1 00:00:00,659 --> 00:00:05,950 Everything I told you up to now was about one chance of cheating. 2 00:00:05,950 --> 00:00:12,222 But of course, an imposing question is, what happens when we give people more 3 00:00:12,222 --> 00:00:17,046 chances to cheat over time? So we created a task like that. 4 00:00:17,046 --> 00:00:20,349 We showed people some dots. On the computer screen and we asked them 5 00:00:20,349 --> 00:00:23,217 are there more dots on the right or more dots on the left. 6 00:00:23,218 --> 00:00:27,242 Tell us which one has more dots, but just so you know, you're getting paid ten times 7 00:00:27,242 --> 00:00:30,824 more if you say right. And this came up to equivalent conflicts 8 00:00:30,824 --> 00:00:33,990 of interest tasks. When you ask a doctor for example, whats 9 00:00:33,990 --> 00:00:36,690 better for this patient, treatment A or treatment B? 10 00:00:36,690 --> 00:00:39,499 But you, the doctor, gets ten times more if you say. 11 00:00:39,500 --> 00:00:43,916 Let them go for treatment A. So what happened with this task we could 12 00:00:43,916 --> 00:00:46,430 do many, many, many of them, hundreds of ,them over time. 13 00:00:46,430 --> 00:00:50,812 And what we saw was that many people started by cheating just a little bit. 14 00:00:50,812 --> 00:00:54,070 Cheating just a little bit. Feeling good about myself, cheating a 15 00:00:54,070 --> 00:00:57,924 little bit, feeling good. But at some point, many people switched. 16 00:00:57,925 --> 00:01:00,230 And all of a sudden, started cheating all the time. 17 00:01:00,230 --> 00:01:04,451 And this switch happened to different people at different times, but it was very 18 00:01:04,451 --> 00:01:06,928 salient that there was this switch occurred. 19 00:01:06,928 --> 00:01:10,180 Our issue was where is this switching coming from. 20 00:01:10,180 --> 00:01:13,838 And one of the thoughts we had was that maybe this is what we call the What the 21 00:01:13,838 --> 00:01:16,046 Hell effect. If you think about it, you've been trying 22 00:01:16,046 --> 00:01:18,862 to feel good about yourself and cheat a little bit, and at some point you cheat 23 00:01:18,862 --> 00:01:21,485 too much. And you can't feel good about yourself any 24 00:01:21,485 --> 00:01:23,776 more. You might say, forget about it, let me 25 00:01:23,776 --> 00:01:26,550 just enjoy it as much as I can, and cheat all the way. 26 00:01:26,550 --> 00:01:31,533 And that's indeed what we see in the data. So if you had a little bit of cheating and 27 00:01:31,533 --> 00:01:34,007 then cheating a lot. And by the way, a lot of people reported 28 00:01:34,007 --> 00:01:36,993 it happened to them in dieting. When you start the day and you decide to 29 00:01:36,993 --> 00:01:40,399 be good on your diet and then you eat the muffin or you eat something bad you say 30 00:01:40,399 --> 00:01:44,480 forget about it. I'll star tomorrow, I'll start next, next 31 00:01:44,480 --> 00:01:47,192 Monday. So if we have this pattern, when people 32 00:01:47,192 --> 00:01:50,705 cheat a little bit and then they switch to cheating all the time. 33 00:01:50,705 --> 00:01:53,920 We thought what ever stop people from cheating all the time? 34 00:01:53,920 --> 00:01:59,902 What would ever reset this fudge factor? And, and we went back and we thought about 35 00:01:59,902 --> 00:02:03,287 religion. And if you think about the Catholic 36 00:02:03,287 --> 00:02:07,238 confession, I'm of course not Catholic. You can ask what is the logic of the 37 00:02:07,238 --> 00:02:10,521 Catholic confession? Now if you actually came to the Catholic 38 00:02:10,521 --> 00:02:14,828 confession from an economic perspective. You would say that if you can sin and get 39 00:02:14,828 --> 00:02:18,200 absolved you should cheat more, because you can get absolved. 40 00:02:18,200 --> 00:02:21,610 In fact, you should sin on the way to confession. 41 00:02:21,610 --> 00:02:25,161 If you want to minimize time in purgatory, the only thing you want is to cheat just 42 00:02:25,161 --> 00:02:28,724 before you confess. But of course this is not what the 43 00:02:28,724 --> 00:02:33,582 religion thinks, and we went to talk to all kinds of Catholic priests about this. 44 00:02:33,582 --> 00:02:37,878 So if we don't think that confession increases cheating, how could it reduce 45 00:02:37,878 --> 00:02:41,192 cheating? One theory is that confession adds to the 46 00:02:41,192 --> 00:02:45,010 cost of the crime. You say to yourself like in the Baker 47 00:02:45,010 --> 00:02:49,105 theory, not only do I have a chance of being caught and go to prison, but I have 48 00:02:49,105 --> 00:02:53,265 to talk to somebody and they would make me feel bad about it and let me, recite, 49 00:02:53,265 --> 00:02:57,547 something 1,500 times, that's not good. Not worth it, let me not cheat. 50 00:02:57,547 --> 00:03:00,833 Not much evidence on there. We thought maybe what happened is that 51 00:03:00,833 --> 00:03:05,136 just like the Ten Commandment experiment, that the moment you get out of confession. 52 00:03:05,136 --> 00:03:08,800 You feel wonderful and pure, and you don't want to disturb that feeling. 53 00:03:08,800 --> 00:03:12,635 So for a little bit of time after that you're behaving more consistent with the 54 00:03:12,635 --> 00:03:16,721 tenets of the religion. Or we thought maybe what's happening is 55 00:03:16,721 --> 00:03:19,372 confession is a particularly useful idea for resetting a new page. 56 00:03:19,373 --> 00:03:23,295 If you cheat all the time, why would you ever stop? 57 00:03:23,295 --> 00:03:27,857 Unless you can start a new page and therefore start again. 58 00:03:27,858 --> 00:03:30,755 So we tried experiments like this. People cheated a little bit, they start 59 00:03:30,755 --> 00:03:34,270 cheating a lot then we stop them. So here's a piece of paper. 60 00:03:34,270 --> 00:03:37,707 Would you like to write something about something bad you have done recently? 61 00:03:37,708 --> 00:03:40,777 People wrote something and we asked them to shred it. 62 00:03:40,778 --> 00:03:44,999 Then we gave them another sheet and said, would you like to ask for forgiveness from 63 00:03:44,999 --> 00:03:48,990 whatever deity you believe in? Please write it down, and shred it when 64 00:03:48,990 --> 00:03:51,974 you're finished. What happened after people did both of 65 00:03:51,974 --> 00:03:54,319 those elements? Saying what they have done and ask for 66 00:03:54,319 --> 00:03:57,582 forgiveness? Cheating went down dramatically. 67 00:03:57,582 --> 00:03:59,640 Now regrettably, later on, it went up again. 68 00:03:59,640 --> 00:04:04,790 But at least for awhile, this idea of opening new pages seemed to be effective. 69 00:04:04,790 --> 00:04:07,847 And I think this is one of the things that we can actually take from religion. 70 00:04:07,847 --> 00:04:13,145 We can say, if religion has evolved to help men deal with some of our challenges, 71 00:04:13,145 --> 00:04:17,382 some of our fallacies, some of our weak points as human beings. 72 00:04:17,382 --> 00:04:22,849 What can we learn from religion and should we incorporate similar mechanism in other 73 00:04:22,849 --> 00:04:26,223 places in society? For example, should we create a 74 00:04:26,223 --> 00:04:30,390 confession-like mechanism for citizens submitting the IRS? 75 00:04:30,390 --> 00:04:35,720 Should we create a confession-like mechanism for some of our politicians for 76 00:04:35,720 --> 00:04:39,708 some of our CEOs? Is there something important in starting a 77 00:04:39,708 --> 00:04:44,622 new page and allowing people to all of a sudden reset the level of honesty and 78 00:04:44,622 --> 00:04:48,505 behave better? And of course, a big scale version of this 79 00:04:48,505 --> 00:04:54,081 was the Reconciliation Act in South Africa, where they basically created the 80 00:04:54,081 --> 00:04:57,783 structure. Where people basically, in, in, in a very 81 00:04:57,783 --> 00:05:03,220 important public stage both ask for forgiveness and started a new, a new page. 82 00:05:03,220 --> 00:05:07,300 And you know, apartheid, apartheid is of course a complex, difficult thing. 83 00:05:07,300 --> 00:05:11,601 It will take a long time for anybody to get over it completely. 84 00:05:11,601 --> 00:05:17,061 But it is taking these ideas of saying you're sorry and starting a new page that 85 00:05:17,061 --> 00:05:22,467 might be important not just in religion, but in civic society, in many cases. 86 00:05:22,468 --> 00:05:26,474 Another question people ask me many times, is what about cultural differences? 87 00:05:26,475 --> 00:05:30,780 We travel around the world, we see all kinds of other cultures, some of us grew 88 00:05:30,780 --> 00:05:34,194 up in a different culture. And you can ask yourself, how is cheating 89 00:05:34,194 --> 00:05:37,470 different in different places. So we try this out. 90 00:05:37,470 --> 00:05:41,029 So I grew up in Israel, so the first thing I did was to try Israel. 91 00:05:41,029 --> 00:05:44,990 So what do you think? Do the Israelis cheat more or less than 92 00:05:44,990 --> 00:05:49,132 Americans? Actually I'll tell you that when you ask 93 00:05:49,132 --> 00:05:53,676 most people whether they think that people in their home countries cheat more or less 94 00:05:53,676 --> 00:05:56,579 than Americans. Most people think that people in their 95 00:05:56,579 --> 00:06:00,919 home countries cheat more than Americans, aside from the Canadian and the northern 96 00:06:00,919 --> 00:06:04,142 European countries. And in general, when we ask people what do 97 00:06:04,142 --> 00:06:07,799 you think, that Israeli cheat more or less than Americans, most people say that the 98 00:06:07,799 --> 00:06:10,440 Israelis probably cheat more but the reality was not that. 99 00:06:10,440 --> 00:06:14,216 We found that the Israelis were cheating just the same as the Americans. 100 00:06:14,217 --> 00:06:18,151 Francesca, my Italian, my Italian collaborator said you want to see 101 00:06:18,151 --> 00:06:22,380 cheating, come to Italy, we'll show you what this cheating is all about. 102 00:06:22,380 --> 00:06:24,584 We went to Italy, we tried the experiment there. 103 00:06:24,585 --> 00:06:29,348 Same results as in the States. We tried Turkey, we tried England, we 104 00:06:29,348 --> 00:06:32,670 tried Canada. So far, we don't find any differences. 105 00:06:32,670 --> 00:06:36,252 Now, we haven't tried everything, but so far, we don't find any differences. 106 00:06:36,252 --> 00:06:41,080 Now, how can it be that in one hand, when you travel in other countries, or you grew 107 00:06:41,080 --> 00:06:45,616 up in other countries, you feel the cheating in different countries is very 108 00:06:45,616 --> 00:06:47,896 different. On the other hand we don't find it in our 109 00:06:47,896 --> 00:06:50,940 experiment. How would you balance these two views? 110 00:06:50,940 --> 00:06:52,660 Is one of them right and is one of them wrong? 111 00:06:52,660 --> 00:06:56,050 No, I think they're both correct and here is what I think happens. 112 00:06:56,050 --> 00:07:01,877 Our experiments are basically without any cultural context. 113 00:07:01,878 --> 00:07:05,490 And therefore they check the basic fudge factor. 114 00:07:05,490 --> 00:07:08,470 And the fudge factor is in it's basic is something we all share. 115 00:07:08,470 --> 00:07:11,949 We can all rationalize about 15% of our actions. 116 00:07:11,950 --> 00:07:15,700 In a slightly moral way, and still feel of ourself as good people. 117 00:07:15,700 --> 00:07:19,738 So that's kind of the basic thing that is similar across many people. 118 00:07:19,738 --> 00:07:24,572 But culture is important, but what culture does, it doesn't change your morality 119 00:07:24,572 --> 00:07:29,207 across the whole range. It attacks a particular domain. 120 00:07:29,208 --> 00:07:33,744 Think about something like plagiarism. Cheating on exams. 121 00:07:33,744 --> 00:07:38,160 University cheating, different cultures have said, oh, this is really important, 122 00:07:38,160 --> 00:07:40,964 you can't cheat much. Other cultures have said that is not a big 123 00:07:40,964 --> 00:07:43,838 deal, you can cheat more. Other cultures have said, this is not a 124 00:07:43,838 --> 00:07:46,450 moral question. It is not a part of the moral domain at 125 00:07:46,450 --> 00:07:49,210 all, cheating in this case is just like playing poker. 126 00:07:49,210 --> 00:07:51,991 Don't worry about it. What about infidelity? 127 00:07:51,991 --> 00:07:53,894 Some cultures have said oh, this is a real big issue. 128 00:07:53,894 --> 00:07:56,229 Other countries have said this is not such a big issue. 129 00:07:56,230 --> 00:07:59,798 What about cheating on your taxes? Issue or not issue. 130 00:07:59,798 --> 00:08:05,090 Illegal downloads. Buying fake counterfeit fashion product. 131 00:08:05,090 --> 00:08:11,162 So basically what I think is going on, is that there is this basic fudge fact that 132 00:08:11,162 --> 00:08:14,240 we all share. But the way it plays out in different 133 00:08:14,240 --> 00:08:18,346 domains are influenced by culture. Where the culture can either shrink or 134 00:08:18,346 --> 00:08:21,450 expand the fudge factor. And in some cases, it can even take a 135 00:08:21,450 --> 00:08:25,100 particular behavior. And say this is not a moral question. 136 00:08:25,100 --> 00:08:29,792 In some countries with cultures with high corruption, bribing a policeman is not 137 00:08:29,792 --> 00:08:33,801 something that is immoral. It is just, this is how you do business, 138 00:08:33,801 --> 00:08:36,868 and so on. So then we need to think about, both about 139 00:08:36,868 --> 00:08:41,568 the fudge factor, as well as about what the effect of culture is on top of that. 140 00:08:41,568 --> 00:08:46,578 I did find one experiment. One experiment that showed a difference in 141 00:08:46,578 --> 00:08:51,147 cheating across cultures. Every time we do these experiments, it's 142 00:08:51,147 --> 00:08:55,762 important to conduct them in a similar way across different places. 143 00:08:55,762 --> 00:08:59,723 So what we do is, we go to other universities and we check students, or we 144 00:08:59,723 --> 00:09:03,108 go to bars. And when we go to bars, what we do is also 145 00:09:03,108 --> 00:09:08,174 adjust the payment scale. So a dollar in the states is not a dollar 146 00:09:08,174 --> 00:09:11,650 in, in India. It's not a dollar, a dollar in England. 147 00:09:11,650 --> 00:09:16,707 It's not a dollar in Turkey. So what we do is we go to bars. 148 00:09:16,708 --> 00:09:21,688 And we create a payment situation, such as answering four questions will equal one 149 00:09:21,688 --> 00:09:25,648 glass of beer. So now, this is kind of our international 150 00:09:25,648 --> 00:09:28,500 currency. Every four questions is a glass of beer. 151 00:09:28,500 --> 00:09:34,242 And the one place we found the difference was when we went to a bar in Washington, 152 00:09:34,242 --> 00:09:38,077 D.C. Where congressional staffers hang out. 153 00:09:38,078 --> 00:09:41,222 And then we went to a bar in New York where bankers hang out. 154 00:09:41,222 --> 00:09:44,952 And the question is, which one of those two groups cheated more? 155 00:09:44,952 --> 00:09:48,801 Now, I have to admit that I expected the politicians to cheat more than the 156 00:09:48,801 --> 00:09:51,167 bankers, but the results were the opposite. 157 00:09:51,168 --> 00:09:56,604 The banker cheated twice, two to one. But I do have one word of caution, which 158 00:09:56,604 --> 00:10:02,264 is these were congressional staffers. So these were junior politicians, and 159 00:10:02,264 --> 00:10:07,326 maybe over time there's room for growth in terms of their dishonesty. 160 00:10:07,326 --> 00:10:42,087 [music].