American economic nationalism has found its economist in Friedrich List (1789-1846), and the choice makes sense. List denied that the existing international division of labor was natural or permanent. He separated a nation’s present wealth from its productive powers, meaning its skills, its technical knowledge, its transport, its schools, its firms, and its capacity to make things it had never made before. A country could trade consumption today for capability tomorrow. A country that accepted free trade while its industries remained immature might never acquire the machinery, the trained workers, the suppliers, the engineers, or the commercial institutions needed to compete with an established leader. It might instead settle into permanent supply of raw materials to a manufacturing rival and call the arrangement efficient.
That argument is serious, and the profession has spent two centuries either ignoring it or rediscovering it under other names. The infant-industry case, learning by doing, agglomeration, path dependence, and most of the recent literature on industrial policy are List translated into the language of models.
The trouble with the American use of List is not that he was wrong. The trouble is that his own scheme assigns the United States to a different chapter.
The National System of Political Economy (1841) is built on a sequence of stages. A nation passes from hunting to pasturage to agriculture, then into manufacturing, then into manufacturing and commerce together. List attaches protection to one passage in that sequence, the transition into manufacturing, and he attaches it under conditions. The duties should be moderate. They should be temporary. They should cover industries with a plausible prospect of standing on their own. They should not extend to agriculture. They suit a country with sufficient territory, population, and coastline to support a full industrial system, and they are useless to a country that has no such prospect. Protection carried too long, List wrote, produces indolence in the firms it shelters.
Then he describes the nation that has reached the top of the sequence. That nation, he says, has an interest in free trade, and it preaches free trade to its rivals for that reason. His model of the case was England. England had climbed behind Navigation Acts and prohibitions and colonial monopoly, arrived at supremacy, and then announced that supremacy was the reward of open competition. List’s charge against Britain was hypocrisy, and Ha-Joon Chang (b. 1963) revived the charge under a better title in Kicking Away the Ladder (2002). But the charge only works because List agreed with the British about what a leading manufacturing nation should do. He expected Germany to arrive at the same place and adopt the same policy.
The United States now occupies the position List assigned to England. It is a mature, high-wage economy at or near the technological frontier in semiconductors, software, aerospace, biotechnology, pharmaceuticals, finance, advanced energy, and scientific research. It issues the currency that made up roughly 57 percent of official foreign exchange reserves in early 2026. It has lost sectors, not a civilization. Its problem is selective industrial erosion inside an advanced economy, not national industrial infancy.
Citing List to protect the leader inverts his stage theory. That does not make protection wrong in 2026. It means the movement is invoking an authority who prescribed something else for its situation, and it has not noticed.
The second difficulty concerns what List wanted protected and how.
His unit of analysis was productive power. He counted engineers, schools, apprenticeships, roads, canals, patents, banks, and firms that could design as well as assemble. He would not have recognized the bilateral goods balance as an object of policy. A deficit tells you nothing about whether a country can build a machine tool, train a die maker, or bring a new chemical process from a laboratory to a plant. The movement has substituted an accounting figure for the thing List cared about, and the substitution is convenient, because the figure moves with a tariff schedule and productive power does not.
His practical career points the same way. From 1819 List served as the organizer of the German commercial association, and his first campaign was the abolition of the internal tariffs that divided the German states from each other. He wanted the Zollverein, which arrived in 1834. He promoted the national railway network and helped bring the Leipzig-Dresden line into existence. He wrote about technical education and public credit. The external duty was one instrument among several, and the several were the expensive ones, requiring a state capable of building things and teaching people.
So the first act of nineteenth-century German economic nationalism enlarged a market. A general duty against forty-six countries contracts one.
The American precedent runs on the same asymmetry. Alexander Hamilton (1755-1804) submitted the Report on Manufactures in December 1791 and argued that bounties, meaning direct subsidies tied to output, were superior to prohibitions and preferable to high duties. Congress declined the bounties and kept the tariff. The American System took the cheap half in 1792 and takes the cheap half now.
The uncomfortable comparison is not nineteenth-century Germany. It is Edwardian Britain.
Joseph Chamberlain (1836-1914) launched the tariff reform campaign at Birmingham on May 15, 1903. His premise was that Britain’s early industrial lead was gone, that Germany and the United States had built larger industrial systems behind protective walls and continental markets, and that British free trade had stopped being a neutral doctrine. It exposed British manufacturers to foreign protection while enriching the shippers, insurers, bankers, and holders of overseas assets who lived off an open system. His remedy was duties on foreign goods and preference for goods from the Empire, a protected commercial bloc large enough to sustain British production, fund social provision, and answer the continental rivals.
Read Chamberlain’s speeches next to American Compass and the resemblance is close enough to be awkward. Cheap imports conceal the destruction of capacity. Free trade has become an ideology serving finance. Industrial weakness threatens military power. National cohesion requires national production. A protected bloc can replace dependence on competitors.
The campaign split his own party. Arthur Balfour (1848-1930) tried to hold both wings and held neither. The Unionist free traders defected. And the argument was lost at the grocery counter, on the price of bread, where the free traders campaigned with a big loaf and a little loaf and asked which one the voter preferred. The 1906 election destroyed the Conservative and Unionist position in the Commons. Chamberlain suffered a stroke in July of that year and never returned to active politics.
The sequel is the part American nationalists should sit with. Britain did eventually get protection. The Import Duties Act of 1932 imposed a general tariff, and the Ottawa agreements of August 1932 delivered imperial preference. Britain’s relative industrial decline continued.
Tariffs were not the cause of that decline, and free trade was not the cause either. Britain faced rivals with larger home markets, carried the costs of empire and two wars, returned to gold at an overvalued parity, and held early leads in industries where the early lead had left aging plant and settled methods. The instructive diagnosis comes from Alfred Chandler (1918-2007). Scale and Scope (1990) traces British underperformance to what he called personal capitalism: family firms that declined to invest in the managerial hierarchies, the distribution networks, and the in-house research that American and German competitors built. British firms stayed small, stayed owned by families, and stayed out of the three-pronged investment that made a first mover durable. Add the technical education deficit that British commissions documented for fifty years and did little about.
No tariff schedule supplies a managerial hierarchy. No duty trains a chemist. Chamberlain proposed to change the price of imports in a country whose deficiency lay in how its firms were owned, organized, and staffed. Aaron Friedberg’s The Weary Titan (1988) shows how hard British officials found it to see their own position accurately, which is the more general warning. A nation in relative decline reaches for the instrument it can operate rather than the one that addresses its condition.
The second comparison is import substitution.
Import substitution began from a real problem and a competent argument. Raúl Prebisch (1901-1986) and the economists at the Economic Commission for Latin America, founded in 1948, held that countries specialized in primary exports faced deteriorating terms of trade, unstable prices, chronic foreign exchange constraints, and technological dependence. Hans Singer (1910-2006) reached similar conclusions. Domestic industry was the route out. Judged against passive specialization in minerals and agriculture, the policy accomplished a great deal. Brazil, Mexico, and Argentina built factories, technical cadres, urban employment, infrastructure, and domestic firms that would not otherwise exist.
The failure came later and had a precise shape. Albert Hirschman (1915-2012) described it in 1968 in his essay on the political economy of import-substituting industrialization. The process ran backward. It began at the last stage of production, final assembly of consumer goods, using imported machinery, imported components, and imported technology. It skipped the learning. The entrepreneur who assembled radios behind a tariff wall never acquired the capability to design a radio or build the machines that made one. Substituting the final good left the demand for foreign exchange intact, because the inputs still came from abroad. The easy substitutions ran out. What remained were the hard ones, requiring capital, scale, and technical depth that the sheltered market had given nobody a reason to develop.
Then the politics closed around it. Anne Krueger (b. 1934) named the rent seeking in 1974 and Jagdish Bhagwati (b. 1934) generalized it. Tariff structures became negotiated rather than designed. Firms learned that a good relationship with the ministry paid better than a productivity gain. The infants aged without maturing. Capacity was visible, so capacity got mistaken for competitiveness. A government could point at factories and payrolls, and neither established that the protected sector could survive competition or earn the foreign exchange to pay for its own imported inputs.
The debt crisis of the 1980s had many causes and it would be careless to hang it on industrial protection. The durable lesson is narrower: a country can build industrial capacity while failing to build productivity growth or export competitiveness, and it can go a long time without finding out.
Here the American position produces an unexpected danger. The United States faces no comparable payments constraint. It borrows in its own currency and the world wants its assets. Argentina learned within a decade that its protected industries could not earn the exchange to buy their own machinery. The United States can finance an uncompetitive industry for thirty years out of an enormous domestic market, federal borrowing, protected procurement, and foreign demand for Treasuries. The privilege the movement wants to preserve is the same privilege that would hide the failure from view. America may escape the crisis and reproduce the underlying productivity failure in a slower and more politically durable form.
The standard answer to both comparisons is East Asia, and the answer has force. Japan, South Korea, and Taiwan protected industries, directed credit, promoted investment, and coordinated technological development. Chalmers Johnson (1931-2010) described the apparatus in MITI and the Japanese Miracle (1982). Robert Wade (b. 1944) documented Taiwan in Governing the Market (1990). Industrial policy worked.
Alice Amsden (1943-2012) identified the feature that made it work. In Asia’s Next Giant (1989) she described Korean support as reciprocal. Firms received cheap credit, protection, foreign exchange, and licenses, and in exchange they accepted performance standards that the state monitored and enforced: output targets, investment commitments, technological upgrading, and above all exports. The export requirement supplied a test the domestic market could not. A company can persuade its own ministry that it is strategically vital. Persuading a foreign purchaser to buy an inferior product at a higher price is harder.
The recent quantitative work supports a qualified version of the story. Nathaniel Lane’s study of Korea’s Heavy and Chemical Industry drive finds that targeted sectors expanded, developed new comparative advantage, raised productivity, and held the gains after the temporary support ended. Plant-level work by Minho Kim, Munseob Lee, and Yongseok Shin finds that the same program worsened the allocation of resources among firms inside the targeted industries. A real achievement, purchased at a cost.
Which puts the institutional question where it belongs. Protection is easy to start. Discipline is hard to sustain.
The American state has formidable promotional capacity. It funds basic research, builds infrastructure, guarantees loans, writes tax credits, and creates markets through procurement. Defense, aerospace, computing, biomedicine, and energy all carry its fingerprints. Its terminating capacity is another question. Congress represents districts and states, not a national production function. Once a subsidized plant exists, its continuation becomes a local interest with two senators attached. Once an industry is designated vital to national security, withdrawal of support becomes an argument about patriotism. Concentrated firms lobby better than diffuse consumers. A politician gains from announcing an investment and loses from closing a failed one nine years later.
The CHIPS program shows that conditions can be attached. It carried milestones, reporting requirements, restrictions on expansion in countries of concern, and clawback provisions. It also shows the limits of the technique. A national security condition is verifiable, because you can determine whether a company built prohibited capacity in China. A commercial performance condition requires someone to decide whether an American plant has become productive enough, innovative enough, or cheap enough to justify continued support, and to say no. Nobody has yet had to say no. The conversion of unpaid Intel grants into a federal equity stake in 2025 moved in the opposite direction, toward ownership, which is a relationship that makes exit harder rather than easier.
This is the core of the List problem. The case for protection rests on the state eventually exposing the protected firm to judgment. The politics of protection give both the firm and the state reasons to postpone the judgment forever.
The third complication is automation, and it separates two promises the movement makes together.
The first promise is restored industrial capacity. The second is restored industrial employment. A modern plant can deliver the first while disappointing the second, and the incentive runs that way by design. A firm that brings production into a high-wage country has every reason to substitute capital, software, and robotics for labor.
The CHIPS numbers are the available evidence. Against roughly $53 billion in authorized support, Bilge Erten, Joseph Stiglitz, and Eric Verhoogen estimate about 15,000 to 16,000 direct jobs in semiconductor production and equipment, and 28,000 to 35,000 indirect jobs in upstream industries and construction, with measurable wage gains in the affected counties. Those are real jobs, skilled and well paid. Beside the scale of public money and private capital, and beside the promise of restoring the industrial Midwest, they are small.
Which does not mean the program failed. Semiconductor fabrication is capital deep and always will be. Its justification is that advanced chips sit underneath military systems, communications, artificial intelligence, transport, and most of the modern economy, and that a country without domestic capacity holds a hostage relationship with Taiwan and South Korea. The honest defense is insurance. The dishonest defense is employment.
The contradiction is unavoidable. A program built to maximize productivity and competitiveness automates and hires few. A program built to maximize headcount preserves labor-intensive, low-productivity firms that will need protection permanently. The movement has not chosen, because choosing costs it either its economics or its voters.
And a factory is not a social order. Restoring steel, machine tools, ships, pharmaceuticals, or semiconductors does not restore affordable housing, stable families, local commerce, technical schools, solvent municipalities, or a household supported on one industrial wage. Those outcomes came from a settlement that included unions, regional banks, regulated utilities, restricted foreign competition, employer benefits, a young population, and a different balance between capital and labor. Industrial policy can contribute to regional reconstruction. It cannot substitute for it. The men who voted for reindustrialization heard a promise about their towns, and the policy addresses their supply chains.
Underneath all of this sits an arithmetic problem the movement discusses only in the abstract.
Reserve currency status draws foreign capital into American assets. Those inflows are the accounting counterpart of the current account deficit. Matthew Klein and Michael Pettis set the argument out in Trade Wars Are Class Wars (2020): the imbalance originates in savings and investment decisions, many of them made abroad, and the trade balance adjusts to accommodate the capital account. On that account a tariff redirects trade among foreign suppliers without touching the underlying relationship.
So the program as stated wants incompatible things. It wants the productive structure of an industrial nation, the consumption possibilities of a free-trading nation, and the financial privileges of a reserve-currency hegemon. Reducing imports while preserving unrestricted capital inflows, dollar supremacy, elevated asset prices, cheap consumer goods, and the global organization of American corporations describes a set of goals that do not fit together.
The test is available and simple. Pettis’s diagnosis implies capital-account measures: taxes on inbound portfolio flows, tolerance of a weaker dollar with higher long rates, and a willingness to give up some part of the reserve privilege. American Compass has adopted the diagnosis. Watch whether anyone in the movement proposes the medicine. Watch whether any elected nationalist tells the holders of dollar assets what reindustrialization would cost them. Until then Pettis is decoration, and the industrial program is being sold as an addition to the American order rather than a subtraction from it.
List gave a rising nation permission to protect its infant industries for a period, moderately, selectively, and on the way to something else. He gave the leading manufacturing nation a different instruction. And the parts of his program that required a competent state, the customs union that enlarged the market, the railways, the technical schools, the public credit, are the parts the movement has left on the shelf.
A policy he might recognize looks unlike a general duty. It would enlarge markets among trusted countries rather than tax them, because no advanced economy reproduces every mineral, chemical, component, machine tool, and subsystem inside its borders, and the relevant production unit is a bloc. It would spend on apprenticeship, technical education, and engineering capacity, which is where Britain lost and Germany won. It would attach performance standards to support and an export test where competitiveness is claimed, and it would say openly that capacity purchased for military reasons is defense procurement rather than economics. It would protect industries without protecting every incumbent inside them, and it would fund new entrants against the firms that already have the lobbyists. It would name, in advance, which privileges of the current order the country is prepared to trade away.
None of that is impossible. All of it is harder than a tariff, which is why the tariff is what exists.
The nationalists are right about the things that made them necessary. Industrial structure has consequences that consumer prices do not capture. National power cannot be reduced to the cost of a television. Markets that cross jurisdictions are shaped by states, currencies, security arrangements, and political choices, and the discipline that treated the global allocation of production as presumptively efficient was making an assumption rather than a finding.
What they have not done is explain why the United States escapes what happened to Britain, to Argentina, and to every protected sector that aged without maturing. They have not shown how a duty at the border overcomes the financial incentives created by the dollar. They have not shown which official will one day tell a subsidized firm in a swing state that its support is ending. They have not shown how an automated plant restores a town. And they have not decided what they will sacrifice.
List climbed those questions before he answered the tariff question. His American admirers have taken the answer and skipped the climb.
