Homelessness has risen in most countries that measure it and more than 1.16 billion people live in slums or informal settlements, while approaches known to help people find and keep a home, such as Housing First and emergency rent help, reach too few.
Updated October 14, 2026. Written from the sources listed below, not independently peer-reviewed.
In 2025, the homeless services system in Los Angeles moved 23,532 people into permanent housing. For every person it housed, roughly how many newly homeless people entered the system?
2.6 million+
People reported as homeless across OECD countries
At least 2.6 million people were reported as homeless or in housing exclusion under national definitions across the OECD, more than 1.4 million of them in the EU. The OECD compiles data for 42 countries; each uses its latest year, from 2020 to 2026 (for most, 2024 or 2025), and definitions differ, so this is not a global total.
Data year 2026. Source: OECD, 2026
745,652
People experiencing homelessness in the US on one night
Counted in January 2025, 3.3% fewer than the record count of 2024. 266,320 people, about 36%, were unsheltered. The count leaves out people staying with family or friends, and HUD notes that the unsheltered number could be larger than reported.
Data year 2025. Source: US Department of Housing and Urban Development, 2026
1.16 billion
People living in slums or informal settlements
More than 1.16 billion people in 2024, over 130 million more than in 2015. UN-Habitat estimates that at least 3 billion people live in inadequate housing. These figures describe inadequate housing, not homelessness.
Data year 2024. Source: UN-Habitat, 2026
Unit: people
HUD point-in-time counts as compiled in the OECD Affordable Housing Database (table HC3.1.A2, columns 'around 2018', 'around 2023' and '2026 or latest year', which is 2025 for the US). They cover sheltered and unsheltered people on one night and exclude people staying with family or friends. The table does not include 2024, which HUD reports as the record year; the 2025 count was slightly lower.
SDG target 11.1 aims to ensure access for all to adequate, safe and affordable housing and basic services and to upgrade slums by 2030. Its indicator counts people living in slums, informal settlements or inadequate housing; there is no global indicator for homelessness itself. UN-Habitat reports that more than 1.16 billion people lived in slums and informal settlements in 2024, over 130 million more than in 2015, and that housing is becoming less affordable: the global ratio of house prices to incomes rose from 9.3 in 2010 to 11.2 in 2023. UN-Habitat judges that improvements under SDG 11 are not keeping pace with rapid urban growth or reaching all communities equally.
In the US, 155,750 people experiencing homelessness on their own, one in three, had chronic patterns of homelessness in 2025, a number 81% higher than in 2013. Almost half (49%) of people homeless on their own slept unsheltered, while 93% of people in families with children were in shelter.
Source: US Department of Housing and Urban Development, 2026In Los Angeles County, 59% of newly unhoused people questioned in the 2026 count surveys named economic hardship as the main cause of their homelessness. It takes an income of US$110,000 a year to afford the county's average rent of US$2,745 a month.
Source: Los Angeles Homeless Services Authority, 2026Women make up on average about 27% of people counted as homeless in countries with point-in-time data, but they are more likely to stay temporarily with family or friends, a form of hidden homelessness that official statistics often miss.
Source: OECD, 2026UN-Habitat finds that the people most affected by inadequate housing often face several barriers at once, linked to income, age, disability, tenure status, displacement, informality, migration, gender, ethnicity or Indigenous identity.
Source: UN-Habitat, 2026Each country is shaded by its latest available value, so you can see where the gap is widest. Hover or tap a country, search for one, or open the full table.
Loading country values…
Values in % of urban population.
No comparable global count of homelessness exists, so this map shows a related housing-deprivation measure: the share of city residents living in households that lack improved water, improved sanitation, enough living space, durable housing or secure tenure, as estimated by UN-Habitat. It is not a count of homelessness: countries with many people experiencing homelessness, such as the United States, show values near zero. Values are the latest year available per country, 2022 for most and 2018 to 2020 for a few; values older than 2018 are not shown, and several countries share identical values, so treat small differences with caution.
Source: UN-Habitat Urban Indicators Database, via World Bank Open Data. Latest year available per country; years differ.
Values load from World Bank Open Data. Simplified Natural Earth borders; they imply no position on disputed territories.
Instead of one “percent solved” score, DO GOOD rates three separate dimensions. A problem can have mature solutions that still reach too few people. These ratings are an editorial reading of the sources on this page, not a measured index.
Has the situation improved in the real world?
LowBetween 2023 and 2026, the reported number of people experiencing homelessness rose in 15 of the 20 countries with comparable data and stayed stable or fell only in Canada, Japan, Korea, Latvia and Cyprus. The number of people living in slums is also still rising.
How mature and validated are the available solutions?
ModerateHousing First is well tested for its main aim: a review of 26 studies from the US and Canada, including randomised trials, found that it improved housing stability by a median of 41% across six studies compared with programmes that required treatment first. Its effects on physical and mental health were similar to those of usual services, and all included studies came from North America.
How widely are effective solutions actually used?
LowThe US had 408,167 permanent supportive housing beds in 2025, yet 266,320 people slept unsheltered on the count night and the number of individuals with chronic patterns of homelessness was 81% higher than in 2013.
There is no internationally agreed definition of homelessness, and differences in definitions, counting methods, coverage and frequency limit comparisons between countries. Even detailed counts such as the US point-in-time count leave out people staying with family or friends and may miss people sleeping outside.
Reported homelessness rose in 15 of 20 countries with comparable data between 2023 and 2026. In the US the count fell slightly in 2025 after a record 2024, largely because fewer people stayed in emergency shelter, while the number of people homeless on their own reached a new high; in Los Angeles County the count rose slightly again in 2026.
Housing is an established social determinant of health. Homelessness has no internationally agreed definition, and countries count it in very different ways, so there is no reliable global total. Across OECD countries, at least 2.6 million people were reported as homeless or in housing exclusion under national definitions, and between 2023 and 2026 (or the nearest years available) the reported number rose in 15 of the 20 countries with comparable data. Far more people have shelter that is not adequate: UN-Habitat estimates that more than 1.16 billion people lived in slums and informal settlements in 2024, over 130 million more than in 2015, and that at least 3 billion people live in inadequate housing.
In the United States, 745,652 people experienced homelessness on a single night in January 2025, and 266,320 of them spent the night in places not meant for sleeping, such as cars, parks, abandoned buildings or stations. The total was 3.3% below the record of 2024, but the number of people experiencing homelessness on their own, outside a family with children, was the highest since data collection began, and one in three of them had a disability and had been homeless for a year or more or repeatedly. In Los Angeles County, the count rose by 1.2% in 2026 to an estimated 73,040 people, 49,002 of them unsheltered, after two consecutive years of decline; LAHSA notes that the change is not statistically significant. Poor health, including mental illness and substance use disorders, is both a cause and a consequence of homelessness, and people experiencing homelessness have high rates of emergency department use and hospital stays.
There is no internationally agreed definition of homelessness, and statistical definitions vary widely between countries. The OECD sorts national data using the European ETHOS Light typology, the United States counts people on a single night each January, and the UN's housing target measures slums and inadequate housing, which is a related but different problem:
Living rough and in emergency accommodation
The narrow meaning of homelessness covers people living rough in streets or public spaces, people in emergency accommodation such as overnight shelters, and people in accommodation for homeless people such as temporary or transitional housing (ETHOS Light categories 1 to 3). The wider term housing exclusion adds people leaving institutions without a home, people in non-conventional dwellings and people staying temporarily with family or friends because they have no housing (categories 4 to 6).
Hidden homelessness
Many people without a home are not in any count: people who stay temporarily with family or friends, who avoid visible public spaces, or who are not in contact with support services. There is no formal definition of hidden homelessness, and countries use the term in different ways. Women are more likely to rely on informal arrangements of this kind.
Point-in-time counts in the United States
HUD's annual count gives a one-night estimate in the last week of January. People are sheltered if they stay in emergency shelters, transitional housing or safe havens, and unsheltered if they sleep in a place not meant for sleeping, such as a car, park, abandoned building or station. People staying with family or friends are not included. Individuals have chronic patterns of homelessness if they have a disability and have been homeless for a year or more, or at least four times in three years for a total of at least 12 months.
Slums and inadequate housing
UN-Habitat counts an urban household as a slum household if it lacks one or more of improved water, improved sanitation, sufficient living space, durable housing or secure tenure; SDG indicator 11.1.1 also considers housing affordability. People in slums usually have a dwelling, but not an adequate one, so this measure is not a count of homelessness.
What works
Housing First offers people with disabilities who are homeless a permanent, subsidised home with support services, without first requiring psychiatric treatment or sobriety; treatment is encouraged but not required. A CDC Community Guide review of 26 studies with 17,182 participants compared it with programmes that required people to be 'housing ready' first: across six studies, Housing First improved housing stability by a median of 41%, and in one study it cut the time people spent homeless by 88%. In the largest randomised trial, Canada's At Home/Chez Soi with 2,148 people with serious mental illness, participants were more than twice as likely as those in usual care to report positive changes in their lives.
Limits
Compared with usual services, the programmes produced similar changes in physical and mental health, and results for substance use were mixed. Most studies followed people for two years or less, all came from the US or Canada, and none covered rural programmes. Programmes also depend on a supply of homes that can be rented at an affordable cost.
What works
Shelters, transitional housing and safe havens give people immediate safety off the street. The US had 405,028 year-round emergency shelter beds in 2025, 91% more than in 2007, and 64% of the people counted in January 2025 were sheltered. Families with children are mostly reached: 93% were in shelter on the count night.
Limits
A shelter bed does not end homelessness, and people can stay for long periods when there is no housing to move on to. People experiencing homelessness on their own are often left out: 49% of them were unsheltered in January 2025.
What works
Short-term rent subsidies help people move quickly from the street or a shelter into an ordinary rental home. In Los Angeles, 10,308 people were newly housed with time-limited subsidies in 2024, the year the region moved a record 28,625 people into permanent housing.
Limits
The approach depends on steady funding and on rents that people can later pay themselves. When funding cuts paused new enrolments, the number housed with these subsidies in Los Angeles fell to 6,590 in 2025, and total placements into permanent housing fell to 23,532.
What works
One-time payments for back rent or other housing costs can keep households at imminent risk in their homes. In a randomised trial in Santa Clara County, California, in which help was allocated by lottery between July 2019 and December 2020, 0.9% of people who received emergency financial assistance used homelessness services within six months, compared with 4.1% of those who did not; after twelve months the figures were 1.9% and 7.2%.
Limits
Even without help, only 7.2% of families in the comparison group used homelessness services within a year, which makes assistance hard to target, and the trial covered families with moderate risk scores in one county. Its main outcome was the use of services such as shelters and outreach, not homelessness itself, and the effect was smaller after the COVID-19 pandemic began, possibly because of federal measures such as eviction moratoria.
What works
Upgrading informal settlements in place (in situ) can improve housing, tenure security, basic services, transport, public space and resilience without displacing residents. UN-Habitat stresses that residents should shape the work as co-producers of solutions rather than be treated as passive beneficiaries.
Limits
Improvements have not kept pace with rapid urban growth: more than 130 million more people lived in slums and informal settlements in 2024 than in 2015. Many cities lack the money and technical capacity to act at scale, and UN-Habitat warns that upgrading must avoid displacing the people it is meant to help.
Promising, but not proven at scale. The stage shows how far each approach has come; it is not a recommendation.
Street and shelter counts miss people who stay temporarily with family or friends or avoid visible public spaces. The OECD's Monitoring Framework to Measure Homelessness, published in 2025, uses a broad definition of hidden homelessness, and household surveys have begun to ask about past housing difficulties: in Eurostat's 2023 survey module, about 5% of adults in the EU reported at least one episode of housing difficulty, and over 80% of them had stayed with family or friends out of necessity. Countries still apply these concepts in different ways.
Beyond the yearly count, some regions now follow people through their service records: how many enter the system for the first time, and how many move from the street to interim housing, on to permanent housing or back to the street. In Los Angeles, these data showed a bottleneck between interim and permanent housing in 2025. Their value depends on how completely services record the people they see.
Because prevention funds are limited, programmes are testing ways to direct them to households most likely to lose their homes. In the Santa Clara trial, eligible families had moderate scores on a standard vulnerability screening tool and could not show that they could pay future rent, and help was allocated by lottery so that its effect could be measured. Results from one county, partly during the pandemic, cannot show how well such targeting works elsewhere.
Implementation gap
Nearly half of all households worldwide spend more than 30% of their income on rent, and the global ratio of house prices to incomes rose from 9.3 in 2010 to 11.2 in 2023. When homes become less affordable, people on low incomes have fewer ways to stay housed or to leave homelessness.
Coordination gap
In Los Angeles, 55,691 people used the rehousing system for the first time, or for the first time in two years, which LAHSA puts at 2.37 people for every person placed in permanent housing in 2025. The region lacks more than 473,000 affordable homes, and slowing this inflow depends on rents, incomes, health care and other systems beyond homelessness services.
Coordination gap
In Los Angeles in 2025, after enrolments in time-limited rent help were paused, moves from interim to permanent housing fell by 36%, from 12,120 to 7,770 people, while moves from interim housing back to the street rose by 24%, from 4,708 to 5,842. Without enough permanent homes to move on to, shelter becomes a waiting room rather than a route out.
Knowledge gap
There is no internationally agreed definition of homelessness. National statistics differ in which living situations they include, how they collect data, how often and where they count, and whether they record one night or a longer period, so national figures cannot simply be added up or compared.
Knowledge gap
All Housing First studies in the CDC Community Guide review came from the US and Canada, most followed people for two years or less, and none covered rural programmes. No study from other high-income countries met the review's criteria, so little is known about long-term health effects or about how the approach works elsewhere.
Implementation gap
Cities are expected to manage growth and deliver services, but many lack the fiscal space and technical capacity to act at scale. Globally representative urban data are often too infrequent to guide decisions, and detail on informal settlements, women, children and people with disabilities is limited.
Unresolved questions, not vetted breakthroughs. Answering any of them would change what is possible.
How can services identify, early enough, the small share of households at risk who would actually become homeless without help, so that limited prevention funds reach them?
What are the long-term effects of Housing First on health, substance use and mortality, and why do health outcomes improve less clearly than housing stability?
Can a common definition and standard survey questions produce comparable counts, including hidden homelessness, in low- and middle-income countries where homelessness and informal housing often overlap?
How far do rents, wages and the supply of affordable homes explain differences in homelessness between cities and countries, and which housing policies change those numbers fastest?
Greet people experiencing homelessness as you would anyone else, ask before you help and use person-first language. Keep the number of your area's homelessness helpline or outreach team at hand so you can pass it on if someone asks for help.
Outreach programmes, meal services and annual homeless counts often rely on volunteers. Working through an organisation that coordinates with local services makes it more likely that your help is safe and useful.
Local decisions on zoning and on new affordable or supportive housing often meet opposition. Attending a planning meeting or writing to local representatives in support is one way to add your voice to those decisions.
If you donate, ask organisations how many people they helped move into housing and how many were still housed a year later.
Ideas visitors sent in for this challenge, read by the DO GOOD team and pinned with permission. The colour shows the kind of contribution.
Loading the board…
What happens next
The DO GOOD team reads every idea and sorts it by colour: solution, research question, existing project, evidence or local action.
If you agree to publication, it can be pinned to this board. Contact details are never shown.
Pinned ideas are not reviewed by scientists and are not endorsed by DO GOOD.
This page is a research overview. DO GOOD does not run a dedicated programme on this challenge, and donations are not earmarked for it. You can join our volunteer work and outreach.
OECD, published 2026
OECD indicator sheet last updated 11 September 2026: at least 2.6 million people reported as homeless or in housing exclusion across the OECD (more than 1.4 million in the EU), data for 42 countries, trends between 2023 and 2026, ETHOS Light categories, hidden homelessness, gender shares and comparability caveats, plus country series including the US point-in-time counts used in the trend.
US Department of Housing and Urban Development (HUD), published 2026
Official US count for January 2025 (released May 2026): 745,652 people experiencing homelessness, sheltered and unsheltered numbers, individuals, families, chronic patterns, changes since 2007, 2013 and 2024, shelter and permanent supportive housing bed inventory, definitions and data caveats.
Los Angeles Homeless Services Authority (LAHSA), published 2026
LAHSA's release of the 2026 Greater Los Angeles Homeless Count (24 July 2026): 73,040 people in Los Angeles County and 45,194 in the City of Los Angeles (changes not statistically significant), unsheltered numbers, survey on causes, permanent housing placements in 2024 and 2025, time-limited subsidies, system inflow and flows between street, interim and permanent housing, and affordability figures.
UN-Habitat, published 2026
UN-Habitat's summary of progress on SDG 11 (SDG 11 Global Report 2026): more than 1.16 billion people in slums and informal settlements in 2024, at least 3 billion in inadequate housing, rent burdens and house-price-to-income ratios, and key messages on exclusion, in-situ upgrading, local finance and urban data.
Journal of Public Health Management and Practice (Peng et al., Community Preventive Services Task Force; author manuscript via CDC Stacks), published 2020
Systematic review of 26 US and Canadian studies (17,182 participants), including the largest randomised trial, Canada's At Home/Chez Soi (2,148 participants, cited from its national final report): effects of Housing First on homelessness, housing stability, health, health care use and outcomes for people living with HIV (three studies), with limitations.
Lab for Economic Opportunities (LEO), University of Notre Dame, published 2023
Research brief (June 2023) on a randomised trial of one-time emergency financial assistance for families in Santa Clara County (July 2019 to December 2020), published in full by Phillips and Sullivan in The Review of Economics and Statistics (doi:10.1162/rest_a_01344): use of homelessness services and shelter entry after six and twelve months.