I. Introduction
Every few months, another headline declares that America is “reindustrializing.” Reshoring announcements, CHIPS Act press releases, a new gigafactory groundbreaking in some state that hasn’t seen a factory built in a generation. The instinct behind all of it is correct: a nation that cannot make things cannot defend itself, cannot control its own supply chains, and cannot claim, credibly, to be a serious industrial power. But the way we talk about the problem is not entirely correct, and the imprecision is starting to cost us.
We keep using “manufacturing” as a catch-all for two entirely different disciplines: engineering a product, and making that product a thousand or a million times over. These are not points on the same continuum. They are different problems, solved by different people, with different tools, different timelines, and different definitions of success. Conflating them is not a semantic nitpick, but rather the reason so much of the current reindustrialization push, for all its energy and capital, keeps producing prototypes instead of production lines.
II. Two Problems Wearing One Name
Engineering is the discipline of getting to “it works.” You are solving for correctness under uncertainty: does the design meet spec, does the prototype survive the test, does the system behave the way the model predicted. The unit of success is one. A brilliant engineering team can build a single flawless unit of almost anything, given enough time and enough hands adjusting it along the way.
Manufacturing is the discipline of getting to “it works, again, at cost, on schedule, ten thousand times in a row, with three different shifts of workers who didn’t build the original prototype.” The unit of success is not one. It is the distribution around the mean: yield rates, tolerance stack-up across suppliers, cycle time, first-pass quality, the behavior of a process when the operator is having a bad day. None of this shows up in a CAD file. It shows up on the shop floor, and only after you have run the process enough times to see where it breaks.
Gary Pisano and Willy Shih, in their work on America’s industrial decline, gave this gap a name: the industrial commons¹. It is the accumulated, largely tacit knowledge of how to actually produce something at volume: the tooling suppliers, the process engineers, the machinists who know why a particular fixture warps under heat, the supply base that can hold a tolerance nobody wrote down because everybody just knows it. You cannot download this knowledge. You cannot buy it in a software subscription. It is built, slowly, by running real production, and it decays quickly when production stops.
III. How America Kept One Muscle and Let the Other Atrophy
For most of the twentieth century, American industrial giants did not separate these two problems. Ford’s River Rouge complex designed and built the car under one roof. GE’s engineers sat down the hall from the people running the line. When something failed in production, the engineer who designed it walked over and looked at it. The feedback loop between design and production was physically short, and that proximity is precisely what built the industrial commons in the first place.
Starting in the 1980s and accelerating through the 2000s, American firms began pulling that loop apart. Design stayed in Cupertino, Seattle, and Boston. Production moved to Shenzhen, and later to a dozen other cities most Americans couldn’t find on a map. This was, for a long time, treated as a purely financial decision: labor was cheaper elsewhere, so production followed the cost curve. What got missed, or ignored, is that the decoupling was not just geographic. It severed the feedback loop that had built and maintained the industrial commons for a century, and once severed, the commons on the American side began to erode: the tooling shops closed, the process engineers retired without training replacements, the machinists’ apprenticeship pipelines thinned out².
What America kept was the engineering half. The country still produces extraordinary design talent: national labs, defense primes, a fabless semiconductor industry that designs the world’s most advanced chips and ships the actual fabrication overseas, a software culture that treats iteration as a birthright. What America lost, in large part, was the muscle memory of turning a design into a million reliable units on domestic soil. We can still engineer almost anything. We have, in too many sectors, forgotten how to make it.
IV. Why the Conflation Produces Bad Strategy
This distinction is not academic. It explains why so much of the current reindustrialization effort underperforms its own press releases.
Subsidize “innovation” without addressing production capacity, and you get brilliant prototypes that never leave the lab, because nobody built the process engineering discipline to scale them. Build a greenfield mega-factory without embedding real engineering talent into its operation, and you get an expensive, rigid asset that can produce exactly one thing, exactly one way, with no capacity to iterate when the product or the geopolitics change³. Treat “build more factories” as a single policy lever, and you will keep funding half of the problem while wondering why the other half never shows up.
The nations that are currently out-executing the United States in manufacturing did not solve this by picking one discipline over the other. They built institutions, often deliberately, that treat production engineering as its own craft: a discipline with its own training pipelines, its own career ladders, its own prestige, sitting in continuous, physical proximity to the factories where the actual making happens⁴.
V. What Rebuilding the Muscle Actually Requires
If the industrial commons was lost through decoupling, it will not be rebuilt through a subsidy check or a ribbon-cutting. It has to be rebuilt the way it was built the first time: by putting engineering discipline back inside working factories, and doing it factory by factory, process by process, until the tacit knowledge accumulates again.
That is a different project than starting new companies to design new products. It requires going to where the residual manufacturing capacity already exists, in the thousands of small and mid-sized American manufacturers that never fully lost the skill, and rebuilding around them rather than around them. It means acquiring operating businesses instead of building greenfield ones, because an operating business already has the one thing you cannot manufacture from scratch: a working process, run by people who know why it works.
This is the problem Ironstead was built to solve. We are not a product company, and we are not a design shop. We acquire profitable, operating manufacturers and go to work on the half of the problem that gets skipped: the discipline, the training, the process. Only once that is in place do we touch the software or the automation. We start with the assets, not the pitch deck, because the industrial commons cannot be simulated. It has to be operated back into existence.
VI. Conclusion
America does not have an engineering deficit. It has a translation deficit: the gap between a design that works once and a production line that works every time, run by people who were trained to keep it working. Closing that gap is not a matter of national willpower or a single piece of legislation. It is a matter of correctly identifying which problem you are solving, engineering or manufacturing, and building the institutions that specialize in each.
Reindustrialization will not be won by companies that are excellent at one and hope the other follows. It will be won by the people willing to do the less glamorous work of rebuilding the industrial commons factory by factory, shift by shift, until making things at scale is, once again, something this country simply knows how to do.
References
- Pisano, Gary P., and Willy C. Shih. Producing Prosperity: Why America Needs a Manufacturing Renaissance. Harvard Business Review Press, 2012.
- Autor, David, David Dorn, and Gordon Hanson. “The China Shock: Learning from Labor Market Adjustment to Large Changes in Trade.” Econometrica, 2016.
- Womack, James P., Daniel T. Jones, and Daniel Roos. The Machine That Changed the World. Free Press, 1990.
- McKinsey Global Institute. “Reshaping U.S. Manufacturing Competitiveness.” 2023.
George Payne 