At its core, economics is about making choices. We face trade-offs. If you want more of this, you must give up some of that.
Non-economists often ignore the trade-offs. For example, protectionists argue tariffs will help American industries compete with their foreign rivals while raising a lot of revenue for the American people.
Not so fast, say the economists. If tariff rates are low, most people will keep on importing. The government will collect tariff revenues on those imports, but the policy will not do much to protect American industries. If tariff rates are high, most people will stop importing. This may help those American industries that would otherwise face foreign competition, but it will not result in much revenue since so little gets imported.
Tariff Revenues
In a 2019 Journal of Economic Perspectives article, Mary Amiti, Stephen J. Redding, and David E. Weinstein considered the initial effects of the tariffs imposed by the Trump administration in 2018. The six waves of tariffs increased the average tariff rate by around 1.7 percentage points and reduced imports by 1.3 to 5.9 percent.
The monthly and cumulative tariff revenues estimated by Amiti, Redding, and Weinstein are presented in Figure 1. The authors estimate the additional revenue raised by the newly-imposed tariffs at around $15.6 billion in 2018. Furthermore, they find that “the US import tariffs were almost completely passed through into US domestic prices in 2018, so that the entire incidence of the tariffs fell on domestic consumers and importers up to now, with no impact so far on the prices received by foreign exporters.” The 2018 tariffs raised some revenue for the American people, but the revenue raised came almost entirely (and perhaps entirely) from the American people.
Of course, none of the tariffs imposed in 2018 were in place for the full year. Indeed, much of the increase in tariff rates occurred in the back half of the year. Since all of the tariffs were in effect by December 2018, we can multiply the December 2018 tariff revenues estimated by Amiti, Redding, and Weinstein ($3.2 billion) by twelve to get a rough estimate of how much these tariffs might be expected to raise per year going forward. Assuming no additional efforts to reduce one’s exposure to tariffs occur in subsequent years, the 2018 tariffs can be expected to raise around $38.4 billion per year — or, $46.8 billion per year in today’s dollars. For comparison, the federal government spent around $6,900 billion in 2024.
The estimated revenue raised by the 2018 tariffs is relatively small at around 0.7 percent of federal spending. Moreover, the revenue raised is largely (and perhaps entirely) paid by Americans. Higher tariff rates have a direct effect of raising tariff revenue. But higher tariff rates also discourage imports, which reduces tariff revenue. At some point, the latter effect dominates: higher tariff rates reduce tariff revenue.
Deadweight Loss of Tariffs
Economists are keen to cite another tariff trade-off, as well. Suppose the objective is to protect American industries. The higher the tariff, the bigger the disincentive to import. However, a higher tariff also raises the price prevailing on the domestic market — and the higher price will discourage some transactions from taking place. Economists use the term deadweight loss to denote the lost gains from trade that result when tariffs push up prices. You can increase protection for American industries, but only if you are willing to accept a bigger deadweight loss.
Amiti, Redding, and Weinstein also estimate the deadweight loss of the 2018 tariffs. Their monthly and cumulative estimates are presented in Figure 2. In total, they find that the six tariff waves reduced the gains from trade Americans realized by around $8.2 billion. As with revenues, we can get a rough estimate of the annual deadweight loss these tariffs might be expected to generate going forward by multiplying the December 2018 deadweight loss estimated by Amiti, Redding, and Weinstein ($1.4 billion) by twelve. Hence, the 2018 tariffs can be expected to reduce gains from trade by around $16.8 billion per year — or, $20.5 billion per year in today’s dollars.
The deadweight losses associated with the 2018 tariffs were relatively small: each US household loses roughly $156 per year. Higher tariffs would offer more protection for American industries but at a higher cost to Americans.
Tariffs and Income Taxes
Interestingly, there is one tariff trade-off many economists seem to overlook. Suppose the objective is to raise a given amount of revenue. You could impose a tariff. Or, you could impose a tax on income. The less you rely on tariffs, the more revenue you will need to raise from other taxes and fees, particularly income taxes, which account for the bulk of federal revenues. That’s how trade-offs work.
The United States spends far too much today to rely exclusively on tariffs, of course. A 100 percent tariff on existing imports — that is, implausibly assuming no one was dissuaded from trading by the sky-high tariff rate — would generate just $4,110 billion (and much, much less under more realistic assumptions). Recall that the federal budget was around $6,900 billion in 2024. Still, there is a trade-off at the margin. We could rely a little more on tariffs and a little less on income taxes, or a little less on tariffs and a little more on income taxes.
Economists who oppose tariffs on the grounds that they generate a deadweight loss are ignoring an important trade-off. Income taxes also generate a deadweight loss. The relevant question is whether the marginal deadweight loss associated with the tariff is greater than the marginal deadweight loss associated with the income tax. It is at least conceivable that, given the relatively low tariff rate and the relatively high marginal income tax rates, the deadweight loss caused by a marginally higher tariff rate would be more than offset by the gains from trade caused by a marginally lower income tax rate.
The estimates from Amiti, Redding, and Weinstein imply that the 2018 tariffs generated around 44 cents in deadweight loss for every dollar raised, in addition to the dollar transferred (almost entirely or entirely) from Americans to their government. For comparison, Martin Feldstein estimated that a one-percent increase in all marginal income tax rates (e.g., from 15 percent to 15.15 percent, 25 percent rate to 25.25, and so on) would have increased the deadweight loss of taxation in 2001 by around 76 cents per dollar raised.
Before concluding that tariffs are a more efficient revenue-raising device on the margin, at least two caveats are in order. First, the marginal deadweight loss of tariffs and income taxes rise with the corresponding rates. That implies that the marginal deadweight loss of additional tariffs would exceed those estimated for the 2018 tariffs. Likewise, the estimates of the marginal deadweight loss from Feldstein should be updated to reflect potential changes from the status quo (e.g., expiration of the Tax Cut and Jobs Act) rather than an across-the-board increase from the 2001 income tax rate schedule.
Second, the marginal deadweight loss of tariffs estimated above does not include any costs associated with retaliatory tariffs levied by other countries. Amiti, Redding, and Weinstein find complete pass-through of foreign tariffs as well, indicating that retaliatory tariffs were similarly paid by those in the country imposing them. They do not estimate the deadweight loss of retaliatory tariffs in 2018, but note that “foreign retaliatory tariffs were also costing US exporters approximately $2.4 billion per month in lost exports” by the end of 2018. Since the corresponding deadweight loss would subtract the opportunity cost of lost exports from the value of lost exports, $2.4 billion per month can be thought of as an upper-bound estimate. Hence, the deadweight loss of retaliatory tariffs realized by Americans could be substantial — particularly in cases where the US imposes higher tariffs on many countries. If the opportunity cost was less than 57 percent of the value of imports, the deadweight loss of the 2018 tariffs per dollar raised exceeded the marginal deadweight loss of the income tax as estimated by Feldstein.
The aforementioned estimates should not be mistaken for rigorous policy analysis. They are rough-and-ready back-of-the-envelope calculations. That they cast doubt on the conventional view among economists should give one pause, though — and prompt economists working in public finance to take a closer look.
Economists usually have a keen eye for tradeoffs. They understand you cannot have your cake and eat it, too. It is surprising, therefore, that they have largely missed the tradeoff between tariffs and income taxes. It is not enough to say that tariffs are bad. One must also show that tariffs are worse than the available alternatives.
US government debt is growing faster than the economy. That is not sustainable. The government must get its budget deficit under control, but the political will to reduce spending is limited. That means the government will need to raise additional revenue. How should it go about doing that? Some want to let the Tax Cut and Jobs Act expire. Others want higher tariff rates. In order to decide which approach is best, we must consider the tradeoffs.
