How to read this site
This dashboard answers one question — how is the Netherlands doing, compared with 10, 20, 30 and 40 years ago and with neighbouring countries — using 44 indicators chosen from three established frameworks: the CBS Monitor of Well-being, the OECD Better Life framework and the Eurostat Quality of Life framework. It is built to be usable by people who disagree with each other.
01Source-first
Only official statistics (CBS, Eurostat, OECD, RIVM) form canonical series. Institute models — like the housing-shortage estimate — are shown, but always labelled as model estimates, never blended with measured data.
02No verdicts
No composite "Netherlands score", no red or green, no causal claims. Trend arrows show direction over the last eight years and nothing more; whether a movement is good is yours to judge.
03Breaks and gaps stay visible
When a statistical method changed (income statistics in 2011, the safety survey in 2021), the chart carries an amber marker at that point. When data simply doesn't exist — dwelling completions 2012–2014 — the line stops. We never interpolate across a hole.
04Revisions are kept
Statistical agencies revise. Every number we ever fetched stays stored with its retrieval date; a revised value is a new row, not an overwrite. The published database makes this auditable.
05Honest comparisons only
A peer-country chart appears only when an internationally harmonized series exists. Where the Dutch definition differs from the harmonized one, both are shown separately and the difference is stated. No harmonized series — no comparison.
Reading the marks
- ↗ ↘ →
- Direction of the last eight years (ordinary least squares over annual values, only shown when statistically significant at 5% — the Monitor Brede Welvaart convention). An arrow is never a judgement.
- A B C D
- Source confidence: A = official statistic; B = official but with known measurement issues; C = institute model; D = weak or experimental.
- ⌇ amber
- The honesty layer: series breaks, method changes, model estimates and short series. Amber never means “bad” — it means “read the note”.
- 1y 5y 10y
- Change versus one, five and ten years earlier, in the indicator’s own unit. An em-dash means the series doesn’t reach back that far — absence is shown, not hidden.
Derived numbers
A few series are computed by us from official inputs — industry’s share of value added, investment and government consumption as % of GDP, real (CPI-deflated) house prices, emissions per capita, net enterprise growth, employer enterprises. Where money is involved we prefer %-of-GDP over CPI-deflation (no deflator choice to defend); the real house-price series divides the house price index by the CBS consumer price index, both 2015 = 100. Each carries a versioned, published formula, and every computed value stores references to the exact source observations it was derived from. Formulas live in the public repository (kpi_ingestion/derive.py).
Roadmap
V1 is the credible core: 44 indicators. Planned next: ~26 fast-follow series (energy costs, wealth distribution, waiting times, teacher shortage), a Dutch translation, “since you were born” views, and curated paradox pairings. Some indicators wait deliberately — mental-health waiting times until the national data gap closes (~2027), water quality until the 2027 KRW reporting round. Missing means not credible yet, not forgotten. See sources for the live pipeline status.