 So I thought this was really interesting. I have a feeling you're conflating risk with adverse circumstances in the sense that suppose that you knew for sure there's going to be five more weeks with high temperature. Wouldn't the same thing happen? So is it really risk and coping with risk or is it adverse circumstances? And the adverse circumstances, the evidence you show about remittances and the correlation with the fraction of people who had immigrated before, that doesn't tell you that it's risk sharing. It could again be income support. And I do think it's really important to not say risk if really you mean income. And then the other thing that you didn't touch upon was what about switching to crops that are more heat resistant? And do you see any evidence on that? And if that is possible, maybe that would be a good source of policy, you know, policy reforms. Do I just do all the questions first? Yeah, yeah, perfect. Yeah, great. I actually have two, two. First of all, super interesting. I've seen this paper presented before, but I love how you guys are able to really pin down exactly the mechanism. What's going on? Very, very interesting. First question is on property rights in El Salvador. I wonder if part of the effect could be driven, especially on the discrepancy between the landless and the landowners. I'm thinking of the paper, like the Amaric, the Genvry and Sadole paper in Mexico that shows I want to give farmers titles to their land. That's when they're more likely to migrate. So they actually want to own land. They don't want to migrate because it would take, like the property rights aren't secure enough for them to do so. So then you see actually the landless one migrating more because they don't have that challenge. So that would be good to clarify. And then of course myself, I'm always curious about the interaction. I'm sure you've run these things just on rainfall. I had expected rainfall to be a big one, but and also the interaction of these two, right? I've kind of heard that like often it's not just the fact that you would have very hot days, but that the hot days that remain dry as well is where the issue was, I believe. But again, you've fully run these things. Thank you. Very interesting. Thanks a lot. So suppose there's significant heterogeneity in our willingness to migrate. So each of us, for each of us there is a threshold, an individual threshold, on the size of the shock that would lead us to migrate, right? And then what could be the case is that people in those regions in which there are tons of migrants, I mean, many people left, maybe the issue is that all those with low thresholds are not there anymore. So the fact that you then observe low migration, low additional migration in those regions might not be due to the fact that they're better insured, but that the population itself is less willing to migrate. I mean, the migrant pool has been depleted. So my question is how do you, I mean, could there be a way to address that concern? Thank you. Very interesting paper. So there are lots of points, but very quickly, why two standard deviations? So you give the choice, but you don't explain it. Do you look only at the adjustment on labor? If the shock is persistent, maybe capital can also change if you have the information. And my third point is similar to the colleague that talked about income. I think, yeah, it's probably income heterogeneity rather than, and what maybe you can look at even if we exclude remittances. If you have a small farm and the big farm, it's not the same. So I don't know if you look at farmers by quantile or by whatever. And my last point is, you said you don't look at the shock for a different, for a long period, but I think if the shock is persistent, maybe two, three, four years, maybe this could push them more to migrate rather than if they know if it's just one year. And thank you. Thank you so much. I think you are like touching all the points we are working on. So let me just go back to first, Anna Maria, please help me with anything that I forget or that I don't answer complete. The persistence of the effects, this is so important, right? Particularly because if you think this international migration, this is very costly, is one shock going to be enough to push you to migrate internationally again with everything that is happening in the U.S. So yes, I think that is very important to address and we don't have it right now, but now we are, we just run those regressions where we are doing exactly what you are saying of we have the temperature shock in T-1 and we are controlling also for the accumulated shock T-2 to T-5. When we have, when we run that regression, we do find that those accumulated effects that also matters, but it doesn't really change the magnitude of T-1. So it also matters, right? Like definitely what happened two or three years ago also matters, but when we compare those magnitudes, it's really what is driving the effect is what you just perceived the year before. So that is, that is one. And I think again, it's because of the context. These are very small farmers that are like, they rely on that crop so much. Capital effects, yes, that is great and the agricultural survey actually is really good. We have explored the different things, but yes, we should look at again what we have like the stock of capital. What we see is just like what is happening with the inputs that they can manage with, but for more like medium term effects, we can see what is happening with the capital. So I can, we can do that. Yeah, Tomás, I think that is a very good point and I would, I want to give more time to the other presentations, but I would love to talk more about it with you. I need to be honest. I'm the one that is like married to those results because I think, yeah, you just have like, it has a very nice story, but we cannot, honestly, I just don't think we can identify it well because of all the problems that you are raising and Raquel and you raise as well. Is this really access to rescoping mechanisms or are these just like regions that are very different? These are regions that are just richer in other ways, right? You have more migrants and because of that you have better income and that's why you are staying is really now no access to that. But I would love to talk more about it if you can think about like any way of really exploiting these effects. I accept this is right now, I think one of the weak parts from like empirical point of view. It's not well identified. There are many things happening at the same time. Yes, rainfall. Yeah, again, I would love to talk more with you about it. We do control for rainfall and I went super fast through that, but we control for rainfall and actually we have been debating, is this a whole like weather shocks paper or is this temperature shock? What we see and what we have seen like later in the literature is that temperature is more of a shock, at least with rainfall they could store it and it's not, it doesn't have such a perfect prediction on corn production whereas temperature has. So we are following that literature and we do control for rainfall every time, but we don't see a clear pattern with rainfall shocks whereas with temperature we do. And I think the property rights is very important and now we are actually playing more with it and now we have like completely people that have no access to land at all, people that are renting land, people that really are the landowners and we have important to generate it there and we are, we were just working yesterday and trying to understand those results, but I think it's very important, particularly if we think about policy implications. Yes, and other agricultural sectors, yes, we were thinking about it and we see no movement at all. We see that they are not, we were surprised by it and we were surprised by not seeing that they go to the non-agricultural sector, but we see no evidence about like movement to these other sectors. But thank you so much for all your suggestions. We are rewriting these papers, all of these is very, very useful. Thank you all. About the increase in reporting crime to the police, I think, so you said it's conditional of, there are surveys, opinion surveys where people are asked if they like living close to a migrant or not, so we've used this in one of the papers on integration and maybe you can use it to check if the people who report or not are correspond to these places where there are high levels or low levels of people hating migrants or who dislike living close to migrants. These are super interesting results and I was wondering if you could also, I'm not sure, but I think maybe the IFLS has a module on victimization and fear, and fear of being victimized, so I don't know if you could see if not only they are being victimized, they are saying more or less like the victimization but whether they are reporting more fear. I think the Mexican Family Life Survey follows the module of the IFLS and they have a whole module of fear of victimization that could be interesting to explore. My question is about the instrument and how can it be explained by state presence in the sense of a rainfall shock is way more important in the area when the state doesn't do enough investment in the handling of that rainfall. So could that be maybe tampering with the actual effect of rainfall? To the instrument as well, I think that the effect of rainfall and migration might not be monotonic because like you have the monsoons in Indonesia that like you have low rainfall, you have more migration, but if you have too much rainfall, you have more migration as well, so do you deal with that trying to make maybe dummies on rainfall for instruments? Mine may not be really a concern very much, I'm not sure. I was wondering about the exclusion restriction. I don't know what's grown in Indonesia particularly, but I was worrying about the extent to which rainfall shocks are correlated across different areas, they're migrating out of those areas and therefore you would have a terms of trade, basically in terms of trade shock within the country. Just very quickly, very quickly, is there a way to reconcile the two results that you have like a quick back of envelope calculation to say what's the share of the effect that you find that's due to reporting, more reporting in a sense. One question that I don't know if you will be able to see it, but we don't know where the victimization, the crime is increasing, why is it increasing? Is it increasing because they arrive to the destination cities and migrants get involved in crime or because migrants are more victimized? Here in Colombia what we have found, a paper by Ana Maria Trivina and Brian, I don't remember his last name, he finds that the ones who were victimized were the migrants. So crimes reports were increasing but it was the migrants. The other thing that I wanted also to raise is what you were mentioning about the reports, the newspaper reports. There is a very nice paper for Germany where they see what happens when a newspaper changed their policy such that now they reported the nationality of the migrants and instead of increasing xenophobia it decreases because then people knew, so I don't know if with the newspaper report you are able to see whether they are reporting that the crimes were committed by migrants. This is more difficult because these are internal migrants but it would be interesting to see whether that is the case. You have more newspapers reporting that crimes when they are committed by migrants, like highlighting that it was migrants. Thank you so much, I love all these comments and I'll be around in case anybody else wants to talk after us as well. Just because it's fresh in my mind, let me start with this last discussion with an unfortunate note that we know nothing about the perpetrator, nothing, there's no information about the perpetrator. That is unfortunate, keep in mind that we are not scraping these newspaper articles, this is existing data set that's been collected by various folks, various folks in the Indonesian government, World Bank, various groups, so the data is what the data is. So we don't know much about anything about the perpetrator. The victim, however, and Maria we do, from the household survey, so yeah, that's definitely on our list and that's why I'm presenting that household survey results a little bit very preliminary, there's just so many more things, you know, migrant origin, gender I can imagine being interested in, skill level like I can imagine of various heterogeneity being really interesting. So we'll definitely do that. Now let me go back in order, yes I'm very familiar with the data you're describing, I have a very old paper using the World Value Survey which I think is the main survey I had in mind and I'm actually going to show a slide from that very old paper, if you could click please on this little attitude button here, it's going to pop up, there's actually three of four main questions from this World Value Survey, only got this one on the slides, one is do you want to live next to a migrant? This is the one, if jobs are scarce, employers should give priority to natives over migrants and it asks do you disagree, neither or agree. This was done for a human development report so we don't have by GDP what we have done by the Human Development Index and one thing that we actually see in all countries a vast majority of people think that priority should be given to natives but actually especially in these medium income countries, the attitudes towards migration actually most negative which we explain in that paper, it's a paper with Jenny Klugemann which we explained by this idea that well in these very high HDI countries sure we hear a lot about negative attitudes towards migrants but at the end of the day there's social safety nets. People don't have to worry about feeding their family as much, there's certainly some of that worry and then when we go here the levels of migration are really small so again you know there's not so many negative attitudes about these yet, it's really in these middle income countries, medium HDI countries where we see a lot of this competition going on. I love your idea to look more into this, the one constraint that I have I think I don't think the World Value Survey has any geographical information other than the national level from what I remember. So I don't think, I love your idea to like sort of look geographically, do you think I'm wrong or? I don't remember which survey but the survey was used and they were, did they do the survey? Oh I would love to use that, thank you so much. So the Gallup and if you have, maybe we can talk afterwards, I'd love to look into this. I'll give you the name of the survey. Please, please don't leave without, I want to make sure because that would be super interesting to kind of like pin that down a little bit better. Let's go back by the way and let me make sure. What else did we have? My of course didn't take, oh yeah so okay I actually have one other thing I want to show. It's a little bit about these questions about the instrument, exclusion restriction etc. So if you could please go back and I'm going to ask you one other thing, I'm going to ask you to please put on a little button, sorry, the button of the testing exclusion restriction because I think it's going to come exactly to some of the words that people had which I think are very valid, particularly on trade and here I have to put a disclaimer that the results I'm going to show you are actually based on my earlier paper with Jeremy Magruder but the first stage is the same so it should be the same result but we just haven't run it yet for this paper. So what do we test? I actually think it's a fairly powerful test. If there was something else like trade patterns that could explain, you know that would lead to sort of similar variation then we would expect that if we use any kind of different weighting scheme rather than the half, quarter, quarter but there's some other weighting scheme maybe that would just produce similar results. So let's test it. Let's just bootstrap our weights a thousand, ten thousand times and let's just see how often do we see results similar to the ones that we have appearing and because if it were the case that our migrant rainfall measure is just picking up on something else, trade, whatever it is then we would expect that a lot of these other weighting schemes which has happened to also lead to very similar predictive power. So we run that here first on the first stage. Our first estimate, our first stage estimate is right here and as you see it's even cut here in the horizontal axis. None of our bootstraps weighting schemes comes even close to it. So what this is telling me is that there's something about specifically where these migrants were coming from. Not trade, not anything else. It's specifically where these migrants or any, could be anything, right? Any kind of correlated productivity, gravity model of areas being particularly close together, none of that is happening. Same with a reduced form and then the last thing that I'll mention is we look at long distance migration as well. So rather than having just any move we can say you need to move over at least 100, 200 kilometers. And if anything we actually see that migration over longer distances has stronger effects, well if there was any kind of trade or anything going on there wouldn't be really a reason to believe that any kind of gravity model would predict that trade is greater if locations are smaller together. Does that make sense? Can I just say that we're going to talk about in terms of trade improvement, i.e. are richer than before. Maybe that richness or that inequality in that area is aggravated, where they're going to. Sure. And that's what's leading to more harm. That is different than what's going on in migration. Yeah, yeah, sure. I mean, it's more like a mechanism of why, yeah. In terms of trade to be growing the price, that was my concern. No, I think it's, well, I'm always a glass half full person. I don't think that's a concern. I think that would be just a way that we could explain our results. I don't know how I would use that to explain the pattern that we see results in the newspapers, but not on the household survey and on crime, that I couldn't explain that. But I do agree that like whatever, and that's why I actually, like the previous paper presents so much, it's sort of very carefully laying out every little step of the causal chain, so to say. All that we have is, there's migrants coming in, there's effects on crime. That's kind of where we're at. I may have more questions, but I may be out of. Okay, let me see what else I had. I have, oh yeah, so actually, yeah, your question, Andrea. I have to admit, I've worked with the IFLS for many years. I do not know of this, of course. I know you have a lot of experience with the AMACS FLS. I will search for this module. I hope the module exists. There is one crime module, IFLS 4, but it was actually so bad that they discontinued it. So the IFLS 4 crime module we have at some point looked into, but it's just not a good module at all. But you seem to be talking about some kind of fear or something like that, right? So, not that I, but the IFLS in general, I love that, and the AMACS FLS too, and I love the people who've worked on it, these are such great surveys. They have people sit down on average two days to answer questions, like these are, we know everything about these people. Monetonicity, again, kind of coming back, like what is the exact weather term. We try a couple of other things. Again, I'm not going to show more bonus slides, but for sure, a lot of rainfall can be bad. We actually don't see evidence of that in the Indonesian context, and that's mostly because it's a lot of rice growing. I have no doubt that it's true, right? I mean, there's floods, et cetera. But for any of the kind of variation that we're looking at, that was never a negative thing. And in fact, David Levine and Dean Yang have a paper that actually does the very similar job as what they're doing in the beginning, where they look at the effect of rainfall in agriculture, like they combine it in a paper. I just cite Levine and Yang, like that basically said, it's a monotonic linear relationship. Use that. I'm a little lazy. All right, any other questions? I'm happy to talk more after. Thank you so much for your feedback. And Maria was wondering whether if there are any other benefits such as tax collection, net of transfers to the migrants. So I think it would be important to see the whole benefits to society at the destination rather than just the benefits for the migrants, yeah. Sure, thank you very much for the presentation. I just maybe have just a question of clarity. I don't know if you might have mentioned it. But I'm quite interested in after the regularization of the emigratory status. One, was that time bound? And then secondly, I guess if it was time bound, what then subsequently happened? Because I think in many places, you know, in the world there's big question marks around specific amnesties or dispensations and what subsequently follows. The second one that I'm quite interested in is your earlier comment. And I'd love to just hear your reflections on this. Syria, Venezuela, and Ukraine. Now if I look at all three of those places, I mean, you know, two of them conflict, right? Syria, Ukraine. But Venezuela, and I would probably add another country, there's Zimbabwe, are direct outcomes of economic sanctions. And effectively, if you think about Zimbabwe, a large part of the working age population is outside of Zimbabwe, so I'd be interested to hear just some of your own reflections whether or not the cause of the displacement is also important. So if the displacement is caused by conflict, should the policy response be different if it's caused, for instance, by say economic sanctions? I have a quick question. I have a quick question, if I may. I'm just curious if you get information on the Colombian side, kind of, if when migrants were more incorporated and they feel like more welcoming, do people actually, like more welcoming, but rely more on the state? Do they also rely more on the Colombian counterparts in the municipalities, like the Colombians, receive them? Well, probably it's beyond the scope of your work, but I don't know if you get something of that. Thank you. Thank you. I was wondering as you finished, and I was talking to another colleague here about that, the sense of, I mean, for integration, we need to, both sides. And this is one side, and I think we could find something interesting here. I was wondering if the qualitative data or in the service, you asked about first the sense of belonging, or sharing values with Colombians, if that changed, or if you have anything similar in the questions. And second, how much exposure to Colombians was, for example, when you are not regularized, you interact more with Venezuelans and with, and stay in the neighborhood and stay there. And on the other side, when you feel that you are integrated, whatever they perceive as integration, you will talk more to Colombians and you will feel empowered to talk and interact more with them. My question regards maybe, maybe it's beyond this scope of the paper, but I would like to know, and that's a question that came up to me, is that maybe this program had some spillover effects over the access and the wellbeing of other Colombians. In the sense that I know that this program it was financed by the national government. So many institutions and municipalities and departments gained new resources for attending these populations. So maybe that had some spillovers in access and quality to Colombians. Yeah, I was surprised to see that take-up of that PAP is actually quite low. So I wonder if you could at least tell us like what predicts a take-up. I guess you can do that in your data just to kind of like look at who are the people who take up and who don't. And the other thing I'm just really curious about being from here, why did the government do this? Like I just love to hear the motivation behind this because this is something that we don't usually see. Just a very basic question, maybe I didn't understand but do you assume that all the people that arrived after the closure would all apply while we've seen that for the others the share was low so I may have missed something. Just to finish, I wanted to build on that. And can you check like if the migrants that came after round close, they are different because I don't know if they hear that they are opening this registration, they might think this is like to help them somewhat and the people that might come could be different. I'm going, I have some minutes. Okay, thank you very much for all the questions and for all the suggestions. Rerata about tax collection, yes, this is very important. Here it's a very short term effect but I didn't show that but you can see it somehow when you compare subsidized versus contributive health system because in the contributive one you are paying taxes. And we have not looked at that but we can do back of the envelope calculations for that. And when they get formalized they pay taxes. So that's important. So in the new version that when we are doing at the second wave we are going to start to look into all those issues. Something that we're finding for example for the second wave which is very interesting is that fertility rates dropped for people, for the woman. They decide to have less kids. So that's important because then you have, the state is going to have less kids to take care of and to provide social services. But yes, we still have to do that and that's something that is very important to do because migrants contribute in many ways and one of those ways is to start paying taxes when they become formal and that's important. What happened after? That's a tricky question here because they could request an extension after two years but when those two years were going to end the program did the new one, the new PEP which is called ETPB. So now they can stay for 10 years. So what we need to know is what's going to happen with this 10 year program. What we are finding that is very important with respect to what you are asking is that we ask whether they want to return and when do they want to return? We ask that question in the survey and we don't find that the program changes that. So willingness to return is very similar to the regular and the regular migrants and we really need to understand that a little bit more. About whether the effects are different whether there are economic sanctions or conflict. I really, we don't have enough data to see whether the results can be different or not. My perception, I have worked for many, many years here in Colombia with forced migrants, internal migrants and what is very important about all these processes of migration is that these processes are quite hasty. They are not necessarily completely voluntary because the conditions are so bad that they need to migrate and they do it in very poor conditions and in very difficult conditions. So regardless of what happened, what I think is more important is the migration process itself. That it was hasty, that you didn't have funds to fund the migration and that the persons were in quite vulnerable conditions but that's just a hypothesis of that. The Colombian side, we have surveys for that. Colombians, Senafobia in Colombia, Peruvian Ecuador who were the countries that received more migrants increased significantly and when we look at the surveys for that, the spike after some time increases a lot. But and very interestingly, so when we ask them, that's a survey that we just did, whether they are willing to provide entrants to migrants, whether limited or not, people are willing to do that and they are also willing to provide state services. So they don't trust migrants, they think that migrants are hurting the economy but on the other hand, they are open to provide social services and to allow them to enter the country. Which is, I don't understand very well why. Yes, it's kind of contradictory but we have to understand it a little better. Sandra, your question is very important. We have questions of that in the survey. We didn't show it here but when we were doing the focus groups, we had focus groups for those that were PEP and those that were not and the difference between the two groups was really stark because the PEP ones were much better. They had Colombian friends and the other ones were completely over the wearer. They were really, they cried in some focus groups. They were all crying. They were very pessimistic. So we do believe that there is a huge impact on integration. In the second wave, we include better questions because in the first wave, we include some questions and I think the questions were not, we are not capturing that but we do believe that that's a very interesting avenue for research. A spillover effect. I'm not sure that there are spillover effects. There is a study, a paper by a student of mine and he finds that municipalities have to spend more for migrants and the tax effect has not kicked in yet and social services sometimes they feel are overwhelmed. So that's kind of, you have both sides. Why is 64, that's a huge mystery, Mike because basically what happens is that the process to get into the registration is very easy. It was online and for the TPE, it was the same thing and the take up, even though people already knew what was happening and the benefits of this, it was at the beginning it was quite low. So there are problems of people that left to Ecuador and to other countries but also problems of people not being able to register and we have information about that. I should have brought it and we asked those questions in the survey. The motivation, the person who did that is at the IDB right now and we are going to organize a lunch with him so he can tell us that. So if you want to come, it's going to be a closed door lunch but we can invite you. And I think that's it because we need to close. Thank you very much.