A cargo ship is seen at the port of Oakland. Oakland, California, U.S., July 22, 2022.

Blog Name

Four years into the trade war, are the US and China decoupling?


US imports of some Chinese products have tanked. Others are higher than ever. How Trump’s selective trade war continues to matter.


Photo Credit: REUTERS/Carlos Barria

Yilin Wang and Julieta Contreras provided outstanding data assistance, and Nia Kitchin, Melina Kolb, and Oliver Ward assisted with graphics.


For many decades China and the United States have been locked in such a tight economic embrace that it is challenging to quantify whether, how, or why the embrace may be weakening.  Are the mounting tensions, bordering on hostility, between the two superpowers causing their economies to “decouple”?

Yes and no. On the one hand, US imports of certain products from China—including semiconductors, some IT hardware, and consumer electronics—have fallen dramatically. Even clothing, footwear, and furniture imports are down.

But on the other, imports from China of laptops and computer monitors, phones, video game consoles, and toys are higher than ever. Demand for these products surged in response to the COVID-19 pandemic. Stuck at home, Americans switched their spending away from services and toward many of these goods manufactured in China.

US and Chinese policymakers certainly seem determined to reduce the two countries’ economic interdependence, built over many decades but now buckling under the weight of their animosities. So far, the decoupling that is—and is not—occurring is partly the result of President Donald Trump’s trade war, the selective way it was waged, and the continuation of many of those policies under the Biden administration. A more recent motivating factor that may be spurring decoupling is the desire for increased diversification of imports to make supply chains for certain goods more resilient. Other drivers include human rights, democracy, and geopolitical concerns.

But the data also show something else. Even if policymakers foresee long-run benefits in disentangling the two economies, their choices come with immediate costs. These costs include product shortages, as supply chains struggle to adjust, as well as inflation, as companies find it expensive to establish new suppliers. Firms and ultimately consumers need to prepare to pay the price for the new policy-induced reality.

Total US imports from China have been down since the onset of the trade war

For 15 months beginning in July 2018, the Trump administration applied tariffs to more and more imports from China. Thus far, the Biden administration has chosen to keep those duties in place.

Overall the trade war has reduced US goods imports from China (figure 1).[1] Imports declined immediately after tariffs were imposed, falling further beginning in March 2020 as global trade collapsed in the wake of the COVID-19 pandemic, and have since recovered only slowly. Today, US imports from China (red line) remain well below the pre-trade war trend (dashed line), as defined (conservatively) by US imports from the world, and have only recently returned to pre-trade war levels of June 2018.[2] China is now the source of only 18 percent of total US goods imports, down from 22 percent at the onset of the trade war.

In comparison, current US imports from the rest of the world are 38 percent higher than pre-trade war levels and are even above trend (blue line). With a few exceptions, these imports were not hit with new US tariffs.[3] They have also recovered strongly following the onset of the pandemic.

After conducting an investigation under Section 301 of the Trade Act of 1974, the Trump administration began by imposing tariffs of 25 percent on products covering roughly $34 billion of US imports from China in July 2018 (List 1) and on $16 billion of imports in August (List 2).[4] When China retaliated, the trade war continued with Trump imposing 10 percent tariffs on an additional $200 billion of imports in September 2018 (List 3), increasing the rate of those duties to 25 percent in June 2019. In September 2019, Trump hit another $102 billion of imports (List 4A) with 15 percent tariffs, subsequently reducing them to 7.5 percent upon implementation of the US-China Phase One agreement in February 2020. (The administration identified another set of products covering most of the rest of US imports from China of more than $160 billion—List 4B—for which it scheduled tariffs to go into effect on December 15, 2019. It never imposed those duties and cancelled them upon the initial announcement of the Phase One agreement on December 13, 2019.)

Trump’s trade war tariffs affected imports in expected ways

As expected, the trade war has had the largest impact on imports from China of products hit with the highest US tariffs. US imports from China of goods currently facing a 25 percent duty (Lists 1, 2, and 3) remain 22 percent below pre-trade war levels (figure 2). US imports of those same products from the rest of the world are now 34 percent higher. US imports from China of products currently subject to 7.5 percent tariffs (List 4A) remain 3 percent below levels in August 2019 (right before imposition of tariffs on those products), whereas comparable imports from the rest of the world are now 45 percent higher.

Yet, US imports from China of certain products have surged. Imports of products never hit with trade war tariffs are now 50 percent higher than immediately prior to the trade war (see again figure 2). (US imports from the rest of the world of those same products are also up but by only 38 percent.) Products not facing tariffs made up roughly 33 percent of total US imports from China before the trade war and have grown to 47 percent today (see appendix table).

US imports are surging for some Chinese products never hit with tariffs

Beginning in 2020, COVID-19 lockdowns led many Americans to work, school, and play from home, which sharply increased demand for certain products, many of which were imported predominately from China. Laptops and computer monitors, phones, video game consoles, and toys are examples—combined they made up 21 percent of total US imports from China before the trade war, growing to 27 percent today (see appendix table). Coincidentally, the Trump administration had earlier decided to not apply trade war tariffs on these and other selected products. US imports from China of these goods have grown rapidly (figure 3, panel a). Though there is some evidence of diversification and changes to foreign sourcing in these products’ supply chains, US-China decoupling is not quite evident here.

Take laptops and computer monitors. Growth in US imports from both China and the rest of the world has been strong, with imports from each up roughly 50 percent since prior to the trade war (figure 3, panel b). Thus, shares of US imports have barely budged—China remains the source of 92 percent of US imports of laptops and computer monitors, with Taiwan and Vietnam each making up roughly 2 to 3 percent (panel c).

Figure 3 US imports from China of products not hit with any trade war tariffs have increased and show little evidence of decoupling

US imports of phones—including smartphones—have increased from both China (14 percent) and the rest of the world (70 percent). While declining in relative terms, 74 percent of US imports of phones are still sourced from China. Most noticeable is Vietnam’s increased share of US phone imports from below 10 to 22 percent. In 2019, Samsung closed its last mobile phone plant in China, moving assembly to lower-cost locations like Vietnam (and India), in part, because it was reportedly also losing out in the Chinese market to Chinese firms like Huawei and Xiaomi. But Vietnam’s gain in US imports has come not only at the expense of China. In 2019, LG also moved its smartphone assembly to Vietnam from South Korea. (Apple has also recently diversified some of its iPhone assembly by firms like Foxconn and Wistron away from China to India.)

Video game consoles tell a similar story. US imports from China have increased by 82 percent.  In 2019, however, Nintendo indicated it was moving some of its Switch assembly from China to Vietnam. (The New York Times reported some Microsoft Xbox consoles are now being shipped from Vietnam in addition to China.) Indeed, US imports of consoles from the rest of the world have grown five-fold since the onset of the trade war. Yet, because that growth started from such a low base, the United States still imports 90 percent of video game consoles from China.

US imports of toys like board games—e.g., chess, checkers, and backgammon—as well as playing cards from both China and the rest of the world are up considerably.  Vietnam has increased its share of US toy imports from 3 to 6 percent. But China remains the source of 83 percent of US imports of toys overall.

During the trade war, the Trump administration deliberately chose not to impose tariffs on these products—most of which were on List 4B—worried that, for such identifiable goods, consumers would suffer price increases and attribute them to the tariffs.[5] Trump said as much in August 2019, “What we’ve done is we’ve delayed [tariffs on List 4B] so they won’t be relevant in the Christmas shopping season…. Just in case they might have an impact on people.” Trump first postponed tariffs on those products until December 15, 2019—long after shipments would have arrived to stock shelves before the 2019 holidays—and then ultimately canceled them altogether upon announcement of the Phase One agreement.

Chinese products hit with 25 percent tariffs have struggled

At the other extreme are the products that Trump hit with 25 percent tariffs. This group is dominated by intermediate inputs and capital equipment—much less visible to households—that firms use to make other consumer goods or to provide services. US imports of these products are down overall (figure 4, panel a). Some are lower despite surging US demand during the pandemic, contributing to shortages and higher costs for firms needing those inputs to continue their operations. Such companies were forced to either continue importing from China even with the tariff or spend to establish relationships with new suppliers elsewhere.

Figure 4 US imports from China of products facing high trade war tariffs show some signs of decoupling

Take IT hardware and consumer electronics in higher demand with the pandemic lockdowns, such as network servers, modems, routers, as well as wireless headphones and smartwatches. US imports from China of these products are down 62 percent since the 25 percent tariffs were imposed, whereas imports from the rest of the world are now 60 percent higher (panel b).[6] China’s share of US imports of IT hardware and consumer electronics has been cut by nearly two-thirds, from 38 to 13 percent (panel c). Mexico is one sizeable alternate foreign supplier of such products. Taiwan has also substantially increased its share of the US import market.

Prior to the trade war, imported auto parts from China were a looming threat to the US industry. Imports from both China and the rest of world fell considerably during the early months of the pandemic, when the auto industry in North America suddenly halted production in response to the pandemic. US imports from the rest of the world have recovered and today are 20 percent higher, whereas imports from China subject to the tariffs have only just returned to pre-trade war levels. Nevertheless, China’s share of US auto parts imports has only dropped from 15 to 13 percent. Unsurprisingly, Mexico and Canada continue to dominate the US market for parts due to the integrated North American auto supply chain.

Social distancing during the pandemic also led households to increase spending on home furnishings. US imports of furniture from the rest of the world have grown by 87 percent; imports from China hit with tariffs remain 21 percent lower. (China’s share of US furniture imports has dropped from 57 to 36 percent.) Much of the new sourcing of US imports of furniture has come from Vietnam.

US imports of semiconductors from China cratered with US tariffs

Semiconductors are perhaps the most telling example of product shortages during the pandemic. The scarcity of chips led automakers to reduce their output in 2021, impacting employment through furloughs in a politically and economically important sector. (In figure 4, panels b and c, data for semiconductors alone are in volume, as opposed to dollar value, terms.)

US imports of semiconductors from China remain 26 percent lower than before the imposition of 25 percent tariffs (panel b).  Though imports have increased recently, as late as June 2021, US import volumes from China remained more than 50 percent below pre-trade war levels. Prior to July 2018, China had 47 percent of the US import market in volume terms (panel c). This share fell immediately after the tariffs were imposed, reaching only 39 percent today. Yet, import volumes from the rest of the world have expanded only by 5 percent (see again panel b).

This problem became clear when just one missing chip kept the manufacturing of products ranging from cars to refrigerators to washing machines from being finished. In volume terms, lost imports of semiconductors from China were not being fully replaced from elsewhere.

One reason was that production could not be substituted between Chinese and other chipmakers. China’s foundries specialize in “more mature nodes,” producing high volumes of “legacy” chips for low profit margins. Leading foreign firms like Taiwan Semiconductor Manufacturing Company (TSMC) or South Korea’s Samsung manufacture more advanced (and profitable) semiconductors and both did not have idle capacity or an interest in switching to less profitable products. This likely explains why the US semiconductor industry also did not significantly expand production, in addition to the fact that it was running at close to historical levels of its capacity utilization rate. Given that legacy chips are not particularly profitable to manufacture, and if the United States does not want to import them from China,[7] then who will produce them? That is the question facing America’s industrial consumers—like the auto sector—of large volumes of legacy chips.

In dollar terms as well, US chip imports from China remain 22 percent below their pre-trade war levels (not shown in figure 4). However, import values from the rest of the world are up 32 percent, mostly due to the price increase caused by heightened US demand for chips, as opposed to increased volumes offsetting lost imports from China.

To summarize, in each of these four examples of products hit with 25 percent US tariffs, US imports from China declined. Reduced imports were sometimes offset by imports from other foreign sources, but not yet in other cases. Yet, these examples were not unique. In value terms, US imports from China of all other products subject to 25 percent tariffs have fallen by 17 percent, even though imports of those goods from the rest of the world are now 33 percent higher (not shown).

US imports of Chinese products subject to lower tariffs have been uneven

A final set of products, covering 20 percent of US imports from China at the onset of the trade war, were initially hit with 15 percent tariffs in September 2019, which were then reduced to 7.5 percent in February 2020. Overall, imports from China of these goods have only recently returned to their pre-trade war levels (figure 5, panel a). The smaller negative impact on these US imports was partly because the tariffs were imposed at lower rates at the outset, imposed later in the trade war, and subsequently reduced. Imports of these products from elsewhere are now 51 percent higher.

Figure 5 Products facing lower trade war tariffs show mixed signs of decoupling

Clothing and footwear are one example, making up about 7 percent of total US imports from China before the trade war.[8] They remain 11 percent below pre-trade war levels, whereas US imports from the rest of the world are 44 percent higher (panel b). The tariffs may have accelerated an ongoing shift in production of some of these goods out of China’s market, as rising labor costs associated with China’s economic development were already moving such industries elsewhere. China’s share of US clothing and footwear imports has fallen from 34 to 24 percent, while the shares of Vietnam and Bangladesh have increased (panel c).

Personal protective equipment (PPE) and related COVID-19 medical products are a slightly different story. China was the source of roughly 50 percent of US imports of many such products before the trade war. That dipped once the tariffs were imposed in late 2019—jeopardizing the preparedness of the US health care system in the face of a pandemic in early 2020. Eventually, the Trump administration did exclude such products from its trade war tariffs, and by April 2020, US imports from China had resumed, before exploding over much of the rest of 2020, given the increase in US demand. (Price increases were the dominant cause of the sharp increase in values, though volumes increased as well.)

Exercise equipment and lithium batteries are two additional products where imports from China have jumped considerably, despite the trade war tariffs.[9] For both, China is also the source of more than 50 percent of the US import market (panel c). The boom for products like exercise cycles, rowing machines, and treadmills may have been temporarily driven by the pandemic and the inability to access private gyms. The growth in lithium battery imports is partly due to the recent increased US demand for electric vehicles (EVs). However, this growth too may fade over time if the new supply chain sourcing requirements found in the August 2022 Inflation Reduction Act—explicitly offering subsidies for automakers that diversify their EV battery supply chains out of China—are successful.

These four examples show the additional difficulties of attempting to assess potential decoupling of US imports from China. This complexity is expected, given the size and diversity of the Chinese economy and its involvement in so many different types of products. (As for the other products on List 4A not shown, US imports from China remain 10 percent below pre-trade war levels, while imports from elsewhere are 50 percent higher.)

Benefits of any US-China decoupling come with costs

Numerous studies have documented the negative impact of the trade war tariffs on the US economy.[10] Tariffs have hurt US manufacturing output, employment, and exports.[11] While those duties may not have caused the inflation pressure that began in 2021, American importers did bear the costs of the tariffs, in the form of higher prices, when they were imposed beginning in 2018-19. (There is little evidence that the tariffs led Chinese exporters to reduce their prices to sell to US consumers.)

The US tariffs that remain in place continue to impede American companies’ access to imports. The tariffs make those companies less attractive customers for Chinese firms, some of which may have had limited supplies to sell to other customers globally. Higher costs hurt American firms’ competitiveness in the US and international markets, relative to their non-Chinese competitors operating outside the United States.

Interpret the evidence with caution

For some products, the evidence here shows the United States increasingly sourcing imports assembled in countries other than China. It is partly the result of “trade diversion”—i.e., the United States now buys more expensive imports from third countries that it once bought but no longer buys from China because of the tariffs. The changes in imports shown here are consistent with other evidence that countries like Vietnam, as well as others in East and South Asia, are now trading more, including with the United States, in response to the US-China trade war.[12]

However, policymakers seeking to achieve “decoupling” need to carefully interpret the evidence documented here. Answers to the most important questions are still unknown.

US tariffs are not the only “cause” of the United States importing less from China. Some labor-intensive production closely associated with much of the clothing and footwear industry was likely relocating anyway, following a trend that was visible even before the trade war. China was losing competitiveness in this industry, relative to other emerging economies, as local wages have increased. (For other products, Vietnam may be rising as a source at the expense of other higher-income countries, such as South Korea.)

The full implications of any “movement” of economic activity that the data reveal also remain imperfect. For example, companies may be adding a separate assembly facility in Vietnam to service US consumers without having to pay the trade war tariffs. The same firms may also be keeping their Chinese facilities to continue to manufacture for the Chinese market as well as for other countries that have not imposed new tariffs on imports from China.

Such redundant investments may have complex and offsetting effects. On the one hand, such investments could improve resiliency if the diversification is useful. If the original US import arrangement involved single sourcing through concentrated suppliers in China, future buyers may find that adding non-Chinese assembly facilities lowers the risk of geographically concentrated disruptions due to climate change (floods, droughts, wildfires), health (pandemics), or geopolitics (military conflict).

On the other hand, redundant investments come with higher costs. There is the initial, one-time expenditure of establishing the new assembly plant. But there may also be additional (and ongoing) costs associated with operating two supply chains, each on a smaller scale than previously when it was all being done in China.[13]

Lastly, these data at most reveal changes only in the final assembly facility that is the source of US imports of a good. Precious little is yet known about any changes to the value-added content of that good the United States is importing. As an extreme example, suppose the final assembly of a consumer electronic product moves from China to Vietnam. The workers involved in the final assembly would change, but if the product continues to derive the same amount of critical intermediate inputs from Chinese suppliers, who now ship to Vietnam for final assembly by, say, the subsidiary of a Chinese-headquartered firm, then how much is really different?

Policymakers therefore need to interpret even this preliminary evidence of some US-China “decoupling” with extreme caution. Policy decisions made today to reduce economic interdependence between the two countries will have profound implications for both economies, and neither will escape unscathed.

Appendix table: US imports from China subject to Section 301 tariffs, by product list (percent of total imports from China)
 Product grouping Percent of total imports from China
Year prior to trade war Most recent year
(July 2017-June 2018) (September 2021-August 2022)
Not subject to tariffs
Total 33 47
Laptops and monitors 8 11
Phones, including smartphones 9 9
Video game consoles 1 2
Toys 3 5
Others not subject to tariffs 12 20
Subject to 25 percent tariff (Lists 1, 2, or 3)
Total 47 34
Selected IT hardware and consumer electronics including data servers, modems, routers, wireless headphones, and smartwatches* 5 2
Auto parts 2 2
Furniture 6 4
Semiconductors 1 1
Others on Lists 1, 2, or 3 33 26
Subject to 7.5 percent tariff (List 4A)
Total 20 19
Clothing and footwear 7 6
Personal protective equipment (PPE) and COVID-19 products** 1 2
Exercise equipment 1 1
Lithium batteries, including for electric vehicles 0 1
Others on List 4A 11 9
Notes: *Smartwatches were technically on List 4A but were not created as a separate 10-digit Harmonized Tariff Schedule (HTS) code until September 2018. For data prior to the trade war they were included with products on List 3 that were subject to 25 percent tariffs. **Subject to exclusions from tariffs. The Excel file accompanying this blog post provides the exact HTS codes associated with each product grouping. Numbers may not sum to total due to rounding.


1. All figures are based on the previous 12-month sum (“trailing sums”) of the value of imports, which helps eliminate seasonality concerns. The exception is semiconductors in figure 4, which reports data of the volume of imports.

2. The approach adopted here is to report indexed data at each month of the cumulative imports of the previous 12 months. Unless noted otherwise, “pre-trade war levels” throughout refer to levels as of June 2018 and “current” imports (or “today” or “now”) refer to imports as of August 2022, the most recently available data. The pre-trade war trend in figure 1 is defined by US imports from the world from August 2016 to June 2018, as August 2016 was the (local) trough of US imports (not shown).

3. Exceptions include imports from some third countries of solar panels, washing machines, steel, and aluminum, as well as some products investigated under antidumping and countervailing duty laws, which did face additional US protection during this period.

4. For ease of exposition, this paragraph refers to US government-announced level of US imports from China for each product list. For example, List 3 was announced as covering $200 billion of US imports even though the product-level data revealed List 3 as covering roughly $184 billion of US imports in 2017.

5. While not shown, US imports from China of other products not hit with any trade war tariffs are up 71 percent while imports of the same from the rest of the world are up only 36 percent.

6. Smartwatches are technically on List 4A, subject to a 15 percent and then 7.5 percent tariff, and ideally would be in the next section. They are also large, covering 88 percent of the value of imports in the Harmonized Tariff Schedule (HTS) 8-digit code (as of today). However, before September 2018, smartwatches were in the same HTS category as other electronic products on List 3 and are thus combined here so as to make comparisons with pre-trade war levels. Thus the data in figure 4, panel a, are a slight overestimate of the products covered by 25 percent tariffs and the data in figure 5, panel a, are an underestimate of products covered by 7.5 percent tariffs.

7. US taxpayers may partially cover the cost of such US production through the funding made available in the CHIPS and Science Act signed into law in August 2022.

8. These are clothing and footwear products on List 4A. Clothing and footwear products on List 1, 2, or 3 or not subject to any Section 301 tariffs are not shown here. Roughly 90 percent of US imports from China of clothing and footwear before the trade war were on List 4A.

9. For the most part, products in these categories were not granted tariff exclusions for lengthy periods of time.

10. See Mary Amiti, Stephen Redding, and David Weinstein. 2019. The Impact of the 2018 Tariffs on Prices and Welfare. Journal of Economic Perspectives 33, no. 4: 187–210; Pablo D. Fajgelbaum, Pinelopi K. Goldberg, Patrick J. Kennedy, and Amit K. Khandelwal. 2019. The Return to Protectionism, Quarterly Journal of Economics 135, no. 1: 1–55; Alberto Cavallo, Gita Gopinath, Brent Neiman, and Jenny Tang. 2021. Tariff Pass-Through at the Border and at the Store: Evidence from US Trade Policy, American Economic Review: Insights 3, no. 1: 19–34. For a survey, see Pablo D. Fajgelbaum and Amit K. Khandelwal. Forthcoming. The Economic Impacts of the US–China Trade War. Annual Review of Economics.

11. See Kyle Handley, Fariha Kamal, and Ryan Monarch. 2020. Rising Import Tariffs, Falling Export Growth: When Modern Supply Chains Meet Old-Style Protectionism. NBER Working Paper 26611; and Aaron Flaaen and Justin R. Pierce. 2019. Disentangling the Effects of the 2018–2019 Tariffs on a Globally Connected US Manufacturing Sector. FEDS Working Paper 2019-086.

12. See Pablo D. Fajgelbaum, Pinelopi K. Goldberg, Patrick J. Kennedy, Amit Khandelwal, and Daria Taglioni. 2021. The US-China Trade War and Global Reallocations. NBER Working Paper 29562, December.

13. Related tradeoffs are explored in Gene M. Grossman, Elhanan Helpman, and Hugo Lhuillier. 2021. Supply Chain Resilience: Should Policy Promote Diversification or Reshoring? NBER Working Paper 29330, October.

More From