A strange burst of numerical rhetoric from Ilya Somin at the Volokh Conspiracy (a blog I generally like, in case I seem to be picking on it here:)
It is, I think, still possible to make a left-wing case that, overall, Israeli policies are, say, 10% worse than French policies. Perhaps even 50% worse. I don’t agree with such claims, but they are not wildly implausible. However, it is utterly impossible for a fair-minded observer with typical left-wing values to conclude that Israel is 100 or 1000 times worse than France. Yet the ratio of left-wing criticism of Israel to left-wing criticism of France is far closer to 100-1 or 1000-1 than 1.5-1.
I suppose one could compute something called a “ratio of left-wing criticism” by having a team of undergrads leaf through editorials for a long time. I’m not sure, though, how one is meant to calculate the ratio of policy badness that Somin’s comparison requires, nor is it clear what one would gain from this comparison. Phrasing things numerically, though, gives the reader the false impression that there’s automatically a meaningful quantitative computation to be made. Don’t use numbers when you’re not actually talking about numbers! If you replace “10%” by “a little” and “100” by “a lot,” you start to be aware that the underlying claim “In order to be considered unbiased, one should complain a lot about things that are very bad and comparatively little about things that are not so bad” is, well, pretty false.