The advance estimate for first-quarter 2025 GDP, released on April 30, created a big stir. The headline number showed the U.S. economy shrinking at an annualized rate of -0.3 percent, compared to a 2.8 growth rate for the full year of 2024. Pessimists were quick to cast the Q1 numbers as the leading edge of a possible recession.

However, others were quick to point out that most of the change in GDP came from an unusual jump in imports. In a widely-read piece for the Financial Times, economist Jason Furman suggested that we should instead  be looking at “core GDP,” which grew at a reassuring 3.0 percent annual rate.

Core GDP is Furman’s reader-friendly term for a statistic that the BLS awkwardly calls “final sales to private domestic purchasers.” That number includes consumption and fixed investment but leaves out imports, exports, government purchases, and private inventory investment. Furman’s reasoning is that the omitted items are often affected by idiosyncratic factors, making core GDP a better predictor of economic growth going forward.

Furman may well be right about core GDP as a more stable predictor of future growth, but I am less comfortable with the idea that the 3 percent core growth in Q1 meant that all was well. To see why, it is worth taking a closer look at how imports affect the GDP numbers in general, and at their sector-by-sector effects in Q1 2025.

Why subtract imports?

To understand what went on in Q1, it helps to keep in mind just what GDP means and why it is measured the way it is. US GDP is supposed to be a measure of productive activity within the geographic borders of the United States. Theoretically, the best way to measure that would be to sum up the value added by each firm or independent contractor in the country. The value of steel sold by Nucor minus the value of raw materials and electricity, the amount the person who paints your house bills you minus the cost of paint and brushes, and so on. But that turns out to be impractical. Instead, the Bureau of Economic Analysis (BEA) at the Commerce Department does it a different way.

Instead of measuring goods and services at the point they are produced, the BEA measures the quantity of goods and services purchased in each sector of economy – consumption, investment, government activities, and exports. To avoid double counting, the BEA is careful only to count purchases by final users. (For example, food bought to eat at home and meals purchased in restaurants, but not the food that restaurants use to make those meals.)

But even after leaving out intermediate goods, there is still a problem in measuring domestic product because each sector’s purchases of final goods are contaminated by the inclusion of imports. It turns out to be impracticable to track imports at the point of sale,  for example, by asking Costco to report how many of the tomatoes it sells are grown in California and how many in Mexico. Instead, the Census Bureau measures imports as they come into the country. The BEA then subtracts total imports from total final purchases. That gives us this familiar equation: GDP = Consumption + Investment + Government Purchases + Exports – Imports.

If we keep all that in mind, we can see why core GDP does not really tell as much as we would like to know about the effects of the Q1 import surge on domestic economic activity. Core GDP would be a clean number for that purpose if all the extra imports in Q1 had gone into inventories, but that does not seem to have been the case. Instead, some of the extra imports appear to have gone directly into consumption and fixed investment. To that extent that was true, the reported values for consumption and fixed investment would include a smaller share of American-made goods than usual. To put it another way, it would mean that “true” core GDP – the parts of Q1 consumption and fixed investment that were attributable to domestic value added – would be overstated.

Where did the imports go?

We don’t have all the data we would need to calculate “true” core GDP for Q1, but there are some clues in the data we do have. Consider Table 1, which compares each sector’s contribution to economic growth in Q1 2025 to the full year of 2024.

The big jump in imports shows up in this table as a -5.03 percentage point deduction from total GDP growth, 4.3 percentage points larger than the impact of imports during 2024. The reason overall GDP did not fall off the cliff is that the effects of the import surge was largely offset by positive contributions from other GDP components. The largest single offset was a huge rise in inventory investment. Inventories contributed just 0.06 percentage points to growth in 2024 but jumped to 2.25 percentage points in Q1 2025 – by far the largest value for that item since the pandemic year of 2021. Still, inventories absorbed less than half of the import surge.

Some of the rest appears to have gone into fixed investment. That component’s 1.34 percentage point  contribution to GDP growth in Q1 was more than double its 2024 value. Detailed data from the Commerce Department shows that almost all of that – 1.16 percentage points – came from information processing equipment and software. One possibility is that businesses were not just importing extra computers to build inventories for later resale but were also beefing up their own IT departments before tariffs made it more costly to do so. Fixed investment in transportation equipment showed a 0.15 percentage point uptick, some of which might also have been tariff-motivated. Residential investment and other business equipment showed little change from 2024.

Consumption also showed a positive contribution to Q1 growth of 1.21 percentage points, but that was a little less than the 2024 baseline, and almost all of it was in services. It is possible that consumers made some tariff-motivated increase in purchases of imported goods, offsetting it with decreases in purchases of domestic goods, but there are only scattered data to support that hypothesis. Sales of imported cell phones and other goods did tick up 14 percent from January to February, according to the latest data available. Imports of cars showed little change, however.

The overall impression, then, is that less than half of the import surge went directly to inventory. A small amount may have shown up in government purchases or purchases of goods that were later reexported.

So, was everything OK?

Where, then, does that leave us? Was the import surge a signal of trouble ahead or did the reported healthy growth of core GDP show that all is well?

The data discussed in the previous section suggest to me that the import components of fixed investment and consumption probably increased in Q1 relative to 2024. If so, that would mean that “true” core GDP did not, after all, grow by the reported 3 percent, but more likely by 1 percent or a little less, the reported rate minus the substantial part of the import surge that did not go directly into inventories. What would that imply for the future?

As a thought experiment, suppose that imports were to decrease in Q2 by the same amount that they increased in Q1, and that inventories were run back down by the same amount, with domestic consumers and businesses as their final purchasers. If so, reported core GDP would not change at all. Overall GDP would not change either – the contribution to growth of inventory investment would turn negative, but that would be exactly offset by the decrease in imports. (Remember that a decrease in imports makes a positive contribution to GDP growth.)

On the other hand, it is possible that some or all of the increased Q1 consumption and investment purchases were originally planned for later but moved up to get ahead of tariffs. If so, demand for those goods might well fall in Q2. Unless those consumers and businesses switched to buying equal quantities of different kinds of domestically produced goods in Q2, aggregate demand would falter, and slow GDP growth would continue. There are two reasons that might happen.

One is that firms and households might have financed their incremental purchasers by borrowing or running down their cash balances in Q1. If so, they might cut spending to restore their balance sheets to normal in Q2. A second possibility is that the prices of American goods might rise due to tariff-driven increases in input costs. Either would tend to slow real growth in Q2 and beyond.

All things considered, then, I do not find the reported 3 percent growth of Q1 core GDP to be all that reassuring. We will need more data to know for sure. Some will come from the second and third estimates of Q1 GDP that will come out in May and June. Those will include the first estimates of Q1 national income and savings, which will help interpret the strength of household and business balance sheets. Price and labor market information for May and beyond will provide further clues. We can also watch the Atlanta Fed’s GDP Now reports, which correctly predicted negative Q1 growth. Meanwhile, we can’t be too confident that everything is OK.

More By This Author:

Why A War On Trade Deficits Won't Reindustrialize America
Why Tariff Inflation Won't Be Just A Blip
Why Tariff Inflation, If It Comes, Won't Be Just A Blip