Data Sources
Every dataset the site consumes, with its provider, refresh cadence and last-updated stamp. Live series refresh on every page load; curated snapshots list the publication date and get an amber banner on their pages once they age past twelve months.
Interest Rates
3 datasetsOvernight / policy rates for the major central banks, proxied through /api/fred.
DFFECBDFRIUDSOIAIRSTCI01JPM156NIUDSOIABIS-compiled 10Y sovereign yields, republished on FRED.
IRLTLT01*M156NGS10Direct BIS SDMX feed for central-bank policy rates and real effective exchange rates, used as a fallback layer alongside FRED and OECD.
Macro Fundamentals
9 datasetsNY.GDP.PCAP.PP.CDFP.CPI.TOTL.ZGSL.UEM.TOTL.ZSGC.DOD.TOTL.GD.ZSHousehold income or consumption Gini from national surveys. Coverage is uneven and survey years differ by country, so the "latest" value can be several years old for some economies. Powers /inequality.
SI.POV.GINILong-run r-vs-g, capital/income ratios, top income and wealth shares, the elephant curve, inheritance flows and top marginal tax rates, transcribed from Piketty (2014), Piketty & Saez (2003), the World Inequality Database, Milanovic (2016) and Saez & Zucman (2019). Historical figures before 1950 are reconstructions from tax and estate records, not survey data.
IMF WEO Oct-2024 general-government debt projections 2019-2029, S&P/Moody’s/Fitch sovereign ratings mid-2025, 5Y sovereign CDS Sep-2025, sovereign default database 2000-2024 (Bank of Canada / Reinhart-Rogoff), Fed/ECB/BOJ/PBOC quarter-end balance sheet snapshots 2007-2025, BIS household debt 2024. Powers /debt.
IMF International Financial Statistics as a secondary source for policy rates and general-government gross debt, used when FRED / OECD coverage is thin.
IFS/*/PMP_IXIFS/*/GGXWDG_GDPLabor
2 datasetsLabour force participation, youth (15-24) unemployment and female participation. ILO modelled estimates redistributed by the World Bank, so values are model output rather than direct survey counts. Powers Chapter 1 of /labor-ledger.
SL.TLF.CACT.ZSSL.UEM.1524.ZSSL.TLF.CACT.FE.ZSILO Global Wage Report 2024 median PPP hourly wages + real growth, OECD/ICTWSS union density 2023, ILO informal employment 2023, UN DESA World Population Prospects 2024 working-age 2000-2050, OECD Employment Outlook 2024 AI exposure + complementarity, ILO gender-LFP 2023, ILO youth-unemployment 2024. Powers /labor-ledger.
Trade
4 datasetsNE.RSB.GNFS.ZSPartner-level goods exports and imports from the UN Comtrade public API, used to compute partner diversity and concentration. Powers the live-data mode of /trading-places. Free-tier rate limits mean coverage can be partial.
Hand-compiled 2025 bilateral goods flow estimates between major economies, reconciled from national statistics offices and UN Comtrade. These are estimates, not an official series. Powers the /trade-network flow diagram.
WTO Applied Tariffs 2024, PIIE US-China tariff timeline 2018-2025, Baltic Dry + Drewry World Container Index snapshots, RCEP/USMCA/CPTPP/EU/AfCFTA membership matrix, USGS + SEMI supply-chain concentration, UNCTAD maritime chokepoints. Powers /trade-ledger.
Governance
1 datasetControl of Corruption, Government Effectiveness, Political Stability, Regulatory Quality, Rule of Law, Voice & Accountability. Fetched via source=3 (dedicated WGI database).
GOV_WGI_CC.ESTGOV_WGI_GE.ESTGOV_WGI_PV.ESTGOV_WGI_RQ.ESTGOV_WGI_RL.ESTGOV_WGI_VA.ESTHousing
1 datasetQ*R628BISQ*N628BISForeign Exchange
3 datasets27 currency pairs (16 USD-quoted, 11 derived crosses) with 30-day sparklines. Proxied through /api/frankfurter.
EUR/*USD/*DEXUSEUDEXJPUSDEXUSUKDEXCHUSDEXCAUSDEXUSAL+7 moreRolling 90-day return-correlation matrix computed client-side from FRED + Frankfurter daily rates. Derived rather than published, so it moves whenever the underlying rates move.
Resources & Commodities
2 datasetsOil (WTI, Brent), gold, iron ore, aluminium, copper, wheat. 30-day sparklines rendered client-side.
DCOILWTICODCOILBRENTEUGOLDAMGBD228NLBMPIORECRUSDMPALUMUSDMPCOPPUSDM+1 moreUS Energy Information Administration international data via /api/eia. Falls back to a static snapshot when EIA is unavailable.
Defense
8 datasetsSIPRI Military Expenditure Database published on the World Bank API. USA slot backfilled from FRED FDEFX (scaled to absolute US$).
MS.MIL.XPND.CDMS.MIL.XPND.GD.ZSMS.MIL.XPND.ZSMS.MIL.TOTL.P1Trend indicator values in constant 1990 US$.
MS.MIL.XPRT.KDMS.MIL.MPRT.KDRanked by arms-related revenue in the most recent SIPRI Top 100 release.
Curated 2018-2024 spending for 15 major spenders. Used as a fallback layer when the live WB fetch is unavailable; live values always win where present.
Estimated 2025 defense spending as % of GDP, from NATO Public Diplomacy Division annual release.
Deployed / reserve / retired warhead counts for the nine nuclear-armed states, per the Federation of American Scientists.
Uppsala Conflict Data Program v25.1. Annual deaths grouped by region plus active state-based conflict counts.
UN Fifth Committee approved peacekeeping budget and active mission count.
Cultural Capital
7 datasetsPublic education spending, tertiary enrolment, adult literacy.
SE.XPD.TOTL.GD.ZSSE.TER.ENRRSE.ADT.LITR.ZSCurated top-100 university list plus per-country top-university mapping.
BX.GSR.ROYL.CDBM.GSR.ROYL.CDCultural / natural / mixed heritage sites by country, plus UNESCO Creative Cities and creative-goods trade metrics.
Creative-industries exports as a share of goods trade, from UNCTAD Creative Economy Outlook.
International tourism receipts and arrivals, from UN Tourism Barometer.
PISA is a triennial 15-year-old assessment, so this is the oldest snapshot on the site by design rather than by neglect. The 2022 wave is the current release; the next wave is expected end of 2026. Reading / math / science mean scores by country.
Climate
9 datasetsAbsolute CO₂ (kt), per-capita CO₂ (t), and methane / N₂O expressed as kt CO₂-equivalent.
EN.ATM.CO2E.KTEN.ATM.CO2E.PCEN.ATM.METH.KT.CEEN.ATM.NOXE.KT.CECoal share, renewables share, fossil-fuel share of total energy consumption.
EG.ELC.COAL.ZSEG.ELC.RNEW.ZSEG.USE.COMM.FO.ZSEG.FEC.RNEW.ZSForest area (% of land) and terrestrial protected areas (% of land).
AG.LND.FRST.ZSER.LND.PTLD.ZSMean annual population-weighted exposure to PM2.5, benchmarked against the WHO 5 µg/m³ guideline.
EN.ATM.PM25.MC.M3Nationally Determined Contributions submitted to the UNFCCC NDC Registry. Falls back to 2030 target where a 2035 pledge has not yet been submitted.
Operating / under-construction / announced / retired GW of coal generation by country.
Cumulative pledged vs disbursed climate finance by donor, plus the annual $100Bn Copenhagen mobilisation series.
Annual counts of floods, storms, droughts, wildfires and extreme-temperature events grouped by continent.
Land-ocean temperature anomaly vs 1951-1980 base period. Snapshot of the annual mean series.
Energy
2 datasetsNet energy imports as a share of energy use (negative = net exporter) and share of population with electricity access. The import series lags by several years for most countries. Powers Chapter 1 of /energy-ledger.
EG.IMP.CONS.ZSEG.ELC.ACCS.ZSIEA Electricity 2025 generation mix, BNEF Global Storage Outlook 2024 battery build-out 2015-2030, IGU World LNG Report 2024 top flows, IAEA PRIS Sep-2025 reactor status + policy stance, EIA International Energy Statistics 2024 + BP Statistical Review proven reserves, IEA WEO 2024 capacity factors, IEA Efficiency 2024 intensity. Powers /energy-ledger.
Health
2 datasetsCurrent health expenditure (% of GDP), life expectancy at birth, under-5 mortality and physician density. Expenditure figures originate in the WHO Global Health Expenditure Database and typically lag by two to three years. Powers Chapter 1 of /health-ledger.
SH.XPD.CHEX.GD.ZSSP.DYN.LE00.INSH.DYN.MORTSH.MED.PHYS.ZSWHO Global Health Expenditure Database 2023 spend-outcome pairs, IHME Global Burden of Disease 2023 DALY series, Johns Hopkins Global Health Security Index 2021 + WHO JEE core capacity, Pharma Intelligence top-15 R&D 2024, WHO Mental Health Atlas 2020 prevalence + treatment gap, dual-burden obesity/undernutrition (WHO NCD Atlas 2022). Powers /health-ledger.
Mobility & Passport
1 datasetCurated visa policy matrix (visa-free / visa-on-arrival / eTA / visa-required) for major passports, plus mobility & advisory scoring. Powers the /cultural-capital passport tab.
Migration
4 datasetsPersonal remittances received (current US$). WB series updated annually; some low-income coverage is sparse.
BX.TRF.PWKR.CD.DTForeign-born population totals, updated via UN DESA into the WB indicator. Powers the Migration Ledger.
SM.POP.TOTLUNHCR data surfaced via the World Bank indicator API. Complemented by the curated UNHCR mid-2025 stock snapshot.
SM.POP.REFG.ORUNHCR mid-2025 refugee stocks by origin, KNOMAD top-25 remittance corridors, UN DESA International Migrant Stock 2024, Eurostat EU asylum applications 2015-2024, OECD talent migration (brain drain/gain), IOM Missing Migrants Project deaths by route 2014-2024. Powers /migration-ledger.
Technology & Innovation
5 datasetsStanford AI Index 2024/25 notable-model counts, Epoch AI frontier model release registry 2018-2025, SEMI + TrendForce foundry capacity Q2-2025, Stanford AI Index + CB Insights AI private-market investment 2018-2024, IEA Electricity 2025 data-centre energy projections, MacroPolo Global AI Talent Tracker, 2023-2025 AI regulation timeline (EU AI Act, US EOs, GAISI, PRC generative-AI rules). Powers /ai-ledger.
Individuals using the internet (% of population) and mobile-cellular subscriptions per 100 inhabitants, covering 2018-2024. The ITU DataHub publishes no open REST API, so /api/itu serves a transcribed snapshot of their published tables rather than fetching live.
Resident and non-resident patent applications by country, covering 2018-2023. WIPO IP Statistics is published as bulk downloads with no open REST API, so /api/wipo serves a transcribed snapshot rather than fetching live.
Individuals using the internet (% of population) and mobile cellular subscriptions per 100 people, originally compiled by the ITU and redistributed by the World Bank. The internet series counts any use in the previous three months at any speed, so it measures reach rather than quality of access. Subscriptions exceed 100 in many countries because of dual-SIM use.
IT.NET.USER.ZSIT.CEL.SETS.P2EU-27 R&D expenditure, EPO patent applications, and high-tech exports as share of total. Fetched from the Eurostat SDMX endpoint.
Every entry above is registered in src/app/data/dataProvenance.ts. When a live feed is proxied through a Next.js API route (World Bank, FRED, Frankfurter, EIA), the proxy adds server-side retry, IPv4-first DNS resolution and CDN caching headers on top of the upstream — see the /api/* routes for details.