How to Build an Industry from Nothing
Poorer Than Ghana
In 1961, South Korea's per capita income was $82. That is not a typo. Eighty-two dollars per year.
Ghana's was $179.
South Korea had almost nothing. It had been colonized by Japan for thirty-five years, devastated by a war that killed millions, and then split in half. The half that became South Korea got the farms. The half that became North Korea got the factories, the mines, the hydroelectric dams.
South Korea in 1961 exported fish, wigs made of human hair, and cheap plywood. Its biggest source of foreign currency was the money sent home by Korean soldiers fighting in Vietnam and Korean miners and nurses working in West Germany.
The country had one natural resource: its people.
Thirty-five years later, in 1996, South Korea joined the OECD — the club of rich nations. Its per capita income was over $13,000. It was producing some of the world's best steel, the world's biggest ships, and semiconductors that powered the global electronics industry.
This is the most compressed and dramatic economic transformation in human history. And it did not happen by accident.
Look Around You
Look at the brands around you. Samsung — that is South Korea. Hyundai — South Korea. LG — South Korea. Kia — South Korea. The memory chips in your phone — there is a good chance they were made by Samsung or SK Hynix, both Korean.
Sixty years ago, none of these companies existed, or they were tiny traders selling dried fish and noodles. Samsung started as a grocery trading company. Hyundai started as a construction firm. LG started as a cosmetics and toothpaste maker.
How does a country go from exporting wigs to exporting semiconductors in a single generation?
The answer tells you more about how economies actually develop than any textbook theory.
The Idea: Infant Industry Protection
The concept behind South Korea's transformation is old. It goes back to Alexander Hamilton in the 1790s and Friedrich List in the 1840s. It is called infant industry protection.
The logic is simple and intuitive.
Imagine a country that has no steel industry. Building one requires enormous investment — blast furnaces, rolling mills, skilled workers, decades of learning. In the beginning, the domestic steel will be expensive and low quality. It cannot compete with steel from established producers in Japan or Germany.
If the country opens its market to free trade, cheap foreign steel floods in. The domestic steel industry, unable to compete, dies in its cradle. The country never develops the capability to make steel.
But if the government protects the infant industry — with tariffs that make imported steel more expensive, with subsidies that help the domestic producer invest and learn, with government procurement that guarantees a customer — then the domestic industry gets time to grow.
Over years and decades, the infant industry learns. It improves quality. It reduces costs. It develops technological capabilities. Eventually, it becomes competitive on its own. The protection can be removed. The infant has grown up.
This is the theory. The question is: does it work?
In South Korea's case, it worked spectacularly. But the way it was done matters enormously.
The Developmental State
South Korea's transformation was driven by the state. Not by free markets. Not by foreign aid. Not by natural advantages. By the deliberate, strategic, sometimes ruthless decisions of the government.
General Park Chung-hee seized power in a military coup in 1961. He was authoritarian, repressive, and utterly focused on one objective: industrializing South Korea.
Park's government did several things that were critical:
It picked winners. The government identified industries that South Korea should develop — first light manufacturing (textiles, garments, wigs), then heavy industry (steel, shipbuilding, chemicals), then technology (electronics, automobiles, semiconductors). This was not left to the market.
It directed credit. The government controlled the banks and directed lending to favored industries and companies. Firms that met export targets got cheap loans. Firms that did not were cut off. The interest rate for strategic industries was often below the rate of inflation — meaning the government was effectively paying companies to borrow and invest.
It subsidized learning. Korean firms were initially terrible at making things. Their steel was expensive. Their cars broke down. Their electronics were crude. The government subsidized them anyway — but with a critical condition: they had to improve. Subsidies were tied to performance. You had to hit export targets. You had to close the technology gap with global competitors.
It invested in education. In 1960, South Korea had one of the highest illiteracy rates in Asia. By 1980, it had near-universal primary and secondary education. By 2000, it had more college graduates per capita than most Western nations. Human capital was the foundation everything else was built on.
It demanded exports. This is perhaps the most important difference between successful and unsuccessful industrial policy. Korea did not just protect its industries from imports. It demanded that they export. Firms had to compete in world markets, even while being protected at home. This meant they could not become lazy or complacent behind tariff walls. They had to keep improving.
"The state was not a passive referee. It was the coach, the banker, the drill sergeant, and sometimes the bully. It pushed, punished, and rewarded — all in the service of one goal: making South Korea an industrial nation."
POSCO: A Steel Will
The story of POSCO — the Pohang Iron and Steel Company — illustrates how Korean industrial policy worked in practice.
In the late 1960s, South Korea decided it needed a steel industry. Steel is the foundation of industrial economies — you need it for construction, shipbuilding, automobiles, machinery, everything.
The World Bank said no. It commissioned a study that concluded South Korea had no comparative advantage in steel. It lacked iron ore, it lacked coking coal, it lacked experience, it lacked everything. The report recommended that Korea continue importing steel and focus on light manufacturing.
Park's government ignored the World Bank.
Using Japanese reparation funds (payment for colonial-era exploitation), the government built a massive integrated steel mill at Pohang, on the southeastern coast. Park Tae-joon, a former military general with no steel-making experience, was put in charge.
The early years were brutal. Everything had to be learned from scratch. Workers were sent to Japanese steel mills for training. Technology was licensed, copied, and adapted. The mill operated at a loss.
But the government kept investing. And POSCO kept learning.
By the 1980s, POSCO was one of the most efficient steel producers in the world. By the 2000s, it was among the largest. South Korea, which the World Bank said should not make steel, became the world's fifth-largest steel producer.
The lesson: comparative advantage is not fixed. It can be created. The World Bank looked at Korea's endowments in 1968 and saw no reason to make steel. Korea looked at its ambitions and made steel anyway.
Samsung: From Dried Fish to Semiconductors
Samsung's story is even more remarkable.
Samsung was founded in 1938 as a trading company, exporting dried fish, vegetables, and noodles. In the 1960s, under government direction, it moved into textiles, then insurance, then construction.
In 1969, Samsung entered electronics — making cheap black-and-white televisions using Japanese technology. The products were mediocre. No one outside Korea wanted them.
In the 1980s, Samsung made a fateful decision: it would enter the semiconductor business. Memory chips — DRAMs — were at the cutting edge of technology, dominated by Japanese and American firms. Samsung had no experience, no technology, and no customers.
It acquired technology by hiring Korean engineers who had worked at American chip companies. It reverse-engineered competitors' products. It invested billions in fabrication plants. And the Korean government supported it — with subsidized loans, with tax breaks, with diplomatic pressure on technology suppliers.
For years, Samsung bled money. Its chips were a generation behind the competition. Industry analysts predicted failure.
But Samsung kept investing through downturns — a strategy called "counter-cyclical investment." When chip prices crashed and competitors cut investment, Samsung doubled down, building new factories and developing the next generation of technology.
By the 1990s, Samsung was the world's largest memory chip producer. Today, Samsung is the largest semiconductor company in the world by revenue, the largest smartphone manufacturer, and one of the most valuable companies on earth.
From dried fish to semiconductors in sixty years.
THE INFANT INDUSTRY LEARNING CURVE
=====================================
Cost per unit
^
|
| $$$$ World Price
| $$$ - - - - - - - - - - - - - - - -
| $$$ (The line to reach)
| $$$
| $$$$
| $$$$$
| $$$$$$
| $$$$$$$$
| $$$$$$$$
+----+----+----+----+----+----+----+----+----> Time
Yr 1 Yr 3 Yr 5 Yr 7 Yr 9 Yr11 Yr15 Yr20
|<-- PROTECTION PERIOD -->|<-- COMPETITIVE -->|
| (Tariffs, subsidies, | (Can compete |
| cheap credit) | globally) |
The infant industry starts with costs far above
the world price. Protection gives it time to learn.
Costs fall through learning-by-doing, scale economies,
and technological improvement.
Eventually the industry reaches world-competitive
costs. Protection can be removed.
KEY: This only works if the industry is FORCED
to improve. Without performance pressure, the
infant never grows up.
Maruti Suzuki: India's Version
India tried something similar with automobiles, though the execution was different.
Before 1983, India's car industry consisted of two models — the Ambassador (based on a 1954 Morris Oxford design) and the Premier Padmini (based on a 1960s Fiat). Both were antiquated, inefficient, and produced in small numbers. There was a multi-year waiting list to buy a car. The industry was protected by tariffs exceeding 100 percent but had no incentive to improve.
In 1982, the Indian government partnered with Suzuki of Japan to create Maruti Udyog. It was a joint venture — the Indian government held the majority stake, and Suzuki provided the technology and management.
The Maruti 800, launched in 1983, was a revelation. Small, fuel- efficient, reliable, and affordable (by Indian standards), it transformed the Indian automobile market. For the first time, a middle-class Indian family could aspire to own a car.
Maruti Suzuki also built a supplier ecosystem — hundreds of small and medium companies that made components for the car. This is the backward linkage effect that manufacturing creates.
Today, India has a significant automobile industry — Tata Motors, Mahindra, Maruti Suzuki, Hyundai India — that employs millions directly and indirectly, and exports vehicles to dozens of countries.
But the contrast with South Korea is instructive. Korea built Hyundai as a nationally-owned champion that developed its own technology, designed its own engines, and competed globally under its own brand. India's most successful car company remains a subsidiary of a Japanese corporation.
Both approaches worked, to different degrees. But one created deeper technological capabilities than the other.
What Actually Happened
Industrial policy — the practice of governments actively shaping their industrial structure — has a mixed record. For every South Korea, there is a counterexample.
India's import substitution industrialization from the 1950s to the 1980s protected domestic industry but without the performance pressure that made Korean policy successful. Indian firms did not have to export. They did not have to improve. Protection became permanent rather than temporary, and many protected industries became inefficient monopolies.
Latin America's experience was similar. Brazil, Argentina, and Mexico pursued import substitution from the 1950s to the 1980s. They built some industrial capacity, but much of it was uncompetitive. When these countries were forced to liberalize (often by debt crises), their protected industries collapsed.
The difference between success and failure was not whether the government intervened — it was how it intervened. The successful cases (Japan, Korea, China) combined protection with performance requirements. The unsuccessful cases protected without demanding improvement.
Protection without discipline breeds inefficiency. Protection with discipline builds capability.
When Industrial Policy Fails
Not all protection works. The line between nurturing an infant industry and coddling an incompetent one is thin and easily crossed.
India's License Raj (1951-1991). India protected its industries with some of the highest tariff walls in the world. But it did not demand exports. It did not force competition. It created a system where you needed government permission to produce, to expand, to import, even to close a factory. The result was a protected, stagnant industrial sector that fell further and further behind global standards.
The Ambassador car is the perfect symbol: designed in the 1950s, still in production with minimal changes forty years later. No competitive pressure meant no reason to innovate.
Argentina's automobile industry. Argentina protected its car industry for decades. But domestic producers never achieved the scale or quality to compete internationally. When protection was reduced in the 1990s, foreign firms took over. Argentina's "national" car industry was a fiction sustained by subsidies.
Sub-Saharan Africa's import substitution. Many African countries attempted import substitution after independence in the 1960s. They built factories behind tariff walls. But the domestic markets were too small to achieve economies of scale. The bureaucracies were too weak to enforce performance standards. The result was small, expensive, inefficient factories producing goods that nobody wanted to buy.
The pattern is consistent: protection alone is not enough. You need protection plus discipline. The infant must be nurtured and pushed.
"There is a fundamental difference between protecting a child so it can grow and protecting a child so it never has to."
The Recipe: What Works
Looking at the successful cases, a pattern emerges:
Start with education. Every successful industrializer invested heavily in education before or simultaneously with industrialization. Literate, numerate workers can be trained. Illiterate workers cannot.
Protect strategically. Do not protect everything. Identify industries with high learning potential, strong linkages to other sectors, and export possibilities. Protect those.
Demand performance. Tie subsidies and protection to measurable outcomes — export targets, quality improvements, cost reductions. Remove support from firms that fail to improve. This is the crucial discipline mechanism.
Invest in technology. Send students abroad. License foreign technology. Build research institutions. Create incentives for firms to invest in R&D. The goal is not just to make things but to learn how to make things better.
Build infrastructure. Factories need roads, ports, reliable electricity, and clean water. No amount of industrial policy compensates for crumbling infrastructure.
Have a plan — and revise it. Industrial policy requires strategic thinking about which industries to pursue and in what sequence. But it also requires flexibility — the willingness to abandon failures and double down on successes.
Japan: The Model
Japan's post-war industrial policy, guided by the Ministry of International Trade and Industry (MITI), is the template that Korea and others followed.
MITI identified target industries — first steel and shipbuilding, then automobiles, then electronics, then high technology. It coordinated investment, directed credit, brokered technology agreements with foreign firms, organized research consortia, and managed the transition from declining to emerging industries.
Japanese firms were fiercely competitive with each other — Toyota versus Nissan, Sony versus Matsushita — but they cooperated on pre-competitive research and technology development, with MITI as the broker.
The system was not without failures. MITI tried to consolidate Japan's car industry into a few national champions in the 1960s — Toyota resisted, and MITI backed down. Not every bet paid off. But the overall direction was clear and sustained: move up the value chain, from simple to complex, from imitation to innovation.
By the 1980s, Japan was the second-largest economy in the world and a technological leader in automobiles, electronics, robotics, and materials science.
China: The Latest Chapter
China's industrialization follows the same logic, adapted to Chinese conditions.
China used several distinctive strategies:
Special Economic Zones. Starting in 1980, China created zones where foreign firms could operate with tax breaks and relaxed regulations. The goal was not charity toward foreign capital — it was technology transfer. Foreign firms brought technology, and Chinese firms learned from working with them.
Forced technology transfer. Foreign companies that wanted access to China's vast market were often required to form joint ventures with Chinese firms and share technology. Western firms complained, but they complied, because the Chinese market was too large to ignore.
Massive infrastructure investment. China built more highways, railways, ports, and power plants in thirty years than most countries build in a century. This gave its manufacturers a logistical advantage that few competitors could match.
Scale. China's sheer size — 1.4 billion people — created a domestic market large enough to achieve economies of scale in almost any industry. Chinese factories could produce at volumes that drove costs to levels no competitor could match.
The result: China now produces more manufactured goods than the United States, Japan, and Germany combined.
Think About It
The World Bank told South Korea not to build a steel industry. Korea ignored the advice and built one of the world's best. What does this tell you about the limits of expert advice?
Why did South Korea's industrial policy succeed while India's License Raj largely failed, even though both involved government direction of the economy?
Samsung went from dried fish to semiconductors in sixty years. Could an Indian company do the same? What would it need?
China required foreign firms to share technology as a condition of market access. Was this fair? Was it smart? What would you do if you were in charge of a developing country's industrial policy?
The successful cases all involved authoritarian governments (Park in Korea, the LDP in Japan, the CCP in China). Does industrial policy require authoritarianism? Can a democracy do it?
The Difference That Matters
Let us be precise about the difference between protection that builds and protection that breeds dependency.
It is not the amount of protection. Korea's tariffs were high. India's were high too.
It is not the level of government involvement. Korea's government was deeply involved. India's government was deeply involved too.
The difference is discipline.
Korea protected its firms but demanded that they export. India protected its firms and let them sell only at home.
Korea gave subsidies but withdrew them from failures. India gave subsidies and made them permanent.
Korea used protection as a temporary bridge to competitiveness. India used protection as a permanent shelter from competition.
Korea's infant industries grew up. India's infant industries remained infants.
This is the hardest lesson of industrial policy: the government must be willing to be tough with the very firms it is helping. It must nurture and discipline simultaneously. It must protect and pressure at the same time.
This requires a rare combination of strategic vision, institutional capacity, and political courage. Not every government has it. But the ones that do have achieved extraordinary results.
The Bigger Picture
The story of how countries build industries from nothing is, at its core, a story about what human societies can achieve through deliberate, collective effort.
South Korea in 1961 was a devastated, impoverished country with nothing but its people and the determination of its leaders. In one generation, it became one of the most technologically sophisticated economies on earth.
This did not happen naturally. It was not the invisible hand of the market. It was not foreign aid. It was not luck. It was the visible, forceful hand of a state that decided to industrialize and then mobilized every resource to do so.
The lesson is not that government always knows best. MITI made mistakes. The Korean government backed losers. China's state-owned enterprises include many inefficient ones.
The lesson is that markets alone do not build industrial economies. They never have, anywhere, at any time in history. Every successful industrialization has involved the state — protecting, investing, directing, disciplining.
The question for developing countries is not whether the state should be involved. It is how to make state involvement effective. How to protect without coddling. How to subsidize without creating dependency. How to pick winners without ignoring losers. How to maintain discipline when the pressure to relax it is enormous.
These are hard questions. But they are the right questions. And the countries that answer them well will be the next South Koreas.
The countries that do not ask them at all will remain where they are.
"Countries do not grow rich by accident. They grow rich by design." — A lesson from the East Asian miracle
In the next chapter, we ask a related but different question: why do some countries make things while others only dig things up? The answer involves a curse — the resource curse — and it explains why being blessed with oil or diamonds can be the worst thing that ever happens to a nation.