World Population Pyramid data methodology
How we source, transform, label, calculate, and quality-check every population profile and ranking.
Primary source
All population totals, age-by-sex distributions, median ages, fertility rates, life expectancy (at birth, by sex, and at 65), infant and under-five mortality, births, deaths, net migration, crude rates, sex ratios, mean age at childbearing, and growth rates come from the United Nations, Department of Economic and Social Affairs, Population Division, World Population Prospects 2024 Revision.
The build pipeline downloads two official compressed CSV files: WPP2024_Demographic_Indicators_Medium.csv.gz and WPP2024_PopulationByAge5GroupSex_Medium.csv.gz.
Estimates versus projections
Years through 2023 are historical estimates in WPP 2024. Years from 2024 through 2100 are medium-scenario projections. The site calls 2026 a projection even when “current” is convenient shorthand in navigation or rankings.
A projection is conditional. Future fertility, mortality, and migration can differ from the scenario. For formal work involving uncertainty, use the UN’s probabilistic intervals and alternative scenarios.
Reference date and units
Population totals use the UN’s 1 July (mid-year) values. Source population fields are published in thousands and converted to people. Five-year age groups run from 0–4 through 95–99, plus 100+.
Full integers are provided so readers can inspect and reproduce sums; that formatting should not be interpreted as exact-person certainty. Because totals, male/female values, and age groups are converted and rounded independently, adding displayed components can differ from a published total by a few people. Headline figures are compactly rounded.
Calculated age shares
We sum male and female population counts and group them as follows:
- Children: ages 0–14
- Conventional working age: ages 15–64
- Older population: ages 65 and above
Total age dependency is calculated as (ages 0–14 + ages 65+) ÷ ages 15–64 × 100. Child dependency uses ages 0–14 in the numerator, and old-age dependency uses ages 65+. All three use ages 15–64 as the denominator. They are age-structure measures, not literal worker-to-dependant ratios.
Descriptive structure labels
Summary cards use three descriptive labels based on rounded age shares:
- Youthful: the under-15 share is at least 32%.
- Older: the 65+ share is at least 18%, or the under-15 share is below 19%.
- Transitional: profiles between those thresholds.
These are navigation aids, not official UN classifications. Exact age shares and the full pyramid take precedence.
Rankings and peak references
Rankings compare all 237 entries classed as Country/Area by the UN, including territories and statistical areas. Most use the 2026 medium projection; the 2050 and 2100 tables state their projection year explicitly. Ranks use stored source precision even when the displayed value is rounded. Equal stored values share the same rank, and the next distinct value skips ahead (1, 2, 2, 4), consistently across ranking tables, CSV downloads, and country profiles. In the peak-year table, places peaking in the same year share a rank and are listed by peak population.
Peak timing is calculated from every annual 1 July population value from 1950 through 2100, not only the decade points displayed in profile tables. A maximum at the 1950 boundary is labeled “1950 or earlier” because this dataset cannot establish an earlier all-time peak. A later maximum through 2023 is labeled historical, although it remains an estimate. A maximum from 2024 through 2100 is labeled projected. If the population is still rising at the 2100 boundary, the site reports that condition rather than inventing a peak beyond the source horizon.
Births, deaths, and migration
Annual births, deaths, and net migration are the UN’s calendar-year components of change, converted from thousands to people. Crude birth, death, and net migration rates are per 1,000 people in the mid-year population; infant and under-five mortality are per 1,000 live births. The births–deaths–migration and population charts plot every year from 1950 to 2100; values after 2023 are drawn dashed on a shaded band to mark the medium projection. The year in which deaths first exceed births is the first annual value where UN deaths are larger than UN births; it is not smoothed.
Distinctive cohorts and similar age structures
A five-year band under age 80 is flagged as a bulge when it is larger than both neighbouring bands and at least 10% above their average, and as a notch when it is smaller than both and at least 10% below their average. Birth years are the reference year minus the band’s age limits, so a 25–29 band in 2026 was born about 1997–2001. Flags describe the shape; they do not identify whether births, deaths, or migration produced it.
“Most similar age structure” compares the share of the total population in each of the 21 five-year bands. The distance between two places is the sum of the absolute differences in those shares; the displayed overlap is 100% minus half that distance. The year in which people aged 65+ first outnumber children under 15 is located to the five-year pyramid step in which it first occurs.
Downloads and chart assets
Country CSV downloads contain every age-by-sex observation used by the interactive chart, every five years from 1950 to 2100 plus 2023 and the reference year, with data-status, source, and revision columns. Ranking CSV files use unformatted numeric values so they can be sorted and calculated without removing thousands separators. Downloadable SVG charts are generated directly from the same checked-in age data; PNG files are rendered from those SVGs in the browser, so the two image formats contain the same values and labels.
Regions and names
Six broad geographic regions and the 21 UN subregions within them are used for navigation; their totals and age groups are the UN’s own aggregates. Common English short names such as “United States” and “South Korea” improve findability; the underlying UN location name and ISO code remain stored in the dataset. Territories and areas included separately by the UN may also have separate pages here.
Updates and quality control
The source pipeline validates the shared schema, sorts age groups, checks required current and future points, and generates static data at build time. A source revision update requires a new build and a review of field definitions, country coverage, and methodology—not just swapping a date in the interface.
Possible errors include transformation bugs, ambiguous short names, or explanatory mistakes. Please report a specific page and value through the contact page. Corrections follow the published policy.