 Hi. Thanks so much. It's a great panel. Sorry, I missed the beginning, though. I have a question for Joana and then Santiago. Joana, I was wondering if you could tell us a little bit more about how your data is being constructed and who's doing the reporting. This is literally new for me, but I'm working on armed group governance. And in situations where armed groups have very large territorial control, what you observe is citizens that are under this control also have very little incentives for denouncing. Where are you getting this information from and what are the incentives for the citizens to report? Santiago may be two unfair questions, but trusting police is very hard to change, even in Europe if you look at trust levels with all institutions, police is always one of the lowest. I'm slightly worried whether these are very short-term effects, very transitional, and whether you're thinking about a follow-up. And also, what are your thoughts about scaling up? Scaling up is these operations. Thanks. We take these questions and then... No, maybe we answer these questions and then we take all those questions. So, this can denounce it's super strong and real, so these are like any citizen can call, and so the key question is why citizen call this can denounce it. So it's going to use it to sit inside the security department, so they have policemen working for them and follow. So for me, they are the key intelligence unit information of the time. And they do also have a great media work, so they are constantly in the newspapers and then the media broadcasts saying, no, you see, because of this can denounce it, the police did this. Based on this can denounce a report, there was a police raid on that place. So they really work on that in order to say to people that it's worth to call them, and indeed the police do use a lot of their reports. And they have several campaigns like to get specific information about specific drug traffickers, like to seizure weapons, apprehension, and so on. So it was an institution that was created in 1995, so it over time and had spent a lot of effort to abuse its reputation. Patricia, and on our side, like we actually haven't discussed like running the survey again, but this is, I think this is something that we definitely need to do at some point. This happened between March and April this year. Maybe we can do it next year, something like that. And about the scaling, so that was the original plan also with the National Police. Maybe the only struggle that we have now is that there's been a massive change in the leadership of the police during the last month. And this actually changed a lot of our counterparts, but this is, I mean the police reform process will continue and we look forward for the police to actually scale this up because it's actually like quite, I mean the marginal cost of doing it is very low. Leopold. So thank you all for the panel, super interesting. I want just to share a confusion that I have with some of your results Santiago, and see what you think. So first of all, I'm not sure if the fact that you have found limited effects in hotspot interventions carries over to other coirantes, which are not hotspots. Maybe the nature of the interaction is quite different. And given that, I'm a bit surprised because you find definitely a first stage in terms of interactions. There are more interactions, but according to reports of the police, not a take up on better procedural justice. Yes, the citizens perceive legitimacy, if I understood correctly, but the policemen are saying that they are not behaving as they are supposed to be behaving. So could it be that they are not behaving that way because they find that way of behavior costly, difficult, but they are interacting more and that's giving trust to the citizens. And they think, as we are doing it, it's fine. I'm not doing what they are telling me, but yes, I'm having to see the citizens more and actually they are more trustful. So that's one possibility. I don't know if that makes no sense at all in the context, but I'm just thinking of that possibility. Or the other is these guys, more than what I was saying, these guys are kind of using the procedural justice thing only that they are now aware of all the things that you should really do when you are doing that. So they are doing something better than what they did before, but they are aware that they are not doing as good as they should, so to speak. So I wonder what is your interpretation of that if the first hypothesis or alternative hypothesis makes any sense, but super interesting in any case. Yeah, no, thank you very much. I think, I mean, I definitely agree with you that we don't have a strong argument to conclude that this is only explained because of this procedural justice, like principles. We believe that it's like the evidence that we have is some sort or to some degree like suggestive that it's not only the increase in directions, but yeah, I think like this is something that we definitely need to explore more because I actually don't have like a good counterpoint to say it's not this hypothesis that you're raising. I think, I mean, I think this is something that we need to explore more, maybe looking at like heterogeneous effects based on maybe, for instance, baseline crime levels, stuff like that that could help us tie these to the hotspots, policing interventions that were run before and so on. So, yeah, thanks. Thank you. Actually, first of all, great session. And I, but it's about Santiago and I think you were essentially just touched on what I was going to suggest. I think, you know, more heterogeneity analysis about who is, who is it that's changing their mind? Once you start very pessimistic about the police, will you change your mind? Or is it people who are like in the middle? That would be sort of super interesting. And then two other things, and you know, related to baseline crime or other personal or read, you know, local area level interactions. One is sort of going back to Patricia's question as well. You know, it would be interesting to see it to like there are two dimensions of persistence here. One is the police behavior. And then the other is the citizen's perception and, you know, trust, etc. So I would be very curious whether the police side is actually persistent. And it may not be that some of these things, they create like a whole thorn effect. And the police do a few things differently after training. You may need bigger training. You may need continuous training. I think that's a sort of a really interesting thing. And then if the police side is persistent, then does that also translate? And then if you get really big effects, one interesting thing to look at would might be spillovers. Like if the police are behaving better in a local area, does that have any spillovers and perceptions in neighboring areas, which would be something that, you know, double whammy, too good to ask. But if these are really important things, which I think they are, it might be true. Oh, thank you. Yeah, I think fortunately, also, we have like a 2018 census. So we have a lot of data to actually try to examine not only crime data, but also like socioeconomic data that we can use to examine more heterogeneity analysis. And definitely like the problem with spillovers might probably be that we don't might not be powered enough to examine that. But we do have like at least the setting to examine spillovers and maybe look for that looking at treatment effects on like neighboring equivalents and so on. Thank you very much. Yep. Thank you very much for the very interesting sets of presentations. I have a question for Rashid. I was wondering, given the large, you know, your observations are quite huge. So I was wondering if you've exploited or explored rather any heterogeneous effects. It would be interesting to see a more nuanced narrative. So if you have, you know, these equality of opportunities, who is responding more, who is more motivated? Is it the females, males, people from urban areas, people from rural areas, you know, even if there's a possibility of knowing their friends, backgrounds or things like that. I'm just thinking this will give you a more nuanced, like, you know, narrative which would make it your framing of the policy message more targeted. Thank you. Thank you very much for the question. It is definitely a question of interest. We looked at it and we didn't find any major heterogeneity. So it affects males as much as female and females. And we also split it by quartile, at least of income among the poorest ones. But that's only something we can do with the difference in differences. And it's also, it seems to affect all of them. And there is a little bit, but the differences are not significant. So without it's not something where we can say a lot of additional things given the space constraint and the many tests that we're saying that we're doing. But thank you. Hi, I have a question for Chelsea. I didn't get whether the whole dynamic is about the product, the region. So what is, what is special about avocados? So I understand the mechanism, but it seems I got the idea that this could happen with oranges or with anything. Or so are we are we meant to stop eating all of these products? So I don't know if you can develop a little bit more. Yeah, certainly. So I think the reason that we started actually with the cross national analysis is precisely this, that we we don't think that there's anything in particular that is special about avocados, with the exception of the change in the export market share that Mexico was able to capture. And the period of time where there was the increased demand that led to an uptick in price. And certainly, you know, as we see that happen with lemons in the Sicilian context, long ago, it could happen with oranges, as you suggest. So if we tie this to policy objectives, I think this would be where and I kind of zoomed through this because of time, the role of maybe international organizations or maybe even domestic organizations and kind of monitoring some of these criminal interventions into these markets, and perhaps alerting international markets or consumers about this would have to happen on kind of a product by product basis. Because exactly this, you know, we don't expect that there's anything particular about a given agricultural product, although I do think that there's room to explore there about especially the labor dynamics related to different forms of production. Thank you. Yeah, hi. And my question goes to Rashid. It's about the Georgian effect that maybe you could explore. In particular, as I informed, the SISBEN has three different cut-offs for rural areas, urban areas, and I think there is another category. Those categories actually give us some idea of the lack of opportunities because, well, individuals in rural areas clearly have, in average, less opportunities than individuals in urban areas. So what I was thinking that was that although your estimator bundles the effects for the population that has one of each of the three cut-offs, maybe an estimation for each of the populations using the particular cut-off could give an heterogeneous effect that talks us more about the importance of the lack of opportunities and perhaps differences when this lack is greater for some groups than for others. Thanks. There is indeed three groups, the 14 largest cities, other cities, and rural areas. One thing is it seems like we have a lot of data, but the regression discontinuity is also very demanding in data. So every time we did some split, we can, again, there were some changes. The rural areas seem to be particularly affected by the motivation effect and without. It is interesting, but again, we didn't have that much power to really dig into this. That was the main constraint here. But there are lots of things changing at the same time. The cut-off is changing, but not that much. The order of magnitude in the move in the cut-off is not big, but I think it's above all that the number of constraints that you're facing where you're in rural areas is much more and perhaps the lack of motivation is really perhaps an even bigger thing. How would you have the right role models around you when you're in an isolated rural area? So I think it's interesting there's a lot more to dig. We're talking with Fabio about the possibility of future Serpilopaga programs, so maybe there will be more to do and to study in the future. Thanks. Thank you very much. Thanks for all the questions. Yeah, thanks everyone.