An important reality about international trade is that not all imports are final consumer goods. According to Dartmouth’s great trade economist Doug Irwin ,
Over half of all [American] imports are either intermediate components or raw materials. These imports are sold as inputs to domestic businesses rather than as goods consumed directly by households.
Recognition of this reality is important for at least three related reasons. First, it disabuses people of the false notion that increases in imports per-capita imply that we Americans are short-sightedly sacrificing our and our children’s futures in order to live irresponsibly high on the hog today. Second, it highlights a vital, positive connection between importing and domestic production. Third, it reveals that, while tariffs and other import restrictions are good for some domestic producers, they harm many other domestic producers even if those other domestic producers do not sell in export markets.
But I here wish to make a different point. It’s a point that not only has direct relevance for the analysis of trade, but it also suggests caution when reading scientific objectivity into seemingly objective facts.
The point is that nearly all imports that are not raw materials are appropriately classified as intermediate components. Thus, the percentage of American imports that together comprise the category “intermediate components or raw materials” is far larger than 50 percent. Indeed, it’s likely well over 95 percent. Except for vacation travel abroad and the consumer goods and services that American vacationers purchase abroad, nearly all American imports – even of goods formally classified as consumer goods – are inputs into the production of producers in America.
Consider, for example, a truckload of individually packaged bed linen imported into America from China. And to make matters simpler (although this assumption is unnecessary), assume – likely contrary to fact  – that no American-supplied cotton or other inputs, at any stage of production, went into making these packages of bed linen. Suppose, reasonably, that these packages of bed linen will be offered for sale to final consumers in Wal-Mart stores throughout the United States.
These packages of bed linen are classified as consumer goods. They are, therefore, not among the greater-than-50-percent of American imports that are conventionally classified as intermediate components or raw materials. Yet I submit that this conventional classification is mistaken. These packages of bed linen, when unloaded on an American dock, are not sold directly to final consumers. They are sold to (or have already been purchased by) Wal-Mart. These packages of bed linen are intermediate goods; they are inputs into Wal-Mart’s production process.
Wal-Mart produces retail services. Among the inputs that a producer of retail services must purchase are inventories to be used to stock its stores’ shelves. So when Wal-Mart imports packaged bed linen, it does not buy these goods as consumer goods; it buys them as intermediate goods – goods that are used by Wal-Mart as inputs into producing the final consumer service that we might call “shopping convenience.” Only by supplying this latter service – shopping convenience – does Wal-Mart earn profits. From Wal-Mart’s perspective, imports from China of packaged bed linen are inputs used to produce its own output no less so than are Wal-Mart’s delivery trucks, warehouses, cash registers, advertising, and corporate stationery.
It follows that when, say, textile producers in South Carolina collude with Uncle Sam to inflict a special tax on Americans who buy imported bed linen, that tariff most directly raises the cost to American producers – in this example, Wal-Mart and other consumer-goods retailers – of producing their final output (“shopping convenience”). Such a tariff is no less one that artificially raises the price of an input used by American producers than is a tariff on steel one that artificially raises the price of a input used by other American producers (such as General Motors and Ford).
Of course, Wal-Mart and other retailers who sell packages of Chinese-made bed linen will, as they say, “pass this higher price on” to the final consumers of bed linen. Final consumers – American households – are forced by this tariff to pay unnecessarily high prices for bed linen. (Note that even those Americans who purchase only American-made bed linen are forced by this tariff to pay unnecessarily high prices for bed linen. The reduced supply of Chinese bed linen causes the total supply of bed linen for sale in the U.S. to fall and, thus, causes the prices of all bed linen – imported and domestic – to rise.) But the very same effect occurs with a tariff on goods, such as steel, conventionally classified as inputs rather than as final consumer goods. A U.S. tariff on steel artificially raises the costs to U.S.-based automakers of making automobiles. Fewer such automobiles are made and the price of new automobiles rises as a result. The cost of the steel tariff is “passed on” to automobile consumers.
In summary, with the exception of vacation travel abroad, almost no imports are purchased directly by final consumers. Almost all imports, even ones that are conventionally classified as “final consumer goods,” are inputs into the production processes of domestic producers. Almost all imports, even ones that are conventionally classified as “final consumer goods,” are in fact intermediate goods. And therefore, almost all tariffs, even ones on imports that are conventionally classified as “final consumer goods,” artificially increase the costs of producing domestically and of otherwise operating domestic businesses.