(see Poster Gallery) Understanding the Service Pathways of Homeless Populations: Insights from the Homeless Management Information System (HMIS) Data

Background: Homelessness is an inequity that has always been a matter of surviving. Covid-19 underscored that homelessness is a public health crisis for all. A “pathway” is an essential concept for understanding homelessness and building solutions. Historically, previous research focused on common pathways through which people enter homeless, but less is known about how people navigate through services while experiencing homelessness. This study explored the service pathways of homeless individuals from a longitudinal perspective to identify service use patterns and gaps in services.

Methods: The individual-level longitudinal dataset was provided by a local nonprofit organization that manages the Homeless Management Information System (HMIS), an information system used to collect client-level data on the use of housing and other homeless services, for the Capital Region of New York (defined as the Albany–Troy–Schenectady Metropolitan Statistical Area with a population size of 800,000 people). Our dataset comprised anonymized information of a total of 6,011 individuals and 18,818 records of all services provided by homelessness service providers in the region from 2012 to 2018. The data captures the administrative timeline of services, including the beginning and end dates of each service, transitions between service types, and basic demographic characteristics. The data also yielded 193 different pathways for those who utilized more than one service during the time frame of 2012 to 2018. The overall characteristics of service pathways (i.e., length, patterns in service use) were analyzed using descriptive statistics. The path lengths were visualized in cumulative distribution to examine how many steps it took for people to exit homelessness by various routes through the HMIS data. In addition, the pathways involving psychiatric hospitals or substance abuse treatment facilities were analyzed separately and compared to the other pathways in the data.

Results: The analysis of individual records showed that 70% repeated visiting shelters or temporary housing options. For example, people who entered the system through an emergency shelter were likely to hop to another emergency shelter instead of moving to a more stable housing environment. Regardless of starting conditions, the ultimate goal (e.g., permanent housing) was either reached quickly or not at all. Those who had substance use issues were likely to reach stable housing options more quickly than the other groups. However, many individuals connected with substance use programs showed an overall pattern of repeating contact with temporary or institutional settings such as hospitals, treatment facilities, or jails.

Implications: This research signals that the service contact points for most homeless populations are still limited to emergency shelters and many people who experience homelessness are visiting the same type of facilities repeatedly over time. Future research can further examine the characteristics of pathways and the homeless populations who repeat specific patterns to identify if any significant factors perpetuate the pattern. In practice, Coordinated Assessment is used to target the most vulnerable; however, considering the patterns of individuals who repeat contact with multiple homeless service agencies, additional earlier intervention plans could be developed.