Geography and method remain visible beside every metric. Jail and homelessness are countywide; consumer prices cover the metropolitan area. The homelessness estimates are discrete Point-in-Time observations with a method change in 2022, not an annual census.
Net housing completions uses the City's permit terminology: new units finaled (a final building inspection or temporary certificate of occupancy) minus demolished units recorded in the same completed-permits source. It covers complete 2015–2025 calendar years; the source's 2026 current-year-quarter snapshot is omitted. It does not include congregate housing or measure the running housing stock.
The derived offense-rate view divides SPD offense records by Seattle population and scales by 100,000. The real budget view shows adopted operating budget per resident in 2025 dollars, using the Seattle-area CPI series. Both use only years with all required observations; missing years are left out and never interpolated.
Actual spending stays separate from adopted budgets until fund and accounting changes can be reconciled across annual financial reports. Median gross rent is rent paid by incumbent renters, not current asking rent or a measure of rent burden.
Renter cost burden is the estimated share of Seattle renter-occupied housing units with a computable gross-rent-to-income percentage where gross rent is at least 30% of household income. It sums B25070's 30.0–34.9, 35.0–39.9, 40.0–49.9, and 50.0%+ categories, then divides by all computed categories. Census's excluded Not computed group contains units with no rent paid and households reporting no income or a net loss. This is not all renters, asking rent, housing supply, or a causal household-wellbeing measure. Standard 2020 ACS 1-year data are unavailable and 2025 had not been released as of the October 1, 2026 audit; gaps are left blank. The source artifact retains the component counts, published 90% MOEs, and approximate derived MOE.