The Soybean Weapon

No single commodity tells the story of U.S.-China trade more vividly than the soybean. In 1999, American food exports to China totaled $800 million. By 2012, they had reached $21 billion. By 2022, they peaked at $35 billion. A 44-fold increase in 23 years, driven almost entirely by a single legume.

The soybean’s rise followed China’s economic transformation. As 400 million Chinese citizens entered the middle class between 2000 and 2015, pork consumption roughly doubled. Pork is the foundation of Chinese cuisine — present in everything from dumplings to stir-fries to hot pot. But pigs are inefficient converters of feed to meat: it takes roughly three pounds of feed to produce one pound of pork. As China’s hog herd expanded to over 400 million animals, the demand for protein-rich feed exploded.

Soybeans are the most efficient source of high-protein animal feed on Earth. Crushed soybeans produce soybean meal (used for feed) and soybean oil (used for cooking). China had the crushing capacity but not the farmland. The United States had the farmland — specifically, the black soil of Iowa, Illinois, Indiana, Minnesota, and the Dakotas, where soybeans grow in rotation with corn across hundreds of millions of acres. By 2017, China was purchasing roughly 35–40 million metric tons of American soybeans per year, representing about 60% of total U.S. soybean exports.

The concentration was extreme. Iowa alone exported an estimated $3.6 billion in soybeans to China in 2017. For a state of 3.2 million people, China was the single most important customer for the most important crop. The same was true across the Midwest: soybean prices, farm income, and land values all reflected the Chinese appetite. When soybean futures traded on the Chicago Board of Trade, traders watched Chinese demand data as closely as they watched weather reports from Des Moines.

When the Trump administration imposed the first round of tariffs on Chinese goods in July 2018, Beijing needed a retaliatory weapon that would cause maximum political pain. The soybean was perfect. It was concentrated in politically important states (Iowa, Indiana, Ohio — all swing states). It was easily substitutable (Brazil, the world’s other major soybean exporter, could expand production within one growing season). And it was massive enough in dollar terms to send a clear signal.

China’s 25% tariff on American soybeans took effect on July 6, 2018 — the same day as America’s first tranche of tariffs on Chinese goods. The impact was almost instantaneous. Chinese buyers stopped placing orders for American soybeans. Shipments that were already at sea were rerouted or sat in port awaiting buyers. Soybean prices at the Chicago Board of Trade fell from $10.50 per bushel in May to $8.12 per bushel in September — a 23% decline in four months.

American food exports to China collapsed from $18.3 billion in 2017 to $8.2 billion in 2018 — a $10 billion decline, or 55%. The drop was almost entirely soybeans. Other food categories (tree nuts, seafood, pet food) were largely unaffected. The soybean tariff alone accounted for the vast majority of the export crash.

The substitution was immediate. Brazil’s soybean exports to China surged. Brazilian farmers, who had been expanding soybean acreage in the Cerrado region for years, suddenly found Chinese buyers competing aggressively for every available bushel. Brazilian soybean prices rose as Chinese demand shifted south. American prices fell. The spread between Brazilian and American soybean prices, normally narrow, widened to over $2 per bushel — meaning American soybeans were cheaper in absolute terms but unsaleable to their largest customer.

The political backlash was immediate. Iowa’s farmers were vocal, organized, and represented by powerful senators. Senator Chuck Grassley called the tariffs “a tax on Iowa.” The American Soybean Association warned of farm bankruptcies. Rural land values, which had been rising for a decade on the strength of Chinese demand, began to slip.

The Trump administration responded with the Market Facilitation Program (MFP), an emergency subsidy scheme that paid farmers directly for losses attributed to the trade war. The first round, announced in August 2018, allocated $12 billion. The second round, in 2019, added $16 billion. Total: $28 billion in farm subsidies over two years — more than double the annual value of the lost soybean trade, and the largest direct payment to farmers since the New Deal era.

The subsidies were distributed based on acreage and production, not actual losses, which created its own distortions. Large farms received the largest payments. An operation with 5,000 acres of soybeans in Illinois might receive $200,000–$400,000 in MFP payments — more than enough to offset the lower prices. Smaller farms and tenant farmers received proportionally less. The payments kept farms afloat financially but did not restore the Chinese market.

Critics pointed out the arithmetic: the $28 billion in subsidies significantly exceeded the $10 billion per year in lost soybean trade. American taxpayers were paying more to compensate farmers than the tariffs were generating in revenue from Chinese goods. The soybean tariff had, in effect, created a transfer payment from American consumers and taxpayers to American farmers — a politically palatable arrangement for the farm belt but an expensive one for the federal budget.

The Phase One trade deal, signed in January 2020, included China’s commitment to purchase $32 billion per year in American agricultural products — roughly double the pre-tariff level. The commitment was widely viewed as unrealistic, but China made a genuine effort, partly because it needed the soybeans and partly because African Swine Fever had devastated China’s domestic hog herd in 2019, requiring even more imported animal feed to rebuild.

The recovery exceeded the pre-tariff baseline. Food exports rose to $24.5 billion in 2020, $30.7 billion in 2021, and $35.1 billion in 2022. The 2022 figure was nearly double the pre-tariff peak — amplified by record commodity prices (soybeans touched $17 per bushel, up from $8 during the tariff trough) and a drought in Brazil that temporarily reduced China’s alternative supply.

But the recovery was fragile. As Brazilian production recovered and commodity prices normalized, American food exports to China declined to $27 billion in 2023 and $22.7 billion in 2024. The 2024 figure remains above the pre-tariff level of $18 billion — a genuine improvement. Yet the concentration risk hasn’t disappeared. American soybean farmers still depend on Chinese demand, and the 2018 experience demonstrated exactly how quickly that demand can be weaponized.

The soybean story carries three lessons that extend far beyond agriculture.

First: concentration creates leverage. When 60% of your export crop goes to a single customer, that customer has veto power over your price and your income. Iowa’s soybean farmers learned this in 2018 at a cost of $2 per bushel — roughly $5,000 per farmer on a typical 500-acre operation. The same principle applies to any sector with a dominant bilateral dependency: if China decided to stop purchasing American semiconductors or civilian aircraft, the impact on Intel, Qualcomm, and Boeing would follow a similar pattern.

Second: substitution is asymmetric. China replaced American soybeans with Brazilian ones within a single growing season. The physical product is identical — a soybean is a soybean regardless of origin. But American soybean farmers could not replace the Chinese customer with equivalent demand elsewhere. India, Europe, and Southeast Asia buy American soybeans, but at a fraction of China’s volume. The supplier with a commodity faces a structural disadvantage in a trade war against a buyer with alternatives.

Third: the bailout cost more than the tariff. The $28 billion in MFP subsidies exceeded the $10 billion annual value of the lost soybean trade. The government paid farmers more to offset the retaliation than the farmers had been earning from the exports in the first place. This is not an argument against tariffs — there may be valid national security or industrial policy reasons for accepting the cost — but it does mean the fiscal cost of agricultural retaliation is higher than it first appears.