With US plans to impose tariffs on another $200bn worth of imports from China, on top of $50bn already imposed or under way, Donald Trump is getting the trade war he has always wanted. He is also threatening further tariffs against Europe — on cars — in addition to the steel and aluminium duties already affecting that trade, as well as against other traditional allies (and only minimally with China, against whom they were ostensibly directed).

It is a commonplace view among economists that protectionism hurts the economy that pursues it. And we have already examined how Trump’s trade war is harming the US in three ways: the direct cost of dearer imports (in particular such important inputs to manufacturing as industrial metals); the possibility of retaliatory tariffs elsewhere; and the disadvantage to US exporters if other countries pursue trade liberalisation between one another.

But as Dani Rodrik wisely reminds us, the dislocations and redistributive effects within a country due to trade policy changes are much larger than the overall effect on the country’s economy as a whole. When an economy lowers trade barriers, some sectors and types of workers lose and others gain in much greater degree than the average effect on typical consumers or producers. The same, of course, must be true when the barriers go up.