Oh come now, that's a truly silly claim.

Social Security was designed to reduce senior poverty (it did work), but spends > $900 billion every year to continue working.

No you cannot spend $20 billion and solve homelessness or poverty. You probably can't even make a double digit % dent in homelessness or poverty for a single year with $20 billion, even with 100.0% efficiency of payout.

Statements like that totally mis-characterize the kind of problems you are putting on our intellectual plates. They aren't point problems like a leak in a gas line. In fact, they're closer to, ironically, a complex series of conflicts in the middle east...

> No you cannot spend $20 billion and solve homelessness or poverty. You probably can't even make a double digit % dent in homelessness or poverty for a single year with $20 billion, even with 100.0% efficiency of payout.

Solving poverty altogether is out of the question with only $20 billion to work with, but since you mentioned homelessness - if we assume there are 1.5 million homeless in the United States, $20 billion would provide $1,100 per month per homeless individual for a year. (Assuming 100% efficiency of payout)

What am I missing in this consideration?

(Also worth mentioning that a Housing and Urban Development official, in 2012, estimated the cost to end homelessness at $20 billion.)

> 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.

The absolutely astronomical cost of housing a homeless person, the amount of graft that goes into a y program designed for charity on that scale, and the knock-on effects on the housing economy by pouring that money in.

$1100 will buy you cheap rent and basic food. It will not buy you the ability to put our homeless into those houses or to run the kitchens to feed them. If you give it as direct subsidy, the money will go straight into the hands of drug lords. If you administer it, you will spend 50% more than that on just the administration costs of tracking down the homeless and offering them services.

The marginal value of a dollar decreases substantially as you give something more funding. We already have shelters and programs - they're ineffective at keeping people off the streets long term. Adding more shelters until there's always large quantities of empty beds will not help. Well meaning though it may be, it's giving a man a fish, one who doesn't want to be taught.

The city of San Francisco is spending 240m on its ~7k homeless so about 30k/y per. It doesn't appear the homeless problem there is getting better at all least being solved.

Right, but the context of the person I was replying to implied "100% efficiency" of the dollars spent. Is the $30k per year per homeless person in San Francisco actually reaching those people? Presumably that's how we evaluate efficiency - what percentage of the money spent actually reaches the cohort being targeted and not eaten up in administrative costs.

As far as I understand a significant amount of the money reaches homeless in form of various services. I doubt SF is just handing cash out though.

Estimated number of homeless in the US is around 500k (554k was the 2017 HUD estimate). That gives $35k-40k/year to spend on them if you only set aside $20 billion.

"solving" homelessness or world poverty is not a one-time expense. You can throw $82 billion dollars worth of food at the starving populations of the world and they'd just multiply and require even more later on, it's not that simple

OP's estimates are prices per year, which makes more sense. But $82 billion is only $11 per person per year; I find it hard to believe that this will be able to make a meaningful difference.

I find it really hard to believe $30 billion is all it would take to solve poverty world-wide.

I don't even see how $30 billion could solve poverty in the U.S. Think about it, if that was true, what the hell are we wasting all the this money on social services? Apparently all it would take is $30 billion. We spend almost a trillion a year in social services, and yet you think another $30 billion is all it would take to solve poverty?

USA spends more combined on welfare than on war per year, so then why isn't poverty solved yet?