Climate Migration and Displacement

Core Concepts

A taxonomy of movement

Not all climate-driven mobility looks the same. The UNFCCC, in a formal decision at COP 16 (2010), established three distinct categories: displacement (predominantly forced), migration (predominantly voluntary), and planned relocation (which can be either). The distinction between forced and voluntary hinges on whether realistic options remain available to the person. If they do, the movement is migration; if they don't, it is displacement.

The IPCC AR6 adds further texture with four forms of climate-linked mobility: adaptive migration (individual choice), involuntary displacement (few or no options), organized relocation (state-planned), and — crucially — immobility, where people are unable or unwilling to leave high-risk areas despite significant exposure.

Slow-onset versus sudden-onset

The distinction between slow and sudden processes shapes almost everything about how displacement unfolds.

Sudden-onset events — hurricanes, floods, wildfires — produce movement that is typically short-distance, involuntary, and temporary. People displaced by a cyclone usually attempt to return home as soon as it is safe.

Slow-onset processes — sea level rise, desertification, salinization, glacial melt, increasing temperatures — produce movement that is more likely to be permanent, potentially cross-border, and large-scale. Rather than a single catastrophic event, slow-onset processes gradually erode livelihoods: first farmers shift from crops to livestock, then seek wage labor, then migrate. The pathway is cumulative and often irreversible.

The scale difference matters. Between 1979 and 2008, storms affected approximately 718 million people globally, while droughts affected approximately 1.6 billion. Slow-onset hazards are not the dramatic headline events, but they operate at an entirely different order of magnitude.

Despite this, research on slow-onset displacement remains predominantly qualitative and case-study based, with few comparative studies. National and international policies have consistently prioritized sudden-onset disasters, leaving slow-onset impacts under-studied and under-resourced.

Where people go

One of the most consistent findings across the literature is that climate-related displacement is predominantly internal — people move within their own countries, not across international borders. When cross-border movement does occur, it tends to be between neighboring countries. This pattern holds for both slow-onset and sudden-onset events.

Origins are disproportionately rural, with destinations either other rural areas or urban centers within the same country. Climate-related migration in low- and middle-income countries follows agricultural and water-scarcity logic — people leave where rain has stopped, and move toward where livelihoods remain viable.

The trapped population problem

A counterintuitive insight from research on climate migration: some climate impacts may actually reduce migration by destroying the resources needed to move. Poverty is the most commonly cited constraint on out-migration. In communities where incomes are already fragile, the up-front costs of relocation — transport, housing, establishing new livelihoods — may be prohibitive.

This creates "trapped populations": people who remain in high-exposure areas not because they choose to, but because they have no viable alternative. The immobility paradox is that the populations most exposed to worsening climate conditions are often the ones with the fewest resources to escape them.

Climate change may create a future with fewer mobile people despite worsening impacts — because it destroys the resources needed to move.

---

The Groundswell Numbers: What They Mean and What They Don't

The most widely cited global figure for climate-induced displacement comes from the World Bank's Groundswell reports: up to 216 million internal climate migrants across six regions by 2050 under pessimistic scenarios. The same model suggests effective mitigation and adaptation could reduce this figure by around 80 percent — to approximately 44 million. The range illustrates both the scale of the problem and the significance of policy choices.

How Groundswell works

The Groundswell methodology combines climate scenarios (RCP 2.6 for low emissions, RCP 8.5 for high emissions) with development scenarios (SSP2 for moderate growth, SSP4 for unequal development). It uses statistical models linking climate variables — temperature, precipitation — to agricultural outcomes and water availability, then estimates migration flows in response to livelihood pressures.

Fig 1

What Groundswell does not capture

The model has significant limitations that are essential to understand before citing its figures:

• Slow-onset drivers only. Groundswell models water availability, crop productivity, and sea-level rise. It does not capture rapid-onset events (acute flooding, hurricanes), extreme heat waves, or climate-conflict interactions — all of which can drive displacement.
• Internal migration only. The Groundswell analysis focuses exclusively on movement within national borders. Cross-border climate migration receives far less analytical attention due to methodological challenges.
• No equivalent of hind-casting. Climate models can be validated by testing whether they accurately retrodict past climate conditions. Climate migration models have no comparable validation framework. The numbers are conditional scenarios, not forecasts.

Under scenarios with weak mitigation and adaptation, other estimates put total climate-driven migration at up to 500 million by 2050, accounting for additional climate drivers not captured in Groundswell. The gap between 44 million and 500 million is itself a statement about uncertainty — and about what policy choices can prevent.

---

Annotated Case Study: The Syria Drought and the Limits of Climate Causation

No case has been cited more frequently in debates about the climate-conflict link than Syria. The narrative runs as follows: a severe drought from 2007 to 2010 caused agricultural collapse, drove rural-to-urban migration, created social tension, and contributed to the outbreak of civil war in 2011. It is a compelling story that offers apparent evidence for the entire climate-conflict chain.

The reality is substantially more contested.

What the initial research claimed. A 2015 study published in PNAS argued that the 2007–2010 drought was the worst on record, that century-long precipitation and temperature trends were consistent with anthropogenic climate forcing, and that the drought increased the probability of agricultural collapse and mass rural-to-urban migration.

What subsequent research found. A 2022 study in Nature Communications Earth found that Syrian croplands recovered quickly after the 2007–2009 drought period, contradicting claims of sustained agricultural collapse. The same research found no clear and reliable evidence that anthropogenic climate change was a factor in the pre-civil war drought. A UN University report concluded that systemic governance failures — not environmental stress — were the primary cause of rural livelihood collapse, and that the scale of migration attributed to the drought was substantially exaggerated in some academic accounts.

What this tells us about methodology. The Syria case illustrates several systematic problems in climate-conflict research. Studies tend to focus on African and Middle Eastern cases where researchers expect to find climate-conflict links — what scholars call the "streetlight effect". Many studies inadequately control for poverty, governance failures, and existing political tensions, attributing outcomes primarily to climate stress. The result is that causal claims about specific conflicts remain highly contested even when a climate signal can be identified.

Why this matters beyond Syria. Climate determinism — the reduction of conflict causation to environmental scarcity — can serve political purposes. Framing conflict as the inevitable result of drought deflects attention from governance failures and political choices. Some researchers have documented how authoritarian leaders invoke climate causation to escape accountability for mismanagement. This is not an argument against studying climate-conflict links — it is an argument for studying them carefully, with explicit attention to confounding factors and power dynamics.

The IPCC AR6 frames climate as a "threat multiplier" rather than a direct cause of conflict: climate stress interacts with weak governance, social divisions, and resource competition to amplify existing fragility. This is the position the UN Secretary-General has formally adopted: climate change does not directly or inevitably cause violent conflict, but its interaction with other factors can amplify conflict drivers in already fragile contexts.

---

Common Misconceptions

"Most climate migrants are crossing international borders to reach wealthy countries." The evidence is consistently the other way around. The IPCC concludes with high confidence that most climate-related displacement occurs within national borders. International movement, when it happens, is primarily between neighboring countries. The image of mass climate migration toward Europe or North America is not supported by current evidence or projections, though it drives a great deal of political anxiety.

"Climate refugees are a legally recognized category with international protections." The term "climate refugee" does not exist in international law. The 1951 Refugee Convention recognizes only those with a well-founded fear of persecution on five enumerated grounds: race, religion, nationality, political opinion, or membership of a particular social group. Environmental degradation is not persecution. Climate-displaced people fall outside this framework and lack the legal entitlements (right of non-refoulement, family unity, right to work) associated with refugee status.

"Climate change is driving a wave of migration from poor countries to rich ones." Climate-driven migration is predominantly from rural to rural or rural to urban areas within the same country. The ability to migrate — even within a country — requires resources that many of the most climate-vulnerable people lack. Poverty can actually suppress migration responses to climate stress, creating trapped populations in the most exposed places.

"If you don't want climate migration, the solution is better border control." The causes of climate displacement — livelihood erosion, failing water systems, crop failures — are not amenable to border control responses. Securitizing climate migration tends to produce exclusionary border-hardening policies rather than adaptive governance frameworks, while doing nothing to reduce the underlying drivers. Critics argue this framing is actively harmful to vulnerable populations and obscures state responsibilities toward climate-affected people.

"The 216 million figure is a prediction of what will happen." It is not. The World Bank Groundswell 216 million figure is a scenario estimate under pessimistic assumptions — high emissions, unequal development. The same models project roughly 44 million under optimistic scenarios with strong mitigation and adaptation. The gap between these figures reflects both genuine uncertainty and the significant role of policy in shaping outcomes.

---

Thought Experiment

A small island nation in the Pacific has been told by climate scientists that its entire territory will be uninhabitable within 30 years due to sea level rise and saltwater intrusion. The government has begun negotiations with a larger neighbor to relocate the entire population.

Consider the following questions — there are no single correct answers:

On identity and continuity: If a state loses its physical territory but retains its government, its people, and its international recognition (as Tuvalu has been working toward), does it remain a state? What does citizenship mean when there is no land to return to? Does nationhood survive relocation?

On consent and agency: The decision to relocate 11,000 people permanently involves trade-offs — community cohesion, access to traditional fishing grounds, familiar institutions — that no outside party can fully value on behalf of the population. What would a genuinely community-led relocation process look like? Who speaks for future generations who had no say in the original decision to stay?

On responsibility: The island nation has contributed negligibly to global greenhouse gas emissions. The country receiving the population is a significant historical emitter. Does this asymmetry create obligations beyond providing territory and citizenship? If so, what kind, and enforced by what mechanism?

On legal gaps: Under current international law, these people are not refugees, and no binding framework governs their relocation. If you were designing a new international legal instrument from scratch, what would you need it to do that existing frameworks cannot?