Sorry MAGA, the GDP Report Is Not Rosy

(This post dovetails with the personal podcast episode Murphy did here.)

The Bureau of Economic Analysis (BEA) recently released its Advance Estimate of real GDP growth for the 1^st quarter of 2025, and it came in at an alarming negative 0.3 percent. (For context, the 4^th quarter 2024 reading was positive 2.4 percent.) Predictably, Trump’s critics cited the report as proof that his tariffs are terrible, while his fans (and Trump himself) asked for more time before rendering the final verdict.

In this vein, I came across a very interesting analysis from economist Peter St. Onge (a colleague whom I’ve interviewed previously regarding the yen carry trade). Peter argued that once you adjust for the quirks in how GDP is calculated, you see that this report actually shows Trump is winning. I happen to disagree with (most of) Peter’s points, but it’s instructive to present his case—and my rebuttal—to explain the underlying economics, and also to shed light on what’s going on with Trump’s tariffs.

Peter St. Onge on GDP Report: “It’s Glorious”

Peter argues that the GDP report is actually “glorious” once we make some adjustments in light of sound economics. Specifically, Peter claims that once we look at the components of GDP that add up to the negative 0.3 percent growth, we see why (in his mind) the situation is actually quite rosy: 

 As this chart indicates, the main driver of the negative figure is a surge in imports. (Imports are subtracted in the formula GDP = C + I +G + NX, where “Net Exports” are Exports minus Imports.) Businesses accelerated their imports because of the looming Trump tariff hikes, and so this led to downward pressure on the official GDP figure. However, Peter argues, we should recognize that the mere flow of goods into the country doesn’t mean the economy is producing less, and so that’s not an economically meaningful calculation. Hence, Peter throws it out as being irrelevant to the welfare of actual Americans (as opposed to a numerical estimate).

The other big driver was the surge in investment, which Peter attributes to companies relocating inside the United States (because of the tariff threat). He views this as unambiguously good for the US economy and American workers. 

Finally, Peter points out that the (modest) cut in government spending technically reduces GDP because “G” is one of its components, yet Peter thinks in reality, government expenditures do not signify genuine economic output. Here again, Peter throws out this figure, which allows him to conclude: “Control for that drop in government spending and the economy grew by nearly 5 percent on the quarter.”

Although I was intrigued by Peter’s provocative analysis, after reviewing the BEA report I concluded that I (mostly) disagree with him, as I explain in the rest of this post.

“Structures” a Tiny Component of Private Investment in 1^st Q 2025

I believe that the most useful data to decompose the constituents of “real” GDP can be found in Table 3 (which starts on page 9 of 17) of the pdf version of the BEA report. Specifically, I will be referencing the right-hand side of this table, which looks at inflation-adjusted dollars (which have been calibrated to match the purchasing power of a “chained” 2017 dollar). Furthermore I’ll be looking at the “Change from preceding period” column so that we can understand the constituents of the overall 0.3 percent drop in real GDP from the 4^th quarter of 2024 to the 1^st quarter of 2025.

As the screenshot below shows, Gross Private Domestic Investment (which is the “I” in the GDP formula C+I+G+NX) grew by $218.6 billion (in adjusted dollars) from the prior quarter. (See green circle). 

However, of that surge, only a mere $0.6 billion—in other words, $600 million—was due to “Structures.” For frame of reference, in the prior period (i.e. 4^th quarter of 2024), real spending on nonresidential “Structures” rose by $4.8 billion. So contrary to Peter’s analysis, it doesn’t seem that companies relocating into the US is the big story here.

Private Inventory Adjustment Is the Driver

The real driver of the big investment number was the surge in private inventories (red line):

As the line indicates, of the $218.6 billion increment in investment, $131.3 billion of that was due to an increase in the growth of private inventories. (To avoid confusion for purists: This is technically a “change in the change,” but the level of the change in private inventories was only $8.9 billion [in inflation-adjusted dollars] in 4^th quarter of 2024, meaning the $131.3 billion “increase in the change” in 1^st quarter 2025 was almost the same thing as the level of the change, which was $140.1 billion.)

Now to be clear, much nonsense has been written about “inventory bounces” and how they can interact with the official GDP statistics, as I’ve explained elsewhere. Even so, the basic logic of the “Change in private inventories” under the “Gross private domestic investment” category makes economic sense. The whole point of GDP is to measure Gross Domestic Product, i.e. how much is produced (of final goods and services) in the country during a particular period.

In practice, economists have better data on sales than they do on production, and so the popular formula GDP = C+I+G+NX measures the spending of those four components. But if businesses order the production of more units and store them in the warehouse for future sales, this period’s “C” won’t pick it up, because the consumers haven’t bought them yet. That’s why the implicit business investment in bulking up their inventory is counted as part of GDP, working its way through the I component. 

In the future, as businesses draw down their inventories and sell the units to consumers, there would be offsetting effects: On the one hand, the consumer spending would raise GDP, but on the other, the negative inventory adjustment would reduce it. This is exactly what we want to happen: Once units are physically produced and added to inventory, they should only be counted in GDP for that period. In the future, as these units are sold off, they shouldn’t count again as part of (future) GDP. 

Without knowing more, we can’t say whether the surge in private inventories in 1^st quarter 2025 is a sign of a productive American economy or not. That’s where the import data come into play.

Subtracting Imports

 As the next screenshot shows, there was a $333.3 billion surge in imports during the quarter. For context, imports actually fell by $17.6 billion in the prior quarter, and the overall level of (inflation- and seasonally adjusted, annualized) imports in Q1 was actually 9 percent higher than the level in the prior quarter. 

Although Peter is right that there is nothing intrinsically anti-productive about imports, nonetheless it is correct to “subtract” them from GDP. This is because the figures of C and I capture all private spending on consumption and investment even if the items were produced abroad. Therefore, these earlier components overcount consumption and investment production in the US, and so we need to subtract out spending on imports in order to correct the overcount. 

In context, it seems very likely that much of the surge in private inventories consisted of businesses “pulling forward” their normal imports ahead of Trump’s tariff hikes. Because final consumer demand wasn’t ready to increase by the same degree, much of these extra imports went into storage. This all makes perfect sense, but it shows that Peter is wrong for simply adding the import number back into GDP as if these facts aren’t relevant.

Government Expenditures

Finally, consider the $14.5 billion fall in government (at all levels) expenditures. (To avoid confusion, only government spending on final consumption or investment is included in GDP; things like Social Security payments are considered transfers, and don’t contribute to GDP.) 

 Here I agree entirely with Peter that the drop per se shouldn’t be construed as a “bad economy,” since government spending is not on the same footing as private-sector spending. Yet even so, it’s probably an overcorrection when Peter simply adds those dollars back into GDP. The reason is that a cut in government spending frees up those resources which can be channeled back—and show up in—the C, I, and NX components of GDP. In other words, because the government spent $14.5 billion less (in real terms) in the quarter, that should have allowed more private-sector spending—perhaps not dollar for dollar, but at least a partial offset.

Conclusion

 As I have emphasized on the InFi podcast, I am not reflexively castigating Trump’s tariff plans as idiotic (the way many of my colleagues have). I recognize there are serious structural problems with the US economy, and the chronic trade deficits of recent decades are one symptom of them.

However, I can’t go along with Peter St. Onge’s analysis of the 1^st quarter GDP numbers. Prima facie, they are cause for alarm, and indicate that the US economy had already been shrinking in the first three months of this year. 

Dr. Robert P. Murphy is the Chief Economist at infineo, bridging together Whole Life insurance policies and digital blockchain-based issuance.

Twitter: @infineogroup, @BobMurphyEcon

Linkedin: infineo group, Robert Murphy

Youtube: infineo group

To learn more about infineo, please visit the infineo website