Tariffs as Taxes

A tariff, for reasons described in earlier posts, generally makes the country that imposes it poorer. So do other taxes. Does it cost us more to raise a million dollars of revenue from a tariff than from an income tax, a sales tax, or some other alternative? What does that mean, what does a tax cost beyond the money handed over?

Excess Burden: The Cost of a Tax

In the ideal economy¹ something is produced if and only it is worth at least as much to the person who gets it as it costs the producer to produce it. If something that costs ten dollars to produce goes to someone who value it at only nine dollars, that is a net loss of a dollar — someone pays the cost of production, someone gets the value of consumption, and between them they are a dollar worse off than if it had never been produced. If something costs ten dollars to produce and someone who values it at eleven dollars does not get it, that again is a loss of one dollar relative to the ideal economic outcome. In the ideal economy, anything that is produced is produced in the least costly way; if something that could be produced at a cost of ten dollars is instead produced at a cost of eleven dollars, that too is a cost of one dollar.

We cannot get the ideal economic outcome but a market economy comes surprisingly close. If a consumer values something at more than its cost of production a producer can profit by making it and selling it to him. If something is being produced at a cost of eleven dollars that could be made at a cost of ten dollars, someone, the present producer or a new competitor, can profit by switching to the less costly production process, so will.

This is a very brief sketch of an argument that requires a semester or two of price theory or a good book² to fully explain; I have not even told you what cost and value mean to an economist (hint: they are costs and values to human beings).

Suppose oranges cost a dollar each to grow (and harvest, transport and sell). The government imposes a tax of a dollar an orange so oranges sell for two dollars. Everyone who values an orange at two dollars or more gets one, everyone who values it at less than two dollars (for simplicity I assume that nobody wants more than one orange) doesn’t. The tax has made every consumer who would buy an orange at one dollar but not at two worse off by the difference between his value for an orange and the cost of producing one. That is the excess burden of the tax, the difference between the cost to the people the tax affects and the amount collected. I am limiting my explanation to the cost to consumers but a tax also produces similar costs to producers.

The relation between the excess burden of a tax and the revenue it produces depends on how much the tax changes what people do. If everyone who likes oranges values them at more than two dollars, a one dollar tax has no effect on how many oranges are produced and consumed and no excess burden. If, on the other hand, nobody values an orange at more than $1.99, the tax reduces consumption to zero, producing excess burden but no revenue.

Tariffs vs Other Taxes

If we impose a tariff and reduce another tax, keeping total revenue the same, we have traded one excess burden for another; our question is which has a higher excess burden per dollar collected. The answer depends on how each changes what people do. If the tariff is imposed on a good that can only be imported, not produced domestically, and for which demand is perfectly inelastic, meaning that consumers buy the same amount whatever the price, the quantity produced and consumed is the same as without the tariff; there is no excess burden. Similarly, if the alternative tax is on something with an inelastic demand (or supply), it too has no excess burden. If, at the other extreme, a tariff is high enough to entirely prevent imports or a tax high enough to reduce quantity produced and taxed to zero, it produces excess burden but no revenue.

Most, perhaps all, real world cases lie somewhere between those extremes, with the cost per dollar collected depending on the supply and demand elasticities of the goods affected and, in the case of a tariff, to what degree it results in producing something ourselves that we could get at a lower cost by trade.

The political pressure for a tariff usually comes from industries that want protection from foreign competitors. To the extent that the tariff succeeds in protecting them, imports of the goods stop or are greatly reduced so little or no revenue is produced but there is still excess burden, so protective tariffs tend to be inefficient taxes.

That argument does not apply to a tariff designed to produce revenue. In the 19^th century an income tax was unconstitutional — that changed with the passage of the 16^th Amendment in 1913 — and would have been hard to administer in an economy where income went largely to self employed farmers, so tariffs provided the largest share of federal revenues. A well designed revenue tariff would target goods with inelastic demand so as to minimize excess burden, as would other well designed taxes, so there is no obvious reason to expect one to be more efficient than the other.

Trump’s Tariffs

Protective tariffs are usually imposed to protect specific industries in exchange for political support from the protected industry and its workers. Some of Trump’s tariffs, such as those on steel, copper, aluminum and automobiles, fit that pattern, but many are broad taxes on imports from a specific country. They have produced significant revenue³ but they do not look like revenue tariffs since the tariff rate depends not on the goods but on the country they come from.

Some supporters of Trump’s tariffs support them because they do not understand the economics of trade, as discussed in an earlier post, but not all; in another post I described arguments for tariffs that do not depend on economic ignorance. Why does Trump himself support them?

He may believe they are good for the country, was happy with the revenue they produced. But the main purpose for Trump’s tariffs, judging by how has used them, is as a weapon, a way of punishing countries that do not do what he wants them to, which explains why he imposes most of his tariffs on countries rather than on specific goods.

A recent example was the threat to impose tariffs on EU countries that opposed annexation of Greenland by the US but the pattern goes back at least to the beginning of his second term:

On February 1, he declared several “national emergencies” regarding fentanyl trafficking and invoked the IEEPA to impose 25% tariffs on most goods from Mexico and Canada and 10% on goods from China. Tariffs on USMCA-compliant Mexican and Canadian goods were quickly suspended, but the “fentanyl tariff” on Chinese goods was raised to 20% on March 4. On February 13, Trump announced plans to impose “reciprocal tariffs” on all countries with trade barriers against the US in April, prompting a wave of diplomatic outreach. (Wikipedia)

The use of tariffs as a weapon is not entirely novel, although Trump is carrying it further and doing it more openly than past presidents. Tariff negotiations sometimes end with agreements on mutual reduction which can be seen as each country using the threat of maintaining a tariff to persuade the other to lower one. The US routinely uses trade sanctions to punish countries, such as Russia or Iran, which are doing things we disapprove of.

The answer to the question this started with is that the tariffs Trump imposed are probably worse, in terms of excess burden, than alternative sources of revenue. Whether you view them as better or worse judged by Trump’s chief reason for imposing them depends on whether you think giving Trump the power to bully other countries is good or bad. The same is true of whether you view the recent Supreme Court decision as good or bad in terms of its short-run effects, since it reduces Trump’s ability to use tariffs as weapons. He can still do it, but with limits that did not apply to the legal justification that the court has just ruled against.

---

My web page, with the full text of multiple books and articles and much else

Past posts, sorted by topic

A search bar for past posts and much of my other writing

1. Ideal here is in terms of economic efficiency, discussed in an earlier post.

2. I have written two, Price Theory: An Intermediate Text, a textbook, and Hidden Order: The Economics of Everyday Life, written for people who want to teach themselves economics. Both can be read for free — the cost to me of your reading them.

3. Although the recent Supreme Court ruling against most of Trump’s tariffs may require the government to give most of the money back.