The discourse surrounding US tariffs has been riddled with political concerns, leading to market instability. However, the implications are even more critical. It affects the future of American manufacturing. President Trump's tariff policy is not solely focused on reshaping trade and penalizing certain nations. It aims at “revitalizing the US manufacturing sector” and heralding “the revival of American industry,” as Trump stated.
Nonetheless, tariffs alone cannot revive the manufacturing base that has suffered from years of offshoring and lack of investment. If America wishes to regain its leadership role, it must look beyond the trade war and establish a genuine manufacturing ecosystem, starting with semiconductors—vital components in modern industry.
Take the manufacturing of a vehicle, for instance. The typical car today relies on thousands of semiconductors, with that number increasing as vehicles become electric and self-driving. If the US is unable to secure its own supply of these vital components, tariffs will not shield American factories from future shortages. More concerning, if the US falls behind in the production and advancement of these chips, dependence on foreign sources could transition from market complications to serious national security challenges.
Semiconductors are not just comparable to cars; they are akin to crude oil of the 21st century. Lacking them, the US economy faces a standstill, and our military risks losing its technological superiority. These tiny microchips facilitate the operation of industrial robots, power smartphones, enable aviation, and even run coffee machines. They are essential for advanced AI systems and the functioning of defense mechanisms.
The critical nature of semiconductors has long been acknowledged by Congress and various presidential administrations, yet policies have failed to keep the United States at the forefront of semiconductor manufacturing. Once, the US commanded 40% of global semiconductor production in 1990, but now it is down to only 10%. Furthermore, neither of these percentages include advanced chips essential for AI development. Intel, once a semiconductor leader in the US, has experienced decline.
To regain its semiconductor manufacturing and innovation supremacy, the US must implement a vigorous industrial strategy. This initiative should focus on mobilizing an Industrial Bank aimed strictly at investing in US manufacturing and enhancing research and development by cultivating a dedicated group of specialists to tackle pressing technology and manufacturing challenges.
The 2022 Chips Act was a step forward but only addressed symptoms, not root problems. It provided funds for research and was aimed at factories like TSMC’s Arizona facility, but failed to solve deeper issues. The US continues to lack the necessary infrastructure, talent, and financial resources to effectively compete with Asia.
Economic history shows that tariffs can have a role in industrial policy, but they do not rebuild industries. Tariffs must be part of a broader suite of policy measures for industrialization which includes targeted investments, manufacturing incentives, supportive tax regimes in strategic sectors, and the dissemination of technical and industrial knowledge.
From the 1960s through the 1980s, Japan, South Korea, and Taiwan successfully utilized tariffs to shield domestic industries while simultaneously enhancing them through funding, training, technological advancements, and production incentives. These nations experienced remarkable economic growth, resulting in the emergence of major corporations such as Toyota, Samsung, and TSMC.
This rationale leads us to reject the idea of removing tariffs and instead encourage a comprehensive overhaul. The US must prioritize semiconductor manufacturing similarly to how the Trump administration focused on steel and aluminum production and more recently, automobile production.
This undertaking requires significant capital investment, alongside research and development, as well as technical expertise. For instance, a single EUV lithography exposure tool can cost around $350 million. A typical factory might employ ten of these tools, all of which require operation and maintenance by highly skilled staff.
To remain competitive, the US requires a Marshall Plan-style mobilization focused on semiconductors, involving intensive capital, intellectual resources, and industrial policies.
Most importantly, funding is available. The costs associated with semiconductor manufacturing amount to billions, especially when establishing a single fab or factory. Recently, TSMC pledged $100 billion for semiconductor production, in addition to a previously committed $65 billion in the US. American firms urgently require access to such capital for production investments.
Secondly, research is crucial. A significant technical challenge in today’s semiconductor sector is managing electron leakage as transistors decrease in size and features become more densely arranged on a chip. The complexity of transistor design also complicates manufacturing processes. TSMC and other leading manufacturers have indicated new processes will be unveiled later this year, but these methods are intricate and necessitate advanced equipment. Future advancements will require physics breakthroughs, achievable only through enhanced research facilities and funding.
Thirdly, education is critical. The semiconductor manufacturing field needs skilled professionals to emerge. Establishing such a workforce involves recruiting at least 200 individuals with advanced degrees in math and science, along with equivalent training and skills applicable to manufacturing.
The decision is clear-cut. Should America continue to appeal to TSMC to construct factories on US soil, or will it regain the capacity to manufacture chips that shape the future? Tariffs can offer protection, but only a robust industrial strategy can lead to true reconstruction.
Time is of the essence—it is crucial to act before the next significant breakthrough occurs in Shanghai instead of Texas.
Macabe Keliher is an associate professor at the Clements School of History at SMU Dallas, focusing on industrialization and manufacturing in East Asia.Dev Gupta, PhD, is the founder and CTO of APSTL LLC in Scottsdale, Arizona. His past innovations in semiconductor production at AZ Motorola and Intel are currently widely utilized in AI hardware. He chairs the IEEE International Roadmap section for semiconductor devices.
