Hell, the REAL ID act is technically illegal, but the feds can use the "we won't let you board a flight without one" stick to force compliance.
The federal government gets its paws into whatever it wants to. It simply doesn't care about homeless issues.
I do agree that, on its face, homelessness seems like a textbook local/state problem. However, any city/state that tries to solve the problem with money and services will be a magnet for homeless people from outside the area, which can easily overwhelm the services provided. How do you fix that?
I'd say this is a textbook border-controlling-entity-level problem...
that's such a non-sequitur. so is the gist that we just accept the federal government overstepping its bounds? that's sounds so defeatist. instead, let's put even more pressure on (all of) our representatives to contain the overstepping.
> "...magnet for homeless people from outside the area, which can easily overwhelm the services provided. How do you fix that?"
that's worth keeping in mind, but there's no compelling evidence that that's going to be a problem. an optimist might posit that it's just as likely that micro-economies develop that allow the former homeless to lift others out of homelessness without further intervention. it depends on how things are implemented of course. let's take a step forward and (continually) assess.
the state is the canonical unit for a problem like homelessness, and each state is free to delegate to local jurisdictions or to coordinate with other states on solutions.
by the way, san francisco has about 8000 homeless folks, and LA has ~55,000 (~40% of the 130,000 homeless in the state). if we wanted to build new housing for all of these folks, we'd need about $50B (state budget is about $200B/year). LA, for instance, spent $0.6B on homelessness last year, or about 3% of what is needed just in LA.
we either continuously pay for homelessness indirectly (and ineffectively), through police, crime, social services, drugs, non-profits, governmental programs, emergency services, etc. or actually tackle the problem directly by raising and allocating that $50B.
(obviously talking about first-order effects here. this is not an in-depth policy paper.)