Why do econ journalists keep making this basic mistake?

Imports don't subtract from GDP. Stop saying they do!

Economics journalists, like any writers, aren’t perfect. Perhaps in a previous age, people thought that everything they read in the news was exactly true; perhaps some still do. But reporting is a human activity, and humans make mistakes. In order to get the true story, you have to read multiple sources, and be skeptical of what you read — and even then, mistakes will slip through.

So the purpose of this post isn’t for me to be a pedantic know-it-all, or to insult the economics journalism profession as a whole, or to call out specific writers. But there’s one simple, elementary mistake that almost all econ reporters make very consistently, over and over. And unlike most mistakes, this one has probably had serious negative consequences for American economic policymaking. Therefore I feel like I have to speak up here, and express my frustration a little.

The mistake econ reporters are making is claiming that imports subtract from GDP. Imports do not subtract from GDP. And yet again and again, economics journalists say that they do.

This week, U.S. GDP data for the first quarter of 2025 (January through March) was released. The data showed that the U.S. economy shrank at an annualized rate of 0.3%. But almost every economics journalist and columnist reported that this decline was due to a surge of imports, as American companies rushed to stock up on foreign-made goods ahead of Trump’s tariffs.

For example, here is the Wall Street Journal:

[T]he main driver of the first-quarter contraction was Trump’s trade war…Businesses rushed to get ahead of tariffs…Imports rose at the fastest pace since the third quarter of 2020…Imports subtract from the Commerce Department’s calculation of GDP, since they represent spending on foreign-made goods and services. [emphasis mine]

And here is Bloomberg:

The GDP figures showed imports surged an annualized 41.3% — the biggest advance in nearly five years. Because these goods and services aren’t produced in the US, they are subtracted from GDP. [emphasis mine]

And here is CNBC:

Imports subtract from GDP, so the contraction in growth may not be viewed as negatively given the potential for the trend to reverse in subsequent quarters. [emphasis mine]

And this is the Washington Post:

This economic slowdown came primarily from a dramatic increase in imports — which count against GDP — as businesses rushed to purchase foreign goods ahead of President Donald Trump’s promised tariffs. [emphasis mine]

All of these writers — and many, many more that I didn’t cite here — got it wrong. Imports simply do not subtract from GDP. That is not how GDP works.

Last September, I wrote a post explaining why imports don’t subtract from GDP. I don’t want to just write that post all over again, so here it is:

Once again: Imports do not subtract from GDP

The short version of the story is this: GDP is a measurement of everything produced within a country’s borders. Imports are produced outside a country’s borders. So imports don’t add to or subtract from GDP. Imports simply aren’t counted in GDP at all.

Let’s think about some examples. Suppose an American buys a TV made in China for $1000. Remember that GDP can be calculated as the sum of consumption, investment, government purchases, and net exports:

GDP = Consumption + Investment + Government Purchases + Net Exports

When the American buys the $1000 TV from China, U.S. consumption goes up by $1000. And U.S. net exports go down by $1000, since “net exports” means exports minus imports. The increase in consumption exactly cancels out the fall in net exports. So the total contribution of the imported TV to U.S. GDP is zero.

Let’s take another example, which is more like what actually happened in Q1. Suppose an American company, Best Buy, decides to buy a Chinese TV and put it in a warehouse, because it knows that tariffs are coming soon. That purchase counts as inventory investment. So investment goes up by $1000. And just like in the previous example, net exports go down by $1000. The two cancel out, and the total contribution of the imported TV to U.S. GDP is zero.

In other words, if a bunch of U.S. companies were trying to stock up on imports ahead of the tariffs, that should register as both an increase in (inventory) investment, and as a decrease in net exports. The two should exactly cancel out. Those imports should not subtract from GDP, since they add to investment even as they also subtract from net exports.

Here’s a simple analogy: Does putting on shoes make you lose weight? No, it doesn’t. And yet when you weigh yourself with your shoes on at the doctor’s office, and you want to know your actual body weight, you subtract the weight of your shoes afterwards. Imports are to GDP what shoes are to your weight on the scale at the doctor’s office — just something superfluous that gets added in for the sake of measurement convenience, and which has to be netted out again later to get the true number.

Putting on heavier shoes doesn’t make you a thinner person, and importing more goods from abroad doesn’t make your country’s economy any smaller.

So it’s just not true that America’s economy shrank in Q1 because “imports are subtracted from GDP”. That’s false. Every time an economics reporter writes those words, it’s an error that should be corrected.

(I should note that The Economist, alone out of all the major news outlets I looked at, got the story completely correct. Good job, guys!)

Why does almost everyone keep getting this wrong? And why does it matter?

Why does almost every economics writer get this wrong? The simplest answer is herd behavior.

Back when IBM was the biggest, most important tech company, there was a saying in stock trading: “Nobody ever gets fired for buying IBM.” Similarly, practically everyone in econ journalism writes “Imports subtract from GDP,” so if you write that too, no one is going to give you grief about it. There’s safety in numbers; you won’t be singled out.

It’s easy to trace the origins of the mistake, too. The basic formula for GDP, which government agencies report and every econ writer knows, is just GDP = C + I + G + NX. So when government agencies report this breakdown, it’s pretty much automatic for econ reporters to write “Consumption increased by this much, investment by that much”, and so on. So if the trade deficit widens, it’s just standard practice to say “Net exports contributed negatively to GDP this quarter.” You’re just dutifully reading off each of the standard components.

And because net exports are just exports minus imports, it’s the tiniest of hops from saying “Net exports subtracted 5% from GDP growth” to saying “Imports subtracted 5% from GDP growth”. And if you say it that way, it must mean imports subtract from GDP, right? Right?

In fact, I can’t entirely blame econ journalists for making this mistake, because sometimes the government makes it too!

Most of the time, the people working for the government get it right. For example, this is from the St. Louis Fed in 2018:

When the Bureau of Economic Analysis (BEA) measures economic output, it categorizes spending with the National Income and Product Accounts (NIPA). Some of this spending, which is counted as C, I, and G, is spent on imported goods. As such, the value of imports must be subtracted to ensure that only spending on domestic goods is measured in GDP. For example, $30,000 spent on an imported car is counted as a personal consumption expenditure (C), but then the $30,000 is subtracted as an import (M) to ensure that only the value of domestic production is counted…[T]he purchase of imported goods and services has no direct impact on GDP.

Yes! The BEA’s explainer from 2015 also gives the correct explanation.

Except not everyone who works for the government gets this right. Whoever wrote the most recent data release for the Bureau of Economic Analysis got it completely wrong:

The decrease in real GDP in the first quarter primarily reflected an increase in imports, which are a subtraction in the calculation of GDP. [emphasis mine]

And they included this misleading chart, with the mistake repeated in the captions below the figure:

Source: BEA (sigh)

Why did whoever wrote the release for the BEA get this so wrong? I don’t know. But if the government is putting out data releases saying “Imports subtract from GDP”, you can hardly expect the average econ journalist not to do it, can you? Ultimately, the buck stops with the BEA; it’s their job to educate reporters about how to report on the data they release.

Anyway, some econ writers are sort of tacking in the direction of getting things right, but aren’t quite there yet.¹ Here’s Jonathan Levin of Bloomberg Opinion:

The latest numbers were overwhelmingly dragged lower by a surge in imports…GDP math seeks to measure the total production within a country’s borders, so statisticians add exports and subtract imports[.]

This is self-contradictory. If GDP math measures the total production within a country’s borders, why should a surge in imports drag GDP lower?

And here’s Ben Casselman in the New York Times:

[T]he decline in G.D.P. in the first quarter was driven almost entirely by a huge increase in imports as consumers and businesses tried to front-run Mr. Trump’s tariffs. That surge shaved nearly five percentage points off G.D.P. growth in the first quarter…

G.D.P….is meant to measure only goods produced domestically, not imports, which are produced abroad. But rather than measure production directly, the government counts all the goods and services sold in the country, and then subtracts the ones that were made overseas. (It also adds in exports, which are produced domestically but sold to foreign buyers.)…That means that, in theory, imports neither add nor subtract from G.D.P. Anything that is imported to the country should show up elsewhere in the quarterly data either as consumer spending or as an unsold product held in inventory — both of which are counted as additions to G.D.P.

The second paragraph here gets it right, but it directly contradicts the first paragraph. If imports aren’t counted in GDP, then how could a surge in imports “[shave] nearly five percentage points off G.D.P. growth in the first quarter”? It cannot.

Anyway, I asked AI this question, just to see if all the people getting it wrong had poisoned the training data. Google’s AI search assistant got it right:

But OpenAI’s top-of-the-line o3 model gave a confused, muddled answer that contradicted itself:

So there’s still a lot of work to be done in terms of expunging this common error from the world of econ reporting, but I do see some hopeful signs of progress.

The next question you may ask is: Why does any of this matter? As an angry graduate student thirteen years ago, I would have written this post out of pure pedantry, but I’m long past that. The only reason I care so strongly about this is that this common error seems to be influencing national policy in a fairly disastrous way.

The St. Louis Fed’s explainer on imports and GDP warned us about this:

[T]he [GDP] equation, as [typically] stated, can lead to a misunderstanding of how imports affect GDP. More specifically, the expenditure equation seems to imply that imports reduce economic output. For example, in nearly every quarter since 1976, net exports (X − M) have been negative…which seems to imply that trade reduces domestic output and growth…This can influence people’s perspective on trade.

Well, yes. Trump’s top economic advisor, Peter Navarro, claims that imports subtract from U.S. GDP. Trump doesn’t speak in such clear terms, but routinely says that countries that run trade surpluses with the U.S. are “ripping us off.”

It’s based on this very idea that Trump and Navarro have saddled our country with boneheaded tariffs that are going to hurt the American economy. And when MAGA types want to defend the tariffs, they often trot out the mistaken idea that trade deficits make our country poorer.

If econ reporters hadn’t continuously said that “imports subtract from GDP” for decades on end, this mistaken idea might not have embedded itself so strongly in the MAGA people’s heads. The tariffs are based, at least in part, on a simple accounting mistake. The least econ writers could do is to not reinforce this mistake every time a new quarterly GDP release comes out.

So what’s really going on with imports and GDP?

At this point, if you’re an econ writer, you might be reading this and asking: “So, what should I report about imports and GDP?” It’s a fair question. When you see imports surge and GDP drop, you get the strong inkling that there’s some kind of important relationship between the two. So what is it?

In fact, there are three plausible candidates for what’s going on there:

• Measurement error

• Companies diverting resources from domestic purchases to import stockpiling

• Imports clouding up the forecasting picture

Let’s talk briefly about each of these.

First, measurement error. A lot of people are arguing that a last-minute surge of imports is failing to show up in inventory investment or consumption yet, so that it looks like imports are subtracting from GDP, even if they’re really not. For example, here’s Ben Casselman:

In practice…the government is good at counting both imports and consumer spending, but often must rely on rough estimates for inventories, especially in preliminary data. The first quarter figures showed only a modest increase in inventories, despite anecdotal reports of companies stockpiling products and materials ahead of tariffs.

And here’s The Economist:

Why, then, did GDP fall? It may have been because of measurement problems. Inventories and consumption are harder to keep track of than imports… The Census Bureau tracks trade as it flows through customs; it estimates consumption and inventories using less precise sources, including voluntary surveys. Mr Trump’s “Liberation Day” tariffs loomed right at the end of the quarter. Perhaps a last minute rush to beat the deadline was picked up at the border but nowhere else. If so, then the GDP figure is likely eventually to be revised up.

In fact, it’s true that inventory investment is typically subject to large revisions in later months, and this month may be especially wacky. So we may eventually find out that the U.S. economy did better in Q1 than we initially thought.

Second, even though imports and GDP have no accounting relationship, it’s possible for them to have a behavioral relationship — which is a very different thing. The economic forces that cause a surge in imports could also cause domestic production to fall at the same time.

Suppose American companies wanted to stock up on foreign goods in Q1 before Trump’s tariffs hit. They might do this by temporarily stocking up on fewer domestically made goods that quarter. Or they might install foreign equipment, and pause their capital expenditures on domestically produced equipment. Either way, this would mean that investment in domestically produced stuff would fall. And that would reduce GDP that quarter.

This is Joey Politano’s hypothesis:

[I]t’s important to clarify that [an] import surge does not itself shrink the economy—GDP is an estimate of domestic production, and is not affected by purchases from abroad…consumers rushing out to buy foreign-made clothes will show up as an increase in imports (a negative “contribution” to GDP) and an increase in consumption (an equal-and-opposite positive “contribution” to GDP) and have no net effect on economic growth.

In [the first quarter of 2025], however, the surge in imports primarily manifested as a surge in business inventories—it was corporations, not consumers, doing most of the stockpiling. The positive contribution from inventory growth was the highest since late 2021, “offsetting” a bit less than half the jump in the trade deficit. Fixed investment of foreign equipment also helped significantly “offset” the increase in imports at the start of this year…

If a construction company rushed to buy a bunch of wood from Canada ahead of tariffs, that would show up as an increase in imports and an increase in inventories which cancel each other out. But if that company slowed down construction so it could afford to stockpile inputs, that would show up as a hidden drag on GDP in the form of lower investment. In other words, tariff front-running can indirectly slow the economy.

Joey is exactly right, and I think this is the most plausible theory for why GDP shrank in the first quarter. The Economist also mentions this possibility.

Note that this isn’t typical behavior, and — as Joey points out — it can’t explain all of the slowdown in Q1. So it’s still not right to say “imports subtract from GDP”. Behavioral effects and accounting identities are very different things.

Finally, Q1 is just one quarter. What we really probably want to know is whether there’s a recession coming, or whether this GDP decline was a one-off. Jason Furman likes to use an alternative measure called Final Sales to Private Domestic Purchasers — basically, consumption plus investment — to forecast future growth.

FSPDP takes GDP, removes government spending, removes exports, and then adds imports in. It’s just consumption plus investment — both of which include imports. In other words, the surge in imports this quarter might predict higher GDP growth in the quarters to come.

On the other hand, the reason FSPDP is a good forecaster of the future economy is that it’s basically just an indicator of underlying aggregate demand. If Trump’s tariffs cause a supply shock instead, FSPDP might fail to forecast the recession that results.

So anyway, here are three useful and informative things econ reporters can say about the effects of imports on the Q1 GDP data:

1. “Inventory investment is slow to measure, so we might see big upward revisions to Q1 GDP, when the last-minute surge in imports finally gets reflected in inventories.”

2. “Companies eager to stockpile imported goods ahead of Trump’s tariffs probably diverted spending away from domestically produced goods and services, which hurt GDP.”

3. “Strong imports are typically a sign of strong underlying demand in the economy, so we could see the economy bounce back strongly in three months.”

These ideas are all somewhat theoretical, but they’re all on a lot more solid ground than “Imports subtract from GDP”, which is just plain wrong.

---

1. Richard Feynman remarked on a similar phenomenon among early 20th century physicists, who were trying to measure the charge of the electron:

Millikan measured the charge on an electron by an experiment with falling oil drops and got an answer which we now know not to be quite right. It’s a little bit off, because he had the incorrect value for the viscosity of air. It’s interesting to look at the history of measurements of the charge of the electron, after Millikan. If you plot them as a function of time, you find that one is a little bigger than Millikan’s, and the next one’s a little bit bigger than that, and the next one’s a little bit bigger than that, until finally they settle down to a number which is higher…Why didn’t they discover that the new number was higher right away? It’s a thing that scientists are ashamed of—this history—because it’s apparent that people did things like this: When they got a number that was too high above Millikan’s, they thought something must be wrong—and they would look for and find a reason why something might be wrong. When they got a number closer to Millikan’s value they didn’t look so hard. And so they eliminated the numbers that were too far off, and did other things like that.

When the standard consensus is wrong, people realize it’s wrong, but they tend not to break radically from that consensus; instead, they take small steps in the direction of the right answer.