By D'Vera Cohn
Washington Post Staff Writer
Monday, March 4, 1991 ; Page D01
More than 48,800 people slept in shelters for the homeless in the District and its suburbs in 1989, a number that officials say is the most accurate estimate of the area's homeless population but probably far from the actual figure.
"Counting the homeless is virtually impossible," said Jack Powers, director of Alexandria's economic opportunities division. "Many homeless people do not seek shelter. If a shelter is full, they're turned away. They're not counted and for the most part not countable."
Attempts to gather statistics on the number of homeless people in the area are fueled by public demands to know the extent of the problem and to judge the need for more government services.
But the task of counting the area's entire homeless population is frustrating, expensive and potentially risky -- both physically and politically. Government officials and advocates for the homeless who often joust with them agree, for different reasons, that counting costs too much effort for too little return.
"The value of this data is looking to see what the trends are, rather than putting so much emphasis on the number," said Harriet Goldman, director of services for the homeless in Maryland's Department of Human Resources. The trend, officials agree, is that the homeless population is worsening with the recession.
The figure 48,837 represents the number of people using the area's shelters, as reported to local or state officials.
According to the Virginia Coalition for the Homeless, 17,081 people were sheltered in Northern Virginia in 1989; state officials consider the number accurate. In Maryland, state officials reported that 12,552 people were sheltered by Anne Arundel, Howard, Montgomery and Prince George's counties in 1989, the last year for which statistics are available.
In the District, 19,204 people were counted in the city's shelters for families and adults in the fiscal year that ran from Oct. 1, 1988, to Sept. 30, 1989.
The officials who compiled those numbers are the first to say they are not highly accurate. People who spent time in more than one shelter are counted more than once. Sometimes, so are people who spent more than one night in a single shelter.
The figure excludes many who fit the usual definition of homeless: people who sought help but were turned away, people sleeping in the streets, people seeking sanctuary in abandoned buildings.
No one has sought to discover how many people are living with relatives, or are a paycheck away from eviction. Some homeless people, if asked, deny they are homeless because they are ashamed, advocates say.
The most ambitious attempt to count the homeless was last spring's Census Bureau one-night enumeration of people in shelters and other public places. But Census Bureau officials say the $2.7 million effort counted only part of the nation's homeless population.
"We weren't in the business of trying to define homeless," said Peter Buonpane, an assistant director of the Census Bureau. "We're going to count components of the homeless population . . . and users can add up the numbers as they see fit."
Figures on the homeless from that count will be released beginning in July.
Some advocates for the homeless and academics complain the count was done sloppily and should have been broader. The failure to do a complete count means "the bureau itself has shirked," said Charles Williams, an anthropologist at the University of the District of Columbia.
But many advocacy groups argue that a more complete count is useless because shelters already are overflowing, and taxpayers are not prepared to pay for more services. They say officials and politicians do not want a broader count because it would force them to acknowledge that services are inadequate.
"It's certainly in their best interest not to keep those statistics," said Susanne Sinclair-Smith, director of the Washington Legal Clinic for the Homeless. Other advocates, including the late Mitch Snyder, who barred census takers from his Community for Creative Non-Violence shelter during the count, fear that any census tally, however incomplete, will be seized on as the official number.