Nothing is hidden here
A composite index that ranks countries earns trust only by showing its working. This page is that entire working: the six indicators, their weights, the sources, how the score is built and how to reproduce or cite it.
Six indicators, six weights
Each indicator is a recognised public statistic. The weight is its share of the composite score. In v0 every indicator points the same way: a higher value means more cracks.
| Indicator | Unit | Weight | Direction |
|---|---|---|---|
| Problematic debt Share of households with debts they can no longer repay. The first crack most people fall through, and the one that pushes the others open. | % of households | 19% | Higher is worse |
| Number of evictions Court-ordered evictions per 10,000 households. A direct measure of how often housing insecurity is allowed to reach its endpoint. | per 10,000 households | 15% | Higher is worse |
| Mental-health treatment gap Share of people with an identified mental-health care need who receive no care. The distance between needing help and reaching it. | % unmet need | 15% | Higher is worse |
| Unmet youth care Share of young people referred for care who wait longer than the recommended term or are never helped. A crack measured early in a life. | % unmet need | 12% | Higher is worse |
| Food insecurity Share of the population reporting moderate or severe food insecurity. The point at which a financial shortfall reaches the kitchen table. | % of the population | 12% | Higher is worse |
| Severe material deprivation Share of the population unable to afford several basic necessities that most people take for granted. A widely used floor for material insecurity. | % of the population | 12% | Higher is worse |
| Social assistance People on social assistance (bijstand) per 1,000 inhabitants. Social-assistance receipt reflects income insecurity, so a higher value counts as more strain on the safety net. | per 1,000 inhabitants | 8% | Higher is worse |
| Homelessness Estimated homeless people per 10,000 inhabitants. A modelled estimate at municipality level, as no per-municipality open measurement exists. Homelessness is an unambiguous sign of a torn safety net. | per 10,000 inhabitants | 8% | Higher is worse |
| Total | 100% |
From four steps to one number
Collect
For every country, six public country-level aggregates are retrieved from the sources below. The connectors are fail-soft: a missing or slow source never breaks a run, and a static fallback keeps the index complete and reproducible.
Normalise
Each raw value is rescaled to a 0-to-100 score and direction-corrected. In v0 every indicator is framed so that a higher value means more cracks, so the scores can be combined without sign confusion.
Weight and combine
The six normalised scores are combined with the fixed weights below into one composite score. When a country is missing an indicator, the weights are redistributed across only the present indicators, so no country is unfairly penalised by a gap in the public data.
Rank and point forward
Countries are ranked from low to high on the composite score. For each country the weakest indicator with the highest weight is identified and linked to the most relevant constructive change, published as the country’s "fix".
The composite score is a weighted average, not a black box. With the six values and the six weights, the score for each of the 40 countries can be recomputed by hand.
How hard each figure is
The Cracks Index is a v0. Not every value is fetched live yet. Some cells come straight from an open data source, others are still a modelled estimate that keeps the index complete until the real source is connected. We do not hide that difference, we label it.
A measured cell is a value genuinely fetched for this area and this indicator from Eurostat, the OECD, the World Bank or CBS StatLine. The source year beside it is the real year of that statistic.
A modelled cell is a curated v0 estimate. It is used when a source for this area is not yet connected or returned nothing usable. It is a careful approximation, not a measurement, and on the country and area pages it is labelled as an estimate.
How to tell them apart
On every country and area page each indicator carries a quiet label: measured or estimate. At the top there is a reliability indication, for example four of six indicators measured, so you can see at a glance how hard the composite figure is.
Coverage grows weekly. Each round the export pipeline fetches more cells live; a figure that is an estimate today may be measured next week. The method does not change, only the share of real data rises.
Where the numbers come from
The Cracks Index adds no new data collection. It rests on established, reputable public statistics, and that is what makes it both reproducible and free of personal data.
Reproducibility
The full pipeline is public: the data connectors, the normalisation, the weighting and the composite score. Anyone can re-run it against the same public sources and arrive at the same ranking. The Cracks Index is published as a living v0 and treats methodological criticism as a contribution, not a threat.
Privacy by design
The Cracks Index runs entirely on public, aggregated country-level data. At no stage is personal data read, stored or processed. It is the first world index of its kind that you can verify yourself, built without touching a single piece of personal data.
Licence and citation
The Cracks Index dataset is published under Creative Commons Attribution 4.0 (CC-BY-4.0). You may reuse, distribute and build on it, including commercially, as long as Fynqo is credited. When citing, use this line:
Fynqo (2026). The Cracks Index, v1-20260518. Snapshot of 2026-05-18. Licensed under CC-BY-4.0. https://fynqo.app/cracks-index/
What people ask first
Does The Cracks Index use personal data?
No. Every value is a public country-level aggregate from Eurostat, the OECD, the World Bank or CBS StatLine. No individual is identified, tracked or modelled, so the index itself needs no GDPR processing basis.
Does the Cracks Clock count real people in real time?
No. The Cracks Clock is a modelled statistical estimate, derived from public annual figures and shown at a readable pace. It is explicitly labelled as an estimate and follows no real event or person.
How can the methodology be checked?
The full pipeline, from data sources to the composite calculation, is public. A critic can re-run it and arrive at the same numbers. The Cracks Index is published as a living v0 and invites methodological review.
Which licence applies to the data?
The Cracks Index dataset is published under Creative Commons Attribution 4.0 (CC-BY-4.0). The data may be reused, distributed and built on, including commercially, as long as Fynqo is credited.
See the ranking the method produces
40 countries, each one composite score, every number traceable to the steps on this page.