To solve a problem, first you have to know the problem's scope. That's the idea behind an initiative to count the homeless begun three years ago by Philip F. Mangano, the Bush administration's homelessness czar. The New York Times reports that the nationwide estimate of homeless people on any given night is around 750,000, 20% of whom are "chronically homeless." The others may be people who've been turned out by a scorned spouse or lover, are sobering up from a bender, or simply mistaken for homeless by the rather unscientific tallying process.

Unscientific to a fault. According to an NPR interview last May with Mangano, the estimate is closer to 2 million, with a chronically homeless rate of 10%.

Such a discrepancy is easy to understand when you consider the methods in place, not to mention the difficulty of identifying someone as homeless. In January, 60 volunteers banded together to count the homeless population of northern Kentucky. They tallied just over 2,000 homeless people in several counties. But who's to say that some of those homeless people weren't merely college students?

This might sound like a joke, but the Times article describes exactly such a case during a recent counting night in New Haven, Connecticut.

Once, when I was working at a fairly expensive Manhattan restaurant, I spotted a man sauntering letharigically toward the bathroom whom I instantly took for homeless. After all, this restaurant catered mainly to financial executives and "ladies who lunch." By comparison, this guy just seemed homeless. He was huge, with ratty-looking, oversized camouflage pants, a black t-shirt that draped over his massive build, and an enormous, disheveled head of hair. I remember thinking, this is awkward... is anyone going to say something? Of course, I was at work, so I was more inclined to worry about such things -- and despite our usual clientele, it wasn't uncommon for pan-handlers, drug addicts, and crazy people to filter in off the streets, whom the managers would swiftly -- but politely -- usher to the door.

It turns out he wasn't a homeless man, but the drummer for very successful band. (I mean, with a few notable exceptions, how often can you instantly identify the drummer of a band?) I hadn't seen him in the restaurant until that moment simply because he had requested a secluded table to conduct a business meeting. Needless to say, I felt like an idiot.

But it proves that appearances can be deceiving. Let's hope the 4,000 communities participating in this nationwide count are more discerning than I was on that embarrassing day back in 2003. An accurate number could mean the beginning of some real measures to help get the unknown number of homeless men, women, teenagers, and children off the streets for good.