What populists don't understand about tariffs (but economists do)

Oren Cass’s shallow and selective defense of Donald Trump’s trade policy proposals in The Atlantic misrepresents what economists know about tariffs. Agreed, Trump’s specific ideas—including minimum duties on all imports and the use of tariffs for commercial and foreign policy objectives—are unprecedented. Still, under Trump’s plan, the well-documented costs of tariffs will swamp the far more speculative benefits that Cass touts.

Cass’s main critique is that economists “consider only the costs of tariffs and not the benefits.” Cass imagines potential benefits of tariffs that he calls “externalities.” In his telling, these include fueling a US manufacturing renaissance, leading to higher middle-class wages and more stable families and communities. He asserts that this manufacturing resurgence will promote research, development, and innovation, leading to economy-wide productivity gains.

While such goals are laudable, an assessment of tariffs as a policy tool must answer three questions:

• How costly are tariffs?
• Would they deliver the desired benefits?
• Would other policy tools be more effective than tariffs?

Our brief answers are:

• Very.
• Negligibly or no.
• Yes.

By hawking the illusion that Trump can ameliorate America’s serious social problems with a wave of the magic wand of tariffs, Cass discourages the tough work that economists and policymakers need to do to craft and implement practical solutions.

Let’s take each of the three preceding questions in turn.

Costs of tariffs. Tariffs are a tax on imports, and they will raise prices for households and, crucially, for businesses that rely on imported inputs to make their products. Not only will prices rise for the imported products, so will the prices of goods produced at home that compete with imports. Simply put, protectionism reduces the gains from trade; we choose to pay more than necessary for some goods (imports and their domestic substitutes) instead of focusing on those goods that we produce more efficiently than foreigners.

Skeptics argue that inflationary effects of Trump’s 2018 tariff hikes were not visible in the data. However, the tariffs Trump now proposes will apply to more than eight times more imports than his last round (about $3.1 trillion based on 2023 data), giving them a much bigger impact on prices while reducing opportunities for switching toward non-tariffed substitutes. Moreover, the US inflation rate did rise in 2018, leading the US Federal Reserve to raise interest rates. Careful product-level studies of the Trump tariffs—such as those here, here, and here—indicate that not only did the prices of tariffed goods rise, they rose by the full amount of the tariffs. In other words, American households and businesses bore the entire burden; none was shifted to foreign exporters.

Benefits of tariffs. Tariffs also have a poor record at creating the manufacturing renaissance that Cass yearns for. On the contrary, Trump’s 2018–19 tariffs harmed manufacturing employment. US exporters facing retaliation lost sales abroad, and some felt compelled to lower their prices as a result. Tariffs on intermediate goods also harmed US export competitiveness.

These negative effects would be even larger under Trump’s proposed universal tariffs, which would hit a wider range of production inputs. Since many of the manufacturing industries Trump hopes to resurrect produce precisely such inputs, it would be counter to his policy’s aims to shield US businesses by exempting intermediate imports from tariffs. So Cass’s big, unrecognized, benefit from Trump’s tariffs—their supposed boost to manufacturing employment—would be unlikely to materialize.

Economists have long known that tariffs on imports not only reduce the demand for imports, they also discourage exports. This effect arises because as more domestic resources are used to produce goods that were previously imported, those resources are drawn away from export industries. There is only so much that the US economy can produce with its current labor force and capital stock; moreover, shrinking the labor force, as Trump aims to do by deporting unauthorized immigrants, would only exacerbate that constraint. If US production of import goods rises but production of export goods declines, there is no guarantee that overall manufacturing employment will expand. This process also works in reverse when trade expands. One study found that higher US imports over 1995–2011 were associated with the loss of 1.4 million manufacturing jobs, but higher US export sales generated 2 million more manufacturing jobs.

Cass believes that the experiences of developing Asian nations and pre-World War I America demonstrate the value of tariffs in spurring manufacturing. However, these economies were all primarily agrarian, with large supplies of rural labor available to move into manufacturing, hardly a model for the United States today. The value of tariffs in these industrializations as opposed to other factors such as export promotion is unclear. Even Alexander Hamilton, the first US Treasury secretary and a strong advocate of policies to promote American manufacturing, preferred relatively low tariffs.

Cass mentions another item on the benefit side of the tariff ledger, revenue. Contrary to his assertion, economists do not assume that “tariff revenue simply disappears.” But he is correct that what happens to that revenue is important. Tariffs are regressive taxes, more costly to lower- and middle-class households. Would a Trump administration use extra tariff revenue to compensate those households? Simply put, no. Instead, Trump has suggested an ever-expanding array of regressive tax cuts, including cuts to corporate tax, extension of the Tax Cuts and Jobs Act income tax cuts, tax exemption of Social Security benefits, and restoration of the full deduction for state and local taxes. While Trump insists there will be enough tariff revenue to do all that and establish a sovereign wealth fund too, plausible estimates of tariff revenue are much lower and leave little scope to compensate poorer households.

Tariffs may also result in political dysfunction that limits the revenue they yield. Not only can they often be levied without congressional scrutiny, but tariffs create a regime that invites lobbying influence. As a consequence, tariff revenues will be dissipated as companies petition for exemptions, and politics will be a major factor in the disposal of the funds. Trump spent $28 billion on aid to farmers during his last trade war to compensate them for Chinese retaliatory actions. Consumers were not so lucky.

Are there better tools? Globalization is not the only factor that has disrupted US labor markets and communities over the past three decades, but economists often understate the difficulties associated with compensating those harmed by disruption.

However, the underlying problem of worker security is much broader than preserving manufacturing jobs, and tariffs are an ineffective and inefficient tool for enhancing this aim. After all, the net decline in manufacturing jobs masks massive job churn in our economy, much of which is domestic in origin. For example, factory closings in Indiana or California alongside openings in Alabama or Georgia will not be addressed by tariffs, nor will the disruptions due to technological change. These disruptions may cause distress, but they will be untouched by tariffs, which will instead penalize consumers and shift activity toward inefficient producers. What is needed instead is a redesign of the US social safety net and a strengthening of US fundamentals, including public investments in infrastructure, research, and people.

If one views more manufacturing jobs as desirable, other tools outperform tariffs. For example, direct subsidies avoid the side effect of higher consumer costs. One could also fix elements of the US tax code that incentivize US companies to produce abroad rather than in America.

Promoting research and innovation does involve a genuine externality: positive productivity spillovers throughout the economy. As a share of GDP, private US research and development (R&D) spending has risen steadily since 2006, but government investment in basic R&D is now below 1 percent, about half the peak it reached in the late 1960s. Direct R&D subsidies and government investment are in order and are much better ways to spur innovation than costly and inefficient trade barriers.

We do not deny that tariffs can be appropriate in some circumstances. For example, countervailing duties in the face of a trade partner’s subsidies, or safeguards to give industries more time to adjust, can be appropriate if done in a manner consistent with World Trade Organization rules.

In contrast, Trump’s proposed comprehensive tariffs, premised on the notion that “the U.S. has been ripped off for years” by the entire world, go far beyond any economically defensible purpose. The siren song of such snake-oil solutions should be rejected; instead, the United States should do the hard work of crafting a fairer and more resilient economy.

NOTE: This essay is a response to "Trump's Most Misunderstood Policy Proposal: Economists Aren't Telling the Whole Truth about Tariffs," published on The Atlantic.com on September 25, 2024. We offered this essay to The Atlantic, which declined to publish it.