Our methodology
Make the source visible. Preserve the definition. Explain the limits.
Where data comes from
We import a small selection of World Development Indicators through the official API. Each indicator records its original code, definition, providers, query URL and retrieval date. Dates labelled “source dataset update” refer to the source dataset, not necessarily the date of every observation.
Integrated sources and coverage
- World Bank · World Development Indicators
- 22 active annual historical indicators. Underlying providers, definitions and country-specific reference years are retained.
- United Nations · WPP 2024
- Ten annual demographic projection series for 2026, sourced 2025 gap values, and the separate World Population clock. Medium variant; these are projections, not observed outcomes.
- United Nations · WUP 2025
- 2026 urban-population projections using national urban definitions.
- Eurostat
- Separate monthly inflation and unemployment series for seven European countries. Monthly periods are never substituted for annual indicators.
- IRENA
- A separate renewable-electricity generation share series for 20 countries, through 2024. Generation is distinct from installed capacity.
- FAO via UNSD SDG database
- 2025 forest area as a proportion of land area for 213 matched profiles. Original country-reporting and estimation flags are preserved; normal status does not certify direct measurement.
- IEA / UNSD energy balances via UNSD SDG database
- Renewable share of total final energy consumption, 2023 and selected 2024 country values. Source estimation flags are retained; this is not renewable electricity alone.
Value types and chart conventions
Solid segments connect values explicitly identified as OBSERVED. Dashed segments include estimates, modelled estimates, forecasts or projections. Dotted segments mean the source status has not been classified; a blank source flag never establishes an observed value. Diamonds mark 2026 only. Missing years remain gaps. Official supplements fill missing country-year cells without overwriting the historical WDI snapshot. Cards, tables, charts and chart exports resolve the same source rows.
Recent monthly releases
Eurostat monthly inflation and unemployment are imported separately for seven European countries. Exact query dimensions, reference periods and source flags are retained. An estimated or provisional monthly figure is not a final annual result. Blank source flags are not treated as proof that a figure is final.
How data is processed
Imports are staged and validated before publication. Country codes, indicator codes, dates, duplicates and values are checked. Nulls stay null; graphs leave gaps instead of inventing intermediate values. Display rounding is identified; CSV exports preserve source precision.
How comparisons work
Country comparisons select the latest common year within the chosen range with the same indicator, unit, source edition and value type. An older comparable year can be used when newer values do not match across countries. The same indicator code does not eliminate national methodological differences. Urban definitions can differ, and crude mortality rates depend on population age structure. We therefore show descriptive series rather than league tables or causal conclusions.
About the editorial archive
The retained archive is outside the current country-and-data navigation. Its selection is editorial and covers health, education and living conditions. It is not an independent systematic review. Overviews preserve links to the original research and distinguish study findings from the separate statistical series used in our charts.
How AI is used
AI assisted the initial summaries and site creation. No language model generates observation values. Summaries are labelled; independent scientific review has not been completed for the retained editorial archive.
Scope and the editorial archive
Sensitive population categories
No race or ethnicity breakdowns are published in this release. Future records must preserve the original source classification, country, period and definition; categories will not be automatically equated across sources.
Corrections and versions
Source payload hashes and import logs allow comparisons between imports. Corrected data must pass validation and be republished; silent overwriting during a failed import is avoided. To report an issue, use the contact and corrections page. The published snapshot and source evidence identify the edition being displayed.
Health information
This site presents research and population-level data and is not medical advice. Correlation does not establish causation.