Thanks for posting Noah. I've been bemused by the plethora of articles in the last few days from the WSJ/Bloomberg/Reuters/PickYourOutlet talking about the drop in GDP being due to imports - I was questioning my own sanity for a while but it really doesn't make sense when you stop and think about it for more than a few seconds. GPT made a complete hash of it as well which was pretty concerning, I think GPT is normally an okay place to send people for a "neutral" factual source but it did highlight for me how that can also go wrong as well.
I'd assumed that the jump in imports is starving other investment; I wouldn't be surprised either to see future revisions to investment/inventories lifting the April number back up.
Thank you Noah. I wonder if access to accurate data is at risk; and if so how that may impact reporting in any form around these measurements. Not a political question - a logistical one. A root source one. A DOGE math one.
Sometimes it’s not so much fun to wonder. Always fun to read your insights tho.
Buying an imported TV subtracts from GDP if you make the additional assumption that the alternative was for a domestic manufacturer to make one additional TV and sell it to you. A lot of people seem to make that assumption. It's not exactly wrong, but it's not right, either.
No. It’s still counted the same, as consumption. The only case where it makes a difference is if you then go on to say that the foreign TV would never have been imported, and therefore your country’s next exports are higher.
You do not count the production of the good in this expenditure based approach, only where it is sold to.
If you included the production of the good, you would then double count it when it is sold. The only time the origin of the TV being purchased matters is if the balance of trade changes.
I think I'm seeing why this is so easily confusing. We expand balance of trade to exports - imports without similarly expanding consumption and investment to show their domestic and foreign contributions. If you expand everything, it's easy to see what's going on.
GDP = C + I + G + NX
= (CD + CF) + (ID + IF) + G + (X - M)
= (CD + ID + X + G) + (CF + IF - M)
The first term is all the domestic contributions. The second term is all the foreign contributions. But every dollar of imports in M is either sold, increasing CF, or held, increasing IF, so CF + IF - M = 0.
GDP = CD + ID + X + G
This is a more fundamental definition. When someone makes something, they either sell it to a private domestic party (CD), sell it to the government (G), take it out of the country (X) or keep it (ID). The reason we use C + I + G+ NX is that those are the quantities we can measure.
Let's expand this to show consumption and investment of both domestically- and foreign-produced goods:
GDP = (CD + CF) + (ID + IF) + G + (X - M)
If a domestic manufacturer makes one additional TV, it increments either CD or ID, depending on whether it was sold or not (increase in unsold inventory counts as investment). GDP goes up, as it truly should; GDP is supposed to be the value of goods and services produced domestically, so one more domestic TV should increase GDP.
If a foreign manufacturer exports one additional TV to the US, it increments either CF or IF, depending on whether it was sold or not. BUT it also increments M, exactly canceling the change to CF or IF. GDP stays the same, as it truly should; the value of domestically-produced goods and services did not change.
So if I decide that I will keep my old TV and not buy a domestically produced TV, does that also decrease GDP? What happens to GDP if I steal my neighbor’s TV? You can’t account for intentions when calculating GDP.
That's right. Not consuming means it will not be included in GDP. If something is produced but not sold (note that warehousing by a business is counted) it is not counted.
Sales are counted as consumption. Inventory ("warehousing", but also goods in transit, on shelves, etc.) is counted as investment. The only way something "not sold" gets left out of the count is if it is destroyed or stolen.
Thank you. I've been raging every time someone on Bloomberg radio said that imports reduce GDP. At least a handful got it right but didn't bother explaining it which just leaves those listening worse off
Imports. Neither did the prior quarter’s GDP. And yet, a reason GDP is lower this quarter than the period before is that more of our total consumption and investments were imports, leaving less C+ I for domestic activity counted in GDP.
As a shorthand means of summary and communication, the media claiming imports reduced GDP is pretty far down the list of bad takes the media offers.
"And yet, a reason GDP is lower this quarter than the period before is that more of our total consumption and investments were imports, leaving less C+ I for domestic activity counted in GDP."
This is incorrect without having a change in net exports.
We’re talking real data here: Q4 ‘24 vs Q1 ‘25, during which real GDP declined. Not hypotheticals.
But sure, since X-M is part of the equation and if both increased by the same amount then there would have been no change in net exports. Also, if it was 20 degrees here yesterday that would have warmer than 18 degrees, but in fact (in the real world) it was 16 degrees.
Yes, buying a Chinese TV for 1000$ does not reduce GDP. It increases consumption and imports.
In contrast, buying an American TV increases consumption and GDP.
And that's Trump's logic: by buying Chinese goods instead of American goods, US GDP is lower than it could have been. And by introducing tariffs, an incentive to buy more US and less foreign goods is created, thereby pushing GDP.
If American TVs were actually still being produced. If only TVs could be created with the same minimal effort with which a $TRUMP or $MELANIA shitcoin are created.
That is the exact question I need help answering….if GDP is 1000 when I purchase a domestic TV (Consumption), then if I buy a Chinese TV, GDP is zero — Consumption=1000 and Net exports = -1000
The question is: if you buy the Chinese TV, what does the US factory do? In the good case, the US factory creates something even more valuable that you can buy. In the bad case, the US factory is idle and the workers unemployed.
This would lead to a simple heuristic: if the US is at full employment, tariffs might do more harm than good and distract the US economy with the production of goods that are more efficiently produced elsewhere. But if there is idle production capacity and idle human resources, tariffs might actually increase GDP.
Suppose I was going to buy an imported $1,000 TV and an imported $1,000 laptop. Tariffs force me to buy a $2,000 domestic TV instead so I don’t buy the laptop. GDP goes up compared to the counter factual, but my wellbeing goes down. And that’s the best case scenario. The worst case scenario is that some of that $1,000 hit that I take prevents me from buying some other domestic thing I would have bought. Or prevents me investing domestically which reduces future GDP. Now GDP goes up in one place but down in another, *and* I’m still worse off.
Thanks for another great clarifying post on the sometimes confusing field of economics. I think the psychological aspect of this misinformed writing is a very real danger and is clearly contributing to some bad policy . I think that the phrase nobody got fired for buying IBM was more a saying in the IT world rather than the trading world although I’m sure it was used in that context 😎
Noah good post again. I am glad you included behavioral factors as well as the technical explanation. Question, what is the stock market, not the Dow, but the broader market, going to do when shortages show up in a few weeks?
Neither the stock market nor corporate purchasing managers wait for “liberation day” announcements nor events before acting. See also- the big jump in imports in Q1 or the stock market correction that started in mid-Feb. They try to anticipate them. The question is how the actual event compares to their anticipation/expectation. April 1 was worse than anticipated (especially the 145 pct Chinese tariffs some days later), so the market went down as investors recalibrated and priced in very negative scenarios (market often overshoots in response to surprises). Target and wal mart and Apple all met with admin officials to plead their cases. The market is now looking past the disruption, seeing Trump back away. The question is how bad the disruption will be relative to what they are anticipating. Big companies have some margin of error due to stockpiling. Smaller companies less so.
I expect a resolution with Canada, Mexico, S Korea, India, Bangladesh, Philippines, and Japan and the UK. EU will be a bigger conflict than people expect. Chinese tariffs aren’t going away- maybe they will be 45 percent instead of 145 percent.
Consumer goods like textiles may be less of an issue because the markups are huge along the distribution chain and the true cost at the border subject to tariffs is a fraction of the consumer price (eg the value of a Nike sneaker leaving Indonesia isn’t $100).
Bigger impact could be on intermediate/industrial goods.
With respect, I think you have it technically right but the journos aren’t wrong. They are just using loose language. In this case, the imports aren’t literally reducing GDP. However, they are resulting in GDP being lower than it would have been.
In your example, Best Buy *changed* its mix of inventory investment, accelerating its purchases of Chinese TVs and delaying its purchases of domestic products to make up for the fact it only had a fixed amount it was comfortable investing.
In the alternative plan, without the tariffs, Investment would have been the same total, but Net Exports would have been higher, since Exports would have remained the same and Imports would have been not artificially goosed.
Now, everyone—the journos and you—don’t mention the probable opposite effect of accelerating US exports in anticipation of retaliatory tariffs. If one factors *that* in, then maybe it is all a wash, depending on the magnitude of the two effects….
General question and comment. What are tariffs paid on? On the "inventory cost" or are they paid on the implied retail cost? If the former then they do not include indirect costs and profit. If true would that not mean that if the TV is sold (consumed by either the public or the government) then GDP would go up by the indirect costs and profit and even the tariff, but if the TV goes into inventory their would be no change in GDP. However if he tariff is paid on the implied retail cost then there would be no increase in C,G or I.
My second question would be how is the tariff cost accounted for in GDP? If a tariff is paid or either the "inventory cost or the retail cost" and is passed on why would that not increase GDP?
Yes, like the famous case of replacing damage from a hurricane increases GDP - but doubtful that the overall welfare of the hurricane hit area is better......
If GDP is supposed to measure production of a country, the formula of C + G + I + NX can be clearly seen by an extreme and silly example case.
Imagine a country that produces nothing and imports everything. Furthermore it’s a small country that has large cash reserves, but that it historically had some valuable natural resource like unobtanium that it previously sold all of to other countries and now the government is just a few people who guard this money and send an equal fraction to each citizen. Since the government doesn’t do much G is ≈0 and since everyone just imports everything they consume (assume they go abroad for medical care, or haircuts) there is no investment so then I is also ≈0. Since nothing is produced exports are 0, leaving only consumption and imports. Since all consumption is of imports and all imports are consumed, NX and C cancel each other out and the equation comes out to GDP of zero, correctly showing the country produces nothing. People living there are happy and idle, but they contribute nothing and are a drain on productivity elsewhere (kind of like seniors on social security) so having a GDP of zero is fitting.
The question remains that even if this fictional country has the lowest possible GDP (since it cannot be negative), that doesn’t imply it wouldn’t be nice to live there. Of course Trump would say it is getting terribly ripped off because of the huge trade deficit, but how would you even describe its imports as a percentage of GDP, which is zero?
Yep, GDP is not a particularly good measure of the overall welfare of the country and it is a bit painful that it is used as a "measure" of that so widely.
It’s worth noting that import do increase GDP because there is domestic economic activity associated with them. They need to be transported and sold. The import subtraction from consumption will be at wholesale price, not the final consumer price.
In the Chinese’s TV example, the net is only zero if the consumer goes to the port and unloads the TV from the ship themselves. And even then it’s still not zero since because transporting the TV themselves would have required some economic activity even if it is just from the gas or electrons to get to and from the port.
The pro-tariff people ignore there is considerable domestic economic activity from imported goods since they have to be transported and retailed.
"Putting on heavier shoes does not make you a thinner person"
No, but putting on a bigger shirt does!
Sure- if you define your weight as ex-shoes you can say there is no difference whether they weigh 2 pounds or 20….until you try to run.
Thanks for posting Noah. I've been bemused by the plethora of articles in the last few days from the WSJ/Bloomberg/Reuters/PickYourOutlet talking about the drop in GDP being due to imports - I was questioning my own sanity for a while but it really doesn't make sense when you stop and think about it for more than a few seconds. GPT made a complete hash of it as well which was pretty concerning, I think GPT is normally an okay place to send people for a "neutral" factual source but it did highlight for me how that can also go wrong as well.
I'd assumed that the jump in imports is starving other investment; I wouldn't be surprised either to see future revisions to investment/inventories lifting the April number back up.
Thank you Noah. I wonder if access to accurate data is at risk; and if so how that may impact reporting in any form around these measurements. Not a political question - a logistical one. A root source one. A DOGE math one.
Sometimes it’s not so much fun to wonder. Always fun to read your insights tho.
Buying an imported TV subtracts from GDP if you make the additional assumption that the alternative was for a domestic manufacturer to make one additional TV and sell it to you. A lot of people seem to make that assumption. It's not exactly wrong, but it's not right, either.
No. It’s still counted the same, as consumption. The only case where it makes a difference is if you then go on to say that the foreign TV would never have been imported, and therefore your country’s next exports are higher.
Wait, no. I didn't forget that. If another domestic TV gets produced, GDP goes up, period. Doesn't matter whether the import happens or not.
I can't understand why this is so difficult when it's so simple.
GDP = C + I + G + (X - M)
You do not count the production of the good in this expenditure based approach, only where it is sold to.
If you included the production of the good, you would then double count it when it is sold. The only time the origin of the TV being purchased matters is if the balance of trade changes.
I think I'm seeing why this is so easily confusing. We expand balance of trade to exports - imports without similarly expanding consumption and investment to show their domestic and foreign contributions. If you expand everything, it's easy to see what's going on.
GDP = C + I + G + NX
= (CD + CF) + (ID + IF) + G + (X - M)
= (CD + ID + X + G) + (CF + IF - M)
The first term is all the domestic contributions. The second term is all the foreign contributions. But every dollar of imports in M is either sold, increasing CF, or held, increasing IF, so CF + IF - M = 0.
GDP = CD + ID + X + G
This is a more fundamental definition. When someone makes something, they either sell it to a private domestic party (CD), sell it to the government (G), take it out of the country (X) or keep it (ID). The reason we use C + I + G+ NX is that those are the quantities we can measure.
Let's expand this to show consumption and investment of both domestically- and foreign-produced goods:
GDP = (CD + CF) + (ID + IF) + G + (X - M)
If a domestic manufacturer makes one additional TV, it increments either CD or ID, depending on whether it was sold or not (increase in unsold inventory counts as investment). GDP goes up, as it truly should; GDP is supposed to be the value of goods and services produced domestically, so one more domestic TV should increase GDP.
If a foreign manufacturer exports one additional TV to the US, it increments either CF or IF, depending on whether it was sold or not. BUT it also increments M, exactly canceling the change to CF or IF. GDP stays the same, as it truly should; the value of domestically-produced goods and services did not change.
So if I decide that I will keep my old TV and not buy a domestically produced TV, does that also decrease GDP? What happens to GDP if I steal my neighbor’s TV? You can’t account for intentions when calculating GDP.
That's right. Not consuming means it will not be included in GDP. If something is produced but not sold (note that warehousing by a business is counted) it is not counted.
Sales are counted as consumption. Inventory ("warehousing", but also goods in transit, on shelves, etc.) is counted as investment. The only way something "not sold" gets left out of the count is if it is destroyed or stolen.
Yes; forgot to include that part.
thanks for re-emphasizing this! the incorrect interpretation in virtually every commentary on the gdp data release has been driving me nuts....
Thank you. I've been raging every time someone on Bloomberg radio said that imports reduce GDP. At least a handful got it right but didn't bother explaining it which just leaves those listening worse off
Sure, GDP doesn’t include
Imports. Neither did the prior quarter’s GDP. And yet, a reason GDP is lower this quarter than the period before is that more of our total consumption and investments were imports, leaving less C+ I for domestic activity counted in GDP.
As a shorthand means of summary and communication, the media claiming imports reduced GDP is pretty far down the list of bad takes the media offers.
"And yet, a reason GDP is lower this quarter than the period before is that more of our total consumption and investments were imports, leaving less C+ I for domestic activity counted in GDP."
This is incorrect without having a change in net exports.
We’re talking real data here: Q4 ‘24 vs Q1 ‘25, during which real GDP declined. Not hypotheticals.
But sure, since X-M is part of the equation and if both increased by the same amount then there would have been no change in net exports. Also, if it was 20 degrees here yesterday that would have warmer than 18 degrees, but in fact (in the real world) it was 16 degrees.
Yes, buying a Chinese TV for 1000$ does not reduce GDP. It increases consumption and imports.
In contrast, buying an American TV increases consumption and GDP.
And that's Trump's logic: by buying Chinese goods instead of American goods, US GDP is lower than it could have been. And by introducing tariffs, an incentive to buy more US and less foreign goods is created, thereby pushing GDP.
If American TVs were actually still being produced. If only TVs could be created with the same minimal effort with which a $TRUMP or $MELANIA shitcoin are created.
That is the exact question I need help answering….if GDP is 1000 when I purchase a domestic TV (Consumption), then if I buy a Chinese TV, GDP is zero — Consumption=1000 and Net exports = -1000
Yes, that's what the equation says.
The question is: if you buy the Chinese TV, what does the US factory do? In the good case, the US factory creates something even more valuable that you can buy. In the bad case, the US factory is idle and the workers unemployed.
This would lead to a simple heuristic: if the US is at full employment, tariffs might do more harm than good and distract the US economy with the production of goods that are more efficiently produced elsewhere. But if there is idle production capacity and idle human resources, tariffs might actually increase GDP.
Suppose I was going to buy an imported $1,000 TV and an imported $1,000 laptop. Tariffs force me to buy a $2,000 domestic TV instead so I don’t buy the laptop. GDP goes up compared to the counter factual, but my wellbeing goes down. And that’s the best case scenario. The worst case scenario is that some of that $1,000 hit that I take prevents me from buying some other domestic thing I would have bought. Or prevents me investing domestically which reduces future GDP. Now GDP goes up in one place but down in another, *and* I’m still worse off.
The real question is how did you pay for the imported goods?
Did you go into debt, or sell off part of your country
I don’t particularly love it, but Noah wrote an okay article about this that it sounds like you should give a read, since you’re here. https://www.noahpinion.blog/p/trade-deficits-do-not-make-a-country
I read it already.
It doesn't contradict what I just said..
By the logic above, don’t import than lead to a reduction in GDP?
Thanks for another great clarifying post on the sometimes confusing field of economics. I think the psychological aspect of this misinformed writing is a very real danger and is clearly contributing to some bad policy . I think that the phrase nobody got fired for buying IBM was more a saying in the IT world rather than the trading world although I’m sure it was used in that context 😎
Noah good post again. I am glad you included behavioral factors as well as the technical explanation. Question, what is the stock market, not the Dow, but the broader market, going to do when shortages show up in a few weeks?
Neither the stock market nor corporate purchasing managers wait for “liberation day” announcements nor events before acting. See also- the big jump in imports in Q1 or the stock market correction that started in mid-Feb. They try to anticipate them. The question is how the actual event compares to their anticipation/expectation. April 1 was worse than anticipated (especially the 145 pct Chinese tariffs some days later), so the market went down as investors recalibrated and priced in very negative scenarios (market often overshoots in response to surprises). Target and wal mart and Apple all met with admin officials to plead their cases. The market is now looking past the disruption, seeing Trump back away. The question is how bad the disruption will be relative to what they are anticipating. Big companies have some margin of error due to stockpiling. Smaller companies less so.
I expect a resolution with Canada, Mexico, S Korea, India, Bangladesh, Philippines, and Japan and the UK. EU will be a bigger conflict than people expect. Chinese tariffs aren’t going away- maybe they will be 45 percent instead of 145 percent.
Consumer goods like textiles may be less of an issue because the markups are huge along the distribution chain and the true cost at the border subject to tariffs is a fraction of the consumer price (eg the value of a Nike sneaker leaving Indonesia isn’t $100).
Bigger impact could be on intermediate/industrial goods.
With respect, I think you have it technically right but the journos aren’t wrong. They are just using loose language. In this case, the imports aren’t literally reducing GDP. However, they are resulting in GDP being lower than it would have been.
In your example, Best Buy *changed* its mix of inventory investment, accelerating its purchases of Chinese TVs and delaying its purchases of domestic products to make up for the fact it only had a fixed amount it was comfortable investing.
In the alternative plan, without the tariffs, Investment would have been the same total, but Net Exports would have been higher, since Exports would have remained the same and Imports would have been not artificially goosed.
Now, everyone—the journos and you—don’t mention the probable opposite effect of accelerating US exports in anticipation of retaliatory tariffs. If one factors *that* in, then maybe it is all a wash, depending on the magnitude of the two effects….
General question and comment. What are tariffs paid on? On the "inventory cost" or are they paid on the implied retail cost? If the former then they do not include indirect costs and profit. If true would that not mean that if the TV is sold (consumed by either the public or the government) then GDP would go up by the indirect costs and profit and even the tariff, but if the TV goes into inventory their would be no change in GDP. However if he tariff is paid on the implied retail cost then there would be no increase in C,G or I.
My second question would be how is the tariff cost accounted for in GDP? If a tariff is paid or either the "inventory cost or the retail cost" and is passed on why would that not increase GDP?
Either way, GDP remains a pretty lousy indicator for overall societal wellbeing.
But it's interesting to see that even the main political/economic indicator of the last century is so misunderstood.
Yes, like the famous case of replacing damage from a hurricane increases GDP - but doubtful that the overall welfare of the hurricane hit area is better......
If GDP is supposed to measure production of a country, the formula of C + G + I + NX can be clearly seen by an extreme and silly example case.
Imagine a country that produces nothing and imports everything. Furthermore it’s a small country that has large cash reserves, but that it historically had some valuable natural resource like unobtanium that it previously sold all of to other countries and now the government is just a few people who guard this money and send an equal fraction to each citizen. Since the government doesn’t do much G is ≈0 and since everyone just imports everything they consume (assume they go abroad for medical care, or haircuts) there is no investment so then I is also ≈0. Since nothing is produced exports are 0, leaving only consumption and imports. Since all consumption is of imports and all imports are consumed, NX and C cancel each other out and the equation comes out to GDP of zero, correctly showing the country produces nothing. People living there are happy and idle, but they contribute nothing and are a drain on productivity elsewhere (kind of like seniors on social security) so having a GDP of zero is fitting.
The question remains that even if this fictional country has the lowest possible GDP (since it cannot be negative), that doesn’t imply it wouldn’t be nice to live there. Of course Trump would say it is getting terribly ripped off because of the huge trade deficit, but how would you even describe its imports as a percentage of GDP, which is zero?
Yep, GDP is not a particularly good measure of the overall welfare of the country and it is a bit painful that it is used as a "measure" of that so widely.
I had a look at the BEA archive and it seems they started using the phrase "imports subtracts from GDP" in July 2000.
I think a lot of reference material and coursework for AP Econ and Econ 101 get this wrong:
https://www.albert.io/blog/how-to-calculate-gross-domestic-product-for-ap-macroeconomics/
The AP itself is correct, luckily, but poor HS teachers could be contributing to the myth.
It’s worth noting that import do increase GDP because there is domestic economic activity associated with them. They need to be transported and sold. The import subtraction from consumption will be at wholesale price, not the final consumer price.
In the Chinese’s TV example, the net is only zero if the consumer goes to the port and unloads the TV from the ship themselves. And even then it’s still not zero since because transporting the TV themselves would have required some economic activity even if it is just from the gas or electrons to get to and from the port.
The pro-tariff people ignore there is considerable domestic economic activity from imported goods since they have to be transported and retailed.