West housing construction report: August 2026 https://www.observedwork.com/resources/west-housing-construction-august-2026 Published / source checked: 2026-10-15 A dated Census report for the West region: housing permits, housing starts, and the limits of using regional totals to plan a supplier territory. Archive edition · August 2026 survey month · September 17, 2026 release (CB26-147) · Sources checked October 15, 2026. August values are preliminary in this release; July values are revised. This preserves that release’s vintage, not the latest revised estimates. The regional comparison West permits reached a rate of 313,000 housing units in August, up 1.6% from July and down 0.6% from August 2025. The starts estimate was 323,000. These are annualized rates, not counts of jobs issued or started during August. All table values are thousands of units at a seasonally adjusted annual rate (SAAR). Percent changes are Census’s published figures, calculated from unrounded data. Measure (thousands, SAAR) | Aug 2025 | Jul 2026 (r) | Aug 2026 (p) | Month change | Year change Permits, total | 315 | 308 | 313 | +1.6% | -0.6% Permits, single-family | 172 | 178 | 169 | -5.1% | -1.7% Starts, total | 307 | 244 | 323 | +32.4% | +5.2% Starts, single-family | 184 | 154 | 198 | +28.6% | +7.6% Census/HUD August 2026 release, September 17 vintage (PDF), tables 1a and 3a: https://www.census.gov/construction/nrc/pdf/newresconst_202608.pdf Why the starts increase is uncertain The total-starts month change is +32.4% with a 90% confidence interval of ±41.6 percentage points: −9.2% to +74.0%. Because it includes zero, this release does not establish that starts increased. The year change, +5.2% ±24.7 points, is also inconclusive. The single-family starts comparisons also include zero. A regional point estimate is a reason to ask better questions, not enough evidence to increase inventory. Census also cautions that monthly movements can be irregular. Permit estimates are subject to nonsampling error; do not apply the starts interval to permits. Census/HUD August 2026 release, September 17 vintage (PDF), tables 1a and 3a: https://www.census.gov/construction/nrc/pdf/newresconst_202608.pdf What the measures cover Permits measure new privately owned housing units authorized in permit-issuing places. Starts come from a separate survey and generally mark excavation for footings or foundations. They do not track the same month’s permits through construction. Neither series measures remodeling orders, all commercial construction, or contractor identities. SAAR removes estimated seasonal effects and multiplies a monthly rate by 12. It is not a forecast or the number built so far this year. One apartment building can account for many housing units. An issued permit does not establish that a project is available for bidding. Census Survey of Construction definitions: https://www.census.gov/construction/soc/definitions.html Census Building Permits Survey scope: https://www.census.gov/construction/bps/index.html One region, not a branch territory The West consists of Alaska, Arizona, California, Colorado, Hawaii, Idaho, Montana, Nevada, New Mexico, Oregon, Utah, Washington and Wyoming. This report does not estimate activity in any individual city or distributor service area. For a building-product supplier, use this as background for a territory review. Write down the counties served, product category, customer segment and local records needed. Then inspect local permit scope and contractor identity before assigning an account for research. No ratio here predicts material purchases or open bids. Census Survey of Construction definitions: https://www.census.gov/construction/soc/definitions.html Release vintage and corrections The downloadable table was transcribed from the archived September 17 release and visually checked against tables 1a and 3a on October 15, 2026. It intentionally retains August preliminary permits alongside starts from the same release. Census subsequently publishes revised BPS estimates separately; do not mix those vintages without labeling the change. For the next edition, check the official schedule, retain the new release file and compare revisions to overlapping months. Correct transcription errors here with a dated note; preserve this historical vintage when later estimates change. This page has no latest-data dashboard or automatic monthly publishing commitment. Revision log: initial edition, October 15, 2026. Census release schedule: https://www.census.gov/construction/bps/schedule.html Census historical releases: https://www.census.gov/construction/nrc/data/releases.html Request evidence for your own territory Distributors and manufacturers can request a coverage review with a geography and work category. ObservedWork will need to confirm source access and scope; this regional report does not establish a live local collection service. Request a B2B territory coverage review: https://www.observedwork.com/contact?topic=coverage&workflow=monitoring&territory=Census%20West%20region%20%E2%80%94%20specify%20cities%20and%20trade Review worksheet [ ] Keep survey month, release date and retrieval date separate. Notes / evidence / actual check date: ____________________ [ ] Use the same geography, unit and vintage for comparisons. Notes / evidence / actual check date: ____________________ [ ] Read uncertainty before calling a change a trend. Notes / evidence / actual check date: ____________________ [ ] Request city-level evidence before making an account decision. Notes / evidence / actual check date: ____________________ Responsible reviewer: ____________________ Unresolved questions: ____________________ Next review: ____________________ Census/HUD August 2026 release, September 17 vintage (PDF), tables 1a and 3a: https://www.census.gov/construction/nrc/pdf/newresconst_202608.pdf Census Survey of Construction definitions: https://www.census.gov/construction/soc/definitions.html Census BPS scope and revision process: https://www.census.gov/construction/bps/index.html Census release schedule: https://www.census.gov/construction/bps/schedule.html