What happens to people who leave shelters
Most people who exit a homeless shelter move into some form of housing — either their own place or someone else's — rather than returning to the street. The exact numbers vary by city and by how the shelter counts exits, but research from the U.S. Department of Housing and Urban Development shows that roughly 60 to 70 percent of shelter exits lead to some kind of housing placement. The remaining exits split between people who move in with family or friends, people whose destination is unknown, and a smaller share who return to unsheltered homelessness.
The type of housing people move into depends heavily on what the shelter offers. Shelters connected to transitional housing programs, rapid rehousing services, or permanent supportive housing can place residents directly into those programs. Shelters without those connections see more exits to doubled-up situations (living with family or friends) or to unknown destinations, which often means people found their own arrangement or left without telling staff.
Length of stay matters. People who stay longer in a shelter — typically 30 days or more — are more likely to exit into housing than people who stay a few nights. This is partly because longer-stay shelters usually have case managers working on housing plans, and partly because people with more complex barriers (mental illness, substance use, criminal history) tend to stay longer and need more support to house.
Key Takeaways
- Between 60 and 70 percent of shelter exits lead to housing of some kind, though the exact rate depends on the shelter's location and what services it offers.
- Shelters with case managers and connections to housing programs see higher rates of people moving into permanent or transitional housing.
- People who stay in a shelter for 30 days or longer are more likely to exit into housing than people who stay only a few nights.
- Exits to unknown destinations often mean people arranged their own housing or left without notifying shelter staff, not necessarily that they returned to the street.
- Shelters in areas with more affordable housing and more rental information programs report better housing outcomes than shelters in high-cost cities.
How shelters track where people go
Shelters report exit data to the federal government through a system called the Homeless Management Information System, or HMIS. When someone leaves a shelter, staff record where they say they are going — into their own apartment, to a family member's house, to transitional housing, to permanent supportive housing, or to an unknown destination. This data feeds into the annual reports that cities and states publish about homelessness trends.
The numbers are only as accurate as what people tell staff. Someone might say they are moving in with a friend when they actually plan to sleep in their car. Someone else might leave without checking out, so the shelter records them as "unknown destination" when they actually moved into housing. This means published exit statistics are a useful direction but not a precise count.
Different shelters use different definitions of what counts as a successful exit. Some count any move into housing as success. Others only count permanent housing or housing with a lease. Some count transitional housing (temporary housing with services) as success; others do not. When you see a statistic about shelter outcomes, the definition matters enormously — a shelter that reports 80 percent success might be counting exits to transitional housing, while another shelter's 50 percent might only count permanent housing.
Housing types people move into after shelter
Permanent supportive housing is the most stable exit. This is subsidized housing paired with ongoing case management and services. People in permanent supportive housing have a lease, pay rent (usually 30 percent of their income), and can stay as long as they follow the lease terms. Shelters with partnerships to permanent supportive housing programs can move residents directly into these units, though waiting lists are often long.
Rapid rehousing is a shorter-term program. The shelter or a partner agency pays the landlord's deposit and first month's rent, and the person moves into an apartment. Rapid rehousing typically lasts 3 to 24 months, with case management support that tapers over time. The goal is for the person to stabilize and eventually pay rent on their own. Success rates are highest when the person has income (from work or benefits) by the time the program ends.
Transitional housing is a middle ground. It is temporary housing (usually 6 months to 2 years) with services like job training, mental health counseling, or substance use treatment. People live in a program facility or scattered-site apartments and work toward moving into permanent housing. Transitional housing works well for people who need time to address barriers but are not yet ready for independent living.
Doubled-up living — moving in with family or friends — is the most common exit overall. This is not tracked as formally as program-based housing, so statistics on it are rough. People who move in with family often do so without any shelter involvement in the arrangement. Some of these situations are stable long-term; others break down after weeks or months and the person returns to the shelter or street.
Factors that predict housing success after shelter
Income is the strongest predictor. People who have a job, disability benefits, or other regular income are far more likely to move into housing and stay there. Shelters that connect residents to benefits (Social Security, unemployment, veterans benefits) before discharge see better housing outcomes. A person with $800 a month in disability income can afford a room or studio in most markets; without it, they cannot.
Having a case manager matters. Shelters with enough staff to assign each resident a case manager see higher housing placement rates than shelters where residents are on their own to find housing. The case manager helps with paperwork, landlord contact, benefits applications, and problem-solving when barriers come up. This is resource-intensive, which is why smaller or under-funded shelters often cannot offer it.
Local housing availability and cost affect outcomes directly. A shelter in a city with abundant affordable housing and active rental information programs will see more housing exits than a shelter in a high-cost city with few vacancies. This is not a reflection of the shelter's quality; it is a reflection of the housing market. A person leaving a shelter in San Francisco faces different odds than a person leaving a shelter in a smaller city.
Barriers like criminal history, mental illness, or active substance use lower housing placement rates. People with these barriers often need longer shelter stays, more intensive case management, and access to specialized housing programs (like permanent supportive housing for people with mental illness). Shelters without these resources see lower placement rates for residents with complex needs.
Why some people leave shelter without a housing plan
Not everyone who exits a shelter has a clear destination. Some people leave because they are frustrated with shelter rules, want to reconnect with people on the street, or distrust the system. Others leave because they found an informal arrangement (a friend's couch, a car, a camp) and did not tell staff. Some leave because they are in crisis or experiencing mental health symptoms that make it hard to stay in a structured setting.
Shelters record these exits as "unknown destination" or sometimes "left without permission." The person may have actually moved into housing; they just did not report it to the shelter. Or they may have returned to unsheltered homelessness. The data does not distinguish between these outcomes, which is why "unknown destination" exits are hard to interpret.
Some people cycle between shelter and street repeatedly. They might spend a few weeks in a shelter, leave, spend time on the street, and return when weather worsens or they are in crisis. These repeated cycles show up in shelter data as multiple exits and re-entries. People in these cycles often have unmet mental health or substance use treatment needs, or they are avoiding something about the shelter or the housing options available to them.
Regional differences in shelter exit rates
Exit rates to housing vary significantly by region. Cities with strong permanent supportive housing programs and active rental information funding report higher rates of housing placement. Cities with housing shortages and high rents report lower rates. A shelter in Minneapolis might report 65 percent of residents exiting to housing; a shelter in Los Angeles might report 45 percent, not because of different shelter quality but because of different housing availability.
Funding also shapes outcomes. Shelters in cities that invest in case management, benefits navigation, and landlord incentive programs see better results. Shelters in cities where funding is tight and staff are stretched thin see more exits to unknown destinations and more cycling back into shelter.
State and local policies matter too. Some states have strong eviction prevention programs that help people stay housed after they leave shelter. Others do not. Some cities have rapid rehousing programs that move people quickly into apartments; others rely mainly on transitional housing. These policy differences show up in the statistics.
What the data does and does not tell you
Shelter exit statistics show where people say they are going when they leave, but they do not show what happens next. Someone might exit to an apartment and lose it three months later. Someone might exit to transitional housing and successfully move to permanent housing after a year. Someone might exit to a friend's house and return to the shelter a week later. The exit statistic captures only the moment of departure, not the stability of the outcome.
This is why some researchers track people after they leave shelter — following up at 6 months, 12 months, or longer to see who stayed housed and who returned to homelessness. These follow-up studies are more resource-intensive and less common than exit statistics, but they give a clearer picture of what actually works.
Exit statistics are useful for understanding broad patterns — which shelters have case managers and housing connections, which regions have better housing markets, which programs work better than others. But they should not be read as a complete picture of shelter effectiveness. A shelter with a 60 percent housing exit rate might be doing excellent work with very difficult populations; another shelter with a 75 percent rate might be serving people with fewer barriers.
Frequently Asked Questions
What percentage of shelter residents actually move into permanent housing?
Research suggests 30 to 50 percent of shelter residents move into permanent housing, depending on the shelter and the city. The remaining housing exits go to transitional housing, rapid rehousing, or doubled-up situations. The exact percentage varies widely by location and how the shelter defines permanent housing.
Do most people who leave shelter stay housed?
Follow-up studies suggest that 50 to 70 percent of people who exit shelter into housing remain housed one year later, though this varies by the type of housing and the person's circumstances. People in permanent supportive housing have the highest retention rates. People in rapid rehousing or doubled-up situations have lower retention rates.
What does "unknown destination" mean in shelter statistics?
It means the person left the shelter and did not tell staff where they were going, or staff did not record their destination. This could mean they found their own housing, moved in with someone informally, returned to the street, or left in a way that prevented staff from documenting their exit. The category does not tell you which.
Why do some shelters have better housing outcomes than others?
Shelters with case managers, connections to housing programs, and funding for benefits navigation typically see better outcomes. Local housing availability and cost matter too — a shelter in a city with affordable housing and rental information will naturally see more housing placements than a shelter in a high-cost city with few vacancies.
How long does it usually take to move from shelter into housing?
It depends on the program. Rapid rehousing can move someone into an apartment within weeks. Permanent supportive housing often has waiting lists of months or longer. Transitional housing placement varies. People without case manager support often take longer or do not move into formal housing at all.