Personal Reflections on the Point-in-Time Count
- Eric Tars, Senior Attorney -
Picture from and can be found here.
Each January, thousands of volunteers across the country spend a cold night on the streets counting people experiencing homelessness for the Department of Housing and Urban Development’s annual Point-in-Time (PIT) count. Although the Department itself acknowledges the count is imperfect and an undercount, the numbers are nonetheless cited over and over by policymakers and the media as evidence of progress, or lack thereof, in ending homelessness.
This year, I participated in a PIT count in Philadelphia run by our friends at Project HOME
. They had a fantastic turnout of more than 400 volunteers who were able to cover all of Philadelphia’s zip codes. But despite that, the experience made even clearer to me that the numbers the count reveals are misleading at best.
My zone for the count was Philly’s upper-northwest neighborhoods of East Falls, Manayunk, and Roxborough, mostly row homes with some big parks and a few commercial strips. Homeless people in these neighborhoods wouldn’t be on sidewalks in front of buildings, they’d be in alleyways or the parks. But the training materials explicitly stated "No parks, alleys, abandoned buildings, etc." out of concern for the volunteers’ safety. My team saw not a single homeless person. That does not mean they don't exist in the zip codes we covered, but that all the likeliest places where they would have been, we were not supposed to go.
Additionally, it was mostly by chance that a few of the people on my team lived in the area we were covering, but our team leader did not. He did not have guidance on where we should look, and it was our best guess as to where to go. It was a large area to cover, so we mostly drove around the commercial strips, looking for people at 24-hour laundromats and convenience stores, and alongside the parks, but we didn’t even venture into many neighborhoods. The lack of consistent, or any, methodology for covering this neighborhood means that meaningfully comparing one year’s data to the next would be next to impossible.
Finally, it was below freezing out, so those homeless individuals who could, I'm sure, were doubling up for the night. These are not long-term arrangements, but it gets them out of our sight (and HUD's definition) at least for the night. Philly has all-night homeless cafes in addition to shelters, so possibly the cold helped boost the numbers of more easily identifiable homeless people in shelter and the cafes. But on balance I think counting in January promotes an undercount.
The exercise was still useful, because it mobilized a large number of volunteers to help Project HOME identify many homeless people and connect them with services and it did collect some useful data. But the experience reaffirmed to me that the methodology (and lack thereof) creates unreliable and incomplete data - built in as part of the process. This would not be a problem if these inadequacies were regularly acknowledged, but as the numbers are tallied, I’m sure you’ll see headlines around the country touting increases or decreases in the count without any discussion of the methodology. That, I believe, does a disservice to those experiencing homelessness as it may prematurely diminish the sense of urgency needed to ensure we all have a safe, dignified place to call home.