Hacker News Reader

Comment by slifin | Hacker News Reader

600k per year per person?

With that amount of funding you may as well buy every homeless person a house

This number is horribly bias and misleading. They are using the money that goes towards all homeless programs- including programs that keep people housed- but then dividing it by the number of shelter beds, despite the fact that the vast majority of the money is not spent on shelters.

It's like taking the entire school budget, dividing it by the number of first graders, and then saying that's how much we spend per student. It's stupid math that is only used to mislead.

That's a pretty serious claim against WSJ and it's editors. Since they have a lot of credibility on HN with users and moderators, do you really think they are intentionally misleading readers?

There are places where you could buy folks many houses for 600k.

What you need are a sort of anti-vagrancy law where the "punishment" is confinement in a very minimal security home where your basic needs are met, folks are nice and caring to you, and you're gently compelled to take whatever steps you need to put your life back together. Some people will stay indefinitely, be allowed to, and choose to.

This would increase the supply of housing, lowering the value of real estate investments, and thus is not permitted. Spending on homelessness in places like California is essentially a corrupt form of taxation on insanely inflated housing prices, where the money gets funneled back through various “nonprofits” and “services” that line the pockets of politicians.

It's not 600k per year per person. It's 606k additional spending since 2016, per additional sheltered individual since 2016.

From what I can tell, SF had about 18,548 shelter beds in 2022[0], which puts it at $36,230/bed/year. In 2016, there were 11,017 beds[1], which gets you $20,332/bed/year.

And of course not all help for homeless people is just paying for housing. Still seems expensive, but not nearly as bad as your math indicates.

[0] https://files.hudexchange.info/reports/published/CoC_HIC_CoC...

[1] https://files.hudexchange.info/reports/published/CoC_HIC_CoC...

This is one of those terrible, misleading statistics. Most of the "homelessness" budgets in large cities go towards preventing people from becoming unsheltered - i.e. on housing support and prevention.

Two-thirds of San Francisco's homelessness budget is spent on housing and prevention: https://hsh.sfgov.org/about/budget/overview-previous-budgets...

By definition, the best spent homelessness budget would be the one that resulted in zero people in shelters right?

> By definition, the best spent homelessness budget would be the one that resulted in zero people in shelters right?

Well, no, otherwise “have no shelters” would be the cheapest optimal strategy.

I don't understand the point you're trying to make.

Fundamentally, the denominator that we ought to be using is "number of people diverted from being unsheltered", not "number of people in temporary shelters".

If a city spends $100M on homelessness and prevents 9,900 people from being unsheltered by housing them and only keeps 100 people in shelters, that looks like $1M per person on the bad metric.

But that situation is obviously better than spending $100M on sheltering 1,000 people (leaving 9,000 on the streets) instead of providing housing support, albeit that the cost looks only like $100K per person that way.

I think the point was that there are ways to have nobody in shelters without it being a well spent budget, and practically we don't know that there is a way to spend realistic levels of funding such that no one is in shelters, so while it may be true of some ideal policy it might well not be true of the best real policy.

I think it's a (marginal) nitpick over your wording, not intended as a defense of the metric you were criticizing, though of course dragonwriter can correct me on that if needed.

One of the problems is that homelessness is a flow problem, not a stock one. If there are X homeless people in a year, then next year something like 50% * X will be newly homeless. So you could house everybody currently homeless, but you’d be back to square one shortly.

Nevertheless, the current strategy does not seem particularly effective.

Right - transient homelessness caused by unexpected and temporary loss of a living situation is the most common form. A large proportion of people in homeless shelters are employed, and will quickly leave the shelter.

i.e. there's a big difference in sensible remedy for a family that's in a shelter for a few nights after fleeing domestic abuse vs. chronically unsheltered, mentally ill

Most of the negative interactions people have are with the unsheltered (i.e. someone street sleeping) rather than the homeless, and it is often unclear even how cities would/should spend money to address those people - that's what Mayor Breed is really complaining about.

Well, not in San Francisco, and not anywhere in the surrounding area without driving house prices even higher.

(Now, building every homeless person a house might be more viable.)

Still doesn’t deal with the fact that much of that is nonshelter costs that would still be needs, even if they would move out of the homelessness budget.

This strategy has in fact been tried in a few places and while it's not a cure-all, it seems to work pretty good.

Low income housing projects have been tried in the US too with mixed outcomes. My guess is that the Baseline social behavior and expectation is just lower in the US than in places like Taiwan or Singapore, where it is incredibly successful

What's frustrating is "mixed outcomes" leads to "oh, the whole idea is bad" which then leads to "let's not do that, or try anything remotely resembling that". Low income housing projects are the answer, there's just a lot of cultural changes and reframing and winning hearts and minds; a lot of that sort of work to be done before housing projects are politically viable.

Yeah, it works in some contexts.

I think it would be an interesting experiment to try a planned project with high support, high requirements, and zero tolerance for crime. Instead of expecting the worst from people expect the best and hold them to a high standard.

Imagine a Gated community where everyone is expected to be in rehab, education, or employment and you get kicked out if you fail to follow through or commit crimes.

The point wouldn't be to a provide a catch-all support for everyone, but provide a safe space and support for those working to improve themselves.

Seeing comments like this get downvoted into oblivion when they're stating basic information without any obvious slant or other violations of the HN rules/norms is really bumming me out.

Reply on news.ycombinator.com