The short version

Claim we investigated

King County's 96% 'stay housed' rate for supportive and other permanent housing leaves out residents who die.

True. A federal rule drops residents who die from the rate; overdose deaths alone would lower it to roughly 94%. A federal housing department rule leaves anyone who dies out of the region's 96% rate. By our rough estimate, a year of overdose deaths in supportive housing alone is about three-quarters of the people it counts as not staying housed.

Quote or cite this fact-check
FactHub rates the claim “King County's 96% 'stay housed' rate for supportive and other permanent housing leaves out residents who die” True (high confidence): A federal rule drops residents who die from the rate; overdose deaths alone would lower it to roughly 94%. Reviewed November 12, 2026. https://facthub.org/n/article/king-county-stay-housed-rates-deaths/7

Key points

What the evidence establishes

  • Established. The region's official 96% 'stay housed' rate follows a federal rule that removes residents who die from the count entirely.
  • Established. The regional homelessness authority, which reports the 96%, doesn't mention that deaths are left out.
  • Established. In 2024, 225 people died of overdoses in supportive housing; the 96% rate counted 299 as not staying housed (2022-23).
  • Context. By our rough estimate, counting a year of overdose deaths would lower the 96% to about 94%; counting all deaths at subsidized housing sites, about 91%.
  • Not shown. Other local rates, including 98% figures from the authority and from DESC, a large nonprofit operator, don't say how they count deaths.
  • Context. The authority's 93% dashboard rate counts every household it served in the period, so as written it may count deaths as not staying housed.
Important context: Supportive housing is long-term rented housing with services for people who were homeless for a long time and have disabilities. Agencies report how many residents stay housed: still living there, or moved to other permanent housing. A common criticism is that these rates, often 93% to 98%, hide residents who die. We checked the region's official 96% rate, which it reports to the federal government and which covers supportive and other permanent housing, and the other published rates. HUD, the U.S. Department of Housing and Urban Development, sets the rule for that rate.

The evidence

The conclusion rests on the combined record. Expand each item to see its role and limitations.

Independence note: This check uses 6 methods across 8 institutions and 1 country or international grouping.
Evidence map
Method Institution / country Year(s) What it shows Key limitation Primary links
Read HUD's System Performance Measures programming specifications, Version 3.3 (January 2022, in force for the 2022 and 2023 federal years) and the October 2025 version: measure 7b.2 and Appendix A (exit destinations). U.S. Department of Housing and Urban Development United States 2022, 2023, 2024, 2025 The region's official 'stay housed' figure is a federal measure (7b.2): people in permanent housing who either stayed or moved to other permanent housing. It covers supportive housing and two other kinds of permanent housing for formerly homeless people, but not short-term rent help (rapid rehousing). HUD's rules in force in 2022 and 2023 removed anyone whose exit was recorded as Deceased from the whole measure, so a death counts neither as staying housed nor as a failure. The same rule removes moves to a hospital, foster care or a nursing home. HUD's current rules, from October 2025, keep it. The rule shows how the measure is built, not how many deaths were removed; the region doesn't publish that number. HUD gives no reason for the rule. We read the 2022 rules from an Internet Archive copy, since HUD has replaced the file. U.S. Department of Housing and Urban Development (HUD) U.S. Department of Housing and Urban Development (HUD)
Read the authority's April 2025 slides on its federal performance measures in full. King County Regional Homelessness Authority United States 2021, 2022, 2023 The King County Regional Homelessness Authority, which runs the region's homeless system, reported 96% for both the 2022 and 2023 federal years. For October 2022 to September 2023 that was 7,842 of 8,141 people, leaving 299 counted as not staying housed. Its slides don't say that residents who died were removed. The slides cover only two federal years; newer federal-measure results were not found on the authority's site. King County Regional Homelessness Authority
Read the authority's Q2 2025 board report and its dashboard's methodology section, read the dashboard's permanent housing figures by hand on November 12, 2026, and read DESC's 2024 summary and 2025 annual report and Plymouth's 2025 annual report, searching each for death, died, deceased and passed away. King County Regional Homelessness Authority, DESC, Plymouth Housing United States 2024, 2025, 2026 The authority's quarterly report says 98% of supportive housing participants keep their housing, and its dashboard shows 93% for 2024 and 2025. DESC, a large nonprofit operator, reports 98% after one year, and Plymouth Housing more than 95%. None says how residents who die are counted. The dashboard divides by all households active in the period, which as written would count a death as not staying housed. The quarterly report counts exits 'to any location', yet rates 2,175 of the 2,214 households it served, without explaining the 39 left out. These rates may count deaths as not staying housed, or leave them out; their published methods don't settle it, and we didn't ask the publishers. King County Regional Homelessness Authority King County Regional Homelessness Authority DESC DESC Plymouth Housing FactHub Research Team
Read the program's 2024 and 2025 annual reports in full, searching for death, died, deceased and passed away. King County Department of Community and Human Services United States 2024, 2025 The county's Health Through Housing program reports that 95% (2024) and 93% (2025) of its supportive housing residents kept their housing. After giving where its emergency housing residents went, it adds that 15 (2024) and 10 (2025) people 'passed away while enrolled', outside the percentages. It doesn't say how deaths count in the supportive housing rate. The death counts sit in the emergency housing discussion, one in an endnote, and it is unclear whether they include supportive housing residents. No cause or building is given. King County Department of Community and Human Services King County Department of Community and Human Services
Read the 2024 and 2025 King County Overdose Fatality Review reports and HUD's 2024 Housing Inventory Count for the region. Public Health - Seattle & King County, King County Medical Examiner's Office United States 2024, 2025 Public Health - Seattle & King County counted 225 overdose deaths in supportive housing settings in 2024, and an average of 16 a month from January 2024 to September 2025. These come from Medical Examiner records, not from the housing agencies. The region had 7,768 supportive housing beds in January 2024, so overdose deaths alone were about 2.9 per 100 beds. These are overdose deaths only; deaths from illness and other causes are not included. They cover all supportive housing countywide, which may include buildings and people outside the federal rate. Public Health - Seattle & King County Public Health - Seattle & King County U.S. Department of Housing and Urban Development (HUD)
Rate = 7,842 / (8,141 + deaths added). Calculations are in the FactHub dataset's saved analyses. FactHub Research Team United States 2022, 2023, 2024 We added death counts to the federal rate's base as people who did not stay housed. With 225 overdose deaths, the share not staying housed rises from 3.7% to 6.3%, and the rate falls to about 94%. With all 487 deaths the Medical Examiner recorded at subsidized housing sites in 2023, it rises to 9.1% (about 91%). Overdose deaths alone were about three-quarters of the 299 people the rate counted as not staying housed. This is an estimate of scale, not a corrected rate: the years differ, overdose deaths leave out other causes, and the all-cause count includes other subsidized housing and only deaths the Medical Examiner investigates. Some deaths may already be in the 299, recorded as exits to an unknown place, which would make the estimate too high. The 2024 overdose count (225) is above the January 2024 to September 2025 average (about 190 a year), so it likely overstates the drop for the rate's earlier year. The overdose review counts deaths 'in supportive housing settings', which may include some people who didn't live there. FactHub Research Team FactHub Research Team King County Medical Examiner's Office
1. The federal rule removes residents who die from the 96% rate Read HUD's System Performance Measures programming specifications, Version 3.3 (January 2022, in force for the 2022 and 2023 federal years) and the October 2025 version: measure 7b.2 and Appendix A (exit destinations).

The region's official 'stay housed' figure is a federal measure (7b.2): people in permanent housing who either stayed or moved to other permanent housing. It covers supportive housing and two other kinds of permanent housing for formerly homeless people, but not short-term rent help (rapid rehousing). HUD's rules in force in 2022 and 2023 removed anyone whose exit was recorded as Deceased from the whole measure, so a death counts neither as staying housed nor as a failure. The same rule removes moves to a hospital, foster care or a nursing home. HUD's current rules, from October 2025, keep it.

Limit: The rule shows how the measure is built, not how many deaths were removed; the region doesn't publish that number. HUD gives no reason for the rule. We read the 2022 rules from an Internet Archive copy, since HUD has replaced the file.

Verify: System Performance Measures Programming Specifications, Version 3.3 (January 2022), archived copy System Performance Measures HMIS Programming Specifications, v1.0 (October 2025)

2. The region reports 96% without mentioning deaths Read the authority's April 2025 slides on its federal performance measures in full.

The King County Regional Homelessness Authority, which runs the region's homeless system, reported 96% for both the 2022 and 2023 federal years. For October 2022 to September 2023 that was 7,842 of 8,141 people, leaving 299 counted as not staying housed. Its slides don't say that residents who died were removed.

Limit: The slides cover only two federal years; newer federal-measure results were not found on the authority's site.

Verify: Understanding System Performance Measures: A HUD Federal Report Overview (April 2025)

3. Other local rates don't say how they count deaths Read the authority's Q2 2025 board report and its dashboard's methodology section, read the dashboard's permanent housing figures by hand on November 12, 2026, and read DESC's 2024 summary and 2025 annual report and Plymouth's 2025 annual report, searching each for death, died, deceased and passed away.

The authority's quarterly report says 98% of supportive housing participants keep their housing, and its dashboard shows 93% for 2024 and 2025. DESC, a large nonprofit operator, reports 98% after one year, and Plymouth Housing more than 95%. None says how residents who die are counted. The dashboard divides by all households active in the period, which as written would count a death as not staying housed. The quarterly report counts exits 'to any location', yet rates 2,175 of the 2,214 households it served, without explaining the 39 left out.

Limit: These rates may count deaths as not staying housed, or leave them out; their published methods don't settle it, and we didn't ask the publishers.

Verify: Governing Board Performance Report, Quarter 2, 2025 System Performance dashboard and methodology (read November 12, 2026) 2025 Annual Report Publications page: 2024 Annual Report 2025 Annual Report Published 'stay housed' rates and how each treats deaths

4. Health Through Housing lists deaths outside its percentages Read the program's 2024 and 2025 annual reports in full, searching for death, died, deceased and passed away.

The county's Health Through Housing program reports that 95% (2024) and 93% (2025) of its supportive housing residents kept their housing. After giving where its emergency housing residents went, it adds that 15 (2024) and 10 (2025) people 'passed away while enrolled', outside the percentages. It doesn't say how deaths count in the supportive housing rate.

Limit: The death counts sit in the emergency housing discussion, one in an endnote, and it is unclear whether they include supportive housing residents. No cause or building is given.

Verify: 2025 Health Through Housing Annual Report (June 2026) 2024 Health Through Housing Annual Report (June 2025)

5. How many residents die: hundreds a year from overdoses alone Read the 2024 and 2025 King County Overdose Fatality Review reports and HUD's 2024 Housing Inventory Count for the region.

Public Health - Seattle & King County counted 225 overdose deaths in supportive housing settings in 2024, and an average of 16 a month from January 2024 to September 2025. These come from Medical Examiner records, not from the housing agencies. The region had 7,768 supportive housing beds in January 2024, so overdose deaths alone were about 2.9 per 100 beds.

Limit: These are overdose deaths only; deaths from illness and other causes are not included. They cover all supportive housing countywide, which may include buildings and people outside the federal rate.

Verify: King County Overdose Fatality Review, 2024 Report King County Overdose Fatality Review, 2025 Report 2024 Housing Inventory Count, Seattle/King County Continuum of Care (WA-500)

Left-out overdose deaths: 225, next to 299 counted as not staying housedPeople King County's federal 96% 'stay housed' rate counted as not staying housed (Oct 2022 to Sep 2023), next to a year of overdose deaths in supportive housing, which the rate leaves out.Left-out overdose deaths: 225, next to 299 counted as not staying housedPeople, King Countyfacthub.org/n/chart/king-county-housing-deaths-vs-not-housedSource: FactHub Research Team; KCRHA, Public HealthNote: Different periods: the not-housed count is for October 2022 to September 2023 and the overdose deaths for calendar 2024.Counted as not staying housed (Oct 2022-Sep 2023)299Overdose deaths in supportive housing, 2024(left out)225
Left-out overdose deaths: 225, next to 299 counted as not staying housedPeople King County's federal 96% 'stay housed' rate counted as not staying housed (Oct 2022 to Sep 2023), next to a year of overdose deaths in supportive housing, which the rate leaves out.Left-out overdose deaths: 225, nextto 299 counted as not stayinghousedPeople, King CountySource: FactHub Research Team; KCRHA, Public HealthSeptember 2023 and the overdose deaths for calendar 2024.Note: Different periods: the not-housed count is for October 2022 toCounted as not stayinghoused (Oct 2022-Sep…299Overdose deaths insupportive housing…225
Overdose deaths in supportive housing in 2024 (225) were about three-quarters of the 299 people the 96% rate counted as not staying housed in October 2022 to September 2023. View source chart
6. Counting deaths would lower the 96% to about 91% to 94% Rate = 7,842 / (8,141 + deaths added). Calculations are in the FactHub dataset's saved analyses.

We added death counts to the federal rate's base as people who did not stay housed. With 225 overdose deaths, the share not staying housed rises from 3.7% to 6.3%, and the rate falls to about 94%. With all 487 deaths the Medical Examiner recorded at subsidized housing sites in 2023, it rises to 9.1% (about 91%). Overdose deaths alone were about three-quarters of the 299 people the rate counted as not staying housed.

Limit: This is an estimate of scale, not a corrected rate: the years differ, overdose deaths leave out other causes, and the all-cause count includes other subsidized housing and only deaths the Medical Examiner investigates. Some deaths may already be in the 299, recorded as exits to an unknown place, which would make the estimate too high. The 2024 overdose count (225) is above the January 2024 to September 2025 average (about 190 a year), so it likely overstates the drop for the rate's earlier year. The overdose review counts deaths 'in supportive housing settings', which may include some people who didn't live there.

Verify: Share counted as not staying housed, with and without deaths People counted as not staying housed versus overdose deaths left out Annual Report dashboard: Decedent Characteristics (read November 11, 2026)

Counting deaths would raise the 'not housed' share from 3.7% to roughly 6–9%King County's federal 'stay housed' measure (96%), Oct 2022 to Sep 2023: share of residents counted as not staying housed, as published with deaths left out, and with death counts added.{"width":960,"height":540,"plot":{"left":34,"top":107,"right":926,"bottom":462},"yDomain":[0,10],"series":[{"name":"Value","color":"#0f6e99"}],"points":[{"series":0,"x":182.67,"y":330.65,"xLabel":"As published: deaths left out","yLabel":"3.7"},{"series":0,"x":480,"y":238.35,"xLabel":"Counting 225 overdose deaths","yLabel":"6.3"},{"series":0,"x":777.33,"y":138.95,"xLabel":"Counting 487 deaths, all causes","yLabel":"9.1"}]}Counting deaths would raise the 'not housed' share from 3.7% to roughly 6–9%% of 8,141 not counted as staying housed, Oct 2022–Sep 2023; death bars are rough estimatesAs published: deaths left outCounting 225 overdose deathsCounting 487 deaths, all causes3.76.39.1facthub.org/n/chart/king-county-housing-not-housed-with-deathsSource: FactHub Research Team; KCRHA, Public HealthNote: Rough estimate. Overdose deaths are for 2024; the 487 deaths are Medical Examiner deaths at all subsidized housing in 2023, a broader group than the rate covers.
Counting deaths would raise the 'not housed' share from 3.7% to roughly 6–9%King County's federal 'stay housed' measure (96%), Oct 2022 to Sep 2023: share of residents counted as not staying housed, as published with deaths left out, and with death counts added.Counting deaths would raise the'not housed' share from 3.7% toroughly 6–9%% of 8,141 not counted as staying housed, Oct 2022–Sep2023; death bars are rough estimatesSource: FactHub Research Team; KCRHA, Public HealthAs published: deaths leftout3.7Counting 225 overdosedeaths6.3Counting 487 deaths, allcauses9.1
The region's 96% federal rate leaves out residents who die. Adding death counts back raises the share not staying housed from 3.7% to 6.3% (overdose deaths) or 9.1% (all deaths at subsidized housing sites). Estimates of scale; years and buildings differ. View source chart

Common questions 7

“Counting residents who die would lower the 96% rate by less than one point.”

Short answer: No: by about 2 to 5 points

Counting deaths would lower the 96% rate to about 94% (overdose deaths only) or about 91% (all deaths at subsidized housing sites), by our estimate. The 225 overdose deaths in supportive housing in 2024 were about three-quarters of the 299 people the rate counted as not staying housed in 2022-23. Years and buildings differ, so this shows scale, not a corrected rate.

See the evidence: Counting deaths would lower the 96% to about 91% to 94%

“A death isn't a housing failure, so it shouldn't count against the rate.”

Short answer: Fair, but then it says nothing on deaths

That is a fair reason to leave deaths out of a housing measure. A resident who dies didn't lose a home or go back to the street, and HUD also leaves out moves to hospitals and nursing homes. HUD's rule gives no reason, though. And a rate without deaths tells readers nothing about how many residents died. Only Health Through Housing gives a death count in the same report, and only in its emergency housing section.

See the evidence: The federal rule removes residents who die from the 96% rate

“Officials say 98% of people in supportive housing stay housed.”

Short answer: Yes, but neither explains deaths

The regional authority's mid-2025 report says 98% of supportive housing participants keep their housing, and DESC says 98% of its clients kept their homes after one year. Neither says how residents who die are counted. The authority's table served 2,214 households but calculates the 98% from 2,175, without saying why 39 are left out.

See the evidence: Other local rates don't say how they count deaths

“The other local 'stay housed' rates leave out residents who die too.”

Short answer: Not shown; one may count them

We can't tell. The authority's dashboard (93%) divides by all households active in the period, which as written would count a death as not staying housed. DESC and Plymouth give no method. Health Through Housing lists its deaths apart from its emergency housing shares, but doesn't say how they count in its supportive housing rate.

See the evidence: Other local rates don't say how they count deaths

“The federal rule makes the rate look as good as possible.”

Short answer: Not in every way

No. Only a few exits are removed: deaths, and moves to a hospital, foster care or a nursing home. People who leave without saying where they went stay in the rate and count as not staying housed. Some of those could be deaths that were never recorded.

See the evidence: The federal rule removes residents who die from the 96% rate

“Health Through Housing reports fewer deaths per resident than countywide supportive housing overdose data show.”

Short answer: Yes: about 1 vs 2.9 per 100

Yes. The county program reported 15 resident deaths in 2024 and 10 in 2025, about 1 per 100 people it housed. Overdose deaths alone in all county supportive housing were 2.9 per 100 beds in 2024. The two don't measure the same people: the program's count sits in its emergency housing section, may not cover its supportive housing, and gives no cause. Per-person rates also run lower than per-bed rates. The gap is unexplained.

See the evidence: Health Through Housing lists deaths outside its percentages

“Local officials wrote the rule that leaves deaths out of the 96% rate.”

Short answer: No, it's a federal rule

No: the 96% follows a federal housing department rule that regions getting federal homelessness grants use for this measure. What local agencies do choose is whether to publish deaths beside the rate. The regional authority, DESC and Plymouth don't; Health Through Housing gives only a yearly count.

See the evidence: The federal rule removes residents who die from the 96% rate

Conclusion

The criticism holds for the region's federal 96% rate. The 96% 'stay housed' rate the regional homelessness authority reported for the 2022 and 2023 federal years is a federal measure, and HUD's rule removes anyone whose exit is recorded as a death. The authority reports the 96% without saying so.

The effect is not small next to what the rate does count. It counted 299 people as not staying housed in October 2022 to September 2023. Overdose deaths in supportive housing alone were 225 in 2024. Counting deaths as not staying housed would put the rate near 94%, or near 91% with all deaths the Medical Examiner records at subsidized housing sites. These are rough estimates: the years and buildings differ, the 2024 overdose count was higher than the yearly average since, and some deaths may already be counted among people who left for an unknown place.

The rating covers the 96% rate only. The authority's newer public figures are 98% (its quarterly report) and 93% (its dashboard). The other local figures don't say how they treat deaths, and the authority's 93% dashboard, as its formula is written, may count them as not staying housed. Whether the '98%' figures leave deaths out is not shown.

There is a fair case for the rule. A resident who dies didn't lose a home, and counting every death as a failure would penalize programs that take in sicker people. But that argues for reporting deaths beside the rate. The regional authority and the large operators don't, and Health Through Housing's yearly count of 10 to 15 deaths looks low next to countywide overdose deaths. Publishing deaths by program, next to the stay-housed rate, would let the public judge both.

How we checked
  1. We defined the claim as whether the region's 96% 'stay housed' rate for supportive housing, the federal measure the regional homelessness authority reports, leaves residents who die out of the calculation rather than counting them as not staying housed. We also checked what the other published local rates say about deaths, without rating them.
  2. We read the federal rule behind the region's official rate in two versions: HUD's January 2022 specification (in force for the 2022 and 2023 federal years) and its October 2025 update.
  3. We relied on more than one kind of source: the federal rule itself, the regional authority's own reports, the county program's and operators' reports, and death counts from the health department and Medical Examiner, which don't run housing.
  4. We gathered every published rate we could find: the regional homelessness authority's federal-measure slides, its Q2 2025 board report and its public dashboard; Health Through Housing's 2024 and 2025 annual reports; DESC's 2024 summary and 2025 annual report; and Plymouth Housing's 2025 annual report. We read each method note and searched each for death, died, deceased and passed away.
  5. We sized the effect with public death counts from Public Health - Seattle & King County's overdose reviews and the Medical Examiner's dashboard, adding them to the federal rate's base, and used HUD's 2024 bed count for a per-bed rate. The calculations are in the FactHub dataset.
  6. We did not ask the publishers how they count deaths, and we did not check Seattle's Office of Housing or the Seattle Housing Authority, which don't publish a stay-housed rate we found.

Editorial note: This check rates the region's 96% 'stay housed' rate and reports what other published rates say about deaths, as of November 12, 2026. It does not judge whether supportive housing works. The estimates of what the rate would be with deaths counted are ours and approximate. No external specialist review has been recorded.

Expert-review status

No external specialist review has been recorded.

Funding and conflicts

Funding: No external funding for this piece.

Ownership: Published by the FactHub Research Team.

Conflicts: None known.

Sources 18

Sources are labeled by kind, with what this fact check takes from each, so readers can check the evidence independently.

  1. Government U.S. Department of Housing and Urban Development (HUD). System Performance Measures Programming Specifications, Version 3.3 (January 2022), archived copy

    The HUD rules in force when the region measured its 96% rate for October 2021 to September 2023, including measure 7b.2, the share of people in permanent housing who stay or move to other permanent housing.

    Reliability: The federal agency that funds homeless housing sets the rule; the document reports no local numbers and gives no reason for leaving out deaths. Archived by the Internet Archive on November 27, 2022; HUD has since replaced it at the same address.

    • For measure 7b.2, people whose exit destination is code 24 (Deceased), 15 (foster care), 6 (hospital) or 25 (long-term care or nursing home) are removed from the whole measure.

      What the source says
      Destinations indicated with an X (values 15, 6, 25, 24) cause leavers with those destinations to be completely excluded from the entire measure (all of column C).
      Metric 7b.2, programming instructions, step 5, p. 34; Appendix A lists 24 as Deceased, p. 35

      Not yet checked

    • Measure 7b.2's base is everyone left after those exclusions, stayers and leavers alike, so people who leave for an unknown place stay in the base.

      What the source says
      Of the selected clients, report the distinct number of stayers and leavers in cell C2.
      Metric 7b.2, programming instructions, step 6, p. 34

      Not yet checked

    • Qualifies: Measure 7b.2 covers residents of permanent supportive housing and two other kinds of permanent housing (housing only, and housing with services) who have moved in.

      What the source says
      Permanent Supportive Housing (Project Type 3), Permanent Housing-Housing Only (Project Type 9) and Permanent Housing-Housing Services only (Project Type 10) clients who are active in the report range with a housing move-in date
      Metric 7b.2, Universe, p. 33

      Not yet checked

  2. Government U.S. Department of Housing and Urban Development (HUD). System Performance Measures HMIS Programming Specifications, v1.0 (October 2025)

    HUD's current version of the same rules, for reporting from October 2025, which keeps the exclusion of residents who die.

    Reliability: The federal agency that funds homeless housing sets the rule; the document reports no local numbers and gives no reason for leaving out deaths.

    • The current rules keep the same exclusion: exits to destinations marked with an X in Appendix A are removed from the whole measure, and Deceased (code 24) is marked with an X in every column.

      What the source says
      cause leavers with those destinations to be completely excluded from the entire measure universe
      Appendix A: Exit Destinations, p. 48-49 (row 24, Deceased)

      Not yet checked

  3. Government King County Regional Homelessness Authority. Understanding System Performance Measures: A HUD Federal Report Overview (April 2025)

    Slides in which the regional homelessness authority reports its federal performance measures for the 2022 and 2023 federal years to its board.

    Reliability: Has a stake in this finding. The regional agency that funds and oversees most of the county's homeless housing, reporting its own system's results. It follows HUD's specification for this measure.

    • The region reported that 96% of people in permanent housing stayed or moved to other permanent housing in both federal years 2022 and 2023; the slides do not mention deaths.

      What the source says
      For Metric 7b.2, the percentage of successful exits or retention in permanent housing remained consistent at 96%.
      Slide: KCRHA System Performance (M7)

      Not yet checked

    • The 96% for October 2022 to September 2023 was 7,842 of 8,141 people, so 299 were counted as not staying housed.

      What the source says
      7,874 8,141 267
      Slide: KCRHA System Performance (M7), Retained PH table: Total row; Positive Exits row reads 7,556 7,842 286

      Not yet checked

  4. Government King County Regional Homelessness Authority. Governing Board Performance Report, Quarter 2, 2025

    The regional authority's quarterly report to its governing board on the programs it funds.

    Reliability: Has a stake in this finding. The authority's own report on programs it funds.

    • Qualifies: The authority reports that 98% of supportive housing participants keep their housing or move to permanent housing; it does not say how deaths are counted.

      What the source says
      98% of participants maintain their housing or exit to permanent housing
      Section 2.1 Permanent Supportive Housing, p. 14

      Not yet checked

    • Qualifies: In Q2 2025 the authority served 2,214 supportive housing households, but its 98% rate is out of 2,175, with no reason given for the 39 left out.

      What the source says
      Q2 2025 2,214 2,175 2,141 98%
      Table 13, p. 15

      Not yet checked

    • Qualifies: The report defines supportive housing exits as exits 'to any location' or staying enrolled, without mentioning deaths.

      What the source says
      for PSH, this becomes total households exiting to any location or maintaining enrollment
      How to Read the Report, definitions, p. 7

      Not yet checked

  5. Government King County Regional Homelessness Authority. System Performance dashboard and methodology (read November 12, 2026)

    The authority's public dashboard of homeless records, with a methodology section on the same page.

    Reliability: Has a stake in this finding. The authority's own public scorecard; it says over 85% of programs in the region report into the records it draws on.

    • Qualifies: The dashboard's permanent housing rate is households who stay enrolled or exit to permanent housing, divided by households active in the period; the method doesn't mention deaths.

      What the source says
      Exits to or maintain permanent housing = Total number of households who maintain enrolled past the end of the timeframe and total number of households who exited to permanent housing during the timeframe
      Learn more about our methodology: Exit to or Maintain Permanent Housing

      Not yet checked

    • Qualifies: The dashboard shows 93% of permanent housing households (including supportive housing) 'permanently housed' in 2024 and 2025: 6,580 and 7,084 households.

      What the source says
      Number of households exiting to or maintaining permanent housing : 7,084
      Dashboard, Annual System Trends tab, Permanent Housing (Including Permanent Supportive Housing), 2025 bar (93%) tooltip; 2024 bar reads 6,580

      Not yet checked

  6. Government King County Department of Community and Human Services. 2025 Health Through Housing Annual Report (June 2026)

    The yearly report of the county's Health Through Housing program, which bought hotels and apartment buildings to house chronically homeless people.

    Reliability: Has a stake in this finding. Written by the county department that runs the program, which has an interest in showing it works.

    • Qualifies: Health Through Housing reports that 93% of its supportive housing residents kept their housing or moved to other permanent housing in 2025, without saying how deaths count.

      What the source says
      HTH's 93 percent overall rate of maintaining PSH housing or moving on to other permanent destinations is on par with the regional PSH housing stability average (93 percent).
      p. 9-10, Percent of HTH Residents who Maintain Housing in HTH or Exit to Permanent Housing

      Not yet checked

    • After listing where its emergency housing residents went, the 2025 report adds that 10 people died while enrolled, outside its percentages, with no cause given.

      What the source says
      An additional 10 individuals passed away while enrolled with HTH.
      Endnote 37, p. 64

      Not yet checked

    • Qualifies: Health Through Housing housed 1,380 people in 2025.

      What the source says
      In 2025, HTH permanently or temporarily housed a total of 1,380 people
      p. 7

      Not yet checked

  7. Government King County Department of Community and Human Services. 2024 Health Through Housing Annual Report (June 2025)

    The program's report for 2024.

    Reliability: Has a stake in this finding. Same publisher and method as the 2025 report.

    • The 2024 report likewise lists 15 residents who died while enrolled as additional to its emergency housing percentages.

      What the source says
      An additional 15 individuals passed away while enrolled with HTH.
      p. 26, emergency housing exits

      Not yet checked

    • Qualifies: It reports 95% of supportive housing residents kept their housing or moved to other permanent housing in 2024, out of 1,281 people served.

      What the source says
      with 95 percent of PSH residents maintaining their housing or moving to another permanent housing destination
      Key findings, p. 5

      Not yet checked

  8. Other DESC. 2025 Annual Report

    The yearly report of DESC, the county's largest supportive housing operator.

    Reliability: Has a stake in this finding. A nonprofit operator funded by the county and city, reporting its own results; it gives no method.

    • Qualifies: DESC says 98% of its clients kept their homes after one year, with no method and no mention of deaths.

      What the source says
      In 2025, we moved 327 people into permanent supportive housing, with 98% of all clients retaining their homes after one year
      Opening letter

      Not yet checked

  9. Other DESC. Publications page: 2024 Annual Report

    DESC's web summary of its 2024 annual report.

    Reliability: Has a stake in this finding. The operator's own summary; no method given.

    • Qualifies: DESC reported that 94.6% of people it moved into supportive housing in 2024 kept their homes, with no mention of deaths.

      What the source says
      with 94.6% of them retaining their homes
      Annual Report 2024 summary

      Not yet checked

  10. Other Plymouth Housing. 2025 Annual Report

    The yearly report of Plymouth Housing, a large Seattle supportive housing operator.

    Reliability: Has a stake in this finding. A nonprofit operator reporting its own results; it gives no method.

    • Qualifies: Plymouth says more than 95% of its residents remain stably housed, with no method and no mention of deaths.

      What the source says
      More than 95% of Plymouth Housing residents remain stably housed.
      How do we know our approach works?

      Not yet checked

  11. Government Public Health - Seattle & King County. King County Overdose Fatality Review, 2024 Report

    The county health department's yearly review of fatal overdoses, drawn from Medical Examiner records.

    Reliability: Written by the health department's overdose review team, which does not run housing.

    • 225 people died of overdoses in supportive housing settings in King County in 2024.

      What the source says
      In 2024, there were 225 fatal overdoses in permanent supportive housing settings.
      Fatal Overdoses in Permanent Supportive Housing, p. 10

      Not yet checked

  12. Government Public Health - Seattle & King County. King County Overdose Fatality Review, 2025 Report

    The 2025 edition of the county's overdose review.

    Reliability: Same publisher and method as the 2024 review.

    • From January 2024 to September 2025, an average of 16 people a month living in supportive housing died of overdoses, about 190 a year.

      What the source says
      an average of 16 fatal overdoses occurred each month among individuals living in permanent supportive housing (PSH) in King County
      Fatal Overdoses in Permanent Supportive Housing, p. 17

      Not yet checked

  13. Government King County Medical Examiner's Office. Annual Report dashboard: Decedent Characteristics (read November 11, 2026)

    The Medical Examiner's public dashboard of the deaths it investigates, filterable by the person's presumed living situation.

    Reliability: The Medical Examiner investigates or reviews about one in five county deaths, mostly sudden or unexpected ones. Its housing service site category mixes supportive, income-restricted and recovery housing.

    • The Medical Examiner recorded 487 deaths of all causes in 2023 among people whose presumed living situation was a housing service site (supportive, income-restricted or recovery housing).

      What the source says
      All Race and Ethnicity Deaths = 487
      Decedent Characteristics tab, Presumed Living Situation = Housing service site, Year = 2023

      Not yet checked

  14. Government U.S. Department of Housing and Urban Development (HUD). 2024 Housing Inventory Count, Seattle/King County Continuum of Care (WA-500)

    The region's count of homeless and supportive housing beds, reported to HUD each January.

    Reliability: Has a stake in this finding. Counts reported by the region's homeless system and published by HUD.

    • Qualifies: The region had 7,768 year-round permanent supportive housing beds in its January 2024 count.

      What the source says
      416 1,407 76 0 7,768
      Summary table, Permanent Supportive Housing row, Total Year-Round Beds column

      Not yet checked

  15. FactHub record FactHub Research Team. Share counted as not staying housed, with and without deaths

    FactHub's calculation of the region's federal rate with death counts added to its base.

    • As published, 3.7% of 8,141 people were counted as not staying housed. Adding 225 overdose deaths makes it 6.3%; adding 487 all-cause deaths at subsidized housing sites makes it 9.1%. The 96% would be about 94% or 91%.

      Not yet checked

  16. FactHub record FactHub Research Team. People counted as not staying housed versus overdose deaths left out

    FactHub's comparison of the rate's not-housed count with a year of overdose deaths.

    • The rate counted 299 people as not staying housed in October 2022 to September 2023; overdose deaths in supportive housing, which it leaves out, were 225 in 2024, about three-quarters as many.

      Not yet checked

  17. FactHub record FactHub Research Team. Published 'stay housed' rates and how each treats deaths

    FactHub's table of the 11 published rates we found and what each says about deaths.

    • Eleven published rates run from 93% to 98%. Two, the federal 96%, leave deaths out by HUD rule; none of the other nine says how deaths are handled.

      Not yet checked

  18. FactHub record FactHub Research Team. Deaths per 100 residents or beds

    FactHub's comparison of Health Through Housing's death counts with countywide overdose deaths in supportive housing.

    • Qualifies: Health Through Housing reports 1.2 deaths per 100 residents in 2024 and 0.7 in 2025; countywide, overdose deaths alone were 2.9 per 100 supportive housing beds (7,768 beds) in 2024.

      Not yet checked

  19. FactHub dataset King County supportive housing 'stay housed' rates and resident deaths

    • Tables: scenarios

      Saved analysis not-housed-share

    • Tables: scenarios, deaths

      Saved analysis counted-vs-left-out

    • Tables: retention-rates

      Saved analysis rates-and-deaths

    • Tables: deaths

      Saved analysis deaths-per-100

  20. FactHub chart king-county-housing-not-housed-with-deaths
  21. FactHub chart king-county-housing-deaths-vs-not-housed

Update history 7 changes, last

  1. First version.
  2. The answer and lead now give the overdose-only estimate (about 94%), with the all-deaths figure (about 91%) qualified (feedback 2516077466783371815-23fb91f33334fb530cd23ae6ea297839). Said that the 96% measure covers supportive and other permanent housing, and added HUD's definition (feedback 2516077466785730361-fd79dc67807b3819d40dd452529b10da). Reworded the question on who wrote the rule and made the size question the lead question (feedback 2516077466781304085-6f96279b660cfb492ffdaa003cd1ad0e). Split a long sentence.
  3. Marked the estimate of the rate with deaths counted as a rough estimate and said the 2024 overdose count likely overstates the drop (feedback 2516077463574179378-91bdaf51b85caefc289cf4d9a9030ff8). Said the rating covers the federal 96% rate, and named the authority's newer 98% and 93% figures (feedback 2516077463576367614-70ad51b1611801165a218368d3608297). Removed the lead question, which repeated the answer (feedback 2516077463571988443-2dd7b793961725617b824be8a1e8415e).
  4. Made the question on Health Through Housing's death count the lead question, reworded so the evidence settles it (feedback 2516077463571988443-2dd7b793961725617b824be8a1e8415e).
  5. Shortened the lead question's first sentence and the description.
  6. Reworded the question on Health Through Housing's death count so the evidence settles it, noted that the two figures cover different people, and removed it as the lead question (feedback 2516077460505853223-bf910551daebe067db7290a837faa2ca).
  7. The lead chart now says on its face that the death bars are rough estimates from other years and a broader group of housing.

Feedback these versions applied

For developers: ClaimReview data

This page includes standardized ClaimReview metadata describing the exact claim, FactHub's finding, publisher, and review dates. You can inspect the record below and confirm that it matches the article you just read.

View ClaimReview JSON 
{"@context":"https://schema.org","@type":"ClaimReview","url":"https://facthub.org/n/article/king-county-stay-housed-rates-deaths","datePublished":"2026-11-13","dateModified":"2026-11-13","author":{"@type":"Organization","name":"FactHub","url":"https://facthub.org/"},"claimReviewed":"King County\u0027s 96% \u0027stay housed\u0027 rate for supportive and other permanent housing leaves out residents who die.","reviewRating":{"@type":"Rating","alternateName":"True"},"image":"https://facthub.org/node-assets-v2/article/king-county-stay-housed-rates-deaths/7/img-morrison/social-1200x630.jpg"}

This metadata helps machines discover the fact-check. It does not prove the conclusion or guarantee special treatment by a search engine.

Challenge this fact-check

Every source above is linked so you can check it yourself. Identify the specific conclusion you dispute and provide verifiable evidence that could change it. Credible challenges deserve review and a visible response.

Submit evidence Send general feedback instead

Technical details: node handle article:king-county-stay-housed-rates-deaths:7