@akhilrao at the end of the partial equlibrium piece i self-link, Nick Rowe wisely argues that good policy is about setting things up so that partial eqm reasoning composes to a good general eqm. 'cuz partial eqm reasoning is, for the most part, what people are gonna do. general eqm reasoning is hard in the best cases, and under ordinary uncertainty about magnitudes of countervailing effects, often the best you can say is you can't predict the final result, so how are you supposed to choose?