the way they're right, I fell down some rabbit hole about it but tldr: over a 50 year timespan, you have a 3x higher chance of becoming homeless than u do of becoming a millionaire people in this thread believe this misleading "snapshot of any given night" cope but this doesn't show probablistic outcomes over time.. I modeled out something to illustrate this, using markov modeling (where someone is now vs where they'll be next year) with this "transition" state matrix: • Housed (A) • hidden homeless (RV, car, any building not intended for residency, couch surfing, hotels, shelters) (B) • street homeless (C) • dead (D) We can set a baseline probability based on 50% of adults in US not being able to afford a $400 emergency and assign a "buffer rate" for state B (e.g. gets evicted, stays w/friend) A "Homeless trap" rate is assigning a likelihood a person is to go from state B to C. We can also more confidently assign a death rate (from B/C to D) since we have disease of despair and partial homeless death data vs typical mortality data by income bracket, so I assign an attrition line for it. If we run a simulation for 100k ppl over a 50 year timeline we get: 1) extended periods in extreme poverty act as an exit vector via premature death: it is 450% more lethal to be homeless than not, which tracks with empirical data. This keeps the number of nightly street homeless low. 2) The picture over time is that u will see ~1-3% of US population physically on the street each night, but over a lifetime ~30% of the population will experience being homeless. 3) the permanent underclass: 92-95% of millionaires now will be millionaires one year from now. However if we look at HUD data for episodic homelessness, if you experience it once, there is an 18% chance it will occur again, once you are classified as "chronically homeless" which is 4 episodes in a 3 year period, you have an 80% likelihood you will remain homeless. 4) the gap between these two isn't nearly as large as some claim it is, bc of our death variable & unseen homelessness states. 5) This model holds the same trend as other models have: the number one predictor of longevity is wealth accumulation. Not a cheery thought but I thought I would point out that sometimes people's vibe reads are entirely correct even if a snapshot statistic disagrees with it.