Ending Forced Displacement: A Plan, and Its Test in 2026

Introduction

At the end of 2025, 117.8 million people had been forced from their homes by war, persecution, violence and human rights abuses, about one person in every 70 alive (Al Jazeera). Most of them never crossed a border. Of those who did, seven in ten have been in exile for five years or more, and many families have now spent decades in camps or on the edges of foreign cities (UN Audiovisual Library).

Displacement belongs among the species-level problems because war, hunger, climate change and poverty all feed it and are fed by it. War produces most displacement; displaced families lose land, work and savings at once, and refugee camps are among the first places where rations are cut when aid budgets shrink. Climate change adds slower pressure through drought, floods and storms, and poverty decides who can move safely and who is trapped.

This essay sets out a plan to reduce the number of people living in displacement and to shorten the time they spend in it, then tests that plan against the events of 2025 and 2026. Education and the rights of women and girls are the subjects of separate essays and appear here only where they bear directly on displacement.

Where things stand

UNHCR’s annual Global Trends report, released on 10 June 2026, recorded the first fall in the global total in a decade (UN Audiovisual Library; Al Jazeera).

Measure (end of 2025 unless noted)	Figure
Forcibly displaced people	117.8 million
Internally displaced by conflict and violence	68.6 million
Refugees, including 6 million Palestinians under UNRWA’s mandate	41.6 million
Asylum seekers	9 million
Returns during 2025	14.7 million (4.4 million refugees, 10.3 million internally displaced)
Refugees in protracted exile (five years or more)	about 25 million, 70% of refugees
Share of refugees who are children	39%
Share of refugees hosted in low- and middle-income countries	nearly 68%
Stateless people	4.5 million

Sources: Al Jazeera; UN Audiovisual Library; L’Osservatorio; UN News.

The largest refugee populations come from Venezuela, Palestine, Ukraine, Syria, Afghanistan, Sudan and South Sudan, and the largest hosts are Colombia, Germany, Türkiye, Uganda, Iran, Chad and Pakistan (Al Jazeera).

Internal displacement is larger than cross-border flight. The Internal Displacement Monitoring Centre (IDMC) counted 82.2 million people living in internal displacement at the end of 2025. UNHCR puts those displaced by conflict and violence at 68.6 million, which leaves about 13.6 million displaced by disasters. The number has doubled since 2016. In 2025, conflict and violence caused a record 32.3 million new internal displacements, nearly 60% more than in 2024 and, for the first time, more than disasters caused. Disaster displacements fell 35% to 29.9 million after an exceptional 2024. Sudan had the most internally displaced people for the third year running (9.1 million), and Iran and the Democratic Republic of the Congo together produced two-thirds of 2025’s conflict displacements (NRC).

The fall in the headline total came from returns, the largest wave on record: 3.6 million to the Democratic Republic of the Congo, 3.6 million to Sudan, 3.3 million to Syria, 2 million to Afghanistan, 718,300 to Ukraine and 415,200 to Myanmar (Al Jazeera). Some reflect real improvement, as in Syria after December 2024; others reflect expulsion or desperation, as in Afghanistan and parts of Sudan.

Part I: The plan

The aim of the plan is to cut the time people spend displaced and the share of them who depend on aid, while keeping the right to seek safety intact. It has eight elements.

1. Fund emergency protection predictably

People who have just fled need shelter, water, food, health care and registration within days, and the agencies that provide them are funded late and one year at a time. The plan calls for multi-year, pre-committed funding for refugee emergencies and for the camps that already exist, with a share held in reserve and released automatically when early-warning systems or border monitoring show large movements. Cash should replace in-kind aid wherever markets function.

UNHCR asked for $8.5 billion for 2026 (Swissinfo). The UN’s system-wide humanitarian appeal for 2026 asks for $33 billion to reach 135 million people, out of 239 million in need (Concern Worldwide).

2. Give refugees the right to work, move and use national services

Most refugees live in countries that restrict where they can live and whether they can work. The plan asks host governments to grant work rights, freedom of movement, access to bank accounts and inclusion in national health and school systems, so that refugees earn and pay taxes instead of waiting for rations. The World Bank’s review of evidence from Colombia, Ethiopia, Jordan and Uganda finds that forced displacement has generally modest or positive effects on host labour markets (World Bank). Colombia, the largest of those four hosts and now the largest refugee host in the world, is the test case at scale (Al Jazeera). Kenya’s Shirika Plan, backed by the Refugees Act of 2021, sets out to replace the Kakuma and Dadaab camps, which have existed for more than thirty years, with settlements where refugees can work and live alongside Kenyans (USCRI; HIAS).

UNHCR has adopted a matching target. Its “50 by 35” goal is to halve, by 2035, the number of refugees in protracted displacement who remain dependent on humanitarian assistance, through jobs, education and lasting solutions (UN Audiovisual Library).

3. Pay the countries that host

Nearly 68% of refugees live in low- and middle-income countries and more than a quarter in the least developed ones (UN News; UN Audiovisual Library). Rich countries should pay for schools, clinics, water systems and jobs programmes that serve both refugees and their hosts, on grant or near-grant terms, and should tie the money to the policy changes in element 2.

The instruments exist. The World Bank’s Window for Host Communities and Refugees has grown from $2 billion in 2016 to at least $2.4 billion in the current IDA21 cycle, half of it as grants, for 19 eligible countries that host at least 25,000 refugees and keep an adequate protection framework (Center for Global Development). The Global Concessional Financing Facility has mobilised more than $1 billion for middle-income hosts, benefiting at least 6.6 million refugees and 8.4 million people in host communities (World Bank). The plan would roughly double these windows and restore the requirement that recipients make measurable reforms.

4. Make returns voluntary, safe and lasting

Return home is what most refugees want, and it is the solution that reduces displacement fastest. It lasts only when people go back voluntarily, to places where they can find housing, documents, water, schools and work. The plan calls for funding recovery in areas of return before and during large returns, for restoring civil documentation and property records, and for letting families visit before moving so they can judge conditions for themselves. It also calls for a firm refusal to push people back into danger. The principle of non-refoulement, which bars returning a refugee to persecution, is, in the words of UN High Commissioner for Refugees Barham Salih, “non-negotiable” (UN Audiovisual Library).

5. Expand resettlement and other safe routes

For refugees who can neither go home nor stay safely where they are, such as survivors of torture or people at risk in their host country, resettlement to a third country is the only option. UNHCR projects that 2.4 million refugees will need resettlement in 2027 (UN News). The plan asks rich and middle-income countries to meet a shared annual target of resettlement places, plus labour, study and family visas for refugees.

6. Run fast, fair asylum systems

People who ask for protection in a rich country should get a decision within months, with a proper hearing and legal advice; those who qualify should be allowed to work and settle, and those who do not should be returned with dignity. Slow systems leave people in limbo for years and erode public confidence. The plan asks for adequate staffing of asylum courts and agencies, a shared European system for distributing responsibility, and agreements with countries of origin on returns that respect human rights.

7. Prepare for climate displacement

Climate change mostly displaces people inside their own countries. The World Bank’s Groundswell study projected that up to 216 million people in six regions could be forced to move within their countries by 2050, and that rapid emission cuts and inclusive development could reduce the number by as much as 80% (World Bank). The plan calls for adaptation funding aimed at the places where livelihoods are failing, planned and voluntary relocation where land will become uninhabitable, and regional agreements that let people move for work before disaster strikes. The Falepili Union between Australia and Tuvalu, in force since August 2024, offers a pathway for 280 Tuvaluans a year to migrate to Australia (Wikipedia).

8. Prevent displacement at its source

Conflict now causes most new displacement, so the most effective long-run measure is fewer and shorter wars, the subject of the companion essay on war. Within the displacement field, the plan asks for better data, since IDMC reports that data availability declined in 15% of the countries it monitors in 2025 (NRC), and for sanctions and prosecutions that treat the forced displacement of civilians as the crime it is under international law.

Sequencing

In the first two years, the priorities are emergency funding, protection of rations and health services in the largest camps, and support for the returns already under way in Syria and Sudan. Over two to ten years, the work shifts to work rights and inclusion, host-country financing tied to reform, and the expansion of resettlement. Over a generation, the aim is short displacement, fewer wars and adaptation that keeps most people on their land.

What makes it stick politically

Host countries accept inclusion when it brings money and when their own citizens use the new schools, clinics and roads. Rich countries accept resettlement and asylum when the systems are orderly and arrivals are seen to work. The plan therefore pairs each obligation with a benefit: development finance in return for work rights, and controlled legal routes in return for donor contributions. Refugee participation in designing programmes, which the Kenyan experience shows is often missing, helps the arrangements last (USCRI).

Where reasonable people disagree

People disagree about how much rich countries owe. Some argue that a country’s duty ends with funding help in the region of origin; others hold that the Refugee Convention obliges every signatory to hear claims made at its border. They disagree about offshore processing and the use of “safe third countries,” which governments present as a way to break smugglers’ business models and critics see as outsourcing responsibility to poorer states. They disagree about when return is safe enough to encourage: aid agencies tend to wait for stable conditions, while host governments facing economic strain and public pressure want returns sooner. They also disagree about camps, which concentrate services in the first months of a crisis but trap people when they last for decades.

Part II: The plan in the world of September 2026

Emergency funding: the aid cliff reaches the camps

The first element of the plan is failing most visibly. Global humanitarian funding fell about 30% between 2024 and 2025, and US support fell from $14 billion to $3.7 billion after the administration’s 2025 review of foreign funding (InfoMigrants). The United States had provided about 40% of UNHCR’s annual budget, and by July 2025 the agency had halted or suspended $1.4 billion in programmes, warning that up to 11.6 million refugees could lose access to aid (Al Jazeera). It cut nearly 5,000 jobs in 2025 and scaled back 185 offices (Swissinfo).

The cuts deepened in 2026. Against a request of $8.5 billion for the year (Swissinfo), in May 2026 Salih told member states the agency expected just over $3 billion for the year, about 15% less than in 2025 (CNBC Africa). In late August the agency set a 2027 budget of $7.14 billion, the smallest in at least a decade, and announced 3,500 more post cuts, with double-digit reductions in the Americas, Asia-Pacific and West and Central Africa (Brussels Signal).

The effects show up in rations. From April 2026, WFP moved the 1.2 million Rohingya refugees in Bangladesh from a flat $12 a month to a tiered system in which about 17% receive $7; programmes for the Rohingya were 19% funded (Courthouse News). When rations fell to $8 in 2023, WFP recorded 12% of children in the camps acutely malnourished even before the cut took effect (WFP). The 2025 cuts also suspended education for 230,000 Rohingya children and reduced gender-based violence services by a quarter (Al Jazeera), losses that the companion essays on education and on women and girls take up. The starvation essay records similar cuts for refugees in Jordan and South Sudan.

In 2026 refugee aid rests on discretionary money in steep decline, and agencies are rationing by triage, which makes the plan’s call for pre-committed funding the gap to close first.

Inclusion: the right policies, without the money to carry them

Element 2 has the most support in principle and the least money in practice. Kenya’s Shirika Plan is the clearest case. A year after its launch, a USCRI assessment in March 2026 credited it with better coordination between national and county governments, but found that for hundreds of thousands of refugees “daily realities have yet to change,” largely because food and service cuts undercut the transition (USCRI).

UNHCR’s “50 by 35” target is the right ambition, since roughly 25 million refugees are in protracted exile (UN Audiovisual Library). Self-reliance saves aid money over time but needs up-front spending on permits, training, schools and infrastructure, which is being cut, and there is a risk that “self-reliance” becomes a label for withdrawing aid before refugees can support themselves.

Host-country financing: intact, with weaker conditions

Development finance has held up better than humanitarian aid: the World Bank’s refugee window survived the IDA21 replenishment at $2.4 billion. As part of a simplification of IDA, however, the requirement that at least 60% of eligible countries make significant policy reforms was dropped, and 13 of 17 eligible countries examined still significantly restricted refugees’ access to labour markets in 2022 (Center for Global Development). The money has survived, but its link to work rights has weakened.

Returns: Syria and Sudan test element 4

Syria. After the fall of the Assad government on 8 December 2024, more than 1.5 million refugees and 1.8 million internally displaced Syrians had returned home by March 2026 (UNHCR Operational Data Portal). UNHCR expected about one million more refugee returns in 2026. More than a quarter of returnees lacked basic documents such as identity cards and property deeds (Syrian Observer), and 15.6 million people in Syria still needed assistance, with 5.5 million people still internally displaced (UNHCR Operational Data Portal). These are the conditions element 4 was written for: large voluntary returns into a country whose housing, water and services cannot yet absorb them, at a time when WFP has halved emergency food assistance inside Syria, as the starvation essay records.

Sudan. Sudan’s war, which began on 15 April 2023, displaced nearly 12 million people at its peak and pushed about 4.5 million across its borders (UN News). By September 2026, IOM counted 5.1 million returns to 76 localities, 82% of them people who had been displaced inside Sudan and about 930,000 from abroad, half of those from Egypt. About 8.6 million people remained internally displaced (Darfur24). Returns are concentrated in Khartoum and Al Jazirah, where IOM found homes, water, health and electricity systems heavily damaged, and some families return because life in displacement has become unbearable (UN News). Meanwhile, violence in Darfur and Kordofan keeps displacing people: the fighting at El Fasher in October 2025 displaced more than 100,000 (IOM). Recovery money is not following the returnees.

Afghanistan. Afghanistan is the counter-example. Between mid-September 2023 and 30 May 2026, about 6.04 million Afghans returned from Iran and Pakistan, driven largely by Pakistan’s deportation plan and Iranian enforcement, and many arrived with no housing, work or documents in a country some had never lived in (IOM DTM). In 2026 alone, 1.23 million had returned by mid-September, of whom 506,410 were deported, 439,000 of them from Iran (UNHCR Operational Data Portal). They count toward the fall in the global total, yet they are forced returns of the kind element 4 rules out.

New displacement in 2026: the Iran war

The war that the United States and Israel began against Iran on 28 February 2026 produced the year’s largest new displacement. By late March, about 3.2 million people had been displaced inside Iran, mostly from Tehran and other cities, and more than a million in Lebanon, nearly one Lebanese in five, as fighting between Israel and Hezbollah intensified; more than 125,000 people crossed from Lebanon into Syria by 17 March (Al Jazeera). UNHCR described much of the displacement in Iran as temporary, and the war has since passed through an April ceasefire, a June memorandum that collapsed on 8 July, and renewed hostilities in September (Wikipedia), so current figures for Iran are uncertain.

The war also struck the world’s largest refugee population. Iran hosts more than 4.4 million Afghans, about 2.4 million of them in Tehran, and more than a million were at risk of deportation by June. NRC reported that the humanitarian response in Iran was 47% funded and the 2025 Afghan refugee response only 18% funded (NRC). The same war drives the Hormuz fertilizer shock described in the starvation essay, so the Afghans it pushes home arrive in a country already short of food.

Resettlement has fallen furthest

Element 5 has gone backwards further than any other. Only about 37,000 refugees left for resettlement through UNHCR-supported programmes in 2025, against more than 116,000 in 2024 and needs measured in millions, and the international community is off track to reach its target of 130,000 places a year by 2027 (UN News). The largest single cause is the United States. Its ceiling for fiscal 2026 was first set at 7,500, then raised to 17,500 by an emergency determination signed on 21 May 2026 that cited violence against Afrikaners in South Africa (American Presidency Project). Through 31 July, the United States had admitted 10,258 refugees in fiscal 2026, of whom 10,255 were South African, compared with 38,102 admitted in fiscal 2025 (Pew Research Center).

Asylum policy: stricter rules in Europe and the United States

Element 6 is being tested by two large policy shifts. The EU’s Pact on Migration and Asylum became fully applicable on 12 June 2026. It introduces seven-day screening at external borders, a mandatory border procedure of up to 12 weeks for nationals of countries with EU-wide recognition rates of 20% or less, and a solidarity mechanism under which member states choose between accepting relocations and making financial or in-kind contributions (European Commission; Human Rights Watch). It also guarantees free legal counselling at every stage. Human Rights Watch warns that it allows detention for up to six months, widens the “safe third country” concept to countries with which applicants have no connection, and ends the automatic suspension of deportation during appeals; it notes that Poland’s suspension of asylum from March 2025 was still in force more than a year later (Human Rights Watch).

The Pact contains the plan’s main elements, faster decisions and a shared European mechanism, along with the risks the plan warns about. Whether the fast procedures produce fair decisions, and whether solidarity moves people as well as money, cannot yet be judged.

In the United States, temporary protected status has been ended for several nationalities, including about 352,000 Venezuelans, with the Supreme Court allowing that termination in October 2025. On 25 June 2026, in Mullin v. Doe, the Court ruled 6 to 3 that the statute bars judicial review of non-constitutional challenges to terminations, clearing the way to end protection for about 330,000 Haitians and several thousand Syrians. About 1.3 million people from 17 countries held the status (Wikipedia: Temporary protected status; Wikipedia: Mullin v. Doe). Haiti’s designation was set to end on 27 July 2026 and Syria’s on 20 July.

Climate displacement: quieter year, same trend

Disasters displaced fewer people inside their countries in 2025 than in 2024, 29.9 million against an exceptional prior year, with storms a leading cause and wildfires causing 694,000 displacements (NRC; L’Osservatorio). The next test is the strong El Niño described in the starvation essay, which threatens drought across southern Africa from late 2026, and element 7’s call for anticipatory action applies directly. Mobility agreements such as the Falepili Union remain small and rare.

Prevention: the wars continue

Element 8 depends on the war plan, and the record there is poor. IDMC’s director called 2025’s conflict displacement numbers the most “staggering” it had ever recorded, and NRC’s Jan Egeland described internal displacement on this scale as “a sign of a global collapse in prevention” (NRC). The wars in Sudan, the Democratic Republic of the Congo and Ukraine continued through 2026, and the Iran war added a new one.

Part III: Revised priorities for the next 6 to 12 months

Protect the camps from the aid cliff. Restore food and health budgets for the largest protracted populations, starting with the Rohingya in Bangladesh, Afghans in Iran and Sudanese refugees in Chad and Egypt, and give UNHCR and WFP multi-year funding commitments so that cuts are not made month by month.
Put recovery money behind the returns already under way. Fund housing, documents, water and livelihoods in the areas of Syria and Sudan receiving the most returnees, so that 2025’s record returns do not become 2027’s renewed displacement.
Press Iran and Pakistan to halt mass deportations to Afghanistan and fund reception for those already returned, treating this as a hunger emergency as well as a protection one.
Restore resettlement capacity. Other resettlement countries should expand quotas to offset the fall in US admissions, with priority for Afghans, Sudanese and Rohingya.
Make the EU Pact work as designed. Staff the border procedures to produce fair, reviewable decisions, and make relocation of people part of the solidarity mechanism alongside money.
Reattach conditions to host-country finance. Use the World Bank refugee window and bilateral aid to reward work rights and freedom of movement, supporting plans such as Kenya’s Shirika with the money needed to deliver them.
Plan for El Niño displacement in southern Africa alongside the food response, including pre-positioned shelter and cash for families who move in search of water and work.

Conclusion

The fall in global displacement in 2025 was real and welcome, and much of it came from Syrians and Sudanese choosing to go home. It also rested on large numbers of Afghans who had no choice, and on returns into damaged cities that may not hold. The Iran war has since displaced millions more, while global humanitarian funding fell by about 30% between 2024 and 2025 and UNHCR expects 15% less in 2026.

Tested against these events, the plan’s emphasis on inclusion and development finance looks sound, since those tools depend less on emergency budgets and give refugees a way out of dependence. The weakest parts are emergency funding and resettlement, where political decisions in donor countries have cut capacity sharply since 2024, and prevention, which depends on ending wars that continue. The largest gain available in the next year is to make the returns to Syria and Sudan last, because those two countries account for millions of the people who went home in 2025 and 2026.

Several situations described here, including the Iran war and returns to Sudan, are changing quickly; this essay reflects reporting available at the end of September 2026.