The Hidden Stats on Homelessness

STEPHEN YATES' NOMADIC LIFE on the streets ended six years ago.

Living on the streets and alleys of Harlem, plagued by alcoholism and mental illness, Yates finally sought help at a local shelter. Upon intake, he provided his name, former address, family history and other personal information.

“Anything you can imagine that they could ask you, they asked,” Yates says. “But I had no place to turn. They could have asked me anything, and I would have given it to them.”

Collecting such personal information lies at the heart of a controversial program that some say will strengthen the way homeless services are delivered, while others argue that it will ignite privacy battles from coast to coast.

In direct response to a 2001 congressional mandate to collect useful data about homelessness, the U.S. Department of Housing and Urban Development (HUD) launched its Homeless Management Information Strategies (HMIS) initiative, which set an Oct. 1, 2004, goal for localities to begin developing communitywide databases to collect and report uniform data from homeless service agencies.

About 60 percent of approximately 500 local communities are implementing HMIS; the other 40 percent are shopping for software to drive their systems.

HMIS consists of standards governing technology, security, privacy, and data collection and transmission. From Feb. 1 to April 30 of this year, HUD plans to collect HMIS data from 80 communities to develop the nation’s first Annual Homeless Assessment Report, due in December 2005.

“We are in an information technology society,” HUD spokesman Brian Sullivan says. “This kind of information helps drive policy and create programs that more effectively house and serve those who need these programs the most.”

But critics argue that HMIS could violate the privacy rights of an already vulnerable population. Storing personal information in a regional database, rather than keeping it inside an agency, raises red flags for some, such as victims of domestic violence and people who have HIV, Yates says.

To ensure that the system properly guards the privacy of people whose personal data the agency collects, Yates serves as a homeless consumer advocate on the New York City HMIS subcommittee.

New York is aware of the need for privacy. As it builds its HMIS system, the city works to comply with privacy laws, such as the federal Health Insurance Portability and Protection Act and New York state’s HIV regulations, says Dilip Kulkarni, assistant commissioner of information technology for the city’s Department of Homeless Services.

Regardless of such safeguards, privacy remains a major concern.

“Background checks for even the most mundane jobs are becoming more common,” says Kevin Barbieux, a homeless Nashville man who writes about the ups and downs of life on the street in a Web log (www.thehomelessguy.blogspot.com) and works with the Nashville Coalition for the Homeless and the Mayor’s Taskforce to End Chronic Homelessness. “Once you declare yourself as being homeless through such a system, that information is going to get out.”

Invisible Masses

On a midday B express train from Brooklyn, a disheveled 30-something woman walks from car to car, repeating her tearful script. “I’m not on drugs,” she insists. “I’m asking for help to feed my kids. I don’t know what else to do. Thank you and God bless.”

One young woman hands her a dollar. The other passengers avoid eye contact. With an estimated 37,000 homeless people in New York City, they hear such stories daily and know that a handful of change will not solve the homeless problem.

New York’s shelter system is one alternative. It consists of about 300 agencies that divvy up $65 million in HUD grants awarded to the city each year. The grants are distributed through the New York City Coalition on the Continuum of Care, a group of homeless service providers, homeless people, community representatives and government agencies. While Continuum representatives meet regularly, each agency operates independently and collects its own data on clients.

For instance, when Yates signed up for transitional housing and rehabilitation, he had to provide one agency with extensive information. When he later applied for long-term housing through another program, he had to repeat his story.

“That’s a very demoralizing situation,” says Loren Hoffmann, a state IT staff member for Wisconsin and former social worker who helped build that state’s HMIS system. Homeless people must bare their souls about why they can't care for themselves or their children. If agencies shared information, Hoffmann says, homeless people wouldn't have to go through that painful process every time they need help.

Sharing data among community agencies also can help agency workers become better long-term case managers, ensuring that clients get referrals to appropriate services, Hoffmann adds.

In New York City, which selected an HMIS application service provider in 2003, a shared database can provide a much-needed bird’s-eye view of services, says Maryanne Schretzman, deputy commissioner of policy and planning at the New York City Department of Homeless Services. Once New York’s HMIS system is operating, all 300 agencies will import their client information into the same data warehouse.

DHS and the NYC Continuum plan to use HMIS data to determine the size and characteristics of the homeless population and their current patterns of service. The data also can be used to realign housing resources and service delivery within the Continuum and to create the links to mainstream programs that are essential to preventing homelessness and sustaining formerly homeless people in permanent housing.

“It’s unclear which agencies are serving which clients, so there may be some duplication and some cases where clients aren’t adequately served,” Schretzman says. “[The shared data warehouse] will help us identify service gaps.”

Wisconsin Reaps Results

Wisconsin, an HMIS pioneer, began building its system before the congressional mandate. Launched in April 2001, the HMIS program is already showing results, says Wisconsin’s Hoffmann.

“For most people, the wino on the street corner with mental illness or substance-abuse issues is the only homeless person they see in the course of their day,” he says. “But that’s generally a very small portion of the homeless population.”

A large percentage of Wisconsin’s homeless population is comprised of families, and between 25 percent and 35 percent are children, Hoffmann adds.

Most families show up looking for food or temporary shelter for a few days. By collecting that information, Wisconsin can ensure it has enough family emergency shelter space and can analyze whether clients are transitioning smoothly to long-term housing. If not, it can use the data to pinpoint where they’re falling out of the system, Hoffmann says.

Such information can also help HUD allocate funds around the country, Sullivan says. There has never been an official, unduplicated count of homeless people around the country, he adds.

“You tend to get this skewed perspective of homelessness,” he says. “More information is a good thing.”

Building Blocks

In Colorado, linking HMIS with the state’s 211 information line for human services will connect shelters with the services being sought by homeless people, explains Tracy D’Alanno, manager of homeless and resource development programs for the Supportive Housing and Homeless Programs at the Colorado Department of Human Services. (See “Protecting Privacy ” at right.)

“It’s incredible the things [HMIS] can do to help homeless persons receive the services they need,” D’Alanno says.

However, Tennessee’s Barbieux is skeptical. As an industry serving a desperate population, agencies that service the homeless have no accountability to the consumer—the homeless themselves. He’s concerned that agencies can easily use the benefits of HMIS to cut costs rather than build services.

New York’s Yates sees the pros and the cons of HMIS. He worries that HUD may use the data to label programs duplicative or unnecessary in order to justify funding cuts. He’s also concerned about the cost of building the database systems.

At his first meeting as a representative to New York’s Continuum of Care, Yates heard about a 10 percent cut in homeless services. Two months later, he discovered that the city was going to spend $2.6 million on an HMIS system. The cuts and the HMIS funding were unrelated, but Yates was disappointed to hear that an expensive system was being implemented in the midst of budget cuts.

“I would have liked to see the money go for beds,” he admits, but concedes that there are significant benefits to getting a better handle on services. “Homelessness is out of control,” he points out.