 Okay, so I want to thank you and you, Wider, for having invited me to participate in this session. And as you have heard, I am directly involved in a rather important project with Wider on gender and development. I feel a little bit embarrassed because, as you will soon see, I am the only man that is going to speak here because there will be two women speakers, female speakers, and two female discussants. So I better talk in the beginning, you know, so that they can beat me up as they want. So the paper I want to present in the short term is in fact a joint work with one of my PhD students called Julia Camiotti and I will have more opportunity to refer to a precise work later. Now, my issue is really about strategies to fight against customs that are oppressive for women. You know, you can think of a number of customs like early marriage, food binding in China, female genital mutilation, and others, crimes of honor, et cetera. I will give a number of examples in my talk, but this is the question. And what I am interested in is that during the last 20 years, there have been a new very determinate approach to tackle those customs, which is the human right approach. And the human right approach is, you know, something that I can call social engineering in the sense that you want to use the public power to change those customs, you know? That's what I mean. Now, this human right approach has been embedded in a number of declarations, especially UN declaration. I think the first one took place in 1979 and it is called a convention to eliminate all forms of discrimination against the women. And later on, there was another one in 1995 that was another landmark, which was called a UN fact sheet on harmful traditional practices affecting the health of women and children. And you know, what characterizes this convention and the approach, the human rights approach that has been derived from it, is the willingness to enact laws that prevent, that ban discrimination of women or a number of practices that are considered to be anti-women and to accompany this legal step with building awareness among the women and the men about the harmful side of these practices. And in a sense, what we want to do is to say, what can we say about this and to bring the light, if I may say, of economic analysis. And you know, to bring the light of economic analysis is pertinent for the following reason. Is that if you see as the dominant model of the human right approach, especially embedded into the UNICEF kind of approach to the question like female genital mutilation, you see that the kind of theory, implicit theory there in mind, correspond to a well known game in economics, which is a coordination game with multiple equilibria which are not equivalent in social efficiency terms. What does that mean? It means the following, that you look at the problem of an oppressive norms for women as a coordination problem. And you have this idea, or you assume, that in fact all the people and in particular all the women do not like this practice and would like it to be abandoned. And so it's just a question to change the expectations of the people about what the others are doing. Why? Because the perspective of social norm has this characteristic that you are concerned with what the others do. And the utility or the well-being, call it whatever way, the level of happiness, I don't know how to call it, that you have depends on how many others do it. And if too many others do it, then you have to do it because otherwise you will be ostracized, you will feel guilty because you are going against your own community, values or whatever. So that's the social norm approach is essentially based on the assumption that there is interdependence between the behavior of the people. And so there is a critical role for your expectation about what the others are doing. Right? No, in this kind of coordination game, people all do something that they dislike, that they hate. For instance, there is a rule that prevents a widow in India to be remarried after the death of her husband. You know what you say? Women do not like that. They would have the possibility to get remarried. But since everybody else followed that rule, it would be a bad thing for you to break it. And it's just a matter of coordination. You would like to change the coordination. Okay? No. The thing is that if things were so simple, you just need to change the expectation of the people, but it shouldn't be too difficult because you just trigger a new kind of, as it is called in game theory, focal point. You launch an idea and people say, yes, why don't we do that? It's much better for us. That's what I want. And everybody says that's what I want, then the thing should change. No. As I say in the second paragraph, unfortunately, things are not so simple. And there is little empirical evidence that support the coordination story. And the main assumption, the idea that people dislike the custom and follow it only because they have pessimistic expectations about other behavior, does not really realistically apply to many realities. And so things are more difficult than it seems. Okay. And still on that point, there is a study by Belmar and co-authors on Senegal, he looks at the data, and the support for female genital mutilation. And what he finds is that almost 90% of the variations in support of female genital mutilation is explained by individual and household characteristics, not by any social process or the behavior of the others, et cetera, which really go against this social norm approach. And more than that, he shows that the more there is in an area, the more support there is for female genital mutilation, the more important is the role of individual and household characteristics as compared to social characteristics, like the community characteristics. That's bothering, no? Wellington in a study, experimental study in Sudan has shown that the kind of behavior implied by the social norm approach is not validated by the data. And now I'm coming to the result of my studio, Julia Camillotti, who did a study in Senegal. I want just to say a few words about this. He studied the impact of the action of a well-known NGO fighting against a female genital mutilation called Tostan. It's a Senegalese NGO, but essentially has an American director and was in beginning funded by, and still largely funded by the US. Okay, no, they start exactly with the model I was describing. Female genital mutilation probably is just a coordination problem. But still they did conscientization work at the village level for a number of years, drawing attention of both men and women on the bad health outcome, if you cut the girls, et cetera. And they had this idea that since it is a pure coordination problem, we can have public declaration whereby village leaders from all the communities that have been worked out, worked on by Tostan, will meet and make a public commitment that they abandon female genital mutilation. Okay, there have been 101 such public declarations in the area of Senegal, especially the southern area and the Fulani area, where Tostan was working. So quite a lot. And Tostan considered that, in fact, this was done and this was a successful enterprise. Now when we look at that more carefully, what do we find? More important results, and that's really detailed in the papers by Julia. First one, Tostan had an impact on female genital mutilation that you can extract as a significant result, but it is a very weak impact. So the impact is quite marginal, significant but marginal. Nothing like a massive abandonment of the practice in the villages which have participated in the public declaration. Right? Now second big finding that she has is that one clear impact of the intervention of Tostan is that the age of cutting has been advanced, that instead of being on average at the age of six, now it is at the age of four. And this is a perverse effect because the health consequences of early cutting are bad. So we can ask, why is it that earlier cut, and there seem converging but qualitative evidence that the reason why women do that is that they can conceal the practice more. Before they would do that in public, nor they would do that privately and in a concealed manner so that this is not much talked about, but they will still stick to the practice. But she shows that the public declaration had absolutely no impact and that in fact villages were not even quite aware of what was happening. You know it was like kind of NGO organizing meeting and people in the village saying, yeah why not? You know, we will have a good meal there and they go and participate in this nice meeting and they come back and they continue as before. And lastly, she has some evidence that if coordination there is, it doesn't take place at the community level. But at the sub-community level like the family network or extended family. Okay, now the very fact that there is no massive abandonment of the practice like female genital mutilation doesn't mean that the social by itself, that the social norm approach is wrong. Why? Because there are two explanations that could still explain why there be only partial abandonment of the practice even with the social norm, logical mechanism. The first one is coordination as I have just said, does not take place at the community level. It takes place at another level. And the second one is that you assume in the simple canonical model of UNICEF that people have homogenous preferences. You assume that everybody has the same aversion to female genital mutilation. So if you drop that assumption, then the problem becomes much more complicated. You assume that people have different aversion in a village, in a community toward female genital mutilation. But then you know the impact of let's say a law that is banning FGC and that is supposed to have as it is called an expressive effect which means a focal point effect that would change the expectations of the people. No, the impact of the law becomes quite complicated and depends precisely on the distribution of individual preferences. And then you have to specify how these preferences are distributed in order to derive results. No, for instance, if you have a distribution that is by model with many people, I suppose it will come back, yes? Okay, now if you have a distribution such that many people are strongly adverse to the custom and many people are not very adverse to the custom. They lack the custom, they have few problems with the custom. Then it's easy to see that there is a unique equilibrium. And this is an equilibrium in which a positive fraction of the women will cut their daughters and the others will not cut their daughters. And you can have any law. If the law has only an expressive effect, nothing will change. Same if you have a uniform distribution, which means that at any level of aversion towards the custom, you find the same proportion of the population. But there are some distribution. If you have many people who have a moderate aversion towards the custom, then it may be the fact that the law can bring down to zero the percentage of people that apply the custom, you know? Because it can be you start with 100% do it and the law has a focal point effect and then it falls down to zero. But that's only in this case, then we have this special distribution. Otherwise, you don't know, you should be skeptical about that. And since we know that very often there is little impact of intervention and of the fact that there is a law, no, we have to shift to something else and say, no, this is not enough in the strategy. We need something that modifies the payoff, the returns if you want, that the people derive from stopping to cut their girls if you want to have a chance or you need to change their preferences. So if you have that, then something can be done. No, modifying preference is done by awareness campaign, you know, things like that, that is done with NGO, can be done by public campaigns, et cetera. Again, if you do that and you modify preference, it can be shown that it all depends on where you start with the distribution of the preference to know whether you can have an impact or no. Maybe you have no impact, maybe you have a slight impact, maybe you have a big impact. So, you know, again, you have to be agnostic and you have to know more about the attitudes of the people. No, modifying payoff can also be achieved through law or through economic policies. No, through law, what does that mean? It's that the law is not just there to ban the custom, but you are punishing the people who violate the custom, okay? No, you are punishing the people and you try to detect the people because if you say there will be a strong jail punishment for people who violate the custom, but you have no way to know who is doing it. You have no detection technology. Then you know your law is of no avail. Okay, so you need to have that. No, the, yeah, shaming environment is something else. No, economic policy is mine. It means what? That you increase outside economic opportunities for women. You give them the possibility to go out of their home by a strategy really targeting women for creating employment for them. Now, a great advantage of the economic approach as compared to the legal approach is really that it avoids head-on confrontation with the tradition and the identity-defining culture in which it is embedded. No, it seems to belong to the economic sphere, to abstract market force that are beyond the control of the people and their law for face-saving among the people who are the upholders of the tradition. You know, people will just say, what can we do? You know, our girls are leaving. You know, they know they want to choose their husband and their mother. Or what can we do? So we try to find a compromise. They feel helpless. So that's the advantage of the economic approach where the legal approach has this disadvantage that it's directly antagonized the custom. And this is all the more annoying as the law is clearly inspired by Western values and human rights approach which is resented by the people as coming from the West and being a direct attack on day culture. So what even the women will explain is that we want to protect our culture. We won't let our culture being destroyed by these values coming from the West. So that's the disadvantage from the law. And the politicians are well aware of this resistance. And the problem is that, you know, they may not credibly enforce the law. They might just say, like Mama Arlo, the minister of the governor of Senegal who presented the law banning FGC in Senegal, he told the people reassuringly, don't bother too much. The law is there but it will not be implemented. Public declaration. I'm left with two minutes. Okay, no. The problem with the economic approach is that it is very good in the sense that it has an effect of increasing the bargaining strength of the women because it provides an exit opportunity. So in that sense, you know, it's a very effective approach. It is indirect so that's what we want. The problem is that it must be the case that women are free to seize these outside opportunities. If you have seclusion norms that restrain the physical movements of the women, then these outside opportunities may be there but they will not be available. And so that's the main problem with the economic approach. And in the sense, you know, you need to think of a strategy as maybe being an approach that combines a legal approach, an economic strategic approach and in fact, awareness building approach. And it is the case that they have to be combined in different ways. And in a sense, this is what the paper is discussing also in the light of empirical evidence. And let me just conclude by saying that, you know, in a paper in which we have a theory with customary authority making decisions about cutting of the girls. If it is still a community project and women make free decisions they have heterogeneous preference, but they have to decide, do we go to the court if we don't want to cut our girls or do we stay within the traditional customary juridical judicial domain. So that's the decision they have to make. What we show is that the impact of having a law banning the cutting of the girls and the impact of expansion of outside economic opportunities is strictly formally analogous. It's the same impact. So you can say, yeah, but so they are substituted. No, they are not substituted. They can add up their effect. You can work on both. And since, you know, the economic approach has some problem, legal approach has some other problem maybe you have to give preference to the one that in the particular circumstance in which you are is in fact creating fewer problem. But you have to be aware that they all have disadvantages. I think I have to stop here. Of course they're much more than I would like to say but the time is short.