The editors of the Wall Street Journal are perplexed over the 3% growth in GDP in the 2nd quarter of 2025:
If you think President Trump’s tariff ructions don’t affect the economy, take a gander at Wednesday’s report for second quarter gross domestic product. The economy grew 3% on an annual basis, but largely because imports collapsed.
This may be the weirdest GDP report ever. The top line growth number looks good, and the White House naturally touted it. This reverses the 0.5% decline in GDP in the first quarter, which was largely explained by a surge of imports as businesses tried to front-run the anticipated tariff barrage. Growth in the first half was a mediocre 1.2%.
Most striking are the second quarter report’s wild internal details. Net exports (exports minus imports) added a remarkable 4.99% to GDP as imports fell 30.3%. Imports subtract from growth in the national accounts because GDP measures domestic production. Imports are produced overseas. But imports are still crucial to U.S. economic well-being because consumers buy them and businesses use them as inputs for what they produce—and often export.
The crazy swing in imports shows how much Mr. Trump’s up-and-down trade policies have disrupted business decisions and left companies scrambling to adapt. This seems to have had a negative effect on private domestic investment, which fell 15.6% in the second quarter after a surge in the first.
Nonresidential business investment contributed only 0.27% to GDP, as businesses rapidly drew down their inventories. Chalk this up as another result of the uncertainty caused by on-again, off-again, on-again tariffs.
I should note that the decline in business investment is completely consistent with what I have been predicting and one of the reasons I have been skeptical about President Trump’s tax policies. I don’t think they encourage investment as much as he apparently does and it is my conviction that the most compelling American economic problem is that we are not investing enough.
I have long questioned the utility of the “expenditure approach” to calculating gross domestic product (GDP). That is GDP = Personal Consumption + Business Investment + Government Expenditures + Net Exports. I believe it is an artifact of an earlier day when governments did not do as much redistribution as is presently the case.
Using the income approach to GDP calculation arrives at the same result as the expenditures approach for the 2nd quarter of 2025. That is unlike the previous quarter in which the expenditure approach showed a slight decline while the income approach reflected a slight increase.
Who cares? (I hear someone ask.) I care because of the differing effects on policy. The expenditure approach encourages government spending. The income approach does not.
If you’re wondering what the income approach to GDP calculation is it’s
GDP = Total National Income + Sales Taxes + Depreciation + Net Foreign Factor Income
I was listening to Hassett claim that tariffs are now shown to be a huge success. I dont buy it. Maybe corporations really were sitting on an extra 15% in profits they really didnt need. You might get Se. Warren to agree, but I dont think that is realistic. I do think it likely that there will be some hesitancy to pass costs on to the consumer when you dont know if the tariffs are going to go away, if you are going to get rebates or if they will suddenly increase. Why piss off the consumer over a few months of losses?
Which gets back to what I think was the claimed purpose of the tariffs, though God knows the claims change frequently (or is that literally?). I thought they were supposed to bring back manufacturing (and maybe jobs) to the US. If that is true then it seems investment should be the core metric and that was down. Seems to me that the likely ceiling on the positive effects of tariffs is that they might raise some revenue. Businesses will decide to just accept the 15% tariffs, 50% for some products, and otherwise carry on as normal. No new surge in investments. Businesses will then work on ways to avoid paying tariffs the way they have corporate income taxes.
Steve
Would changing tax policy to allow companies to bring back overseas earnings without cost improve domestic investment? Or did this change when I stopped looking?
I still think we have a dark matter / dark energy problem — i.e. GDP isn’t measuring the investment craze into AI / data centers. Yet its presence can be felt in the AI mania. And can you have a mania if there isn’t (mal)investment?
Just to put it in context, Microsoft is spending $30 billion this quarter on capex, and projecting $100 billion in their 25/26 fiscal year. Google is projecting capex of $85 billion for 2025, Meta $72 billion. And these are growing 30-40% per year and the hyperscalers are revising their projections upwards every quarter. To put it on a scale, each of these companies is spending more in capex then the 3 major telecom companies combined!
Why isn’t this showing up in the investment numbers?
What curious said. And.
You are correct to advocate the alternative measure for the reason you cite. Its just a better metric. I might note, investment in tangible assets is up year over year. And contra-steve, foreign corps are investing here. Clever to leave that out…..
I managed to piss off Dr Taylor when he posted a note talking about the stock market performance being inferior to Biden……………..19 days into Trumps term. His was a stupid comment. At the time i noted that one probably needed to get to Q3 to really see how this all would shake out. Q1 and Q2 have import to and fro. I’m still thinking Sept/Oct. Maybe that’s why Powell want to wait until the Sept meeting.
But any way you look at it, the so called experts missed it by a country mile. You might almost think they were partisans, not analysts. (snicker)
Most of our companies have had some effects of the tariff issue. But its not nearly what its made out to be in the political sphere. A pain? Yes. Manageable. Yes. No more a problem than the everyday regulatory crap we have to deal with. But one hears nothing about regulatory issues from Chuckie Schumer etc. Imagine that.
I think it is. That’s our very problem. Investment is not what it should be despite the large investments of Magnificent Seven companies.
Agree with Dave. Why would they miss the huge amounts being invested into AI, assuming it’s really happening? I think part of the issue is that they are making announcements about PROJECTED investments not what they are actually investing. Meanwhile, measuring what people are actually investing is not increasing.
Steve
Furthermore, it’s quite possible that the investments of non-technology companies are shrinking even as those of technology companies are growing.
Extend the line from 2010-2015 (more or less straight) to see the change I’m pointing to.
Also note that the Trump first term tax cuts created no “bump” in investment.
Hasn’t history shown that the more the economy is run by a single person things dont go well?
Steve
The AI announcements are real. It shows up in the balance sheets and income statements of the hyperscalers.
Its also fairly obvious if you are talking with “normal” people or using the tools. Adoption is picking up noticeably.
The employment revisions today – OOF!
The announcements are real, but announced intentions dont always become reality.
The jobs numbers are going to make it difficult for the FED. PCE was back up again and with the tariffs being continually (partially) delayed price impacts are not clear yet.
Steve
Icepick (oh, and good to have you back)
Its always good to have more money, but I don’t think that’s the key to investment. Companies invest because they a) see market opportunity or b) see available efficiencies. I know Dave desires more investment. Me too. But you don’t invest unless you see a or b. And the notion that its just easier to buy equities on the exchanges is just bull. Rather, it may be the best use of cash.
If we want more investment we need to foster growth. That means less government intrusion and regulation.
“The employment revisions today – OOF!”
Yes, but this woman has been pulling this stuff during her entire tenure. She is rotten to the core. And good riddance.
So much focus on tech and sexy companies. But there are very mundane companies all across the country, a bazillion of them.
During out diligence there always comes the point where we sit down with management and ask “what do you want to do with this company? And what do you need?’ Its a 5 year plan.
We have our ideas. They have theirs. But the point is to allocate money and human capital to the effort – and have a common view – right from the start. There is almost no investment we make where capex isn’t stepped up.
It might be interesting (or maybe not) to see some of the very unsexy companies we have/do own.
a professional spray painting equipment company
a manufacturer of roof trusses for commercial buildings, particularly warehouses
an injection molder of bottle closures
an injection blow molder of the bottles
a one shot aluminum can maker for personal care products, and beer
an epoxy resin countertop manufacturer
a manufacturer and installation project manager of outdoor signage. Not Joes Pizza. Think CVS or Walgreens
a manufacturer of metal doors and related hardware for commercial buildings
an industrial bakery making pies and cakes by the handy million
a manufacturer of concrete culverts and junction boxes for roadwork
A manufacturer of concrete sills and other horizontal structures for residential housing
a manufacturer of sheet metal forming equipment
a manufacturer of dies to impart decorative designs to concrete walls
an HVAC installation company also providing retrofit and maintenance services
I could go on. You get the point. Not sexy like AI, or cell phones, Facebook etc etc
All require and get capex. Many under attack by imports. But our goal is to invest in $ and human capital to make them prosper…….and stay here. The big companies get the headlines. But there are a bazillion of these small companies. The numbers add up. I think Trump understands this. Just give us a reasonably level playing field. Other countries levy tariffs on us. Why sit back and play patsy?
It appears that Intel is about to stop construction on its Ohio chip foundary despite some $8 billion in expenditures. This despite substantial subsidies. They are looking for a buyer. Also, the TSMC is losing money on its US plant.
Deindustrialization of the West continues unabated.
Drew: Yes, but this woman has been pulling this stuff during her entire tenure. She is rotten to the core.
The head of the Bureau of Labor Statistics doesn’t make the reports. She oversees the teams who conduct the Current Employment Statistics and Current Population Surveys, and the team of statisticians and economists who analyze the data using a specified methodology.
The methodology is meant to capture the markets in real time, but when the markets are in a state of rapid flux, the initial reports are often less accurate. None of this is extraordinary, but intrinsic to the process. The result will be to weaken the validity of the reports, undermining rational policy.
Killing the messenger; however, is a long-time tradition. So there is that.
CuriousOnlooker: GDP isn’t measuring the investment craze into AI / data centers.
Investment is a component of GDP = C + I + G + (X – M), where I is investment. Of course, AI may very well have a major impact on future productivity.
steve: I think part of the issue is that they are making announcements about PROJECTED investments not what they are actually investing.
That’s right. Announcements are not included in GDP, only actually expenditures.
steve: Hasn’t history shown that the more the economy is run by a single person things dont go well?
Markets, while imperfect, are essential for the efficient allocation of resources. And while markets require regulation for proper function, the regulation needs to be done in an orderly predictable fashion based on rational principles, which usually implies a legislative process and expert bureaucracy.
All of the data compiled by BLS is, by report, obtainable by the public and can be reviewed. If you actually follow the commentary on the BLS then you know that they are facing a problem in that, especially after the pandemic, people are reporting their numbers later so the initial report is based upon a smaller set of data. That means larger corrections. So once again, we have claims made, just like we did with the crime numbers recently, with nothing to support the claims. Show your numbers.
Of course we should know what is really going on. When numbers were announced and revised down Trump said that was done to help Biden. When they are announced and revised down that was done to hurt him. This is just pure politics/ego. It also didnt really make sense. As soon as the revisions came out the talk about the FED cut in September, which Trump wants, went from talking about a 25 basis point cut to 50. Alas, Trump cannot fail, he can only be failed.
Steve
She sits on top, Zach. It’s her responsibility. It’s been nothing but outsized (and during Biden) one way revisions.
Drew: She sits on top
The data is the data. It’s not her job to make up numbers. It would be different if Trump provided any evidence whatsoever that she wasn’t doing her job. While Trump sits at the top, firing her will not make the numbers any better.
Drew: It’s been nothing but outsized (and during Biden) one way revisions.
Large adjustments are usually associated with economic disruptions.
As I wrote in my latest post (above), I think that President Trump erred in discharging the BLS chief. Also contra Zachriel it is the BLS’s job to “make up numbers”. Only a minority of what they’re reporting are actually sampled or surveyed data.
He does make a good point about economic disruptions. We’ve been in a period of massive disruptions for the last 20 years that show few signs of abating.
Dave Schuler: Also contra Zachriel it is the BLS’s job to “make up numbers”.
If you think that 2+3=5 is “making up numbers”, then sure; but it’s an odd use of the idiom. In essence, they collect data, then use an algorithm to spit out the result: ƒ(2,3)=5.
The algorithm, by the way, is largely based on standard statistical analysis, such as weighting by industry and region, as well as season.
You clearly know nothing about how the Employment Situation Report is produced. I suggest studying up a bit on it before making foolish comments like that.
It’s not like 2+3=5. It’s like 2+3+various adjustments=7
Alternatively, you’re using “data” to mean something other than data.
Dave Schuler: It’s not like 2+3=5. It’s like 2+3+various adjustments=7
2+3=7 would be “making up numbers”, and would mean no connection between the data (2,3) and the result (7). However, surveys are a form of averaging, and often use weighting and other statistical means to provide a more accurate result. That’s not “making up numbers”.
You posted on the topic again and provided more detail on your view. Have posted more in that thread.
You are being silly now Dave. You are basically saying that there is no value, ever, in any kind of statistical sampling or polling.
Steve
It seems that the data collected by the IRS and the SSA should be able to give an excellent snapshot of how things are going. It could be scrubbed of individual identifiers easily enough. It wouldn’t be as an exact match to reality but would be a much closer approximation.
It probably won’t surprise you that using those data in that way is expressly prohibited by law.