[quinnchr]

Actually, only 35% of San Francisco's homeless population has some sort of Psychiatric condition. According to a survey of SF's homeless the biggest barriers to finding work (as of 2015) were:

  28% said no permanent address
  20% said alcohol or drug use
  17% said disability
  14% said age
  13% said need for clothing/shower facilities
  ...
  9% said mental health concerns

So it does seems like lack of housing is the largest barrier for SF's homeless population. Additionally, mental illness is most prevalent amongst the chronically homeless, which only make up about 25% of the homeless population in SF. The vast majority of homeless people are homeless for months not years, and the many programs to aid them in finding housing and preventing homelessness in the first place have a significant measurable effect on the average number of homeless shelter entries and length of stay.

Source: https://sfgov.org/lhcb/sites/default/files/2015%20San%20Fran...

[twblalock]

Asking mentally ill people if they are mentally ill is not a reliable way to determine if they are mentally ill. The same is true for alcoholics and drug addicts.

Many people will deny their circumstances because they are mentally ill and don't realize it, or are simply in denial, or because they think they will get more help if they hide their problems.

[quinnchr]

Yes this is true of any mental health issue in any population, we can only rely on self-reporting. If you have any sources that show "almost all" homeless people in SF are mentally ill, I'd be interested to read them.

From what I've read, at the national level the incidence of serious mental illness in the homeless population are around 20 - 25%. So the numbers for San Francisco seem reasonable.

[stickfigure]

we can only rely on self-reporting

There's another option - we can not rely on self-reporting and just acknowledge that we don't know. It's not satisfying, but it sucks less than creating public policy based on bad/misleading information. I doubt very much we can fix 28% of homelessness by handing out free PO boxes.

[quinnchr]

Except we do know, we have a pretty darn good idea. We know how many people self-report mental illness. We know how many people actually seek treatment for mental illness. We know the self-reported rate of mental illness amongst other populations and can compare the two.

It's literally like any other disease. We don't know how many people actually get the flu every year. We only know how many people self-report getting the flu, and how many people seek treatment. Self-reported surveys are used all the time in epidemiology and we manage to form effective public policy based on that.

So sure, there's definite bias when it comes to self-reporting, but your suggestion is to ignore what data we do have and form public policy based on what? How you feel? First principles?

[plussed_reader]

Somebody failed statistics...