Denver gave homeless people $1k/mth. Year later, nearly half had housing

Original Article ↗ Hacker News Discussion ↗

Study design and control-group debate

  • Major contention over methodology: no group received $0, so many argue there is no true control group.
  • Others treat the $50/month group as a de facto control or “placebo,” but this is disputed because $50 is still a real intervention, especially for homeless people.
  • Participants were recruited via service agencies, so the sample skews toward people already seeking help, not the full homeless population.
  • Some commenters see the study as preliminary and underpowered for firm conclusions; others criticize it as designed to produce positive headlines.

Housing and employment outcomes

  • All three groups showed large increases in being housed (roughly mid-40% housed after a year), with minimal differences between $50, $1,000, and lump‑sum+$500 groups.
  • A key nuance: the $50 group started with more people already housed, so their improvement (delta) may be smaller.
  • Employment effects were more differentiated: $1,000/month saw sizable employment gains; the $50 group reportedly showed no change or even a drop in full‑time work.

Interpretation and base rates

  • Several note that most homelessness is transient; many would exit within a year anyway, so you must ask “compared to what?”.
  • Because there’s no $0 group, it’s unclear how much of the improvement is due to cash vs natural churn or other supports.
  • Some see the similar housing rates across payment levels as evidence that more cash (in this range) barely improves housing attainment.

Costs, public services, and ROI

  • Reported “public service savings” are questioned; cost declines might simply reflect people exiting homelessness over time.
  • One reading of the appendix suggests total public service costs may have increased modestly, though interpretation is unclear.
  • Others argue even expensive programs may be preferable to current spending on policing, incarceration, and emergency services.

What actually helps: cash vs housing vs services

  • Several claim unaffordable housing is the main driver; therefore “just give housing” is argued as the straightforward fix.
  • Others counter that for chronically homeless people with severe addiction or mental illness, housing alone often fails; problems like property destruction, crime, and drug use in projects are cited.
  • Some advocate institutional care for the most impaired, but acknowledge it’s politically difficult.
  • A recurring theme: attention, trust, and ongoing support (Hawthorne‑like effects) may matter as much as the cash itself.

Implications for UBI and policy

  • Thread is divided: some see evidence that even small, unconditional cash can improve well‑being and employment.
  • Others note that if outcomes are similar from $50 to $1,000, then “just give more money” is not the magic lever for housing.
  • Several distinguish targeted pilots from true universal basic income, emphasizing that UBI would avoid means‑testing but might change prices and incentives.
  • A few suggest focusing more on preventing homelessness (upstream) rather than trying to reverse it once entrenched.

Moral and societal views

  • Some insist any aid is worthwhile and humane; others argue that if success rates are far below 100% or 90%, programs are failing or easily gamed.
  • There is pessimism about ever fully eliminating homelessness, but also discomfort with current inhumane conditions.