By Marcus Williams
Thirty years after the Census Bureau’s first attempt, the Obama administration is poised to grossly undercount those experiencing homelessness during the 2010 census.
Census spokesman, Michael Cook, says the bureau will repeat its caveat from 2000 that the numbers “do not provide a count of the population experiencing homelessness” and should not be construed to be one.
“But, of course, it always is [construed as one], even though they say that,” responds Michael Stoops, director of community organizing at the National Coalition for the Homeless.
The National Coalition for the Homeless publicly opposed a separate homeless count in the past, fearing it would be misleading. After the last count was conducted, they lobbied for the results to be kept private.
In 2000, the census published a count of fewer than 200,000 homeless people living in shelters, less than a third of most major estimates of homelessness at any one time.
“It is easy to count the shelter population, but 42% of the nation’s homeless are unsheltered” says Stoops, “there are a lot of people who are invisible.”

Counting Matters

In spite of the expected discrepancy between those counted and the total number that experience homelessness, the census numbers are likely to influence policy towards the homeless.
Census data that undercounts the homeless may be used by lawmakers to argue for a decrease in funding.
Over $300 billion in federal funding is doled out to the states based upon the official counts, including funding for hospitals, child-care, and homeless assistance programs. Shelters, already struggling to meet demand, fear that an undercount may threaten their budgets.
Repeating most of the methods used in 2000, the 2010 census will attempt to count all of America’s homeless population in one, three-night period at the end of March, after the coldest temperatures have subsided and many of the homeless have returned to the streets. The first night is used to count residents at homeless shelters.
“We may not know how strong the coordination [between the many partners] is until after the fact” says M. William Sermons, the director of the Homeless Research Institute at the National Alliance to End Homelessness. “There does seem to be a concerted effort.”
Shelters have protested the count in the past. According to the National Coalition for the Homeless, the Center for Creative Non-Violence protested the census in 1990, refusing to be counted. The executive director followed up by having a truckload of sand dumped at the Department of Commerce, saying “it is as hard to count the homeless as it is to count these grains of sand.”
Nonetheless, the count went forward and included 228,621 homeless individuals nationwide.
The difference between the census counts and other major counts is often staggering. While the census has never released a separate count larger than 230,000 homeless individuals, a recent count conducted by the National Law Center on Homelessness and Poverty estimated that 842,000 people experienced homelessness on a given night in February, 2007. Another, conducted by the National Alliance to End Homelessness, estimated 744,313 homeless individuals on a January night in 2005.

Who Counts?

Because all of the major counts rely on estimated service-provider use, none is able to include people who experience homelessness without using public provisions like shelters. According to a recent Columbia University study, up to 59% of those who experience
homelessness reside in vehicles.
Furthermore, the Census Bureau has also decided not to include abandoned buildings in the enumeration. They were last included in 1990.
“There are a lot of empty buildings in this country and homeless people are squatting and moving into these buildings,” says Stoops, although he recognized the fear that temporary census workers would likely feel if asked to enter unknown buildings.
According to Allan Mallach, an expert on abandoned building restoration, most large cities have over 10,000 abandoned residential
units. Meanwhile, more than 10% of properties in small towns are abandoned.
To help remedy some of that problem, the Census Bureau is seeking to employ enumerators who are familiar with the populations they are counting. That means hiring some of the homeless or formerly homeless to assist with the count.
Although the Census Bureau doesn’t ask applicants about their housing status, many who have experienced homelessness have been hired thus far, one of whom is Street Sense vendor Jeffery McNeil.
A D.C. area non-profit, Striving to Reach, Educate, and Transform Society, held two census job fairs for the homeless earlier this year. One in five of the one hundred participants had been hired a month later.
After the first night of counting the homeless in shelters, enumerators will count at local service-providers, such as soup kitchens. On the third night, they will visit a list of public places flagged as places where the homeless typically sleep outside. However, the process for listing public locations is fraught with difficulty; it relies largely on supplemental lists submitted by local organizations.

How They Will
Find People

Many local organizations hesitate before assisting with the list. “Somehow after lists are turned over there have been police sweeps [of the locations listed]. Even though the census promises it won’t happen, it has happened and does happen,” explains Stoops.
In 1990, less than half of the 30,000 local governments approached by the Census Bureau for supplemental lists ultimately submitted them.
The Census Bureau claims to have improved the method by which they will enumerate the homeless next year, including a more concerted effort to coordinate the thousands of partners involved.
The Obama administration appointed Robert Groves to direct the Census effort in early April, but Republican objections stalled his nomination until mid-July, just a few months before the decennial census will begin. He was once an advocate for statistical sampling
of undercounted populations to improve the accuracy of counts without having to interview every individual.
During Groves’ confirmation hearing, he unequivocally pledged not to adjust the 2010 census based on statistical samples.
No matter what methods are used and how much money is spent, most agree that it will be impossible to get a perfect count of the homeless.
Michael Stoops has worked full-time for thirty-seven years on advocacy for the homeless, including every census attempt to enumerate them. “If you asked me as a D.C. local, ‘Do I know every nook and cranny where the homeless stay?’ the answer is no,” he says.