Climate Migration: What the Evidence Says About Movement, Displacement and Staying Put

1Background: movement, displacement and immobility

Climate hazards are a growing driver of involuntary migration and displacement, according to the IPCC’s Sixth Assessment Report, which rates this finding as high confidence. It also notes that these impacts are unevenly distributed and will continue to be experienced inequitably because of differences in exposure and vulnerability.^[1] The same assessment makes a distinction that is often lost in public debate: climatic conditions and events are important drivers of migration, but responses to specific hazards are strongly influenced by economic, social, political and demographic processes.^[1]

This brief asks three questions. Where do people move, and how many? How do projections describe the future? And what does the evidence say about the choices and constraints behind movement, including the choice to stay?

2Where people move

Most climate-related migration and displacement observed currently takes place within countries. The IPCC adds that observed international migration associated with climatic hazards is considerably smaller than internal migration and is most often between neighboring states that have labor-migration agreements or longstanding cultural ties. Systematic reviews find few documented examples of long-distance, inter-regional migration driven by climate events.^[1]

The scale of internal disaster displacement is large. The Internal Displacement Monitoring Centre (IDMC) recorded 29.9 million disaster displacements across 140 countries and territories in 2025, a 35% fall from the exceptional 2024 but 13% above the decadal average. Storms triggered 17.9 million movements, about 60% of the total, and a small number of high-intensity events caused significant displacement, notably in Jamaica, Sri Lanka and Viet Nam (Figure 1).^[2] At the end of 2025, almost 13.6 million people were living in displacement because of disasters, out of 82.2 million internally displaced overall.^[2]

Two cautions apply. For the first time, conflict and violence caused more displacements than disasters in 2025, and 42 countries recorded movements linked to both triggers, which shows how hazard and fragility overlap.^[2] And a displacement is a movement, not a person: the same household can be counted more than once, and many movements are temporary evacuations.^[2]

3What projections say

The most-cited projection is the World Bank’s Groundswell series. Combining its two reports, it finds that without early and concerted climate and development action, as many as 216 million people could move within their own countries by 2050 because of climate change impacts across six regions. By region, the report gives as many as 86 million in Sub-Saharan Africa, 49 million in East Asia and the Pacific, 40 million in South Asia, 19 million in North Africa, 17 million in Latin America, and 5 million in Eastern Europe and Central Asia.^[3]

The report says hotspots of internal climate migration could emerge as early as 2030 and spread and intensify by 2050, and that immediate and concerted action on emissions and inclusive, resilient development could reduce the scale of climate migration by as much as 80%.^[3] The report identifies areas that people are expected to move from because of increasing water scarcity, declining crop productivity and sea-level rise, and it uses a scenario-based approach, so the figures describe possible outcomes, not forecasts.^[3]

4Agency, adaptation and the choice to stay

The IPCC’s reading of the evidence is that migration is among a wider set of possible adaptation alternatives and often emerges when other forms of adaptation are insufficient. Involuntary displacement occurs when adaptation alternatives are exhausted or not viable, and reflects non-climatic factors that constrain adaptive capacity and create high exposure and vulnerability. The more agency migrants have, in terms of voluntariness and freedom of movement, the greater the potential benefits for sending and receiving areas, while displacement or low-agency migration is associated with poor outcomes in health, well-being and economic security (high agreement, medium evidence).^[1]

Immobility needs equal attention. The IPCC notes that immobility in the context of climate risks can reflect vulnerability and lack of agency, but can also be a deliberate choice to maintain livelihoods and social and cultural attachments to place. Poorer households, which often live in highly exposed locations, are more likely to be forced into low-agency migration and are also the most likely to be immobile or trapped where migration would be the preferred response. At the same time, governments should be slow to label populations as trapped or to promote relocation without local agreement that in situ adaptation options have been exhausted.^[1]

Movement is not always a failure. IDMC observes that, when supported by effective monitoring, coordinated response and community preparedness, temporary displacement through evacuation can function as a protective measure that reduces human and economic losses.^[2]

5Cities, planning and policy

Much internal movement ends in cities, but the evidence base on urban destinations is thin. The IPCC notes that there is limited literature on mobility as adaptation in urban populations, with the focus on resettlement of flood-prone informal settlements within cities.^[1]

The Groundswell report recommends embedding internal climate migration in far-sighted green, resilient and inclusive development planning, preparing for each phase of migration so that internal climate migration as an adaptation strategy can result in positive development outcomes, and investing in better understanding of the drivers to inform well-targeted policies.^[3] The IPCC similarly judges that links between climate risks, adaptation, migration and labor markets highlight the value of providing better mobility options as part of transformative change (medium confidence).^[1]

6Implications for disaster risk management

Analysis. This section is Disaster Science’s interpretation of the evidence above. It is not drawn from the cited sources.

Plan for the common case, which is internal. Most movement stays within borders, so the planning burden falls on national and city governments: land-use plans, services and jobs in the places people go to, not only border policy.

Treat evacuation and displacement as one system. Temporary displacement is often how risk is managed. Evacuation routes, shelters and return support belong in the same plan as long-term relocation options.

Protect choice. The evidence points to better outcomes when people have agency. Programs should ask what would let people stay if they wish, and what would let them move safely, with documents, assets and legal standing intact, if they must.

Watch the people who cannot move. Poorer households are most likely to be stuck in deteriorating places. Risk assessments that count only those who leave will miss them.

Fix the data. Counts of movements, projections of future migrants and measures of immobility all rest on thin data, so monitoring should be built into programs from the start.

7Uncertainty and limitations

Counts are movements. IDMC’s figures record displacement events, many of them temporary, and exclude cross-border movements. IDMC also notes that data gaps affect some countries.^[2]

Projection assumptions. The 216 million figure is a scenario-based maximum, drawn from two reports with different regional coverage, not a forecast.^[3]

Evidence strength. Several IPCC statements on agency and outcomes carry high agreement but only medium evidence, and the literature on urban destinations and on immobility is limited.^[1]

Scope. This brief synthesizes an IPCC chapter, an IDMC report and a World Bank press release on a report that it did not read in full. It presents no original data and is an editorial synthesis that has not undergone peer review.