Donald Trump, trade wars, and optimum tariffs
Donald Trump’s intuition that US economic might could be used to extract more gains from trading partners is not necessarily wrong
The 1 March announcement by US President Donald Trump of tariffs of 25% on steel and 10% on aluminium, confirmed with two presidential proclamations on 8 March, has provoked fears of retaliation and the possibility of a trade war.
Meanwhile, there has been much ridicule among the commenting class on Trump’s explanatory tweet on 2 March. Trump had tweeted: “When a country (USA) is losing many billions of dollars on trade with virtually every country it does business with, trade wars are good, and easy to win. Example, when we are down $100 billion with a certain country and they get cute, don’t trade anymore—we win big. It’s easy!”
Critics have correctly identified the mercantilist fallacy behind Trump’s formulation: that bilateral trade is a zero sum game, and victory is defined as having a positive current account balance with all of one’s trading partners. One could also point out that the US current account deficit is a macroeconomic phenomenon, reflecting the paucity of US domestic savings relative to domestic investment, and has nothing intrinsically to do with trade policy or putative unfair trading by US trade partners.
However, critics have missed a kernel of truth in Trump’s argument: that the setting of trade policy has a strategic element, and it is not wrong that the US could potentially be the winner of a trade war, if one were to erupt—although victory is not a certain, nor the only, outcome.
To understand this possibility, let us review what trade economists call the “optimum tariff”. Two-and-a-half centuries ago, classical economists such as John Stuart Mill and Robert Torrens noted that a country enjoying market power in trade could improve its terms of trade, and potentially improve national welfare, by using protective tariffs. The optimum tariff is defined as tariff that maximizes national welfare. For a small economy, it is zero; for a large economy, it is positive.
The analytical argument was codified, using the Marshall-Edgeworth “offer curve” apparatus, as long ago as 1906 by economist C.F. Bickerdike in a classic (but today little read) paper, The Theory Of Incipient Taxes. He wrote there: “That is to say, with taxes not exceeding some definite height, there seems to be a certain theoretical correctness in the methods followed by Protectionists.”
It was recognized early that the optimum tariff argument—which assumed no response from one’s trading partners—might be vulnerable to the possibility of retaliation. Interest in what until then had been seen as a theoretical curiosum was revived during the actual tariff war of the 1930s, which was widely and correctly perceived as having worsened the effects of the Great Depression.
Writing in 1940, economist Nicholas Kaldor revived and restated the Bickerdike argument, and suggested, without conclusively demonstrating it, that at least one country (in a two-country framework) could gain from the optimum tariff war, even in the presence of retaliation by the other. The following year, economist Tibor de Scitovszky argued, to the contrary, that in the presence of retaliation, all parties would necessarily lose as compared to the free trade situation.
Scitovszky’s conclusion seemed to match the experience of the 1930s, but the theoretical reasoning underpinning it was faulty. This was conclusively demonstrated in a classic 1953-54 paper by economist Harry Johnson, who, in a strikingly forward-looking and modern analysis, deployed game theory and numerical methods to demonstrate that one country or the other might gain in a tariff war, or both could lose, but both could not win.
So much for theory. On the empirical side, a 2008 paper by economists Christian Broda, Nuno Limão and David Weinstein demonstrated that in the absence of World Trade Organization (WTO) rules preventing such behaviour, countries would systematically set tariffs higher on inelastically supplied goods relative to those which were more elastically supplied, which is a direct corollary of the optimum tariff theory. In particular, they showed that US trade restrictions not covered by the WTO were significantly higher on goods where the US wielded market power. This, again, corroborates the optimum tariff theory.
Could Trump be smarter than his critics, and is it possible that the US may benefit from his proposed tariffs? While his exposition is misleadingly mercantilist, Trump’s intuition that US economic might could be used to extract more gains from trading partners is not necessarily wrong.
However, while the optimum tariff argument does have a fundamental core of theoretical and empirical validity, even in the presence of a retaliation-induced trade war, it is an altogether different question whether it is the basis of sound public policy. Such a judgement must move beyond the specific question of whether a particular set of tariffs, such as Trump’s proposed steel and aluminium tariffs, might lead to net gain for the US, narrowly construed, to a larger consideration of the detrimental systemic effects of such tariffs on the rules-based global trade regime.
On this broader question, Scitovszky’s conclusion that everyone loses in trade wars, although it rests on shaky theoretical foundations, is probably a wise public policy judgement.
Vivek Dehejia is a Mint columnist and resident senior fellow at the IDFC Institute, Mumbai. Read Vivek’s Mint columns at www.livemint.com/vivekdehejia
Comments are welcome at firstname.lastname@example.org