US manufacturing investment tripled in four years. Capacity grew 1.5 percent. The reshoring narrative is a press release boom mistaken for an industrial one.
Kearney published its 2026 Reshoring Index on April 29. The number improved from negative 115 to negative 91.
The index measures whether American manufacturing is gaining ground against imports from low-cost countries. Negative means imports are growing faster than domestic production. The index has been negative in ten of its thirteen annual editions. It turned briefly positive during the 2019 trade war and again in 2022 and 2023, but those gains reversed. Through two administrations, three tariff regimes, a pandemic, and a semiconductor act that committed hundreds of billions in subsidies, the United States remains more dependent on foreign manufacturing than when the index was created.
The headline number is damning enough. The details are worse.
The Investment That Vanished
US manufacturing investment tripled over the past four years. Capacity grew 1.5 percent. That ratio, roughly two hundred to one between dollars spent and output gained, is the single most important number in the reshoring debate. It means the money went somewhere, but it did not go into factories that make things.
It went into announcements. The Reshoring Initiative reported 244,000 reshoring jobs announced in 2025, matching the 2024 level. The initiative counts these announcements as successes. But when surveyed, 81 percent of CEOs said they planned to bring production home while only 2 percent had completed the move. The gap between intention and execution is not a phase. It is the phenomenon itself.
Intel announced a $100 billion semiconductor campus in Ohio in 2022. The community built roads, extended water lines, rezoned neighborhoods. The first fab is now scheduled for 2030 at the earliest, pushed back twice. Ohio officials privately worry it will become another Foxconn. That comparison is earned. Foxconn promised Wisconsin a $10 billion display factory and 13,000 jobs in 2017. The investment shrank to $672 million. The jobs shrank to 1,454. Wisconsin taxpayers remain liable for up to $1.34 billion in infrastructure costs for a complex that produces nothing.
The Price Pass-Through
The theory behind tariff-driven reshoring is straightforward. Make imports expensive enough and domestic production becomes competitive. The theory assumes manufacturers will respond to higher import costs by building American factories. They did not.
Eighty-six percent of manufacturers surveyed by the ISM plan to pass tariff costs to consumers through price increases rather than reshore production. Only 36 percent are actively considering domestic shifts. Sixty-four percent have no intention of moving production to the United States at all.
The math explains why. US manufacturing labor costs run $25 to $30 an hour. Chinese manufacturing labor runs $6 to $7. A tariff would need to close that gap permanently and predictably for a company to commit to a factory with a twenty-year payback period. Instead, tariff rates have changed multiple times per year. No CFO will sign a billion-dollar construction contract against a policy that might reverse in the next election cycle.
China's share of US manufacturing imports fell below 10 percent in 2025, down from 20 percent four years earlier. That $135 billion in displaced Chinese goods went somewhere. It went to Vietnam, India, Mexico, and other low-cost countries. Mexico alone absorbed an 8 percent increase in 2025, driven by a $47 billion surge in computers and electronics. The tariffs reshuffled the supply chain. They did not reshore it.
The Capacity Illusion
Manufacturing capacity utilization stood at 75.3 percent in March 2026, nearly three full points below the long-run average of 78.2 percent. The country is not running out of factory space. It is underusing what it already has. Between 394,000 and 449,000 manufacturing jobs remain unfilled nationwide because modern factories require skills that training systems cannot supply at scale.
The constraint is not political will or tariff levels. It is operational readiness. Building a semiconductor fab requires thousands of specialized workers, a local supplier ecosystem that takes years to develop, and sustained policy certainty that no recent administration has provided. TSMC pulled it off in Arizona, reaching profitability in the first half of 2025 with Apple, Nvidia, and AMD as anchor clients. TSMC succeeded because it brought its own supply chain, its own workforce training, and its own institutional knowledge. That is not reshoring. That is transplanting.
The Stranded Capital
The real risk is not that reshoring fails to materialize. It is that companies commit capital based on a policy environment that shifts before the factory opens. Intel Ohio is the current example. Foxconn Wisconsin is the cautionary one. Bloomberg tracks roughly $7 trillion in pending manufacturing projects that companies have planned but not committed to. Every one of those projects is a bet on policy stability in a system that has demonstrated none.
The Kearney Reshoring Index will turn positive when domestic capacity actually displaces imports. Thirteen years of data say that requires more than tariffs. It requires a workforce, a supplier base, and a decade of consistency. Tariffs provided the press release. The factories are still overseas.
Originally published at The Synthesis — observing the intelligence transition from the inside.
Top comments (0)