At least since President Truman asked for a “one-handed economist” – who presumably would be unable to say, “on the other hand” – politicians have demanded the appearance of certainty where certainty cannot exist.

Economists and other academics tend to respond to this demand if they want to be heard in the corridors of power. They do so in a wide variety of ways: at a recent Leverhulme lecture at the Institute for Fiscal Studies in London, the economic statistician Charles Manski laid out a typology of unreasonable “certitudes”.

A memorable example is the “conventional certitude”, in which a spuriously precise number becomes the focus for all debate. In the US, the Congressional Budget Office estimates how much each piece of putative legislation will alter the budget deficit over the following decade. CBO estimates are about a hundred times too precise: they are reported to the nearest billion dollars, when a range of several hundred billion would be more reasonable for major legislation.