 So maybe I can start in the meantime. Thanks, Carlos, for a super nice and interesting presentation. So what I'm going to present now is a work that we've had with some colleagues from the World Inequality Lab on land inequality. So I'm Yajna Govind. I'm from the Copenhagen Business School, as Carlos has mentioned. I'm also affiliated to the World Inequality Lab. And this is going to be joint work with Luis Beluz, who's at CUNEF in Madrid, Felix Novokmet, who's at the University of Bonn, and Daniel Sanchez, who's at the Paris School of Economics. So I think the title is very self-explanatory. We're going to be looking at land inequality in the developing world. And what I'm going to try to show you with this very simple and straightforward presentation is the fact that we want to revisit the measurement of land inequality in the world today. And I'm going to tell you in a second why we think it's important to do that. So before going there, I just wanted to present a bit of motivation. I mean, it's just the theme of the whole conference, but still I just wanted to point out that land inequality is going to be super important because exactly because the poorest of the individuals in the world are going to depend on agricultural land or land in rural areas. So according to FAU, free out of four of the poorest billion individuals in the world depend on agricultural land for their subsistence. And this link between the livelihood of the poor and the dependence on agricultural land has been shown extensively in the literature. So we believe that studying land distribution and access to land is going to be a very important tool that is at the hand of different countries when we want to tackle poverty alleviation and especially in developing countries. So there are many different aspects with which land inequality is going to be important. You can think of like really many different dimensions. One of which I wanted to point out today is the fact that there is also a very straightforward link between land inequality and economic development that has been shown in the literature. So in the literature, there has been some papers that have shown that the initial distribution of land is going to matter for subsequent growth rates. And you can think of different channels for which that happened. It can be because of the fact that the more you have land inequality, the more it's going to impede the development of the financial sector because the poorest would not have access to this asset, cannot use it as a collateral to ask for credit and so on and so forth. So there has been some evidence that shows that the higher land inequality is going to impede the financial development in different countries and that's going to affect subsequent growth rates. So, and just another big strand of the literature that you might already be familiar with is the fact that land concentration is also linked to agricultural productivity and that is also going to affect economic growth. I don't want to go in details on these. This is just to motivate the fact that land inequality is going to be related to bigger questions like poverty and economic growth. What we are going to do in this paper is just take a step back and try to understand how land inequality is measured and whether it's measured in the best way possible to tackle these questions. So the first premise that we're going to pose in this paper is the fact that we need more clearly defined and more consistently measured land ownership inequality measures in the world and here the word ownership is involved for good reason. It's because it's going to be the central point of this paper is to try to understand whether we can measure ownership inequality in land and not other types of inequality. And why am I really putting an emphasis here? It's because of the fact that the existing literature, the existing cross-country estimates of land inequality that we use today is from Deninger and Squire. It has relied a lot on agriculture census data. And what happens, I mean, I'm sure you all agree that agriculture census data has provided us with very valuable information when we speak about land inequality. What I'm going to speak about today is more the disadvantage of using agriculture census data. So apart from the fact that this agriculture census data in different countries has all collected with different, sometimes different parts of the land. So sometimes governmental land are included, sometimes private land are included and that might change cross-country and overtime, making comparison across countries and overtime difficult. The three main problems that I'm going to focus on when we want to consistently measure land ownership inequality is the following. So the first one is the fact that the definition of the measurement of land according to agriculture census is to measure the size of operational holdings. And what we're going to argue in this paper is the fact that operational holdings does not necessarily mean land ownership. So you can think of the fact that, so basically agriculture census data is going to be collection of the size of farmlands that are under single management. That's the form of definition by FAU. And what happens there is the fact that each of these plot of land, if they are not under single management, are not going to be linked to the same owner. So if a single owner has different plots of lands in different rural areas, these plot of lands are going to be considered as one piece as if each plot of land is owned by one person, separate, one different person. What we would like to, in the discussion of land inequality in terms of ownership, we would really want to know whether these pieces of land are connected to a single owner. And if we think about what I've just said, then that would mean that we might be underestimating land inequality in terms of ownership when we rely on agriculture census data. The second point that we're going to try to revisit in this with this paper is the fact that by definition, the agriculture census data is going to measure the land sizes. And we know that different types of land can have different values. And when we are thinking about land inequality in terms of sizes, it might not be the same as land inequality in terms of the values of the land. So what we're going to try to bring, and I'm going to show you in a second how, is to bring a measure of land inequality in terms of the value of land. The third and final piece of the puzzle is this, I think the most important part of this discussion because of what I've just motivated this talk with is the fact that land inequality is really important for the poorest individuals in the world. And because census data are going to measure existing pieces of land, it's going to miss out on those who do not have any land. So we're missing the most important chunk of the distribution that are people that are dependent on agriculture but do not have any access or do not own any land. So what we want to do with this new measure of land inequality is to make sure that we account for the landless population. Okay, so we do not, I mean, we're not arguing that the agricultural census data is wrong or it's not useful. It's actually definitely useful to understand efficiency. There's been a large literature that relied on that. But what we are going to argue in this paper is the fact that it might not be as useful or as precise when we're speaking about equity. So what we are going to do in this paper is to move to a different source of information, that is the household survey data that a lot more survey data have introduced or has had a lot of agricultural modules. In the past, I'm going to try to exploit these agricultural modules that goes in details on the land that is owned by different households. Okay, so in a nutshell, what we're doing this paper, we exploit survey data which is going to allow us to focus on the land that is privately owned by households. We are going to provide consistent estimates of land ownership inequality across countries in different regions of the world, both in terms of area and value and accounting for the landness as well. And then the countries that are going to be covered in our work is for now, the ones that have the best available surveys in terms of information, in terms of area and value. This is going to limit us to around 15 countries in the world for today because of the fact that we really want to work with the ones where we are sure that the estimates are more or less precise, we're going to work more towards including other countries as we go on. I'm aware that this table is not readable by any of you, but I didn't want you to read anything out of this. It's just to show you that we have used a lot of surveys, survey data and also the agriculture census data in different countries in the world and there are different years of the different surveys, but it's approximately all in the 2000s, so early to late 2010. Okay, so just a few words before showing you how the results look like. So in terms of the methodology, the first chunk that I've spoken about is that we're going to look at land area inequality, which the existing literature has also done, and we're going to be first updating the estimates using census data. So we're going to go with the existing literature, we take census data as they have used and we're going to just update the existing literature. So as I was saying earlier, the existing literature has done systematic cross-country estimates of land inequality, but it dates back to the 1990s, so we don't have consistently estimated land inequality series after that and that's the first contribution we want to bring to this table. The second point is what I've just explained is the fact that we're going to depart from using census data to provide new estimates based on survey data, and then we're going to estimate that using Gini coefficients and top land shares, that's the equivalent of top income shares if you're familiar with the top income shares literature. The unit of observation, I just want to take a second to detail like the few notes of quotient that we have to have at the back of the mind when we're interpreting the results. The unit of observation here is going to be focusing on private owned lands or household farms. This is going to exclude governmental owned lands or communal land if you want to think about it that way or corporate land. So that's a limitation I think of not being able to take into account government owned land or corporate land, but on the other hand, the nice advantage of using the surveys is that we can really consistently, we know what we're consistently measuring household farms across country and over time. I forgot to mention that here these different countries have different waves of the surveys that we could exploit. For now we're only exploiting the most recent and best quality surveys, so it's only one point per country. We don't exploit the time dimension of this. In terms of the measurement of the size of land, what we're going to exploit is that in all of these agricultural modules, we have self-reported size of the plot, we have GPS data and we also have measurements using ropes by the person who's doing the survey. And what we're going to try to do is to try to use all three informations, to try to combine and see if it's all converters to the same information. And then we're going to try to validate this by using other survey data. For instance, the DHS has a question actually about do you own land and what's the size of land? And we're going to try to see whether that corroborates with what we're doing in the LSMS work. Finally, in terms of ownership of land, the question, I mean this for development economists, you might know that this is like a very tough thing to work with in developing countries. It's the fact that land tenure is not a very clear concept. So what we're going to do is to use a combination of questions that are from the questionnaire. So the first one is do you or anyone in the house will own this land? What is the ownership regime? How was this land acquired? So different countries are going to ask this question in different ways. Normally we have four or five questions that are going in the same direction. So we're going to try to exploit the rich informations in this to be able to say something about ownership. The second step is the value inequality. I think that's the most challenging part of this discussion is the fact that we want to estimate the value of land and its distribution. And currently what we're doing is to use the self-reported value. So people are asked if you had to sell this land tomorrow how much you would get on the market from that land. We're aware that this is not probably the best or most accurate estimate of value. So we're trying to work on a more refined measure of that. But I think it gives a more or less nice picture of value of land. Finally, including the land less, here we're going to account for the population dependent on agricultural land that are landless. And there are different ways of doing that as well. One, the way we are using now is to look at who are in the agricultural module but do not declare having any land. Okay, so let me get to the result. What we find is four main takeaways, if you want to remember something from this paper, is the fact that when we estimate land area inequality, so I want to here emphasize that this area inequality from the census and the surveys, they're highly correlated. I was like speeding up on this. So, yeah, so when we use it from, when we are focusing on areas, the inequalities estimates from census and surveys are very highly correlated, so they're very similar. So then you might ask what's the point of doing this? Exercise is the agricultural census is already doing a good job in estimating land area inequality. But what we're going to show next is that land area inequality can differ importantly from land value inequality. So that's one of the main results of this paper. Moving forward, accounting for the landness, I mean, this is no brainer, is going to change the estimates of inequality because you are adding people that don't have any land. It's as if not accounting for the landness would be the equivalent of not accounting for people that don't have any income or unemployed. So here, accounting for them is going to make a difference to the inequality estimates and it's going to actually vary depending on how, what's the proportion of the landness population in these different countries. And I'm going to show you in a second that this might be very different. And finally, and why this whole exercise is very important is the fact that the regional patterns that we're going to see in terms of land inequality, according to what we are going to choose as our benchmark metric in this paper, is going to contradict existing patterns from agricultural census. So if anything, I think this is one of the main point that is that we need to have a discussion around how we are measuring land inequality because it's going to give us different pictures of how the different countries are standing in the world. Okay, so this is the first result that I was showing you. This is the land area inequality using genetic coefficients from census data. So this is the using kind of the existing literature methodology and this is our estimate and what you can see is that the correlation between the two is pretty much very high. The R-square is around 0.88 here. Going further, so this is departing from the Gini estimates we're now using the top land shares. So this is the top 10%, so sorry, this is the top land shares among land owners and in terms of area and then in terms of value. So this is telling you how much in different countries when you account for value instead of looking at area, what's the level of inequality? So what we can see here is that you're going to see different patterns in different countries. The African countries, most of them, if you look at the land value inequality instead of area, it's going to increase our perception of inequality. On the other hand, if you look at Ecuador, Guatemala and some of the Latin American countries, what happens is that there seems to be big plots of land that has a very low value. This might be plots of land that are useful, pastures and so on. So there seems to be something going on that if we only focus on land area inequality, we might think that it's much higher in the Latin American countries, but actually looking at the value of that land gives us additional important information. This is in terms of accounting for the land less. So I have a table in the paper that shows you the different proportions of the land less population. Of course, this by default is going to increase in equality, but what I want to show you here is the fact that it's going to increase in equality to different extents in the different parts of the world. So if you account for, I think the most, I mean, the ones that you might be looking at is China and Vietnam that has a completely different regime in terms of land ownership. That's why accounting for the land less is not going to matter in China and Vietnam. But then if you look at African countries or some part of the Latin American countries, you see that this has a massive increase. So I just want to kind of, this is like the final result slide if you want to in this paper is a combination of all the three parts that I've just explained to you. So this is the first green bar is going to be area inequality. The blue bar is value, land value inequality, and then the red bar is value accounting for the land less. So what I want to show you in this graph is the fact that if you only focus on land area inequality, which is what we have done in the past, we're going to have a certain ranking of these different regions of the world. But then if you move to the last estimate, which is our preferred estimate, so including value and accounting for the land less, you can see that these different pictures are going to change. So now China and Vietnam is going to be among the lowest unequal countries, and that's going to increase a lot more in Latin America. And it's going to basically switch the ranking of China and Vietnam with respect to the remaining countries. Okay, so I want to conclude here by saying that it seems like that land ownership inequality is not exactly the same as land holding inequality, and accounting for the difference in land values and including the land less is going to make a difference. And even though we have a lot of things that we still need to do in this paper, there's a lot of analysis in terms of income distribution and so on that we want to take into account from now on. But I want to say that even if it has some limitations, I think if anything, it brings a valuable discussion to the table, that is to say that we really need to understand what we're measuring and try to reflect on whether the existing literature has been doing a good job in estimating land inequality. Thank you so much for your attention.