Tariffs Can Work—but Not Stupid Ones
Trump’s new steel and aluminum tariffs mistake the means for the ends.
About the author: James Fallows is a staff writer at The Atlantic and has written for the magazine since the late 1970s. He has reported extensively from outside the United States and once worked as President Jimmy Carter's chief speechwriter. He and his wife, Deborah Fallows, are the authors of the 2018 book Our Towns: A 100,000-Mile Journey Into the Heart of America, which was a national best seller and is the basis of a forthcoming HBO documentary.
Tariffs and interventionist trade policy have had their place in the building of the American economy. You can look it up! But to say that they have paid off in particular times and places is different from saying that they’re generally a good idea. Like other inherently damaging steps, from invasive surgery to blowing up decrepit buildings, they are justified only when the alternative is worse.
Tariffs are good or bad, useful or pointless, depending on the balance of disruption they cause versus benefits they bring. By that standard, Donald Trump’s new steel and aluminum tariffs against Mexico, Canada, and the European Union are both pointless and bad.
What’s the modern case for tariffs? For the moment let’s not get bogged down in the “infant industry” argument—the idea that certain businesses or technologies can grow into robust commercial successes, if only they are given help in surviving their early, vulnerable years. (The main objection to such policies is that governments have a bad record of predicting which industries will grow into successes—the so-called picking-winners problem. The main counterargument is that in fact the United States government has quite a good record of picking, investing in, and promoting general technology categories. These range from biotech, fostered by National Institutes of Health research and other public support; to the agricultural innovations that state and federal authorities have invested in since the 1800s; to aviation and aerospace; to the modern infotech, internet, and GPS industry as a whole, crucially spurred by Pentagon contracts and research.) Because no sane person would claim that either steel or aluminum is a vulnerable infant industry, this reasoning can’t be part of the new Trump tariffs.
The more frequent case for tariffs is that it is a tool for getting attention and changing behavior. Country X is doing something you don’t like: Either it’s making it too hard for your own companies to sell to Country X’s customers, or it’s making it too easy for Country X’s own companies to undercut yours in competition around the world. You apply a tariff to offset what you consider an unfair advantage. And these days, you also get ready to convince an adjudicator at the World Trade Organization that you’ve got a legitimate grievance, because Country X is sure to complain about your tariff to the WTO.
Your real goal is not to keep the tariff but instead to change the other country’s behavior—to make its subsidies or unfair practices painful enough that eventually it drops them, and you can get rid of your tariffs as well. Tariffs in this sense are like other resorts-to-force: a means toward altering behavior, never an end in themselves. The best tariffs are the ones you can drop most quickly because they’ve done the job of changing the other side’s mind.
(If you want more backstory on how this logic applied during the 1980s and early ’90s, when Japan’s export-driven mercantilism was in its most powerful phase, please refer to this item from last month. Or to this one from back in that era, which gave illustrations of strategic uses of tariffs to cope with the “Asian developmental model.” Or to this book from 1989, which argued that the right U.S. response to Japanese mercantilism was not become more like Japan but instead to be “more like us,” which was the book’s title—more open, more diverse, more adaptable, and more resolute about combatting class, racial, and ethnic barriers.)
For a tariff to do any good, let alone to seem justified, it must have two elements: a reason for starting, and a reason for ending.
The reason for starting, the cause for applying the tariff in the first place, is whatever the other country has done that you find objectionable. The reason for ending is that the other country has halted or changed that practice. The whole object in applying the tariff is to get back to the position where you can remove it, because the other country has stopped doing what you considered wrong.
By these standards, how do the new steel and aluminum tariffs measure up?
Based on all the evidence now available, they fail on both counts. Badly.
The countries being punished—Canada, Mexico, those of the European Union, and others—aren’t doing anything about steel or aluminum that the United States could plausibly call wrong or unfair. They sell a lot of steel to the U.S. Canada is the largest supplier of imported steel to the U.S.; it’s also by far the biggest customer of U.S.-made steel. (Both patterns are because so many industries in the U.S. and Canada are so tightly integrated, with so many components and subassemblies moving back and forth across the border.) But that’s no more a proof of “unfair” practices than the fact that the U.S. sells a lot of soybeans to Japan and a lot of Boeing airliners to China. I’m quick to disclaim any status as an authority on the metals industry. But from following the trade news over the decades, I’m not aware of serious arguments that the European, Canadian, or Mexican industries have been predatory in their dealings with the U.S. The clinching evidence here comes from the administration’s own statement yesterday announcing the tariffs. It justifies this move not on economic or trading grounds, but for “national security” reasons. It’s worth noting that the exercise of a national-security provision is being applied not to, for instance, Russia, but instead to countries that include America’s two immediate neighbors, its other most important diplomatic and strategic allies, and its most closely integrated economic and trading partners.
So there’s no visible, argued-out, economically plausible cause for this action.
Worse, there’s no objective or goal. Anyone who has heard Trump talk about trade knows that his real grievance is grievance itself. It’s the sense that America has been “cheated” and “screwed” and “played for a sucker,” and that its trade deficit is inherently a sign of unfairness by other countries, starting with China.
As it happens, China supplies hardly any steel or aluminum directly to the U.S. market. Some Chinese steel comes indirectly, mainly through Canada—but steel and aluminum represent such a tiny fraction of China’s economy or exports that these measures barely touch China at all. The test of a tariff’s success is, again, whether it changes the other party’s behavior. But these moves don’t even pretend to put any pressure on China; as for the Europeans and North Americans, it is hard to see what specific change they could make that would address Trump’s underlying grievances. It’s an ultimatum to America’s main allies and economic partners, where the point seems to be the penalty itself, not the changed behavior it’s supposed to induce.
Is the real source of Trump’s anger the size of the U.S. trade deficit? That’s a plausible concern, but the tariffs have no real chance of affecting it. Suppose that the U.S. stopped buying even a single pound of foreign steel—not from Canada, not from Brazil or South Korea (large suppliers), not from Germany or Turkey, not from any place. Apart from what that might do to the U.S. industries that use imported metals, it could theoretically cut the U.S. trade deficit by as much as about $25 billion. That’s about 1/22nd of the total U.S. trade deficit, which is not nothing—but meanwhile, the tax cut that Trump and the Republicans forced through last year will increase the trade deficit by many multiples of that number. (Why? The short explanation is that when the U.S. budget deficit gets bigger, the U.S. trade deficit almost always does too. That’s because foreigners have to lend the U.S. government more dollars to finance the deficit, and they earn those extra dollars by selling more products to U.S. consumers, plus spending fewer dollars on exports from the U.S. That’s oversimplified, but it’s the main idea.)
So if this highly disruptive move, aimed at what should be America’s closest partners, was intended as a salvo against egregious world trading practices, it’s aimed at the wrong targets. If it was imagined as a stand against China—whose practices indeed deserve a revised and tougher stance from the U.S.—it will have laughably little effect, except to estrange what should be America’s natural allies in applying longer-term pressures on China. And as a step against overall trade deficits, it’s laughably ineffective as well.
As explanatory factors, we’re left with pique and petulance. Ignorance as well. Tariffs have their place. But this move is giving them a bad name.