W ith both major-party candidates promising various forms of trade protectionism, it’s worth remembering why the U.S. over time has moved away from tariffs in its economic policy.

A new article from economist Erica York of the Tax Foundation, published as part of the Cato Institute’s Defending Globalization project, illustrates why, even when they might initially seem promising, tariffs always have a catch.

Tariffs are taxes, but the government needs revenue, and maybe tariffs are a good way to raise it. Proponents might say it would be great to fund the government and protect American businesses at the same time. They’ll point to the history of the U.S., where tariffs raised a significant chunk of federal revenue until the 16th Amendment legalized the income tax in 1913.

But a tariff that raises revenue is not going to protect domestic industry. “From the Founding era through the Civil War, tariffs were the main source of revenue for the federal government and thus not so broad and high as to discourage imports altogether,” York writes. If tariffs are a source of revenue, that means the government wants imports to continue to come into the country, so it can keep collecting the tax.

Taxing trade now, in a more globalized economy, does significantly more economic harm to Americans than it did in 1800. And if today’s federal government was the same size as the federal government of 1800 — no entitlement programs, a tiny military, only two federal departments (State and Treasury) — it wouldn’t matter much whether you funded it with tariffs because it would be so economically insignificant.

Revenue/protection is an either/or, not a both/and. A revenue tariff isn’t protective, and a protective tariff won’t reliably raise revenue, because if the protection “works,” imports will stop and revenue will disappear.

No matter what the U.S. does with tariffs, foreign sellers still have agency and can avoid them if they really want to. One example is tariff engineering, the practice of changing how a good is made to secure a lower tariff rate. Another example is rerouting supply chains to avoid taxed countries. Avoidance of U.S. tariffs on China likely runs in excess of $100 billion annually.

York notes that foreign sellers could also reduce their prices in response to tariffs and maintain market access. That would harm their profits and still raise revenue for the U.S., but again, that would mean the tariff isn’t protecting domestic industry, since imports would still be coming at prices Americans want to purchase.

Now let’s consider what happens if a tariff is set at a high enough rate to be protective and foreign sellers don’t get around it or lower their prices in response. York posits a situation where a foreign good costs $100 and a domestic version of the same good costs $115. The government imposes a 25 percent tariff, which would effectively raise the price of the foreign good to $125.

Americans could choose to keep buying the foreign good at the higher price. Then the government would get the extra $25 in revenue — but who pays it? That depends on whether the tax is on a final good or an intermediate good. For a final good, an American consumer pays the tax. For an intermediate good, an American business pays the tax and then can either pass the cost on to consumers through higher prices or absorb the cost through lower profits.

No matter which option is chosen, Americans are harmed, York points out. The consumer has less money to spend elsewhere or save. The business has less money to invest in workers or capital.

But a protective-tariff proponent would say that the whole point of a protective tariff is to stop the imports. Americans wouldn’t pay the tax because they’d switch to the $115 domestic good instead.

That’s still increasing costs for American buyers by $15, even though the government isn’t getting any extra revenue. But a proponent would argue that more money is going to American businesses, which is better for Americans down the line.

York points out that there’s a catch there too: the foreign-exchange market for currency. If fewer Americans buy imports, then fewer dollars go to foreigners. That reduces the supply of dollars on the foreign-exchange markets, which increases the price of the dollar. A more expensive dollar means foreigners can’t afford to buy as many U.S. exports, so U.S. exporters will essentially be the ones paying the tariff. Those businesses will then have less money to invest in American workers and capital.

This is also why it’s not so simple to say that higher tariffs will reduce the trade deficit. They will reduce imports, but they will also reduce exports. And it’s not only because of the currency effects. Tariffs also invite retaliation from other countries, which will shut American businesses out of their markets, thereby reducing exports further.

Empirical research shows that tariffs aren’t cheap from the perspective of jobs either. “American jobs supposedly saved by import protection have come at an extremely high cost to consumers, ranging from an annual average of $256,000 per job in the 1980s to more than $800,000 per job in the 1990s (all in 2017 dollars),” York writes. “More recently, tire tariffs and steel tariffs have both annually cost US consumers more than $900,000 per job.”

Is it smart government policy to force Americans to pay $900,000 per year for a job that’s probably paying, at most, one-tenth that much in wages? Even the low-end estimate of $256,000 per job is a bad deal.

Since tariffs hurt American consumers and businesses, governments that impose them should expect more political dysfunction to result. As York writes, “Granting tariff protection to one industry mushrooms into requests for tariff protection from additional industries, extensions when initial protections expire, and exclusions for specific firms.”

For example, from 2018 to 2020, the Trump administration’s China tariffs spurred 53,000 requests for exemptions from American businesses. Those represented 53,000 cries for help about a policy the American government was imposing, supposedly to help American businesses. Think of how many cries for help there would be if Trump gets his way with a 10 percent tariff on all imports from all countries.

A more protectionist economy is one where it pays to be close to government, not to serve customers or workers. York writes, “More recent empirical research finds, in fact, that US companies facing heightened import competition between 1999 and 2017 responded not by redoubling their commercial efforts (e.g., investing more in research and development) but by substantially increasing their lobbying for government help — a trend concentrated among less‐innovative American firms.”

Tariffs aren’t a political balm or an effective jobs program or a general pro-business policy or a trade balancer or a revenue raiser. No matter which argument proponents use, there’s always a catch.