More than 57,000 men, women, and children are homeless on any given night in Los Angeles County, and this number is increasing. According to the 2017 point-in-time count, homelessness was up 23 percent between 2016 and 2017. This, despite more than 14,000 people moving out of homelessness into permanent housing.
Tragically, the fastest growing group of homeless are age 18 to 24 (up 64 percent), followed by those under age 18 (up 41 percent). Some are runaways, others have been kicked out of their homes or have aged out of foster care and other juvenile services. Some are addicted, or have been abused.
All of the shelters have waiting lists, and even with rent vouchers, permanent housing just doesn't exist. Consequently, only 26 percent of Los Angeles' homeless are provided shelter. The remaining are living on the streets, along river beds, under freeway overpasses, in their cars, behind buildings, in parks or in other makeshift encampments.
For many reasons, not the least of which are health concerns, public works crews have cleaned more than 16,000 homeless encampments and removed 3,000 tons of trash. Unfortunately, this didn't prevent a recent Hepatitis A outbreak among the homeless population.
Homelessness is expensive. Impacting cities across the country, it costs Americans billions of dollars each year. Getting a handle on this problem could not be more important.
How can GIS make a difference?
Not just undaunted but truly optimistic, analysts from agencies across Los Angeles County are using GIS to help combat this humanitarian crisis. They recognize they will need a comprehensive, coordinated and efficient, data-driven action plan that can ensure homelessness is rare, brief, and non-recurring.
Using public data from Los Angeles for demonstration purposes, this case study explores several ways that any city can use GIS to help them address homelessness. The workflows presented below include:
- Creating a risk surface for homelessness and examining the spatial patterns of various risk factors.
- Mapping the distribution and characteristics of the homeless population.
- Weighing different options for locating new homeless resources.
- Developing a framework to quantify and monitor performance, making adjustments as needed.
To learn more about how these workflows might apply to your own city, consider completing the detailed, step-by-step tutorial using the data for Los Angeles, and then repeating those steps using your own data.
Where are people becoming homeless?
Prevention is much less costly than dealing with the complications of long-term homelessness. Consequently, a key component of the Los Angeles Homeless Initiative is early intervention. This requires knowing where people in Los Angeles are becoming homeless.
While the point-in-time count tells you where homeless people are (and you will make good use of this data later), unfortunately, it doesn't tell you where they were when they became homeless.
You will need to predict where people become homeless using the key factors known to contribute to homelessness. These key factors include poverty, addiction, lack of affordable housing, unemployment, mental illness, domestic violence, and high health care costs.
With census tract data in hand, you will create a hypothetical tract with the worst-case values for each of the risk factors. Then, you will rank all other tracts against this worst-case, and map the results (see the workflow for details). Locations with one or more risk factors (the darkest areas on the map below) are more likely to generate homelessness. The risk factors analyzed include poverty, unemployment, disabilities, public assistance, paying more than 50 percent of total income for rent, domestic violence, mental illness, post traumatic stress disorder (PTSD) among Veterans, lack of health insurance, and substance abuse. All of the risk factors are given equal weighting for this analysis. (Explore the interactive maps here).
Where should programs to prevent homelessness be implemented?
While the risk factors and the underlying causes for homelessness are complex, they are also spatial, and this is so important. By applying the Science of Where, you can identify where to prioritize targeted remediation to help prevent homelessness.
Ending veteran homelessness, for example, remains an important priority in Los Angeles. Other remediation opportunities exist to address the lack of affordable housing and unemployment.
One option is to create remediation plans that span the entire county (at great cost). The alternative is to use GIS to identify the locations where specific remediation programs are likely to have their greatest impacts.
Identifying where risk factors are (and where they are not) provides a solid strategy for locating remediation projects designed to prevent homelessness.
Where are the homeless?
Spatially-informed strategies to prevent homelessness are only part of the solution. Additional projects are needed to help existing homeless families and individuals move into permanent housing.
Hot spot maps can show you where homeless communities are, in relation to LA County residents. Compare the map of where homed residents live (left) to where homeless populations are found (right). Notice the concentration of homelessness in downtown Los Angeles.
The hot spot maps below show where both sheltered and unsheltered homeless people concentrate. During the point-in-time count, January 2017, more than 42,000 people (74 percent of the homeless population in Los Angeles County) were living on the streets, in their cars, or in other makeshift situations.
Knowing where the homeless are concentrating is important, but so is recognizing differences among these communities. The homeless in Venice Beach or Hollywood, for example, have different characteristics than the homeless in Skid Row. Consequently, the needs in each of these areas will be different. Skid Row, for example, has the smallest area, but the largest (by far) number of homeless with 57 percent of the homeless accessing shelters. In Venice, only 12 percent of the homeless are in shelters.
To demonstrate these differences, you can create a map identifying
where the majority of homeless people are either staying in
shelters; living in cars, vans, or campers, living in
tents and makeshift encampments, or living on the street. 
You can add other characteristics of the homeless population including veteran status, gender, age, race/ethnicity, and factors contributing to homelessness. Try the interactive maps in this story map or work through the tutorial to create charts exploring these characteristics.
Where should new resources go?
Los Angeles County has outlined an action plan with projects ranging from expanding rapid rehousing programs to building many thousands of new housing units (focusing primarily on supportive housing for the chronically homeless).
Where should these new resources go? Any number of different scenarios could be considered. Each scenario puts value on different objectives.
One option is to promote social equity. If we believe that everyone has a responsibility to share the burden of homelessness, the goal will be to distribute resources equitably. From this perspective, a location with 1 percent of the population should be associated with 1 percent of homelessness. For the map below, adding new resources to the red tracts will improve equity. To create this map, a supply-versus-demand variable was calculated based on the total number of people and the total number of homeless people in each census tract.
A second option is to prioritize access by putting new facilities where existing homeless people live. In the map below, adding new resources to the red tracts will improve access. To create this map, a supply-versus-demand variable was calculated based on the total number of homeless people and the total number of homeless resources in each census tract. The number of sheltered homeless was used as a surrogate for homeless resources.
A third option is to locate new resources in the highest risk areas. This is the strategy the mayor of New York is promoting. If resources are available where they become homeless, people are more likely to remain close to existing communities where their children attend school, where they know their neighbors, and where they are likely to have existing resources. Again, adding new resources to the red tracts and avoiding the white tracts will prioritize areas at highest risk for generating new homeless people.
Another option is to consolidate new resources into homeless triage centers by encouraging centralization of resources. This is the model San Francisco has adopted.
Adding new resources to the areas that already have the most resources will encourage homeless resource hubs.
A final option focuses on the most vulnerable homeless populations, evidenced by numerous 311 calls and crime incidents involving homeless people, as well as high numbers of chronically homeless individuals. These locations become candidates for rapid-response, focused interventions aimed at getting every homeless person precisely the resources they need to move out of homelessness permanently. Research indicates a small portion of the homeless population use a majority of money targeted for homelessness. Addressing these people first, will have the biggest impact on reducing costs.
Other possible options for locating new resources might prioritize access to transportation centers, or perhaps give preference to low-crime areas. If you are considering a Housing First solution, you may want to prioritize locations where permanent housing could be built or where existing housing could be converted into housing for the homeless.
Overlaying the maps for each of the five scenarios mapped above, reveals solutions that optimize one or more objectives.
The map above gives equal weighting to each scenario, but other options are possible. For stakeholders with conflicting priorities, these maps provide neutral ground to discuss different perspectives, encouraging both transparency and collaboration.
What's working?
It will be critical to assess the effectiveness of both the new and the existing programs created to end homelessness. The point-in-time count, required by the US Department of Housing and Urban Development, will be a key component of these efforts. Counts are conducted once every year or two across the United States. The Los Angeles count is the largest in the nation involving almost 8,000 volunteers over three days and nights.
The data collected provides information about the number of sheltered and unsheltered homeless. The survey asks questions about why people became homeless, and tracks demographic information such as race, age, gender, family structure, and Veteran status.
The 2017 and 2018 counts will become the baseline for a multi-billion-dollar program funded by two Los Angeles ballot measures: proposition H and proposition HHH. The money from these initiatives will most likely be used to deploy outreach teams composed of case workers and health specialists; to generate bridge housing to help prepare homeless people for permanent homes; to expand a rapid rehousing program and rent subsidies; to increase supportive services such as job training, substance abuse counseling, and mental health treatment; and to construct 10,000 new units of permanent-supportive housing.
A coordinated effort across agencies and departments is necessary to register all homeless services and to track both utilization and capacity. With linked records across agencies, each encounter with a homeless person will provide data to help quantify the effectiveness of services rendered. Metrics to measure program impacts will answer questions such as these:
- Is the average number of days a person remains homeless going down?
- Is the percentage of chronically homeless decreasing?
- Are outreach programs (for substance abuse, domestic violence, unemployment or institutional discharge, for example) assisting a larger number of people and as a consequence, are fewer people becoming homeless due to these risk factors?
- Has every homeless veteran been placed in permanent housing?
- Does emergency shelter and housing capacity match the number of newly homeless each month?
Reducing the number of days a person remains homeless is critical for managing the well-being of individuals and families. It also greatly reduces costs and impacts on public health and safety resources overall. Long-term homelessness takes a toll on physical and mental health, and is correlated with long-term unemployment. The longer a person is out of the job market, the harder it is to re-enter, and this is expensive from an economic, governmental, and societal perspective.
Next steps
Homelessness in America has reached crisis levels. We will need to leverage every single tool we have available in order to combat it. Initial efforts need to focus on stopping the increase, and must be followed by strategic, monitored, action-oriented plans aimed at ending homelessness. GIS and the Science of Where provide powerful tools to help meet this challenge.
Feel free to download the data and to work through the tutorials created for this case study to see if elements of these workflows might help your own city move toward ending homelessness. Explore applications for counting the homeless, for maintaining an inventory of homeless services, for helping the homeless find local resources, and for engaging the community in reporting homeless activities. The tutorial includes a workflow for setting up a resource inventory and service locator application.