 Hello, I'm Zofan from University of Western Sydney. My question is to the paper on aid and inequality, Latin America, yeah. In the framework, you showed that the link between aid and inequality should be most likely through poverty reductions. But in the empirical executions, you see the direct effect. So I'm wondering why not formulating a kind of two-stage process? So a kind of reduction in inequality is the outcome of poverty reductions. And then before the reductions, it's the outcome of, say, inflow of aid, for example, and see how the outcome would look like. So that's a short question. Thank you. And then back here. Thank you. I'm Christophe Moulin from the University of Ex Marseille in France. I have a question for Rosé-Marie, and another one for James. Rosé-Marie, you showed some very interesting graph about the channels, the possible channels of influence of aid on inequality. And I found that you don't exploit so much all the reflection that you had about this in the estimation output at the end. And I was wondering if there would be a way to do it more. For example, something which is really puzzling in this aid literature is that we would like to look at the component of aid. Surely food aid and military aid, this is completely different for what you want to study. Naturally, the kind of data that you have doesn't allow you to do it always. But for some countries, you have this kind of information. You know it. You know that at the time of a war, when you get aid from the United States sometime, it's not going to be food. Not all the time. Another simple way to do it would be to introduce cross-effect in your regressions. You have a lot of variables. You have a lot of information on these countries. If you, very simple things, the level of aid at the time of civil conflict or military conflict, perhaps means something else than in another time, et cetera. So I would like to encourage you to explore more this dimension that you have already sort of. James, I'd like to emphasize that these new methods are really important for a reason that you don't mention, perhaps by modesty. But there were a lot of data lying in most of the countries of the world about discrete ordinal characteristics which had been collected for ages and which are extremely cheap to collect. And people were wanting to use this data, but the equipment, the simple characteristic of people for monitoring poverty or monitoring social programs. But in practice, it was not really used because it was too messy to complicate it. So one of the huge benefit of this method has been to summarize this very important information by using counts of deprivation. This is what is behind it. Naturally, that works only in situations which are favorable, which are probably most of the empirical situation. But if you have a country where you have people who are dying of hunger, and that's all, and other people who are not dying of hunger but are deprived, I don't know, of saying else on education, in that case, your approach would not work because the ranking would not correspond to what you would do if you had more information on standard living conditions. So we believe that this special case, in fact, doesn't happen very often empirically. We believe that there is a correlation between dimension which is strong enough. Now, what I want to say is that this method of using limited information, partial information, in the form of discrete on order characteristics, is, in fact, an approximation of what you would like to do if you had a very accurate description of the conditions, if you had market, if you could convert everything, for example, in monetary equivalent. So it's an approximation. And I'm wondering if the degree of approximation that you ask when moving from identifying the poor to introducing some transfer sensitivity in your measure, is it the same degree of approximation or are you asking something which is much, much stronger and that perhaps is a bit too much for the assumption? My comment is on the trade paper for an alien inequality. There was a report almost a week ago by BBC that most of the foreign aid to developing world, including Latin America, is badly misused. There's huge corruptions, so that almost supposed to save like 3 million deaths, and yet it does not because of the bad use of the aids. How you are here, we get a sort of foreign aid reduces inequality. How if you allow for the fact that there are negative effects of foreign aid, such as the corruption index or any other proxies that allow for the fact that the aid is also bad, not only for growth, but also my enhancing equality. Are you going to get the same results? Just like the way the allow for some factors then the African army disappears. Thank you. Julie Taruzconi from the ILO. So my question is actually for the first presentation. I was wondering when you look at the effect of EPL and the rigid disease and the changes actually on inequality and growth. Have you thought about taking into account the degree of formalization of employment in a country because I mean, we know that there are some countries where you have a large informal market, so because we couldn't, it was very fast, so I couldn't see if you have thought about that right now. Thank you. Okay, I think it's the easiest on me if I just start actually with Nado and work our way back up the table. There's something I think for everybody. So Nado? Tension of the speed. I'm usually very, very slow. I made the point at the beginning that I mean, this is a total, the exercise is measuring the laws. I mean, there's very little effort to capture any kind of de facto and enforcement issues. And I think I mentioned at the beginning that if they bite, if the formal sector is dominant, then there's a direct, very clear link between the rigidity of these regulations and enable market rigidity, right? And I also, I think I mentioned very quickly too, that that effect changes across the distribution, across levels of per capita income, particularly when you go down per capita income levels, I mean, the formal sector is much larger and therefore the bite of these regulations diminish and of course affects the size. Not sure about the sign of the effect. I mean, that is a very, very important issue. I mean, there's some evidence that the role of enforcement is not that severe, but I don't think it's very much at the beginning of the discussion on enforcement. And it's an important issue that still needs more research. I mean, the index is very humble. I mean, if you want to be not generous, it's ambitious enough as it is, but the focus is only, only, only on the loss, only on the day you are in. Thank you for the questions. I think that the first and the second one are very related. I think it has two dimensions or two parts. One, it's more related to econometric technique or a technical aspect. I'm not the econometrician, David Castels is my co-author, who has dealt with these controversies. I will pass the question to him, but I think that, well, we can try to use to list the square ordinarily or to establish squares, but I think that the main problem, we have to select an instrumental variable. If you don't have a very good instrumental variable for inequality, so you can use or external instrumental variable, you can use internal instrumental variable that is what system GNM is doing. I don't believe that internal instrumental variable prove causality in strictly terms, but it's what we have when we deal with the numbers. So I think it is very interesting to analyze the indirect and indirect effects and we have to classify by types of course. I think the ideal research program is try to analyze which program has designed specifically to reduce income inequality in each country. And I think there's no way now because I suspect that there will be any program is specifically designed for reducing inequality as experimental evaluation. And we only will have the impact of this program, not the average view or the overall effects, but just for some possibly good result, we have detected that there's a high correlation between foreign aid channel to education sectors and the public expenditure in education in Latin American countries. This is not causality, this is not proof, anything is just a correlation, but maybe if we can think about 10% or 15% of overall foreign aid is channeled to education sector. We have to disentangle tertiary or primary education, but thinking in a counter-faltile manner, what will happen to these people or zone or region either foreign aid doesn't enhance the education level in that country or in that region. Of course there will be a lot of cross-effects. We don't have a very, very high correlation in our correlation matrix. So in principle, we don't think that this is a very big problem. And we only, for the third question, we only deal with, we don't deal with corruption. We only contemplate the political regime through a policy two index that is a measure of the quality of the governance or the democracy in Latin American countries. I think that empirically you can detect how the corruption can affect the impact of aid on inequality. Maybe because we don't know how much aid is corrupted. But there is a huge suspect that when foreign aid is channeled through elites or it's captured through the elites of government, this aid is not affecting of course the inequality or even it's improving inequality because the earnings or the incomes of elites has increased. Maybe this will be the case or it could be the case in Nicaragua. I know that in Nicaragua, most of the foreign aid is over the table of the Ortega's president. So he decides what foreign aid is going to every zone or every sector. This is the reason because some donals as Sweden has abandoned Nicaragua. But I think this is more related to political problems that empirical problems. That's what I can say, thank you. Hello. So approximation. Yes, in certain ways, clearly it's a crude measure. It's connected to real conditions though that people are experiencing. Compare that to PVP going up and down with someone's time of day or some weather. PVP has had a very serious problem with the income-based approach because you are so out of control that these wide ranges can occur that are totally unconnected to what people are experiencing on the ground. And measures of this sort are concentrated on to be sure simple and crude indicators and the result is not going to be that sophisticated in many ways, but it may be a better approximation from that point of view. That's just one way of doing it. In particular, in addition to that, the dimensions can be connected to policy directly and that's how it's done in Columbia because the interest was in terms of quality of jobs in the informal or non-informal sector. That's one of the dimensions. Do you have a job in the formal sector or not? Also, quality of housing can be put right into there and that's a big important policy variable for Columbia and that's why it's in there. So that makes it much more tangible and much more interesting, I think for policymakers as a result. It's also the organizing mechanism for the entire ministries, all the ministries dealing with poverty in Columbia. They get together in a room with the president every two months and look at the measure and see how success has been, in each of the dimensions has been occurring or not. That's keeping your eye on the ball to get a job done. That's kind of interesting that you can actually do something about poverty potentially. Try and change income in a short term in using any sort of standard policies. Finally, the discussion of whether it's too much to expect that you'd be concerned with inequality within a measure of this type. I think you're probably right and in fact, I'll go one step further nor Alastair has been talking back and forth with different people as to why do we have these inequality sensitive measures anyway of poverty? Now that of course hits me in my heart, but that's okay, I can put up with it. Her point is that when you have a transfer between a poor person and another poor person, is that really where we should be focusing on policy? No, we should be looking at transfers from out there among the rich to the people who are poor and those transfers don't need to be money, they can be other dimensions in public goods and so forth. So I think that's where the inequality that's important is being realized in that kind of dispersion between the others and the poor. I have a question for her. Good. It's a poison. In fact, I have a question for all three. I noticed that all of my colleagues here were talking about indices like me. That was surprising. The first colleague was talking about an index, I guess it was quite interesting. I'm fascinated with measures of quality of jobs. I just think it's amazing when the World Development Report has an entire volume on jobs and quality of jobs, but never defines what it is. And so I think it's really important to have something that goes that way, but of course this is getting into a backwards way because you're looking at laws. But still, I was wondering if you did any robustness analysis with that and changing assumptions whether counting is good enough or whether you wanna emphasize one or the other types of law more and whether that would affect the results. That seems natural to look at. You had a nice concentration index, my favorite Herfindahl index, kind of a variance but with a little bit of size thrown in. I'll just ask a question. That disappeared in the second part of your discussion with perfect competition, of course. How do you reconcile those two parts of your paper? So I'll answer that. And then finally, quality of aid. It seems like the Center for Global Development has a very nice measure of quality of aid. That's all they're talking about in so many ways. So I would think you could get your hands on that and do a little bit of discussion within your work. Thank you. Thank you, James. So I have actually shown the concentration indices for in the first section of my paper in order to show that when the quota is removed or that there is any liberalization policy, then I want to actually identify or I want to show that what happens to the market structure and that is the first stage or first step of my presentation that the market gets concentrated. And then in the second step, we have shown that what is the final outcome of this concentration? So concentration indices have not been used in the second part and we have only shown the inequality aspects and the theoretical justifications or theoretical substantial of the empirical verification. But concentration indices are constructed in order to show that the structural features of the market has been changed because of this policy reform. Yes, I know the index and even they have a quality of ODN health or in agriculture. But I think that the purpose is to compare across donors. I don't believe that they have a few to introduce in a panel later. I don't know. I know what the law defines as a good job. Yeah, but we look at the pro components of the Botero index to check whether or not they make a big difference which one you pick and it's usually not the case. But I mean, that's obviously because you're looking at just the laws and I'm showing different regions in particular out of gas. Enforcement issues will dictate that some of the components work better or not. Okay, exactly, I mean that's a natural. We started a little late so I'm happy to let people vote with their feet if they want to go. But if there are one or two other questions that people have, I'm happy to extend the session for another five minutes or so just to let people ask any burning questions that they might have for the speakers. So it's speak now or go to your coffee. One of the two. It looks like go to your coffee. So let us thank the speakers very much for a fascinating session.