Brussels Is 2,200 Times More Crowded Than Northern Sweden

The widest spread in European statistics

Population density across EU regions runs from 7,827 inhabitants per square kilometre to 3.5. That is a ratio of roughly 2,200 to one, and by a wide margin the largest ratio this site has measured on any indicator.

The EU average is 109.7 per square kilometre.

Why this ranking is mostly about administrative boundaries

Brussels tops the list at 7,827, followed by Ciudad de Melilla at 6,180, Vienna at 5,094, Berlin at 4,343, Ciudad de Ceuta at 4,169, Budapest at 3,305 and Prague at 2,869.

Each of these is a city that constitutes a statistical region — a NUTS 2 region — on its own. Where a capital is its own region, its density is a city density. Where a capital sits inside a larger region — Madrid, Warsaw, Rome — that region includes surrounding countryside and the figure falls accordingly.

The ranking therefore reflects how each Member State drew its statistical regions at least as much as how its people are distributed. Comparing Brussels with a French or Polish region is comparing a city to a province.

That is a real limit on the indicator, and it is why the interesting comparison is at the other end.

The empty north

Övre Norrland in northern Sweden records 3.5 inhabitants per square kilometre, Guyane 3.6 and Mellersta Norrland 5.3. Pohjois- ja Itä-Suomi in Finland records 6.2.

These are large regions of forest, mountain and, in Guyane's case, rainforest. Övre Norrland alone covers more than a third of Sweden's area and holds one twentieth of its people; Guyane is an eighth of the territory of France and holds one person in every two hundred and thirty.

Unlike the top of the ranking, these figures are not artefacts of boundary drawing. Northern Sweden and Finland are genuinely among the least populated inhabited places in the world outside the Arctic and central Asia, and no plausible way of drawing regions would change that.

What density is supposed to explain, and what can be checked here

Density is usually invoked to explain other things: how much people drive, whether public transport can be run at reasonable cost, how far a hospital is, how crowded housing gets. It is worth separating what this study can support from what it cannot.

It cannot support any of it directly. Density here is measured by region; every indicator it is meant to explain is measured on this site by country. A national figure can neither confirm nor refute a claim about one region inside it.

And the national figures do not line up the way the expectation would suggest. On the share of inland passenger-kilometres travelled by car, Finland is third in the Union at 75.4%, which fits — but Sweden records 67.3%, below the EU figure of 69.2% and seventeenth of twenty-seven, and the Netherlands, the most densely populated large Member State, is second at 77.0%. Severe housing deprivation runs the same way round: of the twenty-six Member States with a 2023 figure, Malta records the lowest rate and the Netherlands the fourth lowest, and those two are the densest in the Union.

None of that refutes the regional intuition, which may well hold within a country. It means the intuition needs regional data to test, and a density ranking is not that test.

Reading the chart

The dots on the figure show rank order rather than proportion. A scale that placed 3.5 and 7,827 proportionally would compress every region except the top three into a single line at the left edge, which would communicate less than the numbers do.