Important Economics Lesson; Spurious Correlations Are Spurious

Email Print

There’s an excellent little new humorous website called Spurious Correlations. Well, OK, humorous perhaps only to economics geeks but humorous all the same. And it’s a site that contains a deep and important lesson about economics (and the other social sciences of lesser importance of course). Which is that spurious correlations are indeed spurious.

Another way of putting much the same point is that correlation is not the same as causality and that’s something we get told all the time but unfortunately not all of us take the care to remember it.

For example, it is possible to show a correlation between the consumption of cheese in the US and the number of people who die by becoming tangled in their bedsheets. We could, I suppose, posit some causal connection by insisting that late night cheese consumption increase nightmares but there’s no actual evidence for that old wives’ tale anyway. Or perhaps that cheese consumption is connected to pizza consumption and that the extra anchovy consumption leads to people being strangled by their sheets? That’s not actually any more convincing, is it?

But the correlation is indeed there:

cheese bedsheets

The entire blog is full of these sorts of correlations, all of which can be seen to be entirely spurious. The finest comic use of such correlations to date has been that mainstay of Pastafarianism. That climate change is caused by the absence of pirates. Well, why not? We didn’t used to have climate change and now we do and we did used to have pirates like Blackbeard and Edward Teach (unless they were the same person?) and now we don’t. So the correlation is there. But there’s no obvious or even possible hidden causality there so that correlation is indeed spurious.

Read the rest of the article

Email Print
  • LRC Blog

  • LRC Podcasts