FactHub curated dataset

Seattle/King County homeless count and shelter and housing beds, 2012-2026

King County Point-in-Time homeless counts with HUD Housing Inventory Count beds by type, 2012-2026, and how King County's capacity compares with 66 other large Continuums of Care.

Coverage: 2012-2025 (HUD); 2026 count from KCRHA initial release

Node: dataset:king-county-homeless-count-beds

What this dataset measures

Shows what shelter and housing capacity King County added while its homeless count grew, using HUD's own inventory, and puts King County's capacity in context with the same 67 large CoCs used in homelessness-rate-vs-rent.

Curated data tables

These related tables form this versioned dataset release. Each table has its own downloadable file and field definitions.

Beds by type and homeless counts, 67 large CoCs

HUD Housing Inventory Count beds by project type for each CoC and year, joined to the PIT count and population.

Unit of observation: CoC-year

View this table

hic_panel.csv text/csv · 169.8 KB (173,908 bytes)
24 field definitions
FieldTypeMeaningUnitsValue type
coc_code string HUD CoC code — observed
coc_name string CoC name as HUD listed it — observed
short_name string Short display name — derived
hud_category_2025 string HUD 2025 CoC category — observed
year integer HIC/PIT year (January count night) — observed
es_beds integer HUD year-round emergency shelter beds beds observed
th_beds integer HUD year-round transitional housing beds beds observed
sh_beds integer HUD year-round safe haven beds beds observed
rrh_beds integer HUD year-round rapid re-housing (2014-2016 includes RRH demonstration projects; blank in 2012) beds beds observed
psh_beds integer HUD year-round permanent supportive housing beds beds observed
oph_beds integer HUD year-round other permanent housing (blank in 2012-2013) beds beds observed
es_seasonal_beds integer Seasonal emergency shelter beds (not in es_beds) beds observed
es_overflow_beds integer Overflow/voucher emergency shelter beds (not in es_beds) beds observed
temp_beds integer ES + TH + SH year-round beds beds derived
perm_beds integer RRH + PSH + OPH beds (only types reported that year) beds derived
pit_count_type string HUD PIT count type — observed
pit_total integer People counted homeless people observed
pit_sheltered integer People counted in shelter people observed
pit_unsheltered integer People counted unsheltered (carried forward in sheltered-only years) people observed
population integer Resident population (ACS B01003); blank 2012-2014 and 2020 people observed
temp_beds_per_homeless number temp_beds / pit_total beds per person derived
unsheltered_share number pit_unsheltered / pit_total; blank in sheltered-only years fraction derived
temp_beds_per_10k number Temporary beds per 10,000 residents beds per 10,000 residents derived
perm_beds_per_10k number Permanent beds per 10,000 residents beds per 10,000 residents derived

Seattle/King County homeless count and beds by type, 2012-2026

King County (WA-500) rows from hic-panel, plus KCRHA's initial 2026 count (no 2026 bed data).

Unit of observation: Year

View this table

king_count_beds.csv text/csv · 1.6 KB (1,625 bytes)
16 field definitions
FieldTypeMeaningUnitsValue type
year integer Count year — observed
source string HUD, or KCRHA initial release for 2026 — derived
count_type string HUD PIT count type — observed
pit_total integer People counted homeless people observed
pit_sheltered integer People counted in shelter people observed
pit_unsheltered integer People counted unsheltered people observed
unsheltered_carried_forward integer 1 if the unsheltered figure is not from a new count that year: HUD carried it forward from the previous count (2014, 2023, 2025) or recorded 0 because no unsheltered count was done (2021) — derived
es_beds integer HUD year-round emergency shelter beds beds observed
th_beds integer HUD year-round transitional housing beds beds observed
sh_beds integer HUD year-round safe haven beds beds observed
rrh_beds integer HUD year-round rapid re-housing beds beds observed
psh_beds integer HUD year-round permanent supportive housing beds beds observed
oph_beds integer HUD year-round other permanent housing beds beds observed
temp_beds integer ES + TH + SH beds derived
perm_beds integer RRH + PSH + OPH beds derived
temp_beds_per_homeless number temp_beds / pit_total; blank in sheltered-only years beds per person derived

Supporting files

Source snapshots, manifests, quality reports, documentation, and other materials used to produce or review this release.

  • build_hic.js Transformation manifest · text/javascript · 4.9 KB (4,976 bytes)

    Node.js script that builds hic-panel from the HUD HIC workbook sheets (exported to CSV) and the PIT/ACS panel.

  • analyze_hic.js Transformation manifest · text/javascript · 6.5 KB (6,702 bytes)

    Node.js script for the cross-CoC bed comparisons in the method notes.

  • csvlib.js Transformation manifest · text/javascript · 1.1 KB (1,139 bytes)

    CSV helper used by the HIC scripts.

  • build_count_beds.js Transformation manifest · text/javascript · 24.9 KB (25,500 bytes)

    Node.js script that builds king-count-beds and the saved analyses in this dataset.

  • findings-hic.md Documentation · text/markdown · 16.2 KB (16,567 bytes)

    Method notes, King County and cross-CoC results, and caveats for the bed-capacity data.

Saved analyses

Seattle/King County homeless count, sheltered vs. unsheltered, 2012-2026

Question: How did the number of people counted homeless in King County, sheltered and unsheltered, change from 2012 to 2026?

Result: Total 8,899 (2012) to 16,868 (2024) and 18,365 (2026, KCRHA initial). Sheltered stayed between about 5,700 and 7,100 in full-count years; unsheltered rose from 2,618 to 9,810 (2024) and 11,829 (2026).

Operation: Extract WA-500 rows; append KCRHA 2026 initial release

Scope: King County, 2012-2026

View this analysis

  • HIC counts beds that exist on the count night, not whether they were staffed or occupied.
  • King County switched its unsheltered count to respondent-driven sampling in 2022; counts before and after are not directly comparable.
  • The 2026 figure is KCRHA's initial release; HUD had not published it as of retrieval.
  • Point-in-time counts are single nights in late January.

King County temporary and permanent beds vs. people counted homeless, 2012-2025

Question: Did King County's shelter and housing capacity keep pace with the number of people counted homeless?

Result: Temporary beds stayed between 6,038 and 7,746 from 2012 to 2025 (6,452 in 2015, 6,991 in 2024). Permanent beds rose from 7,421 (2015) to 11,263 (2024). People counted homeless rose from 10,122 to 16,868. Temporary beds per person counted fell from 0.64 to 0.41.

Operation: Extract WA-500 temp_beds, perm_beds and pit_total by year

Scope: King County, 2012-2025

View this analysis

  • HIC counts beds that exist on the count night, not whether they were staffed or occupied.
  • King County switched its unsheltered count to respondent-driven sampling in 2022; counts before and after are not directly comparable.
  • RRH "beds" are rental-assistance slots that swing with grant funding.
  • PSH and OPH include tenant-based vouchers, not only dedicated buildings.

Change in King County beds by type, 2015 to 2024

Question: Which kinds of shelter and housing capacity grew or shrank in King County between 2015 and 2024?

Result: Emergency shelter (ES): 2828 to 5173 (+82.9%); Transitional housing (TH): 3579 to 1742 (-51.3%); Safe haven (SH): 45 to 76 (+68.9%); Temporary beds total (ES+TH+SH): 6452 to 6991 (+8.4%); Rapid rehousing (RRH): 1326 to 1990 (+50.1%); Permanent supportive housing (PSH): 4675 to 7768 (+66.2%); Other permanent housing (OPH): 1420 to 1505 (+6%); Permanent beds total (RRH+PSH+OPH): 7421 to 11263 (+51.8%); People counted homeless (PIT total): 10122 to 16868 (+66.6%)

Operation: Difference and percent change, WA-500, 2015 vs 2024

Scope: King County

View this analysis

  • HIC counts beds that exist on the count night, not whether they were staffed or occupied.
  • King County switched its unsheltered count to respondent-driven sampling in 2022; counts before and after are not directly comparable.
  • Project-type classification can shift between years (TH vs. ES, PSH vs. OPH).

King County capacity compared with 67 large CoCs, 2024

Question: How does King County compare with other large CoCs on shelter capacity per resident and per person counted homeless?

Result: Temporary beds (ES+TH+SH) per 10,000 residents, 2024: 29.87, rank 12 of 67 (median 14.97); Temporary beds per person counted homeless, 2024: 0.41, rank 53 of 60 (median 0.74); Unsheltered share of people counted, 2024: 58.2, rank 13 of 60 (median 34); Permanent beds (RRH+PSH+OPH) per 10,000 residents, 2024: 48.13, rank 14 of 67 (median 27.92)

Operation: Rank WA-500 among the 67 CoCs (or the 60 with a full 2024 count)

Scope: 67 large CoCs

View this analysis

  • HIC counts beds that exist on the count night, not whether they were staffed or occupied.
  • King County switched its unsheltered count to respondent-driven sampling in 2022; counts before and after are not directly comparable.
  • CoC boundaries and populations follow dataset homelessness-rate-vs-rent.
  • Right-to-shelter places (NYC, Boston) shelter people outside the year-round inventory.

Permanent and temporary beds per 10,000 residents, 2015 vs. 2024, selected CoCs

Question: How did King County's growth in permanent and temporary capacity per resident compare with other large CoCs?

Result: Seattle/King County: permanent 35.1 to 48.1 (+13.1); temporary 30.5 to 29.9; New York City: permanent 34.2 to 67.1 (+32.9); temporary 87.8 to 132.5; Boston: permanent 98.7 to 120.6 (+21.9); temporary 95.1 to 103; Houston/Harris/Fort Bend/Montgomery: permanent 7.6 to 16.2 (+8.6); temporary 5.3 to 4.5; San Diego: permanent 11.1 to 35.3 (+24.3); temporary 14 to 17.4; Metro Denver: permanent 8.5 to 32.8 (+24.4); temporary 15.7 to 25.5; Dallas/Collin: permanent 9.3 to 13.4 (+4.1); temporary 10.4 to 8.6; Los Angeles (excl. Long Beach/Pasadena/Glendale): permanent 20.8 to 41.2 (+20.4); temporary 15.2 to 29.6; San Francisco: permanent 90.2 to 198 (+107.7); temporary 29.2 to 52.8; Portland/Multnomah: permanent 57.2 to 117 (+59.8); temporary 21.2 to 48; Median of 67 CoCs: permanent 16.9 to 27.9 (+8.2); temporary 14.5 to 15

Operation: perm_beds_per_10k and temp_beds_per_10k, 2015 and 2024

Scope: Selected large CoCs and the 67-CoC median

View this analysis

  • HIC counts beds that exist on the count night, not whether they were staffed or occupied.
  • King County switched its unsheltered count to respondent-driven sampling in 2022; counts before and after are not directly comparable.
  • Comparison set chosen to include the regions named in the accompanying article; the full panel is in hic-panel.

Derived from

This dataset uses data from these FactHub dataset versions.

Sources

  1. HUD, 2007-2025 Housing Inventory Count by CoC (AHAR Part 1 data) Year-round, seasonal and overflow beds by project type for each CoC and year. Downloaded from the Internet Archive because huduser.gov blocks scripted requests; contentDigest is the SHA-256 of the downloaded workbook (2,598,803 bytes).

    Retrieved October 10, 2026. U.S. government work (public domain). Public domain; not rehosted here.

    Archived copy captured October 10, 2026

  2. HUD, 2007-2025 Point-in-Time Counts by CoC (AHAR Part 1 data) Total, sheltered and unsheltered counts and count type. contentDigest is the SHA-256 of the workbook downloaded from the archived copy.

    Retrieved October 10, 2026. U.S. government work (public domain). Public domain; not rehosted here.

    Archived copy captured May 30, 2026

  3. U.S. Census Bureau, American Community Survey 1-year estimates, table B01003 (population) Resident population used for beds per 10,000 residents (via dataset homelessness-rate-vs-rent).

    Retrieved October 10, 2026. U.S. government work (public domain). Public domain.

  4. KCRHA, 2026 Point-in-Time Count FAQs (initial release), June 23, 2026 King County 2026 count (18,365 total; 6,536 sheltered; 11,829 unsheltered), table on p.3. HUD had not yet published 2026 data.

    Retrieved October 10, 2026. Public agency document. Linked, not rehosted.

    Archived copy captured October 10, 2026

  5. KCRHA, 2024 Point-in-Time Count FAQ v2 Documents the respondent-driven sampling method used since 2022.

    Retrieved October 10, 2026. Public agency document. Linked, not rehosted.

    Archived copy captured October 10, 2026

Limitations and reproducibility

Reproducible from public sources: Export the HUD HIC workbook sheets to CSV, run hic-build-script against them and the panel from dataset homelessness-rate-vs-rent, then run count-beds-script (Node.js).

  • HIC beds are inventory on the count night, not occupancy.
  • King County's count method changed in 2015, 2017, 2022 and 2026; HUD carried unsheltered counts forward in 2014, 2023 and 2025.
  • Permanent housing beds serve people who are no longer counted as homeless.
  • Rapid re-housing beds depend on grant funding and swing year to year.
  • Descriptive data; no causal claims.

Publications using this dataset

Version details

The exact citing versions, and the version of this dataset each cites.

  • Article seattle-homelessness-spending-results — version 5 (published) cites version 2
  • Article seattle-homelessness-spending-results — version 4 (published) cites version 2
  • Article seattle-homelessness-spending-results — version 3 (published) cites version 2
  • Article seattle-homelessness-spending-results — version 2 (published) cites version 2
  • Chart king-county-beds-vs-count-chart — version 1 (published) cites version 2