It's been an accepted fact in recent years that Orange County's homeless population is one of the largest in America, trailing only Los Angeles and New York.
Officially, the number of people without a roof over their heads was pegged at 34,898 countywide.
But on a rainy night in January 2005, when the federal government asked cities and counties across the nation to count bodies in shelters, parks and on the street, Orange County researchers came back with a startling tally: 2,848. The county's 3,300 shelter beds weren't even full.
Homeless advocates were dismayed. "They don't want people to think there are only 2,800 homeless here. Everyone knows that number is not true," said Greg Robinson, director of Cal State Fullerton's Social Science Research Center, which helped coordinate the survey.
Other estimates for Orange County's homeless population, based on scholarly research, range from 5,800 to 11,600.
Finding the true number is a logistical and political minefield, partly because counting the homeless is an inexact science, and partly because of disputes over how to even define homelessness. But on this most people can agree: The number of people living on the streets each night in Orange County is far below the 35,000 still cited by county officials.
Counting the homeless is hardly a new problem. The conflict erupted in the 1980s, when homeless activists claimed 3 million Americans had no shelter, but researchers said 600,000 tops.
"People disagree about this, and they disagree a lot," said Joel Best, a University of Delaware sociology professor whose book "Damned Lies and Statistics" recounts early feuds over homelessness tallies.
Numbers stir emotions, Best said, because they're used to symbolize "how big a problem is and what sort of moral standing it should have." That, in turn, influences public perceptions, government policies and private donations to charities.
In 1990, the U.S. Census Bureau took so much heat for trying to quantify homelessness that it now refuses to release data on the number of "people without conventional housing."
That's fine by Michael Stoops of the National Coalition for the Homeless. "We spend too much time, energy and money trying to count the homeless" instead of actually dealing with the problem, he said. "It doesn't matter whether there's 600,000 or 3 million because we're not even helping the 600,000."
But other experts say the key to ending homelessness is an accurate portrait of its size and causes. That's why the U.S. Department of Housing and Urban Development requested overnight homeless surveys every two years, which began last year. Although the tabulations don't directly affect federal funding to combat homelessness, they are used to prioritize services.
The problem is, even the government can't make up its mind about who should be considered homeless.
Suppose a family of four gets evicted but a friend takes them in while they get back on their feet. Are they homeless? Not according to HUD's definition of the term, which excludes people doubled up with friends or relatives. But they are homeless under the No Child Left Behind Act of 2002. That standard, which is used by schools nationwide, expands HUD's criteria.
However, city and county homeless tallies, which follow HUD guidelines, never include doubled-up people, a policy that rankles Stoops. As he sees it, rising housing costs have created a boom in the number of people who can't afford shelter, but the problem is underplayed because HUD won't revise its "narrow definition."
In Orange County's case, if last year's one-day homeless census had included doubled-up households, the count would have jumped from 2,848 homeless to 10,700, based solely on the number of schoolkids whose living arrangements at that time were listed as "doubled-up or tripled-up" by local school districts.
Defining homelessness is only half the challenge. The other is finding them. In recent years, officials around the nation have crawled into caves, kayaked down rivers and scoured abandoned buildings to locate homeless encampments. But many still get missed, according to some experts.
Los Angeles County, for example, sent 1,000 canvassers to comb selected areas for the January 2005 count. Many of the paid surveyors were homeless themselves. "If I walk down a street in skid row, what I see is very different from what I see if I walk down the street with a guy who has lived there," said Peter Connery, whose firm ran the census. "He knows who's in abandoned buildings and what doors look locked but aren't."
After Connery's team physically counted 33,000 people, about half in shelters, UCLA researchers projected a countywide total of 65,000.
