China’s flooding of global markets with electric vehicles has created a cascade so severe that The Wall Street Journal reports there are not enough ships to get the 10- 12 million cars it wants to export out of the country.
We have all heard praise for the far-sighted industrial policy that made this happen. Beijing technocrats sagely targeted batteries and EVs, spun up state-owned car companies, showered them with subsidies, dictated the terms of joint ventures with foreign carmakers, and executed a twenty-year plan with a discipline and patience that democracies cannot muster. I have hinted at this myself.
New Research Tells a Different Story
A new paper in The China Journal tears this story apart. Fengming Lu spent three years interviewing the officials, engineers, and entrepreneurs who built China’s EV industry. He and his co-author were trying to solve a puzzle. They knew that in 2014, China’s ten largest automakers were all joint ventures with foreign firms. All but one of the Chinese companies were massive state-owned enterprises (SOEs), not private firms.
The research question was why the story had completely changed ten years later. Take a look at the 2024 table.
The private firms like BYD, Chery, Geely, and Great Wall that took over the market did so despite provincial governments investing heavily in their state-owned competitors. Together with Beijing, they awarded nearly every early EV research grant to state firms. The flagship pilot programs were built around them. But the companies that won were the ones that Beijing had initially ignored.
The researchers wanted to figure out why. They concluded that the EV boom grew not from central industrial planning, but from three intersecting forces. Central rules that favored a few large incumbents had actually locked most Chinese cities out of the car business. As a result, dozens of local governments that were desperate for tax revenue wanted in. They teamed up with private investors to create the companies that actually won. Except, of course, that winning means flooding overseas markets, which imposes massive costs.
“Go To The Mountains”
As is so often true in China, the story begins with Mao Zedong. Not the wily revolutionary, who spent his middle years fighting the Japanese from the mountains of Shaanxi province in northwestern China, but the aging and delusional leader trying to govern the world’s most populous country.
Mao was a strong guerrilla leader, and remained one even when trying to govern. His 1958 “Great Leap Forward” was a campaign to rapidly transform China into a rich, modern industrial society. Mao forced peasants into communes and required them to build primitive metal furnaces to make steel. He urged the wholesale slaughter of sparrows, which he thought ate too much grain. This led to massive swarms of insects that ruined crops and left people starving. The resulting chaos and famine killed between 15 and 45 million people – one of the worst disasters in human history.
By 1964, seeing that the Great Leap had taken China backward and not yet inspired to launch the mass psychosis known later as the Cultural Revolution, Mao fixated on the risk of nuclear war. He declared that China needed a Third Front – a defense area safe from Russian or American atomic weapons. China must “prepare to go to the mountains”, meaning relocate entire industries into the wilds of western China.
China gave it a try. They moved nearly 4 million workers and family members and diverted 40% of all capital construction to the construction of more than 1,100 large- and medium-sized industrial and mining enterprises. They built more than 8,000 kilometers of railroad track in the mountains of western China, which Mao deemed safe from atomic bombs.
The result was a massively fragmented auto and steel industry. Instead of expanding existing coastal steel mills, Mao ordered the construction of an enormous steel complex in Panzhihua, a remote canyon in Sichuan province, solely because it was buried deep in the mountains and hard to bomb.
The “Second Auto Works” (now Dongfeng Motor) was built in the remote Shiyan mountains. The factory’s assembly lines were deliberately scattered across dozens of different valleys so that if one workshop was bombed, the others could keep producing trucks for the army.
The Third Front (and, to a lesser extent, the Great Leap Forward) created a large population of geographically stranded, vertically integrated, locally protected producers that were individually too small to be efficient and too politically costly to kill. Eventually, the surviving companies moved to larger population centers: Dongfeng to Wuhan, Shaanqi to Xi’an, Hongyan to Chongqing, all abandoning the mountain sites the Third Front strategy had put them in. But the seeds were planted: even the smallest and most remote provinces in China were sure they could build a car company.
Provincial and Municipal Ventures
Wuhu is a middling city in Anhui, a middling province. It wanted to get in on car production, but in the 1990s, central policy reserved passenger car production for six anointed state enterprises and their foreign partners. Until the 2010s, Chinese law protected these incumbents by requiring new carmakers to invest at least 1.5 billion yuan. This kept out the riff-raff.
Wuhu began by convincing First Auto Works (FAW) to open a chassis plant there. When the city sold its cement works in 1996, it used the proceeds to buy a used engine line from Ford of Britain, poached FAW’s engineers, and started its own carmaker. It named the firm Chery and appointed the deputy mayor as chairman. Although the carmaker received grants from Beijing and remained a regional SOE through the 2000s, it eventually raised private capital and became China's second-largest automaker. Anhui provincial and Wuhu municipal governments still own a quarter or so of Chery.
This pattern – a Chinese city that could not get state champions to return their phone calls deciding to back local entrepreneurs to start a new car company happened repeatedly over the next thirty years. Li Shufu struggled to build a company that made refrigerator parts, then decorative building materials, then motorcycles and scooters. By 2002, the company, now called Geely, began assembling sedans. In 2010, they bought Volvo Cars from Ford. When most state banks refused to finance the deal, Geely convinced Goldman Sachs and four city governments to invest. Each city got a Volvo plant or a headquarters building for its trouble.
Rules loosened in 2015, spawning hundreds of startups, many of them EVs. In 2019, only 12 firms actually held government licenses to build EVs. But there was a loophole designed to protect incumbents: a company authorized to build gas cars could convert to EVs. The rule was a golden ticket to every failing regional automaker and led to more than 100 new EV companies.
These companies survive in part because city or provincial governments will rescue struggling carmakers. The city of Hefei invested 7 billion yuan to rescue NIO at the bottom of the market. When its investment doubled within months, Chinese media proclaimed it the Hefei Model, and soon every mayor in China wanted to imitate them.
As large state enterprises began losing money, they also lost support from their hometowns. Beijing’s BAIC was China’s top EV seller from 2013 to 2019 and outsold Tesla globally in 2017. It built the electric taxis that enabled Beijing to hit its air-quality targets, but in doing so they abandoned hybrids and made fewer cars that private buyers wanted. Instead of backing them, Beijing shifted its bets to Li Auto and Xiaomi. Likewise, Guangdong funneled billions into XPeng because local regulators doubted that hometown favorite GAC would successfully manage the transition to EVs. In short, Chinese carmakers competed not only for customers but also for municipal and private investors.
Lessons
The researchers highlight three important lessons from Chinese carmaking.
Industrial policies benefit from competition. For the industrial policy debate now consuming Washington, it suggests the power of competing investments. China’s real advantage was dozens of governments competing to build winners with their own money and their own consequences. Its advantage was not an enlightened center, which in the end was too rigid to stop them.
State-owned enterprises rarely work well long-term. The pattern of private-sector collaboration with local governments to outperform more established SOEs is evident in many other sectors. Pharmaceuticals, steel manufacturing, and frontier AI models all fit this pattern.
Nobody wants the hometown company to fail. The approach carries the risk of massive wasted capacity, price wars, and municipal debt. China currently has 130–150 active automobile companies competing in its hyper-competitive domestic market. The paper mentions that Nantong, a city in Jiangsu, sank 6.6 billion yuan into EV sports car maker Sailin and built few, if any, cars.
Even though dozens of cities are determined not to let their company fail, a massive industry shakeout is underway because Chinese customers have stopped buying. Domestic car sales are down a stunning 20% over last year. Most analysts predict that by 2030, only 5-10 car companies will be profitable. Last year, Beijing warned local governments to stop launching EV projects. Next year’s message is likely to be even tougher.
ICYMI
Reuters: The US has used up its precision long-range missiles.
NYT: Royals Harry and Meghan are moving back to England. Aww.
Wired: Coders already have a workaround for Claude’s watermarks.
MIT: Economist Paul Osterman finds 35% of US workers are marginal.
WSJ: Christianity continues to surge in China.
NYT: 22 countries host 53 Level 4 biosafety labs, many in dense cities.
FT: Clean energy is booming despite Donald Trump’s efforts to kill it.
Private investors are useless at longer term return projects.... only the state or state guarantees can provide infrastucture.
In this case... the state did get the ball rolling and private piggybacked as it always does... and then tries to say it did it all on its own.. as it always does. See also ... musk.
Great work, thanks. This shakeout happened in computers and even the early car industry. Castle builders too, for all we know.