Press releases can’t fully capture the scope of the ambition when you stand outside the construction site for TSMC’s semiconductor factory in Phoenix, Arizona. The cranes are huge. There are a lot of earthworks. For what is theoretically a single factory, the surrounding fencing extends farther than one might anticipate. Additionally, it is over budget and behind schedule due to a number of issues that its Taiwanese engineers apparently did not fully anticipate when the project was announced with considerable fanfare in 2020.
These issues include a lack of workers with the specific skills needed, regulatory processes moving at their own pace, and construction costs that kept rising. The situation at TSMC Phoenix is not a failure. However, it serves as an example of something that the global supply chain reshore debate tends to overlook: this is actually difficult and costly in ways that aren’t completely represented in the policy speeches.
Reshoring Western manufacturing makes political sense, and it’s simple enough to put on a bumper sticker. Shortages of semiconductors, medical equipment, and basic industrial components during the COVID-19 pandemic exposed how profoundly supply chains had been optimized for cost effectiveness at the expense of redundancy.
Hospital procurement offices in Birmingham and auto dealerships in Detroit were affected when manufacturing in Wuhan went dark. The lesson seemed obvious: if manufacturing is concentrated in too few locations, the system as a whole becomes unstable. Bring the manufacturing process home. Minimize reliance. The reasoning is sound. The economics of carrying it out are much more intricate.
It is hard to overestimate the cost premium that greenfield industrial building in the US or Western Europe entails over comparable facilities in China or Southeast Asia. Purchasing land in areas with pre-existing industrial infrastructure is costly. Over the previous four years, labor and building material expenses have increased dramatically. Environmental and safety regulations, which are actually significant rather than the result of excessive bureaucracy, add requirements that aren’t present in all competitive jurisdictions.
The permitting procedure then starts. Requirements for community consultation, zoning evaluations, and environmental impact studies are all valid components of democratic governance, but they take time. For many industrial projects in Western markets, it is conservative to estimate that it will take an extra twelve to thirty-six months before a single unit of production is feasible. During those months, there are opportunity costs, holding costs, and financing costs that compound before a machine is turned on.
Reshoring proponents occasionally discuss the labor dimension more optimistically than the data warrants. It is true that during the past ten years, wages in China and Southeast Asia have increased significantly. Direct labor cost parity is still a long way off because of the wage differential between those earnings and what a unionized American or German manufacturing worker makes. Automation—sophisticated robots, computer vision systems, and automated assembly lines that lower the workforce needed per unit of output—is the solution for the majority of businesses who are actively seeking reshoring.
That response is genuine and becoming more and more feasible. It also needs its own long installation schedule, its own specialized maintenance staff, and its own capital investment. The factory that uses automation to manufacture competitively is not less expensive to construct than the Guangzhou facility it replaces. After the initial investment is made, it simply becomes less expensive to run.
Historically, a significant portion of this deficit has been filled by government subsidy programs. Over $50 billion was allocated to domestic semiconductor production by the CHIPS and Science Act. This amount may seem significant when you take into account that the construction and equipment of a single modern chip fabrication plant can cost up to $20 billion. The Inflation Reduction Act attracted investment from businesses that may have found additional capacity in Asia by extending tax credits totaling hundreds of billions for green technologies and EV battery manufacturing.
Under multiple “Made in Europe” frameworks, Europe has invested its own reindustrialization capital in digital manufacturing systems and supply chain infrastructure. In the limited sense that they are encouraging investment that would not otherwise occur at this rate, these subsidies are real incentives. It’s still unclear if they’re big enough to eventually overcome the fundamental cost disparities.
