The correlation table

Everything against everything

45 borough measurements, any two of them, all thirty-three boroughs at once. Pick an axis and the city rearranges itself.

r = 0.62 · clear positive association across 32 boroughs. Association is not cause.

£307,208£808,228£1.31M£37,341£47,850£58,359Where the salaries areThe price of a borough
Point at a borough to name it.

How to read it

r runs from −1 to 1 and says how tightly two measurements move together across the boroughs: near 1 they rise together, near −1 one rises as the other falls, near 0 they are unrelated. The dashed line appears only when the association is strong enough to be worth drawing.

None of it is cause. Two measurements can track each other because one drives the other, because a third thing drives both — wealth drives a great many of these — or because thirty-three points are simply not very many. Boroughs also differ enormously in size, and the City of London, with a resident population of about eight thousand, will sit alone at an extreme of almost any per-resident measure.

Every borough in full →