As extreme weather events, including floods, heat waves, droughts, and wildfires, have grown more frequent and severe, climate migration has become more commonplace. But to what extent does climate change vulnerability impact people’s decision to relocate? Existing literature tends to blame resource scarcity, suggesting that vulnerability to climate change merely accelerates migration. SPA Professor Todd Eisenstadt and coauthors recently published a piece in Global Environmental Politics sharing results from their 2023 national survey conducted in climate-vulnerable Guatemala (funded by AU's School of Public Affairs (Helfat Award), Tulane University, and Loyola University Chicago). They found a stronger relationship between a respondent’s sense of climate vulnerability and their propensity to consider migrating when compared to their sense of economic hardship or poor government response to climate-related extreme weather. Extreme weather events serve as an intervening variable between economic adversity and migration, confirming the value of scholarship on climate vulnerability and adaptation policies.
Read the Full PDF

Drivers of Migration

Though research on the climate–migration nexus has mounted in recent decades, most studies focused on host community perceptions of incoming migrants. Little work considers potential migrants’ perception of climate change as a driver of migration, with most studies attributing it a secondary, catalytic role. Further, few studies have explored citizen attribution of damage from extreme weather events to government action on climate adaptation. While national and subnational governments are pivotal in implementing adaptation strategies, prior research suggests that Guatemalans may lack sufficient information to “blame” the government for climate event-related damage. In other words, internationally-available information does not “trickle down” to citizens on the ground in vulnerable areas.

Twenty-one decision making factors found to affect climate migration were later divided into six categories, including economic, environmental, demographic, political, social, and personal/household characteristics. Factors range from micro-level explanations of individual motives (e.g., the current study) to meso-level discussions of the role of social networks and macro-level explanations emphasizing opportunity structures, income, and employment differentials across contexts.

The migrant literature has focused mostly on economic circumstances, such as wage differentials between sending and receiving countries. Just the past decade has produced a clearer consensus that vulnerability and climate change impact migration. Absent climate events or other external shocks, models indicate that migration levels rise over time as poorer countries develop, up to a “peak,” then diminish. Experience with extreme weather events—acting as an intervening variable or a multiplier—can cause more economic adversity and prompt vulnerable communities to consider migration. Other studies have concluded that rapid-onset phenomena, such as floods, are more likely to lead to short-term displacements and relocations than to long-term migrations. This “trapped populations” argument holds that victims of floods tend to stay where they live, or move only a short distance, due to lack of resources.

Perhaps the most important criticism of climate migration scholarship is its inability so far to isolate the impact of one causal factor over others. Even when studies confirm a relationship between climate change and migration, they are often indirect and heterogenous. While forced migration is a consequence of climate change, migration also happens for other reasons.

Approaches to explaining individual decisions to migrate all describe migration as an economically rational response to circumstances. The more recent structural models of migration, by contrast, emphasize economic factors, migration laws, and other political features of the “receiving” environment. Some have sought to examine the impact of climatic events on migration by accounting for migration costs and returns, and multicausality.

Despite these scholarly contributions, climate migration has remained a contested concept. Its politicization has caused many social scientists to identify climate migration as a leading symptom of climate change without fully exploring its causes and conditioners. A deeper understanding of the link between displacements and vulnerability is necessary for creating institutional mechanisms to enable vulnerable populations to adapt to climate change through a range of policies, which may include encouraging migration.

The Context of Guatemala

The study considers four explanations for migration—climate vulnerability, perceptions of climate change, economic scarcity, and satisfaction with government. Guatemala provides a unique opportunity to assess the impact of climate exposure on the propensity to migrate: 44% of respondents reported notable out-migration from their communities over the last five years. The Climate Risk Index listed it among the world’s 20 most climate risk-prone nations in terms of human and physical damage. Some 235 deaths were attributed to Tropical Storm Alex in 2010, which left almost 210,000 homeless (World Bank 2011). Recent events include Hurricane Mitch (1998), Tropical Storm Stan (2005), and Tropical Storm Agatha (2010), which alone cost Guatemala US$2.4 billion in damages. These storms and floods have intensified recently, due to changes in El Niño and La Niña weather patterns. The 2014 drought affected 70-80% of Guatemala’s basic food crops and 1.1 million people, prompting the declaration of a national emergency.

Although internal, urban migration has been Guatemala’s most prevalent form, international migration has also doubled since 1990, to 692,000 in 2020. While most migration studies in the region have focused on factors like poverty, inequality, violence, and conflict, climate variability and exposure to extreme weather events have been established as drivers. A common causal mechanism through which both droughts and floods have impacted both internal and international migration is food insecurity. Care International (2022) found that 21% of rural Guatemalan households went into debt to buy food, while 38% reduced the sizes of their meals.

The Study

The study involved an original survey, conducted via face-to-face interviews in Guatemala between May and June 2023. The national sample included 950 adult respondents and an oversample of 400 more from climate change-vulnerable areas. Hypotheses addressed vulnerability (“H1a: The more respondents are exposed to climate-related extreme weather events including flood and/or drought, the more likely they are to consider migrating” and “H1b: The more respondents experience harm due to natural hazards, the more likely they are to consider migrating”), climate change perception (“H2: The more respondents believe that climate change is a serious issue, the more likely they are to consider migrating”); economic scarcity (“H3: The more respondents experienced economic difficulties in recent years, the more likely they are to consider migrating”); and government satisfaction (“H4: The more respondents are satisfied with government institutions, the less likely they are to consider migrating”).

Survey questions addressed each respondent’s exposure to extreme weather (i.e., flood, drought), corresponding physical or human damage and loss, related government interventions, climate change perceptions, economic adversity, and government trust. Analyses included estimated logit models, an economic loss index, and a government performance index. Controls were instituted for gender, income, education, and access to drinking water.

Key Findings

• 13.5% of the national sample reported considering migration due to climate events, as did 20% of respondents in the drought oversample and 37% in the flood oversample.
• 22% of respondents in the national sample and 18% in the drought sample reported more flooding over the past five years, compared to 59% in the flood sample.
• 33% of respondents in the national sample and 52% in the flood sample reported more drought over the past five years, compared to 56% in the drought sample.
• Climate vulnerability is higher in both flood and drought zones compared to the national average, with the flood zone residents expressing the highest vulnerability.
• A positive relationship exists between an individual’s climate vulnerability (measured by experiencing harm or damage) and their consideration of migration.
• 57% of respondents in the national sample reported increased migration among community members.
• Only 38.5% of respondents reported waiting to migrate because “the natural disasters are still not so bad.”
• Survey respondents with prior extreme weather experience are more likely to distrust their national government, which may prompt them to consider migrating.
• 60% of respondents knew someone affected––financially or otherwise––by climate events.
• While 32% would move for a higher wage even if this meant more exposure to natural hazards, 68% preferred lower wages but less climate risk.
• Only 17% had heard anything about local government planning to minimize damage from future disasters, and only 21% had even heard the term climate change adaptation.

Results

Respondents with greater exposure to extreme weather events (regardless of their economic or education levels or perceptions of government responsiveness) were more likely to consider migration compared to those not directly exposed. While Guatemala is a country with high vulnerability, and respondents often pointed to a link between climate vulnerability and economic hardship, the strength of these findings implies the need for new policies to treat climate-related migration (and its causes), separately from economic hardship.

Respondents who believe that flood and drought have gotten worse in the past five years are significantly more likely to think about migration than those who do not (H1a), as were those whose houses suffered damage due to climate change events, were forced to temporarily relocate, or suffered economic loss for similar reasons (H1b). While those who worry about climate change were more likely to consider migration, this connection was not statistically significant (H2). Empirical evidence in support of the economic difficulty hypothesis was found (H3). Those who have recently faced economic difficulties are more likely to consider migrating, while those unable to provide food for their families are less likely to, confirming the “trapped population” thesis. Migration requires a minimum of capital, likely unavailable to those who cannot provide for household food needs. While perception of government performance was statistically significant in the predicted direction, perception of corruption was significant, but not in the expected direction. Those who have received assistance from the government are more likely to think about migration, while no significant relationship appeared between government trust and migration propensity.

Together, the evidence from different estimated models provides strong support for the climate harm hypothesis: experiencing physical and financial harm due to climate change-related extreme weather events significantly increases the propensity to consider migration among adult Guatemalans. This finding indicates that the most salient issue in an individual’s decision to migrate is their experience of harm from natural disasters.

This study aimed for a ground-up view of people’s experiences with climate vulnerability, to 1) see whether it drove them to consider migration and 2) to understand whether this varied according to the type of threat faced (floods or droughts). In a response to recent calls to refocus climate migration studies on affected populations and their vulnerability to climate events, it found widespread experience with adverse events tied to climate change.

Confirming the role of climate change experience/vulnerability as an intervening variable and a multiplier of economic adversity, citizens seem to translate climate issues into economic terms. A full 58% of respondents in the national sample reported that their migration intention was to get a better job, followed by feeling safer from violence and crime (10%) and avoiding national disasters (9%). When asked “What are the main reasons you haven’t yet moved?”, 50.1% said “the economy is not yet bad enough here,” 53% stated “I don’t have the savings to move,” and 83% said, “My family is here.”

Guatemalans do not seem to think much at all about the government’s role in responding to climate phenomena. Seventy percent reported that they would “organize response on their own or with their neighbors” when natural disasters occurred, rather than waiting for government (13%) or for nongovernment (5%) emergency responders.

Limitations

Eisenstadt’s study only looks at respondents still living in Guatemala and considering a permanent move, though underlying factors in respondents’ propensities to migrate may hold true across a range of nations suffering from climate migration. One may argue that this misses out on actual “migrants,” preferring an investigation into how Guatemalans living elsewhere did migrate and why. However, finding a representative sample of the Guatemalan migrant diaspora would be prohibitively expensive and challenging, and sacrifice the data from those who chose not to migrate in the end. Future research might recruit representative samples of people who left and those who stayed at home.

Policy and Scholarly Implications

Greater consideration must be given to climate vulnerability and adaptation policies, as opposed to economic hardship, and to the attitudes of those who suffer climate events but do not migrate. While these findings validate the idea that many choose to stay behind because they cannot afford to leave, the cross-sectional research design precluded considering those who actually sought to migrate but could not.

The work contributes to an emerging literature showing that extreme weather events provoke migration (albeit as an intervening variable), in contrast to theories championing its economic causes. Economic hardships resulting from climate events may trigger thoughts of migration: flood victims described damage to their homes and crops and drought victims spoke of debt spirals from crop failures and earlier borrowing. The growing prospect that climate vulnerability causes people to think about migrating, activating their economic logic and forcing migration when possible, should matter to climate analysts and planners.

Conclusions

Increasingly, national tort laws are being used to seek accountability from climate polluters for exogenous damage caused by tropical floods, droughts, glacier melt, and sea level rise. In 2015, the term “loss and damage” was included in the UN Framework Convention on Climate Change (UNFCCC) agreement to refer to climate change-induced harm, whether reparable or irreparable, to which countries cannot adapt. Migration represents perhaps the single most disruptive and widespread phenomenon, but also the likeliest response to long-term vulnerability.

These findings confirm the multicausality of climate migration. While the four climate migration drivers studied—climate vulnerability, perceptions of climate change, economic scarcity, and satisfaction with government—were not statistically correlated, evidence suggests that migration is multicausal and that climate change vulnerability functions as an intervening variable and a multiplier of economic adversity, which can significantly impact the decision to migrate. Moreover, migratory intent is statistically tied to experience with direct or indirect economic damage from adverse climate events.

Without prompting or training to associate climate event outcomes with government- and community-driven adaptation efforts, vulnerable communities will not consider these factors as they decide whether to migrate. Vulnerable nations must make choices about economic investment at least partly based on climate vulnerability and create policies and infrastructure to reduce this risk.