Understanding Global Economic Inequality Data
Our World in Data
- Global economic inequality is not inevitable; institutions and policy choices significantly shape income distribution within countries.
- While wealth gaps within individual nations are notable, massive income disparities between countries remain a primary driver of global inequality.
- High-income countries often misperceive global living standards; the global median income is approximately $290 per month, or under $10 a day.
- Income growth has occurred across the global distribution since 1990, with significant reductions in extreme poverty.
Key Data Sources
- World Bank Poverty and Inequality Platform (PIP): Provides the broadest global coverage, using national survey data to map incomes from the richest to the poorest worldwide.
- Luxembourg Income Study (LIS): Focuses on high-quality cross-national comparability by harmonizing survey definitions, though it covers fewer countries, particularly in the low-income spectrum.
- World Inequality Database (WID): Uses administrative tax records and national accounts to track top-end income concentration (the "missing rich" problem) that traditional surveys frequently undercount.
Measurement Methodologies and Limitations
- Income vs. Consumption: World Bank data often mixes income and consumption metrics. Income-based measures generally produce higher Gini coefficients than consumption-based ones because richer households save more.
- International Dollars: All comparisons use international dollars, which adjust for purchasing power parity (PPP) and inflation, allowing for meaningful cross-country living standard comparisons.
- WID Scope: WID measures "net national income," which includes non-cash benefits like imputed rent and undistributed corporate profits. This provides a fuller economic picture but differs from cash-based household income concepts.
- Data Gaps: Household surveys often struggle to reach the very rich, the homeless, or populations in conflict zones, leading to inherent limitations in all datasets.