The City Without People
From the Great Clearance to Artificial Intelligence: An Alternative Future for Shenzhen’s Population and Economy
I. The First Great Clearance
Beginning in 1661, the eighteenth year of the Shunzhi reign, the Qing government implemented a policy of coastal evacuation to sever connections between the inhabitants of southeastern China and the maritime forces of the Zheng family. Large numbers of coastal residents were forced to abandon their homes.
Xin’an County in Guangdong was particularly hard hit. Villages were destroyed, farmland lay abandoned, salt fields ceased production, and coastal trade was disrupted. In 1667, the sixth year of the Kangxi reign, Xin’an County was abolished altogether and incorporated into neighboring Dongguan County.
In 1669, Xin’an was restored as a county, and coastal residents were gradually permitted to return. But the return of a population did not mean the immediate restoration of society. Many former inhabitants never came back, while agriculture and salt production could not be revived overnight. Local authorities had to recruit settlers and encourage land reclamation to restore the population and economic activity.
We have already examined this history in the second and third essays of this series. It reveals a fairly simple fact: under the conditions of production prevailing at the time, for a place to lose its population was also to lose its productive capacity. Fields needed farmers, salt fields needed registered salt-producing households, fishing boats needed fishermen, and marketplaces needed merchants. Without sufficient people, these economic activities could not be sustained.
One of the most pressing problems facing the local government after the reopening of Xin’an’s coast, therefore, was how to bring people back.
More than three centuries later, Shenzhen, built on the former territory of Xin’an, experienced a population movement in precisely the opposite direction. Tens of millions of people came to the region over the decades, and factories, ports, roads, residential buildings, and commercial centers rose around them. What had once been a border city with a relatively small population developed into one of China’s most important economic centers.
Shenzhen’s history seems to demonstrate a principle: economic development requires people, while economic opportunities attract still more people.
But will this relationship necessarily continue forever?
II. A City That Needed People
Following the establishment of the Shenzhen Special Economic Zone, the city’s population grew rapidly. Around 1980, Shenzhen had only a few hundred thousand residents. By the time of China’s seventh national population census in 2020, its permanent resident population had reached approximately 17.56 million. In 2025, the figure was approximately 18.25 million.
Table 1. Changes in Shenzhen’s Population
| Year | Population | Statistical Notes |
|---|---|---|
| Around 1980 | Just over 300,000 | Early population statistics; definitions differ from those used in later years |
| 2000 | Approximately 7.01 million | Permanent resident population, Fifth National Population Census |
| 2010 | Approximately 10.42 million | Permanent resident population, Sixth National Population Census |
| 2020 | Approximately 17.56 million | Permanent resident population, Seventh National Population Census |
| 2025 | Approximately 18.25 million | Year-end permanent resident population |
Sources: National population censuses and population data released by Shenzhen’s statistical authorities. Early statistics are not entirely comparable with later figures for the permanent resident population and should not be used directly to calculate long-term population growth rates.
Why was Shenzhen able to attract such an enormous population? There were, of course, many reasons, but employment opportunities were undoubtedly among the most important.
In the early years of reform and opening up, large numbers of factories needed workers. Subsequently, construction, commerce, logistics, finance, information technology, and a wide range of service industries expanded, creating jobs of many different kinds. Workers from across China came to Shenzhen, where they found employment, made homes, and raised families. Their arrival became a principal driving force behind the city’s population growth.
Shenzhen’s industrial structure was also changing. The relative importance of labor-intensive processing and manufacturing gradually declined, while electronics, advanced manufacturing, software, finance, and modern services continued to develop.
Yet regardless of how its industries changed, they continued to need people. Factories needed workers and engineers, software companies needed programmers, financial institutions needed professionals, and logistics companies needed drivers, warehouse workers, and managers.
Shenzhen grew to its present size amid this continuously expanding demand for human labor.
III. How Many Workers Does Shenzhen Need?
China’s Fifth National Economic Census, conducted in 2023, provides a relatively detailed picture of Shenzhen’s employment structure.
According to the census, legal-entity establishments in Shenzhen’s secondary and tertiary industries employed approximately 13.02 million people.
This figure is neither the city’s total employed population nor its permanent resident population. It measures people employed by legal-entity establishments, a particular statistical category. Nevertheless, it provides a useful indication of the extent to which Shenzhen’s economic activities depend on human labor.
Table 2. Employment in Major Shenzhen Industries, 2023
| Industry | Employees (approx.) |
|---|---|
| Manufacturing | 3.905 million |
| Wholesale and retail trade | 1.854 million |
| Leasing and business services | 1.292 million |
| Construction | 1.031 million |
| Information transmission, software, and information technology services | 969,000 |
| Real estate | 705,000 |
| Scientific research and technical services | 549,000 |
| Transportation, warehousing, and postal services | 537,000 |
| Finance | 418,000 |
| Education | 372,000 |
| Accommodation and food services | 343,000 |
Source: Relevant reports from Shenzhen’s Fifth National Economic Census. Figures refer to employees of legal-entity establishments in selected major industries within the secondary and tertiary sectors and have been rounded. They do not represent the city’s entire employed population.
This table deserves a closer look.
Manufacturing remains one of Shenzhen’s largest sources of employment, accounting for nearly four million employees of legal-entity establishments. Of these, approximately 1.55 million work in the manufacture of computers, communications equipment, and other electronic devices alone.
But Shenzhen is far more than a manufacturing city.
Wholesale and retail trade, business services, construction, information technology, real estate, research, logistics, finance, and education also employ substantial numbers of people. Many of these jobs do not directly produce physical goods, yet they are indispensable to the functioning of a modern urban economy.
The figures also suggest that any discussion of artificial intelligence’s impact on Shenzhen must look beyond robots on factory floors.
Artificial intelligence may transform software development, financial analysis, design, administration, customer service, and research. Robots and automated equipment, meanwhile, may transform manufacturing, warehousing, transportation, construction, and other industries involving physical labor.
In other words, Shenzhen’s industrial structure makes it a particularly interesting place in which to examine the potential effects of artificial intelligence and robotics on employment.
IV. When Machines Take Over the Work
Machines replacing human labor is nothing new.
Since the Industrial Revolution, mechanization and automation have continually transformed production. Agricultural machinery reduced the demand for farm labor. Industrial machines increased factory productivity. Computers subsequently changed the nature of office work.
But technological progress in the past often created new jobs even as it eliminated others.
As agricultural employment declined, industry absorbed large numbers of workers. As industrial productivity increased, services and knowledge-intensive industries created new employment opportunities.
We therefore cannot simply conclude that because a machine can perform a particular task, the person currently performing that task will necessarily lose their job. Still less can we conclude that a city’s population must consequently decline.
What may distinguish artificial intelligence and robotics from earlier technologies is the range of human activities they could affect.
Traditional machines primarily replaced physical labor, while computers mainly assisted people in processing information. More advanced artificial intelligence systems may increasingly undertake tasks requiring judgment, planning, analysis, and coordination.
If such systems are combined with robots capable of increasingly sophisticated physical action, they may enter areas of production and service that previously required human participation.
Does this mean that artificial intelligence will eventually be able to perform the overwhelming majority of economic activities?
At present, there is insufficient evidence to answer that question in the affirmative. Technological capabilities, actual costs, safety requirements, legal liability, and social acceptance may all limit the speed and extent of automation.
But we can propose a hypothesis.
Suppose artificial intelligence and robotics continue to advance, eventually becoming capable of reliably performing an increasing range of productive tasks at a lower cost than human labor.
What would happen to Shenzhen’s economy?
We do not need to predict how many robots will exist in a particular year, nor must we declare that a certain number of jobs will inevitably disappear.
We need only examine one possibility: a city’s productive capacity may continue to increase even as its demand for human labor ceases to grow alongside it.
V. Economic Growth, Population Decline?
We ordinarily associate economic growth with employment growth.
When businesses expand production, they hire more workers. When commercial activity increases, more service employees are needed. Urban development creates additional jobs in construction, transportation, and public services.
Shenzhen’s history over the past several decades has broadly followed this pattern.
But economic growth does not necessarily entail employment growth.
Imagine a factory that originally employed ten thousand people. After introducing automation, it can produce more goods with only one thousand workers.
The factory’s output has increased, but its workforce has shrunk.
If similar changes occur across many businesses, a city’s total economic output could continue to grow even as its demand for workers declines.
Of course, developments could also move in the opposite direction.
Lower production costs might allow companies to expand. New products and services might create new kinds of employment. Lower prices might stimulate consumption, increasing employment in other industries.
The extent to which technological progress reduces employment must therefore be examined in light of actual economic conditions. It cannot be inferred from technological capabilities alone.
But if automation becomes sufficiently widespread, and if the jobs it creates are insufficient to offset those it displaces, a possibility rarely contemplated in accounts of Shenzhen’s development may emerge:
The economy continues to grow, while the population begins to decline.
According to Shenzhen’s 2020 population census, approximately two-thirds of its permanent residents did not hold Shenzhen household registration, or hukou.
This proportion should not be taken to mean that two-thirds of the population are ready to leave at any moment. Many residents without local hukou have lived in Shenzhen for years, establishing families, social relationships, and stable lives.
Nevertheless, this demographic structure demonstrates the enormous scale of Shenzhen’s population originating from elsewhere in China. Employment opportunities were an important reason why many of these people came.
If Shenzhen’s demand for migrant labor were to decline persistently, fewer newcomers might arrive, while some existing residents might choose to leave.
Even without a massive wave of unemployment, a city’s permanent resident population could gradually decline if departures consistently exceeded arrivals, depending also on births and deaths.
Several distinct steps in this reasoning must be kept separate.
The fact that artificial intelligence can perform a job does not mean that businesses will necessarily adopt it.
The adoption of artificial intelligence does not necessarily mean that total employment will decline.
And a decline in employment does not necessarily mean that the population will shrink.
Yet if these conditions were to occur in succession, Shenzhen might enter a pattern of development fundamentally different from that of the past four decades.
Its economy would no longer depend on a continuously expanding population.
VI. A Second Exodus
We can now return to the history of Xin’an in the early Qing period.
The first Great Clearance was an act of state coercion. The Qing government ordered coastal inhabitants to abandon their homes. The resulting population loss devastated production, and the local economy fell into decline.
After the coast was reopened, the government had to encourage people to return because, without sufficient labor, the economy could not recover.
Now imagine another movement of people out of Shenzhen at some point in the future.
No government orders them to leave. No soldiers drive residents from their homes.
Factories continue producing goods. Ports continue loading and unloading cargo. Logistics networks continue transporting merchandise. Financial and information services continue operating.
But an increasing proportion of the work is performed by artificial intelligence and robots. Businesses no longer need to employ as many people as they once did.
For those who came to Shenzhen primarily in search of employment, the economic reasons for remaining there may gradually weaken.
Some leave. Others simply stop coming.
The population decline need not be sudden, nor must it take the form of a dramatic social crisis. It might unfold slowly over several decades.
This is what I mean by a Second Exodus.
It is not a definite prediction about the future. It is a possibility worth investigating, arising from the relationship between technological progress, industrial structure, and population movement.
There is, however, a fundamental difference between the two exoduses.
After the first Great Clearance, people departed, and production largely came to a halt. The government wanted them back because the economy could not recover without them.
In the hypothetical Second Exodus, by contrast, production might continue even as the population declined.
Factories and ports might no longer depend on large workforces as they once did. The urban economy might not decline at all.
The problem created by the first Great Clearance was that when the people were gone, production could not continue.
The possibility raised by the Second Exodus is that production could continue as usual, even though far fewer people were needed.
VII. A City Still at Work
According to figures released by the Shenzhen municipal government, the city’s gross domestic product reached approximately 3.87 trillion yuan in 2025.
Manufacturing, finance, information technology services, and other modern service industries together constitute the foundation of its economy.
Today, these activities still depend on large amounts of human labor. But if automation were to advance dramatically in the future, the relationship between Shenzhen’s economic output and its population might change.
Table 3. Comparing Two Forms of Population Exodus
| Point of Comparison | The First Great Clearance, Early Qing | The Hypothetical Second Exodus |
|---|---|---|
| Principal cause | Government-ordered coastal evacuation | Automation changes the demand for labor |
| How people leave | Administrative orders and forced displacement | Potentially through voluntary migration and reduced arrivals |
| Productive activity | Largely interrupted or destroyed | Might continue or even expand |
| Relationship between population and production | Population loss reduces productive capacity | Productive capacity may become less dependent on population size |
| Problem facing local government | How to restore population and production | Whether the previous population scale remains necessary |
| Economic outlook | Severe decline requiring reconstruction | Continued growth is possible, though other outcomes are also possible |
Note: The first column summarizes the historical circumstances of the Great Clearance and the reopening of the coast. The second presents a conditional hypothesis, not an event that has already occurred or a quantitative forecast of future population change.
We are accustomed to imagining a prosperous city as a densely populated place.
Office towers are filled with employees. Streets are crowded with vehicles. Factories are full of workers. Shopping malls and restaurants are filled with customers.
People are both participants in economic activity and the most immediately visible signs of urban prosperity.
But a dense population is not a logical prerequisite for increasing economic output.
If production technologies and organizational arrangements change sufficiently, the same amount of economic activity—or even more—could be carried out by fewer people.
This does not, of course, mean that Shenzhen must inevitably become a sparsely populated city.
Even if people no longer need to come to Shenzhen for work, they may still choose to live there for its education, healthcare, family connections, cultural life, or lifestyle.
Society might also develop new institutional arrangements that allow people to obtain the necessities of life without depending entirely on conventional employment.
Such developments might allow the population to keep growing, or to stabilize at a particular level.
We cannot presently determine what will happen.
But if technological progress does gradually separate economic production from human labor, we must at least acknowledge a possibility that was once difficult to imagine:
A city could become increasingly successful economically without continuing to attract large numbers of workers.
VIII. The City Without People
Shenzhen is a city that rose through migration.
Factories needed workers. Construction sites needed laborers. Shops and restaurants needed employees. Software companies needed engineers.
People came to Shenzhen because there were jobs. And as more people arrived, they generated still more economic activity.
Population growth and urban development reinforced one another, producing the most distinctive feature of Shenzhen’s history over the past several decades.
What if that relationship were someday reversed?
We do not know how far artificial intelligence and robotics will ultimately advance. We do not know how many new jobs technological progress will create. Still less do we know how people in the future will organize their lives.
There is therefore no sufficient reason to assert that Shenzhen will inevitably experience a massive population exodus.
But neither is there sufficient reason to assume that a city must forever depend on increasing quantities of human labor to sustain economic growth.
If machines eventually become capable of performing the overwhelming majority of productive activities; if newly created jobs prove insufficient to maintain the existing demand for labor; and if the economic reasons for moving to Shenzhen gradually weaken, then population decline and economic growth might occur simultaneously.
Each of these conditions requires independent examination.
They may not all be realized.
Yet there is no logical contradiction between them.
After the Great Clearance in early Qing Xin’an, the local government had to bring people back. Without enough people, the fields, salt works, and marketplaces could not recover.
Over the following three centuries, this land witnessed the return of former inhabitants, the arrival of new settlers, the expansion of villages, and the construction of a modern metropolis.
Eventually, it became the Shenzhen we know today.
We generally assume that Shenzhen’s future development will continue along the same path: more industries, more wealth, more people, and an ever-larger city.
But perhaps that is merely the pattern of a particular historical period.
After the first Great Clearance, Xin’an needed its people to return before production could recover.
If a Second Exodus ever occurs, Shenzhen may no longer face the same problem.
The people are gone. The buildings remain.
The port is still there. So are the factories. The machines are still working.
The city continues to prosper.
There just aren’t many people left.
Leave a Reply