The United States launched an economic war against China last month. The declaration of hostilities was a dense regulatory filing. The war is coolly bureaucratic, its aims strictly limited. But there’s little doubt that we would call Joe Biden’s new China policy by that name were the U.S. its target instead of its author.
The policy’s details are complex, involving a thicket of export controls on semiconductors, semiconductor manufacturing equipment, and American expertise related to the two. But its upshot is simple: it is now official U.S. policy to prevent China from achieving its development goals.
Like their counterparts in the U.S., Chinese policymakers believe that mastering advanced computing technologies is a precondition for competing in the highest-value sectors of tomorrow’s economy. The world’s most advanced microchips can power innovation in e-commerce, robotics, medical imaging, pharmaceutical research, self-driving vehicles, and myriad other markets. And the U.S. is now officially committed to keeping such chips beyond China’s grasp.
In some respects, Biden’s export controls are a natural extension of his predecessor’s policies. But they also represent a decisive break with globalization as we’ve known it. Democrats once saw economic integration between the U.S. and China as a means of enhancing America’s national security. Now, they see severing ties between the two nations as indispensable for safeguarding the liberal world order.
Whether Biden’s policy will succeed on its own terms is a source of controversy among national-security experts. The wisdom of those terms is less controversial yet no more obvious. The White House is wagering that openly stymieing China’s economic development will do less to encourage Beijing’s most belligerent tendencies than to constrain them. Nothing less than peace between the world’s preeminent powers may be at stake.
The road to economic war
Biden’s export controls aim to stop a phenomenon that has long maddened America’s national-security state: the Chinese government’s use of American tech to enhance its own military prowess.
Many of the world’s most sophisticated martial technologies — including hypersonic missiles and autonomous weapons systems — require American inputs to be constructed, whether in the form of software, components, or expertise. Yet that has not prevented the People’s Liberation Army from making progress in such areas. Given open trade between U.S. and Chinese companies and the close ties between China’s private sector and its security state, the PLA has had little difficulty securing high-end American technology for much of this century.
For four years, American policymakers have been trying to change that. First, under Donald Trump, the U.S. banned the direct sale of advanced, American-fabricated semiconductors to companies with known ties to the Chinese military, including the multinational tech giant Huawei. But this left Huawei and its peers capable of securing advanced chips from fabrication facilities (a.k.a. “fabs”) in Taiwan and South Korea.
So the Trump administration established a new rule: If a fab used U.S. designs, software, or components in producing chips, then it would have to abide by American export restrictions (unless exempted through a hard-to-get license). At present, it is impossible to mass-produce the world’s most advanced semiconductors without U.S. inputs. Thus Trump’s rule effectively choked off Huawei’s access to advanced chips produced anywhere in the world.
These measures took a toll on the targeted firms. But they failed to starve the PLA of cutting-edge semiconductors. Huawei may have been locked out of the market for such wares. But Beijing had little trouble erecting an array of ostensibly independent shell companies that retained legal access to advanced chips.
During Biden’s first year in office, America’s intelligence agencies grew increasingly alarmed by China’s persistent technological progress. They warned the administration that, with the aid of advanced chips, the Chinese Communist Party was making headway toward next-generation weapons systems and decryption technologies.
The events of 2022 only exacerbated such fears. Russia’s attempted conquest of Ukraine — and China’s support for it — made the prospect of a Chinese invasion of Taiwan “seem more real to U.S. officials,” according to the New York Times. Then, last summer, China’s own leading chipmaker achieved a technological breakthrough, producing a semiconductor with circuits “10,000 times thinner than a human hair,” making them roughly as fine as those produced in Taiwan. Conventional wisdom held that the Chinese were still many years away from being able to produce such chips at scale. But then conventional wisdom had previously doubted China’s capacity to manufacture such chips at all.
The Biden administration decided that it needed to erect a new wall between the PLA and advanced semiconductors, one that the Chinese government could not breach by creating shell companies or developing its own domestic chip industry. To do so, it would cease recognizing any distinction between China’s private sector and its military and exploit the global semiconductor supply chain’s reliance on American know-how.
Therefore, in October, the U.S. effectively banned the sale of advanced semiconductors made with U.S. inputs to any entity within China (technically, firms can apply for a license to make such transactions, but those requests will be met with a “presumption of denial”). The Chinese remain incapable of replacing such chips through domestic production. And Biden’s export controls aim to keep it that way. The administration’s rules effectively prohibit the sale of modern semiconductor-manufacturing equipment to China, leveraging the fact that such equipment almost invariably relies on U.S. components. Finally, the measures forbid U.S. persons, be they American citizens or green-card holders, from working in the Chinese semiconductor industry.
How starving China of chips could backfire
If successfully enforced, the administration’s export controls will not merely halt the progress of China’s AI industry but reverse it. China’s supercomputing companies will no longer have access to high-end chips. The nation’s chipmakers will be unable to utilize the world’s most advanced chip-design software or to purchase any cutting-edge semiconductor manufacturing equipment (SME). And its fledgling SME industry will forfeit access to indispensable American components — and, quite likely, indispensable Americans. China’s leading SME maker, Piotech, Inc., has seven executives in key research and development positions; six of them are American citizens. “It’s plausible that in 2030, China would be set back to, say, 2015 levels of technology,” said Dan Wang, a technology analyst with the research firm Gavekal Dragonomics.
For China, averting that plausible future might require building up a wholly indigenous supply chain for advanced semiconductors. Which would be a monumental endeavor.
“The question is: Can China begin to produce the machine tools that it needs to enable domestic fabrication?” said the economic historian Chris Miller, author of Chip War: The Fight for the World’s Most Critical Technology. “And I think the answer to that is: not anytime soon. The specific machines in question are among the most precise and complicated that humans have ever made. And the challenge with this machinery is not simply getting it to work once, but to work with almost perfect accuracy running almost all the time.”
Nevertheless, there are several ways that America’s strategy could falter. The most immediate threat concerns the cooperation of U.S. allies. China may be years away from being able to build advanced semiconductors and SME without America’s help. But the Netherlands and Japan are not. Currently, high-end SME like that produced by the Dutch firm ASML requires U.S. inputs. But if leading firms were sufficiently motivated, they could find alternative components, according to Paul Scharre, vice-president of the Center for a New American Security.
And Biden’s policy effectively creates a massive market for advanced chip technologies that are free of American inputs. That market will only grow over time: Today, only about one percent of all chips are advanced enough to meet the export controls’ standard for prohibition. But as innovation in the chip industry progresses, the percentage of chips barred by the administration’s rules will steadily rise.
U.S. allies are sure to abide by the letter of American law. But whether they will enact similar measures, or discourage their firms from helping China navigate the obstacles that America has laid down is less certain. The Biden administration had initially hoped to impose its export controls multilaterally, and spent nearly a year trying and failing to rally allied participation.
“It’s conceivable that foreign companies could recreate these key products without U.S. components in a roughly five-year timeframe,” said Jon Bateman, senior fellow at the Carnegie Endowment for International Peace. “But the companies would need to make strategic decisions to do so. And their host governments would need to refrain from emulating America’s controls. So diplomacy becomes really important.”
Bateman argues that America’s decision to go forward with unilateral controls risks undermining such diplomacy. “By revealing the maximalism of Washington’s campaign against Chinese technology,” Bateman wrote in a recent piece for Foreign Policy, “the move will sharpen debates in allied capitals about whether U.S. aims align with their own political and economic interests.”
As Bateman’s colleague Matt Sheehan noted in a recent report, although U.S. allies “share a general concern about Chinese technology prowess,” the “exact cost-benefit calculations around equipment sales look different when viewed from a smaller country” that doesn’t see itself “as locked in a battle to be the one dominant global superpower.”
Over the long term, meanwhile, it is possible that Biden’s policy could actually hasten China’s achievement of self-sufficiency in advanced computing. Currently, China imports roughly $400 billion worth of semiconductors per year. If China gets locked out of global markets for high-end chips, a large percentage of that buying power will be redirected to domestic chipmakers.
“So long as Chinese commercial and data center firms have access to foreign chips, they’re going to choose those options over domestic suppliers that are not as capable,” Scharre said. “If you force Chinese data centers to rely on those second-tier domestic companies, those companies will see more profits and revenues, which should increase their capabilities over the long-term.”
It may take a long time for China to develop the machine tools, software, and fabs necessary for producing 2022-level chip technology. But there’s little reason to assume that that objective will escape China indefinitely. Especially, if a “100 percent non-American” SME supply chain arises to serve Chinese demand. In Sharre’s view, the geopolitical competition between the U.S. and China is a long-term one. And proficiency in artificial intelligence will be more valuable in two decades than it is today. A world in which China remains reliant on global supply chains for advanced chips in 2035 is one where the U.S. retains considerable economic leverage over its rival. A world in which China is wholly self-sufficient in the production of the world’s highest-performing chips, on the other hand, is the Pentagon’s nightmare.
A technologically weaker China isn’t necessarily a safer one
Even if Biden’s export controls succeed tactically, they might still fail strategically. Which is to say: They might make China less technologically capable without making the nation less geopolitically dangerous.
By far, China’s most menacing ambition is its commitment to reunification with Taiwan. The United States has good reason to fear that China might eventually seek to take the self-governed island by force. At the CCP’s party congress in October, Xi Jinping reiterated China’s commitment to reunification, and it’s right to do so by “all measures necessary.”
A Chinese invasion of Taiwan would not only threaten the lives and democratic rights of its 23.5 million inhabitants, but also devastate the global economy. Taiwan currently produces 65 percent of the world’s semiconductors, and 90 percent of its advanced chips. In the event of conflict, the world would almost certainly lose access to those chips for a prolonged period of time; if the fighting destroyed Taiwan’s fabs, a global shortage of semiconductors could durably depress living standards. What’s more, such a war would also threaten shipping lanes that roughly one-third of the world’s seaborne traffic relies upon. According to a study from the RAND Corporation, a year-long war between China and Taiwan would shave between 5 and 10 percent off America’s gross domestic product, and between 25 and 35 percent off of China’s — a development that could destabilize the many developing countries who derive critical revenue from commodity exports to the world’s largest nation.
But it’s not obvious that Biden’s policy does much to discourage China from pursuing such an invasion. America’s export controls will limit the Chinese military’s proficiency in the most technologically sophisticated forms of warfare. Yet, for the foreseeable future, artificial intelligence and quantum computers will have no great bearing on a war in the Taiwan Strait.
“If you pour over the military policy discourse on this,” Bateman said, “the factors that people point to as likely decisive are really the bread and butter stuff: What’s the professionalization of China’s military at the time of invasion? What’s its capability to launch an amphibious invasion? What about Taiwan’s air defenses? What’s the willpower of its civilian population like, et cetera, et cetera. So, count me as dubious that some kind of artificial intelligence capability will be in any way relevant to that conflict in the next decade or two.”
If Biden’s export controls do little to deplete China’s capacity to wage a war against Taiwan, they may give Beijing greater incentive for doing so. Hal Brands, a China scholar at Johns Hopkins University’s School of Advanced International Studies, warned in June of a “nightmare scenario” in which China managed to conquer Taiwan while preserving its chipmaking capacities, thereby vaulting ahead “in the race for digital supremacy.” Needless to say, this would be an insanely high-risk path to preeminence in chip technology. But the United States has just thrown a slew of obstacles across the saner routes.
The biggest risk of Biden’s policy, though, may be more abstract. America has now positioned itself as not merely hostile to China’s foreign policy ambitions, but also, to its domestic economic ones. Not long ago, that was a stance that American presidents took pains to forswear.
In his speech hailing China’s entrance into the World Trade Organization in 2000, Bill Clinton acknowledged the fear that an economically mighty China would be a geopolitically dangerous one. But Clinton insisted that trying to limit China’s power by constraining its economic development would be tantamount to affirming the Chinese people’s worst fears about America; namely, that the U.S. did not “want their country to assume a respected place in the world.” Such a policy would therefore “be a gift to the hard-liners in China’s government.”
Many of Clinton’s predictions in that address have aged poorly. And, at this point, China’s hard-liners scarcely need a “gift”; at the last party Congress, Xi Jinping consolidated his power over the CCP. Nevertheless, U.S. policy can still theoretically fortify or mitigate Xi’s most aggressive impulses. And it’s far from clear that Clinton was wrong about the political implications of the United States committing itself to China’s economic containment.
“Anything that makes China feel like it is in a zero-sum contest with the United States endangers the security of Taiwan and U.S. allies in the region,” said Jake Werner, a historian of modern China and Research Fellow at the Quincy Institute For Responsible Statecraft.
Werner argues that the Biden administration’s export controls promote precisely that feeling. The policy’s official aim may be to undermine China militarily rather than economically; the PLA is America’s target, China’s tech sector merely collateral damage. But Beijing sees the measure as an effort to prevail in a zero-sum competition for economic advantage, “Out of the need to maintain its sci-tech hegemony, the U.S. abuses export control measures to maliciously block and suppress Chinese companies,” a spokesperson for China’s Foreign Ministry said upon the policy’s unveiling.
And it isn’t hard to find support for this interpretation of America’s intent. In a September address that foreshadowed Biden’s export controls, White House National Security Adviser Jake Sullivan declared that America’s “comparative advantage” in the global economy “must be renewed, revitalized, and stewarded.” In the realm of key emerging technologies, Sullivan argued that the United States could not settle for a “relative” advantage over its competitors, but “must maintain as large of a lead as possible.”
To be sure, the administration’s export controls could have been more aggressive. The rules preserve China’s access to the bulk of the world’s microchips, and leaves the nation with plenty of room to grow its tech industry. The primary threat to Chinese growth and prosperity remains the imbalances and inequalities of its domestic economy, for which U.S. policymakers bear little responsibility.
But over time, America’s economic war against China is likely to expand to new fronts. Biden has just reset the baseline for what it means to be “tough” on China. The Republican Party will surely seek to stake out a more aggressive stance. At present, there is little political incentive for either party to associate itself with a more dovish position toward China. And the more economic links between the two nations get severed, the smaller the constituency for rapprochement will become. Already, the Biden administration is contemplating further controls on biotechnology, quantum information science, and advanced algorithms.
Did we give economic peace every chance?
In celebrating the integration of the American and Chinese economies 22 years ago, Bill Clinton conjured a global order defined by mutually beneficial exchange. In his telling, increasing China’s access to American capital and technology would “liberate the potential of its people — their initiative, their imagination, their remarkable spirit of enterprise.” And China’s liberation would redound to America’s benefit, making the United States more prosperous and the American-led world order more secure.
History hasn’t proceeded according to plan. Beijing proved capable of liberalizing China’s economy without liberalizing its politics. And the peculiar form of globalization that Clinton championed — one characterized by limited economic planning and inequitable growth — expanded the borders of the American Rust Belt while reinforcing reactionary currents in its politics.
But to Werner, that is no reason to give up on globalization’s highest ideals. In his view, the growing enmity between the U.S. and China derives in no small part from lackluster global growth. “The problem with the global economy is not really that we don’t have enough people competing on AI,” Werner said. “The problem is that we don’t have adequate consumer demand. There’s not enough good jobs and highly paid consumers to keep the global economy going. And that’s why you get these really intense competitions for these very niche but high-value sectors.”
Werner argues increasing the purchasing power of the world’s poor and working classes would engender higher growth rates, broader opportunities for profit-making, and thus a less tense relationship between China and the U.S.
Werner’s vision may be hopelessly utopian. And Biden’s critics could well be wrong. The administration’s export controls may durably degrade the Chinese military’s technological acumen. And America’s declaration of economic war may have little bearing on the belligerence of Xi’s regime. If the Chinese leader is already committed to the conquest of Taiwan and the domination of China’s near-abroad, then the U.S. may be justified in worrying more about limiting his government’s capabilities than attenuating its nationalist resentments.
But if this analysis is correct, the implications seem bleak. The last Cold War killed millions and nearly sparked nuclear war more than once. A protracted, zero-sum struggle between the world’s two most powerful economies carries many of the same risks. “The deeper you drive this enmity,” Werner said, “the more likely it leads to an explosion that makes it impossible to turn back.”