How Did China Change So Much in One Generation?
Reform tested in four small zones, a hundred million factory jobs nobody planned, and decades without reversal. How China changed, and what it cost.
The film plays here on the page · 9:45
Put a photograph of Shanghai in 1993 beside the same view now. Western cities show that kind of change across centuries. This took about one generation.
At one point a 57-storey building went up in 19 days. That is a stunt. The interesting question is how the machine behind it works, because doing it for 40 years is a different thing from doing it once.
An old idea: rule by results
For most of Chinese history, the right to rule was said to come from heaven, and it had terms. The emperor had to keep proving he still held it.
The proof was whether things worked. Rain on time, harvests up, rivers behaving, people fed. A flood or a famine was evidence against the government.
Take the religion out and what remains is legitimacy by performance. You rule because it is working.
Three catastrophic decades
Any honest version starts here. A campaign to industrialise farming overnight produced the deadliest famine in recorded history, with deaths in the tens of millions. Then a decade of political violence closed the universities.
In 1976 China was poorer per person than most of sub-Saharan Africa. Four in five people lived in the countryside.
Reform as an experiment
Half the leadership believed markets were the enemy. So nobody proposed changing the country. They proposed changing four small places.
These were special economic zones, with different rules inside a fence. One was a fishing town of about 30,000 people called Shenzhen. It now has more people than London.
If the experiment failed, it failed in four places and could be quietly reversed. It did not fail, so it was copied for 30 years. Announce a plan and you have to defend it. Run a trial and you can learn from it.
The factories nobody designed
In the countryside, village and township enterprises appeared. They were factories owned by local governments, a legal shape that existed because private ownership was still politically impossible.
At their peak they employed well over 100 million people. They were often badly run. They also absorbed a huge amount of rural labour and taught millions of people what a factory is.
Most were closed or privatised by the late 1990s. Nobody had planned them. They were tolerated, then noticed, then encouraged.
Making things badly
For 20 years "made in China" meant cheap, and that was fair. Every country that got rich started there, including Japan, Korea and Germany.
The product is what the making teaches you: logistics, quality control, maintenance and scheduling. Do it long enough and the same workforce makes harder things, then components, then the machines that make the components.
China now makes more than a third of everything manufactured on Earth.
A long horizon
A high-speed rail network takes longer to build than one term of office. So does a power grid or a generation of engineers. China built more high-speed rail in 15 years than the rest of the world combined, partly because nobody cancelled it halfway.
For years, most of the top leadership held engineering degrees. This describes the system. It is not an argument for it.
The price
None of this is disputed.
- There is no way to remove a government that stops delivering.
- The authority that built the railways also decides what you may say about them.
- Entire regions live under heavy surveillance.
- The one-child policy ran for 35 years, and the demographic bill is arriving now.
- Local governments funded themselves by selling land, and whole districts stand empty.
- Air pollution shortened lives by years, before a rapid clean-up.
A model that works is not the same as a model worth copying whole.
What transfers
Almost none of the useful parts need a one-party state.
- Pilot things somewhere small before announcing them everywhere.
- Let awkward institutions exist if they employ people, and tidy them up later.
- Protect beginners in manufacturing, then remove the protection on a published date.
- Fund the boring infrastructure that pays back after you have left office.
- Train more engineers than the economy seems to need.
The short answer
Pilot small. Employ people in whatever shape works. Make things badly, and do not stop. The price was paid by people who did not get a vote on it, and that belongs in the summary too.
This is the written version of a Mr. Orivo film. It is made to explain and to make you curious. It is not medical, legal or financial advice.


