when we walk by anything we consider three things. we say to ourselves how much money is in the till? what is the chance they will catch us? how much time we get in prison? [laughter] we weighed the cost benefit analysis and decide whether-- it sounds like a crazy model and it is. but realize this is the model driving the legal system. so in some sense it is an odd modeled but it very important model in a practical way because of the role of economics in our society. so, let's check this model.

how would we checked that model? a big part of it is how you spend today. some people tell them we would pay 10 cents for question, others 25 cents per correct question. would it matter how much money people stand to gain? you would think so. it turns out a dozen. a lot of people cheat but they cheat a little bit and don't seem to be sensitive to how much money they stand to gain. what about the probability of being caught? some people we asked to shred half the seats so there was evidence remaining. some people shredded the whole sheet and some people shredded the whole sheet, left the room and-- pay yourself, chitty is. [laughter] again. you would think people would cheat more as the probability of being detected guss done. no, they don't. a lot of people cheat but they chief just by a little. who doesn't seem like the

economic factors are rising, what could be driving it? we thought maybe what is happening is there are two conflicting forces. one forces we all want to look in our mayor and the good about ourselves. at the same time we also want to benefit from cheating so maybe what we do is we find an equilibrium and resting point between those two points. to the level we will be comfortable with our image in the mayor. call it the fudge factor. how little we can cheat and still feel good about yourself. think about tax returns. how could you tell us something like this? how could you test an idea for fudge factor? what we want to do was find things that would shrink the fudge factor in expand a fudge factor and the ideas for both of those came from two jugs. let me tell you the first one. i will give you the jewish version but the guy goes to the rabbi and he says you won't believe that the last week when

allison synagogue somebody's sold my bicycle. the rabbi said this is terrible. next week recited the ten commandments and standing in front of the congregation look back inman me get to that shall not steal c2 lansdown bachelet hen you will know those of you when that would steal the bicycle. next ricci sits in the front and looks at the back and the rabbi and anxious to know of this trick worked catches him at the outside on the way out and said did it work? you would not believe it, it worked like magic. when we got to daschle not commit adultery, i remember where i left my bike. [laughter] where is the experiment in this thing? so, we got people to come to the left and we said we have to experiment for you today. we ftap the people to recall ten

books they read in high school and the other have to recall the ten commandments and by the way nobody can recall the ten commandments in people invented all kinds of interesting commandments. , after they tried to recall the ten commandments we shredded the piece of paper, no cheating whatsoever. it wasn't as if the people who remembered more commitments presumably the more religious loans cheated less and the ones who remembered view cheated more. nobody cheated after thinking about the ten commandments. in fact even when you get the self-declared atheists diss were on the bible, they don't cheat. so there is something about thinking about morality that seems to be shrinking the fudge factor. so we said what about the honor code? we got people to come to the lab and they signed. i understand this service con-- falls under the honor code. no cheating whatsoever.

mit does not have the honor code. [laughter] so they just signed it. but, this actually has an important second point which is we ran this experiment at princeton. princeton has a strong undercoat and they take the students and run them through an honest the boot camp for a week. we called them three weeks after the boot camp and measured how much they were cheating. what you think, were the princeton's students cheating less that the mit students? nothing. no effect. at the same time when the prince and/or mit students sign the honor code nobody cheated. this is a very general lesson. if you think about how do we treat social sustaining changes it is very hard. the week-long ethical training did nothing for the princeton students. of course this was not a random simonson you could say princeton students started warsaw.

[laughter] but if we assume it is a general finding, it suggested things, one it is hard to create cultural change and at the same time hard to create at the moment of expectations. when people sign it, they don't cheat and we all signed the tax forms at the bottom. why don't rerivers that? so we went to the irs and we said we found this thing, why don't you get people to sign at the top? the irs said they had better things to do, understandably. but we did find an insurance company. raise your hand, don't be bashful. not you, but we found an insurance company and it turns out that property claims are a big deal, about $24 billion in losses from exaggerated claims. it turns out the people in the insurance industry pink people cheat very much like we find. it is not that people intent burglaries' but tutorings become

three and stuff like that. we said why don't we get people to sign those declarations on the top. this is too difficult and complex but we have this form are we as people to tell us how much they drove last year. this change is their policy. could you use that form? it is a very simple form. it has one number. some people cited at the top and filled in the number and some people killed in the number and signed the bottom. 2700 miles a year difference just by doing this. think about it, just signing something, a second before you fill the number, not you, other people. dramatically changes the level of cheating. as a side comment i will tell you that i often deal with individuals irrationality. cos irrationality. subwe found these things, the signature on the mileage form

about nine months ago. since then this insurance company who i will not name has not implemented this trivial change of the form. not only that, they are not willing to let us do it with the forms. we were offering them basically a free lunch to reduce deception and because of who knows what reason they were not willing to even tryout so well we talk about human irrationality sometimes people think companies are perfectly rational. i think it is the opposite. what ever is to find is actually much worse. the insurance industry, thank you. [laughter] this is about shrinking the fudge factor. what about expanding the fudge factor? what would we do there? here is another little joke. ridge little baby comes home from school with a note from the teacher that says david stolypin so from a kid sitting next to him. his father is firrea census david, this was not how we raise

to. i am so embarrassed and humiliated. this is not how we behave. you are grounded for i don't know long. david, if you need a pencil just say something. i can bring you a dozen pencils from work. [laughter] now, why is this amusing? it is amusing because we all understand that people have different standards taking 10 cents from the petty cash box compared to taking it pencil from work, even though the cost could be the same. with chip-n-saw we can tell ourselves all kinds of stories. beacon say everybody does it. even if we give the pencil tabarra kidwell leave us alone and we can check our blackberry qa-- all of these stories helpless to justify things. so we tried this with two different experiments.

itis jubah did six packs of coke in the refrigerators and i came back in measured what we technically called a half a lifetime of coke. how long does it stay in the refrigerator? as you can imagine it was very quickly the cope disappears. i compared to when i go to the same refrigerators and leave a plate with six, 1 dollar bills. the kids could have gone to the vending machine, gopnik kogan change. nobody tick a single dollar bill. this is not a good experiment because you often find coax and refrigerators. so we tried another version of the experiment. weiss the same procedure i told the before with these 20 simple math problems. some people shredded it and came to wesson said the eyes of the problem, give me x dollars pickle the third of the people shredded it and said i saw the problem, give me tokens. we paid them in tokens.

the walked wealthy to the side and changed at $4. in a since they traded for a pencil something once for money but this thing was going to be just for a few seconds. what would happen? ours subjects in doubled their change. to think about it. in the economy today we have so many things that are multiple steps removed from money. could it be that somebody who would never think about taking $100 from the petty cash box, even if nobody could never find it could find it morally easier to backdate stock options? it is not asking for more. just change the date a little bit. all of those things make it much easier. the final experiment on want to tell you about is a made up experiment. bin this version, the first one was that we prepaid everybody's so if you were an experiment he would end up with all the money up front and we would ask you at

the end of this bair to pass that the amount of money you did not make. it was the first change. the second change was that we hired an acting student to stand up, 30 seconds after the experiment started and say i sold everything. it is clear that nobody could solve everything in 30 seconds. the experiment said if you solve everything, you are free to go home. so imagine yourself in this situation. you are in a room, doing a task, somebody is cheating in an egregious way. what would it cause, what will it do to your own moral standards? well, what happened was people cheated more. there are two possible explanations. this verse explanation is this guy made it clear you could cheating get away with it so it was about any moral changes, just about the fact that you just proved to me the cheat on these experiments and get away with that.

the second possibility is that this guy created a social exceptions of cheating. it was in about the probability 1t dñ,ñ

[laughter] now, why is this? by the way it went down not just the love how much cheating was there when there was-- it went down below when there was cheating when nobody was standing up. so lenson since the act of watching somebody who you don't feel the affiliated to doesn't only, not increase cheating but it decreases cheating. what happens is basically when somebody from your in group is cheating, nuc them cheating, there is more increased acceptance of cheating. when it is somebody from and out group, edward xcel little bit like the ten commandments. only the bad people from virginia tech commisar the other school we don't want to be affiliated with them on top of this an awareness of cheating. i really don't want to think of myself and this. let me say a couple of comments

about the stock market and a couple of comments about madoff. maybe i will start with made up. what is known to be the legacy of made of? this has been a terrible event by all accounts, but one of the things that worries me the most was the last experiment in may of. think about the financial community. aidid kind of a tight community. everybody knows each other. what happens when someone in your community chiefs and an egregious way? i don't think anybody else in the community will necessarily start a new ponzi scheme but i think there's a good chance there is the more flexibility within greece as a consequence of this thing. maybe not big things, there would be all kinds of little things that would increase. the second thing is that if you look at the economy and and all the experiments i told you about, we tested about 6,000 students. and all of these experiments,

there were about a handful of cheated-- people who cheated a lot. and there were thousands of people who cheated a little. i lost $100 to the people who cheated a lot in thousands of dollars to the good people who cheated just a little bit. in fact the sectionally reflects what is happening in society. blue-collar climb-- is tiny compared to other fraud, so car thefts, all of those together, all of those together cost less than exaggerated insurance claims that people's of met. in fact they cost almost as much as what is called wardrobing. wardrobing is the practice of going to a clothing store, buying clothes, getting them to be smokier torn and refounding them for a full refund. do you lose the same amount of money every year in warming to what we lose in blue-collar crime. there's a huge imbalance.

there are things we pay attention to like bad apples. what we don't seem to be paying attention who are the small ax by many people accumulating together to be much more. in the madoff case, terrible's it is, is nothing in comparison to walled the small cheating that happens in the banking and insurance industry that have led us to this catastrophe. what does it say in a positive way? behavior economics is a depressing view of life. more like, send an super man. but there is a silver lining and the silver lining is understanding what is going on so we can think about how to solve things. in this particular case i think what it tells us is when we put people in the situation of high conflict of interest, people will be biased. if you stand to gain $10 million for reviewing mortgage-backed securities as a good product you will be able to do it. we have enough mental flexibility to justify lots of things especially with their

financial incentives. it is not cheating that we do on purpose. it is just the fact that when something is very good for us we see things in that way. biases our perceptions. the people in our experiments don't think of themselves as bad people. the second thing is the ability to see reality in a different way, gets helped by the fact that when we deal with things with money. money is harder. money-- imagine if you try two of the leeway to this and you have 50 different equations. wouldn't pick the one that issue the best possible outcome for your personal gain? of course. we still don't know how much mortgage-backed securities over. what are the chances that people could deal with that? finally what happens when we deal with everybody around this is doing the same, and the same thing. the odds are that we will take those standards as well. if webistan doesn't think what we need to think about is how do

we eliminate conflict of interest? nobody will ever think that thing a judge by giving him 5% of the verdict the pastor brown is a good deal because we understand a conflict of interest would be terrible. why is it okay to put their physicians and stockbrokers in the same position of being in such terrible conflict of interest? i think we should just understand how a conflict of interest works and how-- try to eliminate it. let me end by telling you something a bit more personal about economics and life. when i finished all of my research on pain of went back to the hospital and i called the nurse's. one of my favorite nurses said, to the interesting things. the first one she said was i did not taken to a counterpanes and she said you were sitting there and it was not pleasant for me either. and nobody wants to be tortured-- torture their

patients. it in the burn department they get to know us very well but we agreed the coal of the treatment was not to minimize her pain. [laughter] it is regrettable but the second thing was that she said you know, my intuition was not the same as yours. i didn't think it was going to help. even for a couple of days to do it the way he proposed was going to cost me, not money but time and taken the and i did not think it would work. therefore it did not want to do the experiment. i think this is a very general lesson. many times in life we have intuitions about things. how many people have interviewed somebody for a job? how many people suffered? very few people think-- how would you ever know, and the majority think they are fabulous. you have a great connection and

understanding of how good they will be. how would you ever know? how would you know? the truth is the only way to know is to hire people you think will not work out and see if indeed you are correct. [laughter] and the only organization by the way that does this is the israeli air force. seriously. [laughter] the israeli air force takes people every year that they think will fail the pilot training which is incredibly expensive. we are talking millions of dollars in the tickets the people and put them into this training and when these people fail, they have just wasted money but when the people succeed they go back and say what have we missed in the test? that is actually what he was telling me about these experiments. experiments are very hard to do. you have to work against your intuition. you have to say i am assuming that i know how to tell people apart by talking to them five minutes and in dating it is even

worse. you only have four minutes in speeded dating to find that if you are a good fit. if you understand those are just intuitions, the bigger thing, for personal, business and government is to say how do we test those things? how we become a bit more systematic? how do we really know that this is just intuition or it is factually correct and i think if we started testing things more we would actually get a lot of benefit and i think that is the real hope of behavior economics. thank you very much. [applause] sorry. [applause] any questions? [laughter] >> are you looking for how to pause?

>> you can shout. >> regarding madoff comment you think the sentencing or the giving back or redistribution or rebidding of taxes or anything like that, do think that will have some kind of effect on the cheating that is probably going on out there right now? >> so, sadly i think, okay. there is something we call the victim effect. it is the idea that we pay much more attention to single offense then statistical occurrences, so do you remember jessica who fell in the well? it turns up baby jessica got more coverage that rwanda. you might wonder why that is. wind tragedy, 10,000 statistics.

one of the problems which made of is it is such a single, cleary offence that i think it will have a huge impact on how people behave. what they do with the money at the end of the day i don't think is going to matter that much. i think he will increase, bisek prediction is he will increase. it is not going to matter that much but let me tell you something i think that could have been positive with the aig bonuses that we have all been reading about. here we have another very clear case. wiki of the aig billions of dollars. who cares about a few million dollars. why do we care? we care. there's no question that we care. we care because we think there is injustice, because we think it is betrayal. there is a proposal of capping executive salaries. is this a good idea? i think it is a terrible idea. let me tell you why.

imagine if they ask you to follow, would you help me change a tire on my car? the good chance is he will say, dan seems like a nice guy, why not? what about if i offered you $5? now you say $5, that is work. i don't work for $5. offer me $150 we can talk. [laughter] that is what we cnn experiment when we ask people to drexel was and all kinds of things. if we paid them a little bit of money people get demoted. its switches us from a social-- and wanted switches, we lose much of the motivation. i thank sadly $500,000 for banking executives is a little bit like offering $5. offering of $5 to help change a tire on my car. in some terms it is more offensive than helpful.

not only should it have been zero because it is this the-- a appropriate thing. the moment you offer them $500,000, i don't work for five find a thousand dollars. this is the gnats. for me, i am offended by think zero is in them interesting number because it could have for trade to things. view are doing it at of the goodness of your hoppy cazayoux understand this is a social problem and the second is to help retrain ranking not as a greedy, kind of profession but basically would have given people the feeling that they are the keepers of society and in the economy which we have learned sadly is the correct way to view them. and could have injected more pride into the banking system so i think for example taking the compensation of ceo's and framing it correctly which i think is to pay people nothing could have had a very good

social impact overall. .. in this situation and expect him to behave morally. it can't happen. it is not part of human nature. 2008, sadly, was a good year for behavioral economics. alan greenspan spending in front of congress and saying it was a

good moment for us. what we really need to think about is why do people significantly do that and let the free market operate and what is the fundamental problem people have to prevent them from operating well if they are not supervising? it would not help just to have this ceo compensation. we also have to have line compensation for everybody. the problem with the bank system is it a is not aligned with the behavior we want. the conflict is basically about that. >> one more. >> i would like to know how an experiment -- and also did you find there were people who did not cheat? >> two very good questions.

how do we know that people cheek? the first thing we know is we have a group that doesn't have a chance to cheek and we have a group that has a chance to cheat and we randomly assign them. if we are doing it here, some of you will randomly not be able to cheat and some of you have a chance to cheat. those who randomly have a chance to cheat, we know you are cheating on it is unlikely to happen by chance if i give the sheet to the people who are smarter in some sense. in other experiments, what we do, sad to say, we take our subjects. we have a shredder in the room. we eliminate the teacher in the middle of the shredder. it is we eliminated peace in the middle of the shredder. [laughter] it is for science.

it is okay. we did this experiment recently in some bars. we went to washington d.c. and tested where the congressional staffers hang out. we went to new york city, banks were bankers, hanging out, wanted to see who cheats more. two things, the bankers sheet more, the politicians seek more, you have to vote, you can't abstain. who votes for bankers? politicians? let me tell you, i thought the same as you. turns out bankers to eat much more. thank you. [applause] ity center in washington

hosted this event. it is just over an hour. [applause]

>> thank you, everybody, for that wonderful welcome, and thank you for your kind introduction. it will make my speech easier. ladies and gentlemen, where i live, in the united kingdom, summer last year was cool and wet, and the winter was the coldest of this century. these affects were not just in our country, they were felt in china and parts of this great continent as well. the gulf of mexico last year, last summer, was 8 degrees celsius cooler than it has been in previous years which is just as well. it did not do the damage that its sister, katrina, did the year before. my friends among clima