As East Asia emerged as the world’s main chip-making hub, the United States’ share of global semiconductor manufacturing capacity has decreased from 37 percent in 1990 to 12 percent in 2021. Despite U.S. dominance in chip design, industry giants including Intel, Micron, Broadcom, Qualcomm, and Texas Instruments produce their chips overseas or contract out production to companies abroad. The lack of manufacturing capability, together with the current chip shortage, has prompted U.S. leaders to treat semiconductors as an essential element of national security, especially in the context of tense China-U.S. relations. In March 2022, for example, President Joe Biden referred to semiconductors as “so critical to our national security… that we’re going to create rules to allow us to pay a little more for them if they’re made in America.” Similarly, Senate Majority Leader Chuck Schumer said, “semiconductor manufacturing is a dangerous weak spot in our economy and in our national security.”
Today, it has become conventional wisdom in the United States that semiconductor manufacturing is a security issue. But that view is mistaken. Dependence on East Asia for manufacturing is unlikely to pose a challenge to U.S. security.
The East Asian countries that manufacture chips, including Japan, South Korea, and Taiwan, are U.S. allies. They share not only democratic values but also geopolitical interests with the United States. In particular, Taiwan remains committed to its partnership with Washington. At the United States’ request, Taiwan Semiconductor Manufacturing Co Ltd (TSMC), the world’s largest contract chipmaker, offered the Commerce Department data on supply chains. It has suspended shipments to such Chinese companies as Huawei Technologies, despite its business interests in China, and begun construction of a 12-inch wafer foundry in Arizona. In line with U.S.-led sanctions, it recently halted deliveries to Russia in response to its invasion of Ukraine. Although TSMC is caught in the middle of a tug-of-war between Washington and Beijing, it has sided with Washington.
Another East Asian country that makes chips is China. However, its technologies lag by at least two generations. While the United States and its allies contribute 92 percent of the total value of the global semiconductor supply chain, China only accounts for 6 percent. Chinese companies, such as Semiconductor Manufacturing International Corporation (SMIC), can only produce 45-nanometer chips, whereas TSMC now produces 3-nanometer chips. Although China endeavors to catch up, the United States and its allies will enjoy a competitive advantage for years to come.
One possible security scenario leading to supply insecurity is a Chinese invasion of Taiwan, which would disrupt the U.S.-centered supply chain. However, such a Taiwan Strait scenario is an exaggeration. While it is true that Beijing refuses to rule out the use of force, there is no sign that China will soon seek to resolve the Taiwan issue by force. The Taiwan Strait is certainly a security issue, but Beijing is unlikely to invade Taiwan if it remains preoccupied with domestic socioeconomic issues or factional struggles. During the years 2000−2008, when Taiwanese President Chen Shui-bian adopted pro-independence policies, Beijing exercised restraint. Since 2016, President Tsai Ing-wen has adopted a pro-status quo policy, and consequently Beijing feels little urgency to resort to force. Indeed, Beijing has continued to promote economic interdependence across the Taiwan Strait as an instrument of political integration.
Therefore, the lack of chip-making capacity poses few threats to U.S. security in the short and medium terms. There is no doubt that national security is maximized when any country controls the upstream, mid-stream, and downstream of the industry, but only dominating part of the supply chain is not as worrisome as U.S. security professionals tend to assume. If few East Asian countries (with China the notable exception) treat their weaknesses in chip design as a security problem, why should the United States worry about its lack of manufacturing capacity? Indeed, Taiwan does not worry much about its weaknesses in chip design. From a Taiwanese perspective, its semiconductor manufacturing is sufficient to operate as a “silicon shield” in its fraught relationship with China.
When the United States and its allies need each other, security problems will be eased by economic interdependence and the global division of labor. The same may hold true for China-U.S. relations, even if China’s chip-making capacity were to progress.
In order to achieve what U.S. leaders call chip-based security, they should shift the semiconductor issue from the security realm to the arena of normal business. Specifically, U.S. leaders should leave decisions on the construction of new chip foundries in the hands of market actors rather than security professionals. Current legislative efforts to provide $52 billion in federal subsidies for chip-making on U.S. soil in the name of security will not resolve the long-standing issue that costs related to labor and regulatory compliance are higher in the United States than overseas. As the Boston Consulting Group and the Semiconductor Industry Association point out, the 10-year cost of a chip fabrication plant in the United States is 30 percent higher than in Taiwan and South Korea, and as much as 50 percent higher than the cost in China. If chips manufactured in the United States are not cheaper or better, there is no guarantee that U.S. automakers and other domestic companies would purchase U.S.-made chips. If this proves to be the case, the security goals that U.S. leaders are seeking would not be met.
Conversely, the private sector knows well how to secure their foundry business. Manufacturers maximize the fabrication of chips wherein costs are relatively low. Moreover, they adapt themselves to boom-bust cycles. At a time when a chip shortage has happened and led to price increases, TSMC and other manufacturers have expanded their production for more profits. If the market functions well, the current chip shortage is estimated to end by the end of 2022 or the spring of 2023, whereas it will take three to four years to build a new fabrication plant. (TSMC’s Arizona plant and Intel’s new fabs in Ohio will only start mass production in 2024 and 2025, respectively. Then the industry may find itself dealing with overcapacity.) When these enterprises help reach market equilibrium, U.S. security professionals will have less to worry about.
In other words, there is no contradiction between normal market activity and national security. In some cases, the private sector resolves security problems better than security professionals. Indeed, a White House report on supply chain vulnerabilities acknowledges that “the private sector must take the lead in addressing the shortage” and identifies the government’s role as one of assisting. Security professionals in the Biden administration should not subsidize semiconductor production in ways China has adopted, but instead give the market a bigger role.
In particular, TSMC’s manufacturing operations in Taiwan, if deepened, would ensure not only U.S. but also Taiwanese security. U.S. security professionals have persuaded TSMC to open new fabs on U.S. soil, but Chinese President Xi Jinping would be incentivized to invade the island if TSMC’s role as Taiwan’s “silicon shield” were weakened. Conversely, Xi would be deterred from attacking Taiwan if Taiwan’s semiconductor industry remains important to U.S. interests, in addition to if Taiwan’s geographic location remaining vital to U.S. grand strategies in the Western Pacific. If Xi understands that the United States would come to Taiwan’s defense to protect chip supply chains, war would become a remote possibility.
In sum, there is no single path to security, and security may be achieved by the market rather than security politics.