climate change and migration case study

• Biodiversity
• Cities & society
• Land & water
• All research news
• All research topics
• Learning experiences
• Programs & partnerships
• All school news
• All school news topics
• In the media
• For journalists

How does climate change affect migration?

April 2021 saw a 20-year high in the number of people stopped at the U.S./Mexico border, and President Joe Biden recently raised the cap on annual refugee admissions. Stanford researchers discuss how climate change’s effect on migration will change, how we can prepare for the impacts and what kind of policies could help alleviate the issue.

In the face of a mounting humanitarian crisis at the U.S./Mexico border, the Biden administration has acknowledged climate change among the powerful forces pushing migrants from Central America. A $4 billion federal commitment to address the root causes of irregular migration acknowledges the need for adaptation efforts to help alleviate the situation.

The challenge is not limited to the border. Last year, weather-related disasters around the world uprooted 30 million people – more than the population of the 14 largest U.S. cities combined – and wildfires displaced more than a million Americans, according to the Internal Displacement Monitoring Centre.

Below, Stanford climate and behavior experts discuss how climate change’s effect on migration will change, how we can prepare for the impacts and what kind of policies could help alleviate the issue. The researchers include  Chris Field , a climate scientist who has led UN Intergovernmental Panel on Climate Change efforts to analyze climate-related risks, impacts and adaptation opportunities;  Gabrielle Wong-Parodi , a behavioral scientist who studies how people react to challenges associated with global environmental change;  Erica Bower , a PhD student who studies human mobility in the context of climate change impacts and served as a climate change and disaster displacement specialist at the UN Office of the High Commissioner for Refugees;  Nina Berlin Rubin , a PhD student in Earth system science whose research focuses on decision-making in the face of climate extremes.

What is a common story of climate migration in the U.S./Mexico?

Bower : Climate change is a threat multiplier – it can exacerbate economic insecurity or political instability, which in turn may lead to migration. In the “dry corridor” of Central America, for example, climate change extremes such as droughts may hinder crop production. Without a consistent source of food or income, a farmer may seek other livelihood opportunities in a nearby city or further north. When combined with poverty or violence, a drought may make the perilous journey north seem to be a more promising adaptation or survival strategy.

How has climate change’s effect on the tide of refugees in the U.S./Mexico border changed in recent years?

Bower : In recent years, climate change has made extreme weather events stronger and more frequent, which may contribute to migration decisions. However, given the multiple reasons why people move, we do not have the evidence required to say with certainty precisely how climate change has affected net migration flows to the U.S./Mexico border.

What does the future hold for climate-related migration at the border? What should we expect, and how should we prepare for it?

Field : In the long run, stopping climate change is a key element of getting the situation under control. In the short run, there are lots of effective ways to decrease vulnerability. These range from improving agricultural practices to strengthening social safety nets. Across the full suite of possible vulnerability reduction measures, it is critical that the solutions are implemented in partnership with local communities and not imposed on them. In regions with high levels of corruption or political or criminal violence, it is much more challenging to make progress on vulnerability reduction.

Bower : We may need to change our approach to welcoming people across our southern border. The definition of refugee in international law is very narrow, and most people fleeing in climate change contexts – including from Central America to the U.S./Mexico border – are generally not recognized as refugees under international or domestic law. Newly proposed legislation in Congress would protect the human rights of people fleeing to the U.S. in the context of climate change.

Leaving is only half the story. Tracking these migration pathways from origin to destination can speak to whether people are moving to safer areas or to areas that introduce new or heightened risks. ” Nina Berlin Rubin PhD student, Department of Earth System Science

How extensive is climate change’s effect on migration within the U.S.?

Wong-Parodi : We are seeing some evidence that people who feel more impacted by wildfires and secondary impacts like smoke are more likely to intend to move to a new state within the U.S. or even out of the country. The question that remains is whether the destination they are planning on moving to is more or less at risk for wildfires or other climate hazards.

Berlin Rubin : Exactly. One might decide to uproot their family and leave behind friends and neighbors in search of respite from wildfires and smoke, only to find themselves living somewhere with just as much wildfire risk – or exposure to other risks like flooding – as where they were before. And the reality is, as these fires become more frequent and destructive, people will have fewer options and less confidence that their destination is actually going to be free from wildfires.

Bower : We are also seeing that many communities are seeking government support to relocate away from sea level rise or flooding. A recent mapping exercise identified 36 cases of community-wide relocation in the U.S. alone since 1970.

How might federal and state policy lessen climate change’s impact role as a driver of migration at the border and within the U.S.?

Field : The U.S. Federal government can play a large role in addressing the core drivers of displacement, through three main pathways. It can decrease its contribution to warming through decreasing its greenhouse gas emissions. It can help other countries decrease their emissions through financial and technical assistance. It can also help poor countries adapt so that their people feel less pressure to migrate. This last pathway has the potential to be especially cost-effective in the near term, even though it cannot address all of the drivers of migration.

Is there any promising research/data gathering that could dramatically change how we understand and react to this challenge?

Field : It is likely that progress in decreasing pressure to migrate will require patience. One area where data can make a big difference is in cataloging investments in adaptation and following their consequences. At this point, we are not systematically cataloging and tracking adaptation efforts. A robust database can be the foundation for a new generation of evidence-based adaptation.

Berlin Rubin : Research at the individual level will help us better understand and characterize the psychosocial and experiential factors that motivate climate-related migration decisions. But leaving is only half the story. Tracking these migration pathways from origin to destination can speak to whether people are moving to safer areas or to areas that introduce new or heightened risks. We need to understand the local context in which people are making their decisions in order to get ahead of these challenges.

Field is the Perry L. McCarty Director of the  Stanford Woods Institute for the Environment , the Melvin and Joan Lane Professor for Interdisciplinary Environmental Studies, a professor of Earth system science and biology and a senior fellow at the  Precourt Institute for Energy . Wong-Parodi is an assistant professor of Earth system science in the  School of Earth, Energy & Environmental Sciences ; and a center fellow at the  Stanford Woods Institute for the Environment . Bower is a PhD student in the  Emmett Interdisciplinary Program in Environment and Resources  at Stanford’s School of Earth, Energy & Environmental Sciences.

Related research

How does climate change affect disease.

As the globe warms, mosquitoes will roam beyond their current habitats, shifting the burden of diseases like malaria, dengue fever, chikungunya and West Nile virus. Researchers forecast different scenarios depending on the extent of climate change.

The effects of climate change on water shortages

In Jordan, one of the most water-poor nations, predictions of future droughts depend on the scale of climate change. Without reducing greenhouse gases the future looks dry, but researchers offer some hope.

Media Contacts

Chris Field Stanford Woods Institute for the Environment (650) 823-5326;  [email protected]

Gabrielle Wong-Parodi School of Earth, Energy & Environmental Sciences (650) 725-6457;  [email protected]

Erica Bower School of Earth, Energy & Environmental Sciences (203) 666-9892;  [email protected]

Nina Berlin Rubin School of Earth, Energy & Environmental Sciences (307) 690-4234;  [email protected]

Rob Jordan Stanford Woods Institute for the Environment (650) 721-1881;  [email protected]

Explore More

Addressing the Colorado River crisis

The future of water in the Southwest was top of mind for participants and attendees at the 10th Annual Eccles Family Rural West Conference.

• Freshwater resources

Hurricane risk perception drops after storms hit

Programs and policies that help households go beyond stocking up on food and medical supplies to invest in longer-term protections could overcome the risk perception gap and support adaptation to rising climate-related threats.

• Human behavior

Looking back on the first year of Stanford Ecopreneurship

A new report shares key accomplishments from the inaugural year of Stanford Ecopreneurship programs, a collaboration between the Stanford Doerr School of Sustainability and Stanford Graduate School of Business.

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• HHS Author Manuscripts

Climate Change and Migration: New Insights from a Dynamic Model of Out-Migration and Return Migration

Barbara entwisle.

University of North Carolina at Chapel Hill

Ashton Verdery

Pennsylvania State University

Nathalie Williams

University of Washington

In popular accounts, stories of environmental refugees convey a bleak picture of the impacts of climate change on migration. Scholarly research is less conclusive, with studies finding varying effects. This paper uses an agent-based model (ABM) of land use, social networks, and household dynamics to examine how extreme floods and droughts affect migration in Northeast Thailand. The ABM explicitly models the dynamic and interactive pathways through which climate-migration relationships might operate, including coupled out and return streams. Results suggest minimal effects on out-migration but marked negative effects on return. Social networks play a pivotal role in producing these patterns. In all, the portrait of climate change and migration painted by focusing only on environmental refugees is too simple. Climate change operates on already established migration processes that are part and parcel of the life course, embedded in dynamic social networks, and incorporated in larger interactive systems where out- and return migration are integrally connected.

Many people foresee the ongoing environmental challenges stemming from global climate change as an important and growing influence on human migration. According to a 2010 Gallup poll estimate reported in the New York Times , 500 million people across the globe expect to move in the next five years because of severe environmental problems ( Benko 2017 ). If climate-related environmental problems induce such mobility, it would represent a fundamental shift in global migration systems and settlement patterns. To contextualize the magnitude, consider that estimates suggest there were 244 million international migrants living outside of their country of birth in 2015 ( United Nations 2015 ) and that an additional 763 million internal migrants were living within their birth country but outside of their birth region in 2005 ( Bell and Charles-Edwards 2013 ). Though expectations do not always result in actual migration ( De Jong, et al. 1985 ; De Jong 2000 ; Coulter, Van Ham and Feijten 2011 ), the possibility that climate change may influence migrant flows equivalent to half of the world’s current lifetime migrant stock is an extraordinary proposition. The media, the general public, and policy makers around the world are increasingly attuned to this unfolding story. Thus, it is not surprising that narratives of climate change impacts on migration in the popular literature – and the images in the public’s eye – are dominated by stories of environmental refugees: people pushed out of their homes and livelihoods by desertification, warming and rising seas, and extreme weather events. The mayor of a town in Bolivia sitting in a boat on the dried-up bed of a lake that used to provide his livelihood; a young woman pushing through water waist-deep as she tours her neighbor’s flooded taro plot in Kirabati; the halving of storm-ravaged New Orleans’s population; and the recent decimation of Puerto Rico offer dramatic examples of migration responses to extreme weather events 1 .

In contrast, the scholarly literature has produced a more nuanced picture relating climate change to migration patterns ( Gioli et al. 2016 ; Hunter, Luna and Norton 2015 ; McLeman 2014 ). Several studies document clear migration responses to climate change (e.g., Bohra-Mishra, Oppenheimer, and Hsiang 2014 ; Bohra-Mishra et al. 2017 ; Dillon, Mueller, and Salu 2011 ; Feng, Kreuger, and Oppenheimer 2010 ; Marchiori, Maystadt, and Schumacher 2010; Nawrotski et al. 2016 ). But in others, response depends on resources (e.g., Call et al. 2017 ; Kubik and Maurel 2016 ; Loebach 2016 ), livelihoods and the availability of adaptive responses in situ (e.g., Morrissey 2013 ; Thiede and Gray 2017 ), and perceptions of the nature of the problem to begin with (Koubi, Stoll, and Spilker 2016). Some studies find no migration response at all ( Entwisle et al. 2016 ; Loebach 2016 ). This could happen because climate change operates on existing migration patterns that already incorporate adaptations to adverse conditions, with new changes to climate exerting little additional pressure, especially if they are incremental rather than catastrophic. However, all of this work, as with nearly all research on climate change and migration, focuses exclusively on those who leave, with little attention to the migration systems in which they are embedded. In this paper, we contribute to this growing literature by looking at migration through a different lens, asking about potential effects of climate change on all parts of the process, including return migration. In doing so, we consider new ways to connect the literature on climate change and migration to migration theory.

Climate change is disruptive ( Call et al. 2017 ). It disrupts processes that are already dynamic inasmuch as migration is related to life course transitions, evolving household strategies, and endogenous changes in social networks at and between places of origin and destination. Typically, studies of climate change impact focus on push-inspired out-migration in response to a single event or series of events ( Hugo 1996 ; Massey, Axinn, and Ghimire 2010 ). This is only one possible outcome, however. Also of interest is the possibility of otherwise expected migration that does not occur, whether it be among those who might leave an origin or among those who may return, and indeed, the balance in return and out- migration in the overall system. By incorporating a life course perspective, including a focus on out- and return-migration associated with the transition to adulthood, and embedding it in a systems approach, we identify a type of climate change impact that heretofore has not been considered: disruptions to established and expected streams of return migration. To our knowledge, we are the first to examine climate change impacts of this nature.

To investigate how climate change may affect migration systems , we study the interrelationships between climate change and the linked processes of out- and return-migration through application of an agent-based model (ABM). Our ABM incorporates land use, social networks, and household dynamics in a specific setting: rural villages in Northeast Thailand where rainfed agriculture is the dominant livelihood strategy. The ABM is grounded in decades of empirical survey and ethnographic data from the Nang Rong Projects and includes detail on demographic behaviors—including out- and return migration—and spatial–environmental processes relating to agriculture, land use, and weather patterns. That said, our intent is neither to reproduce the past, nor to predict the future, but rather, to use the model to explore the implications of the theories embodied within it.

The ABM explicitly models the dynamic and interactive pathways through which a climate- migration relationship might operate. All behaviors and processes occur in an interdependent system, linked by social networks and constrained by available land. Rather than assuming a direct “climate effect” on migration, we specify mechanisms of influence on migration through climate effects on crop yields, and thus livelihoods, in these agrarian communities. Importantly, we experiment on this social system in a counterfactual framework, comparing ceteris paribus migration responses in situations with and without the types of extreme weather events posited to grow in frequency in tandem with ongoing climatic change. Results from our simulations of extreme floods and droughts suggest minimal impacts on out-migration, but marked negative effects on return, i.e., migrants otherwise expected to return do not. Our results also point to the important influence of social networks for both out- and return migration, and they demonstrate the need for greater attention to migration theory in models of climate- migration relationships.

Approaches to migration research

The focus of this paper is rural-urban migration in Northeast Thailand. Typically, research on rural-urban migration in lower and middle-income countries draws from economically focused theories, mostly about labor migration ( Massey et al. 1993 ). Other common theories that are often interwoven in migration studies focus on non-economic aspects such as social networks. While the literature on internal migration and residential mobility in high-income countries increasingly views migration in terms of life course transitions (e.g., Falkingham et al. 2016 ; Warner and Sharp 2016 ), these ideas are not often consulted in studies of migration in lower and middle income countries. We integrate these approaches to migration research in our agent-based model. We review each briefly, indicating the potential impact climate change might have in each.

Economic perspectives on migration

Economically focused theories of migration generally fall into one of two broad categories: the neoclassical model and the new economics of labor migration (NELM) model. According to the neoclassical model, the decision to move is made by individuals based on a comparison of economic opportunity at the place of origin and potential destinations ( Harris and Todaro 1970 ). If opportunity is greater elsewhere, such that that the benefits of migrating outweigh the costs of doing so, the neoclassical migrant will move to that place. To the extent that climate change impacts at the origin tip the balance of opportunity in favor of one or more destinations, the neoclassical perspective predicts that it will stimulate out-migration. Migration in the neoclassical model is considered permanent, until circumstances necessitate another permanent move. Return migrants in this model are those who fail in the destination in terms of employment or earnings ( Cassirino 2004 ).

Whereas individual profit maximization motivates migration in the neoclassical model, risk minimization guides decision-making in the new economics of labor migration (NELM), which specifically emphasizes joint decision-making in small and socially relevant groups such as families and households. In this approach, rural households facing uncertain prospects due to the risks of agricultural production in conjunction with underdeveloped capital, credit, and insurance markets may reduce risks by diversifying their economic portfolios ( Massey et al. 1993 ; Stark and Bloom 1985 ; Taylor 1999 ). One strategy is to diversify spatially, sending one or more household members to another area, subject to a different set of risks than at the origin. For a variety of hotly debated reasons ranging from altruism, exchange, insurance, investment, and normative behaviors within social networks (see Rapoport and Docquier 2006 , Carling 2008 , and Garip, Eskici,, and Snyder 2015 for reviews), these migrants may remit earnings back to the household and may return when they have reached a target amount ( Garip 2012a ; Garip 2012b ; Tong and Piotrowski 2010 ). In this theoretical model, migration is intended to be temporary. To the extent that climate change increases risks in places of origin, absent changes in costs, it will motivate out-migration, although only if other diversification strategies are unattractive, unavailable, or not sufficiently successful. The impact of climate change on return migration in this model has not been well considered. If climate change does not also affect the place of destination, the return behavior of target migrants may be likewise unaffected; the benefits of spatial diversification are not changed. Alternatively, if strategies to reduce risks of agricultural production in the face of climate change impacts in the place of origin become more costly and/or the risks at places of origin are increased, return migration could be delayed. Assumed but unstudied in this theoretical tradition is that conditions on the ground in places of destination or origin are known; how such information is transferred is the subject of many social perspectives on migration.

Social perspectives on migration

Theories of the social forces that influence migration are common ( Massey et al. 1993 ) but rarely consulted in the climate change and migration literature. We argue that these theories can help to better understand both out- and return migration processes in the face of climate change. We seek theoretical guidance from them here and incorporate them in our ABM as described below.

Whether migration is motivated by profit maximization or risk minimization, it is a costly endeavor. Decreases in the costs of migration will increase migrant flows, and vice versa. This issue is where social networks and associated theories of cumulative causation enter the discussion. By providing information or real help with finding a job or a place to live, social network ties to prior migrants reduce migration costs ( Massey et al. 1993 ). As more and more migrants move to a destination, and ties between origin and destination densify, migration costs decrease and it becomes easier for those who follow, at least up to a point ( Garip 2008 ; DiMaggio and Garip 2011 ). Migration becomes less selective as this process unfolds ( Garip and Curran 2010 ). This is the cumulative causation theory of migration ( Massey 1990 ; Massey et al. 1993 ).

Social networks within places of origin may also be relevant, in ways not usually considered in the migration literature. Social networks within places of origin comprise the sets of ties linking members of origin communities to one another, directly and indirectly, through kinship, acquaintance, exchange, and information sharing. They can be important, for instance, because they condition information flows about migrants’ activities in destination, enabling access to knowledge about migration opportunities among those not directly connected to migrants. Such information flows can affect migration above and beyond a community’s ties to the destination. Evidence of this can be seen in tests of survey researchers’ ability to prospectively find and locate migrants in destinations, a capacity constrained by flows of information and knowledge. Such work finds strong effects of sibling network cohesion in the origin village on survey follow-up in migrant destinations, even net of numbers of direct ties from origin village to the destination ( Entwisle et al. 2007 ; also see Wantanabe, Olson, and Falci 2017 ). Further, origin-based social networks may augment material motivations to leave, by acting as the locus within which motivations stemming from feelings of relative deprivation and cultural valuations of migrants become salient ( Mouw et al. 2014 ). Additionally, social networks transfer and enforce obligations to kin at origin, enhancing motivations to remain or return. Just like ties to destination, ties to origin are also directly affected by prior migration patterns from that place ( Entwisle 2007 ). We address further dimensions of these processes below.

Another social perspective on migration—the life course perspective—is almost completely absent from the literature on migration in poorer countries. Yet, migration is part and parcel of the transition to adulthood in almost any country ( Rindfuss 1991 ). In industrial and industrializing economies, young people often move as they complete school, enter the labor market, change jobs, enter marital or other unions, and begin to have children (Bernard, Bell, and Charles-Edwards 2014b; Warner and Sharp 2016 ). The transition may be crisp and well sequenced, predictable as a unidirectional pathway, but frequently it is not. The life course structures the extent, timing, and interconnectedness of moves ( Clark and Withers 2008 ; Findlay et al. 2015 ; Geist and McManus 2008 ) as well as linkages with behaviors of others in the household and family. Migration is linked to other status transitions occurring during the early adult years, such as family formation. It also reflects obligations to kin and others in places of origin and destination. In many rural areas, key times in the agricultural cycle – planting, the harvest – are occasions where such obligations are negotiated as migrants return, or do not, and decide to remain, or embark, at the nexus of individual and household decision-making.

Out- and return migration as an interactive system

Although most research focuses on out-migration, it is not a singular independent process. Return migration is also common, although interestingly, its interpretation varies depending on theoretical approach ( Cassarino 2004 ). In the neoclassical model, returning migrants are perceived to be those who have failed to prosper in places of destination. In NELM, returning migrants are thought to have achieved their earnings target; they were only temporary to begin with. In the life course approach, young people who return are coming and going as part of the transition to adulthood. They are neither failures, nor successes, just “in process” ( Coulter, van Hamm, and Findlay 2016 ). They also respond to kin obligations in places of origin, possibly delaying their departure or hastening their return, while they assemble the resources needed to redefine these obligations as they enter a fully adult status, possibly in the destination but possibly not. The life course approach is focused on status changes through the young adult years, in the context of kin obligations as well as opportunities in origin and destination.

As typically conceptualized, climate change affects conditions in places of origin ( McLeman 2014 ). Climate change may occur broadly, at a global scale, but it can also act within as well as between regions and countries; further, there can be variability in its consequences for local environments (e.g., degree of flooding dependent on local topography). Accordingly, research has focused on the potential for climate change to increase out-migration. Indeed, frameworks for conceptualizing the environment- migration relationship view migration in terms of a decision to stay or go ( Black et al. 2011 ; Hunter, Luna, and Norton 2015 ; Massey, Axinn, and Ghimire 2010 ). Even when migration is viewed in terms of connections between sending and receiving areas, the emphasis is on a single direction of movement: out (e.g., Morrisey 2013 ).

Until now, the literature has overlooked the possibility that climate change may affect return migration. However, migration theory clearly articulates mechanisms by which this would occur, whether it be by altering labor shortage-remittance tradeoffs in the new economics of labor migration framework ( Rozelle et al. 1999 ) or by disrupting the role of out- and return migration as they relate to the transition to adulthood in the life course framework. To be clear, our interest is in comparisons of migration patterns given more and less climate change. This is quite separate from recovery migration, i.e., who comes back among migrants who left specifically because of a discrete catastrophic climate event (e.g., Hurricane Katrina; Fussell, Curtis, and DeWaard 2014 ).

In addition to directing attention to return migration as a potential outcome, we also draw on the idea that out- and return migration are not just singular behaviors; instead they are interconnected steps in a migration system. Life course, demographic, and social networks perspectives describe ways in which these behaviors are integrally connected into a system, as we now explain.

Out- and return migration are integral to the transition to adulthood ( Rindfuss 1991 ). There may be movements back and forth as young people fulfill their obligations to their natal family and accumulate the resources necessary to establish themselves as independent adults. There may be movements back and forth as ties to the natal family are redefined through the transition. Indeed, it is not unusual as part of this process for young people to return home before they leave again ( Coulter, van Hamm, and Findlay 2016 ); a growing literature on “boomerang kids” in wealthier countries shows the prevalence of this phenomenon ( South and Lei 2015 ; Sandberg-Thoma, Snyder, and Jang 2015 ). There is a tendency in the migration literature to view movements as planned, but this is not always so ( De Jong, et al. 1985 ; De Jong 2000 ; Coulter, Van Ham and Feijten 2011 ). Movements back and forth from places of origin occur within a larger system, but from the perspective of lived lives, the timing and duration of these moves is not always as planned or predicted. For example, migration initially intended to be temporary, part of a household strategy in the place of origin, can extend indefinitely as migrants stay longer and build ties to others in the destination ( Korinek, Entwisle, and Jampaklay 2005 ). It is not unusual for migrants to marry earlier than non-migrants ( Jampaklay 2006 ); potential partners are much more available in urban areas, especially when young migrants are housed together. Contingency is an essential element of the transition to adulthood as each pathway is structured by changing opportunities and constraints in multiple contexts in origin and destination.

The demographic concept of exposure to risk is another key. Out-migration creates migrants and it is these migrants who are exposed to the risk of return migration. Thus, increases in out-migration will increase the stock of people exposed to the risk of return and at the same time decrease the stock of people exposed to the risk of out-migration. The same is true for return migration: it drains the stock of migrants and increases the stock of residents of an origin area who are then exposed to the risk of out- migration. Most out- and return migrations (especially, but not exclusively, those emanating from poorer countries) are undertaken by young adults ( Rogers, Raquillet, and Castro 1978 ; Bernard, Bell, and Charles-Edwards 2014a ). Older adults are much less likely to migrate for any reason and thus should not always be considered as part of the stock of people who could be potential migrants during periods of climate stability or change. At the same time, older adults in origin areas serve as conduits of information or key connecting nodes in social networks. Depending on their needs, they may also represent sources of obligation for potential out- and return migrants.

The social networks perspective also explains a dynamic link between migrants, non-migrants, and return migrants. When out-migration increases, the stock of people who can provide information about migration increases and more information and other kinds of help become available. However, if the prevalence of migration becomes too high, this can disrupt the connectivity of a network, breaking the ties along which information flows. Likewise, if return migration is high, then this might create higher connectivity in a network, but decrease the number of people who possess active information about a destination.

Because out- and return migration are so intimately connected, if we want to understand the influence of climate change on out-migration, looking only at out-migration provides only half the story. The influence of exposure to risk predicts a fairly linear and inverse connection between out- and return migration. The life course perspective also predicts a relationship between out- and return migration patterns but suggests that the relationship might be more fluid and less planned and can change over time. The influence of social networks, with concern for linkages, nodes, and connectivity, predicts a relationship, but not in a linear or easily quantified manner. In any case, we expect that a model that can connect out- and return migration at the population level will provide new and different insights than a traditional regression model of individual migration behaviors during periods of climate change. We also anticipate that social networks will play a role in creating these new and different results by changing knowledge of on-the-ground conditions, among other things. In the remainder of this paper, we explore these ideas using an agent-based model.

Case study: Climate Change and Migration in Northeast Thailand

Consistent with much of the literature on climate change and migration, we take a case study approach. Our research draws on a multi-decade set of transdisciplinary research endeavors based in Nang Rong district, Thailand undertaken by a large team of sociologists, demographers, geographers, economists, and others from the University of North Carolina at Chapel Hill and Mahidol University in Bangkok ( http://www.cpc.unc.edu/projects/nangrong/ ). On the social side, the Nang Rong Projects consist of a series of community surveys and household censuses conducted in villages of Nang Rong district in 1984, 1994, and 2000, with additional follow-up surveys of migrants living in destination areas that were identified by origin households in the 1994 and 2000 survey rounds. Additional qualitative work spanning the 1980s to the 2010s was also conducted. The environmental side of the Nang Rong Projects is informed by detailed land use mapping surveys, remote sensing, and extensive Geographic Information Systems integration. 2 The location of Nang Rong is shown in Figure 1 . We focus on migration connecting rural villages in Nang Rong district, in Northeast Thailand to Bangkok, the Eastern Seaboard, and other urban areas in Thailand. We begin by setting the scene.

Location of Study Site

Nang Rong is about the size of a small county in the United States, occupying 1,300 km 2 . Historically, it was a settlement area for migrants from other parts of Thailand, with few established villages prior to World War II ( Entwisle et al. 2008 ). After the 1960s, as natural increase from the region’s demographic transition took off, net migration flows reversed and out-migration became a dominant feature of village life. Migration, both permanent and temporary, is common in Nang Rong ( Fuller, Lightfoot, and Kamnuansilpa 1985 ; Fuller, Kamnuansilpa, and Lightfoot 1990 ; Korinek, Entwisle, and Jampaklay 2005 ). At the time of this study, migration streams to Bangkok and the Eastern Seaboard were well established. Prospective empirical research on Nang Rong shows that, of villagers aged 10–19 in 1994, almost half (43%) were migrants six years later, in 2000; of those aged 20–29 in 1994, the comparable figure was a third (30%). Among former residents of Nang Rong villages in 1994, the same prospective studies show that 23% of those aged 10–19 and 19% of those aged 20–29 were no longer migrants in 2000. As cohorts age, those remaining in origin and destination are increasingly differentiated by the selectivity of the migration process, a process that has long been recognized in the literature (e.g., Lee 1966 ).

The economy in Nang Rong is largely agricultural, based on wet rice cultivation as well as upland crops such as cassava and other lowland crops such as sugar. Irrigation is rainfed, dependent on the annual monsoon. Rains arrive in the late spring to early summer, but precipitation is unpredictable in both timing and amount. Livelihoods depend on these features of the rain’s arrival. This is a challenging environment in which to succeed economically ( Bardsley and Hugo 2010 ) and most farmers live “on the edge.” Figure 2 superimposes the locations of villages (2a) and dwellings (2b) on a rare satellite image of Nang Rong while flooded during the annual monsoon (rare because clouds obscure images during these months). Villages line up along the rivers; dwellings likewise hug the borders of the flooded areas, with arrows pointing to their inundated fields. There is a delicate balance between too much water, and too little, confirmed in interviews with farmers on multiple occasions. Different rice varietals require different amounts of water, whose availability can be difficult to predict (traditional irrigation systems dominate in the region). The anticipation of water availability also influences the choice to grow crops other than rice: with more water, sugarcane is preferable; with less, cassava. Farmers report having “tried many things just to survive” (village interviews in 2010). Because of the marginal environmental setting, combined with the frequency of inundation and droughts, Nang Rong households have likely adapted already to extreme weather events, at least to some extent. For instance, empirical work finds no relationship between village water shortages and migration or remittance behaviors at the household level ( Garip 2014 ; Garip, Eskici, and Snyder 2015 ), although other work at the individual level shows a small negative relationship between village water shortage and out-migration ( Garip 2008 ).

Panel A: Location of Villages Relative to Area Inundated by Monsoon, July 2000

Panel B: Location of Parcels Relative to Village Centers on Inundated Landscape, July 2000

We are interested in the consequences of floods and droughts for migration processes in this setting. Rice is the region’s major crop, for subsistence and sale, and rice cultivation is a key component of our model. With rice, the timing of planting and harvest are organized around the beginning and end of the monsoon season, typically June and December ( Naylor et al. 2007 ). If the monsoon comes early, and is heavy, farmers may have to replace their crop, replanting, possibly more than once. They report: “if you fail twice, you would not try again for rice” (village interviews in 2010). If the flooding is less extensive, they may lose part of the crop. If the monsoon comes late, and is light, fields are less productive. It is in this way that monsoonal variation, and in particular, the pattern of floods and droughts, connects directly to the livelihoods of Nang Rong farmers. This was described to us again and again in fieldwork based in the district. Other crops, too, depend on the monsoon for irrigation, although these other crops are typically planted in plots less suited for rice cultivation. Our model includes cassava and sugar as alternate crops.

Consistent with expectations based on climate change, average rainfall amounts in Nang Rong have declined over the 20 th century; equally important, annual variance in rainfall has increased ( Bardsley and Hugo 2010 ). See Figure 3 . What if these extremes were to increase in magnitude and duration, as anticipated by climate models? What would be the impact on migration processes? To answer these questions, we ran experiments on an agent-based model (ABM) of land use and household dynamics developed for the Nang Rong setting. The focus is on a magnification of trends already underway, not catastrophic climate events that make entire regions uninhabitable (e.g., sea level rise, or devastating hurricanes). The ABM is grounded in the wealth of quantitative, qualitative, social, spatial, and environmental data available from the Nang Rong Projects ( Walsh et al. 2013 ), and code and documentation for it are available from the Carolina Population Center. 3

Trends in Annual Precipitation: Nang Rong

Agent-Based Model

An ABM simulates a population of autonomous, heterogeneous agents that interact with each other and their environment according to a set of prescribed rules (Bruch and Atwell 2014; Macy and Willer 2002 ). The dynamic actions of agents at the micro-level and their responses to the behaviors of other agents and characteristics of their environment result in regularities or emergent patterns at the macro-level. In the literature, ABMs range in complexity from highly simplified “toy models” such as Schelling’s (1971) account of residential segregation to models attempting to replicate many of the features of real landscapes ( Bruch and Atwell 2015 ; Manson et a. 2020). In some cases, the purpose is to elaborate a theory, or theories (e.g., Centola, Willer, and Macy 2005 ; Centola and Macy 2007 ), whereas in others, the purpose is to plan for alternative futures (e.g., Naivinit, et al. 2010 ). Although we start out with highly detailed information about individuals and households situated in a real landscape, our aim is to explore theories. We do not seek to replicate reality or predict any future. Rather, we use our ABM to test the implications of the theories described above in a variety of local ecologies.

ABMs allow for the direct incorporation of feedbacks which are fundamental to the dynamism of human and ecological systems. In our model, these feedbacks are of two types. One involves endogenous relations among key variables: for example, the risk of migration depends on household assets, which in turn depend on loss of household labor through prior migration as well as remittances. The other type of feedback involves interaction among agents, or, the influences that neighbors have upon each other. In our ABM, the behavior of individuals and households depends on the behavior of other individuals and households, structured through social networks. The emphasis is on process rather than individual events. With feedbacks, ABMs provide the ability to analyze the dynamics of an interconnected system over time, making it possible to find emergent and unexpected patterns in trajectories that would not be possible with a statistical regression approach. 4

The ABM capitalized on extensive data and research already conducted in Nang Rong. It built on longitudinal panel survey data that followed all individuals in 51 villages, including out-, in- and return migrants, collected complete social network data (kin and exchange) in villages of origin, georeferenced villages, households, and the locations and attributes of plots farmed by each household, and collected information about what households grew on each of their plots. It drew on a ground-truthed time series of satellite images classified for land cover/land use, a digital elevation model constructed from topographic maps, soil depth and drainage maps, and observations in the field. It was also informed by qualitative interviews with village heads, farmers, and other residents in the district conducted over multiple field visits. Spatially explicit ABMs that fully incorporate explicit geography are rare in sociological applications (Bruch and Atwell 2014; Manson et al. 2020); also rare are ABMs that use empirical measures of networks ( Smith and Burow 2018 ).

The agent-based simulation includes multiple types of agents: individuals, land parcels, and households. Households are a point of integration for the model: individuals form households, embedded in social networks and villages. Land parcels are owned, managed or used by households. Villages are composed of households, and social networks consist of direct and indirect kin ties among members within and between those households, including former members (i.e., migrants). Villages are modeled separately, one at a time. 5 Complete descriptions of the model can be found in Walsh et al. (2013) and Entwisle et al. (2016) . Appendix A presents a detailed overview of the components of the model most relevant to out- and return migration. Here, we focus on aspects key to our arguments in this paper.

Each individual agent has attributes such as age, gender, marital status, and place of residence and can experience demographic, social, and/or economic processes including giving birth (married women of reproductive age, MWRA, only), death, out-migration, marriage, and establishing a new residence locally. When not residing in the community, they can return, remit to the origin household (influencing that household’s assets), marry, or die. Couples can rent and own land and accumulate assets and can pass them on to their kin when they die or reach old age. Subfamilies are groups of individuals (such as a married daughter, her husband, and children) who live within households and can split off to form a new household. The modeling of household formation and dissolution, and the distribution of land and other assets when households split or otherwise dissolve, is based on extensive fieldwork.

Land parcels also have attributes, such as size, distance from the village, flooding potential and topographic setting, land use type, and soil suitability for various agricultural uses. The impact of floods and droughts varies within and between villages depending on these attributes. Land parcels are explicitly located on Nang Rong’s landscape as measured in the geographic and environmental survey data described above. Depending on parcel characteristics, household resources, and environmental factors such as the timing and amount of rainfall, each household makes a choice about whether to farm or not (and where, among their available options), whether to rent additional parcels, what crop to grow (rice, sugar, or cassava), and whether to use inputs such as chemical fertilizer, herbicides, and pesticides. Climate change in the form of floods and droughts affect land use decisions, i.e., what to grow on each parcel. Those parcels in turn experience levels of productivity, specific to the crop in question, which are based on household choices, parcel attributes such as soil suitability and flooding potential, and the timing and amount of rainfall. In terms of the model, the timing and amount of the monsoon each year is one of the factors affecting the yield on each plot cultivated by the household in that year. In this way, climate change affects yield, which in turn affects household assets. These components of the model, described by Walsh et al. (2013) in detail, are based on agricultural and environmental science ( Jones et al. 2003 ; Heumann, Walsh, and McDaniel 2011 ; Heumann et al., 2013 ; Malanson et al., 2014 ) and extensive fieldwork in the region in 2004 and 2010. 6

Importantly, we do not build a direct effect of climate change on out- or return migration into the agent-based model. Our interest is not in catastrophic floods and droughts, which might occasion mass migration out of an area, but in extensions of weather trends already well underway (see Figure 3 ). Accordingly, we operationalize the effects of climate change on migration through a linked set of mechanisms centering on livelihoods. This is illustrated in Figure 4 , which abstracts from the ABM shown in Appendix A to show specific feedbacks in the model.

Major Pathway in the ABM: Weather Affects Migration Through Yield, Assets, and Feedbacks

The primary pathway through which extreme climate events can influence migration patterns in this ABM is the following: the timing and amount of rainfall affects crop yields (rice, cassava, and sugar are modeled separately), crop yields affect household assets, and household assets as well as the characteristics of current and prior household members affect out- and return migration (households with more assets are better able to finance migration and to afford the loss of labor associated with it). There are feedbacks from migration to household assets (remittances increase assets), migration to crop yields (through labor availability), and household assets to crop yields (through inputs such as fertilizer, which must be purchased). Rainfall is assumed to be exogenous, a reasonable assumption given the focus on the experiences of individual villages over annual time steps. Because the plots farmed by households vary in their vulnerability to floods and droughts (elevation, distance from rivers, soil suitability), the impact of climate shocks can vary within villages (see Walsh et al. (2013) for a complete description of this part of the model).

The rules for out- and return migration are based on a probabilistic approach and were derived from a statistical analysis of Nang Rong survey data as well as relevant substantive and theoretical literature ( Entwisle et al. 2016 ). Reflecting our interest in the role of the life course in structuring migration patterns, individuals between the ages of 10 and 29 are eligible to out-migrate and return migrate. In the Nang Rong survey data, few individuals outside of these age ranges leave or return to the village. 7 The coefficients from regression models are used in the agent-based simulation to determine individual specific probabilities of out- and return migration in each simulated year. Individuals are randomly selected to migrate with the chance of doing so proportional to probabilities defined by regression-based out- and return migration equations. These regression-based out-migration and return migration equations were designed specifically to incorporate some of the key predictors of migration reflected in the theories mentioned above as well as characteristics that are important for the research questions herein, such as age, marital status, and childbearing (as life course status indicators), household assets, social networks (measured at multiple levels), and village characteristics. The equations used to determine probabilities of out- and return migration, operational definitions of the measures, and treatment of uncertainty are presented in Appendix B . Consistent with a life course focus, we calculate migration probabilities for agents who are 10–19 years old and those who are 20–29 years old in two separate equations.

Social networks are both cause and consequence of migration in the model ( Entwisle 2007 ; Klabunde and Willekens 2016 ). Potential out- and return migrants move (or not) depending on “pulls” in places of origin and destination. Social ties bear on these “pulls” as both sources of information and help and sources of obligation, as we have described earlier. Social ties are also changed as a result of migration. When a potential migrant moves out of the household and village of origin, a new tie to the destination is added, increasing the flow of information and other resources to other members of the household and village, perhaps encouraging additional migration in the future. At the same time, a tie within the origin is subtracted, which depending on where in the village-based network the prior resident and their household was situated, may affect household centrality, links to well-off households (top 10%), and overall village cohesiveness, all relevant to migration. Return migration leads to analogous changes in the patterning of social ties. Logically, social networks are endogenous, part and parcel of dynamic interrelationships involving migration. With ABMs, it is possible to fully incorporate this endogeneity.

Multiple levels of social organization are potentially relevant to the operation of social network ties in relation to migration. The ABM includes network variables at three levels: individuals, households, and villages. Because of their significance for migration, we focus on kin networks. These are measured in terms of first- to fourth-degree kin ties, where degree is operationalized as steps involving birth or marriage (both of which are also processes in our ABM). An example of a fourth-degree tie is mother’s mother’s child’s child—in other words, first cousin. At the individual level, we include the number of kin ties that potential migrants have to migrants from the village, and that potential return migrants have to village residents. For individual village residents and migrants, we also construct a dependency ratio that measures the number of ties to potentially dependent kin (younger than 10, over 60) in the village divided by the number of kin aged 10 to 60 (the working ages in this context) in the village able to assist them. At the household level, we include a measure of household centrality and a count of ties to households in the top 10% of the wealth distribution. At the village level, we include a measure of network cohesion as well as a measure of migration prevalence. See Appendix B for further details.

While weather→crop yields→household assets→migration is the primary pathway through which extreme climate events can influence migration patterns, because of endogenous feedbacks, model results may not be this simple. Factors motivating migration change each year in the model due to prior migration behavior (e.g., migrant remittances increase household assets) as well as the behavior of other agents (e.g., social ties to destination reflect the migration of others in the household and village). A decrease in household assets may motivate migration; remittances may address the shortfall. Such effects are found in empirical regressions based on snapshots of either out- or return migration in Nang Rong villages ( Garip 2014 ; Tong and Piotrowski 2010 ) and elsewhere ( Garip 2012 ), but their dynamics in a coupled system of out- and return migration, social networks, household assets, and land use have not been explored. Prior migrants from a household or village can make it easier for more prospective migrants to move to a destination and can influence what neoclassical models of migration deem “success,” that is they help retain migrants who have already moved there ( Korinek, Entwisle, and Jampaklay 2005 ; Garip 2008 ). Social ties to the village, especially to potentially needy kin, may speed return or delay migration in the first place. Interrelationships between land use, household assets, social networks, and both out- and return migration are dynamic in the ABM. Macro-micro connections between migration, land use, and social organization are complex and multidirectional ( Axinn and Ghmire 2011 ).

Climate Impacts

In the ABM, we simulate 25 years in Nang Rong villages in annual time steps, initialized with empirical social and land use data from the Nang Rong Surveys conducted in 2000. The models are realistic in that they start with “real” data measuring the characteristics of “real” villagers in “real” villages. However, it was never our purpose to reproduce the past or predict the future. Rather, our goal is to use the ABM to explore the implications of the theories embodied within it, which for the purposes of this paper, are those involving out- and return migration. Specifically, we use the ABM as a laboratory for a series of experimental studies of rainfall patterns and migration, and then the role played by dynamic social networks in producing the observed effects. In each instance, we simulate population processes over a period of time, then make one change and re-simulate, with everything else the same.

We study the impacts of extreme weather events by manipulating the climate in four scenarios, one of which serves as a benchmark. We created these scenarios based on monthly rainfall data for Nang Rong from 1900 to 2008, accessed from the University of Delaware Center for Climate and Land Surface Change ( http://climate.geog.udel.edu/∼climate/html_pages/archive.html#gcd ). Each scenario begins and ends with the same weather pattern, what we call “normal-normal” weather to reflect normal timing of the monsoon and normal amount of rainfall associated with it. Normal in this case is defined as average timing and amounts of monsoon rain based on the historical record. In all scenarios, normal-normal weather is present for simulation years 0 to 10, allowing the ABM to become fully established, and in simulation years 18 to 25, when there may be recovery. The scenarios differ in the middle years (simulation years 11 to 17). The first scenario focuses on droughts; its middle years are marked by a seven-year period of extremely dry weather. The second focuses on floods and contains a seven-year period of extremely wet weather in the middle years. In the third, we examine variability with a scenario whose middle years fluctuate between severe droughts and floods. Each is compared to a reference scenario, containing normal-normal weather during the middle years, which serves as a benchmark. Hereafter, we call these scenarios “drought,” “flood,” “variability,” and “reference.”

The drought, flood, and variability scenarios each include a single period of seven years (running from simulation years 11–17) that experiences extremely dry weather, extremely wet weather, or alternating extremes every two years. In the dry years, the monsoon is late and amounts lower; in the wet years, the monsoon is early and the amounts greater. The dry years are drier than experienced in the historical record, but there is still enough rain to provide some yield of all crops in the model. The flood years are wetter than experienced in the historical record, but do not entirely flood the area and allow for some yield of all crops in the model. By examining seven-year periods of extreme weather preceded by 10 years of normal–normal weather and followed by normal–normal weather in the remaining eight years, we can study the destabilizing effects of climate change as well as look at recovery. Further detail on creation of the scenarios is provided in Walsh et al. (2013) .

Models were run separately for 41 Nang Rong villages. We selected these villages from the 51 Nang Rong Project villages for whom social and spatial survey data were available because they had the highest quality data, including cadastral maps or high-quality field measurement of which households use which plots of land. From the standpoint of the model, each village constitutes its own set of “initial conditions.” Despite the fact that the villages are located in a relatively small area in Thailand’s Northeast, they vary in their susceptibility to weather events. A lowland village close to one of Nang Rong’s rivers will be less susceptible to drought, and more susceptible to flooding, compared to an upland village with limited water supply but good drainage. Each village varies in its location, nature and quality of soils, access to water and susceptibility to flood, as well as population size, economic characteristics, migration patterns, social networks and the like (all initialized with real data). By considering all 41 villages, we accommodate the diversity found in the district. We also ensure that model results do not depend on the particulars of any specific village, as might occur if we relied exclusively on the experience of an exemplar village. To accommodate stochastic elements in the ABM and to gauge the prospects of outcome uncertainty, each climate scenario was run 40 times for each of the 41 villages (1,640 runs for each climate scenario). This aspect of potential variability in results yielded little additional information, so median results are reported for each village.

We show results in a series of graphs, one for each climate scenario. The x-axis shows model years, from 1 to 25. Years 1–10 of the simulation of all scenarios are characterized by normal–normal climate, during years 11–17 we simulate the weather shocks, and in years 18–25 we simulate normal– normal climate in all scenarios. The y-axis portrays out- or return migration as a proportion of those eligible, measured as deviations from the baseline rate for each of the villages, with the baseline defined according to the reference scenario of no extreme weather. In this case, d e v i a t i o n y , s = f o c a l = 100 * ( m i g r a t i o n y , s = f o c a l m i g r a t i o n y , s = r e f e r e n c e ) − 100 , where y indexes the year, and s indexes whether the scenario is a focal scenario (drought, flood, or variability) or the reference scenario. This presentation allows us to counterfactually assess migration dynamics in the village under extreme weather events consistent with climate change (drought, flood, or variability scenarios) compared to what would be observed in the absence of such events (reference scenario). As such, there are minimal differences in village means prior to the onset of the climate event (year 11), differences consistent with slight random variation in the stochastic processes we study. Our interest is the direction of the migration response, either positive or negative, during and after the climate event occurs, where each village serves as its own control. That is, we are comparing how the village looks in each extreme weather scenario to how it would look under the reference scenario. We are also interested in the durability of extreme weather impacts, the persistence of migration responses after the extreme weather subsides and the focal scenario resumes the weather pattern of the reference scenario (years 18 to 25). Further, our interest is in response across all of the villages, not trajectories of experience for individual villages. For each model year, the graphs show a box plot summarizing the distribution of results across the 41 villages for which the simulations were run.

Figure 5 shows the impacts of flood, drought, and variability on out-migration rates, defined as the proportion leaving of those eligible to leave in each year of the model run compared to the reference scenario. The graph contains results from all 41 villages, plotted as a distribution of village-specific deviations averaged over the 40 runs of the model. Of interest is the mean and spread of experience each year and especially the discontinuities in years 11 and 18, the beginning and end of the simulated climate event. Since we focus on within-village deviations between the extreme weather scenarios and the reference scenario, positive numbers indicate a village’s migration is greater in the focal scenario than it is in the reference scenario, negative numbers indicate the opposite, and zeros mean that migration does not differ between the extreme weather scenarios and the reference scenario. Because we obtain the results for each village by experimentally manipulating the climate scenarios in the ABM and explicitly benchmark them against what would occur in that village under more normal climatic conditions, it is helpful to think of these results as changes in out-migration rates that are induced by the different climatic scenarios.

Deviations in Proportion of Eligible Village Residents who Out-Migrate in Each Climate Shock Scenario Relative to Reference Climate Scenario

Results are highly clustered until year 13. Even though extreme weather events begin in year 11, it takes a few years to provoke a response in the model, which suggests adjustment in the face of short- term challenges. After that, the distribution of response expands substantially. In some villages, out- migration increases, but in others, it decreases. The impact depends on which village we select. Averages are little affected, however. Median response across villages hovers around zero. For every village showing climate shocks to have a positive effect on out-migration, there is another showing a reverse effect. There are few differences by scenario: the pattern is the same for droughts, floods, and variability between them. 8 The model results suggest no systematic effect of climate shocks on out-migration in this setting.

An immediate question is whether the weak effect on out-migration is due to relatively weak effects of rainfall patterns on yields and assets. As noted earlier, the effect of climate on migration is indirect in our ABM: the timing and amount of the monsoon affects yield, which affects assets, which affects migration. If there is only a weak effect on yields, or on assets, we would not be surprised to see a weak effect on out-migration. To demonstrate that the lack of an out-migration response is not due to some failure along this pathway, Figure 6 shows the impact on total assets measured at the village level for each of the climate scenarios. The negative impact of floods and droughts is clearly visible in this figure: assets respond immediately to the beginning of the climate shock. In fact, although the situation improves after the end of the shock, except in a few rare instances, total village assets do not rebound to pre-shock levels by the end of the simulation.

Deviations in Proportion of Eligible Migrants who Return to Village in Each Climate Shock Scenario Relative to Reference Climate Scenario

What about return migration? Figure 7 shows the impacts of flood, drought, and variability on return migration, defined as the proportion returning among those eligible to return in each model year, in comparison with the reference scenario. As before, results in all 41 villages are shown. Again, results are tightly clustered around baseline until year 13 (i.e., slightly after the climate shock begins), but in the case of return migration, rates clearly fall after that, even as the distribution of results expands. Indeed, there are only one or two instances of villages where return migration is higher than baseline, and these are not sustained, appearing for no more than a year or two. Whether the climate shock is flood, drought, or variability, the effect is to retard and diminish return migration. This is a persistent effect, remaining beyond the end of the climate shock, throughout the simulated years for most villages.

Deviations in Total Village Assets in Each Climate Shock Scenario Relative to Baseline Climate Scenario

What can we say about the magnitude of the effects shown in Figures 5 and ​ and7? 7 ? Are they reasonable? There are several sources reporting broadly convergent results, although we note that interpretations are tricky because of differences in the time-scale of migration, cohort versus period estimates, and eligibility criteria used. Across all years and climate scenarios, the ABM results for out- migration rates vary across villages from 12.7% to 51.5% (mean 33.1%). In prior work, Garip’s (2008) study of predicted probabilities of migrating among Nang Rong residents shows a similar range to what is observed at the village level in the ABM data. In another of her papers ( DiMaggio and Garip 2011 ), village-level percentages of migrants range from about 20% to about 50% (median approximately 33%). Curran and colleagues find migration prevalence rates in Nang Rong villages with a similar range to that reported here (Curran et al. 2005). Similarly, for return migration, the variation of ABM results across all years and climate scenarios (10.5% to 56.0%, with a mean of 28.4%) corresponds with figures reported in other studies (e.g., 26% in Tong and Piotrowski (2010) ; 27.8% for males and 30.0% for females in Korinek et al. (2006)). The ABM broadly reproduces migration levels found in other studies, even though such reproduction was not an explicit aim given our theory-testing orientation, lending credibility to our efforts.

Now, to the more interesting question: Why does the model show impacts on return migration but not on out-migration? The two findings are likely connected. Nang Rong is a setting where repeat migration is common. In such a marginal economic setting, where out-migration is already well established, there is not much “room” for additional impact. This would be all the more true if the stock of potential migrants were decreasing due to reduced return migration, a feedback in the system. One variable in the migration model is the kin dependency ratio, which reflects the number of older and younger kin residing in the village relative to kin in the productive age ranges. Out-migration increases the ratios, return migration decreases them. If migrants do not return, then the stock of potential future out-migrants is reduced relative to competing needs in the village. Kin dependency ratios act as an equilibrating factor in the model.

How do these results reflect the role of endogenous social networks? As explained earlier, we measure kin and other social ties to both origin and destination in our model, with estimated effects based on regressions of out- and return migration using the longitudinal survey data. (Also see Appendix B .) The relative number of migrants from the village already in the destination facilitates out-migration, as do specific ties to family members who are there. Ties to non-migrants are also relevant, especially ties to potentially dependent family members relative to the number of working age adults (a person-specific kin dependency ratio). These ties encourage out-migration at the youngest adult ages but decrease in size and eventually reverse in effect as potential migrants grow older. They consistently encourage return migration. Note that all of these ties are updated in the model annually, in response to births, deaths, and migration in the previous year.

We can address the question of networks systematically. An advantage of the ABM approach is the ability to conduct experiments to test the importance and impact of specific mechanisms in the model ( Bruch and Atwell 2015 ). We do this by “breaking” the links between different conceptual processes and migration by altering different sets of coefficients in the migration probability model. To “break” links involving social networks, we set all network coefficients to zero and re-run the ABM.

Figure 8 shows the consequences of “breaking” these links for modeling out-migration. No matter what scenario we choose, without the facilitating effects of social networks, climate stress decreases out- migration. How can we explain this? In the absence of social ties, migration becomes a profoundly individual decision, a simple cost-benefit decision. It is “neoclassical.” In our scenarios, the benefits associated with migration are essentially held constant, but costs can vary. There are two sources of impact relevant to costs. One is the ability to afford the costs of a move. Floods and droughts reduce the assets available for this purpose. The other are the costs themselves. The relative numbers of villagers currently living in urban destinations as well as specific ties to former household members living there bear directly on these costs. The more contacts prospective migrants have, the more help they can expect in finding a job and a place to live. Without these contacts, costs of migration are high. In essence, in the absence of social ties, each migrant is a “pioneer.” Breaking the links involving social networks thus increases the costs of migration as the resources needed to finance the trip are decreased as a result of climate events. The upshot is reduced out-migration. 9

Deviations in Proportion of Eligible Village Residents who Out-Migrate in Each Climate Shock Scenario Relative to Reference Climate Scenario when Social Network Effects are Removed

The implications for return migration may also appear surprising. As shown in Figure 9 , without social networks, return migration is not responsive to flood, drought, or variability in our model. The likelihood of return continues along as before. How do we explain this? A likely explanation is that without social ties linking migrants to their home villages, little news about hardships associated with floods and droughts at home reaches the migrants; hence, decisions to return or not are unaffected by these patterns. In addition, there would be no conduits through which home households could leverage social ties in places of destination to exert their influence. It is possible that return migration may still respond to household assets. However, there are two countervailing effects. Assets are greater if migrants do not return, and instead remit. At the same time, greater assets hasten return. Qualitative work from Northeast Thailand recognizes these countervailing pathways. Potential returners note: “There are more expenses if the children stay home. If we go away to work, there are less people home, and it less expensive to feed the family” (quoted in Garip 2014 :677). Others who have returned note that: “It might have been better for me to stay in the village because we had land. When I migrated for work, no one took care of the land, so we had to rent it out” (quoted in Garip 2014 :677). Such response patterns differ by household wealth ( Garip 2014 ), but both sentiments are predicated on feelings of obligation to kin and decision-making that extends beyond the household. In the absence of influences from social ties, on average, the two cancel each other out. When social ties are included in the model, climate events clearly depress return migration relative to baseline.

Deviations in Proportion of Eligible Migrants who Return to Village in Each Climate Shock Scenario Relative to Reference Climate Scenario when Social Network Effects are Removed

Discussion and Conclusion

Climate change can occur slowly when measured in the scale of human lives. In areas such as Nang Rong, Thailand, where weather patterns are already quite variable, it can be difficult to detect a turning point in the extent or frequency of floods and droughts. In areas such as this, households likely have already adapted to periodic floods and droughts, even extreme ones. Transformations in the livelihoods once supported in a given locale take time, raising the possibility of adjustment in the interim. A migration response need not be immediate, unidirectional, or uniform, and indeed, may not occur at all.

Migration processes evolve over the course of lives, embedded in social networks, structured within migration systems. These insights are not always reflected in the climate change and migration literature. Accordingly, our agent-based model acknowledges the role of migration in the life course, especially the transition to adulthood, its role in the household economy, and its role in responding to as well as re-forming social networks in origin as well as destination. The system is dynamic, and climate change has the potential to disrupt any or all parts of it.

In contrast to a literature preoccupied with out-migration, we also studied return migration. From the standpoint of the life course, both out- and return migration are central to the transition to adulthood, with young people leaving and returning to their village of origin as they assemble the resources they need to meet obligations to family that are also are redefined at the same time. From the standpoint of the household economy, household members who migrate to urban areas for work often return after they achieve earnings targets. From both of these perspectives, out- and return migration are connected. Climate shocks may provoke out-migration, but equally, they may discourage return migration. Either can lead to depopulation over time.

Our empirical focus was climate shocks related to the timing and amount of rainfall. Recent literature on climate change effects has considered temperature as well as precipitation (e.g., Bohra- Mishra, Oppenheimer, and Hsiang 2014 ; Bohra-Mishra et al. 2017 ; Mueller, Gray, and Kosec 2014 ). 10 There is evidence to suggest warming in Thailand ( Limsakul et al. 2011 ), with potential consequences for cropping patterns. However, in interviews with Nang Rong farmers undertaken over multiple field visits, temperature was never mentioned as an problem. We selected rainfall as a focus because of its centrality to rice cultivation ( Naylor et al. 2007 ) and the livelihoods of Nang Rong households. 11

We examined three extreme weather scenarios, which are beyond recent experience, but do not negate the possibility of adaptation and adjustment. Whether the climate shock involves flood, drought, or both, the result is the same in our models. This may at first seem puzzling, given the scenarios move water conditions in opposing directions. However, it is important to emphasize that they are all extreme shocks, beyond the case where a little more water would improve yields in a too-dry village or a little less would improve them in a too-wet village. Further, the effects all operate through livelihoods, specifically the productivity of plots. Other possible consequences of extreme weather scenarios are outside our model.

Further, considering the set of villages as a whole, there is little effect on out-migration on the median of all villages. However, there is a pronounced reduction in the likelihood of return migration compared to baseline for each of the villages. Of course, the experience of specific villages may vary, depending on village characteristics such as wealth, the strength of social ties, migration history, the agricultural potential of the land farmed by village households, and vulnerability to floods and droughts. An exploration of village heterogeneity is beyond the scope of this paper but would certainly be a worthwhile undertaking in the future.

To better understand these results, and to gain a better understanding of the ABM in general, we “broke” the links involving social networks. In the absence of social networks, climate shocks reduced out-migration relative to baseline, but with little consequence for patterns of return migration. Clearly, social networks are important to the maintenance of out- and return migration flows in this system. As clearly, patterns of out- and return migration are linked in important ways; after all, those who might return are those who left, and many of those who might leave are those who have previously returned.

Of course, our results and the inferences drawn from them are conditional on a particular agent- based model of household dynamics and land use change for the specific case of Nang Rong, Thailand. It drew on extensive social and spatial data collected over three decades in this setting, analyses of those data, and observations based on numerous field visits. It capitalized particularly on the prospective, longitudinal panel, the follow-up of out- and return migrants, complete social network data, and links between households and georeferenced field plots, the last of which were characterized with extensive geospatial data on elevation, soil, and land use. Compared to some ABMs, ours involves considerable detail. The data used to initialize the model were also quite detailed, drawn from empirical social and spatial survey data, thus permitting us to accommodate a range of initial conditions over the 41 villages simulated. However, from the start, it emphatically was not our interest to make specific predictions, but rather to explore the implications of the theories embedded in the ABM.

What light does our model shed on economic theories of migration? Whether a neoclassical migrant or a temporary migrant moving as part of a household strategy, the likelihood of a move depends on costs and the ability to afford these costs. Ties to the destination embody resources such as help finding a job or place to live that diminish costs of migration. The importance of these ties was dramatically illustrated when out-migration plummeted as social network links in the model were “broken.” With no ties to destinations, potential out-migrants become “pioneers,” completely reliant on their own resources. This result points to the importance of social context even for the individuals who are the focus of neoclassical theory.

In NELM, households are the focus. Here, our concern is with household assets in relation to the behavior of individual members. Household assets are diminished by out-migration, at least in the short run, because of the loss of labor power as well as the resources needed to support migration. In the absence of remittances, the loss of labor power continues to depress household assets over time, other things equal. Remittances can restore assets and potentially even provide a return on the investment, but risks remain in this risk-minimization strategy. The longer household members reside in the destination, the more likely they are to acquire new obligations, e.g., through marriage ( Jampaklay 2006 ). By depressing the likelihood of return, climate shocks begin to undermine a central tenet of spatial diversification as a household strategy.

Economic theories of labor migration are central to the study of international as well as internal migration, which raises a question as to whether our results might shed any light on movements across as well as within national borders. Indeed, Donato and Massey (2016) argue that international migration is shifting from “moving to opportunity” during the 20 th century to “evading threat” in the 21 st , with one of the threats being climate change. As is true of the literature in general, studies of climate change and international migration focus on out-migration. What might we expect for return migration? Given the higher costs and greater difficulty of return for international than internal migration, we might expect an even greater impact of climate change on return migration when international borders are involved. It will depend on the porosity of the borders, of course. An example of porous borders are those between Burkina Faso and Cote d’Ivoire, adjacent countries in West Africa with a long tradition of movement between them ( OECD 2017 ). An example of a hardened border would be that between the US and Mexico. In the latter case, return migration is already so substantially depressed as a result of US policy ( Massey, Durand, and Pren 2016 ) that there might not be much potential for further impact.

Social forces are also relevant. An innovative feature of our model is that it incorporated a life course approach. The literature on residential mobility, and to some extent internal migration in developed countries, is increasingly informed by life course concepts and approaches ( Bernard, Bell, and Charles-Edwards 2014b ; Coulter, van Hamm, and Findlay 2016 ; Warner and Sharp 2016 ). This is not yet true of the literature on rural-urban migration within developing countries or migration across international boundaries. The life course approach offers particular insight on out- and return migration through the transition to adulthood. These processes are linked to other status transitions occurring during the early adult years, including entry into the labor market, marriage, and childbearing. Young people who move to Bangkok may do so as part of a household strategy, but in the course of their stay may meet a future spouse from somewhere else. What started as a temporary move may evolve into a more permanent stay as a consequence of life course transitions. Decisions to leave and return in our model illustrate the life course principles of agency and timing as well as the contingent nature of the transition to adulthood. Our attention to the potential importance of “initial conditions” (simulating 41 separate villages with different migration histories, population compositions, social networks, and landscape features), and indeed, to the particulars of Nang Rong as a setting, captures the principle of historical time and geographic place. The principle of “linked lives” is reflected in network dynamics. As demonstrated in our simulations, out- and return migration depend on the needs (as reflected in the number of dependents in the village) and prior migration behavior of kin.

Social networks are fundamental to the dynamics of our ABM. We focused on kin networks, which change as a result of migration but also other demographic processes such as fertility and marriage. Kin ties are conduits for information flow, within and between origin and destination. They are potential sources of help and assistance, for recent migrants particularly, as well as sources of obligation, especially in the home village, along with social pressure to meet these obligations. Networks are central to understanding the risks and costs of migration. They are also central to understanding migration in the context of “linked lives.” Of course, our focus on kin networks is a simplification. Friends, schoolmates, neighbors, employers, and other acquaintances may also have a role to play, but these alternate ties are beyond the scope of our model.

A major strength of the ABM approach is the ability to run experiments (Bruch and Atwell 2014). We conducted two types of experiments. In the first, we isolated and manipulated weather patterns, simulating the effects of flood, drought, and variability between the two. Everything other than weather patterns was held constant in these simulations, which is reasonable given that villages are small relative to national and international markets for agricultural products in Thailand. The second type of experiment manipulated key linkages in the model, specifically those involving social networks, shedding light on the dynamics of the ABM and the complex pathways producing the effects observed in the simulations. In these experiments, as in all true experiments, “all else” is equal. In the real world, of course, it is not.

While it is not possible to generalize beyond the model and the case, the results do suggest a new avenue for research on the impacts of climate change: return migration. To be sure, there is interest in recovery migration among the people who leave an area because of a catastrophic flood such as caused by Hurricane Katrina or the Indonesian Tsunami, and important research is based on these populations as well as new migrants to affected areas (e.g., Fussell, Curtis, and deWaard 2014 ). The impact highlighted in our study is quite different, however. In ordinary times, many young people in the Nang Rong villages leave for Bangkok and other urban areas; roughly half of them return. Climate shocks have the potential to affect either direction of movement, and it is important to look at both. Definitional issues come into play here, revolving around migrant intent and length of absence. We believe that it is important to look at return separately rather than model the round-trip as the outcome (Call et al. 2018). Whether migration is temporary or permanent is only known after the fact. In our model, climate shocks reduced the likelihood of return relative to baseline although it did not alter patterns of out-migration in any systematic way.

To be sure, it is more difficult to observe migrants who do not return than prospective migrants who leave. The New York Times will be challenged to find iconic photographs and related stories of this potential impact. Yet, in terms of local populations, if young people leave and stop returning, the consequences for social and economic life are no less dramatic.

Acknowledgments.

We would like to thank Brian Thiede, Robert McLeman, the anonymous reviewers, and participants in the Stanford Earth System Science Seminar, Harvard Social Demography seminar, and Brown Population Studies and Training Center Colloquium for their helpful comments, Denise Ammons for expert assistance with graphics, and the Carolina Population Center for general support (P2C HD050924). This paper builds on more than two decades of collaborative research with colleagues at Mahidol University (Salaya, Thailand) and the University of North Carolina at Chapel Hill. Over those years, Nang Rong projects have received support from many sources, including the National Institute of Child Health and Human Development, the National Science Foundation, NASA, the Mellon Foundation, and the MacArthur Foundation. We gratefully acknowledge this support. We owe a particular debt to our colleagues Pramote Prasartkul, Aphichat Chamratrithirong, Aree Jampaklay, and Yothin Sawangdee at Mahidol, and Ron Rindfuss, Steve Walsh, and Peter Mucha at UNC. Most of all, we wish to thank the people of Nang Rong, who made this research possible. Please direct correspondence to Barbara Entwisle, ude.cnu@elsiwtne .

Figure A.1 provides a graphic representation of our agent-based model. The paragraphs below provide additional description for each of the model’s components. Appendix B provides further detail on how migration is modeled. As noted in the text, the ABM is run separately for each of 41 villages.

Initialization:

Initial values are set for households, household members, and parcels in the selected village. For households: a household identifier (ID), a roster (identifying current and prior household members), a list of owned parcels, a list of managed parcels, assets and remittances as well as the identification of subfamilies and a land split trigger. For individuals, these include a Household ID, a person identifier, age, sex, marital status, status as resident or migrant, years gone (if migrant), and spouse, mother, and father identifiers (to identify kin ties). For parcels, these include a parcel ID, size, a list of LULC cells included in the parcel (raster), land use type, duration in that land use, suitability for rice, cassava, sugar cultivation, distance to road, distance to river, and household IDs (multiple in the case of shared management). For villages, a village ID and list of households, and aggregate or constructed measures based on individuals, households, and parcels. Initial values come from the 2000 household survey (and derivatives from prior surveys) and the associated identification and characterization of parcels (Rindfuss et al. 2004). All of these items are drawn from the empirical Nang Rong social and spatial data collected in the 2000 wave.

The model runs household by household within each time step, with values updated at the end of each time step. What follows is a description of procedures for households during a single time step.

Select first household:

Proceed to parcel module:.

Within household, determine whether the household is engaged in agriculture.

If not, owned land is rented and owning household receives a third of the product and renting household receives remainder (based on field interviews).

If the household is engaged in agriculture, for each plot, based on land characteristics, choose crop; based on assets, choose inputs (or not); calculate the yield based on crop, inputs, and rainfall pattern.

Repeat these steps for all parcels.

The yield is valued based on market prices prevailing at the time. The total is added to household assets. See Walsh et al. (2013) for further details on this component of the model.

Proceed to Household Member Module:

Within household, select member:.

Use survival schedule to calculate the likelihood of a death. If a death, adjust roster, subfamily tracker, and sociomatrices.

If female and in fertile ages, use fertility schedule to calculate the likelihood of a birth. If a birth, update the roster, the household subfamily tracker, and the sociomatrices.

If a current resident:

Migrate? See Appendix B for details.

If a prior resident (i.e., migrant):

Return? See Appendix B for details.

For all current residents who are single (including those who have just returned):

calculate the risk of marriage. If there is a marriage, determine postnuptial residence (randomly: 40% chance of moving to and settling in an urban destination; 30% chance of moving to another village; 30% chance of staying in origin village, with equal chances of living with parents or establishing an independent household). If a marriage and postnuptial residence is with parents in origin village, update the subfamily tracker.

For all prior residents who are single (including those who have just migrated):

Calculate the risk of marriage. Update the rosters.

For all prior residents:

Calculate the risk of remittance based on gender, marital status.

Repeat these steps for all members of the household.

Consider the household as a unit. Did the last parent die? If so, then allocate assets according to a set of rules based on fieldwork. If no, then are there one or more subfamilies? If so, is there a split? (Based on the number of subfamilies, presence of children, and length of coresidence.) If so, create a new household and allocate the land.

Update wealth based on productivity of parcels as determined in the Parcel Module and remittances as determined in the Household Member Module. Deductions for use of inputs, loss of labor due to migration, consumption of those remaining in the household, and land conversion (rice to upland crops, upland crops to rice).

Proceed to the next household.

After all households are complete, update the kinship network and other information for the next t.

Graphic Representation of Nang Rong ABM

Agent-Based Model Components Relevant to Social Networks and Migration

The social network and migration components of the agent-based model draw on an analysis of out- and return migration between 1994 and 2000 based on data from the 1984, 1994 and 2000 Nang Rong surveys. The 1994 household survey collected data on complete networks in 51 villages of origin.

Because of their significance for migration, we focus on kin networks.

A complete enumeration of current and previous (1984) household members was undertaken in the 1994 survey. Links to mother, father, and spouse were obtained for each person listed on the household roster. In addition, links to siblings were obtained for each person aged 18–35. We used these data to construct matrices of first-degree kin ties between current and former members of each village (i.e., everyone enumerated in 1984). We derived second-, third-, and fourth-degree kin ties through matrix multiplication. From the individual ties, we constructed measures of the extended kinship system for each village (see Verdery et al. 2012 ).

The model includes network variables at three levels: individuals, households, and villages.

At the individual level, we measure the number of kin ties (first to fourth degree) that potential migrants (village residents) have to migrants, and the number of kin ties that potential return migrants have to village residents. We also measure ties to dependent kin, both young and old. Dependency is typically measured at the household level. However, kin obligations in places of origin may extend beyond origin households to include other relatives living nearby. Further, demands may be felt differently depending on the availability of other kin to help, and these other kin may not be located in a single household. We therefore used the kin network data to create person-specific dependency ratios within three or fewer kinship degrees. Orthodox dependency ratios use ages 15 and 65 as thresholds; we adjusted these to reflect the agricultural nature of the setting, from 15 to 10 at the lower end and 65 to 60 at the upper end.

At the household level, we measure degree centrality within village-based household networks. We also measure whether or not there are kin ties to a wealthy household, defined as those in the top 10% of the village wealth distribution.

At the village level, we measured connectivity as the mean for the sum of households reachable in kin paths of three or fewer degrees, divided by the total number of households in the village, multiplied by 100. We measure migration prevalence as the proportion of 1984 village residents no longer living in the village in 1994.

We estimated logistic regressions of whether village residents in 1994 were no longer resident in 2000 (out-migration) and whether migrants in 1994 (those who were surveyed in the villages in 1984 but were reported as migrants in the 1994 surveys) had returned by 2000. We did this separately for two cohorts, persons 10–19 in 1994 and persons 20–29 in that year. In addition to assets and measures of kin ties, we included age (and age squared), sex, and marital status of individuals; assets and dependence on farming as a livelihood at the household level; and population size, the prevalence of cassava cultivation, and the percent of households owning a pump, a television, and a vehicle. Results shown in Table A1 below confirm that kin ties measured at multiple levels and in multiple ways affect out- and return migration.

The coefficient estimates in Table A1 provided the starting point for modeling the effects of social network ties in the agent-based model. The next step involved converting the predicted probability of moving over the six-year period to a predicted annual probability. The six-year probabilities were rescaled assuming constant annual probabilities but allowing for the repeated nature of migration as an event. 12

The equations predicting annual probabilities of out- and return migration were based on an analysis of all 51 villages in the data set. The agent-based models were run separately by village, initialized with actual data from each of 41 villages with the best spatial data in 2000, and subsequently updated based on predicted values at the end of that year. Values for all of the social network variables were recalculated for each village, household, and household member at each iteration of the model depending on whether that household member moved, whether other household members moved, whether other village residents moved, and whether migrants returned. The calculations also adjusted for births, marriages, and deaths during the year. Kin networks are thus fully endogenous in the agent-based model, both cause and consequence of out- and return migration and other demographic events.

The ABM also includes a stochastic element for each prediction. This is both an acknowledgment of the incompleteness of the equations (i.e., not all relevant determinants of migration behavior are included, and indeed, it is not clear that this would even be possible) as well as potential imperfections in the coefficient estimates reported in Table A1 . In recognition of this stochasticity, each model was run 40 times to obtain a distribution of results for each village; we focus on village averages, which do not alter the story. In addition, in an early stage of model development, the sensitivity of the ABM results to the particular regression coefficients was tested by using the estimated standard errors to develop “strong” and “weak” versions of the expected network effects. Although the ABM results were not identically the same (and were not expected to be), trends were similar.

Logistic Regression of Out- or Return Migration 1994–2000 on Village, Household, and Individual Characteristics: Nang Rong, Thailand

1 https://www.nytimes.com/slideshow/2016/12/28/blogs/photographing-climate-change-refugees-by-drone-and-on-foot/s/28-lens-haner-slide-RBR5.html ; http://www.economist.com/node/7833886 ; https://www.nytimes.com/2017/11/17/us/puerto-ricans-orlando.html .

2 Data from the projects are available from the University of Michigan’s Data Sharing for Demographic Research website: http://www.icpsr.umich.edu/icpsrweb/DSDR/studies/4402 .

3 http://www.cpc.unc.edu/research/tools/abm/nangrong

4 Simulating the effect on out-migration of resource declines due to climate shocks based on a snapshot, as done in a standard regression analysis, presumes that the population at risk is continually replenished, which is not necessarily the case.

5 Distance is not a consideration in the agent-based model. Because households are clustered within villages, they do not differ in distance to Bangkok and other urban destinations. There are differences between villages, but they are quite small, especially in relation to distance to Bangkok.

6 In the ABM’s Landscape Module (described in Walsh et al. 2013 ), land use decisions and outcomes are informed by household characteristics, land suitability (calculated within a maximum entropy model), inputs (three fertilizer amounts), anticipated yield (based on the Decision-Support System for Agro-tecnology Transfer cropping model), and a prespecified climate scenario.

7 Among migrants in the 1994 Nang Rong Household Survey, only 11% initiated their most recent migration when they were outside of these age ranges. Among those in the 2000, the corresponding percentage was 15%.

8 Floods and droughts do affect land use decisions and productivity differently. For example, total sugar yields respond more dramatically to droughts than floods ( Entwisle et al. 2016 ).

9 There are few studies of climate change and migration that incorporate social networks, and therefore few studies with which to compare this result and its interpretation, but there is one. Nawrotski and colleagues (2015 , 2016) report that social networks suppressed climate-induced international migration based on an analysis of data from the Mexican Migration Project. This is opposite of what we found. However, it is difficult to interpret this disparity given so many differences between the two studies in basic design (prospective vs. retrospective), the definition of migration (individual vs. household; internal vs. international), the measurement of social networks (direct vs. indirect), the measurement of remittances (explicit vs. presumed), and the measurement of climate variables (local vs. municipality level).

10 In their investigation of the effects of annual variability in precipitation, temperature, and typhoons on census- based measures of interprovincial outmigration in the Philippines, Bohra-Misha et al. (2017) find temperature effects but not precipitation effects (also see Bohra-Misha, Oppenheimer, and Hsiang 2014). That study is focused at the provincial level and examines migration over five-year intervals. Agricultural dependence also differs between the Philippines and our setting. We are looking at much more fine-grained processes.

11 Further, within the realm of potential warming scenarios, increased temperatures are unlikely to impair the productivity of the crops we model (rice, cassava, and sugar); such crops are more susceptible to low than high temperatures ( Matthews and Hunt 1994 ; Timsina and Humphreys 2006 ).

12 We would like to thank Peter Mucha for contributing this method.

Contributor Information

Barbara Entwisle, University of North Carolina at Chapel Hill.

Ashton Verdery, Pennsylvania State University.

Nathalie Williams, University of Washington.

• Axinn William G. and Ghimire Dirgha J.. 2011. “ Social Organization, Population, and Land Use .” American Journal of Sociology 117 : 209–58. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Bardsley Douglas K. and Hugo Graeme J.. 2010. “ Migration and Climate Change: Examining Thresholds of Change to Guide Effective Adaptation Decision-making .” Population and Environment 32 : 238–62. [ Google Scholar ]
• Bell Andrew Reid. 2017. “ Informing Decisions in Agent-based Models—A Mobile Update .” Environmental Modelling and Software 93 : 310–21. [ Google Scholar ]
• Bell Martin, and Elin Charles-Edwards. 2013. “ Cross-national Comparisons of Internal Migration: An Update of Global Patterns and Trends .” United Nations Population Division Technical Paper Number 2013/1. New York, United Nations . [ Google Scholar ]
• Benko Jessica. 2017. The New York Times . https://www.nytimes.com/2017/04/19/magazine/how-a- warming-planet-drives-human-migration.html [accessed 10/20/2017].
• Bernard Aude, Bell Martin, and Elin Charles-Edwards. 2014a. “ Improved Measures for the Cross- national Comparison of Age Profiles of Internal Migration .” Population Studies 68 ( 2 ): 179–95. [ PubMed ] [ Google Scholar ]
• Bernard Aude, Bell Martin, and Elin Charles-Edwards. 2014b. “ Life-course Transitions and the Age Profile of Internal Migration .” Population and Development Review 40 :213–39. [ Google Scholar ]
• Black Richard, Stephen R. G Bennett, Thomas Sandy M., and Bennington John R.. 2011. “ Migration as Adaptation .” Nature 478 : 447–9. [ PubMed ] [ Google Scholar ]
• Bohra-Misha Pratishya, Oppenheimer Michael, Cai Ruohong, Feng Shuaizhang, and Licker Rachel. 2017. “ Climate Variability and Migration in the Philippines .” Population and Environment 38 : 286–308. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Boha-Mishra P, Oppenheimer M, and Hsiang SM. 2014. “ Nonlinear Permanent Migration Response to Climatic Variations but Minimal Response to Disasters .” Proceedings of the National Academy of Sciences 111 : 9780–5. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Bruch Elizabeth and Atwell Jon. 2015. “ Agent-based Models in Empirical Social Research .” Sociological Methods and Research 44 : 186–221. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Call Maia A., Gray Clark, Yunus Mohammad, and Emch Michael E.. 2017. “ Disruption, not Displacement: Environmental Variability and Temporary Migration in Bangladesh .” Global Environmental Change 46 : 157–65. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Carling Jorgen. 2008. “ The Determinants of Migrant Remittances .” Oxford Review of Economic Policy 24 : 581–98. [ Google Scholar ]
• Cassarino Jean-Pierre. 2004. “ Theorizing Return Migration: The Conceptual Approach to Return Migrants Revisited .” International Journal on Multicultural Societies 6 : 253–79. [ Google Scholar ]
• Centola Damon, and Macy Michael. 2007. “ Complex Contagions and the Weakness of Long Ties .” American Journal of Sociology 113 : 702–34. [ Google Scholar ]
• Centola Damon, Willer Robb, and Macy Michael. 2005. “ The Emperor’s Dilemma: A Computational Model of Self-Enforcing Norms .” American Journal of Sociology 110 : 1009–40. [ Google Scholar ]
• Clark William A. V., and Suzanne Davies Withers. 2008. “ Family Migration and Mobility Sequences in the United States: Spatial Mobility in the Context of the Life Course .” Demographic Research 17 :591–622. [ Google Scholar ]
• Coulter Rory, Maarten van Ham, and Findlay Allan M.. 2016. “ Re-thinking Residential Mobility: Linking Lives through Time and Space .” Progress in Human Geography 40 : 352–74. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Coulter Rory, Maarten Van Ham, and Peteke Feijten. 2011. “ A Longitudinal Analysis of Moving Desires, Expectations and Actual Moving Behaviour .” Environment and Planning A 43 : 2742–60. [ Google Scholar ]
• Jong De, Gordon F, Root Brenda D., Gardner Robert W., Fawcett James T., and Abad Ricardo G.. 1985. “ Migration Intentions and Behavior: Decision Making in a Rural Philippine Province .” Population and Environment 8 :41–62. [ Google Scholar ]
• Jong De, Gordon F. 2000. “ Expectations, Gender, and Norms in Migration Decision-making .” Population Studies 54 : 307–19. [ PubMed ] [ Google Scholar ]
• Dillon Andrew, Mueller Valerie, and Salau Sheu. 2011. “ Migratory Responses to Agricultural Risk in Northern Nigeria .” American Journal of Agricultural Economics 93 : 1048–61. [ Google Scholar ]
• DiMaggio Paul, and Garip Filiz. 2011. “ How Network Externalities Can Exacerbate Intergroup Inequality .” American Journal of Sociology 116 : 1887–933. [ Google Scholar ]
• Donato Katherine M. and Massey Douglas S.. 2016. “ Twenty-first-century Globalization and Illegal Migration .” The Annals of the American Academy of Political and Social Science 666 : 7–26. [ Google Scholar ]
• Entwisle Barbara. 2007. “ Putting People into Place .” Demography 44 : 687–703. [ PubMed ] [ Google Scholar ]
• Entwisle Barbara, Rindfuss Ronald R., Faust Katherine, and Kaneda Toshiko. 2007. “ Networks and Contexts: Variation in the Structure of Social Ties .” American Journal of Sociology 112 : 1495–1533. [ Google Scholar ]
• Entwisle Barbara, Edmeades Jeffrey, Malanson George, Podhisita Chai, Prasartkul Pramote, Rindfuss Ronald R., and Walsh Stephen J.. 2008. “ Village Settlement, Deforestation, and the Expansion of Agriculture in a Frontier Region: Nang Rong, Thailand .” Pp. 165–79 in Millington Andrew and Jepson Wendy (eds.), Land-Change Science in the Tropics: Changing Agricultural Landscapes . New York, NY: Springer. [ Google Scholar ]
• Entwisle Barbara, Williams Nathalie E., Verdery Ashton M., Rindfuss Ronald R., Walsh Stephen J., Malanson George P., Mucha Peter J., Frizzelle Brian G., McDaniel Philip M., Yao Xiaozheng, Heumann Benjamin W., Prasartkul Pramote, Sawangdee Yothin, Jampaklay Aree. 2016. “ Climate Shocks and Migration: An Agent Based Modeling Approach .” Population and Environment 38 : 46–71. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Falkingham Jane, Sage Jo, Stone Juliet, and Vlachantoni Athina. 2016. “ Residential Mobility across the Life Course: Continuity and Change across Three Cohorts in Britain .” Advances in Life Course Research 30 : 111–123. [ Google Scholar ]
• Feng Shuaizhang, Krueger Alan B., and Oppenheimer Michael. 2010. “ Linkages among Climate Change, Crop Yields and Mexico-US Cross-border Migration .” Proceedings of the National Academy of Sciences 107 : 14257–62. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Findlay Allan, David McCollum Rory Coulter, and Gayle Vernon. 2015. “ New Mobilities across the Life Course: A Framework for Analyzing Demographically Linked Drivers of Migration .” Population, Space and Place 21 :390–402. [ Google Scholar ]
• Fuller Theodore D, Kamnuansilpa Peerasit, and Lightfoot Paul. 1990. “ Urban Ties of rural Thais .” International Migration Review 24 : 534–62. [ PubMed ] [ Google Scholar ]
• Fuller Theodore D, Lightfoot Paul, and Kamnuansilpa Peerasit. 1985. Rural-urban Mobility in Thailand : A Decision-making Approach. Demography 22 : 565–79. [ PubMed ] [ Google Scholar ]
• Fussell Elizabeth, Curtis Katharine J., and DeWaard Jack. 2014. “ Recovery Migration to the City of New Orleans after Hurricane Katrina: A Migration Systems Approach .” Population and Environment 35 : 305–22. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Garip Filiz, and Curran Sara. 2010. “ Increasing Migration, Diverging Communities: Changing Character of Migrant Streams in Rural Thailand .” Population Research and Policy Review 29 : 659–85. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Garip Filiz, Eskici Burak, and Snyder Ben. 2015. “ Network Effects in Migrant remittances: Evidence from Household, Sibling, and Village Ties in Nang Rong, Thailand .” American Behavioral Scientist 59 : 1066–82. [ Google Scholar ]
• Garip Filiz. 2008. “ Social Capital and Migration: How do Similar Resources Lead to Divergent Outcomes ?” Demography 45 : 591–617. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Garip Filiz. 2012a. “ An Integrated Analysis of Migration and Remittances: Modeling Migration as a Mechanism for Selection .” Population Research and Policy Review 31 : 637–63. [ Google Scholar ]
• Garip Filiz. 2012b. “ Repeat Migration and Remittances as Mechanisms for Wealth Inequality in 119 Communities from the Mexican Migration Project Data .” Demography 49 : 1335–60. [ PubMed ] [ Google Scholar ]
• Garip Filiz. 2014. “ The Impact of Migration and Remittances on Wealth Accumulation and Distribution in Rural Thailand .” Demography 51 : 673–698. [ PubMed ] [ Google Scholar ]
• Geist Claudia, and McManus Patricia A.. 2008. “ Geographical Mobility over the Life Course: Motivations and Implications .” Population, Space and Place 14 :283–303. [ Google Scholar ]
• Gioli Giovanna, Hugo Graeme, Costa Maria Manez, and Scheffran Jurgen. 2016. “ Human Mobility, Climate Adaptation, and Development .” Migration and Development 5 : 165–70. [ Google Scholar ]
• Harris John R., and Todaro Michael P.. 1970. “ Migration, Unemployment and Development: A Two- sector Analysis .” American Economic Review 60 : 126–142. [ Google Scholar ]
• Heumann Benjamin W., Walsh Stephen J., and McDaniel Phillip M.. 2011. “ Assessing the Application of a Geographic Presence-only Model for Land Suitability Mapping .” Ecological Informatics 6 : 257–69. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Heumann Benjamin W., Walsh Stephen J., Verdery Ashton M., McDaniel Phillip M., and Rindfuss Ronald R.. 2013. “ Land Suitability Modeling using a Geographic Socio-environmental Niche-based Approach: A case Study from Northeastern Thailand .” Annals of the Association of American Geographers 103 : 764–84. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Hunter Lori, Luna Jessica K., and Norton Rachel M.. 2015. “ Environmental Dimensions of Migration .” Annual Review of Sociology 41 : 377–97. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Hugo Graeme. 1996. “ Environmental Concerns and International Migration .” International Migration Review 21 ( S1 ): 21–33. [ PubMed ] [ Google Scholar ]
• Jampaklay Aree. 2006. “ How Does Leaving Home Affect Marital Timing? An Event-history Analysis of Migration and Marriage in Nang Rong, Thailand .” Demography 43 : 711–25. [ PubMed ] [ Google Scholar ]
• Jones James W., Hoogenboom Gerrit, Porter Cheryl H., Boote Ken J., Batchelor William D., Hunt LA, Wilkens Paul W., Singh Upendra, Gijsman Arjan J., and Ritchie Joe T.. 2003. “ The DSSAT Cropping System Model .” European Journal of Agronomy 18 : 235–65. [ Google Scholar ]
• Klabunde Anna and Willekens Frans. 2016. “ Decision-making in Agent-based Models of Migration: State of the Art and Challenges .” European Journal of Population 32 : 73–97. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Korinek Kim, Entwisle Barbara, and Jampaklay Aree. 2005. “ Through Thick and Thin: Layers of Social Ties and Urban Settlement among Thai Migrants .” American Sociological Review 70 :779–800. [ Google Scholar ]
• Vally Koubi, Stoll Sebastian, and Spilker Gabrielle. 2016. “ Perceptions of Environmental Change and Migration Decisions .” Climatic Change 138 : 439–51. [ Google Scholar ]
• Kubik Zaneta and Maurel Mathilde. 2016. “ Weather Shocks, Agricultural Production and Migration: Evidence from Tanzania .” The Journal of Development Studies 52 : 665–80. [ Google Scholar ]
• Lee Everett S. 1966. “ A Theory of Migration .” Demography 3 : 47–57 [ Google Scholar ]
• Leyk Stefan, Runfola Dan, Nawrotzki Raphael J., Hunter Lori M., and Riosmena Fernando. 2017. “ Internal and International Mobility as Adaptation to Climatic Variability in Contemporary Mexico: Evidence from the Integration of Census and Satellite Data .” Population, Space and Place 23 . doi: 10.1002/psp.2047. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Limsakul Atsamon, Limjirakan Sangchan, Sriburi Thavivongse, and Boochub Suttamanuswong. 2011. “ Trends in Temperature and its Extremes in Thailand .” Thai Environmental Engineering Journal 25 : 9–16. [ Google Scholar ]
• Loebach Peter. 2016. “ Household Migration as a Livelihood Adaptation in Response to a Natural Disaster: Nicaragua and Hurricane Mitch .” Population and Environment 38 : 185–206. [ Google Scholar ]
• Macy Michael W., and Willer Robert. 2002. “ From Factors to Actors: Computational Sociology and Agent-based Modeling .” Annual Review of Sociology 28 : 143–66. [ Google Scholar ]
• Malanson George P., Verdery Ashton M., Walsh Stephen J., Sawangdee Yothin, Heumann Benjamin W., McDaniel Philip M., Frizzelle Brian G., Williams Nathalie E., Yao Xiaozheng, Entwisle Barbara, and Rindfuss Ronald R.. 2014. “ Changing Crops in Response to Climate: Virtual Nang Rong, Thailand in an Agent Based Simulation .” Applied Geography 53 :202–12. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Marchirori Luca, Maystadt Jean-Francois, and Schumacher Ingmar. 2012. “ The Impact of Weather Anomalies on Migration in sub-Saharan Africa .” Journal of Environmental Economics and Management 63 : 355–74. [ Google Scholar ]
• Massey Douglas S. 1990. “ Social Structure, Household Strategies, and the Cumulative Causation of Migration .” Population Index 56 : 3–26. [ PubMed ] [ Google Scholar ]
• Massey Douglas S., Arango Joaquin, Hugo Graeme, Kouaouci Ali, Pellegrino Adela, and Taylor J. Edward. 1993. “ Theories of International Migration: A Review and Appraisal .” Population and Development Review 19 : 431–66. [ Google Scholar ]
• Massey Douglas S., Axinn William G., and Ghimire Dirgha J.. 2010. “ Environmental Change and Out- migration: Evidence from Nepal .” Population and Environment 32 : 109–36. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Massey Douglas S., Durand Jorge, and Pren Karen A.. 2016. “ Why Border Enforcement Backfired .” American Journal of Sociology 121 : 1557–1600. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Matthews RB, and Hunt LA. 1994. “GUMCAS: A Model Describing the Growth of Cassava (Manihot esculenta L. Crantz) .” Field Crops Research 36 : 69–84. [ Google Scholar ]
• McLeman Robert A. 2014. Climate and Human Migration: Past Experiences, Future Challenges . NY: Cambridge. [ Google Scholar ]
• Morrissey James William. 2013. “ Understanding the Relationship between Environmental Change and Migration: The Development of an Effects Framework Based on the Case of Northern Ethiopia .” Global Environmental Change 23 : 1501–10. [ Google Scholar ]
• Mouw Ted, Chavez Sergio, Edelblute Heather, and Verdery Ashton. 2014. “ Binational Social Networks and Assimilation: A Test of the Importance of Transnationalism .” Social Problems 61 : 329–59. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Mueller Valerie, Gray Clark, and K. Kosec. 2014. “ Heat Stress Increases Long-term Human Migration in Rural Pakistan .” Nature Climate Change 4 : 182–185. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Naivinit W, Page Christophe Le, Trebuil Guy, and Gajaseni Nantana. 2010. “ Participatory Agent-based Modeling and Simulation of Rice Production and Labor Migrations in Northeast Thailand .” Environmental Modelling & Software 25 : 1345–58. [ Google Scholar ]
• Nawrotski Raphael J., Riosmena Fernando, Hunter Lori M. and Runfola Daniel M.. 2015. “ Amplification or Suppression: Social Networks and the Climate Change-Migration Association in Rural Mexico .” Global Environmental Change 35 : 463–74. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Nawrotzki Raphael J., Runfola Daniel M., Hunter Lori M., and Riosema Fernando. 2016. “ Domestic and International Climate Migration from Rural Mexico .” Human Ecology 44 : 687–699. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Naylor Rosamond L., Battisti David S., Vimont Daniel, Falcon Walter P., and Burke Marshall B.. 2007. “ Assessing risks of climate variability and climate change for Indonesian rice agriculture .” Proceedings of the National Academy of Sciences 104 : 7752–7. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Nickl Elsa, Willmont Cort J., Matsuura Kenji, and Robeson Scott M.. 2010. “ Changes in annual land- surface precipitation over the twentieth and early twenty-first century .” Proceedings of the National Academy of Sciences 100 : 729–39. [ Google Scholar ]
• OECD. 2017. Interactions entre politiques publiques, migrations et développement en Côte d’Ivoire Pathways to Development Series . Paris: OECD. [ Google Scholar ]
• Piotrowski Martin. 2008. “ Intergenerational relations in a context of industrial transition: A study of agricultural labor from migrants in Nang Rong, Thailand .” Journal of cross-cultural gerontology 23 : 17–38. [ PubMed ] [ Google Scholar ]
• Rapoport Hillel, and Frederic Docquier, F. 2006. “ The economics of migrants’ remittances .” Handbook on the Economics of Giving , Reciprocity and Altruism 2 : 1135–98. [ Google Scholar ]
• Rindfuss Ronald R. 1991. “ The young adult years: Diversity, structural change, and fertility .” Demography 28 :493–512. [ PubMed ] [ Google Scholar ]
• Rindfuss Ronald R., Prasartkul Pramote, Walsh Stephen J., Entwisle Barbara, Sawangdee Yothin, and Vogler John B.. 20014. “Household-parcel linkages in Nang Rong, Thailand” Pp. 131–172 in Fox Jefferson, Rindfuss Ronald R., Walsh Stephen J., and Mishra Vinod (eds.), People and the Environment: Approaches for Linking Household and Community Surveys to Remote Sensing and GIS . Boston, MA: Kluwer Academic Publishers. [ Google Scholar ]
• Rogers Andrei, Raquillet Richard, and Castro Luis J.. 1978. “ Model migration schedules and their applications .” Environment and Planning A 10 : 475–502. [ Google Scholar ]
• Rozelle Scott, Taylor J. Edward, and deBrauw Alan. 1999. “ Migration, remittances, and agricultural productivity in China .” American Economic Review 89 : 287–91. [ Google Scholar ]
• Sandberg‐Thoma Sara E., Snyder Anastasia R., and Jang Bohyun Joy. 2015. “ Exiting and returning to the parental home for boomerang kids .” Journal of Marriage and Family 77 : 806–18. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Schelling Thomas C. 1971. “ Dynamic models of segregation .” Journal of mathematical sociology 2 : 143–86. [ Google Scholar ]
• Smith Jeffrey A., and Burow Jessica. 2018. “ Using Ego Network Data to Inform Agent-based Models of Diffusion .” Sociological Methods & Research : 10.1177/0049124118769100. [ CrossRef ] [ Google Scholar ]
• South Scott J., and Lei Lei. 2015. “ Failures-to-launch and boomerang kids: Contemporary determinants of leaving and returning to the parental home .” Social Forces 94 : 863–90. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Stark Oded, and Bloom David E.. 1985. “ The new economics of labor migration .” American Economic Review 75 : 173–8. [ Google Scholar ]
• Taylor Edward J. 1999. “ The new economics of labour migration and the role of remittances in the migration process .” International Migration 37 : 63–88. [ PubMed ] [ Google Scholar ]
• Thiede Brian C., Gray Clark L., and Mueller Valerie. 2016. “ Climate variability and inter-provincial migration in South America, 1970–2011 .” Global Environmental Change 41 : 228–40. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Thiede Brian C. and Gray Clark. L.. 2017. “ Erratum to: Heterogeneous climate effects on human migration in Indonesia .” Population and Environment 39 : 173–95. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Timsina Jagadish, and Humphreys E. 2006. “ Performance of CERES-Rice and CERES-Wheat models in rice–wheat systems: a review .” Agricultural systems 90 : 5–31. [ Google Scholar ]
• Tong Yuying, and Piotrowski Martin. 2010. “ The effect of remittances on return migration and its relation to household wealth: The case of rural Thailand .” Asia-Pacific Population 25 : 53. [ Google Scholar ]
• United Nations, Department of Economic and Social Affairs. 2015. Trends in International Migrant Stock: The 2015 revision (United Nations database , POP/DB/MIG/Stock/Rev.2015).
• Verdery Ashton M., Entwisle Barbara, Faust Katherine, and Rindfuss Ronald R.. 2012. “ Social and spatial networks: Kinship distance and dwelling unit proximity in rural Thailand .” Social Networks 34 : 112–27. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Walsh Stephen J., Malanson George P., Entwisle Barbara, Rindfuss Ronald R., Mucha Peter J., Heumann Benjamin W., McDamniel Philip M., Frizzelle Brian G., Verdery Ashton M., Williams Nathalie E., Yao Xiaozheng, and Ding Deng. 2013. “ Design of an agent-based model to examine population– environment interactions in Nang Rong District, Thailand .” Applied Geography 39 : 183–98. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Wantanabe Megumi, Olson Kristen M., and Falci Christina D.. 2017. “ Social Isolation, Survey Nonresponse, and Nonresponse Bias: An Empirical Evaluation Using Social Network Data within Organization .” Social Science Research 63 : 324–38. [ PubMed ] [ Google Scholar ]
• Warner Cody and Sharp Gregory. 2016. “ The short- and long-term effects of life events on residential mobility .” Advances in Life Course Research 27 : 1–15. [ Google Scholar ]

Programs submenu

Regions submenu, topics submenu, chinese interference in taiwan's 2024 elections and lessons learned, the political pressure behind scotland on the ballot, senegal’s democracy: how did we get here and what shall we expect, commonhealth live with dr. julie gerberding and former senator richard burr.

• Abshire-Inamori Leadership Academy
• Aerospace Security Project
• Africa Program
• Americas Program
• Arleigh A. Burke Chair in Strategy
• Asia Maritime Transparency Initiative
• Asia Program
• Australia Chair
• Brzezinski Chair in Global Security and Geostrategy
• Brzezinski Institute on Geostrategy
• Chair in U.S.-India Policy Studies
• China Power Project
• Chinese Business and Economics
• Defending Democratic Institutions
• Defense-Industrial Initiatives Group
• Defense 360
• Defense Budget Analysis
• Diversity and Leadership in International Affairs Project
• Economics Program
• Emeritus Chair in Strategy
• Energy Security and Climate Change Program
• Europe, Russia, and Eurasia Program
• Freeman Chair in China Studies
• Futures Lab
• Geoeconomic Council of Advisers
• Global Food and Water Security Program
• Global Health Policy Center
• Hess Center for New Frontiers
• Human Rights Initiative
• Humanitarian Agenda
• Intelligence, National Security, and Technology Program
• International Security Program
• Japan Chair
• Kissinger Chair
• Korea Chair
• Langone Chair in American Leadership
• Middle East Program
• Missile Defense Project
• Project on Fragility and Mobility
• Project on Nuclear Issues
• Project on Prosperity and Development
• Project on Trade and Technology
• Renewing American Innovation Project
• Scholl Chair in International Business
• Smart Women, Smart Power
• Southeast Asia Program
• Stephenson Ocean Security Project
• Strategic Technologies Program
• Transnational Threats Project
• Wadhwani Center for AI and Advanced Technologies
• All Regions
• Australia, New Zealand & Pacific
• Middle East
• Russia and Eurasia
• American Innovation
• Civic Education
• Climate Change
• Cybersecurity
• Defense Budget and Acquisition
• Defense and Security
• Energy and Sustainability
• Food Security
• Gender and International Security
• Geopolitics
• Global Health
• Human Rights
• Humanitarian Assistance
• Intelligence
• International Development
• Maritime Issues and Oceans
• Missile Defense
• Nuclear Issues
• Transnational Threats
• Water Security

The White House Report on Climate Migration, Explained

Photo: Sanjay Kanojia/AFP via Getty Images

Critical Questions by Erol Yayboke, Catherine Nzuki , and Sierra Ballard

Published November 5, 2021

The Biden-Harris administration released a major report on climate migration on Thursday, October 21, 2021. Strategically timed to coincide with final preparations for the UN Climate Change Conference of the Parties (COP26) in early November 2021, the report demonstrates a nuanced and comprehensive understanding of the interlinked challenges of climate change and migration. Though light on specific next steps, the report provides much-needed attention and framing. As global attention turns to the climate talks in Glasgow, policymakers would be well-served to have a copy of the 37-page report in their briefing packets.

Q1: What is the report?

A1: The Biden-Harris administration’s Report on the Impact of Climate Change on Migration outlines the inextricable links between climate change and migration, marking a first for any U.S. administration. It identifies how migration will evolve as a critical, and sometimes necessary form of adaptation in the wake of climate impacts. The report presents the underlying factors that could affect the severity of climate migration impacts, including:

• The ability of vulnerable communities in conflict settings and areas of weak governance to navigate more acute climate-related impacts and displacement in the absence of effective protection measures;
• The potential of existing programs to reduce the risk of forced migration and displacement via development of local capacities for climate-informed decisionmaking and adaptation;
• The prevention of further conflict or instability and the protection and resettlement of displaced individuals with adequate support to host communities;
• The level of engagement with and learning from regional and civil society groups that can strengthen response capabilities and minimize the impact of climate migration on individuals and communities; and
• The effectiveness—and adequate support—of multilateral efforts to mitigate negative effects and to support all of the above.

The report calls for the development of strategies that would allow the humane, safe, and proactive management of climate migration flows. Such management, according to the report, should be based on awareness of context-specific environmental characteristics that enhance instability. It specifically accentuates the need to provide necessary resources to mitigate the humanitarian consequence on displaced individuals. Beyond highlighting the need for active foreign assistance and the protection of affected individuals, the report considers the geopolitical implications of climate migration and the principal role of multilateral engagement in response. Ultimately, the report thoroughly contextualizes the risks and opportunities of climate migration in a multilevel review, outlining high-level recommendations to enable future policy action.

Q2: Why was it released now?

A2: The report was published in direct response to President Joe Biden’s February 9, 2021, Executive Order 14013 , Rebuilding and Enhancing Programs to Resettle Refugees and Planning for the Impact of Climate Change on Migration. The timing of this report’s release was no coincidence just 10 days before the president himself prepared to attend the COP26 international climate conference. The report endorses multilateral engagement and in particular the important role of the United Nations. It reaffirms the commitment of the United States to engage with key resolutions and efforts and calls for greater diplomatic action to further key agreements on climate. It explicitly names the Paris Agreement and “mainstream efforts to strengthen resilience and reduce risks of displacement” resulting from events related to environmental disasters and climate change. As just one in a series of reports mandated by the administration's executive orders on climate action in January and February 2021, this analysis reinforces the United States' commitment to climate action, humanitarian principles, and evidence-based decisionmaking. With a growing global sense of urgency over the climate crisis, the timely release of this report ahead of the COP26 summit is a signal of the administration's aspirations to restore the United States as a trusted global leader even on some of the hardest challenges ahead.

Q3: Why is this climate migration report important?

A3: Credit where credit is due: the Report on the Impact of Climate Change on Migration is an important document. It thoroughly and accurately assesses the nuanced interrelationship between climate change and migration. When conducted in a regular, safe, and orderly manner, human migration offers solutions to many of the economic and demographic challenges ahead of us. But climate change is forcing people to make difficult decisions about leaving home on quicker timelines and under more precarious circumstances. In this way, climate change is a threat multiplier, exacerbating the underlying drivers of conflict, straining public budgets, offering new opportunities for smugglers and other bad actors, widening resource inequities, and increasing political and social tensions. For too long, climate and migration issues were treated as siloed and separate issues; this report recognizes that the two are inextricably linked.

Recognizing these linkages, the report offers a framework to bring together stakeholders across the U.S. government to form a coordinated response to climate migration risks and outcomes. The report calls for the establishment of a standing interagency policy process on climate change and migration to coordinate U.S. government efforts with a focus on improving analytics, programming, and investments, and a legislative strategy to ingrain these issues as priorities beyond the current administration. Among other priorities, the process is meant to address climate-related migration in the short term and create strategies to tackle climate impacts in the long term. Action seems far off when one of the most exciting parts of the report is the creation of a policy process, but such a coordination mechanism is typically a first step on the road to major interagency action. In this way, the report establishes the foundation on which future policy action—nationally and internationally—can be taken. This will, of course, be difficult given political constraints and additional global challenges like the ongoing Covid-19 pandemic. The creation of new legal pathways, for example, Temporary Protected Status, which a 2020 CSIS policy brief on climate migration called for, will take time and promises to be highly contentious, especially since the most durable solutions will require congressional action. But the first challenge is always to accurately define the problem, which this report does well. The administration engaged a diverse array of stakeholders over the course of writing it and the final version shows that they were listening.

Critics will say that the report does not go far enough given the months it took to write it. It is true and noteworthy that climate change is an immediate threat having immediate impacts disproportionately on the places least responsible for it. However, while light on specific, actionable recommended policy actions, the report was never meant to mandate policy. That comes next and will be closely watched by all, particularly in instances where U.S. policy alone will be insufficient. Climate migration is a global challenge, one that will require not only multilateral engagement, but updates to how relevant international laws and institutions can adapt to climate realities.

As officials from around the world are now gathered in Glasgow for COP26, this and the other climate-related reports released by the Biden-Harris administration show that climate conversations must move beyond efforts to simply reduce emissions. While critical, these efforts will do little to address the challenges facing vulnerable populations today—challenges that are laid out very clearly in the Report on the Impact of Climate Change on Migration .

Erol Yayboke is a senior fellow with the International Security Program and director of the Project on Fragility and Mobility (PFM) at the Center for Strategic and International Studies (CSIS) in Washington, D.C. Catherine Nzuki is a program coordinator with the CSIS PFM. Sierra Ballard is a research intern with the CSIS PFM.

Critical Questions  is produced by the Center for Strategic and International Studies (CSIS), a private, tax-exempt institution focusing on international public policy issues. Its research is nonpartisan and nonproprietary. CSIS does not take specific policy positions. Accordingly, all views, positions, and conclusions expressed in this publication should be understood to be solely those of the author(s).

© 2021 by the Center for Strategic and International Studies. All rights reserved.

Erol Yayboke

Catherine Nzuki

Sierra Ballard

Programs & projects.

The Great Climate Migration Has Begun

By Abrahm Lustgarten July 23, 2020

• Share full article

Today, 1% of the world is a barely livable hot zone.

By 2070, that portion could go up to 19%.

Billions of people call this land home.

Where will they go?

The Great Climate Migration By Abrahm Lustgarten | Photographs by Meridith Kohut

Early in 2019, a year before the world shut its borders completely, Jorge A. knew he had to get out of Guatemala. The land was turning against him. For five years, it almost never rained. Then it did rain, and Jorge rushed his last seeds into the ground. The corn sprouted into healthy green stalks, and there was hope — until, without warning, the river flooded. Jorge waded chest-deep into his fields searching in vain for cobs he could still eat. Soon he made a last desperate bet, signing away the tin-roof hut where he lived with his wife and three children against a $1,500 advance in okra seed. But after the flood, the rain stopped again, and everything died. Jorge knew then that if he didn’t get out of Guatemala, his family might die, too.

This article, the first in a series on global climate migration, is a partnership between ProPublica and The New York Times Magazine, with support from the Pulitzer Center. Read Part 2 and Part 3 , and more about the data project that underlies the reporting.

Even as hundreds of thousands of Guatemalans fled north toward the United States in recent years, in Jorge’s region — a state called Alta Verapaz, where precipitous mountains covered in coffee plantations and dense, dry forest give way to broader gentle valleys — the residents have largely stayed. Now, though, under a relentless confluence of drought, flood, bankruptcy and starvation, they, too, have begun to leave. Almost everyone here experiences some degree of uncertainty about where their next meal will come from. Half the children are chronically hungry, and many are short for their age, with weak bones and bloated bellies. Their families are all facing the same excruciating decision that confronted Jorge.

The odd weather phenomenon that many blame for the suffering here — the drought and sudden storm pattern known as El Niño — is expected to become more frequent as the planet warms. Many semiarid parts of Guatemala will soon be more like a desert. Rainfall is expected to decrease by 60 percent in some parts of the country, and the amount of water replenishing streams and keeping soil moist will drop by as much as 83 percent. Researchers project that by 2070, yields of some staple crops in the state where Jorge lives will decline by nearly a third.

Scientists have learned to project such changes around the world with surprising precision, but — until recently — little has been known about the human consequences of those changes. As their land fails them, hundreds of millions of people from Central America to Sudan to the Mekong Delta will be forced to choose between flight or death. The result will almost certainly be the greatest wave of global migration the world has seen.

In March, Jorge and his 7-year-old son each packed a pair of pants, three T-shirts, underwear and a toothbrush into a single thin black nylon sack with a drawstring. Jorge’s father had pawned his last four goats for $2,000 to help pay for their transit, another loan the family would have to repay at 100 percent interest. The coyote called at 10 p.m. — they would go that night. They had no idea then where they would wind up, or what they would do when they got there.

From decision to departure, it was three days. And then they were gone.

For most of human history, people have lived within a surprisingly narrow range of temperatures, in the places where the climate supported abundant food production. But as the planet warms, that band is suddenly shifting north. According to a pathbreaking recent study in the journal Proceedings of the National Academy of Sciences, the planet could see a greater temperature increase in the next 50 years than it did in the last 6,000 years combined. By 2070, the kind of extremely hot zones, like in the Sahara, that now cover less than 1 percent of the earth’s land surface could cover nearly a fifth of the land, potentially placing one of every three people alive outside the climate niche where humans have thrived for thousands of years. Many will dig in, suffering through heat, hunger and political chaos, but others will be forced to move on. A 2017 study in Science Advances found that by 2100, temperatures could rise to the point that just going outside for a few hours in some places, including parts of India and Eastern China, “will result in death even for the fittest of humans.”

People are already beginning to flee. In Southeast Asia, where increasingly unpredictable monsoon rainfall and drought have made farming more difficult, the World Bank points to more than eight million people who have moved toward the Middle East, Europe and North America. In the African Sahel, millions of rural people have been streaming toward the coasts and the cities amid drought and widespread crop failures. Should the flight away from hot climates reach the scale that current research suggests is likely, it will amount to a vast remapping of the world’s populations.

Listen to This Article

Migration can bring great opportunity not just to migrants but also to the places they go. As the United States and other parts of the global North face a demographic decline, for instance, an injection of new people into an aging work force could be to everyone’s benefit. But securing these benefits starts with a choice: Northern nations can relieve pressures on the fastest-warming countries by allowing more migrants to move north across their borders, or they can seal themselves off, trapping hundreds of millions of people in places that are increasingly unlivable. The best outcome requires not only good will and the careful management of turbulent political forces; without preparation and planning, the sweeping scale of change could prove wildly destabilizing. The United Nations and others warn that in the worst case, the governments of the nations most affected by climate change could topple as whole regions devolve into war.

The stark policy choices are already becoming apparent. As refugees stream out of the Middle East and North Africa into Europe and from Central America into the United States, an anti-immigrant backlash has propelled nationalist governments into power around the world. The alternative, driven by a better understanding of how and when people will move, is governments that are actively preparing, both materially and politically, for the greater changes to come.

Projected percentage decrease by 2070 in the yield of the rice crop in Alta Verapaz, Guatemala:

Last summer, I went to Central America to learn how people like Jorge will respond to changes in their climates. I followed the decisions of people in rural Guatemala and their routes to the region’s biggest cities, then north through Mexico to Texas. I found an astonishing need for food and witnessed the ways competition and poverty among the displaced broke down cultural and moral boundaries. But the picture on the ground is scattered. To better understand the forces and scale of climate migration over a broader area, The New York Times Magazine and ProPublica joined with the Pulitzer Center in an effort to model, for the first time, how people will move across borders.

We focused on changes in Central America and used climate and economic-development data to examine a range of scenarios. Our model projects that migration will rise every year regardless of climate, but that the amount of migration increases substantially as the climate changes. In the most extreme climate scenarios, more than 30 million migrants would head toward the U.S. border over the course of the next 30 years.

Migrants move for many reasons, of course. The model helps us see which migrants are driven primarily by climate, finding that they would make up as much as 5 percent of the total. If governments take modest action to reduce climate emissions, about 680,000 climate migrants might move from Central America and Mexico to the United States between now and 2050. If emissions continue unabated, leading to more extreme warming, that number jumps to more than a million people. (None of these figures include undocumented immigrants, whose numbers could be twice as high.)

The model shows that the political responses to both climate change and migration can lead to drastically different futures.

In one scenario, globalization — with its relatively open borders — continues.

As the climate changes, drought and food insecurity drive rural residents in Mexico and Central America out of the countryside.

Millions seek relief, first in big cities, spurring a rapid and increasingly overwhelming urbanization.

Then they move farther north, pushing the largest number of migrants toward the United States. The projected number of migrants arriving from Central America and Mexico rises to 1.5 million a year by 2050, from about 700,000 a year in 2025.

We modeled another scenario in which the United States hardens its borders. People are turned back, and economic growth in Central America slows, as does urbanization.

In this case, Central America’s population surges, and the rural hollowing reverses as the birthrate rises, poverty deepens and hunger grows — all with hotter weather and less water.

That version of the world leaves tens of millions of people more desperate and with fewer options. Misery reigns, and large populations become trapped.

As with much modeling work, the point here is not to provide concrete numerical predictions so much as it is to provide glimpses into possible futures. Human movement is notoriously hard to model, and as many climate researchers have noted, it is important not to add a false precision to the political battles that inevitably surround any discussion of migration. But our model offers something far more potentially valuable to policymakers: a detailed look at the staggering human suffering that will be inflicted if countries shut their doors.

In recent months, the coronavirus pandemic has offered a test run on whether humanity has the capacity to avert a predictable — and predicted — catastrophe. Some countries have fared better. But the United States has failed. The climate crisis will test the developed world again, on a larger scale, with higher stakes. The only way to mitigate the most destabilizing aspects of mass migration is to prepare for it, and preparation demands a sharper imagining of where people are likely to go, and when.

I. A Different Kind of Climate Model

In November 2007, Alan B. Krueger, a labor economist known for his statistical work on inequality, walked into the Princeton University offices of Michael Oppenheimer, a leading climate geoscientist, and asked him whether anyone had ever tried to quantify how and where climate change would cause people to move.

Earlier that year, Oppenheimer helped write the U.N. Intergovernmental Panel on Climate Change report that, for the first time, explored in depth how climate disruption might uproot large segments of the global population. But as groundbreaking as the report was — the U.N. was recognized for its work with a Nobel Peace Prize — the academic disciplines whose work it synthesized were largely siloed from one another. Demographers, agronomists and economists were all doing their work on climate change in isolation, but understanding the question of migration would have to include all of them.

Together, Oppenheimer and Krueger, who died in 2019, began to chip away at the question, asking whether tools typically used by economists might yield insight into the environment’s effects on people’s decision to migrate. They began to examine the statistical relationships — say, between census data and crop yields and historical weather patterns — in Mexico to try to understand how farmers there respond to drought. The data helped them create a mathematical measure of farmers’ sensitivity to environmental change — a factor that Krueger could use the same way he might evaluate fiscal policies, but to model future migration.

Their study, published in 2010 in Proceedings of the National Academy of Sciences, found that Mexican migration to the United States pulsed upward during periods of drought and projected that by 2080, climate change there could drive 6.7 million more people toward the Southern U.S. border. “It was,” Oppenheimer said, “one of the first applications of econometric modeling to the climate-migration problem.”

The modeling was a start. But it was hyperlocal instead of global, and it left open huge questions: how cultural differences might change outcomes, for example, or how population shifts might occur across larger regions. It was also controversial, igniting a backlash among climate-change skeptics, who attacked the modeling effort as “guesswork” built on “tenuous assumptions” and argued that a model couldn’t untangle the effect of climate change from all the other complex influences that determine human decision-making and migration. That argument eventually found some traction with migration researchers, many of whom remain reluctant to model precise migration figures.

But to Oppenheimer and Krueger, the risks of putting a specific shape to this well established but amorphous threat seemed worth taking. In the early 1970s, after all, many researchers had made a similar argument against using computer models to forecast climate change, arguing that scientists shouldn’t traffic in predictions. Others ignored that advice, producing some of the earliest projections about the dire impact of climate change, and with them some of the earliest opportunities to try to steer away from that fate. Trying to project the consequences of climate-driven migration, to Oppenheimer, called for similarly provocative efforts. “If others have better ideas for estimating how climate change affects migration,” he wrote in 2010, “they should publish them.”

Since then, Oppenheimer’s approach has become common. Dozens more studies have applied econometric modeling to climate-related problems, seizing on troves of data to better understand how environmental change and conflict each lead to migration and clarify how the cycle works. Climate is rarely the main cause of migration, the studies have generally found, but it is almost always an exacerbating one.

As they have looked more closely, migration researchers have found climate’s subtle fingerprints almost everywhere. Drought helped push many Syrians into cities before the war, worsening tensions and leading to rising discontent; crop losses led to unemployment that stoked Arab Spring uprisings in Egypt and Libya ; Brexit, even, was arguably a ripple effect of the influx of migrants brought to Europe by the wars that followed. And all those effects were bound up with the movement of just two million people. As the mechanisms of climate migration have come into sharper focus — food scarcity, water scarcity and heat — the latent potential for large-scale movement comes to seem astronomically larger.

North Africa’s Sahel provides an example. In the nine countries stretching across the continent from Mauritania to Sudan, extraordinary population growth and steep environmental decline are on a collision course. Past droughts, most likely caused by climate change, have already killed more than 100,000 people there. And the region — with more than 150 million people and growing — is threatened by rapid desertification, even more severe water shortages and deforestation. Today researchers at the United Nations estimate that some 65 percent of farmable lands have already been degraded. “My deep fear,” said Solomon Hsiang, a climate researcher and economist at the University of California, Berkeley, is that Africa’s transition into a post-climate-change civilization “leads to a constant outpouring of people.”

The story is similar in South Asia, where nearly one-fourth of the global population lives. The World Bank projects that the region will soon have the highest prevalence of food insecurity in the world. While some 8.5 million people have fled already — resettling mostly in the Persian Gulf — 17 million to 36 million more people may soon be uprooted, the World Bank found. If past patterns are a measure, many will settle in India’s Ganges Valley; by the end of the century, heat waves and humidity will become so extreme there that people without air-conditioning will simply die.

If it is not drought and crop failures that force large numbers of people to flee, it will be the rising seas. We are now learning that climate scientists have been underestimating the future displacement from rising tides by a factor of three, with the likely toll being some 150 million globally. New projections show high tides subsuming much of Vietnam by 2050 — including most of the Mekong Delta, now home to 18 million people — as well as parts of China and Thailand, most of southern Iraq and nearly all of the Nile Delta, Egypt’s breadbasket. Many coastal regions of the United States are also at risk.

Through all the research, rough predictions have emerged about the scale of total global climate migration — they range from 50 million to 300 million people displaced — but the global data is limited, and uncertainty remained about how to apply patterns of behavior to specific people in specific places. Now, though, new research on both fronts has created an opportunity to improve the models tremendously. A few years ago, climate geographers from Columbia University and the City University of New York began working with the World Bank to build a next-generation tool to establish plausible migration scenarios for the future. The idea was to build on the Oppenheimer-style measure of response to the environment with other methods of analysis, including a “gravity” model, which assesses the relative attractiveness of destinations with the hope of mathematically anticipating where migrants might end up. The resulting report, published in early 2018, involved six European and American institutions and took nearly two years to complete.

The bank’s work targeted climate hot spots in sub-Saharan Africa, South Asia and Latin America, focusing not on the emergency displacement of people from natural disasters but on their premeditated responses to what researchers call “slow-onset” shifts in the environment. They determined that as climate change progressed in just these three regions alone, as many as 143 million people would be displaced within their own borders, moving mostly from rural areas to nearby towns and cities. The study, though, wasn’t fine-tuned to specific climatic changes like declining groundwater. And it didn’t even try to address the elephant in the room: How would the climate push people to migrate across international borders?

In early 2019, The Times Magazine and ProPublica, with support from the Pulitzer Center, hired an author of the World Bank report — Bryan Jones, a geographer at Baruch College — to add layers of environmental data to its model, making it even more sensitive to climatic change and expanding its reach. Our goal was to pick up where the World Bank researchers left off, in order to model, for the first time, how people would move between countries, especially from Central America and Mexico toward the United States.

First we gathered existing data sets — on political stability, agricultural productivity, food stress, water availability, social connections, weather and much more — in order to approximate the kaleidoscopic complexity of human decision-making.

Then we started asking questions: If crop yields continue to decline because of drought, for instance, and people are forced to respond by moving, as they have in the past, can we see where they will go and see what new conditions that might introduce? It’s very difficult to model how individual people think or to answer these questions using individual data points — often the data simply doesn’t exist. Instead of guessing what Jorge A. will do and then multiplying that decision by the number of people in similar circumstances, the model looks across entire populations, averaging out trends in community decision-making based on established patterns, then seeing how those trends play out in different scenarios.

Projected percentage of city dwellers who will live in slums by 2030:

In all, we fed more than 10 billion data points into our model. Then we tested the relationships in the model retroactively, checking where historical cause and effect could be empirically supported, to see if the model’s projections about the past matches what really happened. Once the model was built and layered with both approaches — econometric and gravity — we looked at how people moved as global carbon concentrations increased in five different scenarios, which imagine various combinations of growth, trade and border control, among other factors. (These scenarios have become standard among climate scientists and economists in modeling different pathways of global socioeconomic development.)

Only a supercomputer could efficiently process the work in its entirety; estimating migration from Central America and Mexico in one case required uploading our query to a federal mainframe housed in a building the size of a small college campus outside Cheyenne, Wyo., run by the National Center for Atmospheric Research, where even there it took four days for the machine to calculate its answers. (A more detailed description of the data project can be found at propublica.org/migration-methodology .)

The results are built around a number of assumptions about the relationships between real-world developments that haven’t all been scientifically validated. The model also assumes that complex relationships — say, how drought and political stability relate to each other — remain consistent and linear over time (when in reality we know the relationships will change, but not how). Many people will also be trapped by their circumstances, too poor or vulnerable to move, and the models have a difficult time accounting for them.

All this means that our model is far from definitive. But every one of the scenarios it produces points to a future in which climate change, currently a subtle disrupting influence, becomes a source of major disruption, increasingly driving the displacement of vast populations.

II. How Climate Moves People

Delmira de Jesús Cortez Barrera moved to the outskirts of San Salvador six years ago, after her life in the rural western edge of El Salvador — just 90 miles from Jorge A.’s village in Guatemala — collapsed. Now she sells pupusas on a block not far from where teenagers stand guard for the Mara Salvatrucha gang. When we met last summer, she was working six days a week, earning $7 a day, or less than $200 a month. She relied on the kindness of her boss, who gave her some free meals at work. But everything else for her and her infant son she had to provide herself. Cortez commuted before dawn from San Marcos, where she lived with her sister in a cheap room off a pedestrian alleyway. But her apartment still cost $65 each month. And she sent $75 home to her parents each month — enough for beans and cheese to feed the two daughters she left with them. “We’re going backward,” she said.

Her story — that of an uneducated, unskilled woman from farm roots who can’t find high-paying work in the city and falls deeper into poverty — is a familiar one, the classic pattern of in-country migration all around the world. San Salvador, meanwhile, has become notorious as one of the most dangerous cities in the world, a capital in which gangs have long controlled everything from the majestic colonial streets of its downtown squares to the offices of the politicians who reside in them. It is against this backdrop of war, violence, hurricanes and poverty that one in six of El Salvador’s citizens have fled for the United States over the course of the last few decades, with some 90,000 Salvadorans apprehended at the U.S. border in 2019 alone.

Cortez was born about a mile from the Guatemalan border, in El Paste, a small town nestled on the side of a volcano. Her family were jornaleros — day laborers who farmed on the big maize and bean plantations in the area — and they rented a two-room mud-walled hut with a dirt floor, raising nine children there. Around 2012, a coffee blight worsened by climate change virtually wiped out El Salvador’s crop, slashing harvests by 70 percent. Then drought and unpredictable storms led to what a U.N.-affiliated food-security organization describes as “a progressive deterioration” of Salvadorans’ livelihoods.

That’s when Cortez decided to leave. She married and found work as a brick maker at a factory in the nearby city of Ahuachapán. But the gangs found easy prey in vulnerable farmers and spread into the Salvadoran countryside and the outlying cities, where they made a living by extorting local shopkeepers. Here we can see how climate change can act as what Defense Department officials sometimes refer to as a “threat multiplier.” For Cortez, the threat could not have been more dire. After two years in Ahuachapán, a gang-connected hit man knocked on Cortez’s door and took her husband, whose ex-girlfriend was a gang member, executing him in broad daylight a block away.

In other times, Cortez might have gone back home. But there was no work in El Paste, and no water. So she sent her children there and went to San Salvador instead.

For all the ways in which human migration is hard to predict, one trend is clear: Around the world, as people run short of food and abandon farms, they gravitate toward cities, which quickly grow overcrowded. It’s in these cities, where waves of new people stretch infrastructure, resources and services to their limits, that migration researchers warn that the most severe strains on society will unfold. Food has to be imported — stretching reliance on already-struggling farms and increasing its cost. People will congregate in slums, with little water or electricity, where they are more vulnerable to flooding or other disasters. The slums fuel extremism and chaos.

It is a shift that is already well underway, which is why the World Bank has raised concerns about the mind-boggling influx of people into East African cities like Addis Ababa, in Ethiopia, where the population has doubled since 2000 and is expected to nearly double again by 2035. In Mexico, the World Bank estimates, as many as 1.7 million people may migrate away from the hottest and driest regions, many of them winding up in Mexico City.

But like so much of the rest of the climate story, the urbanization trend is also just the beginning. Right now a little more than half of the planet’s population lives in urban areas, but by the middle of the century, the World Bank estimates, 67 percent will. In just a decade, four out of every 10 urban residents — two billion people around the world — will live in slums. The International Committee of the Red Cross warns that 96 percent of future urban growth will happen in some of the world’s most fragile cities, which already face a heightened risk of conflict and have governments that are least capable of dealing with it. Some cities will be unable to sustain the influx. In the case of Addis Ababa, the World Bank suggests that in the second half of the century, many of the people who fled there will be forced to move again, leaving that city as local agriculture around it dries up.

Percentage of El Salvador’s 6.4 million residents who currently lack a reliable source of food:

Our modeling effort is premised on the notion that in these cities as they exist now, we can see the seeds of their future growth. Relationships between quality-of-life factors like household income in specific neighborhoods, education levels, employment rates and so forth — and how each of those changed in response to climate — would reveal patterns that could be projected into the future. As moisture raises the grain in a slab of wood, the information just needed to be elicited.

Under every scientific forecast for global climate change, El Salvador gets hotter and drier, and our model was in accord with what other researchers said was likely: San Salvador will continue to grow as a result, putting still more people in its dense outer rings. What happens in its farm country, though, is more dependent on which climate and development policies governments to the north choose to deploy in dealing with the warming planet. High emissions, with few global policy changes and relatively open borders, will drive rural El Salvador — just like rural Guatemala — to empty out, even as its cities grow.

Should the United States and other wealthy countries change the trajectory of global policy, though — by, say, investing in climate mitigation efforts at home but also hardening their borders — they would trigger a complex cascade of repercussions farther south, according to the model. Central American and Mexican cities continue to grow, albeit less quickly, but their overall wealth and development slows drastically, most likely concentrating poverty further. Far more people also remain in the countryside for lack of opportunity, becoming trapped and more desperate than ever.

People move to cities because they can seem like a refuge, offering the facade of order — tall buildings and government presence — and the mirage of wealth. I met several men who left their farm fields seeking extremely dangerous work as security guards in San Salvador and Guatemala City. I met a 10-year-old boy washing car windows at a stoplight, convinced that the coins in his jar would help buy back his parents’ farmland. Cities offer choices, and a sense that you can control your destiny.

These same cities, though, can just as easily become traps, as the challenges that go along with rapid urbanization quickly pile up. Since 2000, San Salvador’s population has ballooned by more than a third as it has absorbed migrants from the rural areas, even as tens of thousands of people continue to leave the country and migrate north. By midcentury, the U.N. estimates that El Salvador — which has 6.4 million people and is the most densely populated country in Central America — will be 86 percent urban.

Our models show that much of the growth will be concentrated in the city’s slumlike suburbs, places like San Marcos, where people live in thousands of ramshackle structures, many without electricity or fresh water. In these places, even before the pandemic and its fallout, good jobs were difficult to find, poverty was deepening and crime was increasing. Domestic abuse has also been rising, and declining sanitary conditions threaten more disease. As society weakens, the gangs — whose members outnumber the police in parts of El Salvador by an estimated three to one — extort and recruit. They have made San Salvador’s murder rate one of the highest in the world.

Cortez hoped to escape the violence, but she couldn’t. The gangs run through her apartment block, stealing televisions and collecting protection payments. She had recently witnessed a murder inside a medical clinic where she was delivering food. The lack of security, the lack of affordable housing, the lack of child care, the lack of sustenance — all influence the evolution of complex urban systems under migratory pressure, and our model considers such stresses by incorporating data on crime, governance and health care. They are signposts for what is to come.

A week before our meeting last year, Cortez had resolved to make the trip to the United States at almost any cost. For months she had “felt like going far away,” but moving home was out of the question. “The climate has changed, and it has provoked us,” she said, adding that it had scarcely rained in three years. “My dad, last year, he just gave up.”

Cortez recounted what she did next. As her boss dropped potato pupusas into the smoking fryer, Cortez turned to her and made an unimaginable request: Would she take Cortez’s baby? It was the only way to save the child, Cortez said. She promised to send money from the United States, but the older woman said no — she couldn’t imagine being able to care for the infant.

Today San Salvador is shut down by the coronavirus pandemic, and Cortez is cooped up inside her apartment in San Marcos. She hasn’t worked in three months and is unable to see her daughters in El Paste. She was allowed a forbearance on rent during the country’s official lockdown, but that has come to an end. She remains convinced that the United States is her only salvation — border walls be damned. She’ll leave, she said, “the first chance I get.”

Most would-be migrants don’t want to move away from home. Instead, they’ll make incremental adjustments to minimize change, first moving to a larger town or a city. It’s only when those places fail them that they tend to cross borders, taking on ever riskier journeys, in what researchers call “stepwise migration.” Leaving a village for the city is hard enough, but crossing into a foreign land — vulnerable to both its politics and its own social turmoil — is an entirely different trial.

Seven miles from the Suchiate River, which marks Guatemala’s border with Mexico, sits Siglo XXI, one of Mexico’s largest immigration detention centers, a squat concrete compound with 30-foot walls, barred windows and a punishment cell. In early 2019, the 960-bed facility was largely empty, as Mexico welcomed passing migrants instead of detaining them. But by March, as the United States increased pressure to stop Central Americans from reaching its borders, Mexico had begun to detain migrants who crossed into its territory, packing almost 2,000 people inside this center near the city of Tapachula. Detainees slept on mattresses thrown down in the white-tiled hallways, waited in lines to use toilets overflowing with feces and crammed shoulder to shoulder for hours to get a meal of canned meat spooned onto a metal tray.

Projected decrease in percentage of annual rainfall by 2070 in many parts of Guatemala:

On April 25, imprisoned migrants stormed the stairway leading to a fortified security platform in the center’s main hall, overpowering the guards and then unlocking the main gates. More than 1,000 Guatemalans, Cubans, Salvadorans, Haitians and others streamed into the Tapachula night.

I arrived in Tapachula five weeks after the breakout to find a city cracking in the crucible of migration. Just months earlier, passing migrants on Mexico’s southern border were offered rides and tortas and medicine from a sympathetic Mexican public. Now migrant families were being hunted down in the countryside by armed national-guard units, as if they were enemy soldiers.

Mexico has not always welcomed migrants, but President Andrés Manuel López Obrador was trying to make his country a model for increasingly open borders. This idealistic effort was also pragmatic: It was meant to show the world an alternative to the belligerent wall-building xenophobia he saw gathering momentum in the United States. More open borders, combined with strategic foreign aid and help with human rights to keep Central American migrants from leaving their homes in the first place, would lead to a better outcome for all nations. “I want to tell them they can count on us,” López Obrador had declared, promising the migrants work permits and temporary jobs.

The architects of Mexico’s policies assumed that its citizens had the patience and the capacity to absorb — economically, environmentally and socially — such an influx of people. But they failed to anticipate how President Trump would hold their economy hostage to press his own anti-immigrant crackdown, and they were caught off-guard by how the burdens brought by the immigration traffic weighed on Mexico’s own people.

In the six months after López Obrador took office in December 2018, some 420,000 people entered Mexico without documentation, according to Mexico’s National Migration Institute. Many floated across the Suchiate on boards tied atop large inner tubes, paying guides a couple of dollars for passage. In Ciudad Hidalgo, a border town outside Tapachula, migrants camped in the square and fought in the streets. In a late-night interview in his cinder-block office, under the glare of fluorescent lights, the town’s director of public security, Luis Martínez López, rattled off statistics about their impact: Armed robberies jumped 45 percent; murders increased 15 percent.

Whether the crimes were truly attributable to the migrants was a matter of significant debate, but the perception that they were fueled a rising impatience. That March, Martínez told me, a confrontation between a crowd of about 400 migrants and the local police turned rowdy, and the migrants tied up five officers in the center of town. No one was hurt, but the incident stoked locals’ concern that things were getting out of control. “We used to open doors for them like brothers and feed them,” said Martínez, who has since left his government job. “I was disappointed and angry.”

In Tapachula, a much larger city, tourism and commerce began to suffer. Whole families of migrants huddled in downtown doorways overnight, crowding sidewalks and sleeping on thin, oil-stained sheets of cardboard. Hotels — normally almost sold out in December — were less than 65 percent full as visitors stayed away, fearful of crime. Clinics ran short of medication. The impact came at a vulnerable moment: While many northern Mexican states enjoyed economic growth of 3 to 11 percent in 2018, Chiapas — its southernmost state — had a 3 percent drop in its gross domestic product. “They are overwhelmed,” said the Rev. César Cañaveral Pérez, who earned a Ph.D. in the theology of human mobility in Rome and now runs Tapachula’s largest Catholic migrant shelter.

Models can’t say much about the cultural strain that might result from a climate influx; there is no data on anger or prejudice. What they do say is that over the next two decades, if climate emissions continue as they are, the population in southern Mexico will grow sharply.

At the same time, Mexico has its own serious climate concerns and will most likely see its own climate exodus. One in six Mexicans now rely on farming for their livelihood, and close to half the population lives in poverty. Studies estimate that with climate change, water availability per capita could decrease by as much as 88 percent in places, and crop yields in coastal regions may drop by a third. If that change does indeed push out a wave of Mexican migrants, many of them will most likely come from Chiapas.

Yet a net increase in population at the same time — which is what our models assume — suggests that even as one million or so climate migrants make it to the U.S. border, many more Central Americans will become trapped in protracted transit, unable to move forward or backward in their journey, remaining in southern Mexico and making its current stresses far worse.

Percentage of future urban growth that, according to the International Committee of the Red Cross, is likely to take place in some of the world’s most fragile cities, where risk of social unrest is high:

Already, by late last year, the Mexican government’s ill-planned policies had begun to unravel into something more insidious: rising resentment and hate. Now that the coronavirus pandemic has effectively sealed borders, those sentiments risk bubbling over. Migrants, with nowhere to go and no shelters able to take them in, roam the streets, unable to socially distance and lacking even basic sanitation.

It has angered many Mexican citizens, who have begun to describe the migrants as economic parasites and question foreign aid aimed at helping people cope with the drought in places where Jorge A. and Cortez come from.

“How dare AMLO give $30 million to El Salvador when we have no services here?” asked Javier Ovilla Estrada, a community-group leader in the southern border town Ciudad Hidalgo, referring to López Obrador’s participation in a multibillion-dollar development plan with Guatemala, Honduras and El Salvador. Ovilla has become a strident defender of a new Mexico-first movement, organizing thousands to march against immigrants. Months before the coronavirus spread, we met in the sterile dining room of a Chinese restaurant that he frequents in Ciudad Hidalgo, and he echoed the same anti-immigrant sentiments rising in the U.S. and Europe.

The migrants “don’t love this country,” he said. He points to anti-immigrant Facebook groups spreading rumors that migrants stole ballots and rigged the Mexican presidential election, that they murder with impunity and run brothels. He’s not the first to tell me that the migrants traffic in disease — that Suchiate will soon be overwhelmed by Ebola. “They should close the borders once and for all,” he said. If they don’t, he warns, the country will sink further into lawlessness and conflict. “We’re going to go out into the streets to defend our homes and our families.”

One afternoon last summer, I sat on a black pleather couch in a borrowed airport-security office at the Tapachula airfield to talk with Francisco Garduño Yáñez, Mexico’s new commissioner for immigration. Garduño had abruptly succeeded a man named Tonatiuh Guillén López, a strong proponent of more open borders, whom I’d been trying to reach for weeks to ask how Mexico had strayed so far from the mission he laid out for it.

But in between, Trump had, as another senior government official told me, “held a gun to Mexico’s head,” demanding a crackdown at the Guatemalan border under threat of a 25 percent tariff on trade. Such a tax could break the back of Mexico’s economy overnight, and so López Obrador’s government immediately agreed to dispatch a new militarized force to the border. Guillén resigned as a result, four days before I hoped to meet him.

Number of people projected to be displaced from their homes by rising sea levels alone by 2050:

Garduño, a cheerful man with short graying hair, a broad smile and a ceaseless handshake, had been on the job for less than 36 hours. He had flown to Tapachula because another riot had broken out in one of the city’s smaller fortified detention centers, and a starving Haitian refugee was filmed by news crews there, begging for help for her and her young son. I wanted to know how it had come to this — from signing an international humanitarian migrant bill of rights to a mother lying with her face pressed to the ground in a detention center begging for food, in the space of a few months. He demurred, laying blame at the feet of neoliberal economics, which he said had produced a “poverty factory” with no regional development policies to address it. It was the system — capitalism itself — that had abandoned human beings, not Mexico’s leaders. “We didn’t anticipate that the globalization of the economy, the globalization of the law … would have such a devastating effect,” Garduño told me.

It seemed telling that Garduño’s previous role had been as Mexico’s commissioner of federal prisons. Was this the start of a new, punitive Mexico? I asked him. Absolutely not, he replied. But Mexico was now pursuing a policy of “containment,” he said, rejecting the notion that his country was obligated to “receive a global migration.”

No policy, though, would be able to stop the forces — climate, increasingly, among them — that are pushing migrants from the south to breach Mexico’s borders, legally or illegally. So what happens when still more people — many millions more — float across the Suchiate River and land in Chiapas? Our model suggests that this is what is coming — that between now and 2050, nearly nine million migrants will head for Mexico’s southern border, more than 300,000 of them because of climate change alone.

Before leaving Mexico last summer, I went to Huixtla, a small town 25 miles west of Tapachula that, because it sat on the Bestia freight rail line used by migrants, had long been a waypoint on Mexico’s superhighway for Central Americans on their way north. Joining several local police officers as they headed out on patrol, I watched as our pickup truck’s red and blue lights reflected in the barred windows of squat cinder-block homes. Two officers stood in back, holding tight to the truck’s roll bars, black combat boots firmly planted in the cargo bed, as the driver, dodging mangy dogs, navigated the town’s slender alleyways.

The operations commander, a soft-spoken bureaucratic type named José Gozalo Rodríguez Méndez, sat in the front seat. I asked him if he thought Mexico could sustain the number of migrants who might soon come. He said Mexico would buckle. There is no money from the federal government, no staffing to address services, no housing, let alone shelter, no more good will. “We couldn’t do it.”

Rodríguez had already been tested. When the first caravan of thousands of migrants reached Huixtla in late 2018, throngs of tired, destitute people — many of them carrying children in their emaciated arms — packed the central square and spilled down the city’s side streets. Rodríguez and his wife went through their cupboards, gathering corn, fried beans and tortillas, and collected clothing outgrown by their children and hauled all of it to the town center, where church and civic groups had set up tents and bathrooms.

But as the caravans continued, he said, his good will began to disintegrate. “It’s like inviting somebody to your place for dinner,” he said. “You’ll invite them once, even twice. But will you invite them six times?” When the fourth caravan of migrants approached the city last March, Rodríguez told me, he stayed home.

In the center of town, the truck lurched to a stop amid a busy market, where stalls sell vegetables and toys under blue light filtered through plastic tarps overhead. A short way away, five men sheltered from the searing heat under the shade of a metal awning on the platform of a crumbling railway station, never repaired after Hurricane Stan 14 years earlier. Rodríguez peppered the group — two from Honduras, three from Guatemala — with questions. Together they said they had suffered the totality of misfortune that Central America offers: muggings, gang extortion and environmental disaster. Either they couldn’t grow food or the drought made it too expensive to buy.

“We can’t stand the hunger,” said one Honduran farmer, Jorge Reyes, his gaunt face dripping with sweat. At his feet was a gift from a shopkeeper: a plastic bag filled with a cut of raw meat, pooled in its own blood, flies circling around it in the heat. Reyes had nowhere to cook it. “If we are going to die anyway,” he said, “we might as well die trying to get to the United States.”

III. The Choice

Reyes had made his decision. Like Jorge A., Cortez and millions of others, he was going to the U.S. The next choice — how to respond and prepare for the migrants — ultimately falls to America’s elected leaders.

Over the course of 2019, El Paso, Texas, had endured a crush of people at its border crossings, peaking at more than 4,000 migrants in a single day, as the same caravans of Central Americans that had worn out their welcome in Tapachula made their way here. It put El Paso in a delicate spot, caught between the forces of politically charged anti-immigrant federal policy and its own deep roots as a diverse, largely Hispanic city whose identity was virtually inextricable from its close ties to Mexico. This surge, though, stretched the city’s capacity. When the migrants arrived, city officials argued over who should pay the tab for the emergency services, aid and housing, and in the end crossed their fingers and hoped the city’s active private charities would figure it out. Church groups rented thousands of hotel rooms across the city, delivered food, offered counseling and so on.

Conjoined to the Mexican city of Juárez, the El Paso area is the second-largest binational metroplex in the Western Hemisphere. It sits smack in the middle of the Chihuahuan Desert, a built-up oasis amid a barren and bleached-bright rocky landscape. Much of its daily work force commutes across the border, and Spanish is as common as English.

Downtown, new buildings are rising in a weary business district where boot shops and pawnshops compete amid boarded-up and barred storefronts. The only barriers between the American streets — home to more than 800,000 people — and their Juárez counterparts are the concrete viaduct of a mostly dry Rio Grande and a rusted steel border fence.

To some migrants, this place is Eden. But El Paso is also a place with oppressive heat and very little water, another front line in the climate crisis. Temperatures already top 90 degrees here for three months of the year, and by the end of the century it will be that hot one of every two days. The heat, according to researchers at the University of California, Berkeley, will drive deaths that soon outpace those from car crashes or opioid overdoses. Cooling costs — already a third of some residents’ budgets — will get pricier, and warming will drive down economic output by 8 percent, perhaps making El Paso just as unlivable as the places farther south.

In 2014, El Paso created a new city government position — chief resilience officer — aimed, in part, at folding climate concerns into its urban planning. Soon enough, the climate crisis in Guatemala — not just the one in El Paso — became one of the city’s top concerns. “I apologize if I’m off topic,” the resilience chief, Nicole Ferrini, told municipal leaders and other attendees at a water conference in Phoenix in 2019 as she raised the question of “massive amounts of climate refugees, and are we prepared as a community, as a society, to deal with that?”

Ferrini, an El Paso native, did her academic training as an architect. She worries that El Paso will struggle to adapt if its leadership, and the nation’s, continue to react to daily or yearly spikes rather than view the problem as a systematic one, destined to become steadily worse as the planet warms. She sees her own city as an object lesson in what U.N. officials and climate-migration scientists have been warning of: Without a decent plan for housing, feeding and employing a growing number of climate refugees, cities on the receiving end of migration can never confidently pilot their own economic future.

For the moment, the coronavirus pandemic has largely choked off legal crossings into El Paso, but that crisis will eventually fade. And when it does, El Paso will face the same enduring choice that all wealthier societies everywhere will eventually face: determining whether it is a society of walls or — in the vernacular of aid organizations working to fortify infrastructure and resilience to stem migration — one that builds wells.

Around the world, nations are choosing walls. Even before the pandemic, Hungary fenced off its boundary with Serbia, part of more than 1,000 kilometers of border walls erected around the European Union states since 1990. India has built a fence along most of its 2,500-mile border with Bangladesh, whose people are among the most vulnerable in the world to sea-level rise.

The United States, of course, has its own wall-building agenda — literal ones, and the figurative ones that can have a greater effect. On a walk last August from one of El Paso’s migrant shelters, an inconspicuous brick home called Casa Vides, the Rev. Peter Hinde told me that El Paso’s security-oriented economy had created a cultural barrier that didn’t exist when he moved here 25 years earlier. Hinde, who is 97, helps run the Carmelite order in Juárez but was traveling to volunteer at Casa Vides on a near-daily basis. A former Army Air Forces captain and fighter pilot who grew up in Chicago, Hinde said the United States is turning its own fears into reality when it comes to immigration, something he witnesses in a growing distrust of everyone who crosses the border.

That fear creates other walls. The United States refused to join 164 other countries in signing a global migration treaty in 2018, the first such agreement to recognize climate as a cause of future displacement. At the same time, the U.S. is cutting off foreign aid — money for everything from water infrastructure to greenhouse agriculture — that has been proved to help starving families like Jorge A.’s in Guatemala produce food, and ultimately stay in their homes. Even those migrants who legally make their way into El Paso have been turned back, relegated to cramped and dangerous shelters in Juárez to wait for the hearings they are owed under law.

There is no more natural and fundamental adaptation to a changing climate than to migrate. It is the obvious progression the earliest Homo sapiens pursued out of Africa, and the same one the Mayans tried 1,200 years ago. As Lorenzo Guadagno at the U.N.’s International Organization for Migration told me recently, “Mobility is resilience.” Every policy choice that allows people the flexibility to decide for themselves where they live helps make them safer.

Are you a teacher looking for a way to use this project in your classroom? You can find resources from the Pulitzer Center here .

But it isn’t always so simple, and relocating across borders doesn’t have to be inevitable. I thought about Jorge A. from Guatemala. He made it to the United States last spring, climbing the steel border barrier and dropping his 7-year-old son 20 feet down the other side into the California desert. (We are abbreviating his last name in this article because of his undocumented status.) Now they live in Houston, where until the pandemic, Jorge found steady work in construction, earning enough to pay his debts and send some money home. But the separation from his wife and family has proved intolerable; home or away, he can’t win, and as of early July, he was wondering if he should go back to Guatemala.

And therein lies the basis for what may be the worst-case scenario: one in which America and the rest of the developed world refuse to welcome migrants but also fail to help them at home. As our model demonstrated, closing borders while stinting on development creates a somewhat counterintuitive population surge even as temperatures rise, trapping more and more people in places that are increasingly unsuited to human life.

In that scenario, the global trend toward building walls could have a profound and lethal effect. Researchers suggest that the annual death toll, globally, from heat alone will eventually rise by 1.5 million. But in this scenario, untold more will also die from starvation, or in the conflicts that arise over tensions that food and water insecurity will bring.

If this happens, the United States and Europe risk walling themselves in, as much as walling others out. And so the question then is: What are policymakers and planners prepared to do about that? America’s demographic decline suggests that more immigrants would play a productive role here, but the nation would have to be willing to invest in preparing for that influx of people so that the population growth alone doesn’t overwhelm the places they move to, deepening divisions and exacerbating inequalities. At the same time, the United States and other wealthy countries can help vulnerable people where they live, by funding development that modernizes agriculture and water infrastructure. A U.N. World Food Program effort to help farmers build irrigated greenhouses in El Salvador, for instance, has drastically reduced crop losses and improved farmers’ incomes. It can’t reverse climate change, but it can buy time.

Thus far, the United States has done very little at all. Even as the scientific consensus around climate change and climate migration builds, in some circles the topic has become taboo. This spring, after Proceedings of the National Academy of Sciences published the explosive study estimating that, barring migration, one-third of the planet’s population may eventually live outside the traditional ecological niche for civilization, Marten Scheffer, one of the study’s authors, told me that he was asked to tone down some of his conclusions through the peer-review process and that he felt pushed to “understate” the implications in order to get the research published. The result: Migration is only superficially explored in the paper. (A spokeswoman for the journal declined to comment because the review process is confidential.)

“There’s flat-out resistance,” Scheffer told me, acknowledging what he now sees as inevitable, that migration is going to be a part of the global climate crisis. “We have to face it.”

Our modeling and the consensus of academics point to the same bottom line: If societies respond aggressively to climate change and migration and increase their resilience to it, food production will be shored up, poverty reduced and international migration slowed — factors that could help the world remain more stable and more peaceful. If leaders take fewer actions against climate change, or more punitive ones against migrants, food insecurity will deepen, as will poverty. Populations will surge, and cross-border movement will be restricted, leading to greater suffering. Whatever actions governments take next — and when they do it — makes a difference.

The window for action is closing. The world can now expect that with every degree of temperature increase, roughly a billion people will be pushed outside the zone in which humans have lived for thousands of years. For a long time, the climate alarm has been sounded in terms of its economic toll, but now it can increasingly be counted in people harmed. The worst danger, Hinde warned on our walk, is believing that something so frail and ephemeral as a wall can ever be an effective shield against the tide of history. “If we don’t develop a different attitude,” he said, “we’re going to be like people in the lifeboat, beating on those that are trying to climb in.”

Abrahm Lustgarten is a senior environmental reporter at ProPublica. His 2015 series examining the causes of water scarcity in the American West, “Killing the Colorado,” was a finalist for the 2016 Pulitzer Prize for national reporting. Meridith Kohut is an award-winning photojournalist based in Caracas, Venezuela, who has documented global health and humanitarian crises in Latin America for The New York Times for more than a decade. Her recent assignments include photographing migration and childbirth in Venezuela, antigovernment protests in Haiti and the killing of women in Guatemala.

Reporting and translation were contributed by Pedro Pablo Solares in Guatemala and El Salvador, and Louisa Reynolds and Juan de Dios García Davish in Mexico.

Data for opening globe graphic from “Future of the Human Climate Niche,” by Chi Xu, Timothy A. Kohler, Timothy M. Lenton, Jens-Christian Svenning and Marten Scheffer, from Proceedings of the National Academy of Sciences. Graphic by Bryan Christie Design/Joe Lertola.

Maps in Central America graphics sequence show total population shift under the SSP5 / RCP 8.5 and SSP3 / RCP 8.5 scenarios used by the U.N.’s Intergovernmental Panel on Climate Change, and it is calculated on a 15-kilometer grid. A cube-root scale was used to compress the largest peaks.

Projections based on research by The New York Times Magazine and ProPublica, with support from the Pulitzer Center. Model graphics and additional data analysis by Matthew Conlen.

Additional design and development by Jacky Myint and Shannon Lin.

Learn More About Climate Change

Have questions about climate change? Our F.A.Q. will tackle your climate questions, big and small .

“Buying Time,” a new series from The New York Times, looks at the risky ways  humans are starting to manipulate nature  to fight climate change.

Big brands like Procter & Gamble and Nestlé say a new generation of recycling plants will help them meet environmental goals, but the technology is struggling to deliver .

The Italian energy giant Eni sees future profits from collecting carbon dioxide and pumping it  into natural gas fields that have been exhausted.

New satellite-based research reveals how land along the East Coast is slumping into the ocean, compounding the danger from global sea level rise . A major culprit: the overpumping of groundwater.

Did you know the ♻ symbol doesn’t mean something is actually recyclable ? Read on about how we got here, and what can be done.

Advertisement

Climate Change and Migration

A story that begins with case studies of small communities facing severe threats from climate change effects and continues with a look at large populations made vulnerable by human-induced changes to Earth's climate, leveraging data from a variety of sources such as academic journals and scientific agencies.

*References to Kosovo shall be understood to be in the context of United Nations Security Council resolution 1244 (1999).

Subscribe to our newsletter.

• Coordinator Briefings
• Executive Committee
• Network Members
• Secretariat
• Arab Region
• Asia and the Pacific
• Europe and North America
• Latin America and the Caribbean
• Workplan 2022 - 2024
• Workstreams
• Policy Documents
• Capacity Building
• Calendar 2024
• Migration Week 2024
• Upcoming Events
• Annual Meetings
• Principals Meetings
• Migration Week
• Newsletters
• The GCM in a Nutshell
• Champion countries
• Reporting form
• Submission form
• The Pledging Dashboard
• GCM and GCR Complementarity
• SG Report 2022
• SG report 2020
• Arab States
• UNECE Region
• IMRF side-events
• #Migration2022
• Exhibition 2022
• Repository of Practices
• Discussion space
• Demand-driven facility
• Experts Database
• CLIMB Database
• Migration MPTF
• Stakeholders

The climate crisis, migration, and refugees

• Download the policy brief

Subscribe to Global Connection

John podesta john podesta founder and director - the center for american progress.

July 25, 2019

• 18 min read

The following is one of eight briefs commissioned for the 16th annual Brookings Blum Roundtable, “2020 and beyond: Maintaining the bipartisan narrative on US global development.”

On March 14, 2019, Tropical Cyclone Idai struck the southeast coast of Mozambique. The U.N. High Commissioner for Refugees reported that 1.85 million people needed assistance. 146,000 people were internally displaced, and Mozambique scrambled to house them in 155 temporary sites. 1 The cyclone and subsequent flooding damaged 100,000 homes, destroyed 1 million acres of crops, and demolished $1 billion worth of infrastructure. 2

One historic storm in one place over the course of one day. While Cyclone Idai was the worst storm in Mozambique’s history, the world is looking towards a future where these “unprecedented” storms are commonplace. This global challenge has and will continue to create a multitude of critical issues that the international community must confront, including:

• Large-scale human migration due to resource scarcity, increased frequency of extreme weather events, and other factors, particularly in the developing countries in the earth’s low latitudinal band
• Intensifying intra- and inter-state competition for food, water, and other resources, particularly in the Middle East and North Africa
• Increased frequency and severity of disease outbreaks
• Increased U.S. border stress due to the severe effects of climate change in parts of Central America

All of these challenges are serious, but the scope and scale of human migration due to climate change will test the limits of national and global governance as well as international cooperation.

The migration-climate nexus is real, but more scrutiny and action are required

In 2018, the World Bank estimated that three regions (Latin America, sub-Saharan Africa, and Southeast Asia) will generate 143 million more climate migrants by 2050. 3 In 2017, 68.5 million people were forcibly displaced, more than at any point in human history. While it is difficult to estimate, approximately one-third of these (22.5 million 4 to 24 million 5 people) were forced to move by “sudden onset” weather events—flooding, forest fires after droughts, and intensified storms. While the remaining two-thirds of displacements are the results of other humanitarian crises, it is becoming obvious that climate change is contributing to so-called slow onset events such as desertification, sea-level rise, ocean acidification, air pollution, rain pattern shifts and loss of biodiversity. 6 This deterioration will exacerbate many humanitarian crises and may lead to more people being on the move.

Multilateral institutions, development agencies, and international law must do far more to thoroughly examine the challenges of climate change (early efforts, like the World Bank’s 2010 World Development Report on climate change, 7 had little uptake at a time when few thought a climate crisis was around the corner). Moreover, neither a multilateral strategy nor a legal framework exist to account for climate change as a driver of migration. Whether in terms of limited access to clean water, food scarcity, agricultural degradation, or violent conflict, 8 climate change will intensify these challenges and be a significant push factor in human migration patterns.

To date, there are only a few cases where climate change is the sole factor prompting migration. The clearest examples are in the Pacific Islands. The sea level is rising at a rate of 12 millimeters per year in the western Pacific and has already submerged eight islands. Two more are on the brink of disappearing, prompting a wave of migration to larger countries. 9 10 By 2100, it is estimated that 48 islands overall will be lost to the rising ocean. 11 In 2015, the Teitota family applied for refugee status in New Zealand, fleeing the disappearing island nation of Kiribati. 12 Their case, the first request for refuge explicitly attributed climate change, made it to the High Court of New Zealand but was ultimately dismissed. Islands in the Federated States of Micronesia have drastically reduced in size, washed down to an uninhabitable state, had their fresh water contaminated by the inflow of seawater, and disappeared in the past decade. 13 Despite their extreme vulnerability, the relatively small population (2.3 million people spread across 11 countries 14 ) and remote location of the Pacific Islands means that they garner little international action, for all the attention they receive in the media.

Although there are few instances of climate change as the sole factor in migration, climate change is widely recognized as a contributing and exacerbating factor in migration and in conflict.

In South Asia, increasing temperatures, sea level rise, more frequent cyclones, flooding of river systems fed by melting glaciers, and other extreme weather events are exacerbating current internal and international migration patterns. Additionally, rapid economic growth and urbanization are accelerating and magnifying the impact and drivers of climate change—the demand for energy is expected to grow 66 percent by 2040. 15 Compounding this, many of the expanding urban areas are located in low-lying coastal areas, already threatened by sea level rise. 16 The confluence of these factors leads the World Bank to predict that the collective South Asian economy (Bangladesh, Bhutan, India, the Maldives, Nepal, and Sri Lanka) will lose 1.8 percent of its annual GDP due to climate change by 2050. 17 The New York Times reports that the living conditions of 800 million people could seriously diminish. 18 Diminishing living conditions on this scale and intensity will prompt mass migration—possibly at an unprecedented level.

Northwest Africa is facing rising sea levels, drought, and desertification. These conditions will only add to the already substantial number of seasonal migrants and put added strain on the country of origin, as well as on destination countries and the routes migrants travel. The destabilizing effects of climate change should be of great concern to all those who seek security and stability in the region. Climate and security experts often cite the impacts of the extreme drought in Syria that preceded the 2011 civil war. 19 The security community also highlights the connection between climate change and terrorism—for instance, the decline of agricultural and pastoral livelihoods has been linked to the effectiveness of financial recruiting strategies by al-Qaida. 20

The intersection of climate change and migration requires new, nimble, and comprehensive solutions to the multidimensional challenges it creates. Accordingly, the signatories to the 2015 Paris Agreement on climate change requested that the Warsaw International Mechanism for Loss and Damage Associated with Climate Change (WIM) develop recommendations for addressing people displaced by climate change. 21 Similarly, The Global Compact for Safe, Orderly, and Regular Migration (adopted by 164 countries—not including the U.S.—in Marrakech in December 2018) called on countries to make plans to prevent the need for climate-caused relocation and support those forced to relocate. 22 However, these agreements are neither legally binding nor sufficiently developed to support climate migrants—particularly migrants from South Asia, Central America, Northwest Africa, and the Horn of Africa.

Time to envision legal recourse for climate refugees

As gradually worsening climate patterns and, even more so, severe weather events, prompt an increase in human mobility, people who choose to move will do so with little legal protection. The current system of international law is not equipped to protect climate migrants, as there are no legally binding agreements obliging countries to support climate migrants.

While climate migrants who flee unbearable conditions resemble refugees, the legal protections afforded to refugees do not extend to them. In the aftermath of World War II, the United Nations established a system to protect civilians who had been forced from their home countries by political violence. Today, there are almost 20.4 million officially designated refugees under the protection of the United Nations High Commission for Refugees (UNHCR)—however, there is an additional group of 21.5 million people 23 who flee their homes as a result of sudden onset weather hazards every year. 24

The UNHCR has thus far refused to grant these people refugee status, instead designating them as “environmental migrants,” in large part because it lacks the resources to address their needs. But with no organized effort to supervise the migrant population, these desperate individuals go where they can, not necessarily where they should. As their numbers grow, it will become increasingly difficult for the international community to ignore this challenge. As severe climate change displaces more people, the international community may be forced to either redefine “refugees” to include climate migrants or create a new legal category and accompanying institutional framework to protect climate migrants. However, opening that debate in the current political context would be fraught with difficulty. Currently, the nationalist, anti-immigrant, and xenophobic atmosphere in Europe and the U.S. would most likely lead to limiting refugee protections rather than expanding them.

The SDGs can help, but not without an update to the US response

While there are no legally binding international regimes that protect climate migrants, there are voluntary compacts that could be used to support them. Most notably, 193 countries adopted the 2030 Sustainable Development Goals (SDGs), which address both migration and climate change.

Several of the 169 targets established by the SDGs lay out general goals that could be used to protect climate migrants. SDG 13 on climate action outlines several targets that address the climate crisis:

• 13.1: Strengthen resilience and adaptive capacity to climate-related hazards and natural disasters in all countries
• 13.2: Integrate climate change measures into national policies, strategies, and planning
• 13.3: Improve education, awareness-raising and human and institutional capacity on climate change mitigation, adaptation, impact reduction, and early warning.

To meet these goals, extensive bilateral and multilateral development assistance will be needed. The U.S. must create a strategic approach to focus development assistance and multilateral organizations on those targets—particularly to create resilient societies that can keep people in their communities.

Although the SDGs do not explicitly link climate change and migration, SDG target 10.7 calls for signatories to “facilitate orderly, safe, and responsible migration of people, including through implementation of planned and well-managed policies.” Again, the United States should channel multilateral development assistance to support the implementation of this target.

The scale and scope of climate change demand dynamic and comprehensive solutions. The U.S. must address climate stress on vulnerable populations specifically, rather than funneling more money into existing programs that operate on the periphery of the growing crisis.

U.S. development agencies and international development financial institutions need to redirect their development assistance to incorporate today’s unfolding climate crisis. Significantly more resources will need to be channeled to the new U.S. International Development Finance Corporation (USDFC), USAID, the Green Climate Fund, UNHCR, as well as to other critical international bodies, in particular those that make up the International Red Cross and Red Crescent organizations.

The Obama administration undertook myriad efforts to update the institutions that can address climate. Several of President Obama’s executive orders, particularly Executive Order 13677, which required incorporating climate resilience into decisionmaking on development assistance, took on the climate crisis. For the first time in the Department of Defense’s history, the 2010 Quadrennial Defense Review (QDR) recognized climate change as a “threat multiplier,” with the potential to exacerbate current challenges. 25

While the current administration has deemphasized or opposed climate-friendly approaches, the current security implications of the migration crisis might prompt a re-examination of those policies. There should be bipartisan support, particularly in the security community, for reducing the conditions that accelerate international migration.

The case for scaling up US action to confront the climate crisis

A variety of medium-term investments (five to 10 years) could create more resilience to the effects of climate change. For example, the climate change factors that push migration in Northwest Africa could—at least in part—be addressed by supporting irrigation infrastructure, providing food supplies, fostering regional water cooperation, and supporting livelihood security. 26

Dedicating greater resources to mitigate climate migration is also part of an effective solution. Research is needed to determine the best way to improve the migratory process itself—be it increasing migration monitors, providing safer modes of transport, and consolidating and expanding destination country integration resources.

This discussion is not new: In 2010, Center for American Progress staff were part of a task force that suggested a “Unified Security Budget” for the United States, to address complex crisis scenarios that transcend the traditional division of labor among defense, diplomacy, and development. 27 The need for longer-term, more calculated assessment strategies and investments has only increased over the past decade. The Pentagon already supports a variety of operational missions that respond to sudden onset climate disasters. The Navy, in particular, serves at the emergency hotline for international extreme weather events and mobilized to support the Haitian people after the 2010 earthquake, the Filipino people after the 2013 typhoon, and the Nepalis after the 2015 earthquake.

Alternatively, creating a single dedicated fund (by drawing funds from Operations and Maintenance, Research and Development, and the Refugee Assistance Fund) would allow the United States to streamline and refine its support strategies, address the effects of climate change directly, and rebuild its reputation abroad. Such a dedicated fund should try to emulate and partner with the United Kingdom’s Department for International Development (DFID), Germany’s Society for International Cooperation (GIZ), and Japan’s International Cooperation Agency (JICA). American seed funding in this area could lead to major investments of allies and partners—and in cooperation with the development agencies of these countries can mobilize massive resources at the scale required to confront the global climate crisis.

The strategies to address climate migrants presented here are far reaching, but this crisis will only intensify, and our response to it will define international relations in the 21st century.

Related Content

Walter Kälin

July 16, 2008

October 10, 2008

Reva Dhingra, Elizabeth Ferris

November 1, 2022

Related Books

Robert Kagan

April 30, 2024

Daniel S. Hamilton, Joe Renouard

April 1, 2024

Michael E. O’Hanlon

February 15, 2024

• United Nations. “UNHCR Factsheet: Cyclone Idai.” May 2019.
• Reid, Kathryn “2019 Cyclone Idai.” World Vision. April 26. 2019. https://www.worldvision.org/disaster-relief-news-stories/2019-cyclone-idai-facts
• Kumari Rigaud, Kanta, Alex de Sherbinin, Bryan Jones, Jonas Bergmann, Viviane Clement, Kayly Ober, Jacob Schewe, Susana Adamo, Brent McCusker, Silke Heuser, and Amelia Midgley. 2018. Groundswell: Preparing for Internal Climate Migration. The World Bank. Pg 2. https://openknowledge.worldbank.org/handle/10986/29461
• McDonnell, Tim. “The Refugees the World Barely Pays Attention To.” June 20, 2018. https://www.npr.org/sections/goatsandsoda/2018/06/20/621782275/the-refugees-that-the-world-barely-pays-attention-to.
• The Nansen Initiative. “Disaster-Induced Cross-Border Displacement.” December 2015. Page 6. https://nanseninitiative.org/wp-content/uploads/2015/02/PROTECTION-AGENDA-VOLUME-1.pdf
• “Slow Onset Events.” United Nations Framework Convention on Climate Change. https://unfccc.int/process/bodies/constituted-bodies/executive-committee-of-the-warsaw-international-mechanism-for-loss-and-damage-wim-excom/areas-of-work/slow-onset-events
• “World Development Report 2010: Development and Climate Change.” World Bank. 2010. https://openknowledge.worldbank.org/handle/10986/4387
• “How Climate Change Can Fuel Wars.” The Economist. May 23, 2019. https://www.economist.com/international/2019/05/25/how-climate-change-can-fuel-wars
• Nunn, P.D., Kohler, A. & Kumar, R. “Identifying and Assessing Evidence for Recent Shoreline Change Attributable To Uncommonly Rapid Sea-Level Rise in Pohnpei, Federated State of Micronesia, Northwest Pacific Ocean.” Journal of Coast Conservation (2017) 21: 719. https://doi.org/10.1007/s11852-017-0531-7
• Roy, Eleanor Ainge, and Sean Gallagher. “One day we’ll Disappear: Tuvalu’s Sinking Islands” The Guardian. May 16, 2019. https://www.theguardian.com/global-development/2019/may/16/one-day-disappear-tuvalu-sinking-islands-rising-seas-climate-change
• Deshmukh, Amrita. “Disappearing Island Nations Are The Sinking Reality of Climate Change.” Qrius. May 17, 2019. https://qrius.com/disappearing-island-nations-are-the-sinking-reality-of-climate-change/
• “New Zealand: Climate Change Refugee Case Overview.” Law Library of Congress. July 29, 2015. https://www.loc.gov/law/help/climate-change-refugee/new-zealand.php
• Nunn, Patrick D., Augustine Kohler, and Roselyn Kumar. “Identifying and Assessing Evidence For Recent Shoreline Change Attributable To Uncommonly Rapid Sea-Level Rise in Pohnpei, Federated States of Micronesia, Northwest Pacific Ocean.” SpringerLink, July 13, 2017. https://link.springer.com/article/10.1007%2Fs11852-017-0531-7
• “The World Bank in Pacific Islands.” World Bank. April 8, 2019. https://www.worldbank.org/en/country/pacificislands/overview
• Prakash, Amit. “Boiling Point.” Finance and Development, September 2018, Vol. 55. No. 3. International Monetary Fund. https://www.imf.org/external/pubs/ft/fandd/2018/09/southeast-asia-climate-change-and-greenhouse-gas-emissions-prakash.htm
• Nansen Initiative Secretariat. “Climate Change, Disasters, and Human Mobility is South Asia and Indian Ocean: Background Paper.” April 5, 2015. Pg 11.
• “Climate Change Danger to South Asia’s Economy.” United Nations Framework Convention on Climate Change. August 19, 2014. https://unfccc.int/news/climate-change-danger-to-south-asias-economy
• Sengupta, Somini, and Nadja Popovich. “Global Warming in South Asia: 800 Million at Risk.” The New York Times. June 28, 2018. https://www.nytimes.com/interactive/2018/06/28/climate/india-pakistan-warming-hotspots.html
• Gleick, Peter. “Water, Drought, Climate Change, and Conflict in Syria.” Pacific Institute of Oakland California. July 1, 2014. https://journals.ametsoc.org/doi/full/10.1175/WCAS-D-13-00059.1
• Werz, Michael, and Laura Conley. “Climate Change, Migration, and Conflict in Northwest Africa.” Center for American Progress. April 2012. Pg 8.
• United Nations Human Rights Council. “The Slow Onset Effects of Climate Change and Human Rights Protection for Cross-Border Migrants.” March 23, 2018. Pg 10.
• Specifically, articles 18.H (share information to better map and predict migration based on climate change and environmental degradation), 18.I (develop adaptation and resilience strategies that prioritize the country of origin), 18.J (factor in human displacement in disaster preparedness strategies), and 18.K (support climate-displaced persons at the sub-regional and regional levels). United Nations. “Global Compact for Safe, Orderly and Regular Migration: Intergovernmentally Negotiated and Agreed Outcomes. July 13, 2018. https://refugeesmigrants.un.org/sites/default/files/180713_agreed_outcome_global_compact_for_migration.pdf
• United Nations. “Global Trends: Forced Displacement in 2018.” June 20, 2019 UNHCR. www.unhcr.org/en-us/statistics/unhcrstats/5d08d7ee7/unhcr-global-trends-2018.html
• United Nations. “Frequently asked questions on climate change and disaster displacement.” November 6, 2016. UNHCR. https://www.unhcr.org/en-us/news/latest/2016/11/581f52dc4/frequently-asked-questions-climate-change-disaster-displacement.html
• Department of Defense. “Quadrennial Defense Review Report” February 2010. Pages 84-89. https://history.defense.gov/Portals/70/Documents/quadrennial/QDR2010.pdf?ver=2014-08-24-144223-573
• Werz, Michael, and Laura Conley. “Climate Change, Migration, and Conflict in Northwest Africa.” Center for American Progress. April 2012. Pg 3.
• Pemberton, Miriam, and Lawrence Korb. “Report of the Task Force on A Unified security Budget for the United States.” Institute for Policy Studies. August 2010. https://fpif.org/wp-content/uploads/2010/08/USB_FY2011.pdf

Migrants, Refugees & Internally Displaced Persons U.S. Foreign Policy

Global Economy and Development

Cheng Li, Xiuye Zhao

June 14, 2023

Pavel K. Baev, Jessica Brandt, Daniel S. Hamilton, Patricia M. Kim, Adam P. Liff, Natalie Sambhi, David G. Victor

May 19, 2023

Online Only

10:00 am - 11:15 am EDT

News from the Columbia Climate School

Climate Migration: An Impending Global Challenge

For months, we have watched the crisis at the Mexican border as migrants tried to enter the U.S. In March, the U.S. Customs and Border Protection office estimated that there were 171,700 people attempting to cross the border—the highest number in 20 years. About 30 percent were families, of which one third were refused entry under Title 42, a public health statute.

The number of unaccompanied children arriving and being held in custody in U.S. border shelters hit over 5,700 in March. And this week, five unaccompanied girls between the ages of seven and 11 months were found at the Texas-Mexico border. While a migrant surge occurs every year as people come to the U.S. for seasonal work, the record number of children being sent by themselves is likely a sign of desperate conditions back home.

Most of those coming to the U.S. are residents of Guatemala, Honduras, and El Salvador, which are considered some of the most dangerous places on Earth. These countries are poor and plagued by gang violence, extortion, and government corruption.

Last year, they were hard hit by COVID, and then two back-to-back hurricanes in November, which killed more than 200 people. The storms displaced over half a million people, buried houses under mud, and destroyed 40 percent of the corn crop and 65 percent of the beans. Many people lost homes, access to clean water, and their livelihoods. The governments did not offer much help.

Climate change — as embodied by the hurricanes — may have been the precipitating factor that pushed many to try to cross the Mexican border into the U.S., but it is usually one of many reasons that people decide to move.

Alex de Sherbinin, associate director for Science Applications and a senior research scientist at the Earth Institute’s Center for International Earth Science Information Network, said, “Climate change, if it’s not currently the main driver of migration, tends to operate indirectly, and will continue to do that. But as the number of severe and extreme weather events or climate-related disasters increases, we’re going to see more migration and more of that may be directly triggered by it.”

Most ongoing migration occurs in response to fast-onset events, such as extreme weather, and usually results in short-term displacement within the person’s own country. This is naturally easier than migrating to another country, and occurs more than three times as often as international migration. But recurring temporary displacements can often lead to permanent displacement. As climate change impacts intensify and living conditions in certain areas gradually worsen, affecting land productivity, access to clean water, food security, and livelihoods, more and more people will likely be forced to leave their homes and potentially cross borders into other countries.

Climate change and migration

For thousands of years, humans have lived mostly on lands where a limited range of comfortable temperatures enabled an abundance of food to grow. Today, only one percent of the world is barely tolerable due to heat; but by 2070, extremely hot zones could make up almost 20 percent of the land, which means that a third of humanity could potentially be living in uninhabitable conditions. For every degree of temperature increase, it’s estimated that one billion people are pushed out of the hospitable zone.

The impacts of climate change — sea level rise, heat waves, storms, drought, and wildfires — will influence global migration. The New York Times reported that  40.5 million people across the planet  were displaced in 2020—the most in 10 years—largely due to these impacts. The 2018 World Bank report Groundswell , which the Earth Institute’s de Sherbinin and Susana Adamo worked on, cited estimates that 30 to 143 million climate migrants may be forced from their homes by climate change impacts by 2050. The Groundswell II report, due out in July and covering all developing countries, pushes the upper bound to more than 200 million.

By 2035, the frequency of major hurricanes is expected increase by 12 percent in the South Indian Ocean, 14 percent in the Atlantic, and 41 percent in the South Pacific compared to 1986-2005 averages. Between now and 2100, sea levels could rise between two and 6.9 feet, submerging millions of homes around the world; sea level rise also creates larger storm surges and can cause saltwater contamination of farmland and drinking water supplies.

Drought has displaced 800,000 within their own countries each year since 2017; in the future, dry regions are expected to get drier still. And over the last decade, wildfires around the world have forced more than 200,000 people per year to leave home; 75 percent of them were in the U.S. Moreover, scientists also believe that there will be more compound weather events , such as flash floods and mud slides that occur after wildfires.

In 2017, approximately 23 million people around the world were displaced due to sudden extreme weather events. Another 44 million or so were displaced due to “humanitarian crises,” likely exacerbated by the cascading effects of climate change. For example, climate experts believe the extreme drought in Syria led to its civil war in 2011. And as crops fail and livelihoods are lost, terrorist groups recruit more desperate people and violence spreads. In countries without the resources to deal with climate change impacts and care for their people, conflicts over resources arise. A recent report by the National Intelligence Council predicted that climate change effects will increase migration, which in turn will put a strain on both origin and destination countries and potentially trigger disputes that could become national security concerns.

Climate migrants of the future

The Groundswell report focused on migration in sub-Saharan Africa, South Asia, and Latin America, which together comprise 55 percent of the developing world’s population. It projected that, if we do not take bold climate action to reduce greenhouse gas emissions and help developing countries, 143 million people from these regions could be forced to move within their own countries to flee the effects of climate change. Migration could also accelerate after 2050 due to stronger climate impacts and population growth. If we can act to stem climate change, the number could be reduced to between 31 and 72 million. Extreme heat events, declining water availability, diminishing snowpack that feeds river basins, and sea level rise will drive people from “hot spots”— such as low-lying cities, coastlines, and places with water scarcity and decreasing crop yields. These migrants will gravitate toward places with a more hospitable climate for agriculture and more job opportunities.

Another study found that higher tides due to sea level rise could affect the land that 150 million people live on by 2050. If the melting of the Antarctic ice sheets picks up speed, 300 million could be affected and up to 480 million by 2100. Seventy percent of the people who would be affected live in eight Asian countries: China, Bangladesh, India, Vietnam, Indonesia, Thailand, the Philippines, and Japan.

Sea level rise has already overtaken eight Pacific islands, forcing the residents to migrate, and two more islands will soon disappear. By 2100, it’s estimated that 48 islands will have been submerged.

Parts of the U.S. are increasingly difficult to live in because of drought, floods, hurricanes, wildfires and sea level rise. A collaborative effort between ProPublica and the New York Times predicted that 162 million Americans will experience a decline in the quality of their environment and by 2070, four million could be living outside of their comfort zone. Because California has experienced record-breaking heat waves and rampant wildfires, with the wildfire season getting longer each year, many Californians are already moving to Idaho, Texas, Oregon and Washington.

Another researcher projected that rising sea levels will force 13 million Americans to move away from the coasts, and that in-land cities such as Atlanta, Orlando, Houston and Austin could each get 250,000 new residents by 2100.

Flooding and high tide are jeopardizing the Quileute Tribe’s school in northwest Washington, forcing the tribe to move the school and tribal government seat 2.5 miles away. In Louisiana, residents of low-lying Isle de Jean Charles are being moved 40 miles to higher ground in the first federally funded U.S. community resettlement project resulting from climate change.

Migration can be considered an adaptation to climate change, as those facing increasingly dire living situations who have the means will likely move. Most people don’t want to leave home, but if they feel they have no choice, they will usually first move from the countryside to a nearby city. If these cities lack the infrastructure or resources to support new residents, they could become overwhelmed.

And migrants with few resources and opportunities often end up living in slums, which are more susceptible to the impacts of climate change and chaos. Today half the global population is urban; the World Bank has estimated that by 2050, 67 percent of humanity will live in cities, with 40 percent of them living in slums by 2030.

Governments in many countries are already reacting to the waves of migrants by holding them in detention centers and erecting walls—Hungary closed off its border with Serbia, India constructed a fence on its Bangladesh border, and under Trump, the U.S. built its own wall. The anti-immigration sentiment has also ushered into office more nationalist governments around the world.

Meanwhile, as the wealthy are able to move to higher or cooler ground or to a more resilient location, some of the most vulnerable people without the means to move—like the poor or elderly—become trapped. This becomes a vicious cycle: As people abandon the community, there is less of a tax base to pay for social services, and those who are left behind and need public support suffer more as they become increasingly desperate. The disparity between rich and poor and their ability or inability to deal with climate change will almost certainly end up creating even more social division than exists now.

Climate migrants have no legal protection

Right now, the world is unprepared to meet the challenge of climate migration. No country offers asylum or legal protection to climate migrants. Because climate change cannot always be identified as the sole or principal reason for migrating, climate migrants have little recourse within current international or U.S. laws. They are not considered “refugees” because they do not fit under the 1951 U.N. Refugee Convention, the U.N.’s legal document that protects refugees—defined as displaced people “who have a well-founded fear of being persecuted because of their race, religion, nationality, membership of a particular social group or political opinion ,  and are unable or unwilling to seek protection from their home countries.”

While some experts believe climate migrants should be included under the “refugee” rubric, others argue that doing so would water down a convention that is already ignored by many countries. De Sherbinin said, “The fact is that Western governments have shown very little desire or inclination to receive migrants from other parts of the world—refugees—even by current standards, so to broaden the definition, you run the risk that people or governments will just say, ‘I’m opting out,’ or ‘I’m not even going to be part of this convention.’ It may be more appropriate to have this be a matter of national decision-making, so governments can decide whether they want to expand the definitions under which they receive refugees.”

The U.N. High Commissioner for Refugees does provide relief help and planned relocation guidance to the “disaster displaced,” and in 2019, appointed a special advisor to help shape the agency’s climate change agenda.

The Global Compact for Safe, Orderly and Regular Migration , an agreement adopted in 2018 by 164 countries, but not the U.S., helps countries prevent displacement due to climate change and provide support to those who are forced to migrate. It is the first international agreement to clarify how countries should deal with international migration, but it is not legally binding.

Because there is no organization that oversees migration, displaced people go wherever they can and not necessarily where it would be best for them to go. As climate migration increases, the situation will likely become more chaotic and overwhelming unless legal and social frameworks are created for the displaced. However, given the current nationalistic and anti-immigrant atmosphere, trying to create an international legal framework to protect climate migrants might actually lead to reducing refugee protections, not extending them.

There are no easy solutions

While there are no easy answers to the complex challenge of climate migration, here are some ideas that have been proposed to make it less chaotic and more humanitarian.

-Enable free movement between member states. For example, the Caribbean has Free Movement Agreements for climate migration. During the 2017 hurricane season, the governments allowed displaced people to move to other islands, waived the need for travel documents and work permits, granted indefinite stays to the displaced, and helped with resettlement.

-Create a Western Hemisphere regional compact on permanent displacement  to expand protection and status for those who have been permanently displaced over an international border by climate change impacts.

-Plan relocation of villages and communities from areas where climate change impacts are threatening. In the Pacific Islands, Vanuatu has developed safeguards and operating procedures for relocation which include technical expertise and financial assistance.

-Ensure that cities are better able to deal with an influx of migrants through investments in infrastructure, sanitation and health services, education, and opportunities for skills and job training.

-Fund more research on how climate change will shape migration so that governments can better predict migration and prepare for it.

-Set up a system of criteria and proof to determine if someone has a credible claim that they were harmed by climate impacts, since this is necessary for the application of laws and rules.

-Build climate migration into policies and long-term planning. This would include helping communities stay where they are by investing in resilience, job opportunities, education and social safety nets. For climate migrants, create incentives to move to low-risk, high-opportunity places. For example, Bangladesh is creating “climate-resilient, migrant-friendly” towns to encourage migrants to move to secondary cities rather than to already overcrowded major cities.

-Pay climate reparations. “The historically largest emitters owe something to the countries that are now being impacted most heavily by climate impacts,” said de Sherbinin. “And providing them with assistance is no longer just a matter of giving handouts—it’s actually an obligation. Because we [wealthy countries] set up the whole climate problem [through our emissions] for many of these countries that have contributed so little to it.”

What President Biden is doing

In an attempt to address the factors that make people want to enter the U.S., President Biden has proposed $4 billion in aid to help Honduras, El Salvador and Guatemala deal with poverty, lack of jobs, government corruption, violence and climate change. He has also announced $310 million of emergency funding to help displaced people. These measures, however, could paradoxically increase migration by giving more people the means to move.

Biden has also issued an executive order for the creation of a report on climate change and migration, with recommendations for the resettlement and protection of migrants. The mayors of 15 cities, including New York, Los Angeles, Chicago, Houston and others, have asked to be consulted for the report, as they “deal with the impacts of climate change and migration on a daily basis.” De Sherbinin is also part of an expert group organized by Refugees International to advise the Biden administration on shaping policies for climate migration.

Under Trump, the number of refugees allowed into the U.S. was set at an historic low of 15,000. President Biden has now raised the refugee cap to 62,500, and plans to increase it to 125,000 in 2022. A number of Republicans criticized the decision as irresponsible and dangerous.

“People in the U.S. perceive that migrants are a burden,” said de Sherbinin. “The reality is the data just don’t prove that in any way. Look at migrants who pay taxes—I’m talking about the undocumented—who are working hard contributing to society. They’re not on welfare, they don’t even qualify for welfare. They’re the ultimate ‘sink or swim’ people who are going to have to pull their own weight. Definitely, there’s a need to recognize and dispel some of the myths about migrants being parasites. But I think, frankly, most of the fear and concern is borne out of a racist reaction to people who are Brown and Black and other colors, and don’t speak our language. They are perceived as a threat.”

The U.S. could actually stand to benefit from an influx of immigrants because it, like other industrialized countries, is facing a demographic decline, which could lead to slow growth and a weakening economy. Migrants who come to the U.S. for better job opportunities would stimulate economic growth through their labor, their buying power and the taxes they pay.

ProPublica and the New York Times modeled international climate migration, as well as political responses to climate change and migration. It showed that how leaders respond to climate change and migration will make a huge difference in how much human suffering there will ultimately be. If they fight climate change aggressively, manage migration humanely, and invest in resiliency, there will be less poverty, less migration, and enough to eat, and these outcomes could help the world be a more stable and peaceful place.

Related Posts

A Virtual Reality Film That Makes the Climate Crisis Feel “Real”

Key Ocean Current Contains a Warning on Climate

Using Nature to Help the Climate: 4 Ways That Work

Celebrate over 50 years of Earth Day with us all month long! Visit our Earth Day website for ideas, resources, and inspiration.

Climate change and its impacts pose a series of threat not only to humans but mother earth at large and we need to put an effort to mitigate it

Get the Columbia Climate School Newsletter →

Climate Change and Migration: Two Case Studies on Mexico

Some items may be under copyright. You are responsible for checking Use and Reproduction terms listed for thie item (below). By downloading, you agree to the Terms of Use.

The main objective of this thesis is to empirically assess the impact of climate factors on international and internal migration in Mexico. Climate models have predicted that Mexico is facing long term drying trends and increased frequency and magnitude of weather related extremes in future decades, two climate factors to which migration is sensitive. Given the historical trends of Mexico-U.S. migration and rural-urban migration within Mexico, I evaluate the extent to which precipitation factors have impacted international and internal migration in the decade of the 1990s. Using migration data extracted from the 1995 and 2000 Mexican censuses and disaster data from the DesInventar database, I assess the impacts of disasters derived from precipitation events on interstate and international migration after controlling for other economic and demographic factors. In the case of international migration, I use a multi-level mixed-effects probit model to measure the extent to which frequency of precipitation based disasters in a state influences the probability that a household in that state has members who recently left abroad. I find that in the period 1990-1994, frequency of disasters is positively associated with the probability that a household has sent migrants abroad in that period. I find further that in the period 1995-1999, disaster frequency is negatively associated with the probability that a household has sent migrants abroad in the general case, but positively associated with the probability that a rural household in a dry state has sent migrants abroad during that period. In the case of internal migration, I construct an expanded gravity model using a Poisson pseudo-maximum likelihood estimator to estimate the impact of the frequency of precipitation based disasters at the origin state on the size of interstate migration flows. I find that higher frequency of disasters at the origin state in the years 1985-1989 and 1990-1994 is associated with higher numbers of migrants for the years 1990-1994 and 1995-1999, respectively. That is, disaster frequency appears to affect migration after a lag period. My findings are consistent with previous research suggesting that climate drivers affect migration in Mexico. Further, as climate stressors are expected to increase over time, we can expect an increased potential for climate change to trigger migration. These findings have implications for policymakers when addressing management of climate risk and future migration flows.

Cross-cutting

• Where We Work
• Data and Resources
• 2030 Agenda

Institutional Strategy

• Climate Change and Migration: Two Case Studies on Mexico

Migration updates 

Subscribe to IOM Environmental Migration Newsletter to receive the latest news and stories.

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• News & Views
• Published: 22 June 2020

PUBLIC OPINION

Perception of climate migrants

• Robert McLeman   ORCID: orcid.org/0000-0001-9593-1606 1  

Nature Climate Change volume  10 ,  pages 600–601 ( 2020 ) Cite this article

1215 Accesses

5 Citations

7 Altmetric

Metrics details

• Climate change
• Developing world
• Environmental social sciences

Migrants arriving in cities must overcome many challenges, including gaining acceptance from established residents. New research from Kenya and Vietnam shows urbanites accept climate hazards as being as legitimate as economic, political or social motivations for rural-to-urban migration.

This is a preview of subscription content, access via your institution

Relevant articles

Open Access articles citing this article.

Exploring interactions between socioeconomic context and natural hazards on human population displacement

• Michele Ronco
• , José María Tárraga
•  …  Gustau Camps-Valls

Nature Communications Open Access 04 December 2023

Interactions between migrant race and social status in predicting acceptance of climate migrants in Norway

• Hege H. Bye
•  …  Charles A. Ogunbode

Climatic Change Open Access 15 April 2023

Access options

Access Nature and 54 other Nature Portfolio journals

Get Nature+, our best-value online-access subscription

24,99 € / 30 days

cancel any time

Subscribe to this journal

Receive 12 print issues and online access

195,33 € per year

only 16,28 € per issue

Rent or buy this article

Prices vary by article type

Prices may be subject to local taxes which are calculated during checkout

Linke, A. M., Witmer, F. D., O’Loughlin, J., McCabe, J. T. & Tir, J. Environ. Res. Lett. 13 , L094014 (2018).

Article   Google Scholar  

Connell, J. & Lutkehaus, N. Environmental Refugees? Aust. Geogr. 48 , 79–95 (2017).

Maharjan, A. et al. Curr. Clim. Change Rep. 6 , 1–16 (2020).

Adger, W. N. et al. Lancet Planet. Health 3 , e440–e442 (2019).

Tebboth, M. G. L., Conway, D. & Adger, W. N. Glob. Environ. Chang. 54 , 172–183 (2019).

Gemenne, F. & Blocher, J. Geogr. J. 183 , 336–347 (2017).

Spilker, G., Nguyen, Q., Koubi, V. & Böhmelt, T. Nat. Clim. Change https://doi.org/10.1038/s41558-020-0805-1 (2020).

Dun, O. Int. Migr. 49 , e200–e223 (2011).

Eriksen, S. & Lind, J. Environ. Manage. 43 , 817–835 (2009).

Ng’ang’a, S. K., Bulte, E. H., Giller, K. E., McIntire, J. M. & Rufino, M. C. World Dev. 84 , 55–68 (2016).

Gröger, A. & Zylberberg, Y. Am. Econ. J. Appl. Econ. 8 , 123–153 (2016).

Adams, C., Ide, T., Barnett, J. & Detges, A. Nat. Clim. Change 8 , 200–203 (2018).

McLeman, R. Gt. Plains Q. 26 , 27–40 (2006).

Google Scholar  

Download references

Author information

Authors and affiliations.

Department of Geography and Environmental Studies, Wilfrid Laurier University, Waterloo, Ontario, Canada

Robert McLeman

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Robert McLeman .

Rights and permissions

Reprints and permissions

About this article

Cite this article.

McLeman, R. Perception of climate migrants. Nat. Clim. Chang. 10 , 600–601 (2020). https://doi.org/10.1038/s41558-020-0803-3

Download citation

Published : 22 June 2020

Issue Date : July 2020

DOI : https://doi.org/10.1038/s41558-020-0803-3

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

This article is cited by

• José María Tárraga
• Gustau Camps-Valls

Nature Communications (2023)

• Charles A. Ogunbode

Climatic Change (2023)

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

• The Accelerator
• Student Hub
• Faculty Hub

Salata Institute Funds Eight New Climate Research Projects 

The Salata Institute Seed Grant Program supports new research, encourages new interdisciplinary partnerships, and enables faculty whose work is not normally in climate and sustainability to apply their expertise to the climate challenge. With the addition of these eight awards, the Salata Institute now supports 27 research projects on understudied and emerging climate topics through this program. 

This program is part of the Salata Institute’s aggressive efforts to expand climate and sustainability research at Harvard University. Since its launch in June 2022, the Institute has awarded over $8.68M in climate and sustainability research funding, supporting the work of 62 faculty from across Harvard University.  

Harvard faculty members interested in the Salata Institute Seed Grant Program, which is supported by a gift from the Troper Wojcicki Foundation, can access the current call for proposals here to learn more about the program. Applications for the Salata Institute Seed Grant Program will be considered three times per year, with deadlines of the second Friday of January, May, and September. 

Direct Air Capture of CO2: Redox-Mediated Salt Splitting  

Principal Investigator: Michael Aziz, Gene and Tracy Sykes Professor of Materials and Energy Technologies, Harvard John A. Paulson School of Engineering and Applied Sciences  

Despite global efforts, the world is not yet on track to reduce greenhouse gas emissions enough to avert some of the harshest climate change scenarios. In some regions and sectors of the economy, emissions reductions could take decades to materialize. Direct Air Capture (DAC), or the removal of carbon dioxide directly from the air, could work in tandem with emissions reductions to stem climate risks. But current DAC technology can be energy-intensive and expensive, making it difficult or impossible to scale to the degree needed to have a meaningful climate benefit. 

With seed grant funding from the Salata Institute, researchers led by Principal Investigator Michael Aziz will develop a novel electrochemical process, known as redox-mediated salt splitting (RMSS). The researchers envision this process as the centerpiece of a new chemical cycle for the capture and removal of carbon dioxide from the air. “The proposed electrochemical process holds promise as a platform for scalable carbon dioxide capture operating in a continuous cycle, based on entirely sustainable and non-critical materials, powered by potentially inexpensive carbon-free electricity,” said Aziz.  

The seed grant award will support researchers as they work toward a successful demonstration of RMSS at the lab-scale proof-of-concept stage.  

Quantifying the Magnitude of Zoogeochemical Effects on Vegetation Carbon Stocks  

Principal Investigator: Andrew Davies, Assistant Professor of Organismic and Evolutionary Biology, Harvard University  

Currently, most regional and global carbon cycle models – models that are important to our understanding of the trajectory of climate change – leave animals out of the picture. But researchers in the emerging field of zoogeochemistry point out that animals play a critical role in shaping carbon cycles. Take elephants, for example. “Many tree species in Congo forests rely exclusively on forest elephants to disperse seeds over long distances or to locations favorable for germination,” said Davies. “These tree species have large fruits and seeds that can only be handled by elephants, and become tall, large-diameter trees with high carbon storage.” 

Leaving animals out of carbon cycle models can contribute to uncertainties about the role animals could play in natural climate solutions. 

With seed grant funding from the Salata Institute, Principal Investigator Andrew Davies will seek to uncover the role of critically endangered African forest elephants, Loxodonta cyclotis , in determining the amount of carbon that can be stored in vegetation in the Congo Basin. Davies looks to use field plots, Unoccupied Aerial Vehicle-Light Detection and Ranging (UAV-LiDAR) and satellite data combined with forest elephant distribution data and abiotic (climate, geology and elevation) variables to disentangle the role of forest elephants as drivers of spatial variation in aboveground carbon stocks (ACS).  

Climate Change and Volatility in Food Supply: A New Workshop  

Principal Investigator: Peter Huybers, Department Chair, Professor of Earth and Planetary Sciences, Harvard University  

Between 1950 and the early 2010s, something remarkable happened: despite a tripling of the human population, the number of people without enough to eat was falling. Over the past decade, that trend has reversed.  Rising food insecurity, driven by many intersecting trends, accelerated during the COVID19 pandemic.  

Looking to the future, the potential for increasingly frequent and severe heat waves, sunlight, flood, and drought variations may lead to more frequent and intense food shocks.  The joint and interacting influence of these climatic factors, however, has never been empirically estimated at the global scale, in part due to data constraints. Similarly, the degree to which farmers have adapted to these individual and interacting shocks is, surprisingly, unknown at the global scale. Together, this leaves large uncertainties in the effect of climate change on future food supply and food security.   

With seed grant funding from the Salata Institute, Peter Huybers and Missy Holbrook are convening a workshop focused on acquiring and deploying the necessary data to better understand the implications of climate change for food production. 

A Meeting of the Macroeconomics Minds on Climate  

Principal Investigator: Adrien Bilal, Assistant Professor of Economics, Harvard University     The scale of climate change and the transition to a decarbonized economy raises important questions for macroeconomists. Among the questions that demand a closer look are: What are the economic consequences of climate change and extreme events? How will the comparative advantages of sectors, regions and firms in energy use and generation shift and interact with the climate?    With seed grant funding from the Salata Institute, Adrien Bilal, Assistant Professor of Economics at Harvard University, will host a new workshop to bring together macroeconomists considering these and other climate policy and traditional macroeconomic policy questions. At the workshop, taking place in April at Harvard University, researchers will discuss the current research and ways macroeconomists can strengthen and improve the response to climate change, both by society broadly and, more narrowly, by those responsible for climate policy and traditional macroeconomic policy. 

International Workshop on Climate-Resilient Development in Southeast Asia  

Principal Investigator: Michael McElroy, Gilbert Butler Professor of Environmental Studies,  

Harvard Paulson School of Engineering & Applied Sciences (SEAS) and FAS Department of Earth and Planetary Sciences, Harvard University  

Climate change is projected to cause growing loss and damage to societies around the world in the near future, even if current mitigation goals are met. How to maintain and enhance “climate resilience” – the capacity of socio-ecosystems to survive and develop in a changing climate – is of utmost importance.  

Countries in Southeast Asia are developing social and economic systems and building capacity to cope with global warming and the possibility of a climate “overshoot” (a scenario in which the world warms by greater than 1.5 degrees Celsius for some period). Yet several important knowledge gaps persist. Harvard researchers point to a geographical bias to study resilience in big cities and coastal areas; a lack of focus on connections between climate resilience and environmental health; a lack of modeling in developing contexts; and a failure to reconcile the scientific drivers of climate change with planning on how to effectively deliver assistance.  

With seed grant funding from the Salata Institute, Principal Investigator Michael McElroy along with Co-Principal Investigators Dr. Michael VanRooyen and Liang Emlyn Yang will host a two-day workshop in July to promote research exchange among scholars and inform climate resilient development pathways in Southeast Asia.  

“Promoting climate-resilient development is a key priority in fields like civil protection, urban planning, health care and others. This workshop will bring together a diversity of regions, disciplines, methodologies, and scholars at varying career stages to foster international climate resilience research,” said McElroy.  

African Perspectives on International Climate Law – A Symposium  

Principal Investigator: Gerald Neuman, J. Sinclair Armstrong Professor of International, Foreign, and Comparative Law, Harvard Law School; Director, Human Rights Program, Harvard Law School  

The number of climate change-related judicial proceedings has increased 2.5-fold globally since 2017. Against the backdrop of ineffective political negotiations, litigation with potentially far-reaching implications on the obligations of states will have bearings on overlapping legal fields including environmental law, law of the sea, trade and investment law, human rights law and, more broadly, the law of state responsibility.  

Yet African perspectives, while crucial and having the potential to shape the world’s response to climate change, are at risk of being overlooked in global norm-making processes due to pre-existing power structures.   

With seed grant funding from the Salata Institute, the Human Rights Program at Harvard Law School launched a project that solicits and addresses African perspectives on international and regional legal and policy debates on climate change.  

“In light of the climate change threats in Africa, and the historic underrepresentation of African voices in the development of international law, it is important to work out African perspectives on global developments in climate change law in a way that takes into account regional, sub-regional, state, and local differences,” said Principal Investigator Gerald Neuman, who is working alongside Co-Principal Investigators Alicia Ely Yamin and Abadir M. Ibrahim.  

Consisting of a series of symposia and publications, the project explores African contributions to what can be described as a transnational normative dialogue on international climate law and overlapping legal fields. 

The Changing Composition of Soil Carbon Inputs: Root Exudate Responses to Climate Change  

Principal Investigator: Benton Taylor, Assistant Professor, Department of Organismic & Evolutionary Biology, Harvard University  

The natural ability of soils to store carbon is one of the most important stages of the global carbon cycle and will play a key role in future climate change mitigation. Root exudates – organic compounds like amino acids and simple sugars secreted from roots into the soil – account for 20% of below-ground carbon allocation. Climate change is having a dramatic impact on this process. Yet most research has focused on how rising atmospheric carbon dioxide is changing exudation rates, rather than changes to the chemical composition of exudates. 

“In recent work, we have shown that differences between common exudate compounds have a larger effect on stable soil carbon formation than even a five-fold change in the amount of exudation. As we develop models of soil carbon dynamics under future climate scenarios, it is critical we consider exudate chemistry,” said Taylor.   

To prepare a proposal for a large-scale study on how climate change will impact the composition of exudates, Taylor will use a Salata Institute seed grant to collect the first data on changing exudate chemistry from in-situ, large-scale global change experiments ongoing in Minnesota and Illinois.  

Adaptive Climate Migration: A Case Study of Resettlement from the Solomon Islands to Canada  

Principal Investigator: Hannah Teicher, Assistant Professor of Urban Planning, Department of Urban Planning and Design, Harvard University Graduate School of Design  

By 2050, the current numbers of climate-related migrants will swell, with estimates ranging from 25 million to 1.2 billion. 

Climate migrants tend to experience a host of negative impacts as they relocate, from declining mental and physical health to the loss of livelihoods. A comprehensive, formal strategy to facilitate climate-related relocation could enhance the well-being of climate migrants and even improve community adaptation.  

But formal recognition of climate migrants as a distinct immigrant or refugee category with its own set of policies and programmatic responses would raise complex challenges and a host of questions: Who qualifies? How would places be deemed uninhabitable? Can current legal frameworks, such as the 1951 Convention on Refugees, be expanded without exacerbating anti-immigrant sentiment?   

“In the context of current migration law, crossing international borders due to climate-related events leaves migrants in legal and administrative limbo. With each passing year of record-breaking climate impacts, filling these institutional and legal gaps takes on greater urgency,” said Teicher.  

With seed grant funding from the Salata Institute, Principal Investigator Hannah Teicher, alongside Co-Principal Investigator Michael Hooper, will study a pilot program in British Columbia that is facilitating migration from the low-lying Solomon Islands. The study will entail semi-structured interviews over the summer of 2024 with three groups: Solomon Islanders relocating to Canada; decision-makers at resettlement-related NGOs and public agencies in Canada; and provincial and federal policymakers with an interest in revising climate migration policy. The researchers intend the findings to inform policy agendas and lay the foundation for more extensive comparative international research on climate-related migration and resettlement.  

Melting glaciers in a warmer climate provide new ground for invasive species

A case study on the island of south georgia.

Invasive species have rapidly colonised new ground exposed by melting glaciers in the sub-Antarctic island of South Georgia, according to new research.

Invasive species brought to new territories through human activities are one of the main causes of the ongoing biodiversity crisis. Even on South Georgia, a remote island located in the very south of the Atlantic Ocean, exotic species are present.

In a new study published in the open access journal Neobiota and funded by Darwin Plus, the researchers Pierre Tichit (Durham University, now Swedish Agricultural University), Paul Brickle (South Atlantic Environmental Research Institute), Rosemary Newton (Royal Botanic Gardens, Kew), Peter Convey (British Antarctic Survey) and Wayne Dawson (Durham University, now University of Liverpool) look at how living organisms colonise new ground provided by melting glaciers on the British overseas territory.

Many were inadvertently introduced by whalers and sealers in the 19 th and early 20 th centuries. Like other cold regions of the world, South Georgia has another problem: many of its glaciers are melting at a fast pace because of climate change, leaving behind large areas of newly uncovered bare ground.

The authors surveyed the forelands biodiversity of six glaciers by counting plants, turning rocks, laying traps and using sweep nets, enabling an inventory of the flora and fauna that colonises forelands at different stages of their retreat.

Their results indicate that invasive species will likely spread on South Georgia as fast as glaciers are retreating. Whether this has or will have negative consequences on local species needs to be investigated to help protect this unique ecosystem.

Just a few years after bare ground is exposed by glacier melting, pioneer plants arrive, progressively covering more ground with time and followed by an increasing number of species. The study discovered that not only native, but also exotic plants and invertebrates, are taking advantage of this opportunity. Even more surprising, two temperate plant species from the Northern Hemisphere, annual meadow grass and mouse-ear chickweed, colonised sites faster than any other species.

Scientific expeditions to such an isolated and inhospitable island are challenging. The crossing from the Falkland Islands to reach South Georgia takes several days on a notoriously temperamental ocean. Once on the island, most glaciers are only accessible with small boats followed by hikes through difficult terrain.

• New Species
• Endangered Animals
• Ecology Research
• Global Warming
• Exotic Species
• Environmental Awareness
• Invasive species
• Water hyacinth
• Larsen Ice Shelf
• Zebra mussel
• Water resources

Story Source:

Materials provided by Pensoft Publishers . Note: Content may be edited for style and length.

Journal Reference :

• Pierre Tichit, Paul Brickle, Rosemary J. Newton, Peter Convey, Wayne Dawson. Introduced species infiltrate recent stages of succession after glacial retreat on sub-Antarctic South Georgia . NeoBiota , 2024; 92: 85 DOI: 10.3897/neobiota.92.117226

Cite This Page :

Explore More

• 3D Mouth of an Ancient Jawless Fish
• Connecting Lab-Grown Brain Cells
• Device: Self-Healing Materials, Drug Delivery
• How We Perceive Bitter Taste
• Next-Generation Digital Displays
• Feeling Insulted? How to Rid Yourself of Anger
• Pregnancy Accelerates Biological Aging
• Tiny Plastic Particles Are Found Everywhere
• What's Quieter Than a Fish? A School of Them
• Do Odd Bones Belong to Gigantic Ichthyosaurs?

Trending Topics

Strange & offbeat.

Assessing future changes in flood frequencies under CMIP6 climate projections using SWAT modeling: a case study of Bitlis Creek, Turkey

• Article contents
• Figures & tables
• Supplementary Data
• Open the PDF for in another window
• Guest Access
• Cite Icon Cite
• Permissions
• Search Site

Emrah Yalcin; Assessing future changes in flood frequencies under CMIP6 climate projections using SWAT modeling: a case study of Bitlis Creek, Turkey. Journal of Water and Climate Change 2024; jwc2024646. doi: https://doi.org/10.2166/wcc.2024.646

Download citation file:

• Ris (Zotero)
• Reference Manager

Climate change is altering flood risk globally, with local variations prompting the necessity for regional assessments to guide the planning and management of water-related infrastructures. This study details an integrated framework for assessing future changes in flood frequencies, using the case of Bitlis Creek (Turkey). The precipitation and temperature simulations of 21 global circulation models (GCMs) from the coupled model intercomparison project phase 6 (CMIP6) are used to drive the developed soil and water assessment tool (SWAT) model in generating daily streamflow projections under the CMIP6 historical experiment and the shared socio-economic pathway (SSP) scenarios of SSP245 and SSP585. Five probability distribution functions are considered to calculate the 5-, 10-, 25-, 50-, 100-, and 500-year flood discharges for the historical period 1955–2010 and the future periods 2025–2074 and 2025–2099. The quantification of climate change impacts on the design discharges is based on the medians of the flood discharges obtained for the climate data of each GCM, using the best-fitted distribution functions according to the Kolmogorov–Smirnov test results. The findings illustrate significant increases in discharge rates, ranging from 21.1 to 31.7% for the 2025–2099 period under the SSP585 scenario, highlighting the necessity of considering changing climate conditions in designing water-related infrastructures.

A framework is proposed for assessing possible changes in flood frequencies under the climate projections of CMIP6 GCMs using the SWAT model.

The quantification relies on the medians of flood discharges obtained from the climate projections of each GCM.

The methodology applied to Bitlis Creek indicates that the changing climate may lead to notable increases in flood discharges, emphasizing the need for adaptive measures.

Supplementary data

Affiliations

• ISSN 2040-2244 EISSN 2408-9354
• Open Access
• Collections
• Subscriptions
• Subscribe to Open
• Editorial Services
• Rights and Permissions
• Sign Up for Our Mailing List
• IWA Publishing
• Republic – Export Building, Units 1.04 & 1.05
• 1 Clove Crescent
• London, E14 2BA, UK
• Telephone:  +44 208 054 8208
• Fax:  +44 207 654 5555
• IWAPublishing.com
• IWA-network.org
• IWA-connect.org
• Cookie Policy
• Terms & Conditions
• Get Adobe Acrobat Reader
• ©Copyright 2021 IWA Publishing

This Feature Is Available To Subscribers Only

Sign In or Create an Account

Let the people have a say

Zimbabweans report worsening droughts, favour climate action by the government

• A majority (60%) of Zimbabweans say droughts have become “somewhat more severe” or “much more severe” over the past 10 years. o Compared to 2017, the proportion of Zimbabweans who say droughts have gotten more severe has increased by 27 percentage points. o Rural residents are more likely than urbanites to report increasingly severe droughts (64% vs. 54%) (Figure 2). o Geographically, at least three-fourths of citizens in Midlands (78%), Bulawayo/Matabeleland North/Matabeleland South (78%), and Mashonaland Central (75%) report worsening drought, compared to just 41% in Harare.
• Fewer than half (45%) of Zimbabweans say they have heard of climate change. Among those who are aware of climate change: o Eight in 10 (80%) say it is making life in Zimbabwe worse. o Almost two-thirds (64%) believe that ordinary citizens can help curb climate change, and even more (71%) want the government to take immediate action to limit climate change, even if it is expensive, causes job losses, or takes a toll on the economy. o Only small minorities are satisfied with efforts to date by the government (2%), business and industry (3%), developed countries (4%), and ordinary citizens (11%) to fight climate change (Figure 4). Large majorities believe more needs to be done, including 78% who say the government needs to do “a lot more.”

Six in 10 Zimbabweans say they are experiencing worsening droughts, setting the stage for urgent climate action by the government, according to the most recent Afrobarometer survey.

Zimbabwean President Emmerson Mnangagwa has declared a national disaster in response to droughts impacting farming activities in the country. With millions facing hunger, he has stressed the urgent need for more than $2 billion in aid to ensure sufficient food supplies. The latest Afrobarometer survey in Zimbabwe, conducted in 2022, shows that the proportion of citizens experiencing more severe droughts has almost doubled since 2017.

While climate change is still an unknown concept to more than half of Zimbabweans, those who are aware of it overwhelmingly say that climate change is making their lives worse. Citizens expect far more climate action from the government – even at considerable cost to the economy – as well as from developed countries, the private sector, and their co-citizens.