For all the attention paid to America's semiconductor reliance on Taiwan and the military consequences of selling advanced Nvidia GPUs to China, I've found it unusual how comparatively little effort has been placed in coming up with an alternative to the Shenzhen manufacturing ecosystem. Before reading The Shenzhen Experiment I didn't know much about this region in south China other than that it is responsible for a staggering amount of the world's electronics manufacturing output. This Christmas I received Juan Du's The Shenzhen Experiment: The Story of China's Instant City; while the book didn't explain as much as I had hoped about electronics manufacturing, it is nonetheless excellent.
In the late 1970s, Guangdong province officials wanted greater policy flexibility from the central government to attract exporting industries and foreign direct investment. Their wish was granted on July 19th, 1979 when the party Central Committee authorised what would become Shenzhen's special economic zone (SEZ). This was a mostly rural strip of land a little smaller than Philadelphia, home to about 358,000 people at the time. After Mao Zedong's death, many party leaders were eager to roll back Mao-era economic centralisation to increase the country's economic growth.1 This culminated in the rise of Deng Xiaoping, who is more responsible than any other Chinese leader for the impressive economic growth of the country starting in the early 1980s. The political and economic autonomy granted to the SEZ made it something of a laboratory of liberalising economic reforms: experiments that worked in Shenzhen were applied to the rest of the nation.
Because of this, "China's first" is a phrase that comes up a lot in The Shenzhen Experiment. This includes:
These led to the early 1980s Shenzhen construction boom: in 1985 over 10 million m2 of floor area was under construction, and as of 1987 there were 62 towers taller than 100m in the SEZ. Despite providing political cover for the SEZ and championing liberalisation in general, Deng Xiaoping only visited the city twice in his lifetime. In his first visit to the city during the height of the construction boom in 1984, he said the following:
This time I went to Shenzhen to take a look; the impression it gave me was one of widespread prosperity and growth. The construction speed of Shenzhen is very fast, building one floor every few days and a tall building in no time. The construction crews there are even from the inland provinces. One reason for the high efficiency is the contract system and fairness in administering reward and punishment. Shenzhen's Shekou Industrial Zone is even faster. The reason is they were given a bit of power. They can make their own decisions on expenditure under five million US dollars.
However, many in the party were sceptical about both liberalisation in general and the SEZ's success in particular. While an impressive amount of construction was going on, not much growth was coming from exports and foreign investment but rather from other provinces and SOEs making one-time infrastructure investments in the city. In 1985 vice premier Yao Yilin remarked "It is impossible for the SEZ's economic development to rely on the country's long-term 'blood transfusion'; now, the 'needle' must be unplugged decisively", and outside observers from Hong Kong started to notice that the majority of the SEZ's revenue came from infrastructure and speculative real estate development. Worse yet to the central government, importers were a substantial part of the non-construction economy.2 It started to look like a convenient talking point to those who saw SEZ reforms as betraying the basic principles of the Party. The political blowback was swift, and led to the suspension of 804 infrastructure projects, hotel occupancy dropping 20%, and widespread unemployment in the construction sector in the mid-1980s.
This may have been the end of the Shenzhen experiment if it were not for changes in the foreign-exchange market. The Hong Kong dollar weakened against both the Japanese Yen and the Taiwanese Dollar, so Hong Kong looked north of the border for manufactured goods in 1986. Shenzhen's manufacturers were moving up the value chain, aided both by the market reforms and investment in infrastructure, so by 1987 the city was meeting the ambitious production and revenue targets set by the central government. Industrial production kept outpacing construction as an economic driver of the SEZ, leading to the 1988 conversion of state-owned enterprises into shareholding cooperatives and constitutional changes to recognise private enterprise.
While Shenzhen ended up being an economic powerhouse and a runaway success, the Shenzhen of today is not something that Guangdong provincial officials were trying to accomplish in the late 1970s. This is because emigration was a daunting problem for the region, emphasis mine:
The most severe problem facing the local government was the abandonment of village communes and farming fields in the "Great Escape to Hong Kong," a phrase coined to describe the successive waves of hundreds of thousands of people illegally crossing the China-Hong Kong border. A recently declassified Guangdong government internal report reveals that between 1954 and 1980 there were 565,000 officially recorded crossing attempts.
The state of the Chinese economy at this time drove disparities across the border that are hard to overstate, which drove so many to risk the crossing:
The stark contrast in economic wealth between Chinese Guangdong Province and colonial Hong Kong next door loomed clearly. In 1977, the annual income of a village farmer in Hong Kong was one hundred times that of a farmer undertaking the same work on the other side of the border, while the disparity between factory workers in Guangdong and Hong Kong was even greater.
Hong Kong's immigration policies varied in the 20th century, but between 1974 and 1980, Hong Kong had a 'Touch Base' policy, which allowed for any illegal immigrant to get a residency permit provided they both had relatives in the territory and could get to the city centre without being apprehended.3 Cross-border farming certificates were granted to Hong Kong residents with ancestral farmland in Guangdong, making the border rather porous in some areas and facilitating many escapes to the south.
In 1980, city officials projected a population of 500,000 by the end of the millennium, showing that stopping the bleeding was their primary concern. But in 2000 the city had well more than 6 million people and became a metropolis of 20 million into the 2010s. In most of the rest of the country the state tightly controlled rural migration and employment through the hukou system under stiff penalties, but in the 80s and 90s many firms in the SEZ would break the law to hire informal labor. City officials would look the other way and lie to the central government about how many unregistered rural migrants lived in the city.4
The unregistered migrants were half of Shenzhen's population by the 1990s, but without an urban hukou status they couldn't receive state housing and other welfare benefits. Shenzhen's 'urban villages' that existed before the establishment of the SEZ helped fill the gap and are a major focus of the book; in describing them Du's architecture and urban planning background becomes more obvious.
Residential land use changes in many urban villages went something like this:
It is a good demonstration of what laissez-faire economists might diplomatically describe as 'evasive entrepreneurship', both in policymaking and the marketplace.
A story that Du really effectively tells is that Shenzhen's success is much more than an 'if you build it, they will come' with respect to market liberalisation. Being across the border from Hong Kong really mattered for the development of the SEZ, and despite this the project very nearly didn't succeed. Many wealthy Hong Kong emigrants remained in touch with kin and invested in Shenzhen before other foreigners were willing to. It is hard to see how private firms could have gotten off the ground absent this outside capital. The Shenzhen Experiment shows many other place-specific factors behind Shenzhen's success, and these help explain why it worked out so well while China's 17 other SEZs haven't been remotely as successful.
While every one of Shenzhen's 300 urban villages was impacted by the explosion of growth, the post-1979 fortunes of these villages can range substantially. One village that rode the wave of prosperity was Huanggang.5 In 1979 the local economy was dire enough that Shangwei nearly split off as a separate village, but the commune remained intact through the boom of the 1980s. The commune invested proceeds from land sales to the city into the Shapuwei Industrial Zone, and the villagers themselves converted their homes into residential towers to collect rents. In 1992 the villagers received urban hukou status and formed the Shenzhen Huanggang Real Estate Holdings Company with each villager as a company shareholder, and the redevelopment of the village continued apace with the central business district encroaching into Huanggang. The village had a ten thousand yuan deficit in 1980 but as of 2002 its holdings reached 450 million yuan and paid villager-shareholders a twenty thousand yuan yearly dividend.
But east of Huanggang about 15 km are the Baishizhou villages, which are among the poorest in Shenzhen. The area around today's Baishizhou was the Shahe collective farm, much of which was settled by peasants ordered to move from rural Guangdong province in the late 50s and early 60s. While land claims in Huanggang were unambiguous, it wasn't clear where collective Baishizhou land ended and the state-run Shahe farm began, which prevented the village from starting a shareholding collective once they were granted urban hukou status. Unlike Huanggang, the village wasn't compensated by the city for land transfers during urbanisation owing to this ambiguity. For many Baishizhou residents, their legal status amounted to "rural migrants squatting on urban land" rather than collecting dividend checks from village owned real estate investments.
As an aside, Baishizhou seems more interesting than Huanggang and other wealthier parts of Shenzhen. It remains an affordable place to rent, making it a common first neighborhood to live in for those getting their start in Shenzhen, and the large migrant population means you can find regional food from every corner of China in the village. The introduction to the book describes Du accidentally wandering into a bustling night market, but the book's most absurd Baishizhou detail is American expat Joe Finkenbinder's Bionic Brewery, opened in the village in 2014 because rents were too high everywhere else in Shenzhen.6
I don't think of myself as a particularly fast reader, but I finished The Shenzhen Experiment in three sittings. There is a lot worth saying about the book that is left out here, and overall it does a great job at both the high-level history and telling individual stories about the evolution of Shenzhen. It was the best book I read in 2025, and I'd recommend it to anyone remotely interested in economic development or urban planning.
Mao-era rule of China was an abject tragedy, as outlined in Frank Dikötter's "The Tragedy of Liberation", "Mao’s Great Famine", and "The Cultural Revolution", all published by Bloomsbury Press↩︎
Yuxi Liu's 'Structure and Interpretation of the Chinese Economy' explains how important getting foreign currency into the country was at this stage of China's development, so a large import sector was doubly bad news. As an aside, if you name an essay 'Structure and Interpretation of $topic' as a nod to Abelson and Sussman's seminal functional programming text, it better be a good explanation of $topic. Luckily the essay lives up to the name.↩︎
I originally ended this paragraph with "Cross-border farming certificates were granted to Hong Kong residents with ancestral farmland in Guangdong, making the border rather porous in some areas and facilitating many escapes to the south.", but when someone on Reddit raised some questions I checked back in the text and realised that pages 210-211 described a border that Hong Kongers could cross without much issue but that this didn't extend to Guangdong side of the border.↩︎
I'm playing a little fast and loose here in this paragraph in that the SEZ was only one part of Shenzhen city in the 20th century. In general Chinese cities seem to be larger geographic areas than in the West; today the city is around the size of Orange County, California. A map on page 58 makes the demarcation clearer.↩︎
Unlike most other Shenzhen villages, Huanggang villagers defied 17th-century imperial orders in the late 17th century to abandon the village lest their resources be seized by rebel forces.↩︎
Having the confidence to start gentrifying a neighbourhood in China is quite American, and I mean that in a profoundly positive way.↩︎