How to Make the U.S. Semiconductor Industry Boom Again

Podcast Summary

• Semiconductor Industry: Shaped by Government Policies: Government policies have significantly influenced the semiconductor industry's growth and continue to shape its future, particularly in the US, Taiwan, and Japan.

  The semiconductor industry is shaped significantly by government policies, making it an intriguing topic for exploration. Principal Asset Management, with their global perspective and local insights, recognizes the importance of identifying compelling investing opportunities in this sector. The semiconductor industry's history is marked by government support, particularly in the US, Taiwan, and Japan. This involvement continues today, as governments play a more active role in shaping their economies. The semiconductor industry's intersection with technology, business competition, and government policy makes it a fascinating subject for analysis. In this episode of the Odd Lots podcast, the hosts discuss the history of semiconductor policy in the US and potential policies that could revive the domestic industry. With the help of economist Alex Williams and technologist Hassan Khan, they delve into the macro and micro aspects of the industry, providing valuable insights for listeners.

• Intersection of Industrial Policy, Trade Policy, and Labor Markets through Semiconductors: Industrial policy can help maintain a competitive edge by directing resources to areas with latent expertise, such as semiconductors, and mitigate the negative effects of offshoring on labor markets.

  Semiconductors serve as an important lens through which to view the intersection of industrial policy, trade policy, and labor markets. As a labor market-focused institution, Employ America recognizes the importance of upskilling and economic policy targeted at specific industrial objectives. Industrial policy, by nature, has a labor market focus, as it aims to direct the economy to make the best use of resources even if short-term profit considerations don't align. The semiconductor industry is particularly relevant in this context, as the offshoring of production has led to the loss of latent expertise within the economy. This expertise, housed in both scientific research departments and blue-collar workers, is crucial for maintaining a competitive edge in the industry. The bipartisan support for addressing the semiconductor industry's challenges presents an opportunity for reflection on past debates and the potential benefits of industrial policy interventions.

• The role of human capital in pushing technology forward: During economic downturns, valuable human capital can be lost, leading to inefficient use of resources. Science policy and industrial policy play distinct roles: science policy funds research, while industrial policy translates research into commercializable products and maintains a booming industry.

  Human capital plays a crucial role in pushing technology forward, not just through academic research and invention, but also through practical implementation on the shop floor. However, when industries experience labor market attrition during economic downturns, valuable human capital can be lost, leading to an inefficient use of resources. The distinction between industrial and science policy is important to understand in this context. While science policy focuses on funding research and development, industrial policy addresses the translation of research into commercializable products and the maintenance of a booming industry for the incorporation of new technology into production. The policy mistake in the past was the belief that funding R&D alone would maintain America's technological lead, neglecting the importance of the industrial aspect. This disconnect between research and commercialization can lead to wasted resources and a failure to translate cutting-edge research into practical applications.

• Collaboration between science, industry, and government during the 1960s and 1970s: Strong collaboration between science, industry, and government drove US semiconductor technology leadership, but later efforts were limited to research and trade measures due to Japanese competition. US regained market share in the early 1990s but then internationalized institutions, allowing competitors insights into future directions.

  The collaboration between science and industry in the semiconductor industry during the 1960s and 1970s, driven by military interest and a de facto industrial policy, helped establish the US as a leader in semiconductor technology. However, in the late 1970s and 1980s, as Japanese firms began to compete, the industry primarily sought help through collaborative research and trade measures. The government's role remained limited, and institutions like the Semiconductor Research Corporation (SRC), SEMATECH, and the National Technology Roadmap for Semiconductors were established. In the early 1990s, US semiconductor firms regained market share, but the US then began to wind down these institutions and internationalize them, allowing competitors outside the US to gain insights into the industry's future directions and processes. This history illustrates the importance of a strong collaboration between science, industry, and government in driving technological innovation and maintaining competitiveness.

• Japan vs US R&D Responsibilities in Semiconductor Industry: Japan's focused approach on specific tech targets led to rapid progress but lacked diversity and potential for innovation. US firms' reliance on foundries led to increased collaboration but lacked redundancies and potential for innovation. Both approaches had risks and limitations.

  The distribution of research and development responsibilities among firms in the semiconductor industry, as seen in the case of Japan and the US in the late 20th century, can lead to a self-reinforcing situation of rapid progress but also lack of redundancies and potential for innovation. Japanese firms, driven by MITI's industrial policy, focused on specific technology targets and forced collaborations, leading to significant advancements and market share in areas like DRAM. However, this approach lacked the diversity and potential for innovation that comes with having multiple firms working on different aspects of technology. Contrastingly, US firms, facing increasing costs in traditional manufacturing processes, pulled out of manufacturing and relied more on foundries, leading to increased CapEx spend and collaboration. While this worked well in the near term, it lacked the redundancies and potential for innovation that come with having multiple firms working on different aspects of technology. The failure of Japan's 5th generation supercomputing project and the subsequent economic stagnation in Japan illustrate the risks of relying too heavily on a single approach to industrial development.

• The role of science policy in maintaining industrial dominance: The US semiconductor industry's loss of dominance resulted from a separation of design and manufacturing, influenced by a generational change and financialization. To regain competitiveness, a focus on developing a robust manufacturing ecosystem, including EUV lithography machines, is necessary.

  Industrial policy plays a crucial role in industrial development, particularly when countries aim to catch up with technological front-runners. However, establishing long-term dominance requires not only hitting targets but also imagining new ones, which is where science policy comes in. The US semiconductor industry's inability to maintain dominance can be attributed to a separation of design and manufacturing, with design remaining in the US while manufacturing moved overseas. This shift was influenced by a generational change in industry leadership and the financialized approach that came with it. Today, the US semiconductor industry exhibits pockets of excellence but lacks a robust manufacturing ecosystem, specifically in leading-edge semiconductor manufacturing. The absence of EUV lithography machines in the US is a notable weakness. Understanding the historical context and current challenges is essential to developing strategies for improving the US semiconductor industry's competitiveness.

• Hindered innovation in semiconductor industry due to lack of competition and weak economy: The semiconductor industry's lack of competition and weak economy hinder innovation, resulting in limited learning opportunities and slower technological progress, which can negatively impact employment and the economy in the long term.

  The lack of a competitive ecosystem and weak economic growth have hindered the innovation and advancement of technology in the semiconductor industry. With only one leading company in the US using advanced technology and limited learning opportunities between manufacturers, suppliers, and design firms, innovation is focused on areas where the company already leads. Industrial policy can help by introducing support and supply lines, allowing the industry to build on its strengths. However, a sluggish economy and poor effective demand can discourage manufacturers from investing in new capital goods and technology, resulting in slower technological progress and a less skilled labor force. This lack of investment in new technology and production capacity can have long-term negative effects on employment and the overall economy.

• Building a resilient semiconductor ecosystem: The US needs to gather data, identify weaknesses, and implement targeted policies to build a more resilient semiconductor ecosystem, including purchase guarantees and support mechanisms, while learning from successful ecosystems like Shenzhen.

  The US government needs to take a comprehensive approach to identifying vulnerabilities in the semiconductor supply chain and building a resilient ecosystem. This involves gathering data to understand the current state of supply chains, identifying areas of weakness such as a weakened manufacturing innovation ecosystem and reliance on individual suppliers, and implementing targeted policies to address these issues. The Biden administration's supply chain review is a good first step, but more needs to be done to maintain internal capabilities and support a competitive semiconductor industry. This may include purchase guarantees for specific markets, as well as other support mechanisms. The US needs to learn from places like Shenzhen, where a strong ecosystem for manufacturing innovation allows for the sharing of information and the development of new technologies. By addressing these weaknesses, the US can build a more resilient semiconductor ecosystem and reduce vulnerabilities in its industrial supply chains.

• Investing in Domestic Industrial Capacity: The US needs to prioritize and invest in domestic industrial capacity, especially in the semiconductor industry, to maintain economic competitiveness and ensure resilience against global market shifts.

  The US needs to prioritize reviewing and strengthening its industrial supply chains, specifically in the semiconductor industry, to maintain economic competitiveness and ensure resilience against global market shifts. The current industrial policy proposals, such as The Endless Frontiers Act and the CHIPS Act, while a step in the right direction, do not fully address the need for active monitoring and investment in domestic industrial capacity. The ideal solution would be a comprehensive infrastructure plan, like the one proposed by the Biden administration, which includes significant funding for monitoring and supporting domestic industrial capacity. This is crucial not only for the semiconductor industry but also for adapting the economy to climate change and other challenges. Market fundamentalists arguing against government intervention should recognize that the global semiconductor industry is not a free market, with national governments actively subsidizing and protecting their industries. To maintain a competitive and resilient ecosystem, the US needs to invest in its domestic industrial capacity and adapt to the changing economic landscape.

• Decay of US production know-how due to outsourcing, critical dependence on foreign chip suppliers: Policymakers underestimated the impact of outsourcing high-tech manufacturing, leading to a critical dependence on foreign chip suppliers. Bipartisan consensus to reverse trend, but success depends on clear goals, merging proposals, and supply chain monitoring.

  The outsourcing of high-tech manufacturing to other countries, particularly Taiwan, for cost savings, has led to a decay of production know-how in the United States. This trend, which was not fully anticipated by policymakers at the time, has resulted in a critical dependence on foreign chip suppliers. Now, there is a bipartisan consensus to reverse this trend, but the success of any policy initiative depends on the right choices being made. From a policy perspective, this involves setting clear goals for the US semiconductor ecosystem, merging specific proposals together, and building up supply chain monitoring muscle to anticipate future challenges. It's important to note that this issue goes beyond the semiconductor industry and has implications for addressing future challenges, such as climate change. The next two pieces in this series will explore the industrial policy toolkit for governments and argue against the Ricardian comparative advantage approach to trade policy.

• The Importance of a Multi-Pronged Approach to Technology and Industrial Policy: A successful revival of the US semiconductor industry requires a multi-pronged approach, including funding for R&D and identifying demand, acknowledging the complexity of the industry, and fostering creativity and innovation through government involvement.

  Technology and industrial policy are interconnected, with technology being a tool for innovation and industrial policy serving as a backstop to support and identify demand for new technologies. The discussion emphasized the importance of a multi-pronged approach, including funding for R&D and identifying demand, and acknowledged the complexity of the semiconductor industry and the challenges of reviving the US industry. The speakers also highlighted the potential for government involvement to foster creativity and innovation rather than stifling it. The conversation underscored the significance of getting policy right and the ongoing nature of the semiconductor industry's evolution. In the coming decades, the success of US efforts to revive its semiconductor industry will likely be a topic of ongoing discussion. To learn more about this topic, follow Alex Williams on Twitter @tragicbios and Hasan Khan @hasankhan, and check out their writing on the Employ America medium page or website. Don't forget to follow our producer, Laura Carlson @LauraMCarlson, and the Bloomberg head of podcast, Francesca Levy @FrancescaToday. Listen to Money Stuff, the new podcast from Matt Levine and Katie Greifeld, every Friday on Apple Podcasts, Spotify, or wherever you get your podcasts.