The U.S. Department of Housing and Urban Development calls it a "Point in Time" count.
It’s the number of homeless people counted on "a single night in January," according to the agency’s "Annual Homeless Assessment Report," which was submitted to Congress on Nov. 19.
To get the tally, HUD works with local observers in each state to count homeless people on the streets, in shelters and in other locations.
This year’s assessment, which reported an estimated 107 homeless veterans in Rhode Island, has drawn criticism from Operation Stand Down Rhode Island, a veterans advocacy agency, which says HUDs measurement is flawed and minimizes the numbers of homeless veterans in the state.
A press release issued Nov. 19 by HUD’s New England region office in Boston provides a list of "estimates" on homelessness reported by state and local planning agencies in Rhode Island, including an estimate about the number of homeless veterans.
"On a single night in January 2015, 107 veterans were homeless," it says.
HUD’s 75-page report cites the 107 number in several exhibits, including one table that shows just two of 107 homeless veterans living without shelter, a rate of 1.9 percent. This is lower than any other state, it says.
The news release qualifies certain limitations of the data, referring to the data as a "snapshot" and saying that the number does not reflect who is eligible for HUDs homeless assistance grants programs.
It says the "point-in-time count only captures those persons sleeping in sheltered and unsheltered locations on the night of the count."
So the count reflects all homeless veterans in unsheltered locations that night?
"It’s not an exact science," acknowledges HUD spokeswoman Rhonda Siciliano. "Obviously, going out and counting people on the street you’re not going to be 100 percent accurate."
Siciliano told PolitiFact that HUD’s counting method would probably generate two different numbers if it was done on two nights instead of one.
"It’s a very fluid situation," she said. "People go in and out of homelessness."
Erik Wallin, executive director of Operation Stand Down Rhode Island, is among critics who say the methodology of the assessment significantly minimizes the actual number in this case.
Wallin argues that many homeless veterans are unlikely to acknowledge their background when approached in a shelter or on the street. Some veterans known to the Veterans Administration are not known to HUD, he said.
Wallin’s organization served 57 veterans who were homeless when they approached Operation Stand Down for help in January of 2015. Wallin says that homeless veterans who aren’t in shelters aren’t likely to hang out where other homeless people congregate.
Wallin talks about one veteran who chose to live in his van as Operation Stand Down worked with him to find housing.
Eric L. Hirsch, a Providence College sociology professor who coordinated the count, said he believes the assessment, which was conducted in the wintertime, did not miss a lot of homeless people who were living outside.
"If someone feels we did, they needed to document that," he said. "If they had documented it and there was evidence of it, then they would have been included in the count."
The nature of homelessness -- a condition that many people cycle in and cycle out of -- can make it very hard to quantify the homeless over the course of a full year, especially in warmer climates where homeless people live outside year round.
But for the same reason, the point in time approach might give the impression that fewer numbers of people are affected by homelessness.
"The point in time count is really a snapshot and shouldn’t be interpreted as a prevalence measure," says Dr. Thomas O’Toole, a physician who cares for Rhode Island veterans and also directs the Veteran Administration’s National Center for Homelessness Among Veterans.
HUD’s definition of homelessness also affects the count.
Some critics, including Wallin and Hirsch, also say the definition doesn’t include individuals who are forced to live with a friend or with a relative because they don’t have resources for their own home.
Hirsch said he estimates there could be hundreds of homeless veterans living under such circumstances but HUD doesn’t count this segment of the homeless population due to concerns about who is voluntarily living with someone else and who is forced to do so.
Hirsch provided one other piece of info to consider when evaluating HUD’s claim of 107 homeless veterans in Rhode Island on a night in January 2015.
Rhode Island’s Point In Time count was conducted on Feb. 25, 2015. That’s not a night in January.
Some vocal critics, who are knowledgeable about the extent of homelessness in Rhode Island’s veteran population, question the usefulness and accuracy of HUD’s point in time count.
They weren’t able to provide hard evidence, for example an alternative count based on logs or records, of more than 107 homeless veterans in Rhode Island in January of 2015.
But while HUD’s press release describes how the point-in-time count works, calling it a "snapshot," it could say more about the possibility that the count missed homeless veterans who might not have been in shelters.
The claim is partially accurate but it leaves out important details. For that reason, the ruling is Half True.