Regional Economics: Some Fairly-Recent Must- and Should-Reads

A very good point here: there is good reason not to take a markup-free model as our benchmark from which we begin our analysis of Trump: Ben Golub: Krugman Thinks Efficiency Loss of a Trade War Is Small: "Krugman thinks efficiency loss of a trade war is small (Harberger triangle size) even though trade is now in intermediates along supply chains...

...This view is, I think, wrong because it ignores the most important thing about supply chains: complexity. In reality, unlike the toy model he sketches, MANY distinct intermediates go into a typical good. And this happens at many steps along a supply chain. Borders are crossed many times! When that happens, tariff-induced distortions get scaled up a LOT. Really a lot: in a beautiful recent working paper, Baqaee and Farhi show that when markups and complex production are combined, real efficiency losses from markups (which tariffs also are) are TWO ORDERS OF MAGNITUDE LARGER than previously thought. The combination of (i) distortions and (ii) complex production invalidates simple intuitions such as those behind @paulkrugman's toy model and makes for a potentially a huge multiplier on distortions.

Baqaee and Farhi's paper is here: https://scholar.harvard.edu/files/farhi/files/micro_distortions_draft_new.pdf P.S. Their analysis is for markups in domestic production, but the intuitions apply directly to tariffs, and they're working on a version that "officially" covers an open economy. For nerd interest, the key thing going on is that with distortions (wedges) there's no macro envelope theorem, which is the key thing behind Krugman's Harberger triangles intuition....


#shouldread

Comments