If one does not understand what GDP is and how it is measured, interpretation errors are not surprising–especially when reinforced by ideological wishful thinking.
In a few pieces and posts, I have tried to bring attention to the mistaken treatment of imports as a “subtraction from GDP,” an error committed by many commentators, journalists, and even economists. It was thus encouraging when Scott Wolla, an economist at the Federal Reserve Bank of St. Louis, published a correct explanation in a Bank’s educational publication (see my post “The St. Louis Fed on Imports and GDP,” September 6, 2018).
Wolla strikes again with an article just published in the Journal of Economics Teaching, “The Textbook Treatment of Net Exports: Will the Uninformed Reader Understand?” Summarizing what’s important to understand in the accounting identity used to describe the expenditure side of GDP, he writes:
In short, correctly calculated, imports don’t count negatively in GDP; rather, they have no impact on GDP.
The basic reason, of course, is that imports are not part of GDP (gross domestic product) and, thus, cannot decrease its accounting value. Wolla explains that many introductory economics textbooks provide misleading information (note that Nx represents “net exports”):
The textbook treatment of net exports in the expenditures approach varies. Many textbook authors capture net exports in a single variable (GDP = C + I + G – Nx), while others have broken it into its component parts [GDP = C + I + G + (X – M)] with exports indicated by “X” and imports indicated by “M.” This extra step brings more attention to the individual role that each variable plays in the expenditures approach and is likely to help students differentiate between the two variables. Put differently, using the (X – M) approach brings direct attention to the fact that the two variables behave differently in the expenditure formula. Exports (X) are an addition to GDP. Imports (M) are subtracted from GDP; or, more correctly, they act as a corrective accounting measure used to offset an earlier addition. Textbook authors and instructors need to make this distinction clear to students. [My underlines.]
Wolla points out how mistaken are Peter Navarro (a trade advisor to President Trump) and Commerce Secretary Wilbur Ross, who stated in a policy paper:
The growth in any nation’s gross domestic product (GDP)—and therefore its ability to create jobs and generate additional income and tax revenues—is driven by four factors: consumption growth, the growth in government spending, investment growth, and net exports. When net exports are negative, that is, when a country runs a trade deficit by importing more than it exports, this subtracts from growth.
This statement reveals a faulty understanding of national accounting. As Wolla repeats,
the –M component is included as an accounting mechanism to ensure that the value of imported goods already included as personal consumption expenditures (C), gross private investment (I), or government purchases (G) is subtracted out.
One surprise perhaps is that half of the 20 introductory economic textbooks reviewed by Wolla “did not provide enough information for students to draw a correct conclusion about the contribution of imports to GDP.” Perhaps Navarro and Ross read one of those and did not go any further?
Wolla’s whole article is worth reading.
READER COMMENTS
Jon Murphy
Feb 23 2019 at 11:43am
I ended up creating my own handout for my classes because the textbooks never go into enough details. I always go through GDP very carefully to prove to students imports have no effect on GDP. It’s an easy mistake to make, but it can have massive consequences
Pierre Lemieux
Feb 23 2019 at 11:58am
Good idea, Jon. One surprise in the Wolla article is how many textbooks are careless (perhaps even ignorant) about this accounting issue.
Benjamin Cole
Feb 23 2019 at 8:06pm
Another method to ponder the issue is to consider gross national income (GNI).
Exports add to GNI. It is correct to posit that imports are something of a non-entity regarding GNI.
However, if imports supplant domestically improved product, then imports can be considered to be reducing GNI.
Generally, it is considered good manners not to characterize those with different points of view as ignorant or uninformed.
It just may be that highly intelligent people, well-informed, have different points of view.
Jon Murphy
Feb 24 2019 at 6:37am
No. Even in that case, imports would not generally reduce GNI or GDP. Resources get moved around to more productive uses and more is produced. The doctor who pays someone to mow his lawn or buys food from a food truck can spend more of his time being a doctor.
Gains from trade.
Pierre Lemieux
Feb 24 2019 at 5:05pm
You are right in the sense that one can entertain an economic theory whereby imports decrease GDP (or GNI); I don’t think that there is such a valid theory, but it is not logically impossible. But one cannot base such a theory of a faulty reading of an accounting identity. In a similar way, one can have a theory to explain why Joe lost weight (his wife does not cook for him anymore or he has cancer), but one cannot deduce it from the fact that Joe’s (net) weight has decreased as he is wearing heavier shoes. (See https://www.econlib.org/logic-theory-and-accounting/.) And we may suspect that somebody who believes in that sort of theory (based on misreading an accounting identity–either Y=C+I+G+X-M, or Net weight=gross weight – shoes) might be wrong in other theoretical results.
Benjamin Cole
Feb 24 2019 at 7:46pm
Okay, let us assume new imports supplant domestic production.
Money that used to flow through US factories and thus become income for workers and US businesses, instead flows overseas. There is a reduction in gross national income.
It is true, axiomatically, that exported money returns to the US in the form of investments in stocks bonds and real estate, and to a very minor degree, investments in new productive capacity. But it no longer returns as income, but rather as a claim on assets.
Thus, when imports supplant domestic production, there is a reduction in gross national income.
I will concede the issue is further complicated by actions taken by the Federal Reserve. It might be the Federal Reserve can offset losses to income by expanding the money supply, but that is a whole ‘nother topic.
Pierre Lemieux
Feb 24 2019 at 8:50pm
There is no way you can derive this from GDP=C+I+G+X-M because neither side of the equation includes M.
Benjamin Cole
Feb 25 2019 at 7:14am
I think you are right about GDP= C+I+G (X-M), It should read GDP=C+I+G+X.
And yet when imports supplant domestic production, we do see a decrease in GDP.
How would you devise a simple formula that would accurately capture the decrease in GDP (preferably GNI) that accompanies a supplanting of domestic production by imports?
Remember, I am not asking for you to endorse tariffs, etc. I admire your libertarianism (though I wish you discussed property zoning and criminalization of push-cart vending a lot more).
I am just asking for an an accurate portrayal of what happens to GNI when imports supplant domestic production, holding everything else constant.
Pierre Lemieux
Feb 25 2019 at 11:14am
Two brief and different points, Benjamin.
(1) GDP=C+I+G+X: This is the theoretical formula, when C, I, and G do not include any imports.
(2) To quickly answer your last question, assume first that there is full employment in your country (given technology and institutions). If you want to buy a domestic car, you have to bid it away from another consumer. Alternatively, you can reduce your purchase of other domestic goods and import a BMW (moving your country along its production possibility frontier). Second case: assume unemployment of resources. Because you really want to import this BMW, you work more and produce (and earn) $60,000 more, which you pay to the German producer for the BMW. In none of these cases has GDP decreased. These cases don’t incorporate all the possible cases, but they are representative of the main argument of why imports do not reduce GDP (or GNI).
Benjamin Cole
Feb 27 2019 at 9:13am
“assume first that there is full employment in your country” –PL
Oh, is that all?
When. oh when, in the last 50 years have domestic auto factories been overloaded with demand, that they could not produce more automobiles?
You are aware that US total capacity utilization rates have been falling since the 1970s?
And until very recently, lots of unemployment was the norm in the US, even in recoveries? The Fed, until very recently, declared that the US needed a minimum of one out of every 20 people who wanted work to be unemployed. The Phillips Curve etc. The Fed more than met its mark (5% unemployment).
Even now, we may be finding out that 4% unemployment is not “low” but perhaps a new normal. People are coming out of the woodwork to work. Excuse the clumsy expression.
Egads, “assume full employment.”
Okay, I assume not full employment, and a present capacity utilization rate at under 80% (see FRED data).
So, yes imports reduce GNI, except in conditions that never exist—–ie, full employment.
Gordon
Feb 24 2019 at 5:25pm
I’m curious about this quote you had from Wolla:
“the –M component is included as an accounting mechanism to ensure that the value of imported goods already included as personal consumption expenditures (C), gross private investment (I), or government purchases (G) is subtracted out.”
Isn’t the -M component also adjusting for exports that have imports as part of their intermediate inputs?
Pierre Lemieux
Feb 24 2019 at 8:52pm
M includes all imports, including intermediate products that enter into X. (Wolla explains this in a section of his article.)
Comments are closed.