> What am I missing in this consideration? If you give $2000 a month to every homeless person in say, San Francisco or Dallas or Smalleville, Nebraska: * How

> What am I missing in this consideration? If you give $2000 a month to every homeless person in say, San Francisco or Dallas or Smalleville, Nebraska: * How many of them can afford an apartment with it? * How many landlords will sign them? After all you are giving them money, not proof of income, and landlords are often particular about who they rent to. * How many of them will decline to look for shelter with the money? My point is that economic considerations like this often under-estimate the amount of footwork needed to make money actually useful. You would need to do a lot more than give them rent money, you'd need to move them to lower-rent places, find landlords that would rent to them, or spend considerable resources trying to convince them to even do anything with the money that involved housing. Programs like medicaid are costly because they provide more than just money or medicine; they provide case workers and associated footwork—and those things are needed. There’s a much higher no-show rate among medicaid patients than other cohorts. They don’t sign up, they’re afraid of doctors, they don’t have addresses, and so on. The problems some groups face are deeper than others, and helping those groups may require difficult and not-very-efficient overhead. It’s overhead that is still worth going through to become a better functioning society! But for a statement like "$20 billion can cure homelessness" to be correct, you'd have to believe its not necessary. This kind of work (what medicaid and medicare workers are doing every day for example) is very, very necessary, if you wanted to make a dent in the homelessness population. That is what you are missing in the dollar figure consideration. Why they are homeless, where they are homeless, and the extent, which goes well beyond renting a flat, of what it takes to make a homeless person non-homeless.