Why the United States Is Losing the Tech War With China

Editor’s Note: The United States has taken a series of steps designed to limit China’s access to semiconductors, particularly advanced ones, as part of a broader effort to slow the rise of Chinese power. The University of Connecticut’s Miles Evers argues that these measures are probably doomed to fail, resisted by U.S. businesses and countered by Chinese efforts to build their own semiconductor capacity.

Daniel Byman

***

Semiconductors—or “computer chips”—lie at the heart of the competition between the United States and China. They are one of the most critical sectors of the modern economy and an irreplaceable input in almost every modern technology, from washing machines to missile guidance systems. For this reason, having reliable and affordable access to semiconductors can shape a country’s economic and military power. China is a prime example. Its meteoric rise over the past 20 years is due mainly to the country’s high concentration of global electronics production. Chinese firms like Huawei and ZTE have led the way in quantum computing, artificial intelligence, and 5G technologies.

Over the past several years, the United States has sought to restrict China’s access to semiconductor technologies. Traditionally, Washington only ever limited the export of cutting-edge chips that might have military applications, believing that the United States could “run faster” than China through innovation and free trade. However, after China released its “Made in China” strategy in 2015, concern spread that China’s military and economic power was growing faster than expected and that it was using U.S. commercial technologies to do so. Consequently, the United States needed to “maintain as large of a lead as possible” in semiconductor technologies. This meant restricting sales of advanced semiconductor technologies to Chinese businesses, regardless of their civilian or military applications, and bolstering the domestic semiconductor industry through the $280 billion CHIPS and Science Act.

In a recent article, “Wars without Gun Smoke,” Ling Chen and I explore the strategic logic behind U.S. technology policy toward China and its likelihood for success. We show that, like in past power transitions between rising and declining powers, the United States is waging a preventive economic war to stop China from becoming a peer competitor. However, the position of American businesses in the global value chains (GVCs) for semiconductors puts the United States at a disadvantage in this war. Ultimately, we find the United States has started a fight that it seems destined to lose. 

The Global Value Chain for Semiconductors

Like most commodities, semiconductors are produced through global value chains, in which the stages of the production process are spread among firms in different countries. For semiconductors, the chain typically begins with “fabless” chipmakers in the United States who design the chip’s architecture using electronic automation software. These designers then outsource manufacturing to “foundries” in Taiwan, Japan, or South Korea that fabricate the semiconductor from silicon wafers and other raw materials. The chips are sent to Chinese firms for assembly, testing, and packing into a final product like a cellphone or laptop. Chains like this are generally of net benefit to the world. They promote specialization, reduce costs, and ultimately increase profits.

Each firm that participates in a GVC contributes some “value” to the final price of an item, albeit to varying degrees. High-value businesses specialize in knowledge-intensive activities that involve steep barriers to entry (e.g., research and development, branding, marketing, logistics, and financing). As such, they tend to govern the GVC and reap the most profits. In contrast, low-value businesses specialize in labor-intensive activities that involve fewer barriers to entry and thin profit margins (e.g., raw material extraction, manufacturing, and assembly). In the case of semiconductors, for instance, American “fabless” designers capture 54 percent of the value added in the production process and make almost five times as much in revenue compared to Chinese businesses specializing in packaging, assembly, and testing.

Notably, the structure of GVCs tends to reinforce the distribution of power between states. For development reasons, dominant states like the United States typically have jurisdiction over high-value firms, whereas rising powers like China usually have jurisdiction over low-value firms. For instance, a handful of American chip designers—especially Intel, Nvidia, and Qualcomm—sell almost 50 percent of the world’s semiconductors. Chinese manufacturers buy nearly 60 percent of the global supply of semiconductors. As such, the United States has more regulatory opportunities than China to influence prices and product standards.

For this reason, dominant states can “weaponize” GVCs against rising powers by using export controls preventively to choke off their access to GVCs and slow their rise. The Trump administration realized this in 2019 when it added China’s most prominent electronics manufacturer, Huawei, as well as hundreds of its affiliates, to its “Entity List” of companies barred from purchasing semiconductors that rely on U.S. designs, software, equipment, and components, even if the products were manufactured outside the United States. The Biden administration has since expanded this policy to cover the sale of advanced semiconductors to the entire Chinese tech industry. The measures effectively cut China off from the GVC for semiconductors and the inputs necessary to develop its own advanced semiconductors. 

Will the United States Succeed?

The United States is unlikely to win the tech war. There are two reasons for this conclusion.

First, the empirical record consistently shows that rising powers don’t sit idly by when dominant states disrupt their access to critical resources. Typically, they respond by subsidizing industrial development, pushing their businesses to upgrade into high-value positions to become self-sufficient. Chinese President Xi Jinping has urged the country to prioritize “self-reliance in science and technology” and described innovation in core technologies as the key to surviving “intense international competition.” To that end, the Chinese government has supported the development of an indigenous chipmaking industry: Guangzhou City invested $30 billion in initial funding, and Beijing has readied an additional $143 billion in subsidies and tax credits.

Second, the structure of GVCs makes it difficult for the dominant power to coerce the rising power without igniting business resistance at home and easier for the rising power to upgrade its industrial base in response. At the end of the day, businesses will prioritize their profits over national security. However, whether business incentives align with national security depends on a firm’s location in a GVC. In past instances of economic competition, high-value businesses typically resisted the dominant state’s policies to mitigate disruptions and recover lost profits. Low-value companies, by contrast, often collaborated with the rising state’s policies to reap the potential gains from industrial development.

For instance, American businesses have largely resisted U.S. sanctions against China. They continue to sell their designs, software, and equipment to China using legal loopholes, third parties, and dummy companies not on the Entity List. For example, American chip designer Nvidia developed a new advanced chip for Chinese customers that meets U.S. export control rules. Applied Materials Inc. allegedly sent chipmaking equipment to Chinese foundry Semiconductor Manufacturing International Corporation (SMIC) via third parties in South Korea. Some U.S.-based companies, like KLA Corporation, even have contingency plans to “de-Americanize” their operations in the event of more stringent sanctions.

The situation looks different across the Pacific. Chinese businesses have turned to Beijing for support in upgrading their technologies, including stockpiling equipment, building foundries, replacing proprietary data, and recruiting talent. For instance, the Shenzhen and Guangzhou investment funds helped Huawei create a self-sufficient chip network by pouring money into optical specialists, chip equipment developers, and materials manufacturers. Cash-flush start-ups and manufacturers like SMIC are successfully recruiting talent from South Korea, Taiwan, and the United States by offering massive salaries, stock options, and research funds.

As a result, Chinese businesses have achieved significant technological breakthroughs despite U.S. sanctions. Huawei remains, and the company has become mainly self-sufficient. The company stunned Washington last fall with a new 5G smartphone made with primarily Chinese parts and an advanced 7-millimeter chip manufactured by SMIC. Industry experts anticipate SMIC will be able to mass-produce these chips in the coming months, though the Biden administration remains skeptical. Some experts even expect the company to develop more powerful semiconductors. Chinese scientists are filing patents in record numbers to secure their position in the next generation of chipmaking. 

What’s Washington’s Next Move?

The Biden administration seems inclined to tighten export controls against China further. This would be a mistake because tightening controls does not reshape the underlying business dynamics stymieing U.S. strategy toward China.

Depriving Chinese chipmakers of core technologies will only drive them further into Beijing’s arms. In the best-case scenario, this is more likely to accelerate than slow China’s pursuit of “high-level technological self-reliance.” In the worst-case scenario, it may push Beijing to pursue extreme measures, like invading Taiwan and seizing Taiwan Semiconductor Manufacturing Company. Neither scenario serves U.S. interests in the long term.

At the same time, depriving American chip designers of the Chinese market will only push them further away from Washington. China is simply a better place for tech-intensive industries than the United States. Its population is larger, its wages are cheaper, and its manufacturers have more government support. Producing chips in the United States still takes 25 percent longer and costs 50 percent more than in Asia. So long as this economic reality exists, the Biden administration will waste time and resources playing “whack-a-mole” to prevent American chip-designer companies like Micron and Nvidia from skirting U.S. export restrictions.

The CHIPS act may eventually change business incentives over the long term, but the signs are not promising. Labor and water shortages have frustrated the opening of new U.S.-based foundries, red tape is prolonging chip production times, delays in federal funding are preventing smaller chipmakers and designers from breaking ground, state-funding requirements are squeezing budgets, and the “guardrails” on CHIPS Act funding are discouraging foreign investment in the United States—as well as making it more difficult for American firms like Intel and Nvidia to compete in China. The Biden administration has not adopted any of the proposed actions to address these impediments.

Given these concerns, it is time for the Biden administration to rethink its strategy before it is too late.