Derived Metrics Model

Currency & price base

The direct indicator EN.GHG.CO2.RT.GDP.PP.KD uses PPP constant 2021 international dollars for the GDP denominator. This is the IPCC AR6 WG3 convention (Chapter 2, Box 2.2). PPP-based intensity values are not directly comparable to legacy calculations using nominal current-USD GDP — the switch typically reduces intensity by 30–40% for middle-income countries with large PPP adjustments.

Coverage & sample bias

EDGAR-JRC covers 215 countries and territories (2023 release). The WLD aggregate used here includes all reporting countries. Emissions from international shipping and aviation (~2.5% of global total) are allocated to bunker fuels and may understate true attributable emissions for high-trade economies.

Uncertainty bounds

EDGAR-JRC reports ±5–15% uncertainty on national totals depending on sector and country data quality (see EDGAR Uncertainty Report 2023). GDP PPP conversion adds ±2–5% from ICP benchmark uncertainty. Combined propagated uncertainty is approximately ±8–18% at the country level; the global aggregate is tighter at ±5–10%.

Institutional reference

Methodology follows IPCC AR6 Working Group III, Chapter 2 (Climate change mitigation). For country-level deep dives, consult the EDGAR-JRC country profiles and IEA CO₂ Emissions from Fuel Combustion (paywall). Our indicator aligns with World Bank's published GHG intensity series.

Currency choice: nominal current USD

SIPRI publishes military spending in current (nominal) USD, which reflects actual transaction prices but is sensitive to exchange-rate fluctuations. A country that devalues its currency will show a drop in USD spending even if domestic military capacity is unchanged. For real-terms comparisons use SIPRI's constant-USD series (base year 2021) or the military-spending-pct-gdp ratio, which is currency-neutral.

Sample bias & coverage

SIPRI data covers 173 countries; the WLD aggregate includes all countries with sufficient reporting. Roughly 40 small or conflict-affected states have partial or estimated data. SIPRI flags these with a note in its database. The global per-capita figure is therefore a coverage-weighted average that may understate spending in opaque military states (e.g. North Korea).

Uncertainty bounds

SIPRI estimates ±3–8% uncertainty for countries with strong parliamentary oversight (e.g. NATO members) and ±15–25% for states with limited transparency. The population denominator (World Bank SP.POP.TOTL) has ±0.5% uncertainty. Combined propagated uncertainty for the global per-capita figure is approximately ±4–9%.

Institutional reference

Methodology and definitions follow the SIPRI Military Expenditure Database FAQ and the NATO definition of defence expenditure. SIPRI uses a broad definition that includes all spending on armed forces, ministries of defence, paramilitary forces, and military space programmes.

Government-only scope

UNESCO SE.XPD.TOTL.GD.ZS captures government expenditure on education only. OECD Education at a Glance (2023) shows that private household spending adds 20–60% in OECD countries, and philanthropic and corporate training budgets add further amounts. Total (public + private) education investment globally is therefore substantially higher than this indicator suggests — making cross-country comparisons valid only within the government-funded dimension.

Sample bias & coverage

UNESCO coverage for SE.XPD.TOTL.GD.ZS is approximately 140 countries (2022 reference year). The WLD aggregate extrapolates missing countries using regional averages, which introduces a ~5–8% estimation error in the global total. Low-income countries with fragmented education budgets are the primary source of uncertainty.

Uncertainty bounds

UNESCO estimates ±5–10% uncertainty on education spending ratios for countries with multiple reporting channels (federal vs. state vs. municipality). GDP uncertainty adds ±2–3% (World Bank national accounts). Population adds ±0.5%. Combined propagated uncertainty for the global per-capita figure is approximately ±6–11%.

Institutional reference

Methodology follows UNESCO Institute for Statistics (UIS) definitions in the ISCED 2011 framework. See UIS data centre for country-level education finance profiles and the OECD Education at a Glance annual report for comparative analysis across 40+ countries.

Currency choice: PPP constant 2021 USD

R&D per capita uses PPP constant 2021 international dollars (NY.GDP.MKTP.PP.KD) as the GDP base, per OECD Frascati Manual convention. PPP adjustment accounts for the fact that researchers and lab supplies cost very different amounts in different countries — a US$1M R&D budget buys far more research capacity in India than in Switzerland. PPP-based values are ~54% higher than current-USD values for countries like China and India, reflecting real purchasing power parity.

Sample bias & coverage

UNESCO GERD data covers approximately 100 countries with consistent reporting; the WLD aggregate covers countries responsible for ~95% of global R&D spend. About 50 low-income countries have no GERD data — their contribution is estimated at <1% of global total and excluded from the denominator accordingly. The aggregate slightly overstates per-capita spending for the true global population.

Uncertainty bounds

UNESCO estimates ±5–10% uncertainty on GERD ratios depending on whether the country uses a dedicated survey or estimates from administrative records (see OECD MSTI 2023). PPP conversion adds ±3–5% from ICP benchmark uncertainty. Population adds ±0.5%. Combined propagated uncertainty for the global per-capita figure is approximately ±7–12%.

Institutional reference

Definitions follow the OECD Frascati Manual 2015 (7th edition), which governs international R&D statistics. The OECD Main Science and Technology Indicators (MSTI) database provides the most current country-level GERD data. UNESCO UIS is the primary multilateral collector for non-OECD countries.