China's Chip Breakthrough Demands an American Response

Earlier this week, a click-bait-y title caught my eye. The article, China’s Semiconductor Breakthrough, detailed the People’s Republic of China’s recent innovations in advanced semiconductor manufacturing.[1] Until now, the world’s highest-quality chips—which are essential to accomplish demanding tasks such as cloud computing, machine learning, and 5G—have been produced outside of the PRC. This is deliberate: the United States has attempted to hamstring China’s advanced semiconductor industry by blocking Chinese entities from purchasing essential manufacturing equipment. Only one company in the world—a Dutch company called ASML—produces the advanced Extreme Ultraviolet Lithography (EUV) machines necessary to manufacture top-of-the-line semiconductors, and the United States has lobbied Dutch officials heavily to ban their sale and export to China.[2]  China has therefore been keen to develop its own chip-production methods and has landed on an alternative: Deep Ultraviolet Lithography (DUV). 

Although use of DUV tech to manufacture advanced—10nm and below—chips in volume is more complex, time consuming, and expensive compared to EUV manufacturing, these tradeoffs are apparently outweighed in Beijing’s eyes by the importance of semiconductors to China’s economic and military well-being. The PRC has created massive subsidies for producers to make the costly and inefficient DUV-powered chip fabs economically competitive—which is necessary to enable production at the volumes required for both national and economic security. This alone is not necessarily troubling, but the following conclusion raises concerns: 

"Given that excess inventory has emerged in some areas of the electronics industry, and the market expects that there will be excess production capacity in chip manufacturing after 2023, price competition is inevitable. Foundries using mature processes will not be able to compete with Chinese semiconductor factories that enjoy major policy subsidies. Some second- and third-tier foundries may have to withdraw from the market, which will allow Chinese foundries to dominate the mature process market."[3] – Chen-Jen Wang, The Diplomat.

            The implications for America’s chip sector—and by extension, every chip-dependent sector (e.g. the defense, auto, telecom, and aerospace industries, to name a few)—are severe. Although the United States government is working to accelerate the development of its own semiconductor industrial base—most notably with the recent passage of the CHIPS Act—Chinese dominance of the mature process market will erode America’s ability to maintain domestic production at the levels required for economic and military self-sufficiency. An influx of cheaply made Chinese chips will end the chip shortage and price American manufacturers out of the market. While this is great news for consumers, it is devastating from for economic and national security. Without a government response, the United States will be left dependent on Chinese chips in the near future, which risks handing Beijing an unprecedented degree of leverage over America’s economic and foreign policy decisions. 

            China’s chip breakthrough therefore presents a vexing problem for American policymakers. Assuming inaction is too costly an option to entertain, the policy tools available to U.S. decisionmakers are few. The standard response might include some combination of tariffs, subsidies, stockpiling, or anti-dumping duties—to name a few. The problem with each of these options is that they would likely violate the World Trade Organization’s (WTO’s) rules and would actually do more harm than good. Although the WTO has experienced its fair share of criticism,[4] it has facilitated mutually beneficial trade relations that have markedly improved the standard of living for billions of people. Admittedly, as with most international institutions, its enforcement mechanisms are lacking. Quasi-mercantilist autocracies have long shirked the rules, but compliance is further hampered by American circumvention and hypocrisy. 

Even if we put that aside—after all, the Chinese chip subsidies that prompted this article likely violate WTO rules as well—pursuing such protectionist measures may offer limited benefits and considerable costs. For instance, although American tariffs on Chinese chips would protect U.S. chip fabs from Chinese market manipulation, they would offer little protection to foundries located outside of the United States. Although other nations with large semiconductor industries—South Korea and Taiwan, for instance—are likely to follow America’s lead with respect to tariffs, but even a coordinated U.S./East Asian (and potentially even E.U.) tariff on Chinese-made semiconductors would not prevent their sale in the developing world. Developing nations are especially unlikely to impose import restrictions on semiconductors, since they lack domestic chip industries and face little incentive to do anything other than purchase the cheapest semiconductors they can get—which, by 2024, will all be Chinese. In addition, tariffs on Chinese chips will almost certainly hamper American competitiveness. If American companies in chip-dependent industries are unable to access the cheap chips that their overseas competitors can freely purchase, every derivative good and service will be markedly more expensive if manufactured in the United States. This will create significant pressure to move manufacturing in chip-derivative industries away from the U.S., creating a capital, talent, and employment drain on the American economy. 

            Subsidies may seem like a tempting alternative. Sufficiently large subsidies for America’s semiconductor manufacturing companies would allow them to engage price competition with PRC chip makers, potentially warding off Chinese domination of global chip markets. This, in turn, would preserve America’s ability to manufacture and sell chips at volume—thereby addressing the economic and security concerns mentioned above. However, subsidies for domestic fabs will be incredibly costly. China enjoys a comparative advantage in labor, energy, and bureaucracy when it comes to building out key industries. Labor costs in the PRC are rising but remain below those of the United States.[5] China’s massive coal reserves and expanding coal power-generation capacity will ensure a consistent supply of cheap (albeit dirty) energy with which it can power its network of chip fabs. Moreover, China’s economic governance—‘socialism with Chinese characteristics’, as it is known in the PRC—is extremely permissive for ‘national champion’ industries (like semiconductor manufacturing), granting them political favors, generous land lease terms, and waiving environmental reviews when the CCP higher-ups believe it necessary. Even if the relative inefficiencies associated with DUV technology negate these advantages, the CCP may unilaterally escalate domestic subsidies for chip manufacturers if they deem it advantageous. By contrast, cash handouts to some of America’s largest companies (e.g. Intel) would be a hard political sell to Americans on both sides of the aisle. The CHIPS Act, for instance, has faced opposition because of key omissions that would allow dollars from U.S. subsidies to be spent on expanding American companies’ semiconductor footprint in China.[6] If American support for its semiconductor manufacturers is simply met by additional Chinese subsidies—or if American subsidies are simply spent in China—political will for an ever-more-costly subsidy war is likely to be lacking. Subsidies, it seems, are no silver bullet. 

The other alternatives mentioned above are even less promising. Anti-dumping duties will likely bring along the same problems associated with tariffs but would also rely on a demarcation of what ‘dumping’ consists of, creating the opportunity for PRC circumvention. Similarly, stockpiling is a non-starter until 2024, as it would exacerbate the current chip shortage if done sooner—potentially causing more harm than good. 

Ultimately, some combination of tariffs and subsidies is likely to be optimal. One option would consist of allowing tariff rates to fluctuate with the level of PRC subsidies, while also using tariff revenue to fund domestic subsidies for American chip companies. This seems particularly prudent since it would effectively allow American chipmakers to benefit from increased Chinese subsidies. Such a scheme would in turn create a substantial disincentive for China to escalate its subsidization of domestic semiconductor manufacturers, and if done in coordination with our global allies and partners, may remove this weapon from the PRC’s arsenal altogether.