~18,000
on the county's by-name list
early 2026, up from 14,361 a year earlier
+~412
net, every month
1,277 in vs 865 out
$1.3B
raised by the homeless-services tax
across the region since 2021
372
died homeless in 2024
mostly overdose; average age 48
Homelessness behaves a lot like unemployment: the number isn't fixed — it rises and falls with two flows beneath it. Every month some people lose their housing and others find their way back into it, and the total only grows when more are falling out than climbing back in. In Multnomah County about 1,277 people are added to the by-name list each month, and only about 865 find their way out (Multnomah County). That gap — not the total — is why it grows. Try closing it:
Close the inflow (cheapest)
Pay one-time arrears for a verified crisis; landlord made whole.
Stop releasing people from jail / hospital / foster care straight to the street.
Open the outflow
Lease existing apartments now — homes this year, not construction-years.
Detox & residential SUD beds — only count if you can staff them.
Workforce is the real rate-limiter — a bed you can't staff is a press release.
People on Multnomah County's by-name list, projected 4 years out. Do nothing and it climbs by ~412/month. Move the sliders and watch the scenario bend.
In 4 years, do nothing
37,776
people, up from 18,000 today
In 4 years, your scenario
37,776
move a slider to change this
Does growth stop?
No
still +412/month
What this scenario costs to run, per year
$0/yr
keeps the economic group from falling in
rehouses the economic & episodic groups
for the chronic & severe group
Doing nothing has a $0 program cost — and the list grows to 37,776. The street those people stay on is the most expensive option of all.
A simplified stocks-and-flows model with deliberately visible, contestable assumptions — how much of inflow is eviction- vs. discharge-driven, and how many people each treatment bed durably houses. The net flow varies month to month; the ~412/month figure is from January 2025 (Multnomah County). The cost panel uses rough but sourced unit costs — eviction prevention ~$2,500/household (National Alliance to End Homelessness), shelter and master-leasing from Multnomah County's own cost studies (Multnomah County HSD, Multnomah County JOHS), and staffed treatment at ~$55k/bed/year (French, Popovici & Tapsell, J. Subst. Abuse Treat. (2008)). The big lesson is structural: you can flip the trajectory by closing the inflow alone — before building a single new unit.
At first contact, the decision tree is practical: is there a crime, is there a mental-health hold, or is this a voluntary referral? The third branch is where Portland loses the moment. A person may say yes to shelter or treatment now, but the public system often still means calling lists and hoping the information is current. Official outreach actors include PSR, Portland Solutions, Fire CHAT, and others (City of Portland, City of Portland, Portland Fire & Rescue).
Legal authority is clear, but the back end only helps if court, jail, deflection, or treatment creates an actual service path.
Only available when the person is a danger to self or others or cannot care for themselves. Many visible street crises fall below that threshold.
This is the gap Bed Finder targets: if the person says yes now, the worker needs an eligible option, phone confirmation, hold, and transport before the window closes.
Who is in the field?
The response system is not one team. Public sources identify Portland Street Response, Portland Solutions, Fire CHAT, NW Community Conservancy, and ImpactNW Recovery Navigation as parts of the outreach and navigation landscape.
The loudest myth is that everyone on the street is the same — addicted, chronically homeless, beyond help. The data says otherwise: homelessness is really three populations that need three different things, and matching the wrong fix to the wrong person is most of what wastes money.
372
people died homeless in Multnomah County in 2024
214 from overdose (183 fentanyl-involved), average age 48. The first year-over-year decline since 2013 (Multnomah County Health Department / Street Roots) — the stakes behind the numbers.
People pushed out by a rent hike, a lost job, a medical bill — no serious addiction or mental-illness barrier. Often homeless for the first time and not for long.
Rapid rehousing — a unit plus light, short-term help.
≈ $8,500 per household/year — the cheapest, fastest fix, for the majority.
Park them in expensive permanent supportive housing and you burn scarce, intensive resources on people who didn't need them.
The twin errors — putting the economic group in expensive supportive housing, or the chronic group in bare rapid-rehousing — waste money and cycle people back. Matching the intervention to the person is most of the game.
The plain-language version of the transitional / episodic / chronic typology (Kuhn & Culhane, 1998). Per-person costs are national averages (National Alliance to End Homelessness); about 41% of Multnomah's homeless population was chronically homeless in the 2023 count (Multnomah County).
Housing First is a tool for the right population. It is not a substitute for same-day family shelter, youth-specific placement, detox, psychiatric stabilization, medical respite, jail-release bridges, safe parking, or repeated documented offers for unsafe encampments. A 90% reduction plan has to route each cohort to the first placement that actually fits.
Operating rule
A real offer is not a generic shelter referral. It is an available first step that matches the person's risk, urgency, legal status, health needs, family situation, and likely path out.
First placement
Motel bridge, prevention payment, landlord mediation, rapid rehousing
Capacity needed
Owners
Measure
Days homeless, Cost per prevention, Return rate
First placement
Safe parking, sanitation, vehicle support, housing navigation
Capacity needed
Owners
Measure
Unmanaged vehicles reduced, Housing exits, Sanitation incidents
First placement
Family motel or family shelter with school continuity
Capacity needed
Owners
Measure
Unsheltered family nights, School continuity, Housing placement
First placement
Youth shelter, host home, family reunification if safe, transitional living
Capacity needed
Owners
Measure
Adult-system exposure, School/work path, Stable exits
First placement
Confidential hotel, safe shelter, legal protection, relocation if needed
Capacity needed
Owners
Measure
Safe placement, Legal protection, Confidential housing exit
First placement
Sobering, withdrawal management, residential SUD, recovery housing
Capacity needed
Owners
Measure
Treatment access, Retention, Overdose/ER/jail reduction
First placement
Crisis stabilization, psychiatric evaluation, inpatient or residential care, ACT, PSH
Capacity needed
Owners
Measure
Stabilization, Medication continuity, Street returns
First placement
Dual-diagnosis stabilization and integrated residential treatment
Capacity needed
Owners
Measure
Dual-diagnosis access, Retention, Crisis events
First placement
Medical respite, accessible shelter, adult foster home, assisted living, supportive housing
Capacity needed
Owners
Measure
Discharge-to-street avoided, Readmissions, SNF/respite cost
First placement
Low-barrier shelter or village while documents, benefits, and PSH match are completed
Capacity needed
Owners
Measure
PSH placement, 6/12-month retention, Returns to homelessness
First placement
Specialty court, supervised diversion, jail-release bridge, treatment and housing
Capacity needed
Owners
Measure
Jail-street-jail interruptions, Compliance, Treatment/housing retention
First placement
Barrier-specific offer with pets, partners, storage, alternative models, and legal path if dangerous or incapable
Capacity needed
Owners
Measure
Real offers made, Refusal reasons, Public-space resolution
Fixes look expensive against a baseline of zero. But the street isn't zero: it's emergency-room visits, jail stays, ambulance runs, and sanitation — spread across a dozen budgets so no one sees the total. Study after study finds it costs more to leave a chronically homeless person on the street than to house them. Move the slider:
The street-cost figure is a central estimate; studies range from ~$35k/year (National Alliance to End Homelessness) to far higher for the costliest individuals. Housing-is-cheaper holds across cities: Los Angeles found a 79% cost drop (Economic Roundtable), Utah 91% (NPR).
ER, jail, EMS, sanitation — spread across a dozen budgets
rent + case management in supportive housing
Doing nothing costs more
$10M / year
That's about $20,000 per person, per year that the street costs over supportive housing.
The honest caveat: most of that “saving” is federal health spending (Medicaid-funded ER and hospital care), not the local budget. The fix isn't that housing pays for itself locally — it's pulling the federal payer in to fund the solution that saves it money. The street figure ($40,000) sits in a national range of $35,000–$50,000/year.
Multnomah County's deflection program sends eligible drug-possession cases to the Coordinated Care Pathway Center at 980 SE Pine (Multnomah County). That may be better than a citation-only response, but the outcome definitions matter. The FY26 Q3 snapshot reported 79 law-enforcement referrals and 9 successful 90-day completions among those who reached the window (Multnomah County).
79
Law-enforcement referrals
FY26 Q3, Jan. 1-Mar. 31, 2026.
21
Reached 90-day completion window
The denominator for Q3 90-day completions.
9
Successful 90-day completions
Under the January 2026 completion definition.
1
SUD/recovery only
One completion was in the SUD/recovery-only bucket.
7
SUD/recovery + care coordination
Seven combined SUD/recovery access with sustained PATH follow-up.
The correction matters: in FY26 Q3, do not say success only meant a food pantry or shelter night, and do not say only one person got treatment. The snapshot says one successful completion was SUD/recovery-only, seven combined SUD/recovery access with sustained care coordination, and one was care-coordination-only. It still does not prove residential treatment completion.
The FY25 annual report used a broader completion definition that could include accessing at least one recommended service (Multnomah County). The January 2026 definition changed, so the Q3 numbers need their own explanation.
Here's the throughline made literal. Oregon is short roughly 3,714 treatment beds (OHA / Public Consulting Group) — but it can't even tell you how many of the beds it has are open tonight. A multi-million-dollar bed registry produced only a handful of placements (Willamette Week), because a database isn't a coordination system.
To send someone to a treatment or shelter bed tonight, you need to know it's open tonight. But the public data mostly tracks reports, dashboards, licensed capacity, or program lists. The three rungs that matter to a worker in the field — is it staffed, is it occupied, is it open right now for this person — still go unreported.
City shelter dashboards are useful for participation, utilization, and outcomes, but they are not the same as a universal live placement feed. So a worker in the willingness moment still makes phone calls on stale information and can lose the person — even when a suitable bed sits empty. The cheapest new bed is the empty one you already own but can't see.
We didn't just write this down. Bed Finder is a live tool that answers “where can someone go right now?” — matching a specific person to the beds they're actually eligible for, with facilities self-reporting real openings. It's the coordination layer Oregon doesn't have, built to prove it can exist.
Portland needs a fuller ladder. Overnight shelter, 24-hour shelter, villages, treatment, jail-discharge shelters, hospital step-down shelters, structured recovery cohorts, and supportive housing all do different jobs. City shelter dashboards track participation, utilization, and outcomes (City of Portland), but the policy question is whether the models connect into a real path out.
Immediate bed for the night
Daytime street exposure remains; live open-bed status is not universal.
Stability, meals, daytime access, and service connection
Works best when structured activity and case management are real.
Low-barrier private sleeping space
Can become a dead end without routine, treatment, work, or exit pathways.
Treat addiction as the binding constraint
Provider lists exist, but facility-level open bed counts are not public.
Reentry, court, treatment, documents, work placement
Mostly a proposal locally; needs program and records-request validation.
Safe discharge with medical knowledge on site
Could reduce street discharge and high-cost skilled nursing overuse; cost claims need verification.
Stable housing with the right service intensity
Fails when used as the only answer or when isolated people return to encampment community.
None of this is a mystery. The plan that survives a hostile hearing is sequenced by what's cheapest and fastest first — and the iron rule underneath it all: you can't move someone off the street faster than you can build somewhere to put them. Capacity precedes enforcement, always.
The cheapest 'reduction' is the person who never becomes homeless. Use time-limited eviction prevention for verified financial crises, paid directly to landlords, while preserving tools to remove dangerous or predatory tenants. Then stop institutions from releasing people from jail, hospital, or foster care straight to the street.
When someone says yes right now, a worker needs an eligible option in minutes: anonymous criteria, live or phone-confirmed availability, name check by phone, hold, transport, and outcome. That is the product gap Bed Finder is built to close.
Portland needs more than shelter vs. apartment: overnight beds, 24-hour shelters, detox, residential treatment, opioid treatment, jail-discharge shelters, hospital step-down shelters, structured recovery cohorts, and supportive housing. Each has a different job.
Deflection, outreach, and shelter programs should report the real funnel: referral, engagement, service type, treatment admission, shelter arrival, housing exit, and retention. A contact is not a placement, and service access is not treatment completion.
Scattered-site housing and permanent supportive housing remain essential. But housing is one tier in a continuum, not a substitute for treatment, reentry, hospital step-down care, or structured recovery community.
An honest broker has to hold the strongest version of every objection. Here are the four that come up at every council meeting — and what the evidence actually says.
“Housing First means no rules — that's why it fails.”
Housing First removes the preconditions to *qualify* (you don't have to get sober first) — NOT the rules once you're housed. Tenants still sign standard leases and must meet ordinary obligations. The model sustains ~85–90% housing retention; the out-of-control-building failures are management and over-concentration problems, fixed by staffing and scattered-site placement.
“Housing the homeless pays for itself.”
Overclaimed locally. The biggest cost savings are *federal* Medicaid (ER and hospital care), not the city or county budget — Portland and the county are mostly on the hook for jail, EMS, sanitation, and the homelessness budget. The honest fix is pulling the federal payer in, not pretending it nets out for local taxpayers.
“Just force the addicts into treatment.”
You can't punish a status, can't force treatment without due process, and can't mandate people into beds that don't exist. The legal, effective version is a real treatment pathway: drug courts, deflection that actually reaches SUD care, pre-release planning, and narrow civil commitment where legally justified.
“We spend over a billion dollars and nothing changes.”
Spending is real — but a balance that peaked near $431 million sat unspent across fragmented budgets while the system couldn't see itself, so effort flowed to the visible lever (units built) instead of the binding one (closing the inflow, staffing beds). The highest-leverage fix is making the machine legible.
“Housing First is either the answer or the problem.”
Wrong frame. Housing First is a strong tool for people whose binding constraint is housing instability or chronic disability with services. It is not a detox bed, a jail-reentry plan, a hospital step-down unit, or a recovery community.
Housing First's ~85–90% retention and the entry-vs-rules distinction: National Alliance to End Homelessness; the aggregate-vs-individual critique: Manhattan Institute (Stephen Eide).
Headline figures are read from primary sources — the county's by-name dashboard, the Point-in-Time count, Metro's tax reports, the state's bed studies — and were re-checked in June 2026. The flow model is a simplified teaching tool with visible assumptions. Where a popular figure didn't hold up (a $500M+ unspent balance, a flat '$50k per person'), it was corrected, not repeated.