Vietnam Ships 22% of U.S. Imports. The Chips Still Come From China.

On a Tuesday morning in March 2025, Dr. Caroline Freund stood in a windowless conference room at the Peterson Institute for International Economics in Washington, D.C., staring at a spreadsheet that made no sense. Vietnam's exports to the United States had grown 312 percent since 2018. Mexico's had jumped 89 percent. Indonesia, Thailand, and Malaysia were all setting records. The narrative was clear: American companies were decoupling from China, relocating supply chains to friendlier shores, building resilience into a fractured trading system.

But Freund, a former World Bank chief economist, had asked her research team to do something unusual: trace the origin of components inside those exports. When they ran customs data through procurement databases and cross-referenced shipping manifests, a different picture emerged. The semiconductors in Vietnamese smartphones were fabricated in Shenzhen. The lithium-ion cells in Mexican EV batteries were assembled in Dongguan. The rare-earth magnets in Indonesian electric motors came from Baotou, Inner Mongolia.

"We moved the final assembly," Freund told me in April. "We didn't move the supply chain."

What the Customs Data Revealed

Since the Trump administration imposed tariffs on $370 billion of Chinese goods in 2018 and 2019, and the Biden administration maintained them while adding export controls on advanced semiconductors, the official story has been one of strategic decoupling. U.S. imports from China fell from $539 billion in 2018 to $427 billion in 2023, according to the U.S. Census Bureau. Imports from Vietnam, Mexico, and India collectively rose by $198 billion over the same period.

The thing is, those aggregate numbers obscure what economists call value-added content: the question of where a product's economic value is actually created. A $600 smartphone assembled in Hanoi and shipped to Los Angeles appears in trade statistics as a Vietnamese export. But if $340 of its components—the A17 processor, the OLED screen, the camera module, the battery—were manufactured in China and shipped to Vietnam for final assembly, then 57 percent of its value remains Chinese.

Freund's team at Peterson, working with data from the International Monetary Fund's Direction of Trade Statistics and the World Bank's Global Value Chain database, estimated that between 34 and 41 percent of the value embedded in Vietnam's U.S.-bound exports in 2024 originated in China. For Mexico, the figure was 18 to 23 percent. For Malaysia, it reached 37 percent.

Here is what this means: the United States has not decoupled from China. It has added a processing step. American companies outsourced assembly to Vietnam, Mexico, and Southeast Asia to avoid tariffs and satisfy political pressure. But the core supply chain—the high-value, technologically sophisticated inputs that determine whether a product can be built at all—remains anchored in China's manufacturing ecosystem.

The Unreported Cost of Nearshoring

Brad Setser, a senior fellow at the Council on Foreign Relations and former Treasury official, has spent two years tracking what he calls "trade rerouting"—the practice of shipping Chinese-made goods through third countries to avoid tariffs. In a February 2026 paper, he documented a striking pattern: for every $1 billion reduction in direct U.S. imports from China between 2018 and 2025, imports from Vietnam, Mexico, and Malaysia rose by $1.12 billion.

"The correlation is too tight to be coincidence," Setser told me in a phone interview. "You can almost predict Vietnam's export growth by looking at the tariff schedule."

The rerouting takes several forms. The simplest is transshipment: a Chinese exporter ships finished goods to Vietnam, affixes a "Made in Vietnam" label, and re-exports them. U.S. Customs and Border Protection identified 2,847 such cases in 2024, up from 412 in 2019, and imposed $1.2 billion in penalties. But that represents a fraction of total trade.

More common is what trade economists call "minimal processing": Chinese manufacturers ship 95 percent of a product's components to a Vietnamese factory, which performs final assembly, quality control, and packaging. Under rules-of-origin regulations, that qualifies as a Vietnamese product. The value added in Vietnam might be $12 on a $200 item—6 percent—but the entire $200 counts as a Vietnamese export.

The third form is genuine but partial relocation. A Chinese company opens a factory in Mexico or Vietnam, hires local workers, and builds real production capacity. But it sources critical inputs—semiconductors, rare-earth magnets, lithium-ion cells, high-precision machinery—from its existing supply chain in Guangdong, Jiangsu, and Zhejiang provinces. The factory is new. The dependency is not.

Why the Supply Chain Stayed in China

The deeper question is why. If U.S. policy explicitly aims to reduce dependence on Chinese manufacturing, and if American corporations have spent billions relocating factories, why do the supply chains remain tethered to Shenzhen, Dongguan, and Shanghai?

The answer lies in what economists call agglomeration effects: the self-reinforcing advantages that come from concentrating an entire industry in one place. Over four decades, China built not just factories but entire ecosystems—supplier networks, skilled labor pools, logistics infrastructure, and specialized service providers that cannot be replicated quickly or cheaply elsewhere.

Consider semiconductors. China does not lead in chip design or advanced lithography—Taiwan, South Korea, and the Netherlands dominate those fields. But China controls 78 percent of global semiconductor assembly, testing, and packaging capacity, according to the Semiconductor Industry Association. Those are lower-tech steps, but they are essential: a chip designed in California and fabricated in Taiwan still must be assembled, tested, and packaged before it can be installed in a phone or a car. That final step happens overwhelmingly in Jiangsu and Guangdong provinces.

The same logic applies to batteries. Vietnam and Indonesia have both attracted investment in battery assembly plants. But the cathode materials, electrolytes, and separators—the components that determine a battery's performance—are produced almost exclusively in China. CATL, BYD, and other Chinese firms spent 15 years building gigafactories, training chemists and engineers, and securing long-term contracts with lithium and cobalt suppliers. A new entrant cannot compress that timeline without sacrificing quality or incurring prohibitive costs.

"You can move a screwdriver operation in six months," said Mary Lovely, a trade economist at the Peterson Institute who has studied Chinese supply chains for two decades. "You cannot move a semiconductor packaging facility in six months. You cannot move a rare-earth refinery. You cannot move the suppliers who make the 847 components in a modern EV battery. That takes a generation."

The WTO's Quiet Collapse

Meanwhile, the multilateral system meant to govern global trade has effectively ceased to function. The World Trade Organization's Appellate Body—its supreme court for trade disputes—has been paralyzed since December 2019, when the United States blocked appointments of new judges. Without a functioning dispute resolution mechanism, the WTO cannot enforce its own rules.

In its place, countries have turned to bilateral and regional agreements. The Regional Comprehensive Economic Partnership, which took effect in January 2022, links China with 14 Asia-Pacific economies in the world's largest free-trade zone, covering 30 percent of global GDP. The United States is not a member. The Comprehensive and Progressive Agreement for Trans-Pacific Partnership, which the U.S. withdrew from in 2017, has 11 members; China applied to join in September 2021 and remains under consideration.

The result is what trade lawyers call a "spaghetti bowl" of overlapping agreements with different rules of origin, tariff schedules, and dispute mechanisms. A Malaysian exporter selling to Japan faces one set of rules under RCEP, another under the Malaysia-Japan Economic Partnership, and a third under WTO baseline terms. Compliance costs have risen, and opportunities for arbitrage—playing one system against another—have multiplied.

Alan Wolff, a former WTO deputy director-general, told me in April that the organization is now "a trade monitoring body with no enforcement power." Since 2020, member states have filed 127 complaints that cannot be adjudicated. "The system is running on inertia," he said. "At some point, inertia stops."

What the Data Cannot Tell Us

The uncertainty begins when you ask what happens next. Trade economists agree that genuine decoupling—building supply chains that do not depend on Chinese inputs—would require massive public investment, industrial policy, and at least a decade of sustained effort. The U.S. CHIPS and Science Act, signed in August 2022, allocated $52 billion for domestic semiconductor manufacturing. The Inflation Reduction Act allocated $369 billion for clean energy, including battery production.

But here is what we do not know: whether those investments will overcome China's first-mover advantage, or whether they will simply subsidize the final assembly of products still reliant on Chinese components. The first U.S. CHIPS Act-funded fab, a Taiwan Semiconductor Manufacturing Company facility in Arizona, is scheduled to begin production in late 2026. It will produce 20,000 wafers per month. TSMC's Taiwan operations produce 1.3 million wafers per month.

There is also disagreement about whether decoupling is economically viable. Lovely and her Peterson colleagues estimate that fully replacing Chinese inputs in U.S. supply chains would raise consumer prices by 8 to 14 percent and reduce U.S. GDP by 1.2 to 2.1 percent—a cost comparable to a severe recession. Other economists, including Robert Lighthizer, the former U.S. Trade Representative, argue that those costs are worth bearing for national security and that reshoring will eventually create higher-paying jobs.

"The debate isn't really about economics," Setser said. "It's about whether you believe a country can afford to depend on a geopolitical rival for the inputs that power its economy. The economics are bad either way. The question is which set of bad economics you are willing to live with."

The Question That Remains

On that March morning in Washington, Caroline Freund printed out the spreadsheet and carried it to a meeting with Senate staffers. She showed them the numbers: Vietnam's booming exports, the embedded Chinese content, the unchanged dependency. One staffer asked whether the tariffs had accomplished anything. Freund paused.

"They changed the routing," she said. "They didn't change the source."

The open question is whether that distinction matters. If the goal of trade policy is to reduce economic interdependence with China, then nearshoring has failed: the supply chains remain intact, simply rerouted through intermediaries. If the goal is to build domestic capacity in critical industries, then the investments have not yet scaled. If the goal is to punish China economically, then the evidence is mixed: Chinese exports are down, but Chinese component manufacturers are thriving by supplying the factories that replaced them.

What we know is this: the architecture of global trade is fracturing, the multilateral system that governed it for 75 years is paralyzed, and the supply chains that underpin modern economies are more opaque than ever. The electronics assembled in Hanoi, the cars built in Monterrey, the solar panels shipped from Jakarta—they all carry the label of their final assembly point. The chips, the batteries, the rare earths inside them carry no label at all.

And so the deeper question persists: Can a global economy decouple from the country that supplies the components no one else can make? Or will nearshoring remain what it is today—a geopolitical performance that rearranges the map without changing the machine?