Economists usually inform policymakers with conclusions that come from studying the conditional expectation, i.e. arithmetic mean, of some potential outcome. But there are other means to study, from the same ‘quasilinear' family. And they can support very different conclusions. In trade...