Home / Population and society / Articles / How many homeless people are in the US? What does the data miss?

More than half a million people experienced homelessness in America last year. The Department of Housing and Urban Development (HUD) counted around 582,000 Americans experiencing homelessness in 2022. That’s about 18 per 10,000 people in the US, up about 2,000 people from 2020.

Who is considered homeless?

The HUD’s definition of homelessness includes both sheltered and unsheltered people. Sheltered people are living in domestic violence shelters, transitional shelters, safe havens that serve homeless individuals with severe mental illness, or hotels/motels. Unsheltered people live outdoors, in cars, in abandoned buildings, or in other places not meant for human habitation.

People staying with friends are only considered homeless if they cannot stay there longer than 14 days.

Who is homeless in America?

Native Hawaiian and other Pacific Islanders are 1.8% of the homeless population despite being just 0.26% of the US population. They have the nation’s highest rate of homelessness at 121 per 10,000 people.

native-hawaiians-and-pacific-islanders-have-the-highest-rate-of-homelessness-in-the-us. (2)

Twenty-four percent of all homeless people identify as Hispanic. This is higher than the national proportion of the Hispanic population: 19%.

One reason Native Hawaiian and Pacific Islanders have such high rates of homelessness could be because of Hawaii’s lack of affordable housing. Many native Hawaiians are also on wait lists for home land leases.

Veterans also experience homelessness at a slightly higher rate than the overall population. Twenty out of every 10,000 veterans are homeless, compared to the nation’s overall rate of 18 per 10,000.

Embed on your website
Homeless veterans 2008-2022 (estimated)
Year Number of homeless veterans
2008 62,223
2009 73,367
2010 74,087
2011 65,455
2012 60,579
2013 55,619
2014 49,689
2015 47,725
2016 39,471
2017 40,020
2018 37,878
2019 37,085
2020 37,252
2022 33,129

The US government has established programs such as the HUD-Veterans Affairs Supportive Housing program that helps veterans find permanent housing and access healthcare. Additionally, the number of veterans has dropped. In 2021, there were 16.5 million veterans, compared to 21.8 million in 2008.

In 2022, 60.6% of homeless people were cisgender men, compared to 38.3% for cis women. Around 1.1% of homeless Americans were trans, nonbinary, or questioning.

Embed on your website

How has the homeless population changed over time?

Sheltered homelessness declined over the past decade but unsheltered homelessness rose.

Embed on your website

Homelessness looks very different across states and localities, but Government Accountability Office (GAO) found that rising rent and job losses contribute to homelessness. Policies with encampment and shelter restrictions, as well as situations that change from person to person (such as poverty and experiencing domestic violence) affect homelessness rates.

The drop in sheltered homelessness might be due to a definitional technicality, according to the Council of Economic Advisors. People in rapid rehousing aren’t counted as homeless, but people in transitional housing are, despite there not being a clear justification for one group being more “homeless” than the other.[1] From 2007 to 2018, the number of people in rapid rehousing increased by 109,000, while people in transitional housing fell by 110,000.

How is this data collected, and who does it miss?

There are separate homelessness counts for sheltered and unsheltered people.

An annual point-of-time count of unsheltered people (an unduplicated count on a single night of the homeless people in a locality) happens in the last week of January. HUD chose January because shelter use is highest in this month, so it expects a more accurate count of people who are truly unsheltered rather than including those who intermittently go in and out of a shelter.

Localities approach point-in-time counts differently. Many use a public places count, where volunteers and workers visit locations where homeless people are believed to congregate. Police officers are sometimes used to help identify survey locations or access potentially dangerous areas, such as abandoned buildings. However, police presence can also cause homeless individuals to be less forthcoming with information.

There are several drawbacks of public places counts. First, they rely on a known list of places where homeless people may gather, which could lead to an undercount. People living in cars or checking in and out of motels are missed. People may also deliberately hide to avoid the count.

Some localities take a service-based approach, counting use of food pantries, soup kitchens, and social service agencies, and other non-shelter services. However, this requires significant screening to make sure only homeless individuals are counted. It also misses people who never use these services.

Sheltered homeless individuals are easier to track because they more consistently interact with government resources. The Homeless Management Information Systems track characteristics of people using homeless services such as emergency shelters and transitional housing. Some localities also supplement this data with surveys.

While more consistent than unsheltered counts, sheltered counts also face challenges. Rural homeless services providers struggle with understaffing and lack of technology infrastructure, communication, and transportation over large geographical areas. This also excludes people staying with family members or friends due to lack of other options, children living in emergency foster care or detention facilities, or adults in criminal justice facilities.

Counting homeless individuals is difficult, but more regular quality checks could improve accuracy. A 2020 report from the GAO found that both sheltered and unsheltered counts are likely an underestimate.

Read more about standard of living in the US, the states with the most homelessness and homelessness among students, and get the facts every week by signing up for our newsletter.

2022 Annual Homeless Assessment Report to Congress
Last updated
February 2023
Point-In-Time Methodology Guide
Last updated

Rapid rehousing is apartment-based but focuses less on self-sufficiency, whereas transition housing is usually project-based and promotes self-sufficiency. In practice, people stay in rapid rehousing for around 7-8 months on average, and transitional housing for 13 months on average.