 Very good, thank you. So let me take on the issue of property rights. And we know that the property rights are important. We know that property rights tend to eventually provide security of access, induce investment, induce gains in productivity. And we also know that property rights sort of underlie markets and the rental of land. And hence, eventually lead to a reallocation of land to better users than the ones that we are currently on the land. So many countries have gone into large certification and titling programs. Mexico is one, we are going to take this as a case study. Ethiopia, as was just mentioned, has initiated an important process of certification. I think the jury is still out as to how much change is going to follow. Countries like China are still debating as to whether they are going to get or not into issues of formal certification. Vietnam has made progress on to this. And we hopefully can discuss this in the discussion that follows. The main change here is to go from a situation where land rights are based on use, is by using the land that you establish your rights. As a consequence, it's the behavior that you develop, which means the land use and the labor allocation to the land that determines your rights. In shifting to a situation where the land rights are based on certification and titling, then you go from the rights to behavior. So the question that we want to ask here is how different is behavior from a system which is use-based where use determines behavior from a system where rights are going to determine behavior? And many countries being still in a situation where use determines rights, when we shift to a situation of certification and titling where rights are going to determine use and behavior, then we want to know how much of a change in behavior can we expect, and whether those changes are going to be efficient, and whether those changes are going to have social and political implications, which actually can be quite large. So there are four specific questions that I want to ask out of this. The first one is how much labor reallocation is going to happen once you let people do what they want as opposed to conditioning access to land based on use, especially how much migration out of agriculture is there going to be? Second, heterogeneity, who is going to live, who is going to stay? And can we infer out of these impacts in terms of equity and in terms of gainers versus losers? Third, in terms of land use, is there going to an impact on land cultivated and on the way in which land is being used? And finally, in terms of efficiency, what we have, especially in the case of Mexico, is a coexistence between a private sector and a sector that was in the land reform sector. As you change property rights in the land reform sector and make them more analogous to the property rights in the private sector, is there going to be convergence in patterns of land use away from the former patterns towards the ones that we see in the private sector? So let me give you very briefly the history of land reform. The Mexican land reform happened in two stages. There was a revolution in 1910. There was a land reform that started in 1917 and kept on going until 1992, where land was expropriated from large users, was distributed to small holders under the form of communities, those communities in Spanish are colegidos. The land rights in those communities were quite incomplete. And it's quite important to understand that property rights were actually given for the purpose of political achievements, namely providing access to the land to many people on small plots of lands, preventing them from trading land, obliging them to use the land, but also achieving political control in the sense that those farmers with incomplete property rights were going to be quite dependent on the state for what they would do with the land and for the state in terms of welfare. And as a consequence, they had a very captive audience in terms of political expression and votes. And I'll come back to this at the end. So the condition of access to land was use it or lose it. Namely, the condition was that you had to cultivate the land yourself with family labor. And if you would not do so and leave the land idle for a maximum of two years, then land would be relocated to others. There was a list of candidates in waiting to get access to the land. And if you would not meet the use it or lose it conditions, land would be relocated to someone else. Transactions were not allowed. And in the end, this was a very large reform process that affected half of the Mexican territory and something like 3 and 1 half million farmers. So in 1992, in the context of NAFTA coming up and the need to modernize agriculture, achieve productivity gains, be more competitive in an open border situation with the United States, there was a reform of the Constitution. The Land Reform Act was revised. And there was a land reform whereby property rights were not given outright, but were first granted out under the form of certificates that could subsequently be transformed into titles as the recipients of certificates would so wish. This certification program was called Prosede. It gives certificate of ownership to land plots and a share of common property resources. Intervention was at the erido level, in the sense that the whole erido had to agree in order for titles for certificates to be issued. The certificates then allowed land transactions but it mainly allowed full freedom of doing what you wanted to do with the land, including cultivating it or not cultivating it, deciding to migrate and leave the land follow behind. Under certification as opposed to titles, there were still some restrictions as to to whom the land could be sold, but effectively the most important was that there was full freedom of using the land as you wanted to do so, including with higher labor, including renting it out and including yourself migrating away from the community. So this is to show you the process that took on the right side, you see a cumulative distribution which is the number of communities that have been titled. There were like 27,000 communities representing more than half of the Mexican territory. All the dark spots of this map is land that has been certificated over the years, a massive land reform, in a sense a very smooth process that happened over the space of 14 years and there were conflicts of the way but in the end was able to certify the land moving away from the use it or lose it situation. So in terms of theory and again, Cauchic then challenged us to go back to theory. What we think is a situation where plots were very small and so it's normal to conceptualize use of those plots of lands as increasing returns to scale. There are fixed factors which need to be spread out over a larger area of land, but then imposing a minimum yield, imposing a minimum level of use, and hence imposing a minimum use of labor to those lands transformed it into a decreasing return to scale situation. So the upper curve, the upper curve in red is the situation under use it or lose it conditions and then once you remove this condition then you switch to the blue lower curve which is an increasing returns to scale. You read from that theory sort of two implications. One is that at given plot size which is the horizontal axis you're going to have less labor use and potentially more migration. And the second is once you are free to transact the land the way you want, if they're increasing returns to scale then there should be land consolidation into larger farms. So theory is predicting us that in fact as you move away from this use it or lose it situation towards kind of free use, then you should see a decline in labor use in agriculture, more migration out of agriculture, a reduction in the size of the rural communities and a consolidation of the land away from smaller farms into larger farms to benefit from economies of scale. So the data and I think that's interesting for those of you, Tony was asking this morning what kind of research you can do. The issue of property rights is important to study but here I think what's important to see is that we studied it fully with secondary data. And we combine a multiplicity of sources of data sources to sort of construct a story. And the art of the story is to find cumulative support of complementary support from those different data sources. Or if you want not to find any contradictions that you cannot hide under the rug that would contradict what you put forward as your main thesis, which is the one that I mentioned migration and land consolidation. So we use household panel data that came from Progresa. We use the population censuses. There are Echido censuses, which are censuses of those communities which were run in 1991 and 2007. We use satellite data to see land use patterns from which we can distinguish between especially agriculture, pastures and forest. And then we use the massive records over 45 million farmers of farm subsidies that give us the transfers to farm households but it indicates in so doing the area cultivated and the farm size. So at the end what we have is a, the analysis is going to be one of panel data where the left-hand side variable is going to be for example whether the household is going to have a migrant or not. On the right-hand side what we have is whether the farm has been certified or not or this household has been certified or not. This is a panel so we have fixed effects for the years and we have fixed effects for the Eredos. And then there are a set of time-varying characteristics that are going to be at the household or at the Echido level. So this is a typical panel analysis and the key in terms of using this rollout as an identification strategy as you know is that you have parallel trends for the communities which have not yet been titled and then once titled they eventually are going to have a kink in whatever endogenous variables we are looking at which in this case might be migration. So what we see here is that Eredos are being treated at different points in time. They are all on parallel trends until they are being treated and then suddenly we see a kink in the importance of migration after they have been titled which tells us in a sense that once we run this test that the treated communities can, or the non-treated communities can serve as counterfactual for the treated communities and that gives us a natural experiment sort of identification strategy which allows us to proceed econometrically in the estimation of the equation I showed you where certification is going to have exogeneity with respect to the outcome and we have a causal relationship. So let me just run you through the main results and the first one is on migration. So here we use the household data and what we see is that it is a large increase in migration, the likelihood that a household would have a migrant is 30% higher after certification than before. The size of the, and we also see that there's a lot of heterogeneity in who migrates. If you had weaker property rights before, you had sort of more defensive labor. You had more labor attached to the land with lower productivity because the quality of the land was less. And as a consequence, when you have certification then you're going to see that it's precisely people who had weaker property rights before who had to be present on the land who are going to respond more. Better farm opportunities, what we have just seen the whole non-farm economy, more pool effect and hence as a consequence more migration, worse land quality, there was more misallocation of labor to the land before the certification on poor land and as a consequence there is going to be more release of labor as a consequence of certification. Second, what we see turning to the population census data is an increase, there's a trend towards declining population in the rural communities. This trend is increasing. There's an additional 4% loss in rural communities population on the background decline of 21% over the decade that we analyze here. So more migration, loss of population in the communities. In the Ejido census there's a qualitative analysis, a set of questions that are being asked and one of the questions which is asked is the majority of the young people immigrating from this Ejido and we see that indeed after certification there's more, a higher percentage of those questions which are being answered positively. Then we turn to the Landsat data where we look at the impact on land use and we use the Landsat for two purposes. One is to characterize the quality of the environment and we see that there's more migration six times larger in the lower productivity areas so this kind of suggests efficiency. In the higher productivity areas the large farmer stays, the smaller farmers leave. In a sense the opening of the migration opportunities is more attractive to the smaller than to the larger farmers and then in the high productivity areas we see that both small and larger farmers stay on the land. We have implications not only on labor and labor allocation and migration but also on land use. We see that there's no aggregate impact on land cultivated but we see a relocation of cultivated land away from the low productivity areas towards the high productivity areas which again suggests efficiency. Turning to the farm subsidies data we see that there's migration, there's a loss of labor, there's no reduction in aggregate area, there's a reduction in the number of farms and as a consequence on average there's consolidation of land into the larger farms. And then finally we see that we want to ask the question how is land use being transformed as a consequence of certification and the way we do this is to use the way land is allocated in the private sector which is half of the territory compared to the way land was used in the erido sector and whether once you transform the erido via certification into a situation which is analogous to that of the private sector land use patterns converge towards the land use patterns that you observe in the private sector. So the private sector land allocation is the benchmark that we use to see whether there's convergence to this or that and what we do find is that this is indeed the case. So how do we predict land suitability where we look at how land is used in the private sector using such indicators as distance to the city, slow, painful, altitude, kind of exogenous determinants if you like and then we give a score for each of the three uses which are agriculture, pastures and forestry to every pixel if you like in the erido sector and then on average at the erido level what we see is that there's a convergence towards the land patterns that we observe in the private sector. So let me conclude reminding you in a sense that many countries are still in a situation where titles have not been granted, certificates are sometimes but rarely been given. Land use patterns determine the rights that you have over the land. At the same time we know that those are not necessarily efficient ways of using the land and hence as a consequence a temptation to move away from used based land rights to certificate and title based land rights. Think of China for example where land can still be expropriated by the community to be given to others if land is underused and land can also be expropriated for non-agricultural uses. So here what we see is that indeed the granting of certificates as opposed to a used based land rights situation is an efficient move. It reallocates large amounts of labor away from agriculture. There was too much labor held captive in agriculture other conditions of increasing returns to scale. It induces more migration so this is in a sense efficient. Induces more migration from lower land quality environments. So the lower land quality environments are going to differentially lose more population which one thing is also an efficient move. Induce more migration by smaller farmers in high productivity areas meaning that the land is going to be cultivated into larger farms given increasing returns to scale. This is efficient as well. Shift agriculture towards the better lands. No impact on aggregate land use but the displacement of agriculture away from low productivity towards high productivity areas. Concentration of production in larger farms with increasing returns to scale which is efficient as well. And then reorientation towards private sector, no. So in a sense we can say yes certification formalizing property rights creates efficiency gain. Something that we tend to know. Question is what are the trade-offs? Are there risks that can come about with this? Let me just finish with those three lines here. Number one large migration which is going to bear on the receiving areas could be intermediate towns as we have seen could be the major cities. It could be the border we are in Mexico towards the United States. So clearly there's an issue as to when do you want to run those reforms in terms of timing and the capacity of the rest of the system or your border to absorb the labor which is going to be released. And the second which is not part of the study here which is a paper that we actually had presented here in Vietnam a couple of years ago is that there has been also a political shift. Once you are titled property makes people more conservative in terms of the way they vote. And it's quite interesting that this issue of land and location was started by the left and in the end benefited to the right. So in conclusion going from use-based to rights-based to certification and titling is efficient but it can have social cost. The social cost is does migration which is going to be induced by certification happen at the time when the rest of the economic system has the absorptive capacity. And second it does have implications in terms of political economy. Namely is going to switch political support probably more towards the right. And that's a risk that the government in power has to decide to take it out. So this is the story of property rights in Mexico. Thank you very much.