A Solution to Ending Homelessness Might Be in the Data

Data-driven approaches are eliminating homelessness in communities across the U.S. Is Eugene next?

On a dewy January morning, clipboard-toting volunteers met up in communities across the U.S., including Eugene, to walk through streets, parks, encampments, underpasses and woodland areas — asking those they came across, “Where did you stay last night?” 

Every year, the Department of Housing and Urban Development (HUD) conducts its annual point-in-time (PIT) count. For one day, community members complete a headcount that feeds Congress’ national report for how many people are experiencing homeless for that year. The 2020 count estimated 580,466 people nationwide were homeless, according to HUD. 

But advocates are beginning to wonder how accurate this number is, and why we continue to collect a snapshot of an issue that changes daily. These questions hit home in a place like Eugene, which is a leading city for homelessness per capita, according to HUD’s 2020 data. 

To target gaps within the system, housing advocates like Community Solutions have begun pivoting their data collection away from the once-a-year estimate to a rolling data set that captures homelessness’s fluid motion. The data-driven approach has eradicated homelessness in some communities across the U.S. and could provide a solution for Eugene.

Homelessness to zero

In 2015, the New York City-based nonprofit organization launched its Built for Zero project. Since then, its approach has shown that data analytics can effectively address this growing problem. 

The strategy understands that no two communities are the same. Instead of working individually towards the same answer, cities form a collective team of critical stakeholders such as HUD’s Continuum of Care, which is representatives from organizations that coordinate the implementation of housing services, local governments, housing authorities, veteran services and nonprofit organizations. 

Alongside the Built for Zero project team, the group of community members gather real-time data on their city by creating a monthly data pool where they can identify their homeless community by name — as well as their age, race, veteran status and how long they’ve been homeless. 

This aids local governments in understanding who this population is and where they reach roadblocks to being housed. Alongside improvement coaches and data analytic experts, districts are guided in tearing down those barriers, says Beth Sandor, Community Solution’s principal and Built for Zero co-director.

Built for Zero targets two types of homelessness — chronic, which means an individual has been homeless for a year or longer, and veteran, where someone who has served in the armed forces is living without accommodation. Communities aim to reach a “functional zero” milestone with these two groups, which does not mean there are no homeless individuals within the community. Instead, the population is low enough that it is achievable to find solutions to house everyone. 

For veterans, this means “the number of veterans experiencing homelessness is less than the number of veterans a community has proven it can house in a month,” according to the Community Solutions’ website. For chronic homeless people, it means communities have three or fewer people experiencing homelessness within their region. 

Built for Zero is working with 83 communities across the U.S. With the disciplined data format, five cities have ended chronic homelessness, 12 have ended veteran homelessness, and three have ended both. 

“Data-driven decisions really helped us learn and identify where improvement needed to take place,” says Heather Kimme, the assistant executive director for the local housing authority in Bakersfield, California, a city that ended chronic homelessness in 2020, according to a Community Solution’s report. 

Eugene has the numbers. Now it’s about what we do with them. 

Off West 11th Avenue, multi-colored tents and tarps pinch together in the large empty lot alongside the busy road. Empty tin cans and assorted trash spill over the metal bars of neglected shopping carts, flattened cardboard boxes lay against the grass, flapping against the soft Oregon breeze. The multi-billion dollar company, Walmart, paints a contrasting backdrop to the product of Eugene’s critical housing problem. 

This is one of the many self-constructed communities formed by some of the roughly 3,000 individuals who experience homelessness in Lane County every month. 

Although Eugene exhibited an almost 26 percent decrease in numbers this past year, in 2019, the city led the nation with its houseless community per capita. Right around the time, Lisë Stuart, a senior management analyst for Lane County, was approached by Kaiser-Permanente, a health care company that works with the Built for Zero project, inquiring if the county would be interested in joining Built for Zero. 

Lane County implemented a homeless-by-name list (HBNL), which uses data already tracked from the Homeless Management Information System (HMIS). This system aids community leaders to visualize and determine the who, what, where, when, why and how individuals are experiencing homelessness by collective numbers based on how many people access human services each month. 

While the 2020 PIT count estimated 1,606 individuals were experiencing homelessness in Lane County, the HBNL pinpointed 9,107 unduplicated individuals that accessed a service in the county for that given year. 

“The Point in Time count is like a picture. You can take a picture of your trip to Yosemite and then take it home. But you can’t experience it,” Sarai Johnson, the joint housing and shelter strategist for Lane County and Eugene, says. “That snapshot could never capture everything that’s happening and the workings of what’s going on. So the by-name list is more like a continuously moving picture.” 

The data visualization tools aid county officials in identifying trends, such as where people are experiencing homelessness most. Eugene dominates within the county with nearly 76 percent of the homeless population as of March, and Springfield charts second place with 10 percent. 

Some of the Pallet shelters that sprinkle throughout the county were placed in areas where data stressed the need for them. Stuart says, “We were able to identify where people are currently experiencing homelessness, not that one time a year story.” Pallet shelters are rapidly constructed shelters designed for those facing homelessness or natural disasters.

Right beside the Lane County Behavioral Services Center, a grey and blue building perched on a brick frame welcomes visitors with freshly cut grass and trees sprouting from a lifted garden. The Commons on MLK is a 51-unit Housing First complex by Homes for Good and set to house part of the county’s houseless population. Reviewing data on the homeless beyond the PIT count helped the Poverty and Homelessness Board, a group of community stakeholders who represent the concerns of people experiencing homelessness, move forward with the complex, Stuart says. 

The data also tracks the inflow and outflow, as homelessness is like an ever-moving conveyor belt. Stuart says, “We definitely see in our data that once somebody has experienced homelessness, the likelihood of them returning to homelessness is very high,” Stuart says.

While people are entering the system every month, people are also dropping out as they move into permanent housing. It is also common for some to temporarily disappear, perhaps while couch-surfing with friends or spending time incarcerated.   

However, not all homeless service providers contribute to HMIS, limiting the county in collecting a complete data album. As well, the pandemic has significant impacts on individuals accessing services. Individuals may avoid accessing services in the county to prevent contracting the virus, which can have severe consequences for the homeless, as many do not have health insurance. 

“It’s important for us to know that there are other people experiencing some kind of homelessness under the radar,” Johnson says. 

Don’t count Eugene out 

To prevent the spread of COVID, HUD issued the option for modifications in the PIT count this year, and Lane County Human Services opted to use its HBNL, says Alexandria Dreher, the county’s program services coordinator. They weren’t the only ones; cities with large homeless populations like Los Angeles and Seattle also used their HMIS systems. This invites the question of whether the nation is beginning to shift towards a data system that the nation can count on. 

Sandor says, “There is not a community that doesn’t have data. It’s about how we use that information, to problem solve and iterate the solutions.”

This story was developed as part of the Catalyst Journalism Project at the University of Oregon School of Journalism and Communication. Catalyst brings together investigative reporting and solutions journalism to spark action and response to Oregon’s most perplexing issues. To learn more visit Journalism.UOregon.edu/Catalyst or follow the project on Twitter @UO_catalyst.