American Heavy Industry Is Not Losing to China.  It Is Losing to Its Own Power Bill.

By Dale J Hobbie, Founder, Managing Member & Chief Scientist, Quantum HPC Infrastructure, LLC

The political conversation about American industrial decline has a standard cast of villains.  Cheap Chinese labor.  Predatory trade policy.  Regulatory overreach.  Currency manipulation.  These are real factors, and the policy debate around them is legitimate.  But they are not the structural constraint that is actually stopping American manufacturers from committing capital to domestic production expansion right now, in 2026, on the timelines that the reshoring moment demands.

The constraint is the power bill.  Not the rate on the bill, though that matters.  The certainty of the bill.  The ability to look at a 20-year capital commitment for a new steel mill, a semiconductor fab, an aluminum smelter, or a defense-critical materials facility, and know with contractual confidence what energy is going to cost for the life of the asset.

That certainty does not exist in the United States today for most industrial operators evaluating new domestic production capacity.  And without it, the economics of reshoring do not close regardless of what the tariff schedule says.

Here is the problem stated plainly.  A heavy industrial facility is not a software company.  You cannot depreciate it in three years and pivot.  A greenfield steel mill represents a capital commitment measured in billions of dollars and an asset life measured in decades.  The financial model that justifies that commitment requires assumptions about input costs, and in energy-intensive industries, power is the input cost that moves the model more than any other single variable.  A 20% swing in electricity cost over the life of a facility is not a line item variance.  It is the difference between a profitable domestic operation and a structurally uncompetitive one.

The American grid cannot currently provide the cost certainty that model requires, and the trajectory is not improving.  Wholesale electricity costs near major industrial and data center clusters have risen sharply across every major U.S. grid region in the past three years.  The PJM capacity price increase from roughly $29 per megawatt-day in 2024 to over $329 in 2026 is the most dramatic single example, but it is not an outlier.  It is a leading indicator of what happens when a grid built for a different century absorbs industrial-scale new load without the generation and transmission infrastructure to support it.

The political response to this has been, predictably, to treat it as a grid problem requiring a grid solution.  Build more transmission.  Permit more generation.  Reform the capacity market.  Accelerate interconnection.  These are not wrong answers.  However, they are correct answers on a timeline that does not match the investment decisions that American manufacturers need to make right now.

An industrial operator who needs a facility running in 2028 cannot wait for transmission upgrades that the utility has projected for 2032.  An advanced materials manufacturer who needs to lock in energy cost assumptions for a project financing that closes next quarter cannot base those assumptions on a grid that has demonstrated, repeatedly and recently, that it can reprice by an order of magnitude between one capacity auction and the next.

The operators who understand this are already making different decisions.  They are not waiting for the grid to get better.  They are decoupling their energy supply from the grid entirely, or as close to entirely as their specific industrial process allows, and they are locking in their power economics through on-site generation arrangements that the utility interconnection queue cannot touch and the capacity market cannot reprice.

This is not a new concept.  On-site industrial power generation has existed for as long as heavy industry has existed.  What is new is the combination of factors making it not just viable but necessary at a scale and across an industrial range that would have seemed impractical a decade ago. The technology options have matured.  The grid cost trajectory has made the comparative economics more compelling than they have ever been.  And the policy environment, with reshoring incentives, domestic content requirements, and defense production priorities all creating new demand for domestic industrial capacity on accelerated timelines, has created a class of industrial tenants whose project economics genuinely cannot absorb grid uncertainty.

The manufacturers, developers, and industrial operators who are building their power supply as a fixed-cost, long-duration, on-site resource, rather than treating it as a utility service they receive from someone else at whatever price the market sets, are not making a bet on energy policy.  They are removing energy policy from their risk register entirely.  That is a fundamentally different competitive position than the one their grid-dependent competitors occupy.

The reshoring narrative in Washington is largely about tariffs, tax credits, and supply chain.  Those are the levers that legislators can pull and take credit for.  The energy certainty problem is harder to legislate and harder to claim credit for solving, so it gets less attention than it deserves.  But it is the problem that will determine which reshoring commitments turn into operating facilities and which ones remain announcements.

American heavy industry is not losing to China because Chinese workers are cheaper.  It is losing, where it is losing, because the capital committed to domestic production cannot price its energy input for the life of the asset.  Fix that, and the rest of the competitiveness argument gets a lot easier to make.

The manufacturers who figure that out first will not just be more competitive.  They will be in a different industry than the ones still waiting for the grid to solve their problem.

Dale Hobbie is the Founder, Managing Member & Chief Scientist of Quantum HPC Infrastructure, LLC, a grid-independent power, cooling, and compute infrastructure company based in Avon Lake, Ohio.  He is a former mission critical analyst for the intelligence community.