Methodology

Every figure on buurtreport is derived from Dutch government open data. If the pipeline emits a null value, the tile hides — no number is ever fabricated to fill a slot. Below is the source list, refresh cadence, and how the core composite grade is computed.

Data sources

• CBS Kerncijfers wijken en buurten — population, age, origin, WOZ, income, education (86165NED 2025, 85984NED 2024). Refresh: annual.
• PDOK / BAG cadastre — buurt geometries + BAG-pand identification for address lookup. Refresh: monthly.
• Kadaster / CBS 83625NED — annual gemeente sale price + 85792NED / 85819NED quarterly per-region price index. Refresh: quarterly.
• Politie 47022NED — registered crime incidents by category and buurt, trailing 12 months. Refresh: monthly.
• DUO scholen — primary and secondary school locations, denominations, addresses. Refresh: monthly.
• Leefbaarometer 3.0 — 5-dimension livability index per buurt (national government). Refresh: biennial.
• Kiesraad TK 2021 / 2023 / 2025 — per-stembureau vote share for the LGBT+ climate composite. Refresh: per election cycle.
• RIVM Luchtmeetnet / Klimaateffectatlas — PM2.5, NO2 (live), heat stress, pluvial flood, foundation risk (static). Refresh: live / static.
• OpenStreetMap — amenities (supermarket, GP, pharmacy, park, gym), LGBT+ venues, parking data. Refresh: weekly.
• NS GTFS — nearest train station + walk-minute commute matrix. Refresh: static.

Composite grade formula

The letter grade in the hero of each buurt page is the mean of five per-city percentile pillars: housing (owner-occupied share), schools (distance-weighted school-quality z-score vs the NL mean), safety (crimes per 1,000, inverted), commute (walk-minutes to nearest station, inverted), amenities (median walking distance across supermarket / pharmacy / park / gym, inverted). Cost (WOZ) is shown as its own tile but is a preference axis and is not part of the composite. Emitted only when at least 3 of the 5 pillars are present. The LGBT+ climate composite is a separate axis, documented on the LGBT+ tile.

Population gates

• Under 100 residents — a "Small sample" caveat is shown above the Residents section; descriptive stats still render.
• Under 300 — per-1,000 rates (crime density) hidden; too noisy to publish.
• Under 400 — hero letter grade and NL rankings hidden even when the underlying pillar was computed.

Editorial commitments

No advertising, no lead generation, no user tracking beyond aggregate Cloudflare Web Analytics. No scraped-behind-paywall data. No fabricated numbers to fill a slot. All source URLs deep-link to the specific CBS dataset code, PDOK service, or DUO record they come from — corrections at [email protected].