Here Come The Counters Again

Each year during the last week in January, "the counters" gather in cities and towns around the country. A makeshift army of volunteers and service providers descend on shelters, libraries and empty parking lots during the wee hours of the morning to count the citizens of their community who are enduring homelessness.

Most of these counters, having worked all day, are by now bleary-eyed and tired, bodies and mind grappling with the odd request to be awake, alert and energized at a time when they're usually asleep in the comfort of a real bed. Between yawns, they strike out into that cold dark night, searching for others who are themselves hoping to grab a few winks between the countless hassles, dangers and problems they encounter each night while living on the streets.

While the need for accurate data concerning the number of our community members who are forced to experience homelessness is important for many reasons, two major problems arise from this annual "point in time count, (PITC)."

The count got it's start back in 2005 thanks to the U.S. Department of Housing and Urban Development's (HUD) effort to provide government and agency officials with an estimate of just how many people were experiencing homelessness on a given night. Essentially, Congress got tired of throwing money every year at a situation they had no hard data on and tasked HUD with requiring those communities receiving McKinney-Vento Homeless Assistance Grant funds to conduct a point-in-time count at least every other year.

HUD requires these counts be performed sometime in the last seven days of January. The numbers obtained in the count, they surmised, would help paint a better picture of the homeless situation across America, as well as bolstering the Homeless Management Information System (HMIS) data, which Congress had also asked HUD to have up and running nationally by 2004.

The first problem arises with those who perform the count. While in most areas teams are often headed by an outreach specialist or "skilled" volunteer, large numbers of well-meaning but inexperienced volunteer counters make up the bulk of many the teams. A significant proportion of these volunteers have little idea what to look for, are only vaguely familiar with the areas they are tasked with "searching," and can't see beyond the edge of their flashlight beams anyway. Many are also fearful of what they might encounter as they wander through places no one in their right mind would ever venture into at 3 o'clock in the morning.

It's usually colder than a well digger's arse on any night just about anywhere in the country in January, and many volunteers realize too late that they should have passed on this particular "exciting opportunity" at serving their community. In the true spirit of volunteerism, however, they make the best of it, zip through the "zone" they are supposed to cover, and miss just about everyone they haven't stepped directly on. I remember one team several years ago who actually drove their area, never once getting out of the car because it was "too cold." Although they patrolled an area often frequented by people experiencing homelessness, they turned in a zero on their count form.

The second problem is with the folks living on the street, many of whom know the count is coming and often do their best to remain undetected for a variety of reasons, not the least of which is that police officers have been known to magically materialize right on top of their sleeping spot just a night or two after the count. Hmmm.

Some don't want to be counted for personal reasons unfathomable to the rest of us. Some are intoxicated or engaging in criminal behavior and don't want to be discovered. Some don't like being awakened by people they see as altruistic nose-pickers bent on saving the world one homeless individual at a time. Still others are suffering from a mental illness and/or a co-occurring substance use disorder and are frightened half to death at the sudden intrusion into their space by a pair or team of shadowy figures rapidly approaching while wielding flashlights.

Understand here that my intent is not to malign the countless conscientious volunteers who brave the cold and the danger of the streets and do an excellent job with little or no training for a very difficult task. But in a study done in New York City and published in the American Journal of Public Health in August, 2008, Estimating Numbers of Unsheltered Homeless People Through Plant-Capture and Postcount Survey Methods, researchers placed "planted" people in local shelters and at areas on the street known by count coordinators to have heavy populations of homeless individuals. The study concluded that in shelters, the "plants" were missed 29 percent of the time, while on the street it was much worse; between 39 and 41 percent of the "plants" were missed by counters.

That's a third of folks in shelters, and damned near half of those on the street who were missed in the official point in time count in New York City. If this even remotely represents a national average for missing individuals during homeless point in time counts, it's probably safe to assume that we're underestimating the actual population of people experiencing homelessness by around 35 percent. Translation: whatever the final count number turns out to be, take a third of that number and add it in for a more realistic representation of the actual number of persons on the street that night.

Let me be clear here. The volunteers are not at fault or somehow doing a bad job. Rather, the concern is the way in we perform the count, and therefore, I'm suggesting a minor change in the way we perform the point in time counts.

To me, counting at night is just a bad idea overall, regardless of what time one may start the effort. Therefore, if we were to simply move the time the count starts forward to ensure the start coincides with daybreak in each area, I believe we will see a significant increase in the accuracy of the count. Several additional benefits may also become immediately apparent. Volunteers would almost certainly be more refreshed, alert and energetic with at least a partial night's sleep under their belts. They may also be more apt to enter dangerous or unsavory areas when they can see their surroundings clearly. The expanding daylight should also assist them in identifying sleepers more effectively, especially if folks are beginning to stir. Additionally, they will not be rudely interrupting an individual's sleep and may find responses to their questions more palatable. This alone would make the count a better overall experience while promoting engagement and interaction with those experiencing homelessness. This interaction may help substantially in dispelling stigma and myths associated with homelessness. Volunteers may also be more inclined to thoroughly explore their zone, since daylight usually brings at least a slight increase in temperature and a definite increase in the safety factor for everyone involved.

True, duplication of counted individuals, perhaps the biggest reason cited when explaining why the count is done at 3am, may increase slightly. However, it's unlikely that it will be so significant as to severely skew the count, especially if counters are able to provide physical descriptions of the counted, a real possibility now, since there would be enough light to actually see the person being interviewed. And frankly, at this point, if the New York City study is any indication, we're missing about 40 percent of the folks outside anyway, so getting an occasional double-count on someone is hardly going to make a difference. Furthermore, as HMIS data becomes more accurate, matching count tallies with this data will provide a better reflection of the true number of people actually on the streets and make the issue of duplicity a moot one.

As for increasing the accuracy of the shelter counts, I'll leave that to the good people who are on the inside, since they know their world better than I. It does seem to me however, that if they were to simply station counters by the doors as people exit in the morning or arrive in the afternoon, they'd most likely increase their accuracy, as it would be pretty hard to miss 29 percent of people passing directly in front of you.

Have you ever participated in an annual point in time homeless count? Based on your experience, do you feel it would make more sense to start the count at daybreak rather than in the early morning hours?

Photo: Steven Samra