The dataset is built from 6 core entities: a Source backs every MetricObservation and CitationFact; observations roll up into an IndicatorSeries; ProvenanceRecord captures lineage; and VersionDiff records what changed between versions. Each entity has an authoritative JSON Schema.
- 6 entities; every figure links to a Source.
- Authoritative shapes: the JSON Schemas (/schema-report).
- Machine-readable at /data-model/data.json.
Source
A cited publisher of a figure (e.g. an official agency or intergovernmental body), with access metadata and citation-readiness.
Key fields: id, title, publisher, url, accessedDate. Relates to: MetricObservation.sourceId, CitationFact.sourceId. schema
MetricObservation
A single quantitative observation of a metric for a subject at a point in time, live or sample, with confidence and review dates.
Key fields: id, metricId, value, unit, asOf, dataMode, confidence, sourceId, lastReviewed. Relates to: Source, IndicatorSeries, ProvenanceRecord. schema
IndicatorSeries
An ordered set of observations of one metric over time (a time series snapshot).
Key fields: id, title, snapshots. Relates to: MetricObservation. schema
CitationFact
A source-linked qualitative or quantitative fact attached to a subject, with a confidence level.
Key fields: id, subjectType, subjectId, sourceId, confidence. Relates to: Source. schema
ProvenanceRecord
The lineage of a record: which source backs it, its type, and review metadata. Derived from the records above.
Key fields: recordId, recordType, sourceId, sourceName. Relates to: Source, MetricObservation, CitationFact. schema
VersionDiff
The value-level difference between two materialized versions (record ids and field names that changed).
Key fields: version, previousVersion, valueDelta. Relates to: MetricObservation, Source. schema
More
Schema report: /schema-report · data dictionary: /data-dictionary · machine-readable: /data-model/data.json.