So I took a look at that. First off it only talks about mentally ill homeless people, not all homeless people. And they looked at all crimes. If a homeless person is arrested for homelessness (i.e. vagrancy) they end up with a higher arrest rate - but that higher arrest rate for the crime of vagrancy does not increase crime within nearby communities. I have no doubt that homeless people are arrested more often; they are doing something that is technically illegal (vagrancy) and any crimes they commit in their "homes" are quite visible out on the street. Nor am I surprised that homeless people have a higher base rate of crime (not just arrests) since higher rates of crime are known to be associated with poverty. But this does not support your argument that "a homeless camp near a residential neighborhood is dangerous for that neighborhood." I mean, black men have a far higher arrest rate than white men. Even so, I suspect you would not argue that black people are dangerous for nearby white neighborhoods.