Sourcing Requirements and U.S. Technological Competitiveness

Evaluating the Impact of National Security Guardrails in the CHIPS Act

Available Downloads

The Issue

The CHIPS Act’s national security guardrails aim to limit recipients’ operations and investments in foreign countries of concern, bringing an element of protection to the legislation’s promotion of U.S. advanced technology capabilities. The guardrails’ expansion clawback allows the Department of Commerce to reclaim funding if a recipient firm engages in any significant transaction involving the material expansion of semiconductor manufacturing capacity in a foreign country of concern, while the technology clawback supports reclaiming funding if recipients engage in any joint research or technology licensing with a foreign entity of concern related to national security critical technology. While the guardrails may protect the United States’ national security interests, they nevertheless come with risks. Their clawback measures could further Beijing’s design-out efforts, prove prohibitive to companies which would otherwise be valuable assets to the United States’ attempts to build up its semiconductor ecosystem, and strain diplomatic ties with U.S. partners.

Introduction

On August 9, 2022, Creating Helpful Incentives to Produce Semiconductors (CHIPS) Act of 2022 became law, continuing a wave of U.S. industrial policy. Backed by $52.7 billion in subsidies, tax credits, and related investments, the act seeks to boost U.S. competitiveness and innovation in semiconductor research and manufacturing while protecting national security. U.S. policymakers often tout its enactment—and ongoing implementation—as a crowning achievement, promising a “new Washington consensus” to promote U.S. geoeconomic leadership, revitalize domestic manufacturing, and reduce economic vulnerabilities—primarily in the context of competition from China.

To that end, the implementation of the CHIPS Act comes with national security guardrails that aim to limit recipients’ operations and investments in foreign countries of concern. This brief first explains the content of the guardrails, which comprise both expansion and technology clawback clauses. It then highlights that while the guardrails have clear national security benefits that may ultimately be worthwhile, they also come with drawbacks that U.S. policymakers should consider as they support the country’s advanced technology ecosystem through industrial policy packages.

Foundational Goals of the CHIPS and Science Act

The U.S. government has prioritized driving American competitiveness and innovation in high-growth critical industries, with industrial policy as a preferred means of achieving that goal. The CHIPS Act offers catalytic investments to rejuvenate domestic semiconductor research and manufacturing. While the United States created the modern-day semiconductor industry and has historically led in the technology’s development, the global semiconductor supply chain has become increasingly globalized and specialized over the last few decades, driving the offshoring of U.S. manufacturing capacity to lower-priced, often highly subsidized foreign markets. As a result, the United States has lost its competitive and innovative edge in key segments of the semiconductor supply chain, falling behind Taiwan, South Korea, Japan, and China in fabrication capacity. At the same time, semiconductors have become increasingly critical to economic and national security, enabling key dual-use applications such as artificial intelligence. The CHIPS Act and associated efforts, such as the National Standards Strategy for Critical and Emerging Technologies, endeavor to reverse this trend and restore U.S. leadership at the cutting edge of semiconductor innovation and production across the supply chain.

CHIPS Act investments likewise represent the U.S. aim to “revitalize domestic manufacturing” and “strengthen American supply chains.” Amid shortages of key technologies and inputs, including semiconductors, brought on by the Covid-19 pandemic and China’s economic coercion policies, which leverage China’s dominance in strategic supply chains to further its geopolitical objectives, the U.S. government has raised the alarm about supply chain fragility and lack of diversification. The U.S. government seeks to eliminate “supply chain chokepoints” and insulate itself from potential supply and demand shocks. This viewpoint extends beyond semiconductors: The United States has sought to onshore and reshore production across a host of critical supply chains, including pharmaceuticals and telecommunications.

CHIPS Act funding for domestic manufacturing is also devised to deliver on the “worker-centered trade policy” of the administration under former President Joe Biden. In response to what the administration perceived as the failures of economic liberalization—including deindustrialization, widespread job losses, and economic fragmentation—the administration prioritized industrial policies that advance growth from the “middle out.” The CHIPS Act is crafted to “create good-paying American jobs,” especially in new tech hubs that are historically overlooked. In the process, the CHIPS Act is said to “support equitable economic growth and development” and create new jobs. Likewise, the Infrastructure Investment and Jobs Act (IIJA) directs billions of dollars in investment toward construction in communities that lack access to clean water, high-speed internet, and quality transportation. Meanwhile, the Inflation Reduction Act (IRA) incorporates several prevailing wage and apprenticeship provisions into its various tax credits.

The U.S. National Security Strategy is another foundational element of the CHIPS Act. Owing to the growing nexus between economic policy and national security amid great power competition with China, the United States has increasingly sought to protect its interests through economic security policies. The CHIPS Act is a leading example, with new investments designed to protect U.S. access to key technologies, as are a range of bilateral and plurilateral U.S.-led economic dialogues that focus on nearshoring and friendshoring critical mineral and semiconductor supply chains.

CHIPS Act Provisions and Implementation

During the 116th Congress (2019–21), several proposed bills aimed to boost U.S. semiconductor research and fabrication capabilities. Many of these proposals were ultimately included in the CHIPS Act, either in full or in part. During the 117th Congress, the 2021 National Defense Authorization Act (NDAA) authorized them, while the CHIPS Act of 2022 appropriated their funding. In all, $52.7 billion in investments has been appropriated via the various CHIPS Act programs, with the goal of supporting three central initiatives: leading-edge manufacturing, new manufacturing capacity, and strengthened U.S. leadership in research and development (R&D).

The CHIPS for America Fund appropriates $52.7 billion to the Department of Commerce for semiconductor incentives. Of that, $39 billion is earmarked for subsidies to spur domestic manufacturing capabilities, while the remaining $11 billion is set aside for R&D and workforce development (Table 1). In addition, the act establishes and appropriates funds for three other initiatives: the CHIPS for America Defense Fund ($2 billion), the CHIPS for America International Technology Security and Innovation Fund ($500 million), and the CHIPS for America Workforce and Education Fund ($200 million).

Remote Visualization

In alignment with the CHIPS Act of 2022 and the 2021 NDAA, the Department of Commerce’s CHIPS for America: A Strategy for the CHIPS for America Fund outlines key challenges that the CHIPS program aims to overcome. According to the Congressional Research Service, these challenges include the following:

  • the significant cost disparity between building and operating semiconductor manufacturing facilities (“fabs”) in the United States compared to other countries (a consequence of variations in government subsidies, variable costs, and construction timelines);
  • the reduction in capital investments in U.S. semiconductor manufacturing and technology, thus slowing progress in process innovations and next-generation chip development;
  • the significant expenses associated with building new state-of-the-art fabs, which has led to a “fabless” business model that separates chip design from manufacturing and has created reliance on a few large foundries;
  • minimal visibility into demand forecasts, which has perpetuated boom-and-bust cycles in the semiconductor industry, posing obstacles to domestic investment; and
  • a gap between demand and supply of skilled workers in manufacturing facility construction and operation due to limited large-scale fab and packaging facility construction in recent years.
     

The Department of Commerce is responsible for administering the CHIPS for America Fund and has established two new offices at the National Institute of Standards and Technology (NIST) to lead its work: the CHIPS Program Office (CPO) and the CHIPS R&D Office. The CPO focuses on the implementation of Section 9902 semiconductor incentives program, while the CHIPS R&D Office manages investments appropriated within Section 9906, including the development of the National Semiconductor Technology Center (NSTC), National Advanced Packaging Manufacturing Program (NAPMP), and related endeavors.

 

National Security Guardrails

A particular focus of the CPO has been the formulation and enforcement of the CHIPS Act national security guardrails, which impose several restrictions on any funding recipients’ operations and investments in foreign countries of concern—namely, China. The first of their kind, the guardrails have required extensive follow-up since Congress broadly outlined them within Section 9902. In 2023, after the CHIPS Act’s passage, the CPO sought to provide specifics on the guardrails’ key details and definitions through a proposed—and now final—rule. In the process, it implemented two entirely new regulatory regimes: an Expansion Clawback and a Technology Clawback.

As the name suggests, the Expansion Clawback allows the Department of Commerce to reclaim, or “claw back,” the full funding amount (plus interest) provided to a private sector firm via the CHIPS Act if the firm engages in impermissible practices. In the case of the Expansion Clawback, CHIPS Act funding recipients, broadly referred to as a “covered entity and members of the affiliated group,” may not “engage in any significant transaction involving the material expansion of semiconductor manufacturing capacity in a foreign country of concern” for 10 years from the date of their award. There are exceptions to the Expansion Clawback, but their scale is limited. To enforce these guardrails, recipient companies must comply with stringent reporting and monitoring requirements as established in their “required agreement” with the Department of Commerce. The threat of financial penalties and legal consequences is intended to serve as a strong deterrent against any attempts to violate these terms (Annex I).

By comparison, the Technology Clawback requires the Department of Commerce to reclaim the full funding amount (plus interest) if a funding recipient or its affiliates “knowingly engage in any joint research or technology licensing with a foreign entity of concern that relates to a technology or product that raises national security concerns,” within 10 years of the date of their award. By requiring several levels of volitional acts on behalf of the recipient, the guardrail is intended only to pick up intentional acts (Annex II).

Assessing the Deployment Impacts of the National Security Guardrails


Successes of the CHIPS Act

The CHIPS Act seeks to catalyze greater investment in domestic semiconductor research and manufacturing, and recent evidence suggests that it has begun to achieve these goals. In the year following the act’s signing, the private sector announced new domestic investments in semiconductors and electronics worth over $166 billion. The Semiconductor Industry Association (SIA) now projects that new semiconductor-related investments in the United States will total $327 billion between 2022 and 2032 and that domestic semiconductor manufacturing capacity will triple—the largest projected percentage increase in the world over that time. In parallel, the United States is predicted to grow its share of global capacity from 10 percent to 14 percent—its highest share since 2005 and the first uptick in over a decade. The United States represents over one quarter (28 percent) of total global capital expenditures, second only to Taiwan (31 percent). Notably, before the CHIPS Act, U.S. global capacity was expected to decline to 8 percent and by 2032 capture only 9 percent of global capital expenditures, according to SIA. Even amid the constraints imposed by national security guardrails, the CHIPS Act is widely predicted to accomplish its primary goal: boosting U.S. semiconductor production and innovation.

The Case for CHIPS Act Guardrails

The case for the CHIPS Act national security guardrails is evident. The U.S. government aims to avoid funding firms’ expansion in countries of concern, which could eventually threaten its national security given the critical nature of semiconductor manufacturing. In addition, the guardrails support the U.S. goal of diversifying critical supply chains away from China, lessening countries’ dependence on a nation that has shown willingness to leverage its productive dominance to achieve geopolitical aims. It is nevertheless important for policymakers to understand this tool’s potential drawbacks, especially when assessing how to include it in additional industrial policy packages, as the U.S. government continues to support its advanced technology ecosystem.

Risks of the Guardrails

There is some evidence that the national security guardrails may produce a countervailing effect that partially impedes the CHIPS Act’s effectiveness. For semiconductor firms that receive CHIPS Act funding and maintain significant operations in China, the guardrails may force them to shift property, plants, and equipment out of the country and thus incur additional capital expenses and operational burdens while forgoing future revenue streams. Intel has raised precisely this concern. Examples abound, including the Taiwan Semiconductor Manufacturing Company (TSMC) facility in Nanjing, China, which produces advanced integrated circuits and feared massive losses after investing $2.9 billion in 2021 to expand its 28-nanometer chip capacity. As a result, the guardrails may impose direct costs on the very companies the CHIPS Act is designed to support, partially undercutting the financial benefits that U.S. industrial policy investment aims to provide to firms for investing in the United States.

  • Guardrails are prohibitive to some foreign companies. Some semiconductor firms perceive the guardrails’ associated costs as prohibitive and may limit their participation in CHIPS Act programs or refrain from participating altogether, in turn limiting investment in and expansion of the U.S. semiconductor industry. For example, leading South Korean chipmakers Samsung and SK Hynix maintain a significant portion of their production capacities in China, meaning their adherence to CHIPS Act guardrails may be infeasible. The two firms have signed a nonbinding preliminary memorandum of terms (PMT) with the Department of Commerce to begin qualifying for funding under the act. However, they were the last of the five major chipmakers to do so and have yet to negotiate a long-form term sheet and award documents, which require a “comprehensive due diligence process.” If the guardrails indeed prove overly restrictive, limited foreign participation may emerge throughout the U.S. semiconductor industry, which historically has proved critical to U.S. competitiveness and innovation. The CSIS Renewing American Innovation program has written on the guardrails’ potential impacts on Korean and Taiwanese semiconductor firms. Korean companies face significant risks. Samsung is currently manufacturing 128-layer NAND flash memory at its Xi’an facilities in China, where it has invested $26 billion. These plants contribute to 40 percent of Samsung’s total NAND production worldwide. Meanwhile, SK Hynix is producing dynamic random-access memory (DRAM) in the 10–20-nanometer range at its Wuxi facilities, along with a factory for 96- and 144-layer NAND flash memory in Dalian. Collectively, these operations represent a $20 billion investment and account for roughly half of SK Hynix’s DRAM output. TSMC is in the process of launching a 28-nanometer wafer fabrication plant in Nanjing, with a total investment of $2.9 billion. CEO C. C. Wei described 28-nanometer production as “the sweet spot for our embedded memory applications.” However, Taiwanese analysts are concerned that the proposed U.S. regulations could halt TSMC’s expansion plans in Nanjing.
  • The semiconductor supply chain remains globally interconnected. Even if these countervailing impacts do not meaningfully reduce investment in U.S. manufacturing and innovation, the impact of the subsidies’ guardrails on long-term U.S. positioning in global semiconductor markets may be mixed. The semiconductor industry is extremely globalized and complex, with supply chains involving design, materials, fabrication, and packaging that rely on hundreds, if not thousands, of companies to make a single chip. Additionally, chip markets depend on a wide range of downstream electronics, from smartphones and personal computers to cars. The manufacturing of these products is highly globalized and depends on complicated supply and demand dynamics. As a result, industrial policy efforts aimed at one nation’s semiconductor landscape can have wide-reaching, often unexpected ripple effects through the global chip market. Crucially, from the perspective of assessing the guardrails’ impacts, U.S. semiconductor companies remain deeply interconnected with the Chinese market. U.S. companies rely on Chinese companies and China-based facilities of multinationals for inputs they cannot source at scale elsewhere, such as critical minerals like gallium and germanium as well as services that China dominates—namely, semiconductor assembly, testing, and packaging. An immediate and wholesale shift of U.S. semiconductor supply chains out of China is not feasible, with devastating shortages of key civilian and military electronics an inevitable outcome. Additionally, Chinese companies remain critical customers for U.S. semiconductor companies—including integrated device manufacturers, fabless chip designers, and semiconductor toolmakers. The scale of China’s market for chip manufacturing and downstream electronics integration and for original equipment manufacturing makes the country an indispensable buyer of U.S. semiconductor technologies across the supply chain. As a result, the success of both U.S. and Chinese semiconductor supply chains (as well as their entire electronics supply chains) remains inextricably linked, creating challenges for any attempt to prop up one country’s competitiveness without supporting the development of the other—which is the implicit goal of the guardrails. This means, for one, that U.S. semiconductor companies benefitting from CHIPS Act subsidization will continue to directly and indirectly support Chinese companies, such as critical minerals refiners and outsourced semiconductor assembly and test firms, not to mention the significant operations of U.S. companies within China.
  • Guardrails apply to only one piece of the puzzle. Additionally, given the CHIPS Act’s focus on subsidizing development of fabs in the United States, a fundamental limitation of the guardrails is that they apply primarily to chip manufacturers. The resulting focus of the guardrails—restricting investments by chip manufacturers in Chinese production and R&D—addresses only one piece of the overall puzzle of diverse factors affecting the development of a robust semiconductor ecosystem. U.S. companies in fields like chip design or tools, such as deposition and etching, are crucial to the expansion of China’s semiconductor capabilities. For instance, Chinese-based chipmakers such as Semiconductor Manufacturing International Corporation, Hua Hong, and Yangtze Memory Technologies Corporation depend extensively on relationships with U.S. suppliers of underlying technologies and inputs. As a result, limitations on chipmaking investment will not necessarily limit the development of fabrication capabilities in China, which are supported by extensive government and corporate investment in indigenous manufacturing capabilities, often while working with U.S. suppliers unaffected by the guardrails.
  • Guardrails accelerate Beijing’s design-out efforts. The guardrails could also exacerbate China’s “design-out” efforts, wherein Chinese semiconductor companies reduce their reliance on U.S. firms and manufacturing facilities due to concerns about current and future U.S. restrictions (e.g., export and investment controls) and their impact on the security of relationships with U.S. companies. Design-out efforts have primarily been associated with U.S. export controls, which have created new political and commercial incentives within China to reduce reliance on U.S. technologies. The CHIPS Act’s national security guardrails risk furthering the perception that U.S. firms are unreliable long-term suppliers for China’s information and communications technology infrastructure. Specifically, design-out efforts allow Chinese and third-country (non-U.S. and non-Chinese) companies to supplant U.S. semiconductor champions in China and other key parts of the global chip market. China has invested heavily in its domestic semiconductor industry to fill gaps left by U.S. firms, develop its capacity in emerging technologies, and establish technological work-arounds. More broadly, the guardrails risk fueling perceptions that the United States seeks to “suppress” China’s rise, thus exacerbating geopolitical tensions and encouraging Chinese policymakers and global firms to de-risk, if not decouple. If this happens, U.S. firms stand to forego revenue that supports R&D investment, which is critical in the semiconductor industry given the rapid rate of technological change and the importance of maintaining leading-edge capabilities. Losses in R&D could mean U.S. companies losing the technology leadership they have in key parts of the semiconductor supply chain—the exact opposite of the goal of the CHIPS Act guardrails.
  • Guardrails create diplomatic strain. The strain imposed on diplomatic relations with key U.S. allies is another important consideration in evaluating the CHIPS Act’s national security guardrails. While U.S. negotiators successfully convinced the Netherlands and Japan to adopt new controls on advanced chipmaking technologies in early 2023, these controls do not apply as broadly as U.S. controls. Dutch and Japanese restrictions are still less stringent than the U.S. Entity List in terms of scope and enforcement infrastructure, thus creating gaps in the semiconductor export controls’ coverage. Amid this failure to effectively multilateralize U.S. controls and the reported breakdown of trilateral talks over the inclusion of technologies such as memory and mature logic chips in controls on chipmaking equipment, the United States appears to be alienating key allies while defaulting to a unilateral approach. Similarly, many U.S. allies and partners are skeptical of perceived U.S. protectionism, including stringent guardrails on industrial policy investments. Before the United States creates further national security requirements tied to industrial policy packages, it should assess how these instruments could bar partners from participating in U.S. industrial policy efforts.
     

Please consult the PDF for the annex.

Thibault Denamiel is a fellow with the Economics Program and Scholl Chair in International Business at the Center for Strategic and International Studies (CSIS). Evan Brown is the program coordinator and research assistant for the Economics Program and Scholl Chair in International Business at CSIS. David Korn is a former intern for the Scholl Chair in International Business at CSIS.

This report is made possible by generous support from Amazon.

Image
Thibault Denamiel
Fellow, Economics Program and Scholl Chair in International Business
Image
Evan Brown
Program Coordinator and Research Assistant, Economics Program and Scholl Chair in International Business

David Korn

Former Intern, Scholl Chair in International Business