The United States buys roughly $3 trillion worth of manufactured goods from abroad each year. A new McKinsey analysis suggests that reshoring even a fraction of that production could cost far more than most policymakers have bargained for.
Replacing imports of strategically critical goods would require approximately $2 trillion in capital investment, or about 6% of U.S. gross domestic product. That figure dwarfs the entire annual defense budget and excludes the additional billions needed to train workers, upgrade infrastructure, and secure reliable energy supplies.
The vulnerability cuts across multiple sectors. American manufacturers lack the capacity to produce enough advanced electronics, particularly AI servers, and face significant shortages in specialty chemicals. Textiles and apparel present an especially steep climb, while the country is in a relatively stronger position on fossil fuels and transportation equipment.
McKinsey researchers calculated what they call a "ramp-up index" for dozens of goods, measuring how much new industrial capacity the U.S. would need to fully replace current imports. The results vary wildly depending on the product. Among the roughly $750 billion in annual imports the firm classifies as critical to national security, supply-chain concentration, or sourced from geopolitical rivals, the investment required differs dramatically by category.
Recent policy efforts have begun moving the needle. The 2022 CHIPS and Science Act has spurred domestic capacity expansion in semiconductors. Foreign direct investment into manufacturing has surged. The Trump administration has elevated industrial reshoring to a top priority.
Yet the broader picture remains sobering. Shubham Singhal, who chairs the McKinsey Global Institute, notes that capital spending in AI-related manufacturing has accelerated sharply. "There have been pockets, including anything related to AI, especially with the big tech firms, where there has been a pretty dramatic scaling up of capital expenditures," he said. But across the wider industrial landscape, the spending surge hasn't materialized. "More broadly, if you look at the overall capital expenditure statistics, there hasn't been a dramatic tick up," Singhal added.
The capital question is only half the problem. Building actual manufacturing capacity requires time to construct facilities, develop supply chains for raw materials, and cultivate a skilled labor force. Investors are flooding money into AI because the profit potential appears clear. The same conviction hasn't taken hold in metals, chemicals, and other essential sectors where returns remain uncertain.
A major geopolitical shock, trade war, or supply-chain breakdown could expose just how unprepared the U.S. remains. Even with two administrations pushing reindustrialization, the distance between current capacity and strategic self-sufficiency remains substantial. The question now is whether policymakers and private investors can find a way to close that gap without waiting for a crisis to force the issue.
Author James Rodriguez: "Two trillion dollars is a staggering price tag for a partial fix, and it assumes investors will cooperate when returns outside of AI look murky at best."
Comments