maurice-obstfeld2By Maurice Obstfeld

Versions in: عربي (Arabic), 中文 (Chinese), Français (French), 日本語(Japanese), Русский (Russian), Español (Spanish)

A longstanding challenge for the global economy is the possibility that some countries compete for export markets through artificially low prices. Political leaders and pundits sometimes propose import tariffs to offset the supposed price advantages and exert pressure for policy changes abroad. What proponents often fail to realize is that such tariff policies, while certainly hurting their targets, can also be very costly at home. And surprisingly, the self-inflicted harm can be substantial even when trade partners do not retaliate with tariffs of their own.

When the going gets tough

The World Trade Organization’s Agreement on Subsidies and Countervailing Measures  allows a country unilaterally to apply a countervailing duty on imports of a good whose production has been subsidized. Other policies that do not clearly fall under the strict WTO definition of a subsidy, including currency undervaluation and accompanying macroeconomic distortions, could also have a net effect of pushing export prices down, leading trade partners to cry foul.

Whatever the reason, political discourse perennially focuses on unilateral national action to “get tough” with trade partners whose export prices are perceived to be artificially low. Leading examples are the U.S. Omnibus Trade and Competitiveness Act of 1988—which followed a notable period of dollar strength and rising U.S. deficits—and several subsequent initiatives in the U.S. Congress, including a proposal last year to allow industries to press for countervailing duties against alleged currency manipulators. Of course, such pressures are not limited to the United States; and they are currently taking a toll on global trade, as the forthcoming World Economic Outlook will document. A problem with this approach is that it opens the door to industry lobbying based on criteria likely to be less objective than a measurable financial subsidy. Furthermore, countervailing protection may induce trade partners to impose retaliatory tariffs, possibly setting off mutually destructive trade wars.

There is another big drawback of such tariffs: while they may give some relief to industries and workers that directly compete with the affected imports, they will be broadly contractionary, reducing output, investment, and employment in the whole economy. These negative effects follow even if trade partners do not retaliate, although if they did, the outcome would be even worse.

This prediction may seem surprising: after all, by shifting demand toward domestically-produced goods and raising the prices of competing imports, wouldn’t a tariff both raise output and employment and deliver welcome upward pressure on inflation? That the answer is “no” was pointed out more than a half century ago by Robert Mundell, winner of the 1999 Nobel Memorial Prize in Economic Sciences (and, incidentally, a staff member in the IMF’s Research Department during the early 1960s). Mundell perceived the key reason why a tariff could have these negative aggregate effects: by promising to improve the importer country’s underlying balance of payments position, it causes the domestic currency to strengthen in the foreign exchange market, possibly reducing GDP and employment (and in that case, worsening the trade deficit in the end). The effects will be harsher when the home country’s central bank policy interest rate is at or near zero—a circumstance Mundell did not consider. In that case, the central bank is more constrained in using monetary policy to offset the tariff’s contractionary effect.

The effects of tariffs in two charts

The point is illustrated by the IMF’s Global Integrated Monetary and Fiscal model, which is more complicated than Mundell’s framework and has the advantage of covering multiple regions and including the dynamic effects of policies. To demonstrate the potential effects, we chose a scenario in the charts below that show the response of four U.S. variables after the United States levies a 20 percent tariff on imports from emerging East Asia. The experiment assumes that the Federal Reserve policy interest rate is at zero, while emerging East Asia is not. (That assumption makes no qualitative difference to the results, but it does make the negative GDP impact harsher than it would be if the Fed could cut interest rates to cushion the economy.)

Chart 1 below shows two scenarios, one in which emerging East Asia does not retaliate with its own tariff on U.S. imports (green line in the chart), and one in which it does (red line). Under either scenario, real GDP declines and, as Mundell predicted, the dollar is worth more. Not surprisingly, output in emerging East Asia (not shown) declines. With retaliation, the dollar appreciates by less, but U.S. GDP falls by far more. Also (not shown), real investment falls, due to both the activity decline in the U.S. and higher prices of East Asian intermediate imports used in making investment goods.


One key behind the fall in GDP is that exports actually fall initially by more than imports (see Chart 2). The rise in tariffs on East Asia leads importers to buy from other countries instead, and the dollar’s appreciation reinforces the switch by making those alternative imports cheaper across the board. At the same time, the stronger dollar weighs on all U.S. exports. The main effect of the selective tariff is thus, broadly, to subsidize most imports and tax all exports. That hurts the trade balance, output, and employment. Moreover, in the United States, as in many countries, the export jobs that are lost tend to be relatively high paying.


Who gains from the tariff? Because the United States is a large country, the tariff, if it does not provoke retaliation, raises the prices of U.S. exports relative to its imports, allowing higher real consumption. These consumer benefits, though, are spread widely and possibly thinly and must be weighed against job losses. Producers for the domestic market who compete directly with emerging East Asia also are likely to gain, though all other import-competing industries and exporters lose. Matters are much worse in case of retaliation, because then, everyone loses.

Think twice

Economic policies aimed at attaining an artificial export advantage are a legitimate topic for international consultation and peer pressure. In some cases, unilateral retaliation is sanctioned by WTO rules. But those who promote “getting tough” with foreign trade partners through punitive tariffs should think carefully. It may be emotionally gratifying; it may boost specific industries; the threat may even frighten trade partners into changing their policies; but, ultimately, if carried out, such policies cause wider economic damage at home.