Population density
People per square kilometre of total area.
How it is measured
Computed here as World Bank population divided by the area figure on each profile, which comes from the vendored reference dataset or, where that is unusable, the CIA archive.
Where the number came from
Population density is one of the oldest derived statistics in geography, and it was central to the way nineteenth and early twentieth century writers reasoned about nations, sometimes usefully and sometimes disastrously: the concept of Lebensraum was an argument built on density figures. Modern quantitative geography has largely moved past the country-level number, because it is a textbook case of what statisticians call the ecological fallacy, treating an average over a large heterogeneous area as if it described conditions anywhere within it. The serious contemporary work uses gridded population products such as WorldPop or the Global Human Settlement Layer, which distribute people across square-kilometre cells using satellite-derived built-up area, and from which measures like population-weighted density are computed. The national figure survives as a quick comparator, and it is worth knowing exactly how much it conceals.
Worked example: Egypt at 118 per km² is more crowded than Bangladesh at 1,191
Taken at face value our figures say Bangladesh, at about 1,191 people per square kilometre, is roughly ten times as crowded as Egypt at about 118, and that Egypt is an ordinary middling country comparable to the global middle of the table. The Bangladeshi figure is broadly meaningful, because the country is a densely settled delta and people are distributed across most of it. The Egyptian figure is close to meaningless. Almost the entire Egyptian population lives in the Nile valley and delta, a small percentage of the national territory; the rest is desert with effectively nobody in it. Recompute Egyptian density over the inhabited strip and it exceeds Bangladesh comfortably. The same distortion applies wherever a country carries a large empty region: our figure for Mongolia is about 2.3 per square kilometre and for Greenland a small fraction of one, which tells you those states own a lot of ground rather than anything about how their people live. At the other extreme, Singapore at roughly 8,500 and Macau higher still are city-states whose entire territory is urban; comparing them to countries is comparing a city to a country, and the number will do it without complaint.
Figures read from our own population density table; each row there carries its observation year and whether it came from a live feed or the frozen archive.
Our data source
Computed here, not fetched: the live World Bank population divided by the area figure described on the area page (mledoze/countries under ODbL for most entities, CIA archive for the rest). Because one input is live and the other is static, the density figure carries the population’s observation year.
Refresh cadence: Follows the population refresh; the denominator effectively never changes.
What to watch for
Country-level density hides everything interesting: Egypt averages about 118 people per km² but its population is concentrated on a few percent of the land along the Nile. City-state densities (Monaco, Singapore) are a different phenomenon from country densities. Because the denominator is total area, countries with large inland water bodies or vast uninhabitable terrain are systematically flattered.