Trade lunacy is back
The Trump campaign proposes to use import tariffs as weapons in a trade war. But this strategy backfired badly in 2018. Why would it work in 2025?
It’s election season in the U.S., marred only by the minor complication of a criminal trial involving one of the candidates. Trump is on the campaign trail, and he is going big on trade policy. Or rather, trade lunacy.
Here’s Gavin Bade at Politico attempting to explain Trump’s trade policy:
Trump is considering a 10 percent universal import tariff, the former administration officials said, and one result of that policy could be to make the dollar weaker relative to other currencies.
This is economic illiteracy of a kind I haven't seen since the heady days of Brexit. Or - more accurately - since the last time one Robert Lightizer was the United States trade representative. He is, once again, setting Trump’s trade agenda. And he appears to be no better informed than he was last time.
Why do I say this policy is economically illiterate? Because universal import tariffs don’t weaken the currency relative to the currencies of trade partners and competitors. They are much more likely to strengthen it.
Why does Trump want to weaken the dollar?
The strength of the US dollar versus other currencies hampers US exports, encourages imports (thus weakening US domestic production), and keeps the US trade deficit wide and its debt high. It also has unfortunate consequences for developing countries that issue debt in dollars or primarily trade in dollars. A strong US dollar causes global trade to shrink and can trigger debt crises in developing countries. So there are good reasons, both domestic and international, to want the US dollar to trade lower than it generally does.
The dollar’s strength reflects not only demand for the currency itself, but also for assets denominated in dollars, particularly US Treasuries. When international demand for a country’s debt is high, its currency exchange rate rises. US Treasuries are the world’s premier savings vehicle. No other country can produce an asset of such quality in the quantities needed by the rest of the world. So demand for US Treasuries is always high compared to the debt of all other countries. Hence the dollar’s persistent strength.
However, far from seeing the strong dollar as a measure of the world’s trust in the US economy, Trump regards it as a sign of international malevolence. In his eyes, countries that maintain trade surpluses are conspiring to keep the US dollar too strong in order to build export economies at the expense of the US. To be fair to Trump, he is not alone in this view: Janet Yellen, the current Treasury Secretary, recently said China “exporting its way to full employment” is unacceptable “to the rest of the world”, by which she obviously means the United States.
One solution to the dollar’s exorbitant strength would be for the US to issue far more debt. The US is the only country in the world that can buy everything it needs in its own currency. It has no risk of FX crisis, and the only serious risk of sovereign default comes from its own politicians. So it could simply swamp the world with US Treasuries. This would drive down the dollar exchange rate, benefiting both US exporters and struggling businesses in developing countries. It would also fuel domestic inflation, which the Fed could be hard-pressed to contain - for economics nerds, this is “unpleasant monetary arithmetic” (UMA).
But it wouldn’t fix the US’s trade deficit.
Since the world trades primarily in US dollars, countries that run trade deficits (and don’t have Federal Reserve swap lines) must borrow US dollars to pay for imports. This exposes them to the risk of FX crisis, which can have catastrophic economic effects. In 1997, many countries in Asia experienced such a crisis. Ever since, those countries have adopted a defensive mercantilist trade strategy, repressing domestic demand and emphasising the export sector so as to earn more dollars than they spend and build up US dollar reserves as a buffer against sudden currency devaluation like that in 1997. They hold these reserves in the form of US dollar-denominated assets, particularly US Treasuries. The IMF has encouraged other countries to adopt a similar strategy. And some countries have mercantilist trade strategies for historical reasons, notably Germany, which has an exorbitant fear of debt default and hyperinflation.
Trade is a zero-sum game: one country’s surplus must be balanced by trade deficits in one or more other countries. The trade surpluses run by the rest of the world are matched by an outsize US trade deficit. We could say the US is the “consumer of last resort” for the world.
While this remains the case, the US will inevitably have a trade deficit, however many US Treasuries it issues. Other countries can eliminate trade deficits by crashing their currencies, though I wouldn’t recommend this as a strategy. But the US does not have that option. Although a somewhat lower dollar exchange rate would benefit everyone, a dollar crash sufficient to close the US trade deficit would cause the mother of all financial crises and a global depression - and the US would inevitably find itself bailing out the rest of the world, simply because so many countries hold legal claims on it through their holdings of US dollar-denominated assets.
Anyway, given the dependence of the financial system on the US dollar and the lack of credible alternatives to US Treasuries, it is not at all clear that the US dollar could even fall low enough to close the trade deficit, however lunatic the economic policies pursued by the US administration.
Why raising import tariffs won’t lower the dollar exchange rate
Lightizer wants to stop the US acting as consumer of last resort for the rest of the world. He thinks higher import tariffs would force other countries to abandon mercantilist trade strategies in favour of balanced trade. This would lower the dollar exchange rate, helping the US’s export sector and encouraging the reshoring of production to the US.
But higher import tariffs won’t lower the exchange rate. They are much more likely to strengthen it. Indeed, they did, the last time Lightizer tried this genius scheme. During the US-China trade war of 2018-19, the Chinese renminbi depreciated by 7% versus the US dollar, reflecting a 4.2% appreciation of the dollar nominal effective exchange rate (NEER) and 3.1% depreciation of the renminbi NEER:
Oliver Jeane and Jeongwon Son of Johns Hopkins University calculate that tariff increases (by both sides) during this period had little impact on the dollar, because they were offset by an export subsidy (see below), but caused a 3.2% depreciation of the renminbi:
Overall, therefore, the tariffs benefited China’s exporters at the expense of the US’s - the polar opposite of what Lightizer intended.
It has long been known that that import tariff increases tend to cause the currency exchange rate to rise. In Lerner’s symmetry theorem, the appreciation of the currency negates the effect of the import tariff on domestic consumers and instead imposes it on exports. If the import tariff is accompanied by an equivalent export subsidy (this combination is known as a “border adjustment tax”), the tariff and subsidy cancel each other out, the exchange rate is unaffected and the economic impact is zero - hence “symmetry theorem”. But if there’s no export subsidy, the import tariff acts like a tax on exports.
This paper from Jesper Lindé and Andrea Pescatori discusses several scenarios in which Lerner symmetry does not hold, of which the only significant ones are complete asset markets and exchange rate fixes. The authors say that complete asset markets might not be as unlikely as they sound, but it is nevertheless far-fetched to assume that sufficient households would change their consumption and saving behaviour to negate the effect of import tariffs. So that leaves us with fixing the exchange rate.
Holding down the exchange rate while imposing import tariffs rebalances the domestic economy from consumption to production, and vice versa for the foreign economy. And it can definitely be done. Central banks can cap exchange rates indefinitely. The Swiss National Bank did so very successfully during the Euro crisis of 2011-15, when inflows of capital from frightened investors fleeing to “safe havens” threatened to send the Swiss franc exchange rate to the moon. How did it do it? It bought unlimited euros on the open market and paid for them with newly created Swiss francs. In other words, it did cross-currency quantitative easing (QE). This was, at the time, the largest QE programme in the world relative to the size of the Swiss economy.
But Lightizer doesn’t want merely to prevent the US dollar exchange rate from rising with respect to the currencies of its main trade partners. He wants to lower it. Import tariffs won’t do that. Export subsidies by themselves could, but if you are happy to interfere with your exchange rate anyway, why bother messing around with tariffs and subsidies? Why not just do a gargantuan cross-currency QE programme, buying up the rest of the world’s currencies and paying for them with newly created dollars? Of course, a large proportion of these dollars would immediately be exchanged for US Treasuries, which are easier to store than cash dollars and have an unlimited Federal guarantee. So to prevent yields falling below zero, the US Treasury would have to issue lots more debt… hey, wasn’t this what I said before?
Import tariffs are inflationary
Why does Lightizer get this so wrong? It’s all about who pays the tariff. Lightizer seems to think that US import tariffs are paid by exporters to the US. But they aren't. They are paid by US consumers. There is no increase in the world price of goods, only in the price paid within the US.
Import tariffs make US consumers poorer. This is because they are inflationary, especially in an import-dependent economy. Suppose that a new Trump administration decided to follow through on its threat to impose a 10% ad valorem tariff on all imports. If importers fully passed on this tariff to consumers, there would be an immediate rise in the prices of imported goods. Because the US is import-dependent, this would probably be sufficient by itself to raise the general price level, though by how much would depend on the degree of import penetration in the economy. But this would not be the end of the matter. Faced with much higher prices for imported goods, consumers would try to substitute home-grown alternatives, rather than cut back consumption. The sudden increase in demand for those home-grown products would raise their prices too. However you look at it, raising import tariffs increases domestic inflation, at least in the short term.
As inflation eroded their purchasing power, customers might cut consumption, reducing demand for both imported and domestic goods and thus bringing down inflation. And as demand for home-grown goods increased, domestic producers might increase production, thus bringing down prices. But alternatively, customers might demand higher wages, and domestic producers faced with rising wage demands might opt to constrain supply and ratchet up prices. An inflation-targeting central bank such as the Fed,would be duty bound to tighten monetary policy to prevent such an inflationary spiral from taking off. In 2018-19 the Fed did exactly this, raising interest rates and rolling off some of its QE asset purchases.
Raising domestic interest rates encourages foreign investors to buy dollar-denominated interest-bearing assets such as US Treasuries. This raises the currency exchange rate - unless the supply of dollar-denominated assets increases to match demand. We’re back to issuing more USTs again.
In the absence of drastic measures such as ending inflation targeting or massively subsidising exports, raising import tariffs won’t lower the dollar exchange rate. But what about that trade deficit? If consumers switch away from imports to domestic products, shouldn’t the trade deficit shrink?
Sadly, no. Upwards pressure on the exchange rate due to a general rise in ad valorem import tariffs raises export prices, hampering exporters and worsening the terms of trade. All else being equal, a general rise in ad valorem import tariffs that is not offset by an equivalent export subsidy will widen the trade deficit, though if demand for imports falls significantly, the effect might be small.
The real problem, though, is that falling exports and imports together add up to a reduction in gross trade. This is likely to mean a fall in GDP and a rise in unemployment.
No country is an island. Not even the US.
Thus far, we’ve assumed that the US’s trade partners make no changes to their trade terms in response to the new tariffs. But this is of course completely unrealistic. They will respond in some way. The question is what they might do.
What Lightizer really wants them to do is agree to weaken their currency exchange rates relative to the dollar. Given how bad for everyone a too-strong dollar is, this isn’t that bad an idea. But threatening to raise import tariffs isn’t a sensible way of persuading them to do this. Trump likes to negotiate with a gun to the head, but he needs to think about which head he is pointing it at.
Some countries might choose not to respond to threats of new import tariffs, on the reasonable grounds that you should never interrupt your enemy when he is making a mistake. True, a strengthening dollar would make financing in dollars riskier and more expensive, but if they had plentiful dollar reserves, that wouldn’t matter - and their mercantilist strategies generally mean they do have such reserves.
But trade tariffs are as much political as financial. When your largest trade partner decides to use import tariffs as a weapon, the population will expect a robust response from their government. Doing nothing would be a “courageous decision” (i.e. career suicide). So even mercantilist governments woluld be likely to respond in some way. And if the new import tariffs were accompanied by export subsidies, as seems likely, then mercantilist governments might feel obliged to respond to prevent their exports being undercut by cheap American substitutes.
Perhaps the most obvious response would be to impose equal and opposite import tariffs. As I’ve explained above, import tariffs have perverse effects. In the trade war of 2018-19 China imposed import tariffs on US goods that appeared equivalent to those the US imposed on Chinese goods, but the renminbi nevertheless depreciated much more than the US dollar did.
Had China imposed equal and opposide export subsidies as well as tariffs, the exchange rate effects might have been eliminated. But why would China want to do this when the renminbi depreciation acted like an export subsidy anyway? The Chinese administration appears to have understood international trade and currency effects considerably better than Robert Lightizer.
However, a tariffs-only reponse is, overall, bad for everyone. Lindé and Pescatori found that it significantly reduces gross trade and GDP. If everyone did this, the effect on the global economy would not be pretty:
When the model is calibrated to reflect trade flows and relative sizes of the US and the rest of the world... a 10% rise in import tariffs in both the US and the rest of the world leads to a 1% fall in world trade and a 1/2% fall in world GDP.
Neither the US nor China would be particularly harmed by this, but fragile developing countries would be.
Another way of defending against US import tariffs and export subsidies would be to raise domestic interest rates to prevent the currency exchange rate falling. But why on earth would a country that was pursuing a mercantilist trade strategy and had built up significant US dollar reserves do this , when sitting tight and allowing its currency to depreciate would help its export sector and widen its trade surplus?
Mercantilist countries are hardly going to respond to US import tariffs,with or without associated export subsidies, by adopting policies that are contrary to their own interests. The countries that would be forced to raise domestic interest rates, impose tariffs and/or subsidise exports to defend against US trade aggression would be import-dependent countries with large trade deficits and high levels of external debt denominated mainly in US dollars. But these are not the countries whose ways Lightizer wants to change.
“Cooperate or I’ll shoot”
The Holy Grail for Lightizer would be a cooperation agreement like the 1985 Plaza Accord, in which governments and central banks agreed to intervene in the markets to devalue the US dollar against the Japanese yen, Deutschmark, British pound and French franc.
The US dollar soared in the early 1980s because of President Reagan’s expansive fiscal policy and Fed chief Volcker’s very high interest rates. The US’s trade deficit had reached 3.5% of GDP, and the US was struggling with a stubborn recession and high unemployment. The Reagan administration was ideologically opposed to both protectionism and beggar-my-neighbour currency devaluation. So when Congress, in response to heavy lobbying from US business interests, threatened to push through protectionist measures, the Reagan administration and its trade partners agreed to lower the US dollar.
Perhaps Lightizer remembers Congress’s threat to introduce protectionist measures, and thinks a new Trump administration could use the same threat again to force the US’s trading partners to intervene to lower the US dollar. But the Reagan administration did not demand cooperation under threat of execution, as Lightizer is doing. Rather, it asked its trade partners to help it defang Congress, and its trade partners were willing to do so. The US’s trade partners might not have been nearly so cooperative if the administration itself had threatened to introduce protectionist measures.
Lightizer’s belligerent approach to trade negotiations didn’t work in 2018-19. Doing the same thing again in the hopes that it will succeed this time is insanity. This is not trade strategy, it is trade lunacy.
Related reading:
Currency manipulation is a really bad idea
The high price of dollar safety
President Trump’s Tariffs Will Hurt America More Than China - Forbes