Do NYC and LA constitute the majority of this nation's homeless?

Even though this data (bottom) was generated by HUD, this seemed a little "odd" to me. So I decided to take a deeper dive (with help from Claude).

First off, NYC and LA make up about 1/3rd of all homelessness in the nation. Lot's of folks elsewhere. That said NYC leads the way because they have a right to shelter mandate. Essentially they are obligated to provide a bed to anyone that asks. That provides more shelter and ensures more people get "counted".

LA doesn't have such a mandate but has far better weather which attracts more folks for obvious reason. I've produced with Claude some better charts which are more nuanced (and don't miss Chicago which is #3).

Bottom line: NYC spends about 3-4% of it's annual budget on the homeless. They are the elephant in the room without peer.

Housing cost is the master variable. The single best predictor of a region's homelessness rate isn't poverty, mental illness, addiction, or even weather — it's rent levels and vacancy rates. Expensive, supply-constrained coastal metros (New York, LA, the Bay Area, Seattle, San Diego) have the worst rent-to-income gaps and the tightest vacancies, so more people get tipped over the edge and have nowhere to land. Cheaper-housing metros have a release valve. Chicago's own PIT report makes this point explicitly, noting that research points to structural housing market conditions, particularly rent and vacancy rates, as major factors in the regional variation. This is why a chart of homeless counts ends up looking a lot like a chart of metro housing costs. City of Chicago

Climate sorts who's visible and who gets counted at all. Mild West Coast winters let people survive outdoors year-round, which inflates the unsheltered counts that dominate this list — six of the nine cities are West Coast. Cold-winter cities like Chicago, Detroit, Minneapolis, and Boston push people indoors, either into shelters or, more importantly, into "doubled-up" arrangements (couch-surfing with family or friends). And here's the catch: HUD's point-in-time count only counts people in shelters or in places "not meant for human habitation." It does not count the doubled-up. So in much of the South and Midwest — where housing instability more often takes the form of crashing with relatives — a large share of the housing crisis is simply invisible to this number. The cities lower on the list aren't necessarily doing better; their distress is often in a form the count misses.

Cost to house and provide homeless services.