Global Trade and Industrial Policy: What Each Position Is Protecting

In April 2023, Jake Sullivan — the U.S. National Security Advisor — stood at the Brookings Institution and declared the era of uncritical free trade orthodoxy over. The "Washington Consensus," he said, had left workers behind, hollowed out industrial communities, and created strategic vulnerabilities that adversaries could exploit. What he described as a "new Washington consensus" was blunter: the U.S. government was going to pick winners, subsidize strategic industries, and build domestic manufacturing capacity regardless of whether it was the most efficient allocation of capital according to comparative advantage.

The speech followed two major pieces of legislation — the CHIPS and Science Act ($52 billion for domestic semiconductor manufacturing) and the Inflation Reduction Act (hundreds of billions in subsidies for clean energy and electric vehicles) — that represented the largest U.S. industrial policy intervention since World War II. The EU responded with its own Green Deal Industrial Plan. Japan, South Korea, Canada, and India followed with similar frameworks. Within eighteen months, every major rich-country democracy had effectively abandoned the position that governments should not direct industrial investment — a position those same governments had enforced internationally through the WTO for thirty years.

The WTO's own dispute settlement system had by then already gone quiet. Since December 2019, the U.S. — under both Trump and Biden — had blocked new appointments to the Appellate Body, the WTO's highest court. Without a quorum, it could no longer hear appeals. Trade disputes could be filed but not resolved. The institution built to enforce the rules of the liberal trading order was paralyzed at exactly the moment when the rules were under their most sustained challenge. The debate about global trade and industrial policy is not, in 2026, abstract. It is about which architecture replaces what is breaking down — and what each position thinks that architecture should protect.

What the free trade liberals and comparative advantage tradition are protecting

The efficiency gains from specialization — and who actually benefits from them. The foundational argument for free trade runs from David Ricardo through two centuries of mainstream economics: countries gain from specialization and exchange even when one country is absolutely better at producing everything. If Bangladesh can produce garments at a fraction of the cost of Germany, both countries are richer if Bangladesh makes garments and Germany makes something else — even if German garment workers lose their jobs in the transition. The free trade tradition is protecting the claim that these efficiency gains are real, large, and broadly distributed: that lower prices for traded goods function as a pay raise for everyone who buys them, that the aggregate wealth created by two centuries of trade liberalization is not a fiction, and that policies designed to override comparative advantage impose real costs on real people — most visibly on consumers who pay higher prices, but also on the workers in sectors that lose export access when trading partners retaliate.

The WTO rules-based system as a constraint on the powerful. Beyond efficiency, the liberal trade tradition has always had a governance argument: that rules-based trade institutions protect smaller and weaker countries from the raw exercise of great-power economic coercion. When the U.S. wants to impose tariffs on steel from Vietnam, the WTO's dispute settlement system gives Vietnam a legal mechanism to challenge that tariff under agreed rules — a mechanism that raw bilateral negotiation would not provide. The tradition is protecting the claim that the alternative to multilateral rules is not "fair" bilateral deal-making but power: that countries with larger markets extract better terms from smaller trading partners in ways the WTO framework was specifically designed to constrain. The paralysis of the Appellate Body — which the U.S. caused — is, from this perspective, not a neutral institutional failure but a choice by the most powerful country to remove the constraint on itself.

The empirical case against industrial policy's track record. The free trade tradition also carries a critique of industrial policy as a practice: that governments are poor at picking which industries have the market-failure properties that justify intervention, that "strategic industry" is a label that expands to cover whatever has political support, and that the history of state-directed industrial investment is littered with expensive failures. The Concorde was a triumph of Franco-British industrial ambition and a commercial disaster. Japan's Ministry of International Trade and Industry backed many winners — but also fax machines and analog photography into the digital era. The CHIPS Act's domestic semiconductor fabrication will cost $30–40 billion per plant — three to four times what comparable capacity costs in Taiwan. The tradition is protecting the discipline of asking what the evidence actually shows about industrial policy's empirical success rate, not just its theoretical justifications.

What the industrial policy advocates and developmental state theorists are protecting

The historical record that no wealthy country industrialized through free trade. Ha-Joon Chang's Kicking Away the Ladder (2002) documented a pattern that disrupts the standard narrative: every country that is now wealthy — Britain, the United States, Germany, France, Japan, South Korea, Taiwan — industrialized through some combination of tariff protection, infant industry subsidies, state-directed credit allocation, and technology acquisition strategies that would be illegal under current WTO rules. Britain protected its textile industry for centuries before championing free trade once it had the most productive industry in the world. The United States' first Treasury Secretary, Alexander Hamilton, wrote the Report on Manufactures (1791) specifically arguing that infant industries needed protection from British competition until they could stand alone — and the U.S. was among the world's most protectionist economies for most of the nineteenth century. The industrial policy tradition is protecting the recognition that the countries preaching free trade built their industrial capacity through the tools they now deny to others.

The market failure argument: learning-by-doing, coordination, and spillovers. Industrial policy advocates draw on a body of economic theory showing that markets systematically underprovide industrial capabilities that have three properties: learning-by-doing dynamics (where unit costs fall as cumulative production rises, so whoever gets there first gets an advantage that is hard to dislodge); positive externalities to other sectors (semiconductor manufacturing generates knowledge spillovers to everything from defense to consumer electronics to AI); and coordination failures (where building a supply chain requires many investments to happen simultaneously, which no single firm can coordinate but government can). Paul Krugman's 1987 paper "Is Free Trade Passé?" — written at the height of his mainstream credentials — showed that strategic trade theory provided a rigorous basis for industrial policy in industries with these properties: a government that subsidizes its semiconductor industry might capture a market that would otherwise go to a foreign competitor, generating returns that exceed the subsidy. The tradition is protecting the claim that comparative advantage is not fixed by nature — it is created by investment, and governments can legitimately create it.

The Inflation Reduction Act as a new template: subsidies over tariffs. The Biden administration's industrial policy innovation — carried forward in the post-2024 policy landscape — was to pursue industrial policy primarily through subsidies rather than tariffs. The IRA's clean energy provisions offered production tax credits, investment tax credits, and manufacturing credits that made domestic production economically competitive without explicitly blocking imports. This is a different theory of industrial policy than the mercantilist tradition: not "protect domestic industry by taxing foreign competition" but "make domestic industry so well-funded that it can compete on the merits." The tradition is protecting a more technically defensible form of industrial policy — one that still distorts markets but does so by lifting domestic producers rather than penalizing foreign ones, a distinction with both economic and diplomatic consequences.

What the labor and fair trade critics are protecting

The "race to the bottom" as a structural feature of free trade, not a bug. The labor critique of free trade begins from a different unit of analysis than the efficiency critique: not aggregate welfare but the distribution of bargaining power within labor markets. When a U.S. manufacturer can threaten to move production to Mexico or Vietnam, it gains leverage over its U.S. workers that is independent of any specific negotiation. The credible threat of exit — relocating to a lower-wage jurisdiction — suppresses wages and working conditions in the home country without the production ever actually moving. The fair trade tradition is protecting the claim that this is not an externality to be managed but a central mechanism of how trade integration affects workers: that "free trade" as practiced is freedom of capital to seek the lowest possible labor costs, and that this freedom structurally disadvantages workers relative to employers in ways that aggregate welfare statistics miss.

The labor and environmental standards problem in trade agreements. Trade agreements negotiated since the 1990s have included increasingly elaborate intellectual property protections, investor-state dispute settlement mechanisms, and regulatory harmonization provisions — but labor and environmental standards have been either absent or unenforceable. The North American Free Trade Agreement's side agreements on labor created a commission but no meaningful enforcement. USMCA's stronger labor provisions — negotiated under pressure from the AFL-CIO and House Democrats — represented an improvement, but their enforceability has been contested in practice. The fair trade tradition is protecting the argument that there is nothing "free" about trade rules that protect pharmaceutical patents and corporate arbitration rights internationally but leave minimum wage standards and union rights as purely domestic matters: that the current architecture of trade law is not neutral but specifically organized to protect capital interests across borders while treating labor interests as local.

The distributional record — not "did trade increase aggregate wealth" but "who captured it." David Autor, David Dorn, and Gordon Hanson's "China shock" research found that communities with high exposure to Chinese import competition experienced persistent job losses, wage declines, elevated disability claims, and increased mortality — effects that persisted for decades without the labor market adjustment that trade theory predicted. The aggregate gains from trade were real; they were also captured almost entirely by shareholders and high-income workers in traded sectors, while the adjustment costs fell on manufacturing workers in specific geographic communities. The fair trade tradition is protecting the distributional lens: not the claim that trade is bad, but the claim that trade policy's costs and benefits have been allocated in a specific direction — toward capital and away from labor — and that "the country gains" is a statement that requires asking which people in the country and what they gained.

What the developing economy and infant industry advocates are protecting

The right to use the tools that rich countries used — and now restrict. The WTO's rules — particularly the Agreement on Subsidies and Countervailing Measures — restrict many of the industrial policy tools that developing countries might use to build domestic manufacturing capacity: export subsidies, local content requirements, domestic input mandates. The CHIPS Act's domestic semiconductor subsidies, the IRA's domestic content requirements for clean energy tax credits, and the EU's State Aid framework are all potential violations of these same rules — rules that wealthy countries are simply large enough to ignore, challenge slowly, or negotiate exemptions around. The infant industry tradition is protecting the argument that this is not a coincidence but a structure: that the WTO framework was written by wealthy countries to lock in their industrial advantages while foreclosing the development paths that those same countries used to build those advantages, and that "trade liberalization" has functioned as development policy for the already-developed.

Policy space as a development prerequisite. Dani Rodrik's concept of "policy space" — the latitude for governments to experiment with industrial and trade policies tailored to their specific circumstances — is central to the developing country critique. South Korea's development model involved protecting the chaebol conglomerates from import competition while requiring them to export; Taiwan's involved state-directed credit to specific industries; China's involved required joint ventures, technology transfer mandates, and subsidized credit at a scale that WTO rules would prohibit for smaller economies. These were not free market successes — they were state-directed industrialization strategies that happened to work. The infant industry tradition is protecting the claim that the question is not "state vs. market" but "which institutional arrangements, with which state capabilities, produce development at which income level" — and that the WTO's one-size-fits-all rules foreclose experiments that developing countries need to try.

The hypocrisy problem in rich-country industrial policy revival. The Biden CHIPS Act, the EU Green Deal Industrial Plan, and similar measures represent wealthy countries doing at scale exactly what they spent thirty years telling developing countries they could not do. India's domestic manufacturing subsidies under "Make in India" have been challenged at the WTO. Indonesia's nickel processing requirements — designed to capture more value from its natural resources — were struck down. Vietnam faces pressure over currency practices that the U.S. Treasury uses in its own monetary policy. The developing economy critique is protecting the argument that there is a double standard operating in real time: the wealthy countries that built the current trade architecture are now dismantling the parts of it that constrained them, while the parts that constrain developing countries remain intact.

What the strategic trade theorists and economic nationalists are protecting

The argument that China's trade practices make naive free trade a strategic mistake. Robert Lighthizer's No Trade Is Free (2023) — written by the U.S. Trade Representative who designed much of the Trump administration's China tariffs — makes the strategic nationalist case most clearly: China did not liberalize as it was supposed to after entering the WTO in 2001; instead, it maintained state subsidies, forced technology transfer through joint venture requirements, tolerated intellectual property theft, managed its currency to maintain export competitiveness, and built dominant positions in strategic industries (solar, batteries, electric vehicles, rare earth processing) through a combination of state direction and market scale. The strategic trade tradition is protecting the claim that comparative advantage theory assumes something like fair competition — similar rules, similar subsidies, similar enforcement — and that when one major trading partner operates by fundamentally different rules, the theory's prescriptions do not apply. You cannot "comparative advantage" your way to industrial capacity when your competitor is running a different game.

The political economy of concentrated losses and diffuse gains. One structural feature of trade that both the free trade and industrial policy traditions tend to underweight is the political economy asymmetry: trade creates diffuse gains (lower prices for many goods, distributed across all consumers) and concentrated losses (specific industries in specific communities, visible and politically organized). Steel workers who lose their jobs vote; consumers who pay slightly less for refrigerators do not organize around that benefit. This asymmetry means that democratic governments pursuing aggregate welfare-maximizing trade policy consistently produce political backlash — not because voters are irrational but because the costs are salient and the benefits are invisible. The strategic trade tradition is protecting the political economy insight: that a trade policy sustainable in democracies must grapple with distribution and adjustment, not only efficiency, or it will produce the political backlash that eventually destroys it — as the backlash against the post-1995 trade consensus demonstrably did.

The industrial commons argument: why letting manufacturing go creates capabilities that cannot be rebuilt. Beyond specific industries, strategic trade theorists argue that manufacturing involves tacit knowledge — skills, supplier networks, engineering capabilities, quality control practices — that cannot be easily rebuilt once lost. The Harvard Business School researchers Gary Pisano and Willy Shih documented what they called the erosion of the "industrial commons": when U.S. companies offshored manufacturing to cut costs, they did not just move jobs — they hollowed out the ecosystem of suppliers, skilled workers, and process knowledge that supports product innovation. Apple's decision to manufacture the iPhone in China was not primarily about labor costs; it was about the fact that the precision manufacturing ecosystem for consumer electronics simply no longer existed in the U.S. at the required scale. The strategic tradition is protecting the claim that "we can always re-shore manufacturing if we need to" is not true — that industrial capability is path-dependent in ways that make the free trade argument's implicit reversibility assumption wrong.

What the argument is actually about

Whether comparative advantage is found or made. The deepest disagreement in this debate is about the nature of comparative advantage itself. The free trade tradition treats it as given by factor endowments and technology — countries should specialize in what they are already relatively good at, and trade will allocate production to the most efficient producers. The industrial policy tradition argues that comparative advantage in high-value industries — semiconductors, aircraft, pharmaceuticals, electric vehicles — is created through sustained investment, learning, and state support: there is nothing natural about Taiwan's dominance of advanced chip manufacturing or South Korea's in memory chips. Both positions engage with real phenomena: genuine natural resource advantages exist, and state-created industrial capacity also exists. The question is how much of the high-value industrial landscape is "found" versus "built" — and whether countries that accept the free trade prescription are accepting a static competitive position while others build dynamic advantages.

Whether the WTO framework can be reformed or must be replaced. The Appellate Body crisis reveals an institutional contradiction: the U.S. built the WTO but now finds the system's constraints on its industrial policy more costly than the constraints it imposes on others are valuable. The free trade tradition argues for reforming the WTO — updating its rules to address digital trade, state-owned enterprises, and currency manipulation — while preserving the dispute settlement architecture. The industrial policy tradition is less sure the institution is salvageable: if the rules were written by and for a world in which the U.S. was dominant and China was not yet a major industrial power, and if those rules now constrain developing countries while wealthy countries violate them at scale, reform may be impossible without the wealthy countries accepting constraints they will not accept. What emerges from the current institutional breakdown — a plurilateral framework, regional arrangements, bilateral deals — will define the trade architecture for the next generation.

Whether "industrial policy" refers to the same thing across the political spectrum. One underappreciated source of confusion in this debate is that "industrial policy" means very different things to different advocates. Mariana Mazzucato's developmental state argument focuses on building public sector capabilities and socializing both the risk and the reward of innovation. Dani Rodrik's version emphasizes experimentation, feedback mechanisms, and disciplined accountability for state investments. The Biden administration's version was primarily tax credits designed to shift private investment, with the government setting direction but not owning or operating anything. Peter Navarro's version was tariffs designed to protect domestic producers from foreign competition. These are not the same policy. Their track records differ, their political economies differ, and their compatibility with a multilateral trading system differs. The debate often proceeds as though "industrial policy" is a single thing that either works or doesn't — obscuring the significant variation in what the term actually refers to.

Beneath the surface: not a dispute about whether trade is good — almost everyone agrees that exchange creates value — but about who captures the value, whether comparative advantage is natural or constructed, whether the rules-based trading system can be reformed or is already breaking down, and whether the industrial policy tools that built the rich world's prosperity can legitimately be used now by countries trying to catch up, or only by the countries that are already there.