The evidence is clear: supportive housing works to break the homelessness-jail cycle. But behind the studies and data points on this effective intervention is the tragic reality that the people who need such intensive supports are also the people most likely to die on the street. This group—many of whom have lived outside for decades and have severe mental and physical health challenges and substance use disorders—haven’t received the supports they need to thrive, or even survive.

That’s why the permanent housing subsidy and wraparound services that come with supportive housing programs, through a Housing First approach, are crucial for people experiencing chronic homelessness. Housing First is built on the idea that safe, secure, affordable, and permanent housing must be available before people can work on other challenges, like mental health or substance use disorders.

Through a previous Urban Institute evaluation of the Denver Supportive Housing Social Impact Bond Initiative (Denver SIB), we found supportive housing helps people stay housed in the long term, reduces people’s interactions with the criminal legal system, increases people’s access to office-based health care rather than emergency care, and lowers the public costs of the homelessness-jail cycle.

In this new analysis, we explore how supportive housing affects mortality rates and causes of death among people experiencing long-term homelessness. We looked into mortality data among an even larger group than our original study—532 people who were offered supportive housing (the treatment group) and 529 people who received services as usual in the community (the control group)—randomized into the study between January 2016 and January 2019. What we found emphasizes the stark vulnerabilities among this population and demonstrates some of the differences in outcomes when people are in supportive housing rather than living on the street.

Because the program is targeted to people with significant health challenges, the mortality rate is the same between people in supportive housing and people receiving usual services in the community
 

To be eligible for the Denver SIB, people had to have experienced repeated homelessness and frequent interactions with the criminal legal system. Without access to housing and services, they often rotate in and out of jail, detoxification centers, and emergency health care.

Because people in this population have so many health vulnerabilities, a significant number of people in both groups died over the years we collected these data. We found 10 percent of people in the treatment group and 9 percent of people in the comparison group died within three years of being randomized into the study. This difference is small and not statistically significant, meaning the difference is likely attributable to chance rather than a true difference between the groups.

This 10 percent figure is extremely tragic though consistent with other studies evaluating supportive housing programs. This share is so high because the program is targeted to people with the greatest needs and greatest chances of dying on the streets, and it shows the Denver SIB is well targeted to people most likely to experience long-term health problems.

Supportive housing reduces people’s risk of dying from exposure to the elements
 

Although there is no difference in the overall share of people who died within three years, we did find differences in the primary reason people died, which warrant a closer look.

Among all people in the treatment and control groups in our analysis, combined primary causes of death were listed as the following: drug overdose (2 percent of people), injury (1 percent), liver disease (1 percent), circulatory illness (1 percent), alcohol use disorder (1 percent), exposure (1 percent), and other causes (1 percent). The remaining causes of death represented less than 1 percent of people in either group: respiratory illness, homicide, cancer, drug use disorder, alcohol overdose, COVID, suicide, and diabetes.

The only statistically significant difference among primary causes of death between people in supportive housing and the comparison group is exposure: seven people in the control group died of exposure, while nobody in the treatment group did. This shows the SIB supportive housing prevented people from dying of exposure to the elements, such as extreme temperatures—an especially devastating and avoidable cause of death.

The majority of deaths among people in both groups are not attributable to drug overdoses. A small and equal share (2 percent) of all people in both the treatment and control groups died of overdoses within three years of being randomized into the study. This dispels the myth that a large share of people in supportive housing are dying of drug overdoses.

Supportive housing is well targeted to people with the greatest needs and works better than the alternatives
 

The Denver SIB accomplished its goal of reaching the people with the greatest needs. Supportive housing is an intensive and expensive intervention, and it isn’t the best option for everyone experiencing homelessness, such as people who fall into homelessness for the first time and who might need a one-time support to get on their feet. But for people who’ve lived outside for years and have severe, long-term health challenges, supportive housing brings them into a stable home and protects them from death caused by exposure to extreme temperatures and dangerous weather.

Studies have also shown that other approaches—such as programs that require people to pass drug tests or meet other requirements to qualify for housing—don’t work as well to help people stay housed and access the services they need, and are much more expensive.

Our previous studies show that supportive housing significantly increases housing stability, reduces involvement with the criminal legal system, lowers the use of emergency health care, and increases office-based health care and access to needed prescription medications. This new analysis shows that, because people often enter supportive housing with long-term health challenges like substance use diagnoses, supportive housing can’t necessarily prevent people from dying over the three years we collected data. But dying inside after facing long-term health challenges, rather than dying of exposure to the elements, can ensure people have the chance to be cared for and die with dignity.

Supportive housing might not be able to reverse the decades of damage done by living on the streets, but it can help people with the greatest needs find stability and improve other aspects of their lives.