South Korea Defied the Gods to Build its Steel Colossus

ELI5/TLDR

In the 1960s, South Korea was dirt poor — about $80 per capita, still wrecked from the Korean War. Their dictator-president Park Chung-hee decided the country needed an integrated steel mill, the kind only rich industrial nations had any business building. The World Bank said no. The Americans said no. The Germans and British said no. So Park used Japanese war reparations money to build it anyway, leaning on a former soldier named TJ Park to actually run the project. The result was POSCO, which today is one of the eight largest steelmakers in the world and runs more efficiently than the American mills it once tried to imitate.

The Full Story

A poor country with a fixed idea

When Park Chung-hee took power in a 1961 military coup, South Korea was the kind of country development economists wrote textbooks about — and not in a flattering way. The Bank of Korea estimated that 75% of pre-war steel and iron capacity had been destroyed in the Korean War. Most of the colonial-era mills had been built in the north anyway. Per capita income was around $80. Inflation was rampant, savings were nonexistent, and the country imported almost everything that mattered.

Park’s fix was a forced industrialization push, and at the heart of it was steel. He believed you could not build cars, ships, or weapons without first being able to make your own steel. So he made it the centerpiece of his first two five-year plans, despite every credible outsider telling him this was the wrong sequence. The accepted wisdom was that poor countries should start with light industry — textiles, simple consumer goods — and work their way up. Steel came later, if at all.

The KISA fiasco

The first attempt to fund the mill collapsed in slow motion over several years. Park’s people approached West Germany. They approached an American consortium. They approached USAID and the US Export-Import Bank. Everyone either said no or signed agreements that died in the cradle. The Americans in particular were uncomfortable: Park was a military dictator, and the US wanted him moving toward civilian rule, not getting rewarded with development loans.

In 1966, Park finally cobbled together a seven-member US-European consortium called Korea International Steel Associates, or KISA. The plan: raise $130 million for a 600,000-ton mill. The plan died in 1968 when the World Bank formally declined a loan, citing a feasibility study that said the largest mill South Korea could economically support was 500,000 tons. The Koreans wanted 1-2 million. The World Bank suggested they try light industry instead.

Japanese reparations as a lifeline

Park’s pivot was to Japan, which was politically loaded — Japan had occupied Korea for 35 years and the wounds were fresh. But Park himself had been educated in Japanese schools and trained in the Japanese military, and he was pragmatic enough to normalize relations in 1965. Part of that deal: $500 million in Japanese grants and long-term loans, framed as reparations.

When KISA fell apart, Park decided to redirect those reparations funds into the steel mill. Japan was reluctant at first — they shared the World Bank’s skepticism. But Japanese steelmakers were already oversupplied at home and looking for export markets for their equipment. The math worked. Sell the Koreans the older slabbing technology, hold back the cutting-edge continuous casting process, and pocket the equipment sales. In late 1969, after marathon negotiations, a deal was signed for a $123 million, 1-million-ton mill. Construction would start in April 1970.

TJ Park and the backwards mill

The man running it was Park Tae-Joon — TJ, in the video’s shorthand to avoid the four-Park pile-up. A former soldier and chief secretary to the president, TJ had earned trust by turning around the state tungsten mine. He was given the steel project and a sleepy fishing town called Pohang as the site, chosen for its deep-water harbor, available land, and safe distance from both North Korean artillery and Seoul.

TJ did one genuinely clever thing during construction. Steel mills are normally built “forwards” — iron-making first, then casting, then rolling — which means no revenue until the whole thing is finished. TJ flipped it. He built the rolling mill first, imported semi-processed steel slab, and started selling finished products while the rest of the plant was still being built. POSCO turned a profit in its second year. This was not how things were supposed to work.

He also extracted brutal effort from his workers. Sandstorms, sleeping on roadsides, mass concrete pours to catch up on schedule. There is a famous (and probably real) speech where TJ told his crew that if they failed, they should march out of the office, turn right, and throw themselves into Yeongil Bay — because they were spending money paid in the blood of their ancestors.

Student to master

Phase 1 wrapped on time in July 1973, taking three years against the typical 4-9 elsewhere. Park then set a new target: 10 million tons by 1980. POSCO added almost 6 million tons across phases 2-4, with Korean engineers gradually taking over from their Japanese teachers.

The real leap came with Gwangyang, a second mill authorized in 1979 (apparently in a meeting hours before Park Chung-hee’s assassination). Built starting in 1982, Gwangyang used every modern technology Korea had once been denied — basic oxygen furnaces, continuous casting, computer control — and laid them out in a single innovative line. The result was a steel mill operating at 60% the cost of the average American mill. POSCO began selling into Japan, competing with the same companies that had taught it.

Key Takeaways

Park Chung-hee’s bet on heavy industry violated standard development advice. The World Bank explicitly told South Korea to stick with light industry. He ignored them and was right.
The project only happened because of Japanese reparations money. Western lenders had refused.
Japan helped partly out of self-interest — their domestic steel market was oversupplied and selling Korea older equipment was good business.
TJ Park’s “build it backwards” trick — rolling mill first, then iron-making — let POSCO generate revenue during construction and turn a profit in year two.
POSCO’s 1973 build took three years versus the typical 4-9 elsewhere. The Gwangyang mill, finished a decade later, ran at 60% the cost of an average American mill.
This is one of the cleanest examples of state-led industrial policy actually working. It also depended on a specific combination of an authoritarian government, friendly labor conditions, geopolitical luck, and a competent operator.

Claude’s Take

Asianometry is reliably good at this kind of thing — granular, sourced, no triumphalism, no editorializing. The video doesn’t try to extract a lesson, which is the right call because the lesson is murky. POSCO worked, but it worked because Park Chung-hee was an authoritarian who could arm-twist chaebol, seize the central bank, evict residents from a fishing village by special law, and override the World Bank. The standard development-economics objections were not stupid; they just turned out to be wrong in this one case for reasons that are hard to generalize.

The TJ Park material is the best part. The “build it backwards” decision is the kind of operational creativity that gets lost in macro-level industrial policy stories. Most of what we hear about Korean development is about Park Chung-hee setting targets and chaebol executing. The actual mechanics — getting the rolling mill running first to fund the rest, the back-to-back negotiations with MITI, scrounging coal supplies independent of the Japanese trading houses — is where the real story lives. Score: 8. Solid, well-paced, and it teaches you something concrete about how a country actually builds an industry from nothing.