You need to be very careful when accepting any statistics as true. And if there is only one study of its kind (so you can't compare results), then in my opinion it's right to be concerned by the results, but it is wrong to take any action that may have a damaging effect or side-effect.

You bet that groups with axes to grind, report the heck out of the stats that support their aims. This is not limited to "social programs" or "liberals".

But what I really wanted to say is it is really really hard to design a good experiment or poll that will give completely unambiguous results. Some of the smartest minds in the world have come up with numbers that did not hold up. Are some of the datapoints in this editorialist's "statistics" the result of honest folks being wrong?

(which homeless statistics, and why do you think they are incorrect?)