Hacker News Reader

Comment by davidw | Hacker News Reader

It is a housing problem though. That is what the data show. That's why homelessness correlates so much with high housing costs. That's why places with serious drug problems like Ohio or West Virginia don't have such high rates of homelessness.

None of which is to say that drugs are not a problem, or to say that someone with a drug problem isn't more likely to become homeless.

All of that is explained in the article which you responded to without reading it, judging by the timestamp on your comment. Let alone the video, which goes into more depth and talks extensively about drugs.

View on HN

Places like Mississippi don’t have high homeless problems because they lack services and good weather, so drug addicts who have burned their family/friend networks, at least, hop on a bus to somewhere they can survive (usually a west coast city with mild weather and a progressive voter base that is willing to throw lots of money at the problem). That segment of the problem is 100% mobile, and even self reporting surveys show cause (like 60% of king counties homeless population saying they had previous housing in pioneer square).

If you can’t afford housing anywhere, you just won’t survive very long in lots of places, so survival bias starts to show up.

This is a great narrative, but it turns out that it's not true.

https://www.theatlantic.com/ideas/archive/2023/07/california...

> The overwhelming majority of homeless people surveyed were locals, not migrants from far away: 90 percent lost their last housing in California, and 75 percent lost it in the same county where they were experiencing homelessness. Of the 10 percent who came from elsewhere, 30 percent were born in California. Most of the others had familial or employment ties, or had previously lived in the state.

Again, those are self reported surveys with biased surveyors (who want to push the narrative you believe in) with very weird results. They are basically trying to convince us that the unhoused population is more local than the housed population, sometimes by an order of magnitude, and it really isn’t statistically possible. Something like 75% of the general population are transplants but only 25% of the unhoused population are?

It's unclear to me why this is isn't statistically possible. Could you please explain?

But anyways, this article (paywalled) claims that 48% of California adults were born in the state as of 2022: https://www.sfchronicle.com/projects/2024/california-populat...

In SF and LA those numbers are nowhere near to being 50%. And I guess the wording is important, since many people can crash on a couch for the first few months and then are considered previously housed by the self reporting survey question.

The survey was done across California -- not just SF and LA. Either way, wikipedia claims that 37.7% of SF residents were born in California which is slightly closer to 50% than the 25% figure you quoted (https://en.wikipedia.org/wiki/Demographics_of_San_Francisco#...).

The survey also states that "Most participants (87%) were born in the United States. ... Two-thirds (66%) were born in California."

I still don't understand why the survey is statistically impossible.

Reply on news.ycombinator.com