To help the homeless, there’s an app for that

The gauzy glow of a flashlight inside a blue canvas tent drew Susan Just’s attention. She and three other volunteers had walked along city streets for almost two hours on a chilly January night as part of Sacramento County’s biennial homeless count. The foursome searched an area outlined on a navigational app on their phones, but they had yet to come across anyone who appeared to live outside until spotting the blue glow.

The tent stood on a muddy, tree-lined bank above a busy road a half-mile from the campus of California State University, Sacramento. Shining her flashlight from the sidewalk below, Ms. Just, a retired state worker, offered a cautious “Hello?” A few moments later, two men emerged from the tent, and she conducted short interviews with each of them, recording their answers using another mobile app.

The survey’s organizers introduced the two apps this year seeking to gain a more accurate and detailed census of the county’s homeless population, which the 2017 survey showed had climbed to nearly 3,700, a 30 percent increase from two years earlier. Their embrace of digital tools reflects growing efforts in California and other states to reduce the guesswork in counting and tracking homeless people – advances that in turn could fortify supportive services to aid them.

“You’re not going to find everybody who’s homeless,” says Arturo Baiocchi, a research fellow at the Institute for Social Research in Sacramento who oversaw data collection for the county’s survey. “But the hope is that, with the help of tech, we can get a more complete estimate and develop more ways to reach people.”

San Diego County augmented its search for homeless residents in January with drones and helicopters equipped with thermal imaging technology, a method borrowed from Las Vegas. Map and interview apps were used by organizers in several of California’s largest counties, including Los Angeles, Orange, and Riverside, and in big cities throughout the country, ranging from Phoenix to Charlotte, N.C.

The US Department of Housing and Urban Development (HUD) requires cities and counties that receive federal funding for homeless programs to carry out point-in-time counts of people living in shelters and on the streets, a category that covers tents and vehicles. The agency describes the tally as a “snapshot.” As advocates emphasize the need to protect the privacy and personal data of homeless people, they see the potential of tech to improve an imperfect survey process that studies show chronically underestimates the size of the unhoused population.

“The homeless count helps determine policies at the city, state, and federal levels,” says Kathryn Durant, operations coordinator for the San Diego Regional Task Force on the Homeless. “The more people we can find, the more we can help.”

‘These are my neighbors’

HUD estimates that 130,000 people in California lack permanent shelter, accounting for almost a quarter of the country’s homeless population. Los Angeles, San Diego, San Francisco, and San Jose rank among the 10 cities with the highest number of homeless people nationwide.

The surge of homelessness in Sacramento, while smaller than in the state’s largest metro areas, parallels a trend across the West Coast, where soaring housing and rental prices contribute to the problem. The crisis led Oregon lawmakers last week to pass a statewide rent control measure, the first of its kind in the country.

Sacramento Steps Forward, the nonprofit group that coordinated the county’s two-night survey, recruited 800 volunteers for the effort, compared with 250 in 2017. But even with more manpower and digital innovations, Mr. Baiocchi explains, locating every homeless person remains an ideal far out of reach.

One reason involves HUD’s guidelines. The point-in-time count excludes people temporarily staying with friends, families, or neighbors as well as those living in motels or campgrounds. Another reason is the unrealistic task of canvassing all 994 square miles of Sacramento County.

The census focused on about one-third of the county and covered the areas where most residents live. Baiocchi and his research team culled data from prior surveys, nonprofit groups, public agencies, and law enforcement to map 200 count zones, estimating the density of homeless people in each one as high, moderate, or low. Organizers then assigned volunteers to search the tracts.

Ms. Just signed up out of a sense of civic empathy. “These are my neighbors,” she says. “I want to see them accounted for and cared for.” The experience of her group revealed that applying tech to the census, even if still unable to capture the full breadth of homelessness, can add nuance and context to the “snapshot.”

Just joined three college students to walk through a neighborhood on Sacramento’s east side and part of the California State University campus. None of them had participated in previous homeless surveys, and their uncertainty showed. They seldom strayed from the route delineated on the navigational app and mostly skirted wooded areas, gullies, and riverbanks where they might have encountered homeless residents.

The group had nearly completed its itinerary for the night when Just glimpsed the blue glow on the muddy bank and called up toward the tent, drawing out the two men inside.

She first interviewed Mark Rothenberg, who told her that he wound up homeless in 2016 after the state revoked his driver’s license and he lost his job as a tow-truck operator. When she asked what services the city could provide to better support him, he stammered for several seconds as if surprised anyone cared. “Make it easier to get housing and a job, I guess,” he said.

Volunteers collected answers to a list of questions from more than 600 people during this year’s census, tripling the total from 2017. Baiocchi ascribes the increase to the interview app making responses easier to record than with pen and paper, and likewise, the tool has expedited his team’s data analysis, with results expected in the next few weeks.

He predicts the growing trove of demographic details – beyond enhancing the planning and accuracy of future surveys – will assist advocates and public officials in refining outreach and prevention strategies.

“Getting a more accurate count can help us understand the scope of the problem,” he says. “Getting more information about who is homeless and why can help us understand and address the causes.”

Promise vs. privacy

San Diego County’s homeless population of 8,600 represents the fourth-highest total in the country. Yet the $20 million in federal funding that the county receives for homeless programs places it behind 19 other metro areas. The discrepancy arises in part from the county’s struggle to conduct a thorough point-in-time count across its vast, varied terrain.

Prodded by HUD officials to attempt a wider census this year, county organizers turned to tech. They enlisted law enforcement agencies to fly drones and helicopters armed with cameras and thermal imaging devices to locate homeless encampments in canyons, ravines, and other remote areas. Outreach teams on the ground later traveled to the locations to confirm the number of unsheltered residents and conduct interviews with them.

Ms. Durant, with the regional homeless task force, offers a simple rationale for exploiting tech to track an elusive population. “If you want to help more people,” she says, “you need to know where they are.”

Along with new methods for conducting point-in-time counts, an array of digital tools aimed at aiding the homeless population has cropped up in recent years.

One relies on blockchain technology to create an encrypted digital identity for homeless individuals that enables them to securely store vital personal records online. The technology averts the common problem of damaged, destroyed, or lost documents that can thwart those living on the streets from obtaining housing, medical, and financial assistance.

Other advances include mobile apps designed to improve access to supportive services. In Santa Monica, first responders began using an app last month that enables them to connect with service providers and case managers in close to real time when they encounter a homeless person.

Researchers with the Digital Health Lab at the University of Southern California helped develop the app. Karthik Murali, the lab’s director, suggests that the tool can nurture cooperation in a manner that could lead a police officer to take a homeless person to an emergency shelter rather than jail.

“If a case manager can quickly share a person’s history with law enforcement, then that might mean an unnecessary arrest can be avoided,” he says.

The promise of digital remedies to homelessness remains shadowed by privacy concerns, and advocates warn of misuse of medical, financial, and other personal data. In the view of Molly Cohen, executive director of ShelterTech, a San Francisco nonprofit that has funded installation of free Wi-Fi in homeless shelters in the city, the needs of homeless people must be measured against their right to confidentiality.

“It’s a tough balancing act,” she says. “You want to help people who are homeless or at risk of homelessness but not at the expense of their privacy. Some people don’t want to be helped because of the stigma associated with being homeless.”

At the same time, advocates recognize there are others, particularly children, desperate for assistance. The use of drones to locate those living outside has elicited criticism in San Diego. Durant respects that perspective but points out that the county faces a spiraling crisis.

“There are legitimate concerns about privacy,” she says. “There are also thousands of people on our streets.”