Infant Industry Protection

In the AI Story

The intellectual genealogy runs from Alexander Hamilton's 1791 Report on the Subject of Manufactures through Friedrich List's National System of Political Economy (1841) to the postwar developmental states of East Asia. Hamilton's argument was that the United States could not develop manufacturing if it accepted Britain's free-trade prescription, because British manufacturers had a century of accumulated advantage. The argument was correct, the policy was implemented, and the United States overtook Britain as the world's largest industrial economy by 1890.

The most successful modern application was South Korea under Park Chung-hee. State-controlled banks directed credit at below-market rates to firms that met export performance targets. Import quotas protected domestic producers during their learning period. Technology transfer requirements gave Korean firms access to foreign know-how as a condition of market access. The result was the transformation of a country poorer than Ghana in 1961 into the world's twelfth-largest economy by 2020.

The AI parallel is structurally exact. The leading AI capabilities are concentrated in a handful of firms with massive accumulated advantages: training data scraped before legal frameworks adapted, computational infrastructure built with public subsidies, talent pools educated in publicly funded research universities, and brand recognition that compounds through network effects. A nascent AI industry in Lagos or Dhaka cannot compete with these incumbents on equal terms — and the rules being written ensure that it will not have the chance.

The infant industry argument is not a defense of permanent protection. Chang has been emphatic that protection that breeds complacency rather than competence is failed protection. The Korean miracle worked because the protection was conditional, performance-oriented, and explicitly designed to produce firms that could eventually compete without it. The argument against contemporary economic orthodoxy is not that liberalization is wrong but that immediate liberalization forecloses the development that would make liberalization beneficial.

Origin

The phrase 'infant industry' was coined by Hamilton in 1791. Friedrich List developed it into a comprehensive theoretical framework in opposition to David Ricardo's free-trade doctrine, arguing that comparative advantage assumes a level of industrial capability that developing nations must first build. List's framework shaped German industrial policy in the late nineteenth century and Japanese policy in the twentieth.

Chang's contribution was empirical: the systematic documentation that the wealthy nations all used infant industry protection during their own development, regardless of what theory their economists were preaching at the time. The pattern is so consistent across so many cases over so many centuries that the burden of proof, Chang argues, has shifted to those who claim the policy does not work.

Key Ideas

Sequential logic. Protect first, liberalize second — the opposite sequence destroys nascent industries before they can develop the capabilities required to survive competition.

Performance conditioning. Successful protection includes export targets, quality requirements, and discipline mechanisms that prevent the slide into rent-seeking.

Time-limited application. The infant must eventually leave the nursery. Protection that becomes permanent breeds the complacency that justifies protection's critics.

AI applicability. The frontier AI industry is the paradigmatic case for infant industry protection — massive incumbent advantages, high barriers to entry, strategic importance for long-term economic positioning.