 Thank you so much for having me here. I'm Chelsea Estancona. This is co-authored work with my colleague, Lucia Tiscornia, and this is from Cocaine de Avocados, Criminal Market Expansion and Violence. I'd like to give a quick disclaimer, which is that we are political scientists, so there will not be a deaf and deaf, but there will be an instrument, so you can wait for that and see that towards the end of the presentation. So when we think about criminality and we think about organized crime, we're usually thinking about cartels, possibly gangs, mafias, these sort of larger groups, and we often associate them with illicit behavior, but certainly also with illicit economic activity. So when we think about criminal market capture, probably what springs to mind for you is the process of engaging in the cocaine market, so especially here in Colombia, we think about the process of producing cocaine and criminality related to that. We might think about heroin and opium poppy. We might think about marijuana, et cetera, et cetera. So largely illicit cultivation, and these are markets that criminals have largely been able to capture in part due to the way that they exist, due to their illicit nature, and then of course the illicit nature as well of these commodities. But we also see, throughout a variety of historic and sort of locational contexts, we see criminals engaging in market capture for fully legal sort of everyday goods. So a historic example here, you see a picture of lemons. The Sicilian mafia in the 19th century was able to largely capture the lemon export market into Europe. We also have more modern examples, so those are abalones in the top right there, and we can associate that with South African criminal groups, gangs, that compete for control of abalone exports to largely the East Asian markets. And then probably the example that comes to mind for many people and one that we're going to talk more about today is the example of avocados in Mexico and criminal organizations largely being able to control some of that export market. So these commodities, we might have a little bit more of a theoretical puzzle here, even though we have the empirical examples. These are less likely to be implicitly expensive to have high levels of profit because they are not illegal, so there's more risk implied in that illegal trade, which may amplify the price, and they may also be more closely regulated by the state, which implies a costly entrance into these markets and certainly a costly attempt to capture these markets when criminal organizations do this. So this leads us to our research question, which is under which conditions do criminal groups attempt to violently capture markets for legal commodities? So I have a couple of terms bolded there to hopefully give you a sense of the things that we're going to be interested in, specifically looking at violent behavior as our dependent variable. We have a fairly robust literature that I know many of you are familiar with about criminal expansion in general. So OCG is here as organized crime groups. And one way in which we see this is when criminal groups have an advantage or an expertise in some way. So we might think about them having connections, for example, to middlemen being able to easily export things that are already kind of within their wheelhouse. That's one way that we see expansion. Again, largely this literature is about illicit trade, not illicit trade. But we also see examples of when markets are easier to capture. And one in particular that we're going to focus on here is this low levels of technology. And we're going to equate this to thinking about agricultural products. So you don't have to build a plant. You don't have to maybe bring in products from abroad to be able to produce. We also are going to focus a bit on this territorial concentration component, which is tied to violence. So we often see criminal groups when they need to use an area of territory for production of something, they may seize that area violently. In fact, they often do. They also maintain control violently. So a lot of this depends on what is the ease of entering into a market and the ease indeed of eventually capturing that market. But just because a market is easy to capture does not mean that it's going to be sufficiently attractive. So, of course, we also want to think about what makes a particular commodity profitable, what's going to earn these groups money. And where the literature has looked here is largely about sudden changes in these markets. So, of course, we see international price, which is something that we will have in our paper as well. But we also see changes in policy that may enable better access to market or, of course, changes in scarcity, which then as well would impact the price of these goods. So for us, we're thinking about this as, well, let's put ourselves in the shoes of a cartel leader for just a moment. So we might be thinking about opportunities for immediate profit, which would be these changes in international price, especially if you're able to control a large portion of the market. So that may be part of the short-term goal is a large amount of an export as well as a lucrative export. But then there's also the long-term thoughts about how do you capture these markets? Eventually you might want to be able to control these prices to be able to better manipulate even these legal goods. And this is where we're thinking about market share. So this is what my co-author and I are going to try and add a little bit of complexity here to this literature regarding legal markets or legal goods. I don't have to define this for you, so I'll skip right over that. So we're thinking here about the country in which a criminal organization is located that states market share in the international market for agricultural products. So sticking with our example of Mexico and avocados, Mexico controls a large portion. They have a large amount of the export share of avocados. They're a large producer and they're able to get a lot of those into the market. So that larger export share, especially when demand increases, when price increases, is going to represent a larger profit. That's attractive for criminal groups who are thinking much like firms. Agricultural markets, we argue, may be easier to capture, but they do require that territorial access. You need a place to grow things. So this is where we think about violence. So criminal groups, unlike, well, at least ideally the state, are able to engage in violence to capture territory. Now, of course, that's in an ideal world. So violence, we expect, or we know rather from the literature on illicit commodities, decreases when there is some state cooperation, when there's corruption, and also when there's any kind of collaboration over extraction of rents. So when criminal groups overlap and they're able to collaborate, this is usually where we see less violence. But in turn, this means that when these protection rackets break down, or when there's increased competition among groups, we expect to see increases in violence as these groups compete for the territory to be able to produce or control production and eventually export of these lucrative illicit commodities. So in the context of illicit commodities, we'd expect violence when there is competition with producers perhaps, when there's competition with the state. So again, often these are, or usually these are goods that are taxed, and so we might think about the state having a particular state in protecting these commodities. So just to diagram this very quickly, we're looking at changes in export value, or we're thinking about changes in export value rather, which we think of as both a change in the price, the value of a good, and the amount that is produced, or the export amount. That for us equates to these changes in market share. So again, changes as a country is able to capture more of the share of the international market for a given good. That of course makes these markets particularly attractive, both for the short-term profit that criminal groups can earn, as well as for the long-term capacity to perhaps capture the control of that market. Because these are territorial bound goods, we expect that this will equate to an increase in violent homicides as these criminal groups compete with either other groups or with the state. So our hypothesis here is gonna be that a positive shock to the export value of a state's agricultural goods, remember I said we're speaking specifically about agricultural here, is associated with an increase in homicides that are perpetrated by criminal organizations. So our data here, we're gonna look at this in a cross-national setting, as well as within Mexico more specifically. I'm gonna speed through this a bit, just to try and get you all of this. So we're looking at the homicides per year at the country level. The way that we're going to classify agricultural goods is using the UN trade data. Those zero, two and four are agricultural commodities and raw materials, excluding things like oil, gold, those sort of typical kind of cash things. And so we're eventually going to calculate the country year share of the change in export value. I'll show you that in just a moment. But we're also interested in what is the degree of criminal threat? So remember that this is an argument not just about violence as a result of change in agricultural commodities, but only where criminal groups are active. So we're thinking about how does a country share of the global export value of a commodity change from year to year? And we're going to say that when that is suddenly much greater, when suddenly a given country captures much more of the global market for a newly lucrative good, this is particularly attractive for criminal organizations. So we're going to look at a three-year moving standard deviation of this value that I'm showing you right here and code for any changes greater than two standard deviations. We're trying to follow some of the other literature on shocks rather than doing this. I think I'll just let you look at that really briefly. I can talk more about controls as needed. So I've been bold here. I'm not showing you a table. I'm so sorry. I'm skipping directly to substantive effects. And my legend seems to have cut off, but the blue line here, so along the x-axis is the criminal threat, so the degree to which criminals threaten a specific country. And on the y-axis is the predicted amount of homicides, according to the model. And so what we see here, the blue line is where countries have not experienced these shocks to export value share, and the green line is where they have. So of course we see here that as that criminal threat increases, so in the higher end of things here, we see that those price, or sorry, those export value shocks are more likely to provoke an increase in homicides. And this is a positive and statistically significant effect. So just this is breaking it down into quartiles. Again, for sake of time, I will skip, happy to talk about it later. So I've shown you that there's a relationship between changes in export value share and criminal homicides, but we want to dig a little bit more into this mechanism. So now we're going to look explicitly at the case of Mexico. And we do so because we can track avocado production, and also because we have significant variation of violence. We have a variety of criminal organizations. We have a fairly long history of criminal organization activity. And you can sort of see where Mexico falls in our data as being an average agricultural producer, as well as sort of right in the middle of the export, or sorry, the homicides that we're looking at. This is just a little bit of demonstrating that in fact criminal organizations do operate where avocados are grown, just in case you were curious about that. So our local hypothesis is pretty much exactly the same language as that cross-national hypothesis, but we're looking at the municipality level. So as changes in the municipality share of the avocado export value, as those increase, we expect to see that there will be criminal homicides as increasing as well, as these groups seek to seize this kind of newly budding avocado market. So again, our dependent variable is going to be homicides at the municipal level. We are looking at the municipal share of avocado export value from Mexico. And then we're gonna be using the Cosciandrios data to code for the percent of years prior to our data where a criminal group is present. So trying to avoid any issues with endogeneity there by using the baseline. Again, we have a set of controls. We're using a linear model. For the sake of time, I will continue here. And here, this is exactly the same plot. We do actually now have a legend, so that's fantastic. The green line is the maximum change in a municipality's share of the avocado export value. So a really large sudden increase in the amount and the value of exports of avocados that the municipality is producing. Whereas the blue line is that minimum change. And so what we see here is municipalities that suddenly are really amplifying their avocado exports are seeing a really large increase in the count of homicides. And again, this is dependent on there being a large criminal threat. There being a strongly associated set of criminal organizations that are participating in this municipality. So you may very well completely understandably be saying this is incredibly endogenous, right? So criminal organizations are probably affecting price. Price is affecting the willingness to produce avocados. Here we have an issue. So what we're going to do here is to think about we need a driver that is tied to demand that may not be tied to the way that criminal groups are supplying these avocados. And one thing that we saw is provoking this shift in avocado growth is this change in demand related explicitly to the phenomenon of avocado toast. So this happens right around the mid 2000s, right on 2006. We actually see a recipe post from an Australian cafe owner. And this is where it all starts. And so we start to see a really, you know, you probably go, oops, sorry, to your local cafe and you can purchase an avocado toast even now. But what we do is use Google Trends data to think about how interested are people and perhaps recipes for avocado toast or places to buy avocado toast as an indicator of global demand. So we were able to get this, as I said, it's via Google searches via Google Trends. We have it over a set of years that are sort of akin to the years of criminality in the data in Mexico. So what we do here is instrument for that avocado change and export value share with the popularity of the search term avocado toast. And what we see is again a similar effect that as those searches go up, as the global popularity of avocados is increasing or changing in a positive manner, that then predicts that uptick in homicides as criminal groups are saying, ah, this looks like a very lucrative international market. We can seize the growth of avocados, we can control this. And indeed we have now, several years later, seen instances of cartels in fact manipulating price as a result of what had happened a few years ago when they took over some of these markets. So I'd like to conclude with just a little bit about these findings and situate these in a policy sense. So we do see that rapid increases and export value share are associated with higher levels of criminally driven homicides. When we dig deeper into the mechanism, in the case of Mexico, this seems to hold pretty well. And we can track this with global changes again in the search terms for avocado toast. As a policy driven audience, I think we just need to think a little bit more, a lot of our research on criminality, which of course is tied to development, to strong long-term development. Much of this is focused on illicit goods. And so my co-author and I are suggesting that we really need to renew our focus on agricultural commodities as a whole, both because these have a huge effect on sort of everyday citizens and because this is clearly something that we see in a global sense. And the last thing I'll leave you with is perhaps international organizations may be responsible for thinking about this, but you all as consumers may as well. So maybe rethink your avocado toast. Thank you.