Import substitution is not a new idea, and that is part of the problem. The phrase carries a lot of history. It became an important part of postwar development economics, especially in countries trying to build manufacturing capacity of their own. Underneath all the theory was a plain question: if a country was importing manufactured goods, could some of those products be made at home instead?
I do not think that question was foolish. I think it remains one of the most practical industrial questions a country can ask.
What Is Import Substitution?
Put simply, import substitution means replacing selected imported goods with domestic production where American manufacturers can compete. Not everything. Not at any cost. Not as a slogan. The useful version starts with product level import signals, domestic supplier discovery, and a clear-eyed look at which imported products could realistically become American production.
That definition matters because the older meaning of the term comes with baggage. Economists such as Raul Prebisch believed developing countries needed a path to industrialize rather than remain stuck exporting raw materials and buying finished goods from someone else. His argument was not as crude as the caricature. He did not say countries should stop trading with the world. He understood the need for exports, machinery, technology, and capital goods.
But practice and theory parted ways. In too many places, import substitution became a wall: tariffs, import restrictions, licensing, and protection that lasted too long. Some industries grew. Others got comfortable, stayed small, or became too insulated from the pressure that makes companies better. By the 1960s, even some early supporters were warning that too much protection could create a new set of problems.
That history matters. We should be honest about it. But it should not scare us away from the original question.
That is still import substitution.
What has changed is how precisely we can answer the question.
The Question Is Still Worth Asking
Earlier approaches often worked at the industry level. Policymakers made big decisions about broad sectors without anything close to the product visibility we have now. Today, we can begin with actual demand. We can see what American companies are already buying from overseas, how often they are buying it, where those products are coming from, and which products show up again and again over time.
That is the break from Prebisch's day. We now have national manufacturing data, product level import signals, domestic supplier information, and digital traces of industrial capability that previous generations simply could not see. Imports are still classified through systems such as the Harmonized Tariff Schedule, and the U.S. Census Bureau and Bureau of Economic Analysis still publish regular trade data on imports, exports, and balances. But the bigger change is that those datasets can now be connected with advanced technology, including AI and machine learning. We can look for opportunity not only one product at a time or one company at a time, but across an industry, a state, or the entire nation.
That changes the starting point. We no longer have to talk about reshoring in generalities. We can start with what companies are actually buying, what domestic manufacturers can actually make, and where the two overlap.
The goal is not self-sufficiency. The better word is substitutability: the ability to replace selected foreign supply when the economics, risk, learning effects, resilience, or national interest justify it. That is a narrower claim than self-sufficiency, and it is a more honest one.
A domestic source can be economically superior even with a higher unit price if it reduces logistics cost, inventory, downtime risk, defects, engineering delay, currency exposure, geopolitical vulnerability, or the cost of recovering from disruption. The mistake is treating price as the only variable in a transaction whose consequences may stretch across years.
The easiest way to avoid ideological mistakes is to think like a portfolio manager. Nobody serious asks whether stocks are always good or bonds are always bad. The question is what mix serves the objective under uncertainty. Supply chains work the same way. For each input, a buyer should ask: what is the right mix of cost, geography, suppliers, lead times, technology, and risk?
For a state or nation, the question gets bigger: which product capabilities are worth maintaining, rebuilding, or developing because they help many firms at once?
Some capabilities look ordinary until they are gone. A heat-treatment facility can support aerospace, automotive, defense, energy, tooling, and industrial machinery. A foundry can support dozens of downstream manufacturers. A testing lab, a wire mill, a fabrication shop, a precision welding process, or a machine shop may look like one business on paper. In real life, it may be part of the reason a whole region can say yes to a new opportunity.
The Conn-Weld Example
Conn-Weld Industries in Princeton, West Virginia, is the kind of company that makes this argument real. In a recent interview, Marvin Woodie described a business that began in 1975 with a very specific product: profile wire, or wedge wire, screen panels used in coal washing. The opportunity was practical. Coal preparation plants needed durable screens that could separate material reliably and be replaced again and again. Jim Conley, the company's founder, believed he had a better and faster way to weld those panels. When his employer declined to pursue the idea, he resigned and started building the process himself in Princeton.
On the surface, that sounds narrow: a coal-related consumable, a particular kind of screen, a founder with a welding idea. But that is not the real story.
Conn-Weld's advantage came from learning that the final product could not be separated from the process behind it. If the company wanted better screens, it had to control the wire. If it wanted to control the wire, it had to control shape, quality, and metallurgy. By the early 1980s, Conn-Weld had brought that capability into production at scale. The company was no longer simply welding screens. It was building an integrated manufacturing system.
That system allowed it to diversify. Marvin described how the company's wire-shaping and screen-making capabilities moved beyond coal into food, paper, power generation, and other markets. The product changed, but the underlying capability traveled.
This is where the import-substitution opportunity often hides. A product catalog shows what a company already sells. It rarely shows what the company could make. It does not capture the welding process, the metallurgy, the machine tools, the engineering judgment, the turnaround discipline, or the people who know how to solve adjacent problems.
That sentence stayed with me because it says the quiet part out loud. A lot of American capacity is not missing. It is unseen.
A buyer looking only at Conn-Weld's visible product list might see vibrating screens, screen media, and related equipment. Marvin sees something broader when he walks the plant. He sees employee commitment. He sees technology placed in the hands of skilled people. He sees large-scale fabrication capacity, machining, cutting, welding, and engineering knowledge that can carry into other industries. He sees a company that can build a fourteen-foot-wide machine twenty-eight feet long, including single-, double-, and triple-deck machines. He also sees the opposite end of the scale: precision welding down to tiny screen openings and fine wire applications many firms cannot handle.
That is the point. The economy contains more capability than the market can easily see.
Why Product Catalogs Are Not Enough
That invisibility creates a very practical failure. A manufacturer may be able to produce something adjacent to its current market. A buyer may be importing that same kind of product from abroad. But the two never meet because neither side has a normal way to find the other.
Marvin's examples were not theoretical. He described a company from Minnesota that found Conn-Weld through LinkedIn while looking for help with mobile crushing and screening equipment. That company had been sourcing some equipment overseas and wanted to onshore it. The opportunity did not appear because of a grand national program. It appeared because one buyer finally saw a manufacturer that had been capable all along.
He described another case involving Elgin Power in Beckley. The company began buying from Conn-Weld because Conn-Weld could make equipment faster than prior sources, including underground power centers, motor control centers, and underground control systems. The important fact is not just that Conn-Weld won work. It is that the buyer had not fully understood Conn-Weld could do the work.
This is the least controversial part of the import-substitution agenda, and maybe the most powerful: reduce the cost of finding. A state can integrate data, map suppliers, fund technical assessments, support testing, and convene buyers without ordering anybody to buy anything.
When a transaction emerges, the market still has to decide. If no domestic supplier can meet the need, that is still useful knowledge. It tells the state where a gap exists. That gap may become a recruitment target, an expansion target, a workforce target, or a research target. If a supplier can meet the need, the system has done something more valuable than publish another report. It has helped create output.
The Test for a Real Opportunity
But a lead is not enough. This work needs the courage to say no. A real opportunity should survive at least four questions.
- Is the demand real and durable enough to matter?
- Is there a domestic capability that can plausibly serve it?
- Can the supplier become competitive after a reasonable learning or investment period?
- Does the substitution create value beyond a temporary political announcement?
The fourth question is the guardrail. A company that can produce only while sheltered indefinitely behind subsidies has not necessarily strengthened the economy. A temporary bridge can make sense. Permanent compensation for noncompetitiveness should make everybody nervous.
Marvin made this distinction from the manufacturer's side. For Conn-Weld, a reshoring opportunity cannot be merely a one-off purchase order. It has to justify a relationship. “We don't just need a purchase order,” he said. “We need a relationship.” The opportunity has to be evaluated by both sides. It has to fit technically. It has to make economic sense. It has to be worth building around.
That is the discipline policymakers sometimes miss. Reshoring is not accomplished when a domestic supplier gets named in a press release. It is accomplished when the buyer and manufacturer can make the work hold up commercially, technically, and operationally.
Public policy may be justified at the beginning of a reshoring transition because the first domestic unit can be expensive. A supplier may need tooling, certification, engineering, working capital, or equipment before it can quote competitively. A buyer may incur qualification costs before the new source becomes useful. The question is whether a temporary intervention can create a durable commercial relationship.
That creates a test for incentives: are they helping a company cross a learning or investment barrier, or are they permanently compensating for a business that cannot compete? A good policy has a step-down, a cap, verification, and an end. A bridge is supposed to get you somewhere.
The company's own history shows what durable relationships can produce. Marvin described work Conn-Weld did beginning in 1999 with Duaromco, a company based in Rugby, England. Duaromco had used Conn-Weld equipment in the United States, liked the quality and competitiveness, and built a business model with Conn-Weld to supply equipment and parts for projects around the world. Together they worked on coal preparation plants in Russia and Turkey.
That is not isolationism. It is the opposite. It is a West Virginia manufacturer using domestic capability to serve global projects. This is the version of import substitution worth taking seriously: build enough strength at home that American companies can participate in the world from a position of productive capacity.
What Government Should Do
This also explains why state and federal systems have to be designed around manufacturers as they actually operate. Marvin was not saying economic development has no role. He was saying too many programs are built by people who have never had to create a payroll, finance equipment, make a product, manage technical risk, and warranty what they shipped.
That sentence should be written on the wall of every economic development office. Accountability matters. Public money needs verification. But verification should not become a maze only specialists can navigate. For a capable manufacturer, the state's first job is to understand the plant, the people, the financial capacity, the equipment, and the opportunity, then reduce friction where a real project exists.
Marvin suggested a better posture: come audit us. Visit the plant. Look at the books. Look at the equipment. Talk to the people. Understand what the company can actually do. Then help connect that verified capability to real demand.
That is far more useful than asking manufacturers to decode a dozen programs, portals, forms, and eligibility rules before anyone has studied whether the opportunity is commercially real. Economic development should not begin with paperwork theater. It should begin with market discovery.
The older version of import substitution often began with protection and hoped domestic industry would follow. The new version can begin with information. Start with the import ledger. Find real demand. Identify domestic capability. Test the match. Support the transition only where a lasting commercial relationship can emerge.
The old version often assumed a domestic industry had to be created behind a wall. The new version can begin with an existing manufacturer that does not know a nearby buyer is importing something within its capability set.
The old version was usually centralized. The new version can be distributed across states, firms, associations, universities, banks, and markets.
The old version too often measured success by replacement alone. The better measure is output, productivity, resilience, and relationships that survive after the incentive ends.
Import Substitution Is Broader Than Reshoring
Import substitution also needs to be separated from simple reshoring. Reshoring usually means bringing production back after it moved overseas. Import substitution is broader. It includes cases where a domestic company never made the exact imported item before but has adjacent capability. It includes cases where a new product category emerges and the question is whether domestic firms can participate before foreign dependency hardens. It includes cases where the substitute is not identical, but is functionally adequate, technically superior, faster, safer, or less exposed to disruption.
That broader definition matters because America should not be limited to rebuilding what it once made. Some of the largest industrial opportunities of the next generation will come from products, materials, energy systems, defense needs, medical technologies, and infrastructure components whose markets are still forming.
This is also different from a local purchasing campaign. Buying local can strengthen communities, but an industrial strategy cannot run on sentiment. A manufacturer buying a critical input should not choose a local supplier merely because the supplier is local. It should choose that supplier because the supplier can perform. The patriotic argument may open the door. Performance keeps the contract.
What Communities Lose When Capability Disappears
This is where the argument stops being abstract. Manufacturing matters because it produces more than goods. It produces skills, tax base, supplier demand, household formation, career ladders, and a community's confidence in itself.
When I asked Marvin what Princeton would lose if Conn-Weld disappeared, he did not talk only about jobs. He talked about the payroll that supports homes and cars. He talked about property taxes, local spending, schools, hospitals, car dealers, and the regional economy. The plant's value is not contained inside the fence line. It moves through the community.
That does not mean every plant can or should be preserved at any cost. But policymakers should understand what is being lost when a manufacturing capability disappears. The loss is not only current employment. It is accumulated knowledge, supplier relationships, trained workers, tax base, and the belief that a young person can build a serious life close to home.
That belief is not abstract. Marvin described speaking to students in elementary schools, middle schools, high schools, career and technical programs, and colleges. He asks them why they want the futures they describe. One student told him she wanted to travel the world. His response was that Conn-Weld could help create that future from West Virginia.
That is what a manufacturing employer can give a young person beyond a job: a place to learn difficult skills, solve real problems, work in global markets, earn a good living, and still build a life in a community where a home is within reach. That matters. It matters a lot.
Import Substitution as Discovery
Import substitution for a new era is not a retreat into the past. At its best, it is a way to make productive futures visible before they are obvious to everyone else.
A country should not try to make everything. But it should know what it is buying. It should know what it can make. It should know which capabilities matter. And it should know when the gap between demand and capability is small enough to close. That means helping buyers find domestic suppliers, helping suppliers evaluate opportunities, and helping both sides cross the early barriers that prevent a durable relationship from forming.
The goal is not to abolish trade. The goal is to reduce unnecessary dependence and increase productive choice.
That is what makes this a new era of import substitution. The old version began too often with protection and hoped domestic industry would follow. This version begins with information. It starts with the import ledger, looks for real demand, identifies domestic capability, tests the match, and supports the transition only where a lasting commercial relationship can emerge.
Bring Manufacturing Home.