Half of humanity lives in seven countries

There are 8.16 billion of us in 2024. Line up the countries by population and it takes only seven — India, China, the United States, Indonesia, Pakistan, Nigeria and Brazil — to account for more than half.

Population is the most lopsided distribution in world statistics, and lists bury the lopsidedness — "India 1.45 billion, China 1.42 billion, United States 345 million…" reads as a ranking, not a shape. A treemap makes the shape physical: each country is a rectangle whose area is its population, so the two giants aren't just top of a list — they're most of the picture.

The seven that make half

India (17.8% of humanity) and China (17.4%) alone are 35.2% of everyone. Add the United States (4.2%), Indonesia (3.5%), Pakistan (3.1%), Nigeria (2.9%) and Brazil (2.6%) and you cross 51.4% at country number seven. On the chart those seven carry a bright halo, so you can see the exact tiles that hold half the species.

Treemap of world population with large haloed India and China tiles among region-coloured country tiles
World population by country, grouped by region, 2024. Open the interactive version → Source: UN World Population Prospects 2024 (CC BY 3.0 IGO).

Where everyone is

Group the tiles into regions and one continent dominates: Asia is 58.9% of humanity — nearly three in five people. Africa is 18.6% and, crucially, the only region still growing quickly; Europe is 9.1%, North America 7.5%, South America 5.3%, and all of Oceania is 0.6%. The twenty-first century's demographic story is written in the gap between the Asian giants that are plateauing and an African bloc that is not.

Why the chart is built the way it is

A treemap earns its place only when magnitudes span orders of magnitude — and populations run from 1.45 billion to a few thousand, so they do. Countries with at least 20 million people are named in their own tile; the long tail of smaller nations folds into one honest "Other [region]" tile per continent, italicised, rather than shattering the chart into 230 unreadable slivers. Big tiles get a name and a number, medium tiles get the country's ISO code, and tiles too small to label cleanly get nothing rather than a smudge — every label stays readable at 100%. Regions are outlined and labelled with their share, and the seven "half" countries carry the accent halo so the headline survives inside the graphic.

What the data made us check

The first thing a treemap owes you is that the parts add up — so we checked. And the first pass didn't: the 233 countries summed to 8.13 billion against the UN's 8.16 billion world total, a 0.38% hole. The culprit was punctuation. The UN files a few places with commas inside their names — "China, Taiwan Province of China", "China, Hong Kong SAR", "China, Macao SAR" — and a naive comma-split shredded those rows, quietly dropping about 31 million people, Taiwan among them. A quote-aware parser put them back and closed the gap to 0.000%. Two framing notes stated on the chart: figures are the UN's medium-variant estimate for mid-2024, and regions are UN continents.

What your data needs to look like

One row per entity, its value, and the group it belongs to. The agent builds the nesting and sizes every tile.

countryregionpopulation
IndiaAsia1450936
NigeriaAfrica232679
BrazilSouth America211999

Novice tip: bring raw counts, not percentages — a treemap sizes tiles by the value, and it checks that your group totals sum to your grand total before drawing. If a few entities are tiny, don't drop them: label a "fold small into Other" threshold and the chart keeps them honestly counted instead of silently missing.

The takeaway

"8 billion people" is a number too big to feel. Seven rectangles holding half of it is not. The concentration is the story — and the countries in those seven tiles are where the century's questions about work, ageing, migration and growth will mostly be answered.

Have a magnitude-and-group story in your numbers? Drop it into the AK.VIZ agent for three chart proposals, see more of the work, or read about bespoke commissions.

Have a dataset in mind?

Get three chart proposals on your data, free.