China Never Goes Dark

I had the window seat on the flight out of Shanghai. We took off from Pudong in the early evening, banked north, and for the next two hours I could not stop looking down. Anyone who flies regularly knows the usual rhythm of a night departure from a large city. The grid blazes beneath you for a few minutes, the suburbs thin into scattered clusters, and then you are over darkness, the occasional town passing like a small ship at sea. That is what cities normally look like from altitude. A bright centre, a fading edge, and then nothing.

This was different. From Shanghai the light simply did not stop. We crossed the Yangtze, passed over what I assume was Nanjing and the long chain of cities that line the river, then the Shandong coast, then the approaches to Tianjin and Beijing, and the carpet beneath the aircraft stayed continuous the entire way. There was no edge. Even when we deviated west toward Europe after Beijing, the ground only dimmed by degrees rather than going dark. I have flown over the eastern United States at night and over the Rhine valley, and neither holds a candle to what runs along the Chinese coast. The closest analogue I can think of is the view of the Boston to Washington corridor, except that the Chinese version is several times longer, several times wider, and was almost entirely built inside a single human lifetime.

That observation is the reason for this piece. The questions it raised are simple to ask and surprisingly involved to answer. First, how thick is this urban network really, and how does it compare to the places I know best, Europe and the United States. Second, when did it happen, because the landscape I was looking at did not exist in any meaningful form forty years ago. And third, what does living inside a country this densely and recently urbanised actually do to the economy, to business, and to the way people behave. What follows is my attempt to work through all three, drawing on official statistics, satellite research, planning documents, and the kind of detail that only shows up when you read past the headlines.

Part I: How thick is the network, and how does it compare?

The short answer is that the carpet of light I saw was not an illusion of altitude. It is one of the most concentrated build outs of urban fabric anywhere on the planet, and the numbers behind it are genuinely difficult to hold in your head.

Start with the count of cities. By the most widely cited tallies, China has somewhere between 113 and 130 cities with more than one million residents, depending on whether you measure by administrative boundary or by built up urban area. To put that in perspective, the United States has roughly ten cities above that threshold by city proper, and the European Union has around thirty six. China now has more million plus cities than the United States, the European Union and South America combined. The country is also home to eighteen megacities, defined as urban areas above ten million people: Chongqing, Shanghai, Beijing, Chengdu, Guangzhou, Shenzhen, Wuhan, Tianjin, Xi’an, Suzhou, Zhengzhou, Hangzhou, Shijiazhuang, Linyi, Dongguan, Qingdao, Changsha and Hefei. Four of those carry permanent populations above twenty million. Chongqing’s municipality, which includes a large rural hinterland, sits above thirty million.

The raw count understates the real story, because the defining feature of the Chinese urban system is not the number of cities but the way they have fused into continuous regions. The light I saw from the plane corresponds almost exactly to three super clusters that researchers studying night time satellite imagery have identified as the brightest, most contiguous zones in the country. They are the Beijing, Tianjin and Hebei cluster, often abbreviated Jingjinji; the Yangtze River Delta around Shanghai; and the Pearl River Delta in the south around Guangzhou and Shenzhen. Studies of light pollution using the long running DMSP satellite series describe a continuous illuminated stretch running from Shanghai to Nanjing along the Yangtze, which is precisely the section I flew over first.

These are not metaphors for size. Each cluster is comparable to a major national economy. The Yangtze River Delta alone holds more than 240 million people, roughly half the population of the entire European Union, on about four percent of China’s land area, and it produces close to a quarter of the country’s gross domestic product. More than forty percent of Chinese exports pass through it, and it attracts close to thirty percent of realised foreign direct investment into the country. Around two thirds of German companies operating in China are located there. Taken together, the three world class clusters account for something on the order of forty percent of national output. Planners explicitly designed them to compete with the Tokyo Bay area, the New York tri state region and the San Francisco Bay area, and at the level of economic mass they now do.

What turns these clusters from a collection of large cities into a single connected machine is the rail. This is the part that European and American readers consistently underestimate, because nothing in their own experience prepares them for it. China’s high speed rail network passed fifty thousand kilometres of operating line at the end of 2025, when the Xi’an to Yan’an link opened. That single network is longer than every other country’s high speed rail combined, accounting for roughly two thirds of all such track in commercial service on Earth. It grew from about thirty eight thousand kilometres in 2020 to over fifty thousand in 2025, a third in five years. It now reaches thirty one of the country’s thirty four provincial level divisions and serves ninety seven percent of cities with an urban population above five hundred thousand.

The operational reality this produces is the thing that matters. Within a five hundred kilometre radius, the major clusters function as one to two hour travel circles. Trips of around a thousand kilometres between regions are routinely done in four hours, and journeys of about two thousand kilometres increasingly allow a same day round trip. In 2025 the national rail system carried 4.255 billion passenger trips, and at peak the high speed network alone can move up to sixteen million people in a single day, which is to say it can move the population of an entire megacity through its turnstiles between sunrise and the following sunrise. The state plans to reach sixty thousand kilometres of high speed line by 2030, organised as a grid of eight vertical and eight horizontal corridors.

So when I describe the network as thick, I mean it in a specific and measurable sense. It is dense in cities, dense in people per city, fused into continental scale regions, and stitched together by a rail system whose scale has no comparison anywhere else. Europe has historic cities of great beauty connected by good but fragmented rail across many borders. The United States has enormous metropolitan sprawl connected almost entirely by road and air. China has something neither of them has: a continuous, planned, high capacity urban lattice along its eastern third, which is exactly what was glowing under the aircraft.

Part II: When did the landscape actually shift?

The honest answer is that almost all of it happened inside the last forty five years, and the bulk of the visible skyline inside the last twenty five. This is the fact that European observers find hardest to internalise, because our own cities accreted over centuries. China compressed the equivalent process into roughly two generations.

In 1980 China was an overwhelmingly rural country. Only about nineteen percent of the population lived in cities, a lower share than much of the developing world at the time. The urban population was around 191 million. The rate then climbed steadily, crossed the symbolic fifty percent line in 2011, the first year in the country’s history that city dwellers outnumbered rural residents, and reached 67.9 percent at the end of 2025, with roughly 940 million people living in urban areas. Somewhere between 600 and 700 million people moved from the countryside into cities across those four and a half decades. There is no migration of comparable scale anywhere in the historical record.

The mechanism was a sequence of deliberate policy decisions, each of which you can still read in the skyline. The starting gun was Deng Xiaoping’s decision in 1980 to designate four special economic zones, the most famous of which was Shenzhen, then a cluster of fishing and farming settlements of around thirty thousand people next to Hong Kong. The zone offered cheap land, tariff reductions and simplified rules for foreign investors at a time when the rest of the country remained closed. The result is the single clearest illustration of the whole phenomenon. Shenzhen’s population today is around 17.8 million. Its gross domestic product rose from 2.7 billion yuan in 1980 to roughly 3.46 trillion yuan in 2023. It is now home to Huawei, Tencent and BYD, and is routinely described as China’s answer to Silicon Valley. A person born there in 1980 has watched a fishing town become one of the largest and most technologically advanced cities on the planet within their own working life.

A second decision broadened the criteria for what counted as a town or city in 1984, which by itself roughly doubled the recorded urban population almost overnight as existing settlements were reclassified. Then in 1990 came the decision that produced the most photographed skyline in Asia. The land east of the Huangpu River in Shanghai, an area called Pudong, was farmland, fish ponds and warehouses. The old Shanghai saying held that a person would rather have a bed in Puxi, the historic west bank, than a room in Pudong. On 18 April 1990 the central government designated Pudong for development as a testing ground for reform. The Lujiazui financial zone that rose there, almost the same footprint as Lower Manhattan, now holds the Oriental Pearl Tower, the Jin Mao Tower, the Shanghai World Financial Center and the 632 metre Shanghai Tower, the world’s first cluster of three adjacent supertall skyscrapers. Pudong grew at an average of roughly fifteen percent a year between 1990 and 2017 and now produces about a third of Shanghai’s output. The farmland I would have flown over in 1989 was, by the time I made this trip, the brightest single node on the coast.

Two further decisions turned this from a coastal experiment into a nationwide construction machine. Deng’s 1992 southern tour publicly endorsed the market direction and removed political doubt, and the 1994 tax sharing reform restructured government finance in a way whose consequences still dominate the property story today. That reform centralised tax revenue in Beijing while leaving local governments responsible for most spending. To fund themselves, local governments turned to the one asset they controlled, which was land, selling long leases to developers and using the proceeds to build. This “land finance” model is the hidden engine behind the speed of the build out, and it is also the source of the imbalances I will come to shortly. China’s accession to the World Trade Organization in 2001 then poured export demand into the coastal cities, and the 2000s became the decade of relentless vertical construction.

The scale of the physical effort is best captured by a single statistic that the historian Vaclav Smil popularised and Bill Gates repeated. China used more cement in the three years from 2011 to 2013, roughly 6.6 gigatonnes, than the United States used in the entire twentieth century, roughly 4.4 gigatonnes. More recent figures show China still consuming as much cement every two years as the United States did across the whole of the last century, and producing well over half the world’s total. More than twenty million people were relocating into cities every year at the peak, a figure larger than the combined downtown populations of New York, Los Angeles and Chicago, year after year. The most recent stage was formalised in the 2014 National New Type Urbanization Plan, which shifted the official emphasis from building ever more concrete toward integrating the people already living in cities, a change of direction I will return to in the social section.

The takeaway is that the landscape I saw is young. The Bund opposite Lujiazui is roughly a century old. Almost everything else glowing under that aircraft was poured after 1990.

Part III: How does the density reshape the economy?

A country this urban, this fast, and this connected does not simply have a larger version of a Western economy. The density itself becomes an input, and it shapes the structure of business in ways that are visible once you know to look for them.

The first effect is agglomeration, the well documented tendency for productivity to rise when firms, workers and suppliers cluster tightly together. China has engineered agglomeration at a scale no one else has attempted. Pudong is the cleanest example. As companies, research centres and infrastructure concentrated there, the advantages compounded: access to a deep talent pool, integrated supply chains, proximity to capital and to government. Around eight hundred integrated circuit firms set up in Pudong alone, employing close to two hundred thousand people, and more than three hundred regional headquarters of multinationals and over a thousand financial institutions now sit within it. The economic logic of the cluster is that each additional firm makes the next one more likely, which is why these regions pull such disproportionate weight relative to their tiny share of national land.

The second effect is that the density made possible a consumer and logistics economy that leapfrogged the Western model entirely. When several hundred million people live in dense vertical neighbourhoods served by good roads and rail, the unit cost of delivering anything to them falls dramatically. This is the structural reason China developed the most advanced online retail and delivery ecosystem in the world rather than importing the Western template of suburban big box stores and car based shopping. The same density that crowds people into towers also puts a courier within a few minutes of millions of doors. Layer onto that the near universal adoption of mobile payment, where more than eighty seven percent of internet users were paying by phone as early as 2021 through Alipay and WeChat Pay, and you get a commercial fabric where ordering, paying and delivery are frictionless at a scale that simply is not replicable in a low density country. The cashless transition is so complete that visitors often struggle to spend physical cash at all, and China became the first major economy to issue a sovereign digital currency in 2021.

The third effect, and the one investors most need to understand, is that the same model carries a structural fragility. The land finance system that funded the build out depended on property prices rising forever and on local governments selling ever more land. For two decades it worked, and property swelled to account for somewhere between a quarter and thirty percent of gross domestic product, with roughly seventy percent of household wealth tied up in homes. Then it broke. In 2020 Beijing moved to deflate the bubble by restricting developer borrowing, and the largest developers, carrying enormous short term debt, began to fail. Evergrande defaulted on more than three hundred billion dollars of liabilities and was ordered into liquidation, leaving around eight hundred thousand presold homes unfinished. Country Garden, the largest developer by volume, followed into distress. By late 2023, developers accounting for some forty percent of national home sales had defaulted.

The visible residue of all this is the phenomenon of the “ghost cities” and the “rotten tail buildings”, the lanwei lou, the half finished towers that stand abandoned across the country. Estimates of vacant housing run as high as sixty five million units, roughly a fifth of the urban housing stock, and a former deputy head of the statistics bureau remarked that the most extreme estimates put the number of empty homes at enough for three billion people. Beijing announced a rescue package of around forty two billion dollars in 2024 to buy unsold homes and convert them to affordable housing, but analysts at Goldman Sachs have estimated that simply finishing what has already been started would cost on the order of a trillion dollars. For an investor, the property sector is the clearest case where the speed and scale that built the lattice also built the bill that is now coming due.

So the economic picture is genuinely two sided. The density delivers real, durable advantages in productivity, logistics and consumption that the country will keep. It also produced an overbuilding cycle whose unwinding is the dominant macro story in China today.

Part IV: How does it reshape daily life and social behaviour?

This is the part that the statistics describe least well and that matters most, because the way several hundred million people came to live in cities was governed by an institution that has no Western equivalent, and that institution shaped the texture of daily life as much as any skyline did.

That institution is the hukou, the household registration system introduced in its modern form in 1958. Every citizen holds a registration tied to a place and classified as either rural or urban, and that registration historically determined where a person could access public services: schooling for their children, healthcare reimbursement, pensions, housing support. The crucial feature is that for most of the reform era, a rural registered person could move to a city to work but could not easily convert their status, and their entitlements did not travel with them. This produced what scholars call “semi urbanization.” Around three hundred million rural migrant workers, sometimes counted as high as 376 million in the 2020 census, live and work in cities while remaining formally entitled to services only in villages hundreds of miles away. They built the towers and staffed the factories, and they did so while paying more for medical care than their neighbours and often unable to enrol their children in local schools.

The social consequences are profound and largely hidden from the view I had at altitude. Tens of millions of children were left behind in rural villages in the care of grandparents while both parents worked in distant cities, a generation of “split families” that has no real parallel elsewhere. The wage gap between rural and urban registered workers persists even after accounting for education and experience, and rural migrants remain concentrated in informal and lower paid work with weaker protection.

And yet the same system produced an outcome that is genuinely remarkable and that any honest observer has to acknowledge. Walk through Shanghai or Beijing, each home to more than twenty million people, and the vast informal slums that ring comparably sized cities across the developing world, the settlements of Mumbai, Lagos, São Paulo, Dhaka or Cairo, are conspicuously absent. China urbanised six hundred million people without producing the megaslums that have accompanied migration on this scale almost everywhere else. The hukou system was central to that, because by regulating the pace and direction of migration it prevented a chaotic flood into a few coastal cities whose housing and services could not have coped. A second buffer mattered just as much. Because rural land was never privatised and remains collectively owned with use rights allocated to households, migrants kept their entitlement to land and housing in their home villages. When the 2008 financial crisis cost roughly thirty million migrants their jobs almost overnight, they were not cast into urban destitution. They went home to land they still held. That structural safety valve is invisible until a shock arrives, and then it is decisive.

The state is now actively dismantling the old barriers, because the semi urban model that suited the export decades now suppresses exactly the thing the economy needs most, which is domestic consumption. A migrant uncertain of services and clinging to precautionary savings does not spend like a settled urban citizen. The 2014 plan began the shift, the July 2024 third plenum committed to deeper reform, and in May 2025 the State Council announced that migrant workers could enrol in social insurance in the cities where they work regardless of where they are registered. Cities under three million people have had hukou restrictions effectively removed, and the megacities have been pushed to lower their thresholds. Whether this succeeds depends on whether local governments, already fiscally strained by the property bust, can afford to serve the populations they would be absorbing. It is one of the most consequential domestic policy experiments underway anywhere.

Beneath the institutional layer sits a cultural one that has surfaced loudly in the past few years and that tells you how the pressure of this environment feels from the inside. The dominant words in Chinese online culture are neijuan, usually translated as “involution,” and tang ping, “lying flat.” Involution describes the experience of a hyper competitive society where everyone works harder and harder simply to stay in place, with the rewards shrinking even as the effort climbs: parents pouring money into tutoring, students sacrificing everything for the gaokao entrance exam, white collar workers staying late under the “996” schedule of nine to nine, six days a week, to signal loyalty. Lying flat is the counter movement, popularised in 2021 by a former factory worker named Luo Huazhong who posted that he had quit, cut his expenses to the minimum, and refused to keep running. The post went viral because it named something an entire generation felt. A related term, runxue or “run philosophy,” captures the impulse to emigrate. These are not fringe phenomena. They are the vocabulary an urbanised, educated, pressured population uses to describe the cost of living inside the machine that the lights represent.

So the social reality is layered. A vast migrant workforce that built the cities while held at arm’s length from them, an institutional system that avoided the slums of the developing world at the cost of deep internal inequality, a state now racing to convert that workforce into consumers, and a younger urban generation increasingly questioning whether the relentless competition is worth it. None of that is visible from a plane. All of it is what the plane was flying over.

What all of this tells us

I started this piece because of a window seat and an unbroken field of light, and the work of explaining that light leads to a fairly clear set of conclusions.

On the question of thickness and comparison, the answer is that the contrast with Europe and the United States is real and structural, not a trick of perception. China has built, in four and a half decades, the densest and most connected urban lattice in the world, fused into continental scale regions and stitched together by a rail network larger than every other country’s combined. Nothing in the Western experience is equivalent, and any analysis that models China as a bigger France or a more crowded America starts from the wrong picture.

On the question of timing, the lesson for anyone trying to understand the country is that this is a young landscape carrying young debts. The skyline is almost entirely post 1990, and so is the financing model that built it. That youth is why the property unwinding is so severe. The system never went through the slow seasoning that aged Western cities did, and the bill for compressing a century of construction into forty years is being presented now, in the form of the developer defaults, the unfinished towers and the strained local government balance sheets.

On the question of what density does, the picture is genuinely two sided and an honest assessment has to hold both halves at once. The advantages are durable. A dense, connected, cashless, logistically frictionless consumer market of several hundred million people is an asset that does not evaporate, and it is the foundation on which the country’s strongest companies in online retail, payments, electric vehicles and consumer technology were built. The fragilities are equally real: a property sector that grew to a quarter of the economy and then broke, a migrant workforce only now being granted full citizenship of the cities it built, a demographic decline that means many of those towers may never be filled, and a younger generation visibly tiring of the competition. The same density is the source of both the strength and the strain.

For an investor watching from Europe, the practical point is that the urban lattice is not the story by itself. It is the stage on which the real stories play out. The companies that learned to operate inside this density, the logistics networks, the payment platforms, the consumer brands that reach a courier distance from a billion doors, carry advantages that are hard to replicate in any lower density market. The companies and local governments that bet on the property cycle continuing forever are carrying the opposite. Reading China well over the next decade will mean separating the businesses whose models depend on the durable feature, the density and connectivity, from those whose models depended on the temporary one, the permanent rise in land prices. The lights will stay on. What they are paying for is the open question.

Disclaimer: This publication is for informational and research purposes only. It does not constitute investment advice. Past performance is not indicative of future results.