good afternoon everybody let's start our meeting with our guest mrs carla hoff who will talk about spite stigma and economic development you may think that these are three words that are far apart one from the other especially if you follow the usual approach to studies on economic development but you will probably understand from professor hoff's presentation that there are important connections there are very deep connections between these three elements which refer to deep cultural psychological social and historical characteristics within which individuals make their own life decisions and the possibility that exclusion marginalization and poverty occur so you will listen to a way to interpret research on poverty and a new interpretation of the role of the economist with tools and viewpoints which may not be known to the large audience but which are a very fertile field of research at the moment a field which is expanding rapidly our guest is a highly qualified witness of this kind of studies she is visiting researcher at the princeton university and she's also senior research economist at the world bank her studies have already been published in numerous illustrious magazines and she's also co-author of a book with professor durlaff with the title poverty traps those poverty traps in which individuals or social groups may become trapped so i would like without further ado to leave the floor to our guest and at the end we will have some time for a q a session so thank you very much for including me in your last year's conference i've never been to a festival of economics thank you for inviting me i've never been to a festival of economics i think it's a brilliant idea it's fun for me it's fun for a lot of other people my talk is closely related to john's talk this morning and also towards talk a moment ago in that it builds on relationships between psychology and economics so just to set the frame there's a field of psychology which begins with the subject of preferences there's a field of economics which takes preferences is given and it begins with their consequences and that was until the field of behavioral economics or much much earlier all the work in the 17th and 18th centuries which was completely ignored in economics in the 20th century until the behavioral revolution the old work emphasizing psychology like the new work in behavioral economics emphasizes the link between economic structure and psychology if you seek esteem the structure of your economy will influence those activities which are viewed as estimable so though certainly the deep psychology of man is the same everywhere the drive for esteem will take you very different places depending upon where you are so there's a crack in the wall between economics and psychology distributional preferences in particular i'm going to argue in my talk can be learned what i mean by distributional preferences are other regarding preferences altruism spite there was a famous paper 28 years ago entitled economist free ride does anybody else do you know this paper a non-representative sample knows this paper in the u.s it got a lot of play so the idea was economists played and exp so as to students were asked to play an experiment in which they could contribute voluntarily to a public good or not and it was discovered that the econ majors followed the assumptions of economics were completely selfish they didn't contribute but the non-econ majors they violated that assumption and they gave so that was a crack in the wall that got a lot of attention and a lot of laughs it seemed to be that economics were economics was a subject about economists rather than about people in general but it did raise the question are communists different in these games because they have self-selected into the field of economics or that's the blue line or are they different because of their exposure to the culture of economics every field has its ideology is this okay oh thank you for answering yes this can be participatory you can yell out answers so that's great so um so this man in the audience says it's not that you self-select it's that you learn well i i would now like to mention in the bottom of this slide a natural experiment at yale law school which doesn't prove that you don't self-select but it does prove that you do learn in other words it proves that the red element is present and this natural experiment takes advantage of the fact that first year yale law students are randomly assigned to teachers in the required courses and some of these teachers have phds in economics and some of them don't a year after these students participated in the courses a sample of them was asked to play games and lo and behold the students that had been randomly put into classes taught by instructors with phds in economics played significantly different than those who had not the students of the phds in econ were more selfish and more concerned with efficiency and this is after just one year's exposure two courses the question i address here is how does social structure in particular how does your affiliation to a particular part of the social structure shape your social identity and in particular your distributional preferences and i will ask how do they affect cooperation and i will argue that cooperation is very important for development but to address this i can't just play games with different kinds of people because then i would face the same difficulty as was faced in evaluating the question economist free ride does anybody else to evaluate this i have to go someplace where social status is assigned by birth rather than earned obviously people who achieve a great deal who at the top of a meritocratic system are going to have somewhat different characteristics so that wouldn't be a good place to go so i go to rural north india in rural north india there are stark historical differences in the rights between high status and low status caste the low status casts were untouchables they couldn't go into public places they couldn't enter into any but menial occupations they couldn't accumulate wealth and they were relentlessly stigmatized you have your caste membership by birth and although certainly cast outcome like almost every other outcome in society is a political process those casts that are at the top of the caste hierarchy the brahmins for example have been there for hundreds if not thousands of years and the untouchables have pretty much been stuck for hundreds if not thousands of years this died i think did it die no okay um and so this is just a wonderful place to go to avoid the problem of meritocracy and yet because the caste system was efficiently outlawed with india's independence in 1940s in 1947 and there was land reform all over the country there is an overlap in the distribution of wealth income and education between high and low cast so it's not that all high casts are richer than all low casts and in my papers i showed that explicitly and i'll show that here as well so we can compare the effect of cast status holding constant education and wealth and thus we can make an argument that it's the cash status itself that is going to drive very differences very large differences in behavior in three experiments i will describe so this brings me to my paper my paper is about spite and its effect on coordination stigma and its effect on norm enforcement and thus implicitly it's about development it's about the way social structure which affects which reflects differences in power a long time ago shapes who people are and thus their capacity for solving the kinds of problems that come up all the time in any society and are perhaps particularly important in a poor society i go to uttar pradesh which is in north india experiment one is about coordination it's a repeated game in each period a player has an endowment of six monetary units so he's given six of these little tokens and he's allowed a binary choice he can put all six of them so notice this has five columns so he's going to be making his decision by putting tokens each period into the first row so if he decides to cooperate he'll put in all six of his tokens and if his partner makes a different choice remember i said there's a binary choice you can do you can cooperate by investing all in a joint project or you can defect by keeping four tokens in your pocket and contributing just two if the partner who's in another location contributes just two then a messenger will come and show this man show this player that his partner has contributed just to and in that case of miscoordination the poor first player will get a very low return he started with six and his whole payoff is just three tokens and he will see that so we call this the losers payoff he started with six he cooperated he put all six in the joint project but his partner didn't cooperate so he only gets to have two now after each period he learns the decision of the partner and i've explained that if he cooperated he gets a high payoff oh now let me explain so suppose now he's going to be trusting and he's going to say to himself well i'm going to give my partner another chance so again he's going to put all six tokens in now if his partner is nice and cooperates his partner will put also six tokens in the messenger will show that of course these players never come face to face or know each other's names and then he will get a very high payoff i don't want to count them all but he will get 10 so he started with six tokens and his payoff will be 10. so if a player cooperates he gets a high payoff if his partner cooperates but a low payoff if the partner defects now if he defects he always gets the same no matter what that's strange mutual cooperation is an equilibrium if i believe the other person will cooperate of course i want to cooperate and then we get really high payoffs but mutual defection is also this is very strange why did it i don't know it doesn't work anymore it's okay um but mutual mutual defection is also an equilibrium there are six pairings for five periods and it's all anonymous the experiment here's the key point the experiment mimics situations of coordination on drainage for example if i build a drain and you build a drain and the other people on our path build a drain then the path is not money but if i'm the only one that builds a drain i lose my effort and still the path is money timing of of of planting forming a cooperative forming a new convention a new institution all these problems arise every day and the canonical example is from david hume to pick up um a an author that took tour uh discussed the canonical example is rowing a boat if if we both row we go someplace and that's a high payoff for both of us if i row and you don't grow i lose my effort and we don't go anywhere if i don't grow i don't lose my effort we don't go anywhere i don't care whether you grow or not so i'm happy if i don't row if i think you're not going to grow if i don't grow i always get the same i go nowhere but i don't lose any effort so the player sees the full history of the game in the game box and lo and behold most low-cast pairs achieve success and coordination and that's also true with fixed pairings of students in the united states and in barcelona and in many other parts of the world where these games have been played so i have to explain the picture the dark areas are about failure there are two ways to fail one way is i row and you don't or you row and i don't that's the worst another way to fail is that we both don't grow then we don't get what i call the losers payoff but on the other hand we miss a chance to go someplace in our boat we miss the chance for a high payoff we miss the chance to build a new convention a new institution success is the yellow so failure is on the left hand side of each and the proportion failure is a minority in period 5 and it's an even smaller minority in period 10 for the low-caste pairs does anyone else not the high caste i was truly amazed by this result the high cast pairs who are the more educated the more sophisticated and with their hands on power they should be used to solving problems they can't coordinate even after five periods very very few of them are both rowing the boat after 10 periods this is the second fixed pairing they're doing a little better but still a very large fraction more than 50 percent are failing to coordinate and the mixed status pairs are in between how in theory do people coordinate so like a french person i first have to ask how people do things in theory the challenge in this kind of game is to build a reputation for cooperating if you build a challenge for cooperating then your partner will in his own self-interest wish to cooperate too and then you'll go both get the high payoff so let's look at period by period outcomes this is for periods one through five on the left hand side here the low cast remember that yellow represents success and notice the success rate is higher every single period the um the blue area look at the key the blue is where one is rowing and one is not and the red is where neither one is rowing that is to say both are defecting so red is both defecting so you get fewer and fewer cases of both defecting and more and more cases of both cooperating and that's just the way it's supposed to work and that's the way it works with university students in the west not so with a high cast they simply can't get it together there's no progress whatsoever what is wrong with them they do not differ very they do not differ significantly in the probability of cooperating in the initial period in the initial period you're really revealing your type you're revealing your trustworthiness and there's only a slight slightly smaller probability of cooperation by the high than by the low caste but in the later periods they act differently when do they act differently a high-cast person who obtains the loser's payoff in the first period of a pairing is 35 percent less likely than a low-cast person to continue to cooperate in the next period so he cannot establish a reputation for cooperation why is that for a long time i had no idea i thought they were crazy this establishes the result this is i pool all of the results for the second period of a fixed pairing so period one and period six are your first times with a new partner and i find controlling for everything i get a 35 difference in the probability of cooperation if i'm a high caste and nothing else is significant after a history of mutual defection after a history in which i defect and my partner cooperates or when both cooperate there's no difference significantly between a high and low caste it's only after getting the losers payoff that is i cooperated my partner didn't that high cast distinguished themselves so is this a possible response to loss of face suppose you initially earned the loser's payoff so you rode the boat your partner didn't you got a low payoff low in absolute terms and low compared to him do you perceive this as an insult would you be insulted if you were playing this game and you put in all your tokens and your partner didn't well if you did you might want to punish him but you might look at it another way you might perceive this as an honest misunderstanding if he didn't trust that i would cooperate if he didn't know i was such a wonderful person then he might even though he's a wonderful person have defected simply in response to thinking i would defect so i'm going to correct his misunderstanding i'm going to cooperate one interpretation of the outcome we obtained is that the high cast perceive an insult the low caste perceive a misunderstanding so the high cast punish in the second period which means they cannot establish a reputation for cooperation and the low caste perceiving a misunderstanding they continue to cooperate in order to establish trust and we saw that they did ultimately establish trust a year later i decided to test this idea i tested this idea to provide more direct evidence that we're measuring differences in preferences rather than just differences in trust so we go back to the same set of villages although we have different subjects now i'm going to describe experiment 2. experiment 2 is very simple it's a dictator game as tor described but it's even simpler than his game you don't get to pick whether you want to give 10 euros or 9 euros or 8 euros or 0 euros to your partner you just get to pick one way of dividing money or another and it's anonymous and it's double blind so that the only pressures you feel in making this choice come from inside yourself not from being watched this is how the setup looks for the players they will say okay there could be um a hundred rupees this 250s for for your um for yourself and 100 for your partner or there could be 150 and 100 or you know all different kinds of choices so i will give you four examples that we did more which do you want 90 rupees for your partner and 90s for yourself or 70 for your partner and 90 for yourself now 90 is a lot of money it's more than one day's wage it's almost two days wage for an unskilled worker so this is real money i asked my daughter and she said well of course 90 90. who wouldn't want 90 90 why do you want to hurt your partner he's someone like you who's living in your very own village forty percent of the high cast choose the spiteful option they choose the second option 70 for their partner 90 for them in other words they they're not they're not willing 40 are not willing to have the partner get the same another choice you could get a hundred or you could give 110 or if you give up a hundred then your partner will lose 80. so here the second choice means you are paying for the privilege of hurting somebody else okay so clearly a totally self-interested person would say 110 is bigger than 100 i'll go for 110 and who cares about my partner but a spiteful person would say if i get 110 it's true i have more money but my partner has even more than i do now again this is anonymous this is just another person in the village however you know life is really about symbols and meaning as much as it is about anything else and if you're a spiteful person you won't like that so forty percent of the people choose the second option if they're high cast and only 29 if they're low cast now i'm going to show you two last choices and these are about charity not about spite here i'm going to ask whether you're willing to give up 10 in order that your partner would get 50. you could have 130 the chair or you could give up you could give up 10 and have 120 and then your partner would get 50 more and this is exactly the same you could give up 10 160 to 150 and then your partner would have 50 more what's the difference between these two options think of yourself as doing the charitable thing suppose you do you do the charitable thing then in the first case you have less than your partner but in the second case you have the same so a charitable person would choose the same option irrespective would be the same in these two cases but someone who was concerned with relative payoff would make very different choices he might want to be charitable as long as he's not behind but if he's behind then he doesn't so this is the outcome for the high caste they don't want 50 percent does not want to be charitable when the outcome of charity is to put their partner ahead but only 27 doesn't want to be charitable in other words 73 want to be charitable if it's 50 50. so it seems they like the idea of charity they like the idea of giving up only 10 to help someone by 50 but only if that doesn't make the other person a bigger lion than they are but the low cast they don't even see a difference they act just the same so the conclusion so far is that for low-caste individuals social preferences that we estimate from these games are very similar to those estimated for university students in the us and barcelona there's no spite and there's a little charity but high cast individuals have different social preferences they are more spiteful now this game was double blind so i cannot link the behavior of individuals to any characteristic other than their cast but i have from the earlier experiment individual characteristics and i know that even controlling for individual characteristics the high cast played the game very differently i show in a model that i won't show here that spiteful preferences can explain the absence of coordination by the high caste and here's a way to make sense of it in a feudal environment or in a traditional society he who has political power soon acquires wealth as a kind of consequence but in a commercial and industrial society the best way to make money is to make money it's quite possible to do this without acquiring power so here's the point i argue that living as the elite in a traditional feudal society has made the high caste very concerned with relative payoffs and that concern is a barrier to coordination thus opportunities for everyone to become better off are missed because a response to a mistake is taken as an insult the insult is punished the punishment blacks reputation formation without trust you don't get to full coordination here's my last experiment this is an experiment about another kind of cooperation now i'm not concerned about coordinating you know like you drive on the left i'll drive on the left or you drive on the right i'll drive on the right you row the boat i'll go the boat now i'm concerned about norm enforcement two individuals a trustworth and a trustee play a sequential exchange game and then there's a third party who could punish and the three players that interact come from distant villages they never see each other they play the game in their own villages the experiment mimics informal sanctions of opportunists cheaters and free writers informal sanctions play a major role in the restraint of opportunism in markets and in collective action you face this every day when john algon the other this morning was looking at the world values survey he reported results for a question do you think it's ever justifiable to cheat on government benefits if you think it's not justifiable you will insult or criticize or reproach your friends who cheat you might not want your children to play with children of cheaters you might not want to marry a cheater these informal sanctions can be very powerful because people face these things every day and we are going to ask is the preference for unselfish enforcement of norms the same for the low and high caste treatment one the trustworth and the third party punisher are from the same specific cast so here's the idea there are three people who play this game person one has a hundred monetary units he decides whether or not to trust them to person two then the money is tripled and person two has a binary choice he can give half the money back to person one or he can keep all of it for himself then person three can punish person two or not but punishment is costly to person three so the idea is to elicit the preferences towards punishment we don't really care about persons one and two we care about the willingness to punish and we have two treatments in one case the punisher is the same descent group the same endogamous caste as the victim and this is the result that we get we get that there's much more punishment by high caste people than by low-caste people so it's just like you are um uh going to the market and giving someone money to buy his cow and then he brings in a cow that turns out to be sick or lame and then you ask for your money back and he says no and then you complain and if you're high cast the high cast people will stand behind you and if you're low cast they'll stand behind you much less well you may say it costs money to punish high cash people have more money so it's just an income effect if you live in a mud house you're very poor and among all the people who live in the mud house the high cast punished more than the low cast among all the people who live in better houses the high cast punish more than the low caste if you own if you own low land you're poor we look at just those people who own low land the high cast punish more than the low caste we look at just those people who own above median land the high caste punish more than the low caste we can predict income we can predict consumption differences based on the education and wealth information we have and we find no relationship whatsoever between wealth extra you know above average or below average well this is above average or below average consumption and and punishment there's no relationship at all so the hypothesis that we that we come up with is that well perhaps the high cast have greater in group affiliation a lot of literature shows that people favor their own community i learned today at lunch that there's a recent scandal in trento where politicians were favoring trento businessmen for some contract to build roads i hope i haven't gotten this wrong that's an example of sort of favoring the in-group now in order to dominate a group as thoroughly as the low castes were dominated you need to pulverize it you need to atomize it you need to stigmatize it and all these things the exclusions from public health from public celebration the bans on marriage ceremonies make sense because they hindered the low cash from developing a positive and a cohesive group identity so i'm going to report one last result treatment two now the only difference is person one who's trusting and person three who can punish are not from the same cast so it's like talkers and brahmanas you know they're both high caste status but they're not the same descent group and now what i find is that this is the new result this you saw before this you've seen before this is the new result the gap vanishes almost there's no difference between high cast and low cast the high cast hardly punish more than the low cast so it seems that the low cast whether they are from the same group as the victim whether they're from the same group as the victim or not from the same group as the victim they don't care whereas the high caste they care a lot so there are two interpretations that aren't really distinct one is strictly an issue of in group affiliation the high caste people have had the opportunity to develop in-group loyalty the low-caste people have not a second interpretation is that the low-caste people do not even have the capability to punish those who violate a cooperation norm they do not have the sense of arrogating to themselves the power to judge and to punish when they themselves have not been able to self-organize they themselves have not been able to freely form their own culture they have always been until independence and really until quite recently under the thumb of the high caste so culture the glue of society the thing that makes norm so powerful that requires space and freedom it cannot be taken for granted so the possible consequence is the difficulty in enforcing contracts and sustaining collective action a perpetuation of a social hierarchy even if all other obstacles to the low caste are removed in summary the contributions of this work are that it identifies new influences on distributional preferences whether you are at the top or the bottom of the caste system affects your spite whether you are at the top of the bottom of the caste system affects your capability to enforce a cooperation norm and the second contribution is to suggest new reasons why distributional preferences matter usually we say they matter when it comes to voting on tax policy or benefits but here we see they matter for something much more basic they matter for the ability to solve a simple problem of whether you will row the boat and i will row the boat they matter for a simple problem of whether you will punish cheaters so in conclusion we argued that concern of high cash with loss of face and gender spite social exclusion of low caste hurts in group affiliation and capability to punish and this is my final slide this is important because if there's capability to punish those who cheat there's capability to solve a collective action problem if you can solve a collective action problem it doesn't matter whether each one of you is a little fish you can all gang up on the big fish but if you cannot solve a problem of collective action the little fish will remain the little fish and not just because they're little have less wealth less political connection less access to violence but for a fourth reason their social preferences are different they're pulverized they're less able to solve a collective action problem thank you very much no please okay all right my co-authors so that slide has disappeared but uh this work is um all these experiments of course i don't do them all by myself i do them with really wonderful people and um those people's names will be appearing so ernst fair has been a big uh a big source of inspiration mayor ashtomata is the most energetic indian i have ever met and priyanka pandy is the most powerful administrator i have ever worked with um and here are some of our papers okay so we're very thankful professor hoff for this very clear presentation you provided with a lot of food for thought i'm sure that there are a lot of questions that people may ask you so we have some time to take questions from the floor thank you very much manufacturing let's wait see um intervention reflecting your presentation mirrors a study that was published 2008 by novak nowak on nature on non-cooperative games as applied to a group of students of the mit for the cooperation a possibility was observed between very smart youths who would prefer to adopt a common line non punishing but rather deserting or defecting and the similar reaction was uh registered also in nature amongst bats or vampires who at the end of the day would throw up a part of the blood they had a drew note to to give it to the batter which hadn't taken any blood so that they would have in or or the vampire rather so that was an implicit cooperation in nature again amongst bats and vampires my question is as follows in a number of western civilizations um do you think that educators can be trained and educated to train uh the um youngest to uh have cooperative uh behaviors i'm referring to the less advantaged areas in europe where these youngest have no education to cooperation whatsoever whereas we see that in other areas such as entrentino area there is a strong capability to develop cooperation in this area where we are here at the moment and then we'll give an answer what about the next question would you like to get it in english yes you you what do you think about a different phenomenon which could be called the nobles oblige which is the apparently the exactly the the country you ask two strong players to decide about the exclusion of a weak player they could allow the weak player to participate in a for example in a division of sum of money if they share some code or sort social norm of fairness within the within the this category which apparently is what happened when in a western society people that share the idea of fairness decide to for example to punish this is a second aspect but to punish another player that for example act in unfair ways with the third party for example social responsible investment typically is this kind of behavior we share some our rule of behavior and people in the western society a small part of them punish funds that or companies that do not behave according to the same kind of uh social norms with third-party weak player for example uh people in developing countries this is a a different kind of this in a sense this could be the the higher level caste because this is the cast but which richer people that apparently want to implement social norms also by punching members that behave in an unfair way with the weaker participant or weaker player is this working okay so the first question um to repeat it the question was can you train people to cooperate this of course is a trillion dollar question i myself would like to work in this area before i give a completely speculative answer let me describe an experiment i learned about from pontus strimling and his co-author who who was visiting princeton this last year it was a very interesting experiment there were two stages to the game in stage one individuals having the opportunity to contribute to a public good there's no punishment they're generously they're not whatever they want and then in the second stage the people who were generous are put in one set of groups and the people who were not generous are put in another set of groups they don't know this of course and now they're given the opportunity to build up a set of rules eleanor ostrom style they get to choose three things the mandatory contribution how much to punish someone who doesn't make the mandatory contribution and how much to reward a person who monitors and there's only one efficient outcome as in any public good game the efficient outcome is always you contribute everything because it gets doubled or tripled and then it's shared among everybody so individually you want a free ride but collectively your interest is in 100 contribution because it's like a money machine and the remarkable result was that those kids these are university students who were generous come up with the efficient set of rules and those who were stingy do not so let's go back to the question can you train people to cooperate here we have people whose self-interest lies in cooperation on rules that will make sure nobody cheats on his taxes nobody cheats on paying the bus ticket nobody cheats on the train ticket so that in fact there's a bus system a train system and public parks for everyone they have this opportunity to cooperate and those who didn't do it of their own free will in the first game don't seem to be able to settle on an efficient set of rules in the second game so one interpretation is that people don't want to choose rules that make themselves look bad or another interpretation is people just don't like rules so if you would have done it anyway you're happy to have those rules but if you wouldn't have done it anyway you don't want them so can you train people to cooperate not easily i think that certainly you can train children to cooperate but will that last to adulthood i think what we see is that people are conditional cooperators that means if they're in an environment where others do not cooperate they will simply stop cooperating you can train them from here to high heaven but i think that it's in the nature of people and certainly in my games people were the trustee person one was very was very nice and person two often was very nice but you know that over time these voluntary games have a universal result voluntary cooperation unravels the reason it unravels is because people only want to cooperate if they see others if they see any cheaters they you know any cheaters could ruin can ruin things plus people often want to cooperate but a little less than the other guys if everyone's cooperating a little less in period t compared to what they saw others doing in t minus one you wait long enough and people are contributing nothing so i think that we're hardwired to be very sensitive to how much others are cooperating and that means that we have a system-wide problem and just training a few people is not going to do the job i think that it's going to require changing a whole system that's why i think the example of the nordic countries that were once corrupt and full of labor conflict and today have achieved such successful institutions is is very much worth studying it points to the idea that institutions shape how cooperative people are which shapes the institutions they choose which shape our cooperative are and so to go back to my first slide this business of how you think and how the economy works are really not different problems they are interrelated this is such an interesting topic i could go on forever but i guess i should go on to the second question no bless oblige this question was a little um diffuse and so i'll just pick out certain elements that i i think i'm capable of responding to the the the question was about fairness and social responsibility behavior so now i'm going to make an infomercial and it's not going to be for my own work it's going to be for a wonderful book that comes out this year by the economic historian joel mokhir called the enlightened economy and it is about the following question why did england lead the industrial revolution why not france why not italy the argument in this book is that england more than other countries had developed an ideal of gentlemanly behavior which was not about drinking fine or riding fine horses but about being fair to business partners can you imagine and if you wanted social esteem in the higher echelons of society you had to look like a gentleman and you had to belong to clubs for gentlemen and these clubs monitored you and gossiped about you and they might even throw you out if they found you guilty of ungentlemanly behavior so even though at this stage of the game england did not have the formal institutions to enforce complex contracts much less the kind of globalizing economy that england was becoming these social institutions and this social ideal of gentlemanly behavior this idea of fairness and the social esteem to be gained by being fair i mean think of this you get esteem by never cheating and by contributing to the hospital into the canal and there were all kinds of voluntary groups that created public goods including regulations of professions there are all kinds of new professions coming about and they all self-regulated and they sort of competed by who could be a better person now we like to talk about the market as being this magical device that aligns private interest private material interests with social interests but we know that it only does that if there's a price for every single action which there never is these social norms caused people to enforce and to internalize the alignment of private actions with the social good and the modern world was born so it's a really really exciting idea i don't know if it's true because i'm not an economic historian but it's a great weed if i can i would like to make a question myself i would like to make a very general question so i hope that is not too too naive and as as usual with economics i would like to go back to adam smith and as a result of your studies how would you or where would you draw the line between the idea that basically the the most parsimonious way to think about society and institutions for good or society starting from non-cooperative selfish behavior or the other way around and think that uh selfish and non-cooperative behaved competition uh is an exception that could be curbed and made compatible with the good of society which is in a sense the uh the uh the point that some historians of thought make that others mates has been overthrown sorry turned upside down it started from the idea that the general attitude of people is towards cooperation and being cooperative and moral sentiments etc and but then economists have pushed the idea that competition and selfishness are first of all dominant and second are not harmful so this is a very general question but i think it could be interesting to to try to to draw some conclusion of if it's possible from your studies on these very general statements about the deep tendencies in society and how to to to model it my reading of adam smith is that he thought the fundamental factor driving behavior was the desire for social esteem and he viewed the west as an exceptional society in aligning social esteem with consumerism and wealth acquisition and he spoke a great deal about slave societies where one got self-esteem one got socialist deep by acquiring more and more slaves and he thought those societies would never go anywhere because obviously more and more people would just get enslaved that part of adam smith has been pretty much forgotten the part about social esteem has been has been now recognized and and is widely discussed but what is often not discussed are what the social is what the sources of social esteem are it's taken for granted that certain things will be estimable and certain things not but there's no reason to take that for granted that is an outcome of a social process ah charles darwin loved to talk about the peacock's tale and for a long time darren was really troubled by the peacock's tail do you know if it rains and the tail gets too wet the peacock can die because the tail is so heavy when it's wet he can't move and he starts to death the life of a peacock is not easy similarly the life of a person who's high caste is not easy there's many things he can't do can't even cut his own toenails it's viewed as polluting think how inconvenient that would be and he has to constantly worry about not losing face and that gives rise to these weird behaviors that we found in our game so i think there's no question but that if we want to understand deep forces explaining differences in society we simply have no choice but to think about what people think are the sources of social esteem and how those are shaped by their society an optimistic result of the sociological work of a young indian recent phd at harvard named paramita sanya is that when you give women exposure to micro-finance and for the first time they get out of their houses they have some control over money they meet other women and they get a sense of personal efficacy that that gets them to be more involved in village life and more active in playing the role of social enforcer playing the kinds of roles that joel mokir talked about that the optimistic thing i think about these findings and experiments is that we are all potentially pro-social and we all do care very much about how we fit into society but the difficult thing is shaping a whole population in a society i can take out any individual and adopt him and make him be very pro-social and cooperative if i take him to the right place but if i have to deal with a billion people all interacting that's the challenge and i think some of the most exciting work um going on today is work that deals simultaneously with the determinants of structure and the determinants of norms solving them together if there are no other questions i believe that we can conclude our meeting here thank you ever so much to professor hoff so thank you very much