When a Treasury Secretary travels to Asia to discuss supply chains, a specific type of institutional urgency usually arises. It didn’t feel like a diplomatic formality when Janet Yellen told the gathered audience at LG Sciencepark in Seoul that the US could not permit China to abuse its hold on vital technologies. It seemed like a long-awaited warning.
Yellen’s main point was fairly simple: nations that rely on a single dominant supplier for anything necessary, such as rare materials, electric car batteries, or semiconductors, are leaving themselves vulnerable. That lesson had become viscerally real due to Russia’s invasion of Ukraine. The same argument was being made, albeit more slowly, by China’s methodical accumulation of manufacturing dominance in cutting-edge chips.
In Washington, the concept of “friend-shoring”—moving production and procurement toward reliable partner countries instead of just going with the cheapest option—was already gaining traction. However, Yellen gave it weight by presenting it as a legitimate economic defense rather than as industrial nostalgia. The initial real funding for the concept was supposed to be the approximately $50 billion in semiconductor funding that was being pushed through Congress at the time. It was another matter entirely whether that money could fulfill its promise.
To understand the scope of what reshoring semiconductor manufacturing truly requires, it’s important to take a step back. A chip manufacturing plant is not a warehouse that can be disposed of in eighteen months. Hundreds of ultra-high-purity chemicals, some of which are measured in parts per trillion, must arrive in this controlled environment on schedule, within specification, and without interruption. Production slows down if one input is missed.
A large portion of that chemical supply, including nearly all of the ultra-high-purity hydrogen fluoride needed to etch silicon, has long been imported by the United States from Asia. High capital costs, a lack of technical know-how, and a schedule that doesn’t fit neatly into any political cycle are all obstacles to building domestic capacity for those inputs.
According to McKinsey’s analysis of U.S. semiconductor supply chains, domestic supply is just insufficient to meet the demand that would result from a significant reshoring push for over half of the chemicals that factories require. The U.S. semiconductor market is predicted to almost double in value to approximately $140 billion by 2030. The necessary chemicals to sustain that growth are not present.
The talent issue is another issue that receives very little attention. An additional $250 billion in investment could be absorbed by the sector, generating about 160,000 new jobs for engineers, technicians, construction workers, and tradespeople. However, only about 1,500 engineers are joining the semiconductor industry annually. The industry might require 88,000 by 2029. Declarations of policy do not close that gap. It closes through consistent, unglamorous investment in community colleges, training pipelines, apprenticeship programs, and hiring from related industries like advanced manufacturing and pharmaceuticals.
Washington’s understanding of the distinction between funding an ecosystem and funding a factory is still lacking. Although companies like Intel and TSMC have made significant investments in U.S. factories, the underlying supply chain—which includes the chemicals, tools, and talent—remains heavily reliant on international networks that were not designed to be easily replicated. Taiwan, for its part, has already refuted claims that 40% of its chip capacity could be easily moved to the US, stating that the notion is unfeasible.
All of this does not imply that the movement to reshore semiconductor manufacturing is incorrect. Reducing single points of failure in such crucial industries makes sense strategically. Building resilience, however, is not the same as believing that decades of accumulated industrial geography can be condensed into five years with a few billion dollars in federal incentives. Instead of attempting to rebuild everything domestically from scratch, Yellen’s instinct to include allies in the solution—South Korea, Japan, the Netherlands, and others with deep chip expertise—is likely the more practical course of action.
The supply chain pivot has a genuine goal. And so is the challenge of carrying it out.
