The Wall Street Journal story on the GDP drop in the first quarter of 2022 echoes at least three false or misleading ideas that have passed into conventional wisdom. Some are offspring of John Maynard Keynes who, although an elitist dilettante, was not unintelligent and was certainly a genius (a rather evil genius, it turned out) in creating the new conventional wisdom under which we labor (See “U.S. GDP Falls 1.4% as Economy Shrinks for First Time Since Early in Pandemic,” April 28.)
First, the story says (reducing the claim to its logical essentials):
The labor market is a key source of economic strength right now … as employers cling to employees amid a shortage of available workers.
The reader is invited to believe that economic strength comes from a shortage (of labor). The solution of the enigma is that there is no shortage of labor but simply an increasing price of labor, that is, increasing wages or benefits. A shortage, as microeconomic theory understands it, would be a situation where nobody can hire more labor even by bidding up its price along with other employers. There is a shortage of labor as much as there is a shortage of diamonds: you can always get some if you are willing to pay the market-clearing price.
The second false, or at least misleading, idea is that
consumer spending [is] the economy’s main driver.
This is misleading because consumers are not the main driver of the economy, they are the only driver. It is only because people want to consume that they are motivated to work and produce. At least, such is the case in a free economy where consumers are sovereign. If businesses can also be called a “driver,” it is only because they try to satisfy consumer demand or other businesses who try to satisfy consumer demand. The idea underlying the quoted claim seems to be that if people want to consume produced goods and services, these things, that is, GDP, will fall like manna from heaven.
The third idea is demonstrably false, has the most complex ramifications, and may be the most difficult to debunk in a few words. The statement is:
The drop in GDP stemmed from a widening trade deficit. Imports to the U.S. surged and exports fell.
The Economist was a bit more prudent by describing the GDP drop as “a reflection of strong imports”— the fuzzy “reflection of” being, in this context, the usual hint at “caused by” without saying it. Similarly, the Financial Times said at first that the contraction was “reflecting growing trade imbalances,” but then lost it and squarely stated that “GDP was pulled lower by a growing trade deficit … as import volumes and prices surged.”
Anybody who has looked at national accounting even just in a good introductory macroeconomic textbook will know that this is demonstrably false from a strictly accounting viewpoint. GDP—gross domestic product—does not include imports by definition. Hence a drop in GDP cannot “stem from” more imports. Redefining GDP so that imports reduced it would require an altogether different national-accounting framework and new underlying theories of how the economic world works.
I have tried to explain before why the idea that imports automatically reduce GDP is false. The reader might want to have a look at my few related posts and articles, perhaps starting with “Imports as a ‘Drag on the Economy,’” October 20, 2020. Scott Wolla, an economist at the St. Louis Fed, has written a good short piece explaining the same thing: “How Do Imports Affect GDP?” Page One Economics, September 2018.
Instead of repeating what I and others have said before, let me try to use an analogy. In business financial accounts, the fundamental accounting identity that asset plus liabilities equals shareholders’ equity implies that, on the flow side, a company’s profits is equal to its (own) revenues minus its (own) expenses. We could redefine profits as its revenues minus its expenses and minus, say, the Vatican’s expenses on candles (even if latter don’t subtract more from a company’s expenses than imports from GDP, but accounting is largely a matter of conventions). This definitional change would upset the whole structure of financial accounts, but it would allow the statement that company X’s profit reduction “stems from” an increase in the Vatican’s candle expenses.
To justify this new accounting, we would need a substantive theory explaining why it is useful to conceive of Vatican candle expenditures as reducing any business concern’s profits. (Perhaps executives’ wrath at a more powerful popery disturbs them mentally into wasting money?) Such a theory may not be more difficult to defend than the idea that producing GDP (and thus income) in order to have something to exchange for cheaper imports (remember that ultimately, behind the veil of money, products exchange against products) reduces GDP.
The relations between accounting and substantive, non-truistic theories is a interesting topic, to which I have not done justice.
READER COMMENTS
James
May 1 2022 at 6:37pm
GDP = C + I + G + X – M
Isn’t this equation used most broadly because it’s easier to measure these 5 components separately than it is to measure GDP directly? And in this accounting, imports “do” subtract from GDP, since GDP is the measure of all goods and services produced domestically. If you hold domestic consumption and investment and government spending and exports constant, imports increasing will reduce GDP, which isn’t an accounting identity but is instead a real world shift where some amounts of increased imports are displacing domestically produced goods.
So in the most recent quarter, GDP fell “due to” imports rising, which reflecting C + I increasing more than G decreased, but that increase coming from a shift in composition towards foreign-produced goods and services and away from domestically produced goods and services.
You can argue that instead some other metric should be targeted*, but if your accounting formalism is GDP, then increasing imports “holding all else constant” does reduce GDP.
*An alternative measure that wouldn’t decrease if imports increased would be Gross Domestic Consumption**, GDC = C + I + G
**Note that “Consumption” in GDC would then have a broader definition than the “C” term in the GDP or GDC equations.
Jon Murphy
May 1 2022 at 7:41pm
Which leads to the question “why even have imports at all?”
The answer is: because C, I, and G all have imported elements! C is both consumption of domestic and consumption of imported. You need to subtract out Imports to prevent double counting.
Daniel Kian Mc Kiernan
May 1 2022 at 9:06pm
That well-known equation seems to but does not indicate that domestic production drops in response to imports.
Imagine that each of us is as productive as possible, period. We might import nothing; we might import an enormous amount; neither changes the fact that each of us is as productive as possible. Likewise if some of us have lower levels of productivity.
Whatever is produced, the resources available for private consumption, private investment, state expenditures, and exports are the sum of domestic production with imports
C + I + X + G = GDP + M
When we try to measure domestic production, we shouldn’t count imports as if domestically produced.
Simply put,
GDP = (C + I + X + G) – M
Algebraically, that is
GDP = C + I + G + (X – M)
but grouping the variables in that way is grossly misleading.
While this equation does not say that imports reduce domestic production, it doesn’t say that imports do not affect domestic production. When we stop abusing an accounting identity, and consider the results of trade based either upon comparative advantage or upon windfalls offered by foreign technocrats and demagogues, then we realize that domestic production should be expected to increase as a result of imports (regardless of whether these exceed exports).
Pierre Lemieux
May 2 2022 at 7:10am
James: As Jon says. You will have a complete explanation in the links I give in my post.
Jon Murphy
May 2 2022 at 8:21am
Note that your mechanism violates your condition. At first you’re saying we hold C, I, G and E constant. But your mechanism requires that they are not constant: they fall (“imports are displacing domestically produced goods”). Thus, the fall in GDP you describe comes from the decline in C, I, G, and/or E, not I.
James
May 2 2022 at 7:15pm
That’s not true, Jon. C is total domestic consumption, of both domestically produced goods & services AND imported goods & services. You can in fact hold C+I+G constant and have M rise and thus GDP fall.
Oversimplified toy economy:
In year N, an economy has a GDP of $1 billion, of which $10 million is domestically produced steel.
At the start of year N+1, all steel customers in the economy sign contracts at the same price to import steel from abroad instead of locally. The steel mill goes out of business. The economy thus replaces that $10 million of steel with $10 million of imported steel and (in this toy example) nothing else changes because government transfers from a sovereign wealth fund allow the laid-off steel workers to have identical consumption patterns while producing nothing for a full calendar year of adjustment.
Thus, C+I+G is the same in year N and year N+1, while GDP is $10 million lower.
In the real world, this only happens when imports are lower price than the domestically produced goods they are replacing, so imports tend to increase aggregate domestic living standards, assuming long-term scarring problems for laid off workers are smaller than the productivity gain from using cheaper imported goods.*
*The China Shock produced large, long-term scarring for the manufacturing workers affected in many localized towns, but overall economists are confident that the China Shock raised average US living standards because of increased productivity gains exceeding the costs of the spike in unemployment even when second-order affects like the exacerbated opioid crisis and reduced human capital formation are taken into account.
Jon Murphy
May 2 2022 at 7:20pm
Mathematically, no. If C is both domestic and imported consumption (which, as you correctly note, it is), then subtracting out M at the end mathematically means M has no effect.
Jon Murphy
May 2 2022 at 7:25pm
Allow me to prove my point mathematically.
GDP = C+I+G+(E-M)
GDP = (C_d+C_m)+(I_d+I_m)+(G_d+G_m)+(E-(C_m+I_m+G_m))*
By cancellation:
GDP = C_d=I_d+G_d+E.
Imports have no effect on GDP, QED.
*_d mean domestically produced. _m means imported)
James
May 2 2022 at 7:34pm
“GDP = C_d[+]I_d+G_d+E”
Yes, exactly.
And C_d, I_d, G_d, and E are all functions of M with complicated relationships.
In the long run rising M tends to increase all four variables for most countries, which is why the world is richer with more trade (trade unlocking comparative advantage).
In the short run, it is completely possible for M to reduce GDP by reducing domestic production, as happened in the toy example I gave.
Your accounting identity is misleading you because it is oversimplified. If imports had exactly and truly zero effect on GDP, then America in 2023 could produce the exact same set of goods that it produced in 2022 without allowing a single atom to cross its borders. We’d somehow manage to produce batteries without lithium from Chile, iPhones without Chinese labor, avocado toast without avocados…
Now do you see your error?
Jon Murphy
May 3 2022 at 4:29am
Not in the accounting, which is the point here. Economically, an increase in imports could decrease domestic production through the substitution effect. Or (what’s more likely and empirically the case), imports increase domestic production through comparative advantage and gains from trade.
But, by definition, imports have no effect on GDP, which is an accounting formula. GDP is not a production function, which is what you are trying to make it.
Daniel Kian Mc Kiernan
May 2 2022 at 11:36pm
A toy model in which the state responds to imports by intervening to idle resources does not show that imports could idle domestic resources and thereby reduce GDP.
The state could respond to an increase in imports by banning beef, interning members of an ethnic minority, or starting a war with Iceland. Imports would not be the cause of these policies.
Pierre Lemieux
May 2 2022 at 2:25pm
James: The crucial point to understand is that, whatever theory of the macroeconomy one espouses (that imports or government expenditure have consequences that reduce GDP or that the Vatican’s expenditures on candle has consequences that decrease GDP, or whatever), if Y=(C’+M)+(A-M), it is incorrect to say that M reduces Y. (Note that C’+M=C in the GDP accounting identity you are mentioning.)
James
May 2 2022 at 7:27pm
“if Y=(C’+M)+(A-M), it is incorrect to say that M reduces Y”
Now you’re getting confused by the accounting identity.
The full logical statement is actually:
If Y=(C’+M)+(A-M),
AND
If M is a fully independent variable whose changes do not affect C’ and A,
THEN
It is incorrect to say that M reduces Y
However, it is often the case that C’ and A are both functions of M (and many other variables), which means that if the equation for Y is represented in terms of M, then an increase in M can cause Y to rise (or fall). This is how non-zero imports can cause an economy to grow — in the long-term a rise in M allows an economy to expand C’ and A through unlocking productivity gains from comparative advantage.
But similarly, in the short run an increase in M can easily cause Y by shrinking C’, which is the scenario I outlined.
Pierre Lemieux
May 2 2022 at 9:19pm
James: It is important not to confuse a national accounting identity with trade theory. Journalists like those of the WSJ, the Economist, and the Financial Times don’t say (although some might believe it), “Since protectionism is good, imports reduce GDP [where I further assume that GDP=”welfare”].” Journalists say “imports reduce GDP because I have once rapidly seen an equation [in fact, a national accounting identity, true by definition] that states Y=C+I+G+X-M.” It is a bit (the analogy is not perfect) like telling a shareholders’ meeting, “Don’t increase your company’s debt because debt reduced shareholders’ equity. The proof is very simple, it is in the fundamental equation [identity] of accounting: Assets – Liabilities = Shareholders’ Equity.” The fundamental equation of accounting (because is is an identity) will remain true whether more leverage does, for other reasons, increase or decrease profits.
Jason
May 2 2022 at 5:44pm
James and others,
I think some of the confusion comes from us forgetting what that infamous equation (Y = C + I + G + X – M) is for. It’s about trying to balance spending with production. An increase in imports can reduce spending on domestically produced goods and services, but imports cannot directly reduce domestic production in the short run. Consider a situation in which all spending on imports is undertaken with funds from the sale of real assets to foreigners, leaving all domestically sourced consumption to be funded by income. An increase in imports would not have any effect on the amount produced domestically, but it would introduce a wedge between domestic spending and domestic production.
In the medium term, a dramatic shift in consumer spending (substituting foreign production for domestic production) would induce a slow down or shuttering of domestic production, but the impact would be indirect. I can see a surge in imports leading to a reduction in GDP at some point in the not too distant future, but I doubt it would happen within a single quarter.
Thomas Lee Hutcheson
May 1 2022 at 10:58pm
Huhhh? People don’t work in part in order to save?
Pierre Lemieux
May 2 2022 at 7:02am
Future consumption.
Mark Brady
May 2 2022 at 2:36pm
Not necessarily.
Pierre Lemieux
May 2 2022 at 2:58pm
Mark, Can you give me an example where somebody saves not to increase his consumption or the consumption of somebody he will give his savings too? Thinking about it, I found a possible one: a Protestant or a Puritan who would save and invest only for religious reasons. But it is because he wants infinite consumption in the after-life.
robc
May 2 2022 at 6:10pm
Even then, eventually someone will consume it.
Daniel Kian Mc Kiernan
May 3 2022 at 11:46pm
Simply distinguishing investment from consumption is perfectly suitable for most of our purposes, but not perfectly correct.
Let us first consider building muscle. It might seem to be an investment for the purpose of later exertion. But it can have other purposes, including the flow of a sense of physical efficacy.
In some families, especially within some cultures, each generation seeks to leave the next more materially well-off, and encourages that next generation to do likewise. Sometimes, inheritance will be directed, on the expectation that one member of the generation would invest too little on behalf on the next. What is implied is a never-ending, ever-growing investment, seemingly never to be consumed.
But the sense of ever-growing efficacy is a flow consumed by each participant. The investment is itself a treasured family heirloom.
Mark Brady
May 2 2022 at 2:48pm
Keynes an elitist. Yes. But where’s your evidence that Keynes was a dilettante?
“In later use generally applied more or less depreciatively to one who interests himself in an art or science merely as a pastime and without serious aim or study (‘a mere dilettante’).” I quote from the entry in the Oxford English Dictionary published online March 2022.
Henri Hein
May 2 2022 at 6:08pm
Another definition of dilettante is “dabbler,” which I think describes Keynes pretty well. To me, Pierre’s pejorative just reminded me that he was not an actual economist.
Daniel Kian Mc Kiernan
May 2 2022 at 11:46pm
Although I’d agree that Keynes were not a good economist, I’d want to know what were meant by “economist” when someone claim that Keynes were not an economist at all. For example, some very fine economists did not receive degrees in economics.
Pierre Lemieux
May 3 2022 at 11:13am
Mark: The term dilettante is probably too strong.
Mark Brady
May 4 2022 at 1:21am
Pierre, what word would you choose to use?
Daniel Kian Mc Kiernan
May 8 2022 at 8:24am
I don’t know what word Pierre would choose, but I would suggest something such as “unconscientious”. Keynes, who treated his guesses as direct apprehensions of objective plausibility, forwent careful study; and he was willing to resort to sophistry to advance what he regarded as best policy (as determined largely by the aforementioned guesses).
Comments are closed.