 OK, so I'm going to talk about another large, very large police reforms about land rights, and that was in Mexico. And this is going to report on several papers, and so I have a piece of work that we've done together, with Alain de Genvry, who is here. Kyle Amaric, who is now at Taf University. Marco Gonzales Navarro, as you see Berkeley with us. Also, and Daley Kussman, who is now Northwestern, so I'm picking up on several papers and studies that we haven't done. The motivation for most of these studies, where the property rights reform come from that conventional wisdom, right or wrong, that we need sufficiently complete property rights for development. This is very important for development. So I'm using expressively sufficiently complete property rights, not necessarily full property rights, but that is security and no arbitrariness in the property rights reallocation and the change. There's also that idea that land reform, which will provide to small land owners that sort of property rights sufficiently secure, can be the source of agricultural growth, at least in that sector, and property reduction. And yet, and Alain just mentioned that in the introduction, most land reform that we've seen throughout the world have given excessively incomplete property rights, and then not only incomplete, but very insecure in terms of not only with restriction, but very insecure in terms of whether the rules can change over time, whether there is arbitrariness throughout the process. Now, in the case of Mexico, we have just observed a huge, relatively recent land reform, which gives us an opportunity to see how a major change and increase in security of property rights has an effect. So I'm going to present to you the result of three studies that we did on the impact of that Mexican property rights reform. Very briefly, I'm going to recall the experience of Mexico for that land reform, talk a little bit about the empirical strategy that we use in all these studies. It's the same in the data that we have. And then look at three issues, labor reallocation and migration, something on land use, and finally some political consequences of the land reform. So briefly about the reminder of what's been the experience of Mexico with land reform. The first land reform started pretty quickly after the revolution in Mexico, and then all the way to 1992. And it created a massive land reform, redistributive land reform, creating 32,000 new agrarian communities, which we are called heridos, or indigenous communities in some cases. It's huge in the sense that it gave land to 3.5 million households. And it covered more than half of the territory in Mexico. So this is at scale, which I think is really important to look at. Now in those heridos, and I'm going to focus on the heridos more than the indigenous communities, which had different property rights. In the herido, there is, in addition to the housing plots, there's two types of land. Each household has individual plots, which are agricultural plots meant to be for production, agricultural production, which were held in user-fruct. And then pastures and forests, which represent a large share of the land, which is held in the heridos, were held in common property. Now the reform give, and Alain would say purposefully, I don't know, but the reform will give very incomplete property rights to the beneficiaries. And the idea was, I think it's been documented as an instrument of political control. On the individual plots, for example, which is under the responsibility of the household, there was a requirement that there would be direct cultivation. So you are not allowed to rent the land out. You're not even allowed to hire the workers. It has to be, and that was the relatively small plot, it has to be cultivated by the household and strictly by the household, and no renting out. There was also a requirement of continuous cultivation. If you let your land idle for more than two years, then it could be taken out, or a few could be taken out and given to somebody else. Use it or lose it. And then the land could be only inherited by one child. And the benefit of that is that there was not that divisions to tiny little plot, but it also means that nobody could be incorporated in those heridos, which also created a lot of marginal people around the heridos as well. For the part which is in common property, all the pastures, which are important, and the forest, there was no formal sharing or rules of sharing of the benefits. So in a way, some heridos were well-managed and could be real common property resources. In many cases, we see extensive over extraction. And in many cases, something that looks very much like the tragedy of the common with encroachment and over extraction. The consequences of the first land reform, I think by many of the political economic writers, is that it was very effective in term of political control. The votes were delivered to the party bosses. And particularly, it's very easy to observe the votes of the heridos, because they all vote in the same electoral sections. So you get a very clear idea of what percentage of the heridos voted for you. And it had the pre, which is the party that did the first reform, remaining power throughout that period until 2000. But seen from the economic point of view, I think this political control was achieved with a very high efficiency and welfare cost. There is clearly excessive labor in agriculture in Mexico. When you look at where Mexico is on the curve of the share of the population, which is in rural areas, relative to GDP per capita, Mexico is way above the overall average of the countries at their level of GDP per capita. There's a study by Melisa Del, a couple of years back, that shows that even now, we can see that the municipalities, which had more heridos relative to the others, lagged in industrialization. And then we have documented that there is extensive poverty in the herido sector. So this is a high cost, and that sort of led probably to the second land reform. So then in 1992, the President Salinas, and this was probably in the context of upcoming NAFTA, the Negotiation of Agreement with the US, and also the entry of the membership of Mexico to the OSCD, wanted to attempt at completing the land reform and giving full property rights to the hereditary. So this program, called PROCEDE, which is a program of certification of land plots and assignment of corporate shares for the common property rights to the heridos member. The certificate for the individual plots, it means that the owners now could freely use the land or not use it, if they choose not to. They would not lose it for no matter. They could hire labor. They could rent out. So there was separation of ownership and land use. Now they had a completely choice on land use. They could also sell the certificate to other members of the communities with an assembly approval. They could not sell it outside. They had to before obtain a full property right, full title, which is called Domino Pleno, and to have unrestricted sale. And actually not many of them, very few of them, have requested that Domino Pleno sell now. On the CPR, on the forest, and the pasture, then the decisions of the shares were given to all the hereditors. But the decision was in the hand of the community as a whole. The heredo has a role. So they could decide now to distribute the land if they choose to incorporate new hereditarios. They could decide the conversion of land use, which were not allowed before. Before it was very restricted. And they could decide at the heredo level how they wanted to dispose of the land, except, of course, under the minimum restrictions that are in the forest, Trilo, in Mexico. The rollout of Procedé took only 13 years. It was absolutely remarkable. It was fast. It was huge. It was orderly. I mean, this is sort of, I think, an example of a success. And I'd like to mention that every time we talk about Procedé, we have to recognize the big success when they happen. It started in 1993 with only a couple of heredos. Only in 1999, a few years afterwards, you already have 17,000 heredos, which are certified. And then by 2006, most of them are. It's 27,000 heredos. And then the land reform, the Procedé, was closed. Because the remaining heredos that were not were either indigenous community, which had more difficult property rights regimes to solve. Or the heredos, which were at very particular conflicts. So essentially, in 13 years, 27,000 heredos were certified. And it was done with a lot of discipline, resolution of the conflict, and as a group. So make sure that there was no abuse. And I think it has been really a very success story. We're going to use that rollout. So this is our empirical strategy. We're going to use this rollout to analyze the impact of the acquisition of these property rights on the heredos. So we're going to have panels of heredos. And we have a lot of observations, 27,000 panels of heredos. And match them with the localities, which are next to the heredos, to look at the population changes. Match them with the electoral section to look at the electoral result. And then we use the land use maps that we have in Mexico for looking at the evolution of land use. And those are three different papers. Now, the validation of the approach, we are pretty seriously looking at parallel trends and also a lot of robustness check. But I'm dispensive of that. So in terms of the data, we're going to use, and I'm sure you do the result afterwards, for the migration, we are going to use two types of data, either the progress data, and this is a panel over three years. This is the first wave of the progress data, 97 to 2000. So it's a small data set. It's only 7,600 households. In only a small number of heredos, but it has the advantage that we have a lot of detail because we have a full household survey. At the other extreme, we're going to look at the population census. So that's covered the whole of Mexico looking at population census, but this is only two years, 90, which was before, and 2000, which is when we are about two-thirds into the project. Landsat data, which have been interpreted by the Ministry of Agriculture in Mexico. And then use also the data from ProCampo. So this is a subsidy program in Mexico, and Mexico takes for the crop a compensation for NAFTA, and Mexico takes the transparency very seriously, and you can find on the web the list of every single farmer who received ProCampo, which is all the farmers in Mexico, and then you can have their name, I think, and they can have the area that they cultivate. So we can actually look at the evolution of farm size based on that data, and find the electoral results that we are going to use. So let me take those three issues at the time. Labor migration. So what we find here is that the certification led to substantial outmigration. Now in those rural areas, the background of migration is pretty high. Over the three years that we look at some progress data, 5.3% of the household had one additional migrant that left the household and the village. And we see that with the certification, 30% outmigration increased by 30%. That was in progress. If we look at the population census data, we have also, again, a very large outmigration out of these small localities. Background of 21% migration has decreased in population, so it's not like a decrease in population between 90 and 2000. And we see an increase corresponding to the certification of 4% additional points. We find some heterogeneity in that migration and which goes with the idea that a lot of the migration or a lot of the returning of the population was due to the incomplete property right. That we see that there is more migration when there were weaker property rights before. That means it has retained population more. We see also more migration. We had lower land quality and less migration when there is more quality and more migration when you have better off wage opportunity outside. In terms of land use, that's another study that we did. What do we expect to happen? Well, we could have decreased, because now you can finally leave your land idle without losing it. You can see no change if it's just there is rental and sales market, or you can see an increase, which a lot of people would say, because you have security and functioning of land market. Now, using those landsat data that we have over those three years, 93, 2002, 2007, we see in aggregate no change in cultivated land, area in agricultural product. What's really interesting is that we see actually an increase in cultivated land in the area which higher productivity as measured at the level of the municipality, and then a slight decrease in the area with lower productivity. So that suggests some efficiency gains of land use through land reallocation, land selection. And using the pro-campo data, and actually also some of the progress data, we see some consolidation of land in larger farm, and we see that the household which have either, which have better land or more land tend to migrate less. So there is also some suggestion of efficiency gains through the farmer selection, better farmer selection on larger farm. The current work is actually very ambitious, is to reconstruct the land use data on an annual basis based on landsat images and machine learning technique to interpret those landsat, to interpret those images. But it's not, we have pretty many results, but I won't talk about that. Political consequences that we can observe. So as I mentioned to you, it was President Salinas under the pre, which initiated the reform. Who benefited politically from the reform? So we looked at the election result, and every three years there is election for the Congress in Mexico. So we have during this period, 94, 7, 2000, 2003. We matched the electoral section to the localities and to the eridos, and so we have 19,000 electoral section that we can observe over days. And what we're going to be looking at is that whether the electoral result that we see, the share that goes to a party, is influenced by the share of the population of that locality, which is associated with an eridore which received a title, okay? So it's not, so there's sort of two shares now. There's a share of the population which is in eridos, but there's also the share which is, which received the certification. Now let me interpret that figure, and I think that's going to be the main result in that. So what we see here, what we have represented here, a very simplified way, the result that we have. So we have here the four election, and then we have a select a group of eridos in each of these bars. We always have the left one is those eridos which have been just certified before the election, and on the right, the eridos which will be certified immediately after, so sort of discontinuity here. And then we follow these, so now they were certified before that second election, and those are the future, those are going to be certified afterwards, and the same as here, okay? So a couple of things to look at on that graph. First, there's a trending increase in the vote share for the, this is the vote share for the panel, which is the right, the party to the right, and not the pre which have given the other. So first we see an overall increase in the share to panel, which is the trend that happened during this period. You see that both looking at the overall trend across all the different groups, and you see also in each, between each election, you see on the group of eridos particularly. So that's a larger, but what you see also which is really interesting is that the difference that we see at each, in each election between those who just got the property right and those who don't yet have, but are going to have them in the next two years. And you see every time, every time those who just got the property right voted a bit more in favor of panel, 1.2 to 2.9%. And that's an average 7% increase in time. Okay, so we see that, I'm going to skip that, and I guess I have to conclude after Anna was saying. And there's an interesting discussion about whether that, let me just take a second to do that, whether that was an error or not. How come the pre gave those property rights and those eridarios had so little gratitude that did not even for the pre? One argument which I think is a very strong argument for that is that the, in fact, when you do a one-time transfer, there's no need to have gratitude. You need gratitude to government only when you have a recurrent transfer to make sure that it comes back. So actually I think it was to be expected in a way. Okay, let me, this is my concluding slide. The reform that we see induced large outmigration of labor and population, concentration of agriculture, concentration of production in larger farms. So probably some efficiency gain due to both labor and land reallocation. It was at the same time a huge transfer in terms of wealth to the population. So it's likely some equity gains although we didn't measure much. Okay, but it was politically costly. Also something we don't mention that here at all but we have to be careful is that this large outmigration mean that we need, I mean it can only be good if there is sufficiently, I can create unemployment, it can only have positive effect if we have the capacity, the labor absorption capacity in the country for them. So it's not all rosy. Thank you.