Census methods could provide lift to hidden homeless

SF Public Press
 — Feb 1 2010 - 4:43pm

The 2010 Census could help address one of the thorniest problems in dealing with San Francisco’s long-standing homeless problem — getting an accurate head count. 

The city’s homeless figures have ranged between about 6,500 and 8,600 people in the last decade, but the real number is anybody’s guess. The sketchy knowledge of who is living on the street has been a big impediment to perennial attempts to solve the crisis.

Temporary census workers will spend three days at the end of March interviewing homeless people at their usual gathering places, including shelters, soup kitchens, parks and highway underpasses. The census workers will ask questions similar to those asked of people who do not reply to questionnaires delivered to households.

The official number of homeless people in the city matters because it can potentially affect the number of representatives for state and federal legislative because they’re drawn based on population. It also impacts federal, state and city grants for social service programs for the homeless.
 
Part of the problem has been that for each count, the methodology changes, and so does number of workers and time dedicated to the count. These tallies have been conducted by the city and an array of private nonprofit service and advocacy groups, each with its own political agenda and definitions of homelessness.
 
Politicians and advocacy groups have also been known to use different numbers depending on the audience. And no one in the government is quite sure of the real number.
 
“Part of it has to do with conflicts between academic estimates of the homeless population and community and activist estimates,” said Chris Bettinger, who teaches sociology at San Francisco State University.
  
Bettinger said the Census Bureau, which has only been including homeless people in its counts since 1980, could not guarantee that this year’s numbers will be definitive, or even better than other methods. The advantages are that the counting of the homeless across the country will be somewhat standardized and conducted by paid staff, not volunteers.
 
Population and politics
 
The homeless count has been a subject of controversy since then-Mayor Willie Brown declared to media outlets such as the New York Times in 1998 that more than 4,000 people were homeless in San Francisco, but then told the U.S. Department of Housing and Urban Development that number was about 12,000 when seeking federal funding. Some advocacy organizations have questioned the official estimates, saying the methods used to gauge homelessness run the risk of undercounting.
 
Sonny Le, a Census Bureau representative, said the bureau would spend three days in March on training.
 
“Thanks to the stimulus funding, we currently have over 150 staff in the Bay Area who have been working closely with service providers to identify potentials sites and locations for these specific operations,” Le said.
 
He added that the national enumeration would be conducted at transitory locations, such as RV parks, campgrounds and hotels. In addition, walk-in sites will be established to assist anyone who has not received questionnaires because of unusual living arrangements, including lack of a domicile.
 
An imperfect science
 
When Robert Livingston worked the 2000 Census, he was staying at Hospitality House, a shelter in San Francisco’s Tenderloin neighborhood. He is now 50 and living in his own apartment.
 
Livingston recalled a meeting in which census workers strategized with Mayor Willie Brown's staff on how to include street people in the count. The result was a disorganized, defensive approach.
 
“Most of the census plans I remember seemed to be how to keep the enumerators safe while they did the count,” he said. “I don’t think hardly anyone had a clue about how to do the count.”
 
In his work for the 2000 Census, Livingston handed out questionnaires to people lining up outside Glide Memorial Church and interviewed people at a single-room-occupancy hotel on Bush Street. He also drove with his group to an obscure lot in North Beach where they found people sleeping underneath the brush and counted them on the spot.
 
Cities conduct their own count every two years, as a requirement for McKinney-Vento Act funding from the Department of Housing and Urban Development. However, those data collection techniques are admittedly imperfect — and wildly inconsistent, even from year to year within a single city.
 
Mixed results from volunteers
 
For example, San Francisco relies on volunteers for counting and devotes just one night of training before sending them to the streets that same evening. They’re given checklists and are discouraged from making contact with people living outside. Also, the counts take place in January, when homeless people attempt to keep themselves warm and dry and out of view from the police.
 
Last year, the city’s Human Services Agency unleashed more than 400 volunteers, some from community organizations. But detailed demographics on homeless people were gathered by trained interviewers at shelters, hospitals and jails.
 
San Francisco placed the 2009 homeless figures at 6,514. Since the city earnestly began counting its homeless in 2000, officials often explained away bumps in 2005 and 2007 studies as the result of improved coverage by more volunteers. They use 2002 figures — a count of 8,640 — as a reference illustrating a decrease in visible homelessness while glossing over the intervening years.
 
Last year, in a San Francisco Chronicle interview, Dariush Kayhan, Mayor Gavin Newsom’s homeless policy director, said fewer homeless people were found in the Richmond, Marina, Pacific Heights and Castro neighborhoods than in previous counts. However, Bayview-Hunters Point has seen a slight increase.
 
“We have brought many invisible homeless people” into social service programs, Kayhan told the Chronicle, “folks who were burrowed into bushes and hills throughout the city in parks.
 
Counting takes money
 
The U.S. Census Bureau admits that time constraints and limited access to service-based locations provide only a snapshot of homeless America. Neil Donovan, executive director of the Washington, D.C.-based National Coalition for the Homeless, agrees.
 
“It depends on the willingness from the community to invest in the time and resources,” he said. Donovan said that methods used locally and nationally are flawed.
 
Despite the census questionnaire’s shortcomings, enumerators may use it to ask for a name, age and race or ethnicity, allowing homeless people the opportunity to self-identify. They are also required to inform respondents that their information is protected by strict confidentiality rules.
 
“The federal government not only has a lot of resources at its disposal, but it also has the power of law on their side,” Donovan said. The census count is required every 10 years under the U.S. Constitution.
 
The National Coalition for the Homeless advocates the use of statistical sampling for increased accuracy in all counting operations, a method that Congress has repeatedly rejected. In a position statement, the organization said the data could be extrapolated from such sources as eviction and foreclosure rates, school district estimates of homeless children and reports of calls made to centralized service assistance hotlines.
 
In the same statement, the group also said such data should be used to improve efficacy in delivering services, not just to satisfy academic curiosity.
 
But Donovan does not believe the claim that a federal undercount is a deliberate attempt to misrepresent homeless numbers. “I’d attribute it to apathy more than anything else,” he said. “This is not willful neglect, just a lack of will to care for this population.”