An alternative take on Trump’s favourite word
Since nothing seems to be off the table when it comes to trade policy, I thought I might put forward a modest two-step trade policy reform to make America great again.
Step one, levy a substantial tax on all exports, from aeroplanes to soyabeans. Such a tax will squeeze America’s most internationally competitive industries. That’s fine. America will never be great again unless the US soyabean industry suffers.
Step two, tax a generous selection of the inputs that corporate America uses to make its products — for example, the sugar that goes into American candy, the aluminium that goes into American drinks cans and, above all, chips and computers. When American businesses are overpaying for the inputs to the products they make, they’ll get tired of winning.
I’m not sure President Trump and his advisers would go for these ideas, although you never know. Trump tells us that he’s a “tariff man”, and has certainly been talking a lot about levying them. (Forgive me for skirting the specifics. Old-fashioned print deadlines mean that in the brief space of time between me writing these words and you reading them, any given tariff could have been introduced or abolished, or both.)
It is true, although far from obvious, that my two proposals — tax US exports and tax the inputs of US business — are actually identical to Trump’s signature policy of a tariff on imported goods.
The fact that a tax on imports is effectively a tax on exports is a famous result in economics. It was formally proved by Abba Lerner in 1936 but it was obvious long before then that there must be an intimate connection between exports and imports.
Back in 1640, Henry Robinson, a merchant who often wrote about economics, explained, “it is worth remembrance that a great part of foreign commodities brought for England are taken in barter of ours, and we should not have vented ours in so great quantity without taking them.”
A customs official named James Deacon Hume made the same point two centuries later, “[I] did not expect to hear it denied that every import must be paid for by an export.”
As trade policy expert Douglas Irwin explained in an elegant essay “Three Simple Principles of Trade Policy” (1996), the fundamental idea behind Lerner’s symmetry theorem is that a tariff on imports (such as computers) and a tax on exports (such as soyabeans) both have the same effect: they make it more attractive to try to produce computers instead of soyabeans. More people and more investment will flow to the US computer industry, and away from the US soyabean industry.
It is not always bad policy to whack your own soyabean industry in the hope of nurturing a laptop sector, but it is worth being clear-eyed about what is going on. While it is natural to frame this as some kind of struggle between foreign computer manufacturers and US computer manufacturers, it is more accurate to frame it as a struggle between US computer manufacturers and US soyabean farmers. Relatively efficient industries like soyabeans or crude oil will always tend to flourish, unless the government intervenes to tax their exports. Or introduce a tariff on imports — it’s all the same.
Imagine that some entrepreneur builds a magnificent laptop factory on the coast near Los Angeles. The factory uses a production process inspired by Dr Seuss: pour soyabeans into a huge hopper in the factory roof and out come laptops.
One day, a member of the FT’s intrepid investigations team infiltrates the factory to discover the truth: the factory isn’t a factory at all! It’s a port. Ships sail off to South Korea and Taiwan laden with soyabeans and come back with laptops on them. And if you want to stem the flow of imports, and favour a real laptop factory? Just tax the exports of soyabeans.
My second policy suggestion was to tax selected inputs of US businesses, making it more expensive to make products in the US. It is easy to see, but also easy to forget, how this policy is a side-effect of tariffs. Last year, half of all US imports were either business supplies or capital goods, and tariffs on such goods simply make US businesses more expensive to run. Tariffs on sugar hurt the US food manufacturing industry. Tariffs on aluminium and steel make life hard for US factories producing planes or cars. Tariffs on computers hurt pretty much everything.
Of course, tariffs on imported consumer goods straightforwardly make life more expensive for American consumers. But Irwin noted that this well-worn point seemed to lack much political salience. It felt superficial to complain about the rising cost of living when jobs were on the line. That was 1996. Maybe things are different now; the US voter may no longer feel so sanguine about a more expensive weekly shop.
Trump’s tariffs may play well politically, and he has decisively shifted the conversation about tariffs in the US. But economically speaking, tariffs are a strange policy lever to pull.
One comfort is that we’ve seen these policy confusions before. The Lerner symmetry theorem notwithstanding, it is perennially hard to grasp that a tax on imports and a tax on exports amount to the same thing. The US Constitution, after all, bans any tax on exports, while the first substantive act of Congress was a tax on imports. Trump is an unprecedented figure in some ways; in others, he’s reading from a garbled script as old as the republic.
Written for and first published in the Financial Times on 21 February 2025.
Loyal readers might enjoy the book that started it all, The Undercover Economist.
I’ve set up a storefront on Bookshop in the United States and the United Kingdom. Links to Bookshop and Amazon may generate referral fees.
