How we measure returnee pressure
What "returnee" means here
Returnees are Afghans who came back from abroad, documented or undocumented, between 2021 and 31 October 2026. People displaced inside Afghanistan are not included, because the 2026 source does not record them. Population figures are IOM's, which already include returnees who arrived by the end of 2025; 2026 arrivals are added on top. The share of residents therefore answers a simple question: of the people living in a district today, how many came back from abroad since 2021? We do not subtract population outflows from each district, because such data are not available.
Counting returnees
IOM's mobility data place 5,800,123 returnees from abroad in Afghanistan's districts between 2021 and the end of 2025[2]. The 2026 border response dashboard records 1,064,441 undocumented and 323,371 documented returnees between January and October 2026[3], and gives the intended province and district of return of undocumented returnees. Documented returnees are placed by the province pattern of UNHCR's assisted returns[4]. Within each province, 2026 documented arrivals are spread across districts in the same proportions as 2021 to 2025 arrivals. The two periods do not overlap, so they can be added.
Two kinds of pressure
The share of returnees in a district describes social and economic absorption, that is, whether there are enough jobs, rentals, social services and aid to accommodate the returnees. It says little about physical or spatial crowding. To capture spatial crowding, this report divides returnees by the area of inhabited land in each district, measured from the 1-km population grid built for ARMO 2026[5,11]. Only land where people actually live is counted, so large empty deserts and mountains do not dilute the figure. The two measures are only weakly related (rank correlation 0.30 across ranked districts), which is why both are needed.
Share of residents = returnees 2021 to Oct 2026 ÷ (IOM population 2025 + returnees 2026) Spatial crowding = returnees ÷ km² of inhabited land Days needed = poverty line per person × 6.1 ÷ unskilled daily wage Days of wage for rent = monthly rent ÷ unskilled daily wage
What this report does not claim
The analysis is descriptive. It does not estimate the effect of returns on prices, wages or services, and a district's position in a ranking is not a judgment about its administration. Where data are weak, the report says so and does not fill the gap with assumptions presented as facts.
Methods and assumptions
| Item | Assumption or method |
|---|---|
| Returnees 2021 to 2025 | IOM mobility dashboard, round 3, column "returnees from abroad, 2021 to 2025", by district[2]. Internally displaced people and IDP returnees excluded. |
| Undocumented returnees 2026 | 1,064,441 undocumented returnees allocated by the intended province and district shares on the border dashboard[3]. Province totals are rounded to whole people so that they sum exactly to the national total; district totals sum exactly to their province. |
| Documented returnees 2026 | 323,371 documented returnees allocated by each province's share of UNHCR assisted returns, 1 January to 7 November 2026[4], then split across districts by each district's share of 2021 to 2025 returnees. |
| Residents | IOM total population for 2025 (which includes earlier returnees) plus 2026 arrivals. IOM's national total of 51.5 million is higher than official estimates; using it gives conservative (lower) shares. Population outflows are not subtracted because no such data exist by district. |
| Inhabited land | Area of populated cells in the Kontur population grid used for ARMO 2026 (H3 cells, about 0.82 km² each)[5,11], assigned to districts by geoBoundaries district boundaries[10]. Three new districts without boundaries (Chinarto, Patoo, Sharak-e-Hayratan) have no crowding value. |
| Typology | Above or below the national share (13.6%) and the national returnees per km² of inhabited land (60). Rankings use districts with at least 20,000 residents (377 of 401). |
| Health capacity | MoPH registry export of 9 November 2026[6]. Facilities with at least one recorded health worker, excluding 202 private outpatient clinics: 3,446 public facilities and 28,576 health workers. Registry district names matched to IOM names; new sub-districts such as Marjah, Gereshk, Babaji, Takhta Pul and Kshata Maiwand assigned to their parent district, and 118 facilities in other new sub-districts counted at province level only. |
| Facilities opened | Registry column "opened in last 5 years" for the same facilities. Closed facilities have no recorded staff and so fall outside the sample; closures are therefore not measured. |
| Earning capacity | Poverty line August 2026[8], urban and rural lines weighted by the urban population share in the ARMO grid; WFP unskilled non-agricultural daily wage, August 2026 province median[7]. Household size 6.1 from IOM[1]. A full working month is taken as 26 days. |
| Rent | NSIA average monthly rent for a four-bedroom house, August 2026, concrete and non-concrete, 20 provinces[9]; the other 14 provinces take the average of the 20. |
| Children | National age shares from IOM's survey (38% aged 6 to 17, 19% under five) applied to every district[1]. No reliable national list of schools by district is openly available[13], so school capacity is not measured. |
| Health benchmark | WHO benchmark of 44.5 doctors, nurses and midwives per 10,000 people[12], used for context only. |
| Stress test | A further 1,000,000 returnees distributed by the 2026 province and district pattern. "Extra workers needed to hold ratio" = added returnees ÷ residents × current health workers; facilities likewise. Households = added returnees ÷ 6.1. |
| Limitations | Dashboard values were transcribed from screenshots and checked against all province and national totals. All indicators are descriptive and make no causal claim. Health worker counts may be incomplete. School capacity is not measured. |
Bibliography
- [1] International Organization for Migration (2026). Returnee Resilience Overview, Afghanistan, Round 3. Kabul: IOM Displacement Tracking Matrix. https://dtm.iom.int/afghanistan [a]
- [2] International Organization for Migration (2026). Afghanistan Mobile Population Dashboard, Round 3. Kabul: IOM Displacement Tracking Matrix. https://dtm.iom.int/online-interactive-resources/afghanistan-mobile-population-dashboard [b]
- [3] International Organization for Migration (2026). Afghanistan Returnee Border Response Dashboard. Kabul: IOM Afghanistan. https://app.powerbi.com/view?r=eyJrIjoiZjQyYzRiOGEtNzdhMC00NjRmLTliMWUtMDI5Zjg1YmM4MzU4IiwidCI6IjE1ODgyNjJkLTIzZmItNDNiNC1iZDZlLWJjZTQ5YzhlNjE4NiIsImMiOjh9 [c]
- [4] United Nations High Commissioner for Refugees (2026). Afghanistan: Assisted Returnees by Province of Destination, 2026. Kabul: UNHCR. https://data.unhcr.org/en/country/afg [d]
- [5] Wahdat, A. Z. (2026). Afghanistan Rural Marketplace Opportunity (ARMO) 2026. Data for Afghanistan. https://dataforafghanistan.org/armo [e]
- [6] Ministry of Public Health (2026). National Health Facilities Registry Dashboard. Kabul: MoPH. https://dhismfl.moph.gov.af/ [f]
- [7] World Food Programme (2026). Afghanistan – Food Prices. Humanitarian Data Exchange. https://data.humdata.org/dataset/wfp-food-prices-for-afghanistan [g]
- [8] Data for Afghanistan (2026). Afghanistan Poverty Lines for August 2026. https://dataforafghanistan.org/reports/2026/aug/Afghanistan-Poverty-Lines-August-2026-DfA.pdf [h]
- [9] National Statistics and Information Authority (2026). National Consumer Price Index, August 2026. Kabul: NSIA. [i]
- [10] Runfola, D. et al. (2020). geoBoundaries: A global database of political administrative boundaries. PLoS ONE 15(4): e0231866. https://www.geoboundaries.org [j]
- [11] Kontur (2023). Afghanistan: Population Density for 400m H3 Hexagons. Humanitarian Data Exchange. https://data.humdata.org/dataset/kontur-population-afghanistan
- [12] World Health Organization (2016). Global Strategy on Human Resources for Health: Workforce 2030. Geneva: WHO. https://www.who.int/publications/i/item/9789241511131
- [13] Humanitarian OpenStreetMap Team (2026). Afghanistan Education Facilities (OpenStreetMap Export). Humanitarian Data Exchange. https://data.humdata.org/dataset/hotosm_afg_education_facilities [k]
Notes
- [a] Published September 2026. Household and individual indicators (household size 6.1; 58 percent under 18, of whom 38 percent of all returnees are aged 6 to 17 and 19 percent under five; employment, shelter, documentation and service access) come from this survey of returnees who arrived between July and December 2025. They are applied nationally and are not province estimates.
- [b] Round 3 data were collected in November and December 2025; the total population view by province and district was used. Values for all 34 provinces and 401 districts were transcribed from dashboard views, because the dashboard does not offer a download. District sums were checked against province rows and the national total, and all checks matched exactly. Only the column "returnees from abroad, 2021 to 2025" is used; internally displaced people and IDP returnees are excluded because the 2026 source does not cover them.
- [c] Filtered to 1 January to 31 October 2026. The dashboard reports 1,387,812 returnees for the period: 1,064,441 undocumented and 323,371 documented. Undocumented returnees from Pakistan draw on IOM Afghanistan from January 2025 and IOM Pakistan for September 2023 to December 2024; undocumented returnees from Iran draw on IOM Afghanistan; documented returnees draw on UNHCR Afghanistan. The dashboard gives the intended province and district of return of undocumented returnees as percentages to two decimals; these were applied to 1,064,441, with province totals rounded so that they sum exactly to the national figure.
- [d] Assisted returnees by province of destination and month, 1 January to 7 November 2026, from Pakistan, Iran and Tajikistan. The file covers 205,816 assisted returnees, fewer than the 323,371 documented returnees on the border dashboard. Only the provincial distribution is used: each province's share of assisted returnees is applied to the 323,371 documented total.
- [e] The ARMO population grid distributes official provincial population totals over Kontur population cells (about 0.82 km² each). Here it is used only to measure inhabited land and the urban share, never as a population count, so the difference between IOM and official population totals does not affect the ratios.
- [f] Master Facility List: Facilities, History and Workforce tables; export of 9 November 2026, latest history snapshot 22 September 2026. Facilities are those with at least one recorded health worker, excluding 202 private outpatient clinics: 3,446 public facilities from regional hospitals to sub health centers, family health houses and mobile teams, with 28,576 health workers. A further 3,886 active facilities have no recorded staff (mostly private laboratories, hospitals and clinics, but also some public units) and are excluded. Because closed facilities carry no staff, the sample contains openings but no closures. Registry district names were matched to IOM district names; 118 facilities in newly created sub-districts with no clear parent are counted at province level only.
- [g] Includes daily wage rates for non-qualified non-agricultural labor and qualified labor by province, with data to 15 October 2026. The August 2026 province median of the daily wage for non-qualified, non-agricultural labor is used, to match the month of the poverty line, the inflation data and the rent data.
- [h] Dataset prepared by Data for Afghanistan: poverty lines by province and urban or rural area at August 2026 prices (food, non-food and total, Afghanis per person per month), updated from the 2019 to 2020 base lines with provincial food and non-food price indices. The food line is 1,669 AFN per person per month everywhere; non-food lines differ by province and by urban or rural area. Seven provinces have rural lines only. A province line is the urban and rural lines weighted by the urban share of population in the ARMO population grid.
- [i] Provincial inflation and rent data are all taken from the August 2026 National CPI. The release gives year-on-year inflation for August 2026, national by item and for 20 provinces, and average prices by province, including monthly rent for a four-bedroom house. Fourteen provinces have no provincial index; for them the national rate of 7.8 percent is shown and marked as national. Housing (12.1 percent) and health (19.0 percent) inflation are national only. Rent for a four-bedroom concrete and non-concrete house is published for the same 20 provinces; the other 14 provinces are given the average of the 20 (6,439 Afghanis a month non-concrete, 9,210 concrete).
- [j] Afghanistan ADM1 (province) and ADM2 (district) boundaries, gbOpen release.
- [k] Snapshot of 2 November 2026, 3,459 features, ODbL license. Reviewed and not used for measurement. The 3,459 features are a small fraction of the country's schools and are unevenly distributed (one district holds 204).