Economic theory predicts that, except in certain edge cases, tariffs will raise the domestic price of imported goods and services in a country.  The way economists present the effects of tariffs to students is generally through a simple supply and demand model (for example, see this discussion of tariffs that is fairly representative of the textbook presentation).  In this model, we posit the direct exchange of goods between buyers and producers (foreign and domestic). 

One of the logical results of this model is that the tariff burden will be shared between buyers and producers, with their respective shares determined by how sensitive each party is to a change in price.  Consequently, if the consumer does not pay all of the tariff, some portion of it will be borne by foreign producers, resulting in a possible net welfare gain if the producer surplus gain plus the government tariff revenue gain from foreign producers is greater than the welfare loss of consumers.

There are a lot of practical problems with this so-called “optimal tariff” model, and consequently many economists reject its usefulness for policy purposes.  I won’t rehash those arguments (interested readers may find a useful summary here).  Rather, what I wish to highlight is that the textbook supply-and-demand model, while exceedingly useful, is limited in crucial respects when discussing the practical effects of some policies.  Not understanding those limitations can lead to incorrect conclusions.

The main shortcoming of the model for the discussion here is that it flattens down the process of trade too much.  The simple supply and demand model posits direct exchange between the consumer (end-user) and the producer.  As a literal translation of the model, that implies each consumer goes to the factory/farm/etc., where the goods are being produced, buys directly from the producer, and transports the goods back themselves.  Reality is not so simple.  Transaction costs arise during the exchange process and various middlemen exist to reduce those costs.  For example, rather than buy my coffee right from the coffee company, I buy it from my grocer, who bought it from a wholesaler, who bought it from an importer, who bought it from the roaster.  Rather than a single exchange between me and the producer, there are four exchanges.  Each producer is a consumer at different stages of the process.  

These middlemen make the conversation about tariffs (indeed most taxes) a bit more complicated.  We need to examine how much prices change at each stage of the exchange (called “pass-through”).  If some actors in the exchange are more sensitive to a change in price, then they will be less likely to see their prices increase.  Those who are more insensitive to a change in price will see their prices rise.  Consequently, we do not want to look at just one set of prices (eg. Consumer Price Index, Producer Price Index, etc.), but the whole schedule of prices in the exchange.

Looking at just one price can lead to faulty conclusions.  For example, let’s say that a 10% tariff is placed on imported widgets.  Widgets are a common item with many substitutes.  Widget end-users have many options and thus are very sensitive to changes in price.  Widget retailers, on the other hand, are very insensitive to price.  Let’s further assume, for the sake of argument, that widget importers are insensitive to price.  In this scenario, the burden of the tariff would be borne by the retailers and importers.  Their margins would shrink.  The consumer would see very little price increase.  If one were to only look at the consumer prices, one would erroneously conclude that the tariffs had no effect on price.  Indeed, one could even conclude that the tariff was being paid by the foreigners!  What they would not see is that the tariff is being paid for by other domestic members of the exchange.  

A recent Wall Street Journal article demonstrates this problem (“American Compares Are Stocking Up t o Get Ahead of Trump’s China Tariffs,” 20 November 2024).  Two American business owners discuss the problem of raising prices to their consumers and how it affects their business:

The Trump/Biden tariffs have been characterized by such pass-through effects.  A 2021 paper by Cavallo et al finds that the tariffs are being borne almost entirely by US firms.  Looking at (inflation-adjusted) retail prices is insufficient to tell us whether or not tariffs are being harmful.  We must look at the whole exchange process.

PS: This just-released paper looks at the longer run effects of the Chinese tariffs.  The authors find “There is no consistent evidence of reshoring but evidence of nearshoring to border nations. Despite the significant reshaping, China remained the top supplier of directly imported goods to the US in 2022.”