Imports as a “Drag on the Economy”
By Pierre Lemieux
A Wall Street Journal story of last week, “The Verdict on Trump’s Economic Stewardship, Before Covid and After,” makes many good points. It also falls into some popular economic errors. Here is an obvious one:
Trade itself turned out to be a drag on the economy. U.S. export growth slowed starting in 2018 as Mr. Trump’s tariff battles ramped up. The U.S. trade deficit, reflecting an excess of imports over exports, grew to $577 billion in 2019 from $481 billion in 2016.
We are told that imports or a trade deficit necessarily constitute “a drag on the economy.” This elementary error stems from the misunderstanding of a national-accounting identity: GDP = C + I + G + X – M. This identity is often misunderstood as meaning that M (imports) constitutes a “drag” on GDP because it subtracts from GDP measured as the sum of personal consumption expenditures (C), gross private investment (I), government expenditures for goods and services (G), and exports (E).
I have blamed the Wall Street Journal and other journalists before for repeating this myth: see my Regulation article, “Are Imports a Drag on the Economy,” Fall 2015. Perhaps one can find a serious economic argument to the effect that imports reduce GDP—although I and most economists since David Hume, Adam Smith, James Mill, or Jean-Baptiste Say don’t think so. But if such a serious argument exists, it is not that the trade deficit (X-M) subtract from GDP in an automatic, accounting, arithmetical manner as some people imagine is shown by the accounting identity above.
The demonstration is simple. National statisticians (the Bureau of Economic Analysis in the United States), in one of their ways of measuring GDP (from the expenditure side), subtract M because it is already included in their measures for C, I, and G. Consumption expenditures, as measured in the national accounts (in the United States as elsewhere) already incorporate imported consumption goods; investment expenditures already incorporate imported capital goods; and government expenditures already incorporate imported goods and services (foreign consultants, for example). Why do statisticians subtract M? Because imports, by definition, are not part of GDP (gross domestic product) and must not be included in any measure of the latter. M cannot reduce the measure of GDP because it is not part of it.
For another statement of my argument, see my “A Glaring Misuse of GDP,” Regulation, Winter 2016-2017. In still another Regulation article (“Peter Navarro’s Conversion,” Fall 2018), I summarize and illustrate the argument:
Imports have to be removed because they are not part of GDP, which is gross domestic production. … Think about the guy on the scales who subtracts 1 lb. to factor in the weight of his shoes; his weight doesn’t change if instead he subtracts 2 pounds because on that day he is wearing heavier shoes. Likewise, American output doesn’t change because more imports are both added and subtracted.
An economist with the Federal Reserve Bank of St. Louis, Scott Wolla also pointed out this misleading error: I reported on, and linked to, Wolla’s article in another Econlog post: “The St. Louis Fed on Imports and GDP,” September 6, 2018.
Many former college students who took a macroeconomic class and glanced at the accounting identity GDP = C + I + G + X – M in a (perhaps not so good) macroeconomic textbook make the same error. The Wall Street Journal should not.