G20 Macro Panel Pack: a data pack of three World Bank WDI series joined into one long CSV, plus a wide cut, the source files, and three latest-year slices. 20 entities, 19 G20 national members plus the EU27 aggregate, 1960 to 2025, one row per entity-year. The panel holds 1,320 rows and 6 columns: iso3, entity, year, population, rural_pct, life_expectancy. It was built by matching iso3 plus year across three source files: total population (SP.POP.TOTL, 1,320 rows), rural share of population (SP.RUR.TOTL.ZS, 1,320 rows), life expectancy at birth (SP.DYN.LE00.IN, 1,300 rows). 3,940 upstream rows become 1,320 panel rows. The join dropped nothing and found no unmatched key: 1,300 rows carry all three indicators, and the 20 rows with an empty life_expectancy are all year 2025, where the WDI mortality series has no value yet. No imputation, no interpolation, no estimate added by us. Rural share keeps the API's own precision. 11 files ship in the zip: g20-macro-panel.csv, g20-macro-panel-wide.csv (20 rows by 199 columns, population_1960 through life_expectancy_2024), the three source CSVs byte for byte, g20-population-2025.csv, g20-rural-2025.csv, g20-life-expectancy-2024.csv, README.md, LICENSE.txt, description.txt. pandas, DuckDB and Excel all read these files directly. World Bank Open Data, CC BY 4.0, retrieved 2026-09-10 from the WDI update of 2026-07-13.