On Tariffs, and Large Country Assumptions

Suppose the US puts a 10% tariff on imports from a foreign country. Will import prices (inclusive of tariffs) rise 10%? It depends on the elasticity of supply of said imports. If the elasticity of supply is less than perfect, then import prices will rise less than 10%. To see this, consider the most basic tariff graph in the known universe (from Feenstra-Taylor) – if you can’t understand it, abandon all hope for comprehension of tariff policy.

In the left hand side panel, we have the home domestic market, in the right hand side the international market. The demand for imports is derived from the gap between the home domestic demand curve and the home domestic supply curve. The supply of imports is derived from the corresponding graph for the foreign country, where the X* curve is the gap between foreign domestic supply and foreign domestic demand. The price at home and in the international market is the same under free trade, PW.

Now suppose home government applies a tariff, t. If the supply of foreign exports (imports into the home market) is less than perfectly elastically supplied, then the new price of imports in the home market rise to P*+t. Because the amount of imports is now lower, the quantity supplied from foreign must be lower, which occurs with lower price P*. The price increase (inclusive of tariffs) rises by less that the amount of the tariff, t. A terms of trade effect from the tariff. pushes the foreign country to sell to home at a lower price. In other words, whether home or foreign absorbs the tariffs is an empirical question (this responds to Mr. Bruce Hall’s point).

In this post examining the impact of iron & steel tariffs, I show that a small gap develops post-tariffs between import prices (ex.-tariffs) and the PPI for iron and steel, and a much bigger in early 2021 (the relevant figure is reprised below as Figure 2).

Figure 2: PPI for iron and steel (blue), and import price for iron and steel (brown), both in logs 2018M03=0. NBER defined recession dates shaded gray. Orange denotes imposition of Section 232 tariffs. Source: BLS via FRED, NBER, and author’s calculations.

In principle, in the simplest exposition (as in the textbook graph in Figure 1 above), the gap should be relatively constant (the composition/weighting scheme for the import price and the PPI are not the same, so that complicates comparison). However, if the slope of the demand and supply curves change depending on macroeconomic conditions (and remember early 2021 the global economy is roaring out of the Covid slowdown), and where one is on the curves, then the gap between the two measures could widen.

I mentioned whether the US or foreigners bore the brunt of US tariffs was an empirical question. In Amiti et al. (2020), it was determined that US consumers (broadly defined including firms) bore almost the entirety of tariffs — except (interestingly) for steel.

My apologies for assuming people understood the underpinnings of tariff analysis in my previous post. I hope that people now understand the import prices (inclusive of tariffs) can move for other reasons than just the level of tariff rates. For more textbook exposition, see my slides on tariffs/quotas under perfect competition, and under imperfect competition.