Sustainability Videos & Lecture Series

Charles Redman:

I'm Chuck Redman, Director of the School of Sustainability, and happy to welcome you to a Wrigley Lecture. This is a series that we usually co-sponsor with other organizations on campus or off campus to bring people who are of interest to as wide a group as we can of people who want to explore what sustainability is and how it's being done. And often, this takes us to sister disciplines or schools or colleges.

And in this case, we have the honor to have someone who is important both in the academic world, but also in the public and private world. And we're co-sponsoring this event with Arizona Public Services service and with the WP Carey School of Business as well as the Global Institute of Sustainability. So I just wanted to make that pitch for collaboration among units in bringing interesting people who can speak across those units.

And we're extraordinary fortunate. It's maybe the event of the millennium for us to have Sir Partha Dasgupta here to speak with us today. Partha is well-known for a whole variety of things. But we'll call him a resource or development economist or an economist of a whole variety of other persuasions.

He has a long list of important publications, a number of which I've assigned and I know Ann Kinzig has just assigned. And many of others of you have assigned it often. Because these are important works.

But the list of things that I was going to highlight are his memberships, which are sort of a recognition of the importance of his ideas and opinions in both an academic and in a public sector. And so they go from areas that are academic, like the Econometric Society, to being a Fellow of the British Academy and a Fellow of the Royal Society, a member of the Pontifical Academy of Social Sciences. That takes us out of England I think.

And then we're a Fellow of the Third World Academy of Sciences. Maybe that takes us at least to Wales-- a member of the Royal Swedish Academy, a Foreign Member of the US National Academy of American Philosophical Society and, most important of all, knighted by Queen Elizabeth in 2002. So we're all excited and having heard you already and spent some time, I know this will be important and stimulating on social capital and natural resources. Thank you.

Sir Partha Dasgupta:

Well, thank you very much. Can you hear me?

[APPLAUSE]

Well, thank you very much. It's been a terrific two days. I've enjoyed visiting you. You live in an enormously attractive terrain. And I have friends here anyway from the past. So it's been quite wonderful. And the theme of my talk today, it's very tentative.

Much of what I say I sort of half believe. And the half that I don't believe is due to doubts, not because I disbelieve it. You'll see what I mean. Because in a way, it's not settled in my mind. And even although the subject has been discussed at great length by very eminent thinkers, social thinkers of some of the most eminent ones of today, I think they claim more understanding than is warranted by either their writings or by the evidence available. So I can see that there are a number of students here.

I can guess from the generation. I mean, at my level, I see students very easily at my age as it were. So this is a field which is open to entry, the notion of social capital, which has been much discussed in recent years. And I've been thinking about it for some time.

And along the way, I'm going to give you an indication of why I got interested in the concept. We economists don't spend much time analyzing concepts. We are, by training, much more opportunistic.

If something helps, if something works, then we pursue it. And if it's found to be not very productive, then we dump it. And I can give you instances of both of these practices.

But what we don't do is to think about concepts and classify them and define them and so forth. Because in some sense, our training, at least my training, has been to have a very loose notion, and then work with it and see, and sharpen it as we make use of the concept. So I won't spend an awful lot of time.

In fact, I will argue that thinking about social capital is probably a bad idea. It's better to ask a different question to which the notion of social capital may be very pertinent. And I will want to arrive at it that way rather than actually talking about it and saying why it's so important.

Obviously, I would not be speaking here on this topic if I didn't think it was an important matter. But I don't think thinking about definitions is a very great help. And I'll try and persuade you of that.

Now, having said that, I won't really know an awful lot about what we really ought to know. And I think that's also a mark of increased understanding, to be able to say I now know why I don't know this. Because this itself, the fact that I can recognize this, quote-unquote, suggests that I know a little bit about the thing.

OK. The idea of social capital sits awkwardly in contemporary economic thinking. Even though it has a powerful intuitive appeal, it is proven hard to track as an economic good. Among other things, it is fiendishly difficult to measure.

This isn't because of a recognized paucity of data, but because we don't quite know what we should be measuring. Comprising different types of relationships and engagements, the components of social capital are many and varied and, in many instances, intangible. So I put up there two working definitions, one from Robert Putnam in his famous 1993 book on Italy, where he talks of social capital as "features of social organization such as trust, norms, and networks that can improve the efficiency of society by facilitating coordinated action," unquote.

He has another definition which comes along in 2000 after a few of us talked him out of the first definition. But here it is. So let me quote that. "Social capital refers to connections among individuals, social networks and the norms of reciprocity and trustworthiness that arise from them," quote, unquote. That's from Bowling Alone, famous book written in 2000.

Series of exchanges between Sir Partha Dasgupta, Chuck Redman, and others trying to figure out microphone:

I'm afraid you're going to need to use this.

I'll need to use this? Is that right? OK.

[INAUDIBLE]

Sorry?

We have a wireless one if he wants to walk around. Yeah, that would be better. I thought that was the wireless. What I've got is-- [INAUDIBLE] because people are having trouble hearing you. OK. [INAUDIBLE] Well, I can stay here. That's all right. There's nothing on the wireless that's on. Sorry? That's not on. No, this is for the video. It was. It's on-- For the video. Yeah. This is on if you wanted to [INAUDIBLE]. OK. [INTERPOSING VOICES] --for the wireless. [INAUDIBLE] But it's working for the video. [INAUDIBLE] Chuck says wireless. There's no wireless mic for the room? No, he has one [INAUDIBLE]. [INAUDIBLE] So what should I do now? Can we use both of these, so we get the video recording as well as-- Yes, of course. OK.

Sir Partha Dasgupta:

Absolutely. So I think you lost me at the time I was quoting from Robert Putnam's second definition, 2000 Bowling Alone, the collapse and revival of American community. "Social capital refers to connections among individuals, social networks and the norms of reciprocity and trustworthiness that arise from them."

Francis Fukuyama 1999, The Great Disruption has the following definition. "Social capital can be defined simply as an instant [INAUDIBLE] set of informed values or norms shared among members of a group that permits them to cooperate with one another. If members of the group come to expect that others will behave reliably and honestly, then they will come to trust one another. Trust acts like a lubricant that makes any group or organization run more efficient."

Finally, since we are in the business of giving you definitions or, at least, I am at this point, a number four, Samuel Bowles and Herbert Gintis 2002 in The Economic Journal-- and I've put that up. "Social capital generally refers to trust, concern for one's associations, and willingness to live by the norms of one's community and to punish those who do not," unquote.

Now, the problem, as I said, I will not go to spend time defining the object. But my point in giving these definitions to you, offering them to you for consideration, is to show how dangerous it can be to start from a warmed glow feeling towards an object which you haven't defined, and then you pack into it everything you like or every sensible person likes. So the common weakness of all these definitions is that they encourage us to amalgamate incommensurable objects, non-comparable objects.

For example, in the first of Putnam's definitions, beliefs, behavioral rules, and such forms of capital assets as interpersonal networks are conflated. And it's just as well to check that that is correct. You have, "such as trust--" that's beliefs.

I trust you, because I believe that you will do, e.g., what you said he will do. That's a belief. Norms are, of course, behavioral rules. Do this if that happens. Do that if she does this, and so forth-- conditional actions.

And networks, of course, that's a very different object altogether. And you have conflated them all. And you are wondering what to do with that object.

Some would be what we would call endogenous. Whether you trust somebody you or not is an outcome, perhaps, of relationships in the past, past behavior, inferring, placing oneself in the other person's shoes and see whether it's in his interest to carry out what he said he would do, and so forth and so on. And norms are, of course, behavioral rules.

In the game theory language, they're strategies. And they get to be formed. They have to be chosen.

They are to be explained why they occur. Just as economists like to try to explain the formation of prices, likewise one wants to explain the birth of trust or the birth of norms or the rules they play for coordinated action. So putting them all together, perhaps, isn't very helpful.

Let me kick off by saying what I think would be a lean definition. And it would be to think of social capital, whatever that object is, as a system of interpersonal networks. In other words, I'll take the third bit of the Putnam definition and see whether it's worth starting from there to make sense of the social world that we observe.

And therefore, trust and social norms typically arise once networks are formed. They may not even arise by the way The fact that people know each other doesn't mean that they trust one another.

There are case studies on that as well. But all of that is to be explained, rather than be part of the thing. We economists-- and I keep on using the expression "we economists," because I wouldn't want you to think that I'm approaching this problem area in any other guise than as an economist.

We tend to keep the object of interest neutral, and then ask whether it's doing a good thing or a bad thing. Steal, is that good? Well, yes, if it's doing something to improve well-being. Is it bad? Yes, if it's used for killing one another and so forth.

So we don't want to necessarily assume that the object is good. And in the context of social capital, that vague notion that we have at the back of our mind, there is a good reason for believing that it has some extremely nasty properties. And I will work to share that with you as well. And, well, I'll do that when I try and give you an explanation as to why I got interested in this area of discourse.

Now, Putnam's original motivation certainly in his 1993 book, was as follows. Most of you will probably know it. But let me just remind you. Certainly, there may be the odd person who hasn't heard about Putnam's motivation in the 1993 book.

He's a Republican. I mean, not in terms of party politics. He's a Democrat, I think. But he's a Republican in the old fashioned sense. He's a Democrat.

[LAUGHTER]

That now goes to show how words can change their meaning. I'm using it in the 18th century sense of the term. And he believes the importance of citizenship and citizens engagements, too.

And what he tried to do was to show two things. One, that network formation and coordinated action among people who know one another can help improve the performance of government, in this case. It's a very 18th century disciplining government. There are civil servants. They need to be told what to do. Because, otherwise, they should be kicked out.

And you need for that, instead of the marketplace, there is a political action space. And there, the idea of social capital as interpersonal networks played a role. And he wanted to show that states in Italy which had greater social capital in the sense of having a denser network of engagements-- and I'll come back to that later-- were performing, had been performing, better economically in terms of, let's say, GDP growth rate.

And he also wanted to show that that came about, because civil servants were behaving better in those states where people were more engaged in associational activities. Now, the empirical route he followed I thought was actually genius. Instead of trying to measure it, he simply said, well, I'm going to look at memberships of choral societies. You know, Italians like to sing. And they like football-- so football clubs and so forth.

So he had some crude measures of association, and then tried to create an index and then run regressions across 20 states. And I thought it was actually an ingenious book. He proved what he set out to prove to the extent that one is able to prove anything in this kind of feeble activity.

I, myself, got interested in a different form before reading his, which was to try and understand non-market activities in rural communities in Africa and South Asia. I come from South Asia myself. And I'm not, by any stretch of the imagination, a rural person. I'm thoroughly urbanized.

But rural communities in South Asia, India, for example or Pakistan, Bangladesh, and even more so perhaps in Africa where roads are scarce or even if they are there, they're extremely unfriendly for transportation purposes as it where-- they're not meant for transportation. They're meant for who knows what. So rural people live on the local, natural resource base.

That I knew pretty early on in my research work. And that is to say the direct utilization of the natural resource base, local natural resource base, by which I mean woodlands, ponds, rivulets, threshing grounds, and then also public goods like watershed management practices, tanks, village tanks, which are designed with cooperative action to collect water and so forth. All sorts of cooperative ventures are entered into. And they're not done through the market.

The market is also very cooperative, by the way. We economists try and tell the story of market as a competitive story. But, of course, the ingeniousness of the construct is that that competition is actually simulating a cooperation to bring about an efficient allocation of resources, for example. OK?

But this was happening differently in the context of poor countries, at least rural parts of poor countries. And I thought there must be a strong connection between non-market resource allocation mechanisms and management of the local natural resource base. And that was an interesting thing to think about, because we knew from economic theory that these environmental resources are particularly vulnerable to market failure.

So maybe there are non-market institutions there to guide the allocation of those goods. I'm giving it to in a sort of textbook fashion, because that's how my mind was working at the time I started thinking about it, which was about 1975. And then I started modeling game theoretically these common property resources.

Because I realized that in principle it was actually a reaction to a famous article by Hardin, Garrett Hardin, which was a very good article, because he really had in mind open-access resources. But he used an unfortunate example to illustrate his point.

He had in mind the atmosphere and the open seas. And he wanted to argue that a property which belongs to everybody belongs to nobody. Because people just enter, and it will be depleted. Freedom of the commons brings ruin to all.

But unfortunately, instead of using the right examples like atmosphere for sick carbon emissions as we now recognize to be the case or the open seas, he talked about grazing land. Now, the one thing I knew about grazing lands, at least in South Asia from the little literature I'd read, was that they're not open access. They have communities or groups.

Villages have historical rights to the land on which they're grazing. That's about all I knew. And I then started constructing a theory which I'll be talking about now in a few minutes and then published the theoretical results in a book.

But I had no examples of what I was writing about. That came later. I then realized later, there was a literature on common property resource. And now, of course, it's a gigantic literature.

There are examples all over the place of the non-market mechanisms or communitarian arrangements for managing common property resources. But you should realize these are local, geographically contained, common property resources. And that makes sense, because then people can be excluded. And the ownership-- they're not private property. They're not state property.

They are community's property without necessarily a deed saying it's this community's. It's historically evolved. And if an outsider comes to take some resources off, his leg gets broken and so forth. There's retaliation there.

All right, so that was my interest. And the issue is, if it is the case that they are being managed reasonably well or there are examples of common property resources which are being managed reasonably well, what kind of institutions are leading to that? And the game theoretic model I constructed was to say, well, suppose they cooperate. Then what would they do? What kind of rationing mechanism, what kind of finds would they institute for sanctioning behavior and so forth?

OK. So that's the background. And in doing that, I realized that the right question to ask when thinking about social capital is to not talk about social capital at all, but to ask a very different question, a much more general question in economics or in any discourse. And I'm going to do that with you first.

And then we'll arrive at the notion of social capital or interpersonal networks and the value of studying them as a derived enterprise as opposed to picking that off as the object of inquiry. OK. I'm having some trouble getting all these papers organized, but here we are. So here's the question.

Imagine that a group of people have discovered a mutually advantageous course of actions. So there are n people. And I'm going to give you some examples.

The examples are useful, because it will give you a hint that the problem that I'm going to be posing is universal. It's not restricted to poor countries. It's not restricted to common property resources. It's not restricted to any co-operative venture.

So here's an example, the most monumental one. Citizens see the benefits of adopting a constitution for their country-- sort of classic problem in political science. Here are the others. The undertaking could be to share the costs and benefits of maintaining a communal resource-- irrigation system, grazing field, coastal fishery. There's a huge literature now on it. But I'm just giving you a bunch.

Construct a jointly usable asset-- drainage channel in a watershed, collaborate in political activity, civic engagement lobbying, which is the Putnam example. Do business when the purchase and delivery of goods can be synchronized. Credit, insurance, wage labor-- there's no synchronization.

If the wage is paid after the task is performed, then there's a synchronization issue or lack of it. Credit by definition, it's not synchronized. You borrow, and then you pay back and so forth. Insurance-- same thing.

A third-- then you have entering marriage, a mutually advantageous collective action as it were. Create a rotating saving and credit association. In Ethiopia, you have these where you buy some invisible object, but you don't have the wherewithal for it.

And everybody needs it. It's a private good, a bullock, a tractor or whatever. Then you put money into a pot. And when the amount reaches the point where one bullock can be purchased, by lottery one of the members purchases it.

But the deal is that the person who has purchased it has to continue paying into it, OK, and then the next man, and the next man. Of course, it saves hugely on resources. In other words, you don't have idle cash lying around not being spent.

Initiate reciprocal arrangement. I help you now that you're in need with the understanding that you will help me when I'm in need. Create a partnership to produce goods for the market. And then there are also instantaneous, where the synchronization problem is resolved, purchase something across the counter.

Then of course, there are these tacit collective action issues. Tacit agreements-- being civil to one another. They range from such forms of civic behavior as not disfiguring public spaces and obeying the law more generally to respecting the rights of others. These are sort of implicit contracts, if you like, that we have with one another.

Now, there are two problems that immediately arise. I'll state first, and then pass over it, because I won't have time to discuss it. But I'll talk about the second one, which I think is where most of the action in social capital-- the issues that in some sense trigger our interest in some amorphous notion of social capital make its presence felt, which is imagine next that the parties have agreed to share the benefits and costs in a certain way.

So I'm assuming the first problem, the problem of bargaining. OK? We see mutually advantageous courses of action. But one course of action make me much better off than it will make you better off.

You both are enjoying the benefits, but I get a disproportionate share. Another course of action would give you the disproportionate share of the benefits, me the less. Now, question is, how do we resolve that tension?

There's a rich theory of bargaining. I'm going to ignore that for the moment. I'll just suppose that the agreement has been reached.

The instrumental problem that arises is in the next one. And that's the one which I think is what lies at the heart of the whole enterprise. But even if it didn't lie at the heart of the enterprise, it's a darned good question to think about.

And that's the spirit in which I want to give the rest of the lecture. Under what circumstances would the parties who have reached agreement trust one another to keep their word? Remember, the examples of the non-synchronization is the key one.

Credit-- I lend you money with the promise that you'll pay me back at some agreed upon rate. If I don't trust you to pay me back, then I won't lend you. Now, you are deeply honest. And you want to assure me that you will pay me back. The problem is, how do I know if you are not bluffing? And if that happens, then, of course, potentially a mutually beneficial course of action will not be carried out. So the parties have a genuine interest in setting up institutions, so context in which they can trust one another to keep their word.

And the question is, what might the context be? And what I will do is I thought there might be four basic broad types of contexts in which people could trust one another to keep their word. And those contexts would define three different types of institutions. And you will see the sense in which out of four, I get only three.

First context would be when the parties care about one another, mutual affection. Why might that be? That would be the case. I think the household is a very good example of an institution where the agreements-- very often they're not agreements, they're habit.

Who does the washing up? Who does the cooking, and so forth, which day of the week? You don't even write down a contract, right?

I mean, you sort of do it. And the reason you trust one another-- at least at the first cut, a great deal of the work that is done in answering the question-- is the fact that people care about one another, the people involved. And so it happens, works with, at least in my experience with children, certainly with one's spouse where it can be completely sanguine that the agreement that has been reached in the morning as to who picks up whom at what time and so forth are carried out.

Of course, there will occasions when it won't be carried out, but that's because something went wrong. The car stalled or something. But there will be an explanation for it.

I won't have very much to say about this case. It's an interesting case, obviously. But in some sense, in some deep sense, it's restricted to a very small institutional unit.

So let's say the household may be the typical one. Of course, it will be also true amongst friends, people who care about one another. So the fact that they care about one another, that parties care about one-- that has to be common knowledge, by the way.

A must believe that B knows that A cares about B and that she knows that he knows that she knows that he cares about her and so forth. That has to be the case. So it has to be something like common knowledge. And then, of course, the system works. That's easy enough.

The second which is now very much in the forefront of economics research under the guise of behavioral economics sees people as being fundamentally good chaps, decent fellows, not cut throats. That is to say, through the selection process, pressure of selection, we are disposed to be prosocial. Now here, of course, a great deal of the great world's literature is founded on that.

That can cause, by the way, enormous great tragedies. When I read the Greek classical plays, there is the prosocial disposition which is the cause of so much of the tragedies that befall the heroes of the classical plays, which is if I've given you my word, period, I'll do it. Now, there I do it, not because I'm fond of you, but because I have to shave in the morning tomorrow.

And I look at myself. And I say, what kind of person am I that broke his word. So it's something like an integrity which is at the heart of it.

So in some cultures when you say, well, they shook hands — that act commits a person to doing what he said he would do. Now, the interpretation, the pro-social disposition, is that he knows that that action of shaking hands commits himself never mind whether the other person believes it or not. So far as he's concerned, if he's shaken hands, given his word, that's it, period. OK?

A great deal of behavioral economics is built around that nowadays, these games in which you play the prisoner's dilemma. And it turns out that the way the game is played is not in accordance with this tradition or classical interpretation or expectations. All right, the problem with it, it's not my place to question it or anything.

That we have a propensity to have something like a self-esteem goes without question. The problem that it raises is not that it's not true. We probably do all have pro-social disposition. But the problem is that's not public knowledge or the extent of the prosocial disposition of a person is not public knowledge.

I don't have a sign saying that the point at which Dasgupta cannot be trusted to keep his word is $5,000. If that were known, then this could work. Because you keep the opportunity cost below 5,000, everything's fine.

But who knows? We all look alike. We all are pro-socially disposed. But we also know that there is a statistical distribution of types.

There'll be extreme cases like Gandhi, possibly Mandela, for whom the prosocial disposition is enormous. If he says he will do something, then he'll do it. And we all know that, or we think we all know that.

But what about the rest of us? We all probably have a price for most transactions. But the problem is we don't know what the price is.

So society can't really rely on prosocial disposition as a way out. It maybe can give a kick start to the procedures that I'll be now talking about. But it will not be the end of the story. It might be the beginning of the process. And saying, yes, we might be able to create an institution where things will work. Because once we think through it, set up the penalties and rewards, then the trigger mechanism for kicking that through, kicking that entire process, may be the prosocial disposition that we all experience.

So now I come to the third and fourth. And they are the ones which will-- I'm going to give it in sort of slightly stylized-- obviously, these are going to be overlapping-- structures. No question about it.

But it helps to think about these things as very distinct, these categories. I've started by talking about laws and norms, because I want to think about three and four, two other mechanisms of context in which the agreement the people might feel disposed to believe that the agreement will be kept. We know now how we are to share the benefits and burdens of the agreement.

One is external enforcement. And I'm going to put it in a crude variety for dramatic purposes and say the rule of law operates through this externally enforcement. External enforcement means that there is another body-- e.g., the state-- who is entrusted with the task of ensuring that agreements are kept.

So violation of the agreement by the members in question will be, if it's detectable, if they're observable-- sorry, if they are verifiable. In the case that I'm thinking about, say, in the legal sense, then if they're verifiable in the court of law, then parties know the consequences of violating the agreement. And the agreement, of course, in this case, in the extreme case that I'm looking at will be in the form of a contract, written contract, and courts of law which will adjudicate misbehavior or breach of contract.

That's the one which we economists write about mostly. When we talk about, let us say, in the first year textbooks that a household does not break its budget constraint, then there is an implicit assumption that the household is not stealing goods from his neighbor. Because the budget constraint gives a well-defined notion of property rights of that person.

So we're not looking at a society in which an awful lot of thievery is going on. OK? And so if the economist is asked, well, how come you're making that assumption? You say, oh, well, at the back of my mind, of course, I have a government which will penalize people for doing anything like that.

And the penalty will be sufficiently large that it will not be in the interest of anybody to violate it. So that's an easy one. But it's a problem.

The problem is, and the reason why, I'm going to now argue that four is the real machinery that we all have, all societies have, for resolving this problem. The reason three is not an answer, is not a deep answer, is that the state is also composed of people, state agents, civil servants, if you like, the police, the law courts, and so forth.

Now, if they're agents, then the right thing to do is to model them in as well as part of the larger game which you're trying to understand. Because then you have to ask if there are some costs of enforcing the contract, why should these people trust those characters to carry out their duty, the higher order contract that they are to adjudicate over disputes, over malfeasance, and so forth? And if one looks, obviously, of the world around us, we see plenty of evidence of where three is not actually working.

Mobutu's Zaire was a classic case of a state which didn't really exist in the sense that we think of the state existing. So we need an explanation for how the external enforcement can be made to work. What context could we bring to bear, could we think of, where the external enforcer could be trusted?

In other words, the parties we have had the first agreement can now go to the external enforcer and say, here's the agreement. A is to do this. B is to do that. C is to do this, and so forth and so on. Now, you now know what to do.

Now, presumably, everybody has now to have trust, this external enforcer. And if they do, then and then only will they be able to trust one another. So in some sense, we have bypassed the problem by shifting it to some other object.

And we need to have an explanation of how they, the external enforcer, can be disciplined, if you like, which is, of course, where Putnam came in. Because he was interested in showing that civil servants behave better in places where citizens are active rather than passive. OK? So he didn't have the theorem that we have. But game theory itself-- a perfectly good theorem which gives you some lead to resolving the problem of the external enforcer.

But it has problems. Everything has a problem in the social sciences. So I'll first sketch for you the way the four works, mutual enforcement works and how that will explain how external enforcement can be made to work. Because it's the same logic.

I will then show you that the power of this theory is that it also explains conditions under which it won't work, context it won't work. And then I will want to argue that even in context where it can work, it may not work. And that's the really hard one.

And that's where the social scientists, certainly economists like myself, have reached a wall. And I cannot penetrate it. That's sort of the result, unanswerable question.

OK. What's mutual enforcement? Mutual enforcement starts with the obvious fact that we all live through time, that engagements and possibilities, our business, are not a one-shot matter. For textbook purposes, you can look at a starting economy, one period model.

But actually, we live through time. There is a tomorrow and a day after and a day after and so forth. Opportunities come and go over time.

And so take the simple case where those people who started off on an engagement, they've agreed on a cooperative action which they wish to carry out. Imagine now that that opportunity will come up tomorrow, day after tomorrow, the following day, and so forth. Or make it a year if you think of unit of time as being a year.

So it's a repeat situation. Now, it doesn't have to be repeat. I'm giving you the simplest version. So repeat means, essentially, the same options are occurring periodically, regularly and periodically, and for a long while.

Now, mutual enforcement starts with the idea, with the very deep idea, that one way you can punish people for misbehavior is to withhold cooperation in the future for those who have violated the agreement. That's a cost. It'll be a cost to you, of course.

But it's a cost to the relevant party for our purposes, namely the person who has violated. And if he knows that he will withhold cooperation in the future and if he cares about the future, that's where really the deep result comes from, then he will think twice before violating the agreement. So the mutual enforcement is different from any external one, because the entire problem is solved internally, inside the community in question, the group in question. No third party is required.

Now, you can see how, suppose there is a method by which the mutual enforcement thing can be made to work, then that'll explain three in principle. Because in getting to grips with three, the external enforcement, the external enforcer becomes part of the game. And so the mutual enforcement then amounts to a punishment that people can impose or inflict on the external enforcer if the external enforcer does not behave.

In democratic societies, ideal democratic societies, the idea is you vote the nasty, these rascals, out of power. Rascal is, by the way, a 19th century expression in the United Kingdom for any politician. They're rascals. Sort of it's tinged with a little affection, by the way. It's not entirely derogatory.

They're powerful. They're rascals. You know, they have their own interests. And if you're not careful, they're going to gyp you. They'll put taxes up too high, waste the money on something wrong. But then we'll vote them out of power.

So you can see in democratic societies in principle you can use the mutual enforcement for the external enforcer. But, of course, external enforced, in other contexts, it could be not the state, could be a warlord in a war torn society, could be in village chieftains as very often. It depends on the context, depends on the kind of adjudication that is required.

You can have a state operating reasonably well. You can simultaneously have certain types of agreements, local agreements, which are enforced by, let's say, the village chieftain or the religious leader and so forth. Their zones of operation differ. For tax purposes, it will be the state maybe. But resolving a quarrel between a husband and wife, maybe it's the local religious leader who comes and calms them down or something. OK?

I don't need to belabor this point, because it will strike you as being really rather obvious. But it's important to recognize that three is not a deep answer. Four is the deep answer.

Question-- why have I talked in terms of laws and norms is because future enforcement can be seen as being triggered or maintained or sustained by the use of social norms. Social norm have two features to them. When I say something is a social norm, I mean the following.

I have to mean the following. Otherwise, it doesn't have any meaning. One is that it's a strategy, a rule of behavior. I do this if that happens. I don't cut the branch off a tree if it's March and so forth. Or I do not take fruit from the tree until it's ripe and so forth, all sorts. I'm giving you sort of seemingly bizarre examples, largely because they are rampant in rural communities.

When I first read about the practices the rules of behavior regarding foraging from the local commons in Africa, it was just amazing. A whole universe opened up. Because its ingenious rules were-- at least as reported by the anthropologist, by the way. And I'm assuming they were correct.

And I became convinced they were correct, because they made ecological sense. So they probably were correct. And then so these are the rules.

But that doesn't make it a social norm. It becomes a norm if it has-- of course, it has to be, roughly speaking, universalizable. So it doesn't matter whether it's you or me. We both follow the same behavioral rule.

Now, of course, that's not literally true. Upper caste norms may be different from the lower caste norms in Hindu society. But the upper caste guy needs to know the lower caste norms and vice versa. Of course, otherwise, those mutually beneficial agreements won't work.

But the key thing that's required for a norm to be a norm is that it should be in the interest of each party to follow the norm if all others are following the norm. Otherwise, you won't see it there. It won't exist, if you like.

So the trick is to find rules of behavior, such that if each believes that the others are following it, then it will be in the interest of the party to follow it. So a typical norm indicates a mutual enforcement could be take the simplest case which game theorists have studied at great length. And I'll use it now mostly for my purposes-- would be the grim strategy, which says begin by cooperating.

The norm says, we have reached an agreement. We kick off by showing faith. We do the right thing. We do the thing we said we would do.

But we observe what others have done as well. And next period, I withhold cooperation in the case of these completely if anyone violated the agreement in the first period in the initial round. And I withhold forever.

Let's suppose that's the norm-- sorry, that's the strategy. The trick is then to show that it becomes a norm. It can be a norm under certain circumstances. That is to say it showed that there are circumstances in which following that rule is in the interests of each if each believes that all the others are following that rule.

Now, next step is the key one. And it's the neat one. Suppose the answer is, yes, I have discovered a set of circumstances, I being the investigator, such that it is in the interest of each party to that agreement to follow that rule of behavior if he or she believes that all others will do so. And that's true for everybody.

So it's mutually self-sustaining. It's sort of lifted by its own bootstraps through a system of beliefs. Then remember the norm says you kick off by being nice, right?

Now, if everybody's following the norm, then everybody begins by being nice. Next period, they find that everybody had been nice, therefore continue to be nice. The next period, they find that the previous two periods everybody had been nice, so continue to be nice and so forth. So that belief system can take this whole community through this cooperative venture without any problem. OK?

Then there are other kinds. You don't actually see grim, this particular strategy which is very unforgiving, as you can see, that one violation would mean the destruction of the agreement. You don't see that.

But actually, to stand out in Yugoslavia, it has been observed in certain-- there's a economist who went and found that something akin to grim exists in a community which is very near a market town. And the implication is that you need very remorseless punishment. Remorseless here means break the agreement forever if the first violation or something akin to that if there are periodic opportunities or malfeasance.

But in communities which are pretty much like self-sustaining rural communities, let's say, in Africa, grim would be an extremely dangerous norm to have. Because one mistake, and if that mistake is not seen as a mistake or is not seen as an extenuating circumstance, then that thing will collapse. And you won't see that society maybe. It will just disappear, the village.

So what they have are much more intricate norms-- maybe punish a bit for a while, and then bring the person back into the fold and so forth. I won't go any further into the nature of these norms, the character of these norms. Some of them are very, very intricate.

My wife and I went four years ago, I think, to Australia for the first time and the only time so far on a long tour. I had some lectures to give. And we went to this magnificent wetlands of Kakadu National Park, which is in the north. North, meaning it's near the equator. Australia, remember, it sort of down under.

And it's just fantastic. And we were in a boat with a whole bunch of other dignitaries from Darwin, the neighboring town. The mayor was there for some reason or the other. It was completely uncoordinated, actually.

We were in the same large boat in which there was this aborigine, who is describing their culture. He showed us the weapons they had. These were hunter gatherer people. OK? But there was no metal. It was just wooden spears and so forth.

And so he described how malfeasance within the clan would be punished by somebody-- exactly. You're quite right-- spearing the thigh of the miscreant. OK? And he was very proud of the fact that they had perfected the art in such a way that you don't want to dismember the person permanently. Because then he becomes a useless member of the community.

So it has to be just done that just so. It hurts like hell. But you can recover after some herbs in a couple of weeks or so. Let's give that as a start.

So I couldn't resist. When I heard that, sort of immediately my game theoretic ears popped up. And I said, well, look, hurting somebody is not exactly costless unless you dislike the person intensely. And here, of course, we're not looking at the mutual affection deal. It's neutral people.

So the question is, what happens if the person who is supposed to do the spearing, you know, doesn't do it, because I guess he's chicken livered? Is that expression? OK. He's not. Because I was hoping he would answer the way, actually, he did. He said, oh, then he gets speared.

[LAUGHED]

So then I ask, well-- of course, then I knew where I was going to head. Because I wanted an example of a norm like that. So I said, well, what happens to the chap if the chap who was supposed to spear the chap who didn't spear the original miscreant, what happens to him? He said, he gets speared, too.

Now, I didn't ask whether the intensity of the spearing was a declining function of the distance from the original miscreant. But the point remains that you've got to have these convoluted processes to have a social norm work. Now, it is true that then I asked him, is this an unending sequence? He says, I don't know what you mean by an unending sequence, but it doesn't end.

Now, he knew perfect English. At which point, I heard my wife speaking to the mayor by saying, don't mind my husband. He's just a mad academic.

So the point of this is that there are very, very intricate sets of norms being practiced. Not all of them are, presumably, of that kind. But it was very nice to have an example like that.

It was an anecdote. And when we went to Sydney for the lectures I was supposed to deliver, I got in touch with the professors of anthropology to ask where I could find documentation of that.

The particular lady who is the professor of anthropology, she was the world expert on norms of the practices of the aborigines of that region, the Kakadu region.

She said, well, I don't know. I can't tell you whether it's even been in a book. But it will be in some book or the other. So I said, well, I need a reference, so that I can put it in my paper if I write one on that.

So she said, can't you trust this guy who told you? So I said, but it hasn't-- actually, I was puzzled about it. What she wanted to say was the fellow who has written it in the book also heard it from a chap like that. So why do you have to, you know, try and "be scientific," quote unquote by quoting a reference? That has no meaning.

I think there was a great deal of truth in what she said. Anyway, so that's the point of it, all right. So that's where we are with the mutual enforcement.

Now, two points follow. One is that notice that the punishment from misbehavior takes place in the future-- tomorrow, next year, year after, and so forth. If I'm very myopic, then I might not care that the punishment will be in the future, that there will be a punishment. Because it's happening in the future if I'm very impatient, if I discount the benefits of the future at a high rate.

So then it's quite clear that mutual enforcement will not work if the parties are all discounting the future, their private benefits, at a very high rate. OK? Because give me any set of numbers, and I'll produce a high enough discount rate such that it's not in the interest of any party to follow the norm. And remember, the grim norm-- now, this is another point that you should remember.

The grim norm, the punishment goes on forever. So it's the most Draconian punishment device, the norm that I've just now said. One error-- out.

So therefore, if I can find a discount rate such that the grim norm doesn't work, it must be the case it won't work for any other norm. Because any other norm will be more forgiving. OK? All right, so we know that.

Now, what do we do with that observation? The discount rate that I mentioned is tempered by the prospect of future encounters between this group. Remember, I said there's tomorrow. Then there's day after and so forth.

Now, in a stable society, the probability of that opportunity arising among the same group of people is relatively high. So never mind the impatience interpretation of a discount rate. Now, think of the discount rate as picking up something like the probability of a repeat engagement.

It tells us that in a mobile society, four is not going to work very well. Because the chance of a future encounter is reduced. You wouldn't follow mutual enforcement, for example, in Los Angeles in an apartment house. Because your neighbor today may not be here tomorrow and, most likely, will not be here.

So you can't operate on that basis of mutual trust, because the probability of repeat encounter is reduced. Of course, for a nomadic tribe moving along, that's fine. They're migratory, but the community remains the same. So I mean uncorrelated movements is what I have in mind.

So that's the first point. The second point-- and that's the bad news point, which is now suppose the discount rate is low enough that mutual enforcement works. Let's say through the grim norm just for the sake of argument.

So there is this prospect. There is a social norm. That's laid out. But it requires that each of us believes that the others are following the norm and that belief is common knowledge. Then and then only will we be hooked onto that.

If I believe that all others are following the norm and the statement of the norm is you kick off by being nice, then, of course, it's in my interest to behave. Because it's an equilibrium, or it's a social norm. So that's fine.

But suppose in that same scenario none of us believe that the other will follow the norm. Then what? The norm is written down. Or hasn't been written down-- it's come through tradition.

But for some reason or the other, we don't trust one another to follow the norm. Then even though that prospect of cooperation exists, non-cooperation is also a possibility an outcome. Now, that's the thing which is where I'd really cannot go any further with.

I have no understanding of the process. And for that matter, I don't believe anybody has by the way. It's not just my own ignorance.

One reason we don't know is because we haven't really bothered to frame the problem in the way I've just now framed it, so that we find this a genuine issue. How are beliefs created? It's all very well to say the circumstances are right to have cooperation.

But in those same circumstances, you could have non-cooperation, simply because you don't trust one another to follow it. Now, in other words, in the game theoretic language, you have multiple equilibrium, an equilibrium in which non-cooperation is the outcome and another equilibrium in which cooperation is the outcome. And of course, there could be other equilibria, too. But let's just look at these two extreme cases.

So you could have two islands identical in every way. One is cooperating. The other is not, simply because things are happening in people's minds which differ across people.

Now, it's not entirely true that we know nothing about it, about these beliefs. Rumors, propaganda, they help to influence our beliefs. And these tipping points that we sometimes observe about this community-- which is, roughly speaking, at peace with one another and then in a very short period of time, convulse into unruly behavior at the best of times or civil war in the worst of times. People who woke up in the morning as friends, by noon they're at each other's throat.

That kind of shift is explainable through-- when I say explainable, I mean we now know two things. We know that that can happen through simply a change in the beliefs that people entertain about one another. And two, that you can't predict it. Because we don't know how beliefs are.

And so the sociologist looks at it and says, god, we should have been able to predict that. You know, we saw symptoms of it. But that may be a post-rationalization. It could be in principle.

Because in theory, it could be unexplained some cues that people are picking up which are not in the public domain, which makes people mistrust one another and end of cooperation. OK. So that's one thing. I'll now make a few concluding remarks.

Four is the key to understanding cooperation in my judgment amongst sort of non-family members. Prosocial disposition is not really about institutions. It is about how we are and then what follows from that.

But three and four are the key things, but four is the one. But four assumes a well-defined group of people. And, presumably, they know one another.

Historically, that's not very difficult. If you look at structures in Africa or South Asia, then, of course, people are born embedded in a network. Most of the networks people have, certainly in South Asia, the networks that you have even now has remnants of your family background-- your father's closest friend, your mother's relatives, people you met when you were a child.

And I believe we should think on networks as capital assets. I believe creating new links involves resources as we know for a fact. We go out to dinner together, bonding. Expressions like bonding, that involve expenditure of resources-- if nothing else, time getting to know one another and so forth and so on.

All of that requires investment, private investment if you like. But a second feature of these networks I think is extremely important is that they're irreversible up to a point. So we have expressions like we take our relatives for granted. We take our nearest and dearest for granted.

What do we mean by that? We really mean that [AUDIO OUT] usually, but most of us probably have experiences of not keeping in contact with our parents or close relatives and so forth. But there's a rationale for it. And I've just now given you one.

The second thing is that they are-- so they are durable. Sorry. The other thing is they can be redirected. So I've become your friend. I can't take that past investment and carry it to you and say, well, now we are friends.

I have to start anew, as it were. Unlike, say, table which can be used for one purpose and then taken it elsewhere and then used for another purpose. So it's not shiftable.

These are non-shiftable. So one thinks about these capital investments that people make in getting to know one another and so forth along those lines.

The third thing about it, and that's where I'll end, is I began by saying people have discovered a mutually beneficial course of action. So the agreement is one where everybody is benefiting. That's the easiest one to think about.

And that's the one which pretty much all writers on social capital have been emphasizing. And I think that's a mistake. That's a great mistake.

And that's one reason why the subject has had such a warm glow about it. I was talking to some students earlier today saying that we economists have spent sort of a whole lifetime talking about market failure or the weaknesses of markets. Political scientists talk about the weaknesses of government.

But those who work on social capital seem to sort of have a warm glow about the whole thing, how nice. Because people are engaged in these, you know, celebratory things and so forth. I think that's a very great mistake.

But there is an intellectual reason behind it in my judgment, which is the following, that the key gain which has been studied, if you like, the metaphor which has driven this whole literature, has been the prisoner's dilemma and the repeats of the prisoner's dilemma game. Now, the prisoner's dilemma game has one feature which many people don't appreciate. You cannot do badly by cooperating.

There is no scope for it, because the equilibrium of the game in a one shot game is a minimax strategy. That's extremely important. All agreements must be of a kind where at least no party is worse off than he would be by not cooperating.

But there are plenty of games, like the common property game that I constructed in 1975, I began with. If you write down the model, you'll find that it's not a prisoner's dilemma game. Of course, it's inefficient. The non-cooperative equilibrium is inefficient.

And so the repeat, the mutual enforcement, can make everybody better off. But there are scope. There is a scope for social norms to be there which makes some members of the community worse off on average than she would have been if there had been no cooperation.

And I use she advisedly, because if there's any sense to the notion of exploitation, that theorem has it. That is, it's possible for you to have a corporate outcome in the sense that norms are in place. Ask people why they are behaving one way rather than another, and they'll say, oh, this is how we-- may have it internalized. This is how we are. This is part of our culture.

Another way of saying it is it's in my interest to do what the norm says I do given that all others are following the norm. But if you look at the time average of the payoff of some of the parties in it, the average could be worse than if there had been no agreement at all, and they had a sort of commons dilemma. That can happen in games of proper games.

And the one I constructed in '75 certainly has that feature. Now, the reason I like that result is because it gives a sharp definition of the notion of exploitation. It's sharp, because it says, yes, the community rules are exploiting a person.

Because that person is worse off than she would have been if there weren't these rules. Maybe others would have been worse off. That's OK. But she wouldn't have.

So it takes us away from the original casting of the problem, where I said there's a mutually beneficial course of action. This one is not mutually beneficial. And yet it can be sustained as a repeat game true norms.

And the other reason I think it's good to-- salutary to know this kind of a result is that it suggests-- and that's how I'll close, by reminding you-- that we tend to have this warm glow about communitarian relationships. But we need to be very circumspect before we arrive at that point of view. Because it could be that you are missing out on some unfortunate subgroups within the group in question.

A general, characteristic between, as talked about, laws and norms, now I could think of, if you like, if you want to classify-- laws are what drive markets. Contracts, written or tacit, they're contracts. And norms guide communities. So if you like, I'm trying to give you a theory underlying communities versus markets as alternative resource allocation mechanisms.

One further weakness of communities is that they are exclusive. They're not inclusive. They depend on networks. Therefore, there is a genuine sense of us as opposed to them.

Now, markets are typically criticized for being inhuman. Because my money is as good as yours. What kind of a world is that? Where are people and so forth?

I would argue that that's one of the great strengths, moral strengths, of the market. Because it's inclusive provided, of course, you have wealth or income to be able to buy the thing. But then we know the market has deficiencies as regards to solving inequalities issues.

We know that. That's part of the bread and butter of economics. The similar thing hasn't happened in the community literature.

And what I wanted to do was to bring out the dark side of communities through a theoretical enterprise. And the kind of communities I come from-- or let's put it this way, in the larger set-- we had a caste system in India, for example, where some are untouchables here. And it's very difficult to square that with the idea that it's mutually beneficial.

Now, in some sense, it's damned difficult to prove that it isn't, by the way. I should tell you that. Because you have to think about counterfactuals.

You observe that there are duties and obligations arising across the castes. It's not the case that the untouchable doesn't receive benefits from the interaction. He does.

If he needs a loan, he gets it from the upper caste who has some money. That's part of his obligations. That's part of the mutual enforcement deal and so forth.

And he knows his place. And she knows her place and so forth. And there is a temptation without any further thought to see that as a very beneficial arrangement.

At one level, it is. If there's no other possibilities, this maybe the best on offer. But in a world in which you can move from four to three through good governance, if it's possible, the mix of four and threes I think in some sense-- where the heart of the social sciences should be now. Because you'll never get rid of four. Because we know, otherwise, you can't discipline the enforcer, the external enforcer. So you need four.

The question is where should four operate. On what kind of parts of our activities? And where should the external enforce have his say?

So if you like, this classification of the activities that we are engaged in in our lives from them when we wake up to when we go to bed and even when we are asleep, of course, then we recognize that one is involved, two is involved, three is involved, and four is involved. And what we need to know is the strengths and weaknesses of three and four, of one, of two, and then to have some kind of a sense of how to think of a societal structure in which some activities fall in the domain of four, some on three, some you rely on one, and so forth. But that's a big open set of issues. Thank you very much.

[APPLAUSE]

Charles Redman:

[INAUDIBLE] will take some questions [INAUDIBLE]. Aaron, speak up.

Audience member:

What about the [INAUDIBLE] that different classes have different access to controlling the violence on the part of the external enforcer. There are [INAUDIBLE] people access to organizing that on [INAUDIBLE]?

Sir Partha Dasgupta:

The question is there is unequal access to the control mechanisms for coercive purposes. The state has the power of coercion. I think that's what you have in mind.

Audience member:

Well, different classes have a different relationship with the state. Some have a more direct control on its violence [INAUDIBLE].

Sir Partha Dasgupta:

Yes. That's absolutely right. And in some sense, I'm bypassing that issue. Because I began by saying that suppose they have reached agreement on the sharing of the benefits and burdens.

The question is, how can they trust one another to carry it out? I didn't address the first question. The question that has been suppressed is the question that you want an answer to. How does that get played out?

And I haven't discussed it at all, largely because I'm not a very great expert on bargaining theory. That's A. And B, bargaining theory itself is in a very unsatisfactory stage, state.

But it doesn't require rocket science to get some sense. One of the things is that I think the past has an enormous weight on the resolution of that problem, the problem that you're raising. Whereas, the way I was thinking of three and four, in some sense the past doesn't matter. But of course, in practice it does matter.

Don't trust that cost. Don't do business with them. They become the other, because of some past misbehavior, let us say, or maybe imagined misbehavior. Who cares? We can construct some story.

But it helps sometimes to not think about the past dominating the present. Because in some sense, some of the greatest tragedies of human history are because we are too wedded to the past. So I'm not a great believer in reading too much history, by the way, for that reason.

Maybe it's safe not to know too much about the past. Because you're carrying a lot of dead load of grievances. That was a flippant remark, by the way.

[LAUGHTER]

Audience member:

It seems to me that, to a certain extent, mutual enforcement can lead to a never-ending sequence. In your example of the spearing the leg, if everybody in the culture then [INAUDIBLE] spear the leg, there's incentive to allow that to collapse. Because nobody wants to be speared in the leg.

Sir Partha Dasgupta:

Sorry, nobody wants to what?

Audience member:

Be speared in the leg.

Everybody in the society is charged with spearing the next person, and the next person, and the next person. Eventually, that cascades to everyone in the society. And there's an incentive to allow that whole system to collapse. Because nobody wants to be stabbed in the leg. So you could certainly say, doesn't mutual enforcement have to have finite--

Sir Partha Dasgupta:

No. But remember-- no, no, no, no. No, the equilibrium will be that the person who is charged with stabbing does the stabbing. Because the idea is that the cost to him of stabbing is less than the cost to receiving a stab.

If he's aware of that, he won't do it. But you need that to be credible. So you need to threaten the guy by saying, oh, but you know there's a credible reason why you should worry about being stabbed next time. Because if you don't stab, you'll be stabbed. It's a credibility issue that's being solved by this unending chain. But the cost of enforcing must be lower than the price you would pay when you are punished. Otherwise, none of this would work, including external enforcement.

Audience member:

I was wondering in general, the idea of the social capital is where you have a set of norms, conventions, traditions, interdependencies, trust and habits. When you have good ones, you could get a positive developmental outcomes, right? But in using your model, how would you know whether the norms and the habits are good or bad? Is it also by measuring the outcomes?

Sir Partha Dasgupta:

Yeah.

Audience member:

In which case--

Sir Partha Dasgupta:

I mean, at least the way I'm looking at it here--

Audience member:

Right.

Sir Partha Dasgupta:

--the answer is yes.

Audience member:

So in a case, it sort of can lead to some kind of circular logic, where you have the positive outcomes. You have the positive norms. And where you have the bad outcomes, then you have the bad norms, right?

Sir Partha Dasgupta:

Well, just like anything else, you assess-- I mean, unless you're a deontologist, you would assess the value of actions on the basis of expected consequences of the actions.

So similarly, here you would be-- but bear in mind one thing. One reason why norms could be stultifying-- see, one of the reasons why I think third world economists very often get a warm glow out of norms is they see that without them the community would not survive. Great. And that's probably true. I'm sure that's true.

But the next question arises is in a changing world where there are alternative opportunities appearing on the scene, should people be in a position to get a bus ride on those opportunities-- e.g., access to a market, wage employment, which makes people better off? Now, there could be sanctions to departure. Exit option may be costly if the norms say that.

So you have a problem. The person wants to migrate to make his fortune. But he is needed in the community. And he may have to pay equilateral, like leaving his family behind.

And then the understanding is that he will be sending remittance and so forth for the public goods on offer in the community in question. Now, of course, it could be if those are high costs, then people won't leave. And that that could be restrictive to the people in question.

Audience member:

But the problem is that these norms and conventions are supposedly historically constructed and territorially embedded. Therefore, we can lock ourselves in a kind of historicity that this place is doomed. And the second one is that we confuse cause and effect. If we are looking at negative outcomes as being the measurements by which the norms are measured, and you say, well, this community is poor, therefore the habits are bad, then we tend to [INAUDIBLE] almost blame the victim.

Sir Partha Dasgupta:

No. I think you're reading more than what I said. I would not infer that the norms are bad, because the community is poor. The community could be poor, because it's isolated.

Audience member:

Or it's good.

Sir Partha Dasgupta:

Let's take-- no, why go for that? They're isolated, so isolated they can't trade with the rest of the world. There are no roads. And so it's a very restricted set of exchanges that it can take and [INAUDIBLE] It allows them to survive, but that's about it.

The norm is fine. It's enabling them to survive. But they can't do anything, because they-- I mean, this is Jeff Sachs's view of African tragedy is that these are isolated enclaves with lousy neighborhoods, landlocked countries with lousy neighbors. So they can't trade.

And they're locked into a permanent state of-- but see a norm is not enshrined in a tablet. A community, in principle, can have a norm which they have been following, and then opportunities open up. They recognize the new opportunities, and then they sit down and construct.

All right, now let us allow this young man who is really bright in his primary school, let's try and collect some money to send him to secondary school in the town. Now, that has some implications on society, some dislocation. He may or may not come back. He may become a bit uppity, you know, and so forth and so on.

But a recontracting of the engagement with changing circumstances is the way forward, if you like, for that community. And you might observe it as so, that it's a community where the actions on which the norms are operating are now changing. That's the point.

The norm is not the issue. It's the actions that are being supported by the norms that are the issue. And those actions can change as communities, in principle, recontract.

I'm using recontract now in a loose sense. But the norm is still the same, that, you know, we will punish you if you do that. Because you've got equilateral left here and so forth. So if a community is capable of observing that there are gains to be had, then it can, in some sense, move out into it. But if it doesn't, then it's stuck.

Audience member:

If we think about your commentary on markets and what economists do with respect to markets-- we identify, focus our attention on market failures and [INAUDIBLE], worry about the interventions, the policy instruments that we might do to correct or to try to compensate for the failure of-- the classification of norms based on outcomes and recognizing some as less desirable than others then leads on to ask the question, well, what are the equivalent of the instruments that we might consider to induce the communities to rethink the way in which they're using their norms?

Sir Partha Dasgupta:

Very good question.

Audience member:

And how do you do that?

Sir Partha Dasgupta:

The question is if a community through the norms are choosing actions which are suggestive of, say, exploitation, how it ended, what's the way out for the community. And I think sort of my own feeling is, of course, that I would like to see for reasons of coverage more and more of a dependable external enforcer.

Because market exchanges have the great virtue in this ideal of form, obviously, of enabling a greater range of people with whom you can transact. Insurance is a very good example of one case where the bigger the geographically dispersed coverage and the character of the people who are covering themselves, the more diverse that is, the more risk you are being able to shed. If you're a bunch of farmers in a small community and you try and mutually insure one another, say, because of harvest failure, you're not going to get much insurance. Because they're highly correlated risks.

So in many other examples, you might want to think that, well, with modern communications and so forth, you more and more want to go to three. But, again, I wanted to end by-- remember, I sort of suggested that what we need is something like an understanding of what should continue to remain within four. Because without that, you don't have the disciplinary action over three which you need, which is where Putnam came in.

So I suspect my own feeling is I guess my answer to your question would be I would expect a good three to deliver in improving the performance of four when four is misbehaving in a particular context. So in some sense, you need four for three. But I think you need three to get the allocation right or less wrong in four.

And we do it. After all, the state comes in and says the caste system is illegal. Now, you may still continue to practice it. But it does have a disciplinary action effect, no question about it.

If nothing else, you observe that untouchables are now joining the civil service. And you have to sit next to them. And if you don't like sitting next to them, you don't get the job. You have to resign.

So it may take time. But that's how it's played out. And you in your own history have seen it in recent decades I guess. And that's where the power of three comes in.

Audience member:

[INAUDIBLE] But it does seem to be economists have generally said-- and I think this is fine-- that we have to worry less about norms and the set of incentives that would make people behave as if they ascribed to a particular norm even if they don't. So I think, actually, that was the best defense of democracy that I've heard.

Because this doesn't really work that well if it's not a Democrat institution. But what I wanted to come back to was, well, I'm happy to hear you say cooperation and trust is not always a good thing. I mean, to me one of the weaknesses in resilience and sustainability literatures generally is this warm fuzzy glow they attach.

And we have colleagues in the Resilience Alliance who have said trust is critical for resilience. And my response to that is warranted trust is useful for resilience, as warranted distrust is useful for resilience. So you should trust when it makes sense to trust. And you should distrust when it makes sense to distrust.

And that would make you the most resilient. But you have examples where particular individuals in a community are worse off in a cooperative environment than others. Whereas, others still are better off given a cooperative environment.

And this is actually coming out of Shade's work, The Evolution of Cooperation. But one of the popular explanations now is this altruistic punishment. And I understand your concern about saying let's evaluate norms based on outcomes.

But it seems to me just as dangerous to say let's evaluate norms based on when we feel good about them or not. And so all sorts of folks are saying here's a level of altruistic punishment that establishes cooperation. So it's that a good thing?

And then if you look at the resources available to those groups relative to what would've happened without altruistic punishment, they're all worse off. Not just particular individuals, but all of them. It's more of a comment than anything. But I don't know if you've looked at that literature on it.

Sir Partha Dasgupta:

Well, it depends on what the context is. I'm not belittling the original question with which I started. But I want to argue that that's not universally the situation on the basis of which cooperation proceeds.

I can't prove it. Maybe discrimination against women still makes them better off or have in all society. How do I know? But this theory, at least the theory that I'm instructing, tells you the possibility of some groups, subgroups, being essentially exploited. And by the way, that's the only interpretation I can give to exploitation.

Audience member:

I think there's also the possibility that cooperative outcomes actually make everyone worse off.

Sir Partha Dasgupta:

Oh, that can happen.

Audience member:

Not just sets.

Sir Partha Dasgupta:

Absolutely. It can happen. Sure. I mean, one can prove that as well. But the asymmetric treatment is more. Because somebody would be smart enough to say, well, here's a chance of one subgroup moving up or something.

Audience member:

But the problem becomes marginal change versus sort of macro-level changes. So it may be that there's no improvement with only marginal change. And if that's all that one achieve, you're in a local maximum but not a global maximum.

Audience member:

My question comes from yesterday. How will the world mutually enforce greenhouse gas emissions and the use of fossil fuels?

Sir Partha Dasgupta:

Well, the way it's operating, I mean, one scenario is this cap and trade scenario. So the cap is based on a mutual agreement. It's enforced through individual governments. Because the cap then allocates it according to the market system.

So presumably, it will be the individual countries who-- governments-- will be monitoring whether their rights are being respected, that you are actually meeting just the right amount that you have bought the right to. So that'll be three. Three is operating on it. But the cap itself is coming out of not a mutual enforcement, but a mutual agreement which hasn't happened yet. But that's the hope.

Audience member:

[INAUDIBLE] the world if you assume global warming is going to be really bad, wouldn't it have to be mutual? Because you can't have one country externally enforcing another country to put the cap on emissions or [INAUDIBLE].

Sir Partha Dasgupta:

Oh, that could happen, sure. Sure. In principle, that can happen. We've seen that happening over wars that take place over resources, certainly amongst communities. We've seen that happen. Sure, that could happen.

But the benign scenario of cap and trade is that it's really three which is operating. Because the actors are not the governments. The actors are the firms, the households and so forth, who are emitting.

And it's their behavior which needs to be controlled. Because the agreement presumes that they're all benefiting given that they're taking into account the interest of the future. So it's under three.

OK. Thank you once again.

[APPLAUSE]

This presentation is brought to you by Arizona State University's Julie Ann Wrigley Global Institute of Sustainability for educational and non-commercial use only.