U.S. semiconductor export controls are a double-edged sword. When controls work, they help prevent advanced chip technologies from falling into the hands of bad actors and other U.S. adversaries. However, these same policies strain the very businesses that propelled the United States into technological leadership in the first place. In limiting foreign semiconductor capabilities, Washington also limits its own.
Worse yet, controls do not always work as intended, especially when they are pursued unilaterally. When Washington placed controls on semiconductor manufacturing equipment in 2022, it didn’t bring its allies along with it. What followed was a months-long struggle to convince U.S. allies to implement mirroring controls. In that time, U.S. businesses were barred from selling to China while companies in the Netherlands and Japan delivered the very same chipmaking tools to Chinese ports in record quantity.
Generally speaking, this dilemma can also be applied to all 21st century critical and emerging technologies. If the United States hastily tries to hurt Chinese innovation, it will only end up hurting itself.
An argument currently being made by some in Washington is that pressing forward with stricter controls is the best way to counter China’s semiconductor ambitions. This line of reasoning fails to consider the full consequences of such a policy. To establish long-term technological superiority over China, Washington must resist the urge to pursue extensive trade restrictions. Instead, policymakers must set the United States up for long-term success by implementing policies that protect only the most sensitive technologies, while allowing the U.S. domestic technology industry to flourish through engaging in appropriate foreign business.
In the 21st century race for technological superiority, the United States must realize that running faster will always be more effective than tripping up its opponents. Knowing this, U.S. regulators and policymakers should focus a majority of their semiconductor-related efforts on R&D, cooperating with allies, and letting homegrown technological champions thrive – wherever appropriate – in international markets. Doubling down on flawed, overly broad economic security regulations only serves to hurt the United States while strengthening China.
‘Damming Half the River’
One of the chief concerns with imposing further semiconductor controls is that Washington would likely do so unilaterally. The United States would thus be “damming half the river” or implementing controls without ensuring that economic partners and allies, which also export this technology, will enact mirroring restrictions. Without a full blockade, the river still flows around the half-constructed dam. Likewise, the United States would end up restricting its homegrown industry without making much of an impact on the target of their controls.
The October 7, 2022, export control package – the first major set of China-related semiconductor controls released by the Biden administration, restricting tools meant to manufacture cutting-edge and slightly-older generation chips – exposed the United States to this threat. Companies in Japan and the Netherlands initially continued to sell semiconductor manufacturing equipment that enabled the very capabilities Washington sought to restrict. U.S. businesses lost out on revenue and suffered reputationally, while Chinese firms still got their machines.
When Japan and the Netherlands finally agreed to implement mirroring controls, China bulk-imported the machines right up until the restrictions went into effect. Now, Chinese firms like Huawei and SMIC use these machines to produce the very chips that U.S. policymakers sought to prevent them from acquiring.
Ultimately, this misstep damaged U.S. companies, strained relations with U.S. allies, and gave China time to adapt. Some executive branch policymakers now recognize this as an issue and are working to build multilateral coalitions for future controls. However, they will have to contend with the political pressure to take rash unilateral action against China’s development.
The Right Way to Use Export Controls
The Biden administration has made defending critical and emerging technologies (CET) a key part of its economic security policy – and for good reason. CETs such as artificial intelligence (AI) are dual use, meaning that they have applications both in civilian life and on the battlefield. To borrow an often-cited example, powerful AI models can be used by medical scientists to quickly, and cheaply, invent new drugs to fight uncured diseases. The same AI model can also be used by bad actors to design novel chemical and biological weapons. In today’s state of geopolitical competition, states aim to develop their own CET capabilities and limit those of countries of concern.
Export controls are a key tool in this technology race, but misusing them can be more damaging than helpful. Limiting the exports of items comes at a price to the domestic economy; when the controls do not have the intended effect, the costs may outweigh the benefits.
First, if a CET-producing firm that sells its product into a major foreign market finds its exports restricted, it will face revenue cuts and financial challenges. Second, export controls compel Chinese companies to “de-risk” away from U.S.-based suppliers altogether, given the possibility of future restrictions. Even if some U.S. semiconductor-related exports do not face curbs today, Chinese buyers – who are afraid of the aggressive U.S. regulatory posturing – are incentivized to turn to other sources of supply.
Moreover, the creation of a strong, unilateral U.S. export control program creates the incentive for multinational suppliers of chipmaking tools to move their operations beyond the reach of U.S. regulators. Foreign-based companies, seeking to lower the burden of complex compliance, will rid their organizations of U.S. goods and services in order to be able to easily sell to China. This phenomenon will inevitably lead to job losses and a less competitive innovation landscape in the United States.
In effect, a country that leverages export controls will hurt the domestic industry that made them a technological leader in the first place.
Thus, to maintain a strong industrial base, it’s critical to limit controls to only the most necessary – and effective – instances. A great example of this is the control policies on extreme-ultraviolet (EUV) photolithography machines. EUV machines are (1) only produced by one company, Netherlands-based ASML; (2) incredibly difficult to replicate; (3) nearly impossible to smuggle given their fragility and size; and (4) a critical tool in the production of highly advanced semiconductors, the enablers of dangerous emerging technologies like advanced computing and generative AI.
The Trump administration successfully pressured Dutch authorities not to ship EUV machines to Chinese customers back when they were first introduced to the market. Now, updated Dutch export regulations prevent this technology from being exported to China. This has granted the United States and its Western allies a significant, medium-term lead in advanced chip production.
However, this success story is difficult to replicate throughout the semiconductor supply chain. Not every chipmaking tool is as easy to control as ASML’s EUV machines.
Those eager to act are suggesting tightening regulations around other CET inputs, many of which have multiple producers in multiple countries. This could be a grave mistake. If the United States places export controls on products that are either manufactured elsewhere or easily recreated by foreign companies, then the controls will have little effect on adversaries like China.
Many chipmaking tools do not share the unique characteristics of ASML’s EUV lithography machine. If Washington chooses to tighten regulations, China will simply source tools from elsewhere – Japan and the Netherlands are key equipment producers, as well as South Korea, Germany, and Israel – or even make the tools themselves.
If Washington imposes new controls, U.S. diplomats would likely put pressure on its allies and partners to adopt mirroring restrictions. However, the United States has already spent its political capital in this area. The Netherlands, Japan, and South Korea as well as other key allies are reluctant to continue tightening semiconductor controls on China after two consecutive years of upheaval. These nations first want to assess the impact of current curbs, both on China’s capabilities and their own industries, before forging ahead. It would be a blunder for Washington to draw ire from allies at a time when robust economic partnerships are critical to Washington’s interests.
Lastly, imposing additional trade restrictions on semiconductors would pose several problems from an enforcement perspective. A broad expansion of the tools in the U.S arsenal (chief among them export controls and outbound investment bans) would make restrictions less effective. The U.S. government’s capabilities are finite. When it comes to export controls, the Department of Commerce’s Bureau of Industry and Security is already stretched thin. Likewise, the Department of Treasury is still in the process of determining how to implement the novel outbound investment mechanism ordered by U.S. President Joe Biden last August. Broader restrictions may further burden those capabilities.
When to Use FDPR
A common argument for those that prefer a unilateral approach to export controls is to make full use of the Foreign Direct Product Rules (FDPR). This rule implements statutory authority for the United States to control a product extraterritorially – even those that are fully manufactured in a foreign country – if the product utilizes any U.S.-origin technology. Since U.S. companies are critical to many parts of the semiconductor supply chain, much of the world’s chipmaking industry is covered.
However, the use of FDPR has two major drawbacks: bad politics and loopholes.
First, the assertion of extraterritorial control is particularly unpopular with U.S. allies and serves to hurt the U.S. in the international political arena. While FDPR is a short-term fix to a lack of multilateral support, it damages the United States in the long term by weakening its alliances.
Second, FDPR is not a silver bullet. While the regulations give Washington vast authority over the export of chipmaking tools around the world, they cannot cover everything. In fact, opportunistic multinational companies seeking to continue selling to China can re-tool their supply chains to rid them of U.S. inputs, freeing them from Washington’s oversight.
Excitement is growing in Washington around the effectiveness of FDPR, and some are eager to employ its use whenever the United States can’t get its way through diplomacy. However, FDPR is a brute force tool that must be used sparingly. Otherwise, the United States risks blowback from both its allies and industry.
Washington must also realize that FDPR isn’t a catch-all solution. If policymakers continue to build regulatory complexity into using U.S. inputs, they will further incentivize the creation of supply chains that eliminate U.S. influence altogether. In this bleak scenario, the United States pays the price for controls, but sees none of the benefits.
Policy Recommendations
The post-Cold War era has ended, and with its demise comes a new, intense phase of global competition anchored by technological rivalry. The United States cannot, and should not, try to backtrack on its original export control policies to return to the status quo with China. However, it also should not focus on significantly expanding trade restrictions. Rather, the United States must pivot to a new set of priorities that will help it achieve its economic security goals.
First, the U.S. needs to increase efforts to multilateralize economic security tools – both to promote and protect technological capabilities. Working with partners and allies would help Washington ensure that third countries do not fill in the gaps left by unilateral trade restriction policies. It would also enable like-minded economic powers to coordinate the development of their respective critical and emerging technologies landscapes, for example, by undertaking complementary state-led investments and enabling the exchange of talent and know-how.
There are several pathways to multilateralization of economic security tools. Collaborating with allies by expanding existing bodies such as the G-7 (adding Australia and South Korea) or AUKUS (adding Canada, South Korea, and Japan) is one. Replacing the now obsolete Wassenaar Arrangement with an entirely new regime dedicated to CETs is another.
Second, the United States should learn from its early export control mistakes and keep trade restrictions at the advanced end, rather than implementing broad actions. That move will avoid further pushes from Chinese firms to indigenize their supply chains, enabling U.S. businesses to retain revenue and granting them a de facto “phase-out” period during which they can expand to new export markets abroad. Additionally, Washington should make every effort to avoid the overuse of FDPR. Diplomacy, rather than coercion, is in the United States’ long-term interests.
While there are still many details to be ironed out by policymakers, the direction is clear: to win the 21st century technological competition with China, the United States must think in the long-term. Washington should pursue – in a multilateral setting – only the most necessary, most effective controls. In doing so, the United States can ensure its policies are fit for purpose, and allow the flourishing of U.S. domestic technology champions that will propel innovation forward.