Climate Change, Migration, and Foreign Aid in Central America

- So thank you all for coming this afternoon. We're really excited today to have Sarah Bermeo from Duke University presenting at our Refugee Seminar Series. Professor Bermeo is a political economist. She's an Associate Professor of Public Policy and Political Science. She's also the co-director of the Duke Program on climate related migration. Some of the work she'll be discussing today. Her broader research lies at the intersection of international relations and development. Focusing on foreign

aid, migration and climate change and how these areas intersect. In addition, she has a number of top notch journal articles, some great public facing publications from institutions like Brookings Institution and the Monkey Cage. And she's the author of a book Targeted Development: Industrialized Country Strategy in a Globalizing World from Oxford University Press. So Professor Bermeo has graciously offered to accept questions throughout the talk. Those of you who are watching online, please so

rt of enter your question in the chat function of Zoom and Teresa will be able to pass it along to the speaker. Without further ado, we're really excited to have Professor Bermeo here today. Thank you. - Thank you so much. (audience applauds) Thank you for the invitation to be here and also for that very kind introduction. As you heard, I'm going to be discussing some work that I have in progress. It is part of a relatively new research agenda for me that is leading on to some follow on projects

, but this is the one that is at the stage where we can present it. And I'll talk a little bit about some of our next steps at the end. So this project grew out of an interest of my own in understanding migration from Central America as it started to increase in the mid 2010, so around 2014 there was this uptick in migration from Central America. It went back down. And then 2018, 2019, we saw a huge increase. And both myself and other people, like I freely admit this, in 2018 I was still writing

pieces attributing migration from Central America to violence. And what we started to see though, the people who were actually coming to the US Southern border, many of them were coming from rural areas and many of them were arriving as part of family units. And so this to me seemed new and I wanted to explore it some more. It was kind of difficult to explore with the data that were available at the time. So what I'm gonna do today is kind of first talk a little bit like big picture, some of th

e interests we're seeing in climate related migration, which is part of the broader research agenda here. And then kind of delve down into the Central American case. Talk about both the findings that we have, but also some implications those might have for policy, particularly US policy toward the region. And then talk about some... Give a little bit of information about some of the areas where we're going next with this. So there has been a huge and I would say growing interest, not just from t

he academic community but even probably more so from outside the academic community on climate related migration. So we've seen kind of newspaper story after newspaper story, The century of climate migration: why we need to plan for the great upheaval. Like kind of this language around it that's really like this is going to be a massive thing that we need to think about. What are climate migrants? I actually don't know the answer to that. And where are they moving? Again, I actually don't think

we know the answer to that for the most part either. Climate migration growing but not fully recognized by the world. I think that's probably pretty fair. In the face of climate change, migration offers an adaptation strategy in Africa. So this is actually something that is a growing understanding among academics is thinking about migration itself. So you think about adapting so that you don't have to migrate, but thinking about migration itself as an adaptation strategy to climate change. The l

ooming threat of sea level rise in Bangladesh. I mean this is actually one of the places that gets the most attention when it comes to thinking about the intersection of climate change and migration, because sea level rise in Bangladesh is really causing large numbers of people to be moving particularly toward Dhaka where they're living in pretty precarious both economic and climate situations once they move. And the huge floods that happened in Pakistan affected I think something like a third o

f the country or something like that... Most people will probably remember this 'cause these are big headline news. World Bank report, the Groundswell report that came out in 2022 where the World Bank was really focusing... This was the second of two reports on this issue. Really focusing on trying to project where climate migration, where we were supposed to see the most climate related migration in the future as well as where people would be likely to move both from and to. And then the Biden

administration coming out with a report specifically on the impact of climate change on migration. So really starting to focus both internationally at the World Bank, also domestically in the US policy on this idea that there is a link between climate change and migration. I would argue, we actually don't understand it very well. And part of that is because there are huge data challenges in kind of unpacking this, and we'll talk about that as we go along. So what are some of the drivers? Again,

this is just big picture, but the one I'm gonna talk about most today is the top one there. But just for people who might be interested in the topic more broadly, what are some of the drivers of climate related migration? One is this kind of obviously warming but also precipitation changes. And by precipitation changes, those actually can come in numerous different forms. So some places are experiencing lower than average rainfall, some places are experiencing shift in precipitation patterns. So

maybe overall they get the same amount of rain, but it doesn't fall at the same times. An increase in extreme events of large amounts of rainfall, which can be particularly devastating if you actually follow... If large amounts of rainfall follow a particularly dry period because then the ground is hardened and can't absorb it, so you actually get more instances of flooding and landslides. What I'm gonna focus on today in Central America, one of the big things that we are seeing is impacts on a

gricultural yields. And there've been some studies now not linked to migration but just some studies on agricultural yield that show that these are happening like much more intensely. The studies that have come out in the last year or two suggest these impacts on agricultural yields are happening much faster and more intensely than what was predicted even say 5, 10 years ago by the models that were happening. And that this is affecting both the amount of the yields of crops but also the nutritio

nal value of crops. And so this is having a big impact, particularly, and this is unfortunately so much of what we see with climate change, particularly in the areas where people are already poor, where yields were probably already relatively low, and it tends to be clustered around the tropics. And that's being coupled as you might imagine with big increases in food insecurity. Because if you're have less yields, you're gonna have a lot less food security to go with it. It's not on the slides h

ere, but like globally just to kind of similarly be talking about small holder farmers in Central America, to give you a sense, these are not my statistics, I actually don't know 100% how reliable big aggregate statistics like this are. But the best estimates out there is that globally there's about 500 million small holder farmers. And they collectively produce about a third of the world's food supply. So when you think about that, right? The idea that small holder farmers who are on the front

lines of climate change are something that we should all be concerned about, even if it's not for altruistic reasons. We're talking about a really big portion of the global food supply, which as if we learned anything from things like the Russian invasion of Ukraine, impacts on the food supply in one place end up having ripple effects throughout the global economy for food. So I would argue, if I'm talking to government or international organization policy advisors, one of the reasons we really

should care about this is that this is going to be a big global impact for everyone not just... I personally think we should care about it just for the sake of the small holder farmers, but from a big grand policy perspective, there are also reasons to care about this for global food security. Some of the others that I'm not really actually going to talk about in my talk today, but just kind of for completeness because other people might be interested in other aspects of climate change that coul

d drive migration. Ocean warming is having impact on fisheries. Again, small scale fishers are heavily affected by this. Sea level rise is driving people from the coast-inland. Floods, storms, heat waves, fires. And one of the things, and this actually becomes pretty key, climate change is intersecting with a lot of existing drivers of migration, which is maybe particularly then tragic for the people who are already having these kind of negative impacts, and then you layer climate change on top

of it. It also makes it a particularly tricky issue to study from a data perspective, because how do you disentangle these issues of climate change when most people don't migrate just for climate change. There are, I think you could argue, maybe like small island states might be an exception to that. But climate change coupled with good policies, good governance, low levels of violence, doesn't necessarily cause people to migrate and particularly doesn't necessarily cause people to migrate inter

nationally. So one of the things that I'll say when I'm giving a talk to kind of a broad audience is people may leave their homes for climate change but they don't leave their countries because of climate change. If you're leaving your country, there's something else going on that is intersecting with climate change. Again, I said small island states might actually be an exception to that rule. They are not at some places that I've studied particularly myself though. Okay. So now to ground this

a little bit in kind of academic theory, there's nothing new about studying the relationship between migration and income. And a lot of what we're talking about with say climate shocks to small holder farmers is a change in income. Their farms are not going to produce as much as they produced before. So we should be thinking about, how does this way of thinking about particularly kind of slow onset climate change as opposed to a wildfire or something that kind of comes through very quickly and c

reates more of a natural disaster situation, the impacts of climate change in agricultural sector are more impacts on people's income, lifetime wellbeing, their projections about how they're going to be doing in the future. And there is a fairly rich literature on the relationship between migration and income, which I want to argue, actually doesn't tell us much about climate related migration. And the reason is, so let's kind of look at that. I believe everything on this slide actually, except

for the very last part maybe. So I think that there is overwhelming evidence that as countries go through kind of a normal growth process, we see this inverted U shape relationship between income and migration. And when you think about it, it's extremely intuitive, right? It's also been documented by lots of people who I put up there. But it's also just intuitive. And when countries are really poor, people may want to migrate but lack the resources and connections to do so. As countries get a li

ttle bit richer, they still are looking out there and seeing other countries where they could be better off, where the returns to their new education could be higher than where they are at home, and now they have the income, the money to actually migrate. And then you get to a certain level of development and people have the income to migrate but no longer have the desire to do so because they can build the type of life that they want within their home country, right? So I actually think this is

... I do not wanna dispute this at all, right? I think this is actually very robust finding in the literature. However, most countries most of the time are seeing an increase in their income, right? Incomes in countries tend to grow over time. So when we're looking at these cross national relationships between income and migration, it's not surprising that we pick up this kind of aggregate effect, right? But what happens if income slides backwards, right? If we really believe that this is a rela

tionship, this inverted U. As people get poorer, they should be less likely to migrate if they were on this kind of upward sloping portion of the inverted U. And what kind of drove me to this as an academic question is that is just not what we're seeing in Central America. We are seeing people losing their farm income and deciding to migrate. So it was not explained by what we were seeing kind of in the economic literature, which to be fair, was not actually claiming that this was explaining bac

ksliding, but they were making, and this actually comes up a lot in some of the policy work I do, these claims that because foreign aid is supposed to improve development, there's actually lots of debate about whether it does or not, but the people who are making foreign aid decisions are doing so in part because they think it improves development. And if that were true and you were giving it to countries and they were getting richer, you might actually be financing migration journeys. So you co

uld actually see an increase of migration as foreign aid increased. Assuming foreign aid were actually having development benefits for the country. And so this is where I started to take issue because this is really sinking into the policy advice. Like maybe we don't wanna increase foreign aid. That's gonna give people the development and the ability to move. And I was like, that is not what we're seeing. And so I kind of sat back and thought about it a little bit. And at every stage, one of the

key things we know about migration is most people don't move, right? So what makes the news, what we analyze with people who do move, but most people don't move. And there's actually a growing body of literature on immobility and why people don't move. But if you think about that, it actually then creates a selection effect. At any given income level, the people who wanted to move at that income level and had the ability to do so have left. And left behind is a group of people that if everythin

g stayed the same, actually weren't planning to migrate, right? And now some of those people face a change in their expectations of their livelihoods where they are. So when they thought their country was on the upward trajectory, they were like, yeah, we're gonna stay, we're gonna work, we're gonna contribute, we're gonna just keep getting better off where we are. Yeah, we might be better off if we move to Europe or the US or Australia or Japan or something right now, but we see everything goin

g in the right direction where we are, we're not going to move. If that changes and they no longer see that as the case, then that has a potential to influence their decision on migration, right? What I think too is, I don't actually know how to test this one, so this is more just from my anecdotal observations of what I'm seeing. In the Central American case, I believe it is true from kind of a lot of time spent analyzing this and speaking with people in the field. People are facing the opportu

nity... The choice between migration today or not migrating at all. Because if they take their very little savings or ability to raise capital and reinvest it in their farm, which may not produce again next year, they may no longer have the ability to finance their migration journey. And so they're facing a choice of do I reinvest in something that's failed multiple times or do I decide to migrate today because I think there's a very low chance that I'm actually going to be able to stay? And if

I don't migrate today, I might not have the money left to migrate tomorrow 'cause I'll have spent my limited savings and capital accumulation reinvesting in my farm. Again, I don't actually know how to prove that, but that is kind of anecdotally what I would say we're seeing. Anyway, from the policy perspective, I would argue that the implications of this for foreign aid are actually huge because there's a difference between thinking about preventing people from backsliding and helping people mo

ve up the ladder, right? A lot of people, if they have hope that things can get better where they are, we know most people most of the time don't migrate, right? And so if they have hope that things can get better where they are, they're more likely to stay and try that one more year. And so I think there are kind of these big implications for how we think about that relationship between foreign aid migration, particularly in the climate change arena. Okay. So all of that is kind of like academi

cs love to sit around and draw graphs and theorize. What about some actual data? So this is joint work with some co-authors. Many people probably know David's name. Gabriela is a fantastic graduate student that we have at Duke. She's really, really quite excellent, joint with the law school. And so this is some work that we have done. So let me give you a little bit of a background on Central America and then I'll dive into the data that we actually use. So this isn't a region that is predicted

to experience kind of increasing impacts of climate change going forward. It is already seeing them, and they're expected to continue and to get worse. The precipitation patterns there are definitely changing. If you talk to kind of any farmers on the ground, they just see this, it's definitely happening and you can also pick it up in the data. So we're kind of seeing the data or showing something. And this is being validated by the farmers on the ground. In particular, I don't know, does anyone

here know anything about Central American growing seasons? I didn't before I started looking at this. Many places in Central America have two seasons, some have three, but most of the places that we're looking at, most of the farmers have two growing seasons. One that is in the planting in the early spring and then harvesting in like June, July. And another that plants in like July, August and then harvests later in the year. This is particularly relevant, like why am I giving you an agricultur

e in Central America lesson? It's particularly relevant because there's this period in the middle, in the July, August timeframe, it's called the canicula. And it's not new. I don't wanna say it's always been there 'cause I don't know what always means, but like for the time that we have data, there's always been a dry period in July and August. But it is becoming drier and longer. And the impact that this has is that it affects the harvest from the first planting as well as the planting for the

second season. And so you can end up losing the majority of two crops in a row if this canicula period is changing and becoming more intense. And that is what the data are showing us. What we saw in the 2014, 2015 period was a drought in Central America. I'd say this is one of the things that first kind of piqued my interest in it, but not in an academic type of way, just in an I've studied Central America before. And in 2014, 2015, the famine alarm bells were starting to go off. They were one

step away on the kind of the people who declare famines, right? One step away, literally on the scale from declaring a famine in Central America. Like famines don't happen in the Americas like that. Like not like climate caused or farming caused famines like that. And so this was something new. And then 2018, again, there was a very substantial drought. And so you started to see really strong impacts on farm yields and on the food insecurity indices for these countries. So this was like a big de

al. Particularly, Guatemala and Honduras. In Guatemala, it's just under 50% of the people are in rural areas, about 47% at last estimate. In Honduras it's in the 40s, not quite as high as in Guatemala. So a large number of people living in rural areas. Many of them indigenous. Again, particularly in Guatemala. And the food security situation at baseline, miserable. These are some of the... Prior to this, right? Prior to these droughts, some of the baseline statistics in areas of Guatemala, rural

areas, indigenous areas, some villages had something like 70% of children were experiencing stunting. So were not growing correctly before the age of five. 70%, like that is... Guatemala is actually one of the worst countries in the world on some of these statistics for child nutrition. And it's very unequal society. So this is being concentrated particularly in the rural areas. And agriculture makes up about 30% of the economy in Honduras and Guatemala. A higher share of employment, but in ter

ms of economic activity, that's about what it is. And then when you look at the agricultural practices that are in place there, there's a lot that could be done to improve them just based on current knowledge for what works. And so there's really this kind of, this doesn't really... This is gonna impact more of the policy implications for later on, but I think it's important to note that you're experiencing these impacts but you're also not nearly deploying the resources the best that you could

even given current knowledge. Okay. So here's what we're seeing. We got data from Freedom of Information Act request that looked at... That provided data on subnational place of birth for migrants from Guatemala, Honduras, and El Salvador for the period 2012 to 2019. So why that is important is because... And we were particularly looking at, I should say family units. So people who migrated with at least one child family member with them. So these are total numbers of migrants, but the orangeish

, reddish, whatever color that is bar is showing, these are total migrations from Guatemala over time from 2007 to 2019. Our data starts in 2012. So that's where we can start tracking the numbers of those that were part of family units. And you see that that is rising not just overall, but also as a share of the total number of migrants who are coming. And significantly so. I just kind of put up there on the side, I looked up what the more recent numbers would be. There's a lot in the news. This

is actually puzzling. It's never puzzling to me really when things are in the news that don't sound completely accurate but... (laughs) People keep saying, oh, migration... And in the news, it's like migration from Central America is going down and now we're seeing more people from Venezuela and from Nicaragua and from Haiti. You know the latter part of that is true. (laughs) But it is still true, I mean, Guatemala is still the third largest sender of migrants to the United States. Honduras is

the fourth largest sender of migrants to the United States. These are relatively small countries, so that is actually a big deal. So have they come down slightly comparatively to where they were in 2019? Yes. But I mean, we're still talking. So we were at maybe like 270,000 in 2019, migrants coming from Venezuela. We're still at 220,000, which based on the historical numbers is still really, really high. So yes, maybe they've come down a little bit, but it's still a big deal. And the percent of

family units here, it rises to about like 70% in 2019. It's come down, but they're still the largest group coming from Guatemala are families, people coming with children. So the blue bars are data that are just freely available on the website. And then we were able to put the family unit data in based on what we had. Again, these are at the country level, we're gonna dive down deeper in a minute. Picture from Honduras looks the same. That's basically the only thing you wanna take away from this

slide is that same thing is happening across, actually really similar numbers in 2019. But also that huge growth in family unit migration as a percentage of total migration, and also still really big right now. The only one that actually has fallen off in 2023 is El Salvador has fallen off some. And you see the numbers for El Salvador while going up never really reach those numbers that we saw, that 270,000 or so in 2019 from Guatemala and Honduras. We're not talking about those type of levels

from El Salvador. El Salvador is actually a less agricultural economy as well. Okay. So that's to kind of give you the picture. One of the things that surprised me when I started looking at this is really what's going to turn out to be the almost overwhelmingness of the impact of agriculture and of rural location on driving these migration flows. I actually didn't expect that as much. What we think of as kind of the traditional drivers of migration from Central America, which I think are still t

rue by the way. Poverty and lack of economic opportunity, food insecurity. One that I always have to put up there and always say this is okay, the number one way to get yourself uninvited to talking at government things is this next one. Violence related to drug trafficking. All of these drugs are bound for the United States. So this is completely being driven by the demand in the US markets and by policies that the US and Mexico took to tamp down on drug trafficking routes through Mexico. That

caused them to relocate through Central America. And saw huge spikes around 2006, 2007 in homicide rates. And Honduras and El Salvador were kind of vying for the dubious distinction of most homicides per capita in the world in the mid 2006, 2007 range. It's also true. So there's really kind of this like US market implication there where drugs coming from... Mainly cocaine from South America up through Central America to the US market. Estimates of this are hard, but around like 90% of the cocain

e in the US market comes through those type of routes, like through Central America routes. And guns. There's no domestic weapons manufacturing in Honduras, El Salvador, Guatemala, the guns are from the United States. Either leftover from the Cold War or trafficked illegally southward on the border. We always think about security at the border coming northward. These are often guns purchased legally in the United States, but then trafficked illegally down to Central America. There's a lot of ext

ortion, there's a lot of corruption, there's a lot of poor governance. So those are things that I think are definitely on the table. But then climate change. Kind of all of those things were preexisting and then you lay climate change on top of it, particularly the droughts that we're talking about. Here is the dry corridor of Central America. So like that brown thing. (laughs) And it is named that way because it does not get a lot of precipitation. But it oddly you might think, given that it do

es not get a lot of precipitation, has a lot of agricultural activity, particularly small holder farmers and subsistence farming or at least partially subsistence farming. Farming for yourself, but maybe then also having a job on a bigger farm or something somewhere. You can see it covers almost all of El Salvador and good chunks of the other countries. I don't have data on Nicaragua though we are doing a project right now that is in Nicaragua. Okay. So I said at the beginning, one of the reason

s that I think we don't know a lot about this, about the relationship between climate change and migration, is because it's difficult to measure. Part of that is because climate impacts is, I think everybody knows, vary within a country. So one part of the country may be affected by something that is not affecting the other part of the country or at least not affecting it in the same way. And data on international migration particularly, it's almost always at the country level. Internal data on

migration is almost non-existent. For countries that actually do a regular census every 10 years or so, you might get some snapshot. And we'll often ask in those questions too, like if people have moved, where have they moved from? So you might get some snapshot, but still it's at 10 year increments. And not every country is doing a high quality census every 10 years. El Salvador hasn't done one since 2007. Any census, not just a high quality one. (laughing) They haven't done any census since 20

07. And the one thing that people, researchers started to do was survey, to ask people if they were migrating, why were they migrating? It turned out that this maybe a new wave of surveys would be better, but for the most part, these surveys didn't ask enough questions to ascertain what was really going on. So people you say, why do you migrate? Well, I migrated because I didn't have a job. But then if you asked another question they might say, well, I was looking for a job in the city and I cou

ldn't find one. But then you ask another question and it turns out that they left their farm. And then when you ask another question, it turns out that they left their farm because it hadn't been producing the same way that it had been producing before. So you have to kind of really have detailed dig down type of answers. And those type of surveys really haven't been done. And it's not even clear that even respondents would necessarily trace that causal pattern. So we were excited because we had

these subnational data. So what we could do is we could say where in Honduras, El Salvador and Guatemala were people moving from? And what had been happening there? Had these been places, so I don't have to ask them if they were moving because of agricultural stress, I could measure the agricultural stress. Or more clearly, the FAO could measure the agricultural stress and I could use the indicator. So we were able to get these data, they have a subnational location of birth. We decided to do t

his at the department, which is their version of a state so the first subnational level. And analyze these against subnational data on agricultural stress. And just to give you a sense of where people are coming from. This is Guatemala. These are numbers based on those Freedom of Information Act data that we have. And what they show is total numbers of migrants that showed up as part of a family unit at the US southern border between 2012 and 2019 and were counted by the US, right? So they actua

lly were count... They came in contact with a customs and border protection official and they were counted, right? These are the numbers. So by the three highest sending departments, these are the highest percentage. The departments on the left side, the departments with the highest percentage of out migration. And let me put that in perspective, that 7% right there means that between... And this is based on population of 2012 census. 7% of the population number of Huehuetenango in 2012 showed u

p between 2012 and 2019 at the US southern border and was counted by a US CBP official. There's a huge under count in the total amount of migration. Because it doesn't count internal migration at all. It doesn't count migration to Mexico, which is actually huge. Mexico is a huge destination for Guatemalan migrants. And they actually have some guest worker programs and other things in Mexico for Guatemalans. And it doesn't count people who either migrated somewhere else or came to the US and were

not counted. The vast majority of this town were actually counted people who were just turning themselves in at the border. So those are really big numbers. And in Honduras we actually... I'll go back to that in a second. In Honduras we actually see percentages like more departments with those really high numbers. Those are really big, right? So I wanna show... I don't wanna do this for every country 'cause it'll take too long, but just to give you a sense. In Guatemala, the three departments w

ith the highest number of apprehensions at the US southern border, so these are not the rates, but these are the three highest numbers. If you look at them, they were almost all like... The lowest was 68% rural population. These are highly rural areas that people are coming from. This is not the picture that I think even still unfortunately the US government has in their mind about where in these countries migrants are coming from. So just really quickly. Same type of picture in El Salvador. And

now let's get beyond the descriptive data and actually look at the analysis. So for those who care about this stuff, if you don't, that's fine. (laughs) The dependent variable here is we... I use two different... Or we use two different dependent variables. One is the natural log of the apprehension rate, which is the number of people per 100,000 population of that department. And another is just a straight up log of the number of apprehensions. Okay? So you might actually think as I would say

I did that you would get a lot of movement in the apprehension rate, but that the number of apprehensions might not really be higher in rural areas because they don't have as many people. So they might be sending a higher percentage of their population, but maybe it's not necessarily an overall higher number of people. Subnational data on lots of things are hard. Okay. I'm just gonna be clear about that right now. We chose here to separately analyze the three countries because I think there's ac

tually differences across them. We use a measure of the agricultural stress index from the FAO, which actually compiles it at the department level. So we didn't even have to do anything with that, it was there. So agricultural stress here is actually a pretty high stress category. It's not just slight variations. It's like drought or just under drought. So it's the percentage of cropland in a department that is experiencing this fairly high level of agricultural stress. And we look at, if you re

member earlier I talked about this period of time called the canicula, which is really important? So we look at kind of the maximum monthly stress within a year that a country had. But then we also zoom in on that canicula period and say, does it matter more if it happens during this time of this dry period? We control for the homicide rate at the department level. And that we do have at a yearly basis. The percentage of the population or the ratio of the population is rural compared to the tota

l population that we only have from a point in time. So I only have one number per department because it's from a census. Wealth Index from the Global Data Lab, which is also only one point in time number, so only once per department. And then we put in a year trend variable because we wanted to just account for yearly trends that might not otherwise be picked up in the data, especially because a lot of these are not yearly, but our data on apprehensions are. So our apprehension data are by depa

rtment year. So we do have yearly variation on that. And so here's the actual regression model that we run, where I is department and T is time. And random effects models. You can ask me about fix effects or maybe I'll just talk about them afterwards. Robust standard errors are clustered on the department. The data are from 2012 to 2019. And like I said, we use two different dependent variables. Okay? All right. Are we good? You said I could take questions during but nobody's asked, so I think m

aybe they're all asleep. Okay. (laughs) Excellent. All right, so here are the results of the regressions. Don't worry if the numbers don't make a lot of sense 'cause I'm gonna show you some marginal effects in a minute. But first just kind of raw, what pops out. The top two rows are agricultural stress ratio, which is the maximum monthly shocks that we see, right? Or amount of stress that we see. And then the second one is limited to that July and August period. So the maximum that we see there.

And you see that both of them are very positively associated with increases in... There's a one year lag here. So having agricultural stress in your T minus one has a big impact in the number of... Well, at least it's highly correlated with the number of people showing up and being counted as part of family units at the US southern border the following year. Okay? What we really see too is huge impacts for the rural ratio. Something I did this time around that we haven't done in our previous pr

esentations of these data, I took population out of the regression on apprehensions 'cause I just wanted to see what it looked like when you didn't even control for population. So this raw apprehension number or the logs of the raw apprehension numbers on those two right hand side columns are not controlling for population. So that's just saying, if you had higher agricultural stress or you were in a more rural area, you are sending more migrants to the US, not even controlling for the fact that

you have a lower level of population. And one thing I have to be really careful about here, the homicide rate does not seem to be picking up a lot. I don't actually think that's because homicide rate doesn't matter. I think that's because that's kind of what was driving the regular level of migration that we were seeing anyway. And I also would argue, though I'm not kind of a crime or a violence scholar, there's some threshold effect that you hit where slight variations around it don't matter t

hat much. So the cities in these areas are pretty violent, but they stay similar levels of violence throughout this time. So you might not be picking up the fluctuations. - [Moderator] She's asking if you want to address the possibility that multiple apprehension may be observation for the same individual. - So yeah. It's actually a great question and it's one of the reasons that we have not gone back. So originally, when we got the data, they only went through 2019. A multiple apprehension star

ts to become a much bigger deal from 2020 onward. So while it can be a partial issue here, we're less concerned about it because in part, people coming with children were much less likely prior to 2020 to be turned back. So this is all before the remain in Mexico policies went into effect. And so it was, if people were deported, they were deported all the way back home. And making that journey again with kids would be actually quite difficult and unlikely to show up here. But they also were just

much less likely coming in that period with children to be deported. Because most people were making asylum claims when they got to the border. And so with children under US law, children cannot be held more than three weeks, 30 days, something like that, in detention facility. So they were being released. And this is actually what led to the Trump administration's 2021 policies of separating children from parents because they weren't able to send them back. So since this is all family units, a

ll people with kids, it's much less likely that they would've been returned back. That does change after COVID and after the remaining Mexico policies go into effect. But not in these data. - [Moderator] Yeah, just as a two finger, just how are you thinking about this? So this is a really, I think policy relevant population to look at, which is this family unit migration relative to regular migration. But is it possible that the reason why the rural is picking up so much here is that that's the

form that rural migration takes, whereas city migration is something... I'm just wondering how you're thinking about this in terms of- - [Moderator] Yeah. So one of the things I'm thinking about is I need those data. So I've actually FoIA those data now to try to get a better sense so that we would be able to go back and compare kind of single adult migration to... So in this time period, this becomes the largest amount of migration. But it's a good question that I can't answer, right? Yeah. - T

he second one is, if you've looked into Mexican government statistics because of people... The people who are likely to stay in Mexico versus continue on too is the other- - Yeah. So not as part of the analysis. I did kind of just do a little bit of a look at like, because the Mexican government does also keep statistics on who they are encountering. I mean what a lot of people in the United States don't know is that huge numbers of people on their way to the US do actually get detained in Mexic

o and deported back south. Like before the US had policies with the Mexican government to do that. So you certainly saw, I don't know the exact numbers of it, but certainly, I mean almost all of these people or all of them, actually all of them travel, 'cause this is just southern border. All of these people traveled through Mexico. Some percentage of the number who set out would've been detained in Mexico and sent back. Yeah. Yeah. - [Participant 1] I was just trying to think, something you sai

d earlier that more and more people are making a decision on why they can move right away or to invest in rebuilding something, and once they invest they can no longer migrate. So could it be that one year lag might actually be too long of a period? Meaning that, if anything, these are conservative estimates because some people might move right away or have to move right away. - Yeah. So it's a really good question and actually thank you for asking it because something that I should have noted.

So US fiscal years for when they've apprehended people actually end September 30th. And so you actually have to lag it or you could be picking up October through December migration. So yes, that could be true, but since I don't have it on a monthly basis, I can't disentangle that. Yeah. I do look at different lags, and you do tend to see still at the two year lag mark, a positive effect. But then nothing. Like if you go back further than that, I'm not seeing anything. Of course I only have like

eight years worth of data. So that might be limiting my ability to get good lag structures too. - [Participant 2] I have a question. So you showed that there were some regions where people were migrating much more than other ones. - Yeah. - [Moderator] Isn't that picking up also like inequality actually in those regions? I mean they are poorer I mean compared to other ones as well- - Yeah. I think one of the key things that kind of struck me about this is to kind of see that huge increase in mig

ration over time. A lot of these people were coming from areas where in just kind of basic economic models, we would've expected them to have already migrated. (laughs) Like the prediction of who would be left in these areas based on economics and proximity to Mexico and the United States should have been quite low. And what we'd actually seen was not a huge amount of migration out of these areas before that. And so a lot of it is... So indigenous community is very tied to their land, tied to th

e local communities. I mean attachment to place for other reasons. People have been on these farms, their families have been on these farms for generations. But kind of just from an economic wellbeing perspective, we actually would've expected to see people migrate before and hadn't seen that. I actually think it really speaks to the level of desperation that these droughts ended up having throughout this time period on these areas. Because when you think about like what we were starting to see

in US government policy. US government policy toward migrants during these times was becoming tougher and tougher, more and more crackdowns on people coming and yet more and more people coming. The route through Mexico is extremely violent, lots of extortion, robbery attacks, and yet people are not only coming but bringing their families with them. And it really does kind of speak. And even after the child separations, during the Trump administration started happening, people are still coming. A

nd it was definitely not because there was a welcome sign out on the US. I mean US policy was becoming tougher and tougher and more and more people were still coming. And I really think it kind of speaks to the desperation that people are feeling. There's something that we don't capture here and I actually don't know how to disentangle it, even if we do get the more recent data. In 2020, there are two major hurricanes that hit in Central America. Two category four hurricanes within two weeks of

each other, Eta and Iota. I don't know what that kind of does. During this period, we don't actually have an instance of kind of major storms. The climate events in this period were more drought related. Okay. So what I'm gonna do is just kind of very quickly show you Honduras and El Salvador. Honduras, this is where I think it's actually quite interesting. We pick up the impact of agricultural stress, but only when we actually zoom in on that canicula period. Those are P values in parentheses.

So the others that may be showing something, but we wouldn't consider them significant, but they're not like going in the wrong direction or anything. But really when we zoom in on the canicula period in Honduras is when we see it. And then El Salvador, I have to admit this one actually surprised me a little bit because it is a less agricultural economy, but we still see this impact both at the monthly level but also during that canicula period as well. And so I thought that was actually to me q

uite interesting. One thing we don't know for sure, like I just know numbers of people and where they're coming from, I don't actually know if these were the farmers, right? And we anecdotally know because lots of people have, lots of groups have interviewed them and talked with them. But the actual numbers I'm counting, I don't know. So what this is saying is if there was agricultural stress in an area, more people left that area, but it could be having ripple on effects that are affecting the

urban areas within locations too. So it could be food insecurity. If you're relying on the nearby rural areas for some of the food into the urban centers that this is actually having an impact to. Okay. So here are some of the marginal effects. This is zooming in on the canicula period. And saying what happens if agricultural stress rises from its level of the 25th percentile to its level of the 75th percentile? What change do we see, kind of holding other things constant? What change do we see

in the apprehension rate or the number of apprehensions? And you see that like across the board, both for apprehensions and for apprehension rate, like these are big increases. It's not just like a statistically significant result, these are substantively quite large numbers that we're seeing, in terms of increases of people moving out. And obviously, if you look up at like the 90th percentile that actually lists, it's almost asymptotic in some ways. So it goes up a lot. Okay. So what do we see?

I said, at the beginning I set this up as, there's a discussion amongst policy makers on foreign aid about the link between foreign aid development and migration. There's certainly in the US government a discussion about foreign aid policy to Central America and how you can or maybe can't use foreign aid to decrease migration. So you would think that given that foreign aid might be flowing to the places where migrants are coming from, and it's not. This is foreign aid, total US foreign aid. And

then agriculture and development food aid. So the blue bar, total aid, the green bar, agriculture and food aid, which is obviously a percentage of that total aid. And then the line, the gray line, is the percentage of US foreign aid going to agriculture or development food aid as a percentage of total US foreign aid. It is astonishing to me that that line has gone down over time. And I've told the US government this. (laughing) I cannot fathom this. As we are seeing a huge mass exodus from rura

l areas in Central America, predominantly agricultural areas, the focus of our foreign aid has been shifting away from agriculture. I genuinely don't know why this is. Yeah. - [Participant 3] What are the areas that are shifting- - Yeah, it's a great question. The US government spends a decent chunk of money on governance initiatives. This doesn't include military aid. So there could be other aid that's going on as part of whatever we call the war on drugs these days and that type of thing, righ

t? But that would not be included here in any of the numbers. So on the development side, they are spending money on things like education and health, but a large chunk of money is going into governance and civil society promotion. I would argue that there's absolutely no evidence that there's been any success in that, in spending on that. So it's not that I have anything against promoting governance and civil society. I think that would be a good thing if we knew how to do it effectively. One o

f the interesting things is, I think if you kind of looked globally, foreign aid has an ability to kind of build political or governance capacity. This does not seem to be one of the things that it does particularly well. Things that it has had more success at are health, especially things like vaccines or like small child health and agriculture. Like there are successful agricultural programs that you can draw on, including actually one of my follow on projects is working to survey people who h

ave been in areas that have had water smart agriculture programs rolled out by Catholic Relief Services, a big international NGO funded by the Buffett Foundation, not funded by the US government. And they're actually showing pretty significant results. And yet I've asked the people there, could you scale this? And they're like, sure we could scale it, we don't have enough money to scale it. So I think there is scaling potential out there within these agriculture communities. There are practices

that have been shown both globally but now also within Central America, those programs are talking about only being rolled out for like five years or so. Showing really significant results including during the 2018 drought. Being able to keep more farmers above the level where they would need external help than farms that didn't have these practices, and yet we're still not scaling up. Except for the fact that I don't know how to churn the... I don't know if anyone else had to churn that ship th

at is called USAID and actually get it to focus on new things. It's not clear to me why we're seeing this pattern. This seems completely against the US declared interest in the region. I mean this means that the people who are desperate enough to need the most help are also not getting the aid, right? Which is the small holder farmers. Okay- - [Moderator] Sorry, when- - Yeah, absolutely. - [Moderator] From the chat it says that, how is that comparing of the agricultural budgets of the three coun

tries? So the budget of the Ministry of Agriculture of Guatemala, Honduras and El Salvador- - You know what, that's an excellent question and I don't have any answer to it. That's gonna be one of the things I'll look up if I can actually find it, which I'm not 100% sure. In general, these are countries that have relatively low taxation and low government expenditure. And particularly in Guatemala, a long history of ignoring the rural, primarily indigenous populations. So my guess would be that n

ot a lot of money is being spent there. COVID also really showed kind of the lack of infrastructure in lots of countries, but I was looking particularly in these three. For being able to cope with a crisis which would include something like a drought. Certainly we saw it after the hurricanes in 2020 and with the COVID crisis, that there's just not the kind of governance capacity to deal with those types of situations. But thank whoever it was because that's a fantastic question and I will look i

t up. (laughs) Or if she already knows, by all means. (laughs) Okay. Oh, was there another? Yeah, no? Okay. This is my last slide. So kind of in conclusion, I think there's pretty strong evidence right now that within Central America, we're seeing a high correlation between areas that experience high levels of agricultural stress and migration to the United States. And this is predicted to just keep increasing, right? Climate change is having an impact now. These are areas that are... Pretty muc

h all the main climate models expected to be high impact areas for climate change. So without help, we should expect it to get worse. But there is evidence that outside, and which I haven't shown you here but is in some of the other work, that outside agricultural assistance can actually make significant differences for small holder farmers. And so here are kind of these next steps. One is we have already conducted the surveys. We're in the process of like getting them in and analyzing them now,

of small holder farmers. In addition to these three countries, Nicaragua and the Oaxaca area of Mexico, where we have been able to ask them what their exposure is to these level of water smart agriculture trainings or demonstrations or just information, there's been radio campaigns. What we're trying to get at there is does increased exposure or more intense exposure to these programs have an impact on... They've already shown it has an impact on like soil health and things like that, but does

it also have a follow on impact on other household decisions, including migration? And so we're looking to be one of the first that has really looked to tie a specific development program to migration decisions. But we're also looking there, which I think is actually really cool and not related particularly to this, but on information dissemination. So how do you get information into the hands of, we think 500 million small holder farmers worldwide. We're gonna have to figure out if we really wa

nna help them, ways for knowledge diffusion. Because a lot of the things that they're doing in Central America on training are things that people in the farming community and other places have known for a long time, like don't burn your crop stubble. Like this is like agriculture 101, right? Because it's depleting the nutrients inside the soil. Other things like planting cover crops. None of the interventions actually that they're doing right now are high cost interventions, 'cause of course the

re's no foreign aid going to it. This is not like new irrigation systems or something like that. It's planting cover crops, low tillage to retain soil moisture, not burning crops, double planting trees to provide strategic shade and to prevent erosion. Like all things that also once done once can often be maintained by the farmers themselves. So fairly low cost intervention except that the training is actually pretty expensive. So if there are ways you can get... Find ways to disseminate the kno

wledge. Because when they actually go out and work with farmers, the uptake is quite high. Farmers are not resistant to it. That was my first question. And it's like, well, do they say no, this is the way we've always been doing it? They're like, no. They say, can you help us change all of our crops over to this? So uptake seems to be high, but also very hands-on intensive in some of the programs. And so there are ways you can kind of do that with a less financially and people power intensive ap

proach. And then this is my more... I have not actually started on this yet, but this is my like later in this year project, is to use these data that I was just talking about today and see if we can see the climate migration models that have been predicting, or I don't know if predicting is even the right word, estimating anywhere between like a couple hundred million and 2 billion people will move globally due to climate change. Obviously, they can't all be right. (laughs) Maybe none of them a

re right. I actually think there's a big problem because a lot of them are using historical migration patterns to project the future. And one of the things we're seeing here is that this current wave of migration does not look like the previous waves of migration in Central America. So can we at least for these three countries, look at some of these models that are out there and see which perform better based on what we actually observed. Because again, most climate shocks are subnational and th

ere's very bad data on subnational migration. And so that is actually kind of for me, a fun project that's on my horizon to look at. Okay. So that is all I have and happy to take any more questions. Thank you guys. (audience applauds) - [Participant 4] So I have a question just on what you were last saying, which is kind of like how the climate driven migration fits in with this broader, I would say system of migration hemispheric? - Yeah. - [Participant 4] Because you kinda start this interesti

ng observation which is that climate migration often means displacement within a country, not across countries. And we see this kind of pattern. And then we see some important shifts in this pattern of migration from Central America to the United States, but of course it's not new. And so I'm kind of curious to see how you see the expansion of the dry quarter or the increased severity of the dry quarter fitting in with that longer term pattern. - Yeah. I mean I think the real answer is we don't

know yet. But there's a lot of discussion, I don't even wanna say debate, but like discussion about climate migration in both policy and academic circles. And kind of, if you think of like a spectrum, like natural disaster displacement versus what we think of as traditional economic migration, where does climate migration fall on that? And I think that the answer is gonna end up being it depends. It depends on like the type of... So if it's a one-off hurricane, even if that hurricane is more lik

ely because of climate change, it's probably gonna more resemble natural disaster migration. If it is affecting your ability to see into the future a way to earn a living where you are, then it's going to lead to a more long term... So natural disaster migration is usually very short in both space and duration, right? People don't go very far and they tend to come back, at least large numbers of them. Whereas this seems, especially if you're taking your kids with you, like a more permanent form

or at least attempted permanent form of migration. But I think it's going to vary a lot based on geography of a country. And literally geographic size, not population size, right? So if you have a large landmass, there are more places where people... There's more variation in climate impacts and more places where people can go internally to be free from or at least less affected by climate change. I think also one of the things that's very hard to test, I don't think we can test it with these da

ta, maybe with survey data we could. What I'm seeing right now, I can just mention it very briefly. The data we're getting from Nicaragua, one of the questions that we asked, we surveyed like 700 people. One of the questions that we asked was has anyone from your household migrated in the last three years? We asked that question both externally and internally. The external number is way higher than the internal number. And so I think that in some cases, it's just going to depend on the attractiv

eness of internal migration options, which I would argue is actually lacking for many reasons in Central America, including in El Salvador and Honduras, really high levels of gang activity in the cities, but also just Central America's going through the youth bulge right now. So there's not a lot of extra economic opportunity floating around if you do move from the rural to the urban areas. But in other countries, climate might affect them at a different time in their development trajectory. The

y might have better institutions in place, they might have lower levels of violence. And so those internal options just might be more attractive. It's not clear to me where you go in some of these places to escape climate change. Like you're on a narrow strip of land with oceans on both sides of you. Yeah. Yeah. - [Participant 5] But this also takes a long time because people who are in drought for years and years and years go to cities because there's nowhere else they can go. And then they sta

y there for a while hoping for the best that something will give somewhere. - Yeah. I mean I think this is- - [Participant 5] It's not sort of straight to the US border- - Except that that's what we're seeing right now in Central America. And this actually is probably a way to slightly answer your question too. I think the very fact, as you said, that this is not new migration, the fact that those networks already partially existed has been what has allowed this to happen. So like Migration Poli

cy Institute did a survey of people coming from Huehuetenango in Guatemala who had migrated. And almost none of them had attempted internal migration first because they know what their cities, they know what the opportunities are like in their cities, but they also know from previous waves of migration what the opportunities are like in the United States. And we do know just from the migration literature that migrants tend to follow migrants, right? And so I do think the very fact that some of w

hat we're seeing in Central America is not doing that internal migration. Because I think you're right. In many situations that is what we see. Maybe even historically in Central America, that is what we see. I don't have the data to look into that, but it's not what we're seeing right now. And that I think was a surprising pattern 'cause we were not sure what would show there, but it really does seem like within a very short space of time of having these stress, you're moving out. So yeah, we w

eren't sure. But yeah, we were a little surprised by it too. Yeah. - [Participant 6] So I have two questions. And this was really, really interesting stuff. So I have two questions. Some of the early work from (indistinct) and subsequent work from Dave Donaldson that emphasizes how connection to a free market can be an antidote in the face of some kind of natural crisis like this, unless there's some actor intervening, preventing people from getting access to aid that they might otherwise be abl

e to get on a free market. So I'm really curious about what might be the barriers. In this case it seems like the big barrier is a lack of precipitation or wrong amounts of water at the wrong times. What are the ability of these small holder farmers to use the political system to advocate for water rights and that would maybe preclude the necessity to move? And my second question relates to the sort of... Okay, my second question is if we are thinking about climate change as the problem here, if

that's the root of the problem and we're only expecting that to get worse, are we just buying a little bit of time by introducing new agricultural techniques or are we just delaying the inevitable? Or do we need to look for other solutions other than helping these people stay in place? Or is this going to be enough for them to stay in place even as climate change hits- - Yeah. I'm smiling 'cause a high ranking US government official asked me that question like two weeks ago. (all laugh) The sec

ond question. (laughter continues) All right. So on your first question, there are kind of varying levels of political institutions in Central America, right? In terms of ability to participate in your political system. Corruption is very, very high across these countries. Governance scores are not particularly high. It's interesting, El Salvador has an extraordinarily popular president who is fairly authoritarian in his tendencies, including he wasn't supposed to be able to run for reelection a

nd ran for reelection, right? I mean huge. But people really like him, right? And now both Guatemala and Honduras have recently elected new administrations. And in the Guatemala case, with very strong support from the indigenous communities. That has historically not been the case, so it's actually kind of interesting to see what happens going forward. The lifespan of environmental activists in Central America is not particularly high. Like there's really well known cases of people just getting

killed, right? By being especially indigenous environmental activists. So is that a measure of ability to participate politically? I'm not sure. Also people being displaced from their land. There's a lot... I would not say that farming communities have particularly strong influence on political outcomes. So InSight Crime has done studies on like municipal level, where lots of mayors of towns are on the payroll of the drug cartels that are moving things through. So I'm not sure exactly what recou

rse you have. I'm in the process actually of trying to... (laughs) So far not successfully. Trying to work with survey data from the LAPOP survey based in Latin America, but the America Barometer Survey for these three countries. I have to do things with the data that I don't quite know how to do yet in order to make them look representative of the populations. But that will allow me, assuming that works out okay, to look at their questions that have things like faith and municipal governance an

d other things like that and see if that affects here. On your second question, I mean it's a really good question. Is this just a bandaid? I guess my immediate reaction is even if it's short term, it still might be worth doing because no one wants these people right now, right? Like I don't know what the alternative for these people is besides having an option to grow things on their farm. Because the United States is not going to revise their migration policies anytime soon. Mexico has become

increasingly tough on migrants coming. And one of the interesting things that I find is I've given talks like this to areas in the US that are particularly rural. This resonates with US farmers, many of whom are actually on the right side of the political spectrum. I actually think there could be really strong political support for helping smallholder farmers in Central America, especially if you can tie it to migration. I don't know what the other options are for people right now because there'

s not a lot of... There are not a lot of job opportunities within these countries. They're not ripe for a huge amount of urbanization with lots of growing jobs. Even within urban areas, people get an education and struggle to find a job that is commensurate with the education level that they have. So even if it were just to allow the countries to kind of get through the youth bulge, get to another stage on their own development trajectory and then have people moving in from rural areas. But on t

he flip side, it is possible that some of these areas will be able to farm. And shifting crops, looking at more commercial crops, looking at farmer co-ops that would allow them to produce in larger quantities so that they can have better negotiating power with markets. All of these things are possible, but it might be growing different things than what they have grown before. But it is not true that it is impossible to grow things there. It is true that you need to use different agricultural tec

hniques and maybe even shift crops. So I guess this is one of the things that I love about things about international development. And as countries develop, more people move to the cities and less people are in rural areas, but somebody has to grow food. And most countries want to have some level of domestic food production. And so I do think there is a role for this even long term, but it might not look the way that it does now. I do think that even short term, something that allows people to f

eed their kids who don't have any place else to go is probably worth doing. Yeah. - [Participant 7] Thank you so much. This is very interesting. I'm wondering, so because it's so difficult to tease out climate as a driver of migration, especially if it's like a compounding factor, obviously there are implications for why you would want to identify climate migration as like a distinct situation in terms of like mitigating emissions or adaptation from sending locations. But do you feel that... Or

like what do you feel is the importance of distinguishing between migration generally and climate migration from like a US policy perspective or in receiving locations, let's say? - Yeah, I think it's a very good question. I think there are a couple things. I mean, from one perspective, it shouldn't matter, right? From one perspective, a migrant who is coming in is a migrant who is coming in. And then you can differentiate based on skill levels and other things, but most people would consider th

ese are lower education level migrants and so why should it matter whether they're coming for climate reasons or not? I guess I would say a couple of things. One is that climate impacts seem to be causing at least here, like big increases over a short period of time, compared to just like a normal life cycle of economic migration type of thing. So being able to prepare for large increases in migrants. And potentially, most of these migrants too didn't want to leave, right? Like they had chosen t

o stay where they were. And then they got hit by droughts and so they decided to leave. So that can allow us to think about early interventions in places where people don't want to migrate, where we could potentially say this looks like an area that is ready to potentially experience a big ballooning in climate migration, we could invest in them there. From a US perspective in terms of migrants coming in, and I think it's actually important to keep several things in mind, migrants on net are goo

d for economies, right? I try to be very careful about this 'cause this is actually a very... I don't think we should be going around saying, well, we don't care if they get driven out of their country 'cause they're gonna be good for our economies, right? Like that's not the message that I want to send. It's certainly not the message that resonates with the American people right now who are thinking that they want lower levels of migration. But the economic evidence is that migrants are actuall

y quite good for economies, become entrepreneurs and start small businesses and file patents at much higher rates than native born populations. And even when you think about impact on services and taxation within one generation, usually become net contributors to the system even though there might be a short time horizon where that's not true. It's also much more true if they're allowed to work and contribute to the economy. So changes that allow people to actually have jobs. And Columbia has be

en experimenting with this a lot with Venezuelan migrants. And that has made a huge impact on the ability of the Venezuelan migrants, not only to support themselves but also to contribute economically to the country. So those are things that we might want to be thinking about. But there's also, when it comes to actual US policy, changes in US policy that even if we kept the same number, even if we kept the same number of visas, started thinking about allocating them based on need that included c

limate, would be a potential way to decrease... So one of the number one ways to decrease quote "illegal migration" is to increase the pathways those people have to come legally. We saw this historically with Mexico when the US changed migration policy toward Mexico. And so we know that this can work. It's also true that because US migration kind of quotas are based on population, and these are relatively small countries, they get very few legal entries into the United States. The idea that this

number of people could kind of go back and wait in line, their grandkids would still be waiting in line. Like there's just not enough legal opportunities per year. But then there's other countries that don't even reach their quota because they're like in Africa or something and not even trying to send a lot of migrants to the US. So thinking about US policy being directed toward legal... Rethinking how we calibrate legal migration across countries might actually be somewhat helpful in this case

. Yeah. Do we have time? Okay, yeah. Okay, excellent. - [Participant 8] Thank you so much. This was really fascinating stuff. I'm just trying to wrap my head around the differences between the countries or the lack thereof. - Yeah. - [Participant 8] Because I guess if this is a story about people learning or getting networks internationally and moving directly to international migration rather than going to the cities, or if this is a story about different crops being affected differently, or di

fferent types of governance and trust in governance. I would expect those things to be somewhat different across countries, if not even across regions. And therefore I would expect more difference across countries than we're actually seeing. Right? So I would just love to hear your thoughts on it. - Yeah. So actually, let me just push. So say a little bit, what differences would you expect? - [Participant 8] Not knowing the context enough, I would expect that agricultural stress might affect dif

ferent crops differently. - Okay. - [Participant 8] And that those crops are spread differently across the countries or the regions. I would expect that gaining networks in the US rather than in a city close by in the same country if nothing else would happen at different times in different countries. And I would also expect that at least marginally, there's some difference on the extent which people trust local government or something like that. And I think that's also part of your story. - Yea

h, I mean it's actually... The last part is not actually part of a story that I can tell yet. I actually hope it will become part of a story we can tell, but I don't know what it will show. It's one of the things I wanna test, I don't know what the answer will be. I actually think that the lack of finding differences here is part of the story. So a couple of things. So these three countries, I mean they literally share borders, right? So they're located very close to each other and within reason

able proximity, like similar proximities to the United States. All three of them have histories of migrants going to the United States, and of return migration from the United States, right? So I don't know that there's a huge difference like that way, but one of the things that really strikes me, if it were the migrant network story, like if that were the main thing driving this, that didn't change in 2014, right? Like that's been true since the 1980s or so, this kind of back and forth migratio

n between the US and Central America. Nothing magical happened in that in 2014, what happened was there was a drought, right? And the same thing with 2018, what happened was there was a drought, right? And so the thing that did change was this huge agricultural stress effect that really hit this dry corridor that runs across these three countries. So I actually think that that lack of difference is part of the story. That they've got this similar climate... Do they grow different crops? Absolute

ly. And is climate change affecting the different crops differently? Absolutely. So one of the... A department in Guatemala that had that biggest level of out migration, a big part of that is a coffee growth story, which is different than the story I was saying, but it's still a climate impact story and it's still this kind of same drought that was allowing diseases to flourish there that hadn't flourished there before. So yes, it's not the same impact but the same weather phenomenon are affecti

ng different crops differently but still affecting them at the same time, right? And so I think that's actually, like I said, I thought El Salvador wouldn't show, so that to me was like, oh wow, something really is going on here that is not as country specific but is actually being driven by this common phenomenon that hit across there. - [Moderator] So the last question I think from the audience here on Zoom. So they said from Christian Espinoza, he said, you mentioned the Guatemalan indigenous

communities a few times. He's currently attending from an indigenous community in Huehuetenango, where he's carrying out field work. Will this research show a strong correlation between agricultural stress and migration? And I'm wondering if you could comment on the mechanism of climate induced migration. In his research, there is no direct link aligned between crop failure and migration, but a set of complicated feedback loops on local household economies. And one something which is a bit rela

ted to that is that if you know, what is the level of US aid government funds' aim at ethnic problems? And then another question. - Okay. So first of all, I'd be happy to follow up offline, 'cause I'd like to know what they're looking at in order to kind of look at those mechanisms. Because one of the reasons that we actually did wanna just look at the raw impact, like the actual climate impact, is because it's very difficult to trace through all of these things. So we literally wanted to say if

there's an impact in these regions, do you know an agricultural impact? Do you see an increase in migration? I will say I didn't show the results here. I think they are somewhat less interesting, but I still think they're important. We have put department fixed effects in this regression or in these regressions. One of the reasons I say they're less interesting is because some of our variables don't vary over time, so then they drop out. Really the only ones that stay in are agricultural stress

and homicides. But the results are still quite strong, so this does seem to be a difference across time within departments that we're picking up. And the survey evidence that I've seen that has come out of places like Huehuetenango, has shown that large numbers of the people who do migrate from there do mention impacts on agriculture for why they are migrating. But I'd have to know a little bit more about what they are finding. But in terms of, what was the second question? - [Moderator] The US

A if you know that... If their governance funds are aimed at ethnic problems as well. - I'm sure they say that they are, whether or not... So one of the things that a lot of international NGOs and advocacy groups have really been pushing the US government to do more of is to work directly with local organizations and they don't. And so they will contract with like a big group that will then go in and work with local populations. And I'm sure they'll put language in there in order to win the cont

ract that says that they will... But the US government is not really monitoring this, right? Is not really checking to see if this is being disproportionately targeted to indigenous or ethnic minority. I mean I wanna be very careful. In Guatemala, indigenous population is almost the majority of the country. So I don't even wanna say ethnic minority communities. USA does not go to the central governments, but it usually goes to large scale organizations that then can distribute it subnationally.

The US government agencies are not working directly with the local partners. And so I don't know. - [Moderator] And last question from Gina Davis. She said El Nino events historically vote for below average precipitation in Central America especially in the tri-corridor, and below average crop yields for staples such as maize in the region. The 2014/15, and 18/19 years in your research highlighted as high stress years were El Nino years. But climate change is exacerbating and changing climate mo

ld such as ENSO and El Nino. Did you notice in your research whether there was a longer term correlation looking back before to 2010, between migrant peaks and El Nino years? - That's actually a really good question. We have not gone back historically. I will say that this year the projections for agricultural yield are down again, which would be in line with that. Whether it will lead to an increase in migration, obviously I don't know yet, but there is this kind of strong trend. You know what'

s something that's really interesting, I'll end on this just kind of interesting side anecdote. I was at a presentation that I thought was gonna have nothing to do with anything that I do, but you just go to presentation sometimes, right? And the person ended up talking about the ocean oscillation weather patterns like off of the British Isles and the impact that that has on agriculture in the UK. And it sounded very similar though obviously happening in the developed countries, so the impacts o

n people are not the same. But very similar to the types of things that we see with the El Niño southern oscillations in America. So I was like, oh wow. So it might be a neat thing to explore going forward is thinking about other places where we would expect similar climatic patterns and what we might see there in terms of, I shouldn't say similar climate, but this idea of these rotating annual types of patterns and what impacts those might have on both crops and then on migration. Thank you- -

[Moderator] Thank you very much for... (audience applauds)