Click here to read the original Cautious Optimism Facebook post with comments
6 MIN READ - The Cautious Optimism Correspondent for Economics Affairs and Other Egghead Stuff concludes his miniseries on what goes into GDP.
![]() |
Gross business receipts vs. consumer spending |
We’ve all heard it over and over from politicians and media: “the consumer is two-thirds of the economy,” or sometimes “consumer spending is two-thirds of the economy.” And the statistic is accepted as an article of faith by the business press while used by certain politicians and academics to minimize the importance of the business sector.
Why exactly do they run around saying this? And is it true?
As you might guess, the answer is a little complicated.
GDP MATH
First let’s cover where the “two-thirds” number comes from. You may recall reading in an earlier installment of this GDP series that Gross Domestic Product is calculated as consumption spending + business investment spending + government purchases + net exports, or:
Y = C + I + G + Nx
If we go to the Bureau of Economic Analysis webpage and break down these elements for the full year 2023 we get the following:
(C)onsumption spending = $18.8 trillion
Business (I)nvestment spending = $5.0 trillion
(G)overnment purchases = $4.7 trillion
(Nx)et exports = -$800 billion
GDP = $27.7 trillion
Based on these figures, consumption spending is indeed 67.8% of GDP. And business investment is a surprisingly small 18.1% of GDP.
And there it is (according to politicians and the business press): proof positive that the consumer is “two-thirds of the economy.”
But is it really that simple?
PROBLEMS WITH GDP
A vocal minority of economists argue that official GDP numbers are misleading and understate the size of business investment activity. For one, salaries aren’t counted but instead only spending on capital investment for things like factories, buildings, machines, etc…
Incidentally, excluding salaries is part of the more popular “expenditure approach” to GDP, but the less-used “income approach” to GDP includes employee compensation. However we’re not going in depth into the accounting of “expenditure” versus “income” GDP in this article.
Another controversy that the Economics Correspondent would like to address in more detail is the intermediate “value added” math used in GDP.
That sounds like a mouthful, but in plain English it means this: GDP calculates business investment spending by counting only the total investment that goes into the final stage product. It doesn’t include investment spending on all the intermediate goods that work their way up the supply chain to the finished good.
A simple example might be a car. General Motors might make a steel chassis and produce its own engine and transmission for a car it intends to sell. Additionally GM will have investment costs for factories and machines that help workers produce these components and assemble the car.
GDP includes all these costs in investment spending.
However before GM bought the steel for the chassis, raw iron first had to be extracted from the earth. The raw materials company had business investment expenses of its own.
Guess what, the raw materials company’s investment expense is not included in GDP (will explain why in a moment).
The same is true for the refining company that purified it into productive metal. Its purchase of the ore and all investment expenses for its plant are not included in GDP.
Then the steel mill that combined the iron and coke incurred expense to purchase those precursors plus investment expense for the mill and furnaces. None of this spending is included in GDP either.
In economics terms, only investment spending on the final product is counted, but investment in goods produced during the intermediate stages of production which ultimately find their way into the final product are not counted.
Hence the controversy. How can business investment be only 18% of GDP when only investment spending at the final goods level (General Motors’ inputs, plant and equipment) is counted, but spending for the steel mill, the refiner, the raw ore extractor, and countless other firms that produce the myriad of precursors for wheels, tires, seating, windows, electronics, fluids, glass, etc… is not?
OFFICIALLY THE ANSWER IS…
Well the mainstream economics community has a good answer. They want to avoid double counting.
If the refiner pays $250 for a block of ore, then adds its own investment to turn it into refined iron and sells it to a steel mill for $500 which in turn combines it with other inputs and sells the steel for $750, economists don’t want to add $250 + $500 + $750 = $1,500.
To count $1,500 in business investment expense would count the $250 paid for the block of ore three different times as it's sold once to the refiner, a second time to the steel mill, and a third time to GM.
The same for the refined iron. Economists don’t want to count its $500 twice: once when sold to the steel mill and a second time when the steel is passed on to GM.
On and on, they say, down the production chain… double and triple and quadruple counting could inflate the numbers. By the time the chassis is produced for, say, $1,000 (the Correspondent is guessing, he doesn’t really know how much a chassis costs to produce), the chain of businesses could have “invested in” the $250 block of iron ore three or four times.
In economics terms the logic is this: all the investments by intermediate stage companies will ultimately be reflected in GM’s final product: the car. GM paid, say, $20,000 for all inputs (there’s your single counting for intermediate stage investments), added $10,000 more of its own investment spending and, after a few transfer costs like interest on debt and taxes, sold the final product for a profit at $35,000.
With traditional GDP only the investment inputs added by GM—plus all the inputs from intermediate firms which are reflected solely in the prices GM, and only GM, paid suppliers for those inputs at the last stage—are counted using this math. There is no double counting of investment.
VERDICT?
As mentioned before, the mainstream economics community accepts the “don’t double count” logic pretty much unquestioningly.
The Economics Correspondent agrees with the logic too, but believes that leaves us with a not fully satisfying number that can still be misleading in other ways.
In the affirmative, yes, the investment dollars for a block of iron shouldn’t be multiplied two, three, four, or five times just because it passes through the hands of several firms.
Where the math gets misleading is it completely understates the total amount of business activity and gross investment spending that goes on in the economy. The math looks as if GM is the only company spending any money when in fact there are countless firms “behind” GM that are collectively spending even more and also employing a lot more people.
It also understates the impact on the economy of disruption in the business sector. If investment spending fell by 5% then according to the standard formula just 5% of 18% is a 0.9% fall in GDP: a blip on the radar.
But a 5% reduction in all (gross) business spending is a much larger number. In the real world the scope of business losses, bankruptcies, and idling of workers would be enormous, far outstripping the suggested 0.9% slump (see attached chart during 2008 and 2020 recessions).
Why? Because there are a whole lot more dollars flowing through the business sector overall than the GDP model portrays.
(Also see comparative size of recessions in attached chart from economist Mark Skousen’s column: “Business—Not Consumers—Drives the Economy.”)
And let’s not even get started on how the salaries and compensation of all those employees isn’t even counted. We even noted in an earlier article that government spending on employees and contractors is considered a “purchased service” and counted in GDP, yet all those private sector salaries and other compensation are not. This makes government worker salaries look “productive” for the economy while private sector salaries are mathematically excluded.
What, private sector employee's aren't productive but government workers are?
Imagine how much larger business investment would look as a share of the economy if private payrolls were included.
There is an alternative measure of economic output that attempts to right some of these wrongs called Gross Output or GO, and in 2014 the Bureau of Economic Analysis even began including crude statistics on GO although they’re buried deep in the website and basically require a search to locate. However the BEA is tacitly admitting GO has certain advantages by putting it on their site.
Conclusion: In the first article on GDP the Correspondent said the formula is pretty good, but not perfect. Understating business activity and excluding expense on salaries are the two biggest reasons why.
In the end, traditional GDP is correct not to double, triple, and quadruple count investment expenditures as intermediate goods flow through the various stages of production, but the press should be a lot more careful when concluding from the simplified math that “the consumer is two-thirds of the economy.”
Of course in a world where people only want to have to digest a single headline, good luck explaining the nuances of where GDP math falls short. And don’t hold your breath waiting for the media, in its present wretched form, to grasp the math either let alone not censor it even if they did understand.
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.