Opportunities and Pitfalls for U.S.-EU Collaboration on Semiconductor Value Chain Resilience

On May 15 to 16, the U.S.-EU Trade and Technology Council (TTC), a multifaceted bilateral working group formed in 2021 to promote joint development of democracy and market-oriented approaches to technology and innovation, convened its second meeting in Paris. Following the TTC meeting, the two sides announced plans to develop “a transatlantic approach to semiconductor investment aimed at ensuring security of supply and avoiding subsidy races.” The parties released a detailed protocol for joint efforts to identify vulnerabilities in the semiconductor value chain, create a monitoring and early warning system to alert each other on looming shortages, and implement subsidy and incentives policies that are transparent and conform to WTO subsidy disciplines.

In effect, the United States and the European Union have proposed a coordinated implementation of industrial policies in a vital strategic sector involving massive public outlays on both sides of the Atlantic—an unprecedented development in postwar transatlantic relations. Given, in particular, the long history of bilateral friction with respect to subsidies in sectors such as commercial aircraft, agriculture and steel, the current accord, if realized, would be remarkable, reflecting a shared recognition of the need to work together to address urgent shared vulnerabilities.

Shocks in the Semiconductor Ecosystem

Semiconductor chips are the basic building blocks of the modern world, yet both the European Union and the United States have allowed themselves to become dependent on external sources—some of them now problematic—for production of many types of chips as well as other elements in the production chain. The shocks that the Covid-19 pandemic delivered to the chip sector, combined with rising demand, have resulted in “shortages across multiple economic sectors with potentially serious societal and economic consequences”:

  • Shortages of chips have buffeted the automobile manufacturing sector and many other industries in both the United States and the European Union, an issue that domestic production capabilities have proven unable to address in the near term.

  • Gaps have emerged in the semiconductor value chain in both regions, such as advanced packaging capability and various exotic materials. The war in Ukraine has underscored these vulnerabilities, giving rise to, for example, shortages of neon gas, of which Ukraine is a major supplier and which is required to operate lasers used in the most advanced semiconductor lithography machines. Some experts point out that without neon, “chip production comes to an abrupt halt.”

  • Such vulnerabilities are becoming apparent against the background of a rising and increasingly assertive China, which has committed enormous resources to the development of its semiconductor industry and the deployment of semiconductor-enabled military systems which could prove decisive in a military showdown. The Ukraine war underscores this point—a single Javelin missile, for example, requires about 250 semiconductors, and shortages of chips are reportedly constraining the ability of U.S. defense contractors to supply weapons to embattled Ukrainian forces.

Opportunities and Challenges for Cooperation

At first glance, the prospects for U.S.-EU collaboration to address an acknowledged joint risk are excellent. The parties have shown they are fully aware of the present danger, and that much of the foundation for collaboration is already in place. In microelectronics, U.S. and European semiconductor industries have already built a multifaceted, interdependent ecosystem of two-way investments, joint research, technology transfers, and supply arrangements. European firms have participated in U.S. chip manufacturing research projects (Sematech, Albany’s College of Nanoscale Science & Engineering, the U.S. national labs extreme ultraviolet, or EUV, consortia), and U.S. firms participate in European research organizations, such as IMEC in Belgium, Fraunhofer in Germany, and CNET-Leti in France. European firms have chip production facilities in the United States (Infineon, X-Fab, BAE Systems) and U.S. firms make chips in Europe (Intel, GlobalFoundries, ON Semiconductor, IXYS, Analog Devices). Companies from each region operate major chip design centers in the other region. The U.S. semiconductor manufacturer Intel is undertaking a major investment program in advanced semiconductor manufacturing in both the United States and the European Union with the support of public authorities in both markets.

However, despite the solid base such integration represents, any collective U.S.-EU effort to create a more resilient semiconductor value chain is sure to face substantial challenges. If history is a guide, past attempts at EU-U.S. government-to-government collaboration are not encouraging. Those initiatives, such as the New Transatlantic Agenda (1995), the Transatlantic Economic Partnership (1998), and the Transatlantic Trade and Investment Partnership (TTIP) foundered, reflecting substantial differences over regulatory policy and acrimonious sectoral disputes, particularly in agriculture.

Complexity and Cooperation

At the most prosaic level, even with excellent cooperation between European and U.S. public authorities, simply identifying risks in the semiconductor value chain poses a significant challenge. This reflects the complexity, multiplicity, geographic dispersion, and limited transparency of the industry chains involved. The fabrication of semiconductor devices requires precision machines, software, and exotic materials, with each of these input elements having its own distinct, often intricate supply chain.

The most advanced semiconductor lithography machines—EUV tools made by the Dutch company ASML—consist of several modules incorporating hundreds of thousands of components sourced from multiple tiers of nearly 800 global suppliers. The modules are built at 60 ASML locations around the world and shipped to the Netherlands for assembly. Other machines and specialized materials used in the semiconductor manufacturing process have their own complex supply chains, featuring obscure and esoteric materials and components, with the roots of the chains often not fully understood by the original equipment manufacturers themselves or by their primary suppliers.

Such sprawling chains have proven to be prone to disruption by the vagaries of events that in theory have been foreseeable but, in reality, the industry was not prepared for. As the TTC’s May 2022 statement on semiconductors acknowledged, during the past two years, chip supply has been “disrupted by a series of events, such as factory fires, winter storms, energy shortages, droughts, and COVID-19 related shutdowns. This was compounded by dislocations in global logistics, transportation networks and shortages of raw materials and intermediary products.” Indeed, the range and variety of the disruptions underscores the reality of systemic vulnerabilities.

Gaps in U.S.-EU Value Chains

Thus, a threshold requirement for any bilateral effort to reduce chip supply chain vulnerabilities is to understand what those weaknesses actually are, ideally through cooperative “supply chain mapping.” At the broadest level, certain major gaps in the U.S. and EU chip value chains are quite evident and offer a point of departure for the TTC initiative:

  • Neither the United States nor the European Union has any current indigenous manufacturing capacity with respect to the most advanced semiconductor devices, which are based on seven-, five- and three-nanometer (nm) design rules. Intel is investing in seven nm fabs in the United States and the European Union, which it expects to bring online in 2023 and beyond. However, for the moment, all of the capacity for making chips at the seven nm and smaller nodes—which is essential to the development of advanced chips for artificial intelligence (AI)—resides in Taiwan and South Korea.

  • The TTC working group on secure supply chains identified other critical semiconductor shortages, “in particular legacy logic chips, analogue chips, and optoelectronic chips, as well as the substrates and raw materials [for chips] used in critical industries and economic sectors including automotive, healthcare, industrial automation, communications, and energy.”

  • The United States is weak in key areas of semiconductor lithography, the process by which complex circuits are printed onto semiconductor substrate. With respect to EUV lithography, essential for manufacturing the most advanced chips, as well as other advanced lithography technologies such as argon fluoride laser (ArF) immersion and dry scanners and krypton fluoride laser (KfL) steppers, the United States has no onshore capability.

  • Europe lacks advanced and trailing-edge (several generations behind cutting-edge) logic fabs and has no foundries producing chips at process nodes under 22 nm.

  • The European Union lacks chip design capabilities for advanced logic semiconductors, although European firms enjoy strength in the design of chips for power electronics, analogue, and radio frequency devices. The European Union has been trying to shore up its chip design ecosystem for a number of years, but with disappointing results so far. With the decline of Nokia, Europe no longer has a major company like Qualcomm, Apple, or Samsung seeking the most advanced semiconductor designs.

  • Both the United States and the European Union are extremely weak in the so-called back end of the semiconductor production process: assembly, testing, and packaging (ATP). These functions have largely been outsourced to East and South Asia, primarily for lower-cost production. However, this calculus may be changing. In particular, “advanced packaging” involving new technological approaches, such as 3D stacking of chips and “heterogeneous integration” involving the combination of separately manufactured components into a system-in-package, are becoming important elements in improving chip performance and energy efficiency. They also open opportunities for the reshoring of these functions in more advanced facilities.

Challenges in Mapping the Value Chain       

While mapping relative regional positions may make sense, examining chip supply chains to identify and address specific risks lurking in the myriad channels that comprise those chains is a daunting task. The TTC working group addressing semiconductor risks pledged that the parties will “take action to increase transparency and monitoring of the value chain.” This may make sense at the governmental level, but the semiconductor companies themselves are hypersensitive about disclosing detailed information about their supply chains and suppliers. 

In 2021, the Biden administration launched an intensive review of supply chain vulnerabilities and established a Supply Chain Disruption Task Force to respond to shortages of semiconductors and other key products. The U.S. Department of Commerce’s Bureau of Industry and Security (BIS) supported this effort with requests for voluntary provision of supply chain information from individual companies. Companies were asked for information on supply chain inventories, raw materials, equipment purchases, delivery times, procurement practices, backlogs, and other similar information. Some companies were reluctant to comply, fearing the loss of trade secrets. The administration warned that it could use the Defense Production Act or other measures to compel compliance, but these coercive tactics prompted a backlash. One Korean observer commented that “semiconductor companies’ confidential data is something unshareable even with the U.S. government, even if it promised to keep the information secret.”

There is little reason to expect that U.S. and European chip firms will be any happier to share confidential supply chain information with both the U.S. government and the European Commission. As recent EU-U.S. friction over digital data privacy issues has underscored, Europeans are extremely hesitant over delivering data to the U.S. government that may be accessed by U.S. national security agencies. Conversely, U.S. companies are leery of antitrust policies and activism by EU regulatory bodies.

Potential for Conflict

Even if the United States and European Union are able to pry sufficient data from the private sector to identify critical choke points in the semiconductor value chain, the question of how the parties will collaborate to address such risks remains. The United States and the European Union are embarking on parallel policy tracks involving the commitment of large public resources to strengthening indigenous semiconductor capability. In the United States, legislation being reconciled in Congress provides for $52 billion in federal outlays, while the European Union’s initiatives envision new and previously committed public investments of comparable magnitude. In the Paris meeting of the TTC, the parties pledged to avoid “subsidy races,” indicating that they recognize the danger of an incentives competition that often sees rival governments bidding against each other for local chip investments. However, while both sides have established guidelines to rationalize the use of chip subsidies, it is not clear that friction over subsidies can be avoided given that political leaders in both regions are beholden to domestic constituencies that favor major public investments that are both national and local in impact.

Historically, most U.S. chip subsidies have been provided by state and local governments, reflecting the economic value of fabs to local economies and the realities of locational competition both globally and at the state level. State and regional governments may assert that they are not bound to bilateral international commitments at the federal level. It is unclear whether and how federal authorities can prevent local undermining of their efforts. On the EU side, the European Commission has indicated that it will relax its legal restrictions on state aid (subsidies) to enable public funding of semiconductor investments. In the past, however, even very rigorous enforcement of EU state aids rules could not halt periodic subsidy competition among the member states. In this situation, the EU rules will be less stringent with respect to internal subsidies. Furthermore, the United States, having no state aids rules of its own, will not be bound by EU state aid law, regardless of whether it is enforced or relaxed. The risk of conflict will be particularly acute in situations in which investment in a particular capability is only economically or technological feasible at a single site. Both parties will vigorously compete to ensure that site is built within their own national boundaries for purposes of security, technological competency, and local economic benefit. It should be recalled that the WTO subsidy rules, which TTC representatives envision as a sort of normative umbrella to govern the joint EU-U.S. chip effort, have proven largely ineffectual as disciplines on subsidies.

This is not to argue that bilateral efforts to manage competition and the levels of subsidies it entails is not worthwhile. However, the record of success is not extensive, while the pressures for progress are. Indeed, it may not be a good use of political capital to block subsidies competition that is, in effect, already taking place and is, understandably, welcomed by the industry. Fundamentally, it may be worth asking what is broken about the current system—one that involves major financial and tax subsidies to attract high-tech manufacturing, but one that is also driven by geographic competition in talent, regulations, educational excellence, and infrastructure. Intel’s unprecedented investments across Europe, including a research center in France, investments in assembly and testing in Italy, investment in existing facilities in Ireland, and a massive new facility in Germany, have been transformative for the European microelectronics ecosystem. The United States should be careful what it attempts to fix.

Paths to Progress

Such concerns notwithstanding, there is reason to hope for the success of the TTC, given sufficient political will. The most recent TTC meeting, which produced the semiconductor accord, reflected a broader trend over the past year that has seen breakthroughs in longstanding U.S.-EU disputes in areas including digital commerce, taxation of multinationals, commercial aircraft subsidies, data privacy, and regulation of U.S. large tech platforms. In addition to chips, the TTC is pursuing detailed and ambitious agendas in diverse thematic areas, including AI, rare earth magnets, climate and clean tech, technology standards, solar supply chains, cybersecurity, and data governance. The two sides appear to have agreed to coordinate their export control and investment screening policies in technology sectors, a very positive development. Fortuitously, agricultural policies, which have been a longstanding source of bilateral controversy, fall outside of the purview of the TTC, but they are perhaps also an indication of its limits.

Finally, the surprising—and welcome—degree of cohesion and solidarity between the United States and the European Union in response to Russia’s invasion of Ukraine bodes well for joint efforts to address other shared security threats, including chip dependency. As an IBM executive commented in the wake of the May 2022 TTC workstream on semiconductors, after Russia invaded Ukraine, “people on both sides of the Atlantic got reminded that the U.S.-European relationship is the essential partnership—economically, politically and now militarily—in the world.”

Recognition of this reality is a promising sign of expanded cooperation. Nonetheless, the shift towards greater cooperation might as well be guided by a more nuanced recognition of the complexities and uniqueness of the industry, the key role of competing “national” firms, and the difficulty of assuring cooperation on investments across diverse actors with equally diverse interests. This is not to say cooperation should not be encouraged, but it is unlikely to be realized to its desired extent. Indeed, modest goals that are achievable might well be a better option than goals that are too ambitious and ultimately unachievable. It is critical that the newfound spirit of EU-U.S. cooperation not pass too quickly from the congratulatory phase to the recriminatory.

Sujai Shivakumar is director and senior fellow of the Renewing American Innovation Project at the Center for Strategic and International Studies (CSIS) in Washington, D.C. Charles Wessner is a senior adviser (non-resident) with the CSIS Renewing American Innovation Project. Thomas Howell is an international trade attorney specializing in the semiconductor industry.

Commentary is produced by the Center for Strategic and International Studies (CSIS), a private, tax-exempt institution focusing on international public policy issues. Its research is nonpartisan and nonproprietary. CSIS does not take specific policy positions. Accordingly, all views, positions, and conclusions expressed in this publication should be understood to be solely those of the author(s).

© 2022 by the Center for Strategic and International Studies. All rights reserved.

Image
Sujai Shivakumar
Director and Senior Fellow, Renewing American Innovation Project

Thomas Howell

International Trade Attorney specializing in the semiconductor industry