Which industries will China dominate?

A few key principles.

Photo by Mohammad Fathollahi on Unsplash

Nothing like the current era of international industrial competition has ever been seen before. Nothing like modern distributed supply chains even existed until the last few decades. The competition between the U.S., Japan, and Europe over cars and semiconductors in the 80s and 90s is probably the closest parallel, but the differences are vast — this time, the competitors are not allies, but rivals, on opposite sides of a new cold war. And the impacts of new technologies — especially green energy and AI — is highly uncertain. The factors at play include:

• Export controls

• Government-directed reshoring efforts and other industrial policies

• Massive R&D into emerging technologies

• Rising Chinese costs and other pre-existing economic trends

So the question of which industries China will dominate, which will go to the developed democracies in the U.S., Europe, and East Asia, and which will be offshored to developing countries like India and Vietnam, is a very hard one to answer. But it’s also a very important question, whether you’re an investor, a businessperson, or a government policymaker trying to affect the outcome. The flip side of “Which industries will China dominate?” is “Which industries will the U.S. and its allies fail to dominate?”.

In principle, you might be able to tackle this question with a very complex econ model and a whole lot of data — or perhaps with machine learning. I’ll certainly keep an eye out for efforts in that direction.

But in the meantime, I think a few basic contours and principles are starting to make themselves clear. China’s two key advantages are disruption and scale; this will allow it to dominate many new industries where existing stores of knowledge aren’t a big advantage, and older technologies where producing large amounts cheaply is key.

China will dominate where it can leverage disruption

The idea of technological disruption was laid out in Clay Christensen’s famous book The Innovator’s Dilemma. Basically, the idea is that companies with big lucrative existing markets tend not to allocate resources toward new technologies that can initially only capture small niche markets. For example, Microsoft focused on its desktop products instead of the nascent search market, which was then captured by Google. Google, in turn, focused on solidifying and milking its lucrative position in search ads, and largely ignored smaller markets in AI. The most infamous recent case of the innovator’s dilemma was probably Intel, which focused on making ever-faster CPUs for cloud servers, and thus ignored new opportunities in low-power chips for phones, GPUs for AI, and foundry manufacturing.

If it can happen for companies, it can happen for countries as well. If you read the book Can Japan Compete?, you’ll learn how in the 1970s, Japanese companies took leadership in the machine tool industry, displacing their older American rivals. The American companies were slow to move into the new market for cheap CNC (computer-controlled) machine tools. Japanese and German companies, having no incumbent position in older types of tools that they could milk for cash, poured their resources into the newer technology. Initially, CNC tools were cheaper and only suitable for lower-value tasks, but eventually they got much better and displaced the older servo-controlled tools. When that happened, Japanese and German companies won, and American machine-tool dominance passed into history.

It’s a good bet that China and its companies will be able to pull off a similar feat in many industries. The most striking example is in auto manufacturing. For decades, China strove mightily to develop its own car companies that could compete with multinationals like Toyota and Volkswagen, but failed again and again. But in the initially small market for electric vehicles, Chinese companies like BYD and SAIC were able to gain a foothold. With no lucrative internal combustion vehicle market to protect, they were free to avoid playing catch-up, and instead focused their resources on something novel.

This is one reason China has suddenly become a leading auto exporter. While some of this is about selling lower-end gas-powered cars to Asia and Latin America, China’s sudden surge of auto exports to Europe is largely about its dominance in EVs. Bloomberg’s Tom Hancock reports:

In Europe, the China-made vehicles sold are mostly electric models from Tesla Inc. and Chinese-owned former European brands such as Volvo and MG, and European brands like Dacia Spring or the BMW iX3, which is produced exclusively in China. A raft of homegrown marques like BYD Co. and Nio Inc. are ascending as well, with ambitions to dominate the world of new-energy vehicles…

China has been leading the charge on…EVs. Local carmakers have found the electric platform relatively easy to master compared with the complex internal combustion engine.

“The switch to battery means the motor is no longer a differentiator,” says Alexander Klose, executive vice president for overseas operations at Aiways Automobiles Co., a pure-Chinese EV maker, which has sold several thousand vehicles in Europe. Technologically, “it’s created a level playing field,” he says.

You can already see clustering effects at work here too; now that China is the center of EV manufacturing, Tesla and European EV makers are producing their cars there too.

(Of course, it’s worth mentioning that domination of the car industry is never very absolute; cars are very heavy, and local tastes can change quickly, so most cars are made close to where they’re consumed. Europe is really the main anomaly in terms of importing lots of EVs from abroad, and that probably won’t last. So even if China does become the world’s dominant carmaker, Europe, the U.S., and Japan will still make some.)

The disruption dynamic has been evident in electronics, too. MacroPolo has a recent article about how in the 2010s, China leapfrogged producers of the older CCD image sensors and went directly to the new CMOS technology; now that CMOS technology has pushed out CCD and risen to dominate the market for image sensors, China is well-positioned to expand its share of one the key inputs into smartphone cameras and (eventually) autonomous cars.

Drones are another example. Even as the U.S. focused on high-end, high-performance military drones like the Reaper, China’s DJI made cheap little quadcopters for the relatively niche toy drone market. Now, those “toys” are proving increasingly crucial on the battlefields of Ukraine, and DJI could end up disrupting the drone industry.

I would not be surprised to see China replicate this trick in the near future. Any time there’s a cheap new technology in a niche market, expect to see China do well. There’s a flip side of this coin, though. Where China can’t leverage disruption — i.e., in established markets with lots of proprietary technology where what Christensen called “sustaining innovation” is the key to success — China will struggle to break in.

This is the basic dynamic in semiconductors, where production of cutting-edge logic chips requires massive investments in extremely complex, specialized technologies like extreme ultraviolet lithography (EUV). Not only has China repeatedly failed to catch up despite absolutely massive government investment in the sector, but now U.S. export controls are going to make their job even harder. The U.S. has successfully persuaded the Netherlands and Japan to join its export control regime, meaning China will now have to reinvent even more basic fabrication tools, to say nothing of the hideously complex EUV.

It’s also the dynamic in aerospace. China has poured a lot of resources into making its own jet engines, but as Barry van Wyk reported last year, progress has been very incremental and China is not yet near the frontier:

The full installation of China’s own jet engines is not going to be easy or quick. As one foreign observer put it, making jet engines is “more art than science,” and the leading Western manufacturers have “tacit knowledge” that their Chinese competitors are still lacking.

Of course, China will try to steal as much as it can, and import equipment through third parties, much as the USSR did in Cold War 1. And it’s pouring research effort into both semiconductors and jet engines. But it’ll continue to be an uphill battle, as a recent report from China’s Institute for International and Strategic Studies noted:

IISS researchers acknowledged that China faces a critical shortage of talent to make integrated circuits, operating systems and industrial software…U.S. sanctions of Chinese telecommunications firms such as Huawei particularly have hurt China’s information technology sector…The U.S. is “the absolute leader” in space transportation, human spaceflight, satellite navigation and communications and deep space exploration, the Chinese researchers asserted. China, with an independent technology and equipment system, ranks in a second tier along with Europe and Russia.

This is important, because it means that Chinese dominance of an industry isn’t just a push-button thing. One big argument against semiconductor export controls — which even some of my favorite China-watchers made — was that the U.S. was shooting itself in the foot, because now China would catch up and displace American and allied chipmakers much faster than it otherwise would have. In this view, Chinese dominance of an industry is merely a matter of will — China has a ton of money and a ton of smart young engineers, so all it has to do is basically just decide which things it’ll dominate this week.

But in fact, this view of Chinese industrial dominance both overrates and underrates Chinese capability. Domination of mature high-tech fields is about more than just throwing money and brains at the problem — it requires time, because the buildup of tacit knowledge in a highly complex field like chipmaking or aerospace is a slow process even for the smartest people on Earth. And it requires specialization, because each high-tech company gets really really good at making one specific thing; this is a very different situation from lower-tech mass manufacturing, where process engineers can quickly pivot from one type of product to another. China can’t just press the button and create either of those things.

At the same time, the idea of China as a giant resource-allocating machine underrates the inventiveness of China’s engineers and the entrepreneurialism of its founders. Just as hungry young Japanese companies saw an opportunity in the nascent CNC machine tool market and eventually managed to disrupt the dominant players, Chinese engineers and entrepreneurs are seeking out opportunities to leapfrog the big established companies in the developed world. China succeeded at leapfrogging the U.S. in many IT applications — mobile payments, video apps, and arguably even social media itself. But China’s government refused to let its companies use this success to disrupt the global IT industry, cracking down on its internet companies in an attempt to divert talent into fields like semiconductors. Even if that ham-fisted reallocation was successful, it meant diverting resources from industries ripe for disruption to those where China is still playing catch-up. When it comes to disruption, China’s worst enemy is its own government.

Anyway, OK, that was a bit of a rant. Let’s move on to China’s other big advantage: scale.

China’s competitive advantage is scale, not cost

Americans tend to think of China as a place where producing things is very very cheap. We got this idea in the 2000s, when the “China price” was so low that American companies (and European and Asian companies) shipped many of their factories to China. But the Chinese cost advantage has eroded severely in recent years, as wages have risen. Now, Chinese labor costs more per hour than in many parts of Asia. Of course Chinese workers have become much more productive over the years. But unit labor costs — which take rising productivity into account — were about even between the U.S. and China as of 2016, and as of 2019 Mexico became a cheaper place to make things than China.

One thing China does have is a dense network of suppliers. If you’re a manufacturer, locating in China is convenient because every component or material you need will also be made in China, so in theory it’ll be pretty easy to source. For this reason, some analysts scoff at the idea that multinational companies will move production out of China to “Altasia” (i.e., the rest of Asia). It would cost too much, they say, to ship all those components around the region.

But this is also wrong, I think. First of all, it’s a lot cheaper to ship things over water than over land, and the factories of “Altasia” are clustered around ports. Second, during the era when the “China price” was world-beating, China mostly only did final assembly for things like electronics — Japan or the U.S. or Korea or Taiwan would make the high-value components like chips and screens, and then ship them to China, and Chinese workers would slap them together into a phone and then ship it back to wherever the phone was sold. And that was a very cheap, easily managed process. Moving assembly to Vietnam or India, and shipping Chinese-made components there, shouldn’t be any harder than what everyone did in the 2000s. In fact, I expect Chinese companies themselves to offshore more and more labor-intensive assembly work to countries like Vietnam, partly for cost savings and partly to get around tariffs and other trade restrictions.

China’s real advantage is not cost or supplier networks, but scale. If you want to make truly massive numbers of EVs or phones or whatever, then to make them outside of China, you might have to source them from a patchwork of other countries — Malaysia and Mexico and Vietnam and the U.S. and India and Poland, etc. Whereas if you’re OK producing in China, then you just press one button, and bam, all your factories go in China, and you’re done. This is probably why Tesla made a big bet on China production back in the 2010s. That advantage hasn’t gone away at all.

One factor behind Chinese scale is its huge population. But there are other factors — the ability to easily get lots of land on which to build factories, for example, which is exactly the opposite of how things work in the U.S. Another is China’s massive energy resources, including both clean energy and very dirty coal.

Scale is why China dominates a lot of primary industries — things like mineral processing and fertilizer. These are low-margin activities, meaning that you have to do them in absolutely massive amounts to be profitable in a competitive market. China wins these industries because it can do things in absolutely massive amounts. (100 years ago, this was the U.S.’ key strength, but no longer.) Developed countries are going to have a hell of a time competing with China in these industries without massive subsidies, unless they all manage to pool their resources somehow.

But the most important factor behind China’s massive scale is probably talent. China has absolutely massive numbers of well-trained, highly competent engineers. In a recent article for Foreign Affairs, Dan Wang explains that these engineers, rather than subsidies or central planning or low wages, are China’s secret sauce when it comes to manufacturing:

Beijing’s manufacturing-driven approach has become critical to its ability to challenge the West in advanced technology. To understand why, it is crucial to recognize the forces that go into successful innovations. Producing new technology can be likened to preparing an omelet: ingredients, instructions, and a well-equipped kitchen are helpful, but they will not in themselves guarantee a good result…An additional element is required: practical experience—skills that can only be learned by doing. These skills can be referred to as process knowledge, and they are part of what has helped China become a major tech innovator…

The country’s most significant technological achievement over the past two decades has been its development of a vast and highly experienced skilled workforce, which can be adapted as needed for the most tech-intensive industries. For example, Apple still counts on China as the only country that can call up hundreds of thousands of highly trained workers on short notice[.]

That massive skilled workforce provides another kind of scale that goes beyond simple high-volume low-margin production. It allows companies to set up and staff new factories very quickly. It allows companies that want to develop new products, or modify their existing products, to get them to market very quickly. It’s the kind of scale that saves time, not just unit cost.

This sort of scale is what will allow China to start competing in trailing-edge industries very quickly. MacroPolo reports on how China is trying to build a dominant position in the well-established silicon carbide chip industry (which regulates power in devices like EVs):

SiC isn’t a new technology, as the technical bottleneck is largely about scaling—and China has a good track record of scaling the living daylights out of existing technologies…While the benefits of switching to SiC chips are obvious, scaling is the main constraint…These constraints on reaching economies of scale, a result of SiC’s comparatively immature and complex production process, will likely lead to demand and supply mismatches as soon as 2023…However, this projection may need to be taken with a grain of salt because it doesn’t account for new Chinese entrants that are aiming to overcome the scaling bottleneck and narrow this gap…

The main obstacle to scaling SiC chips is making ever larger SiC wafers…While increasing wafer size is a time-consuming and iterative process that requires considerable capital, it is also not rocket science. It is not nearly as difficult as reducing nodes at microscopic levels on the most advanced chips that a company like TSMC produces.

The article goes on to discuss how China’s silicon carbide chipmakers are outsourcing their R&D to Chinese universities instead of trying to do it in-house, and are concentrating their own efforts on practical implementation. This is exactly the kind of mass continuous minor innovation that China’s huge engineering workforce is well-suited to do.

Solar and batteries seem like similar cases. It’s generally believed that the reason solar panels and batteries have come down in cost by such an amazing amount is that they’re subject to learning curves, which are driven both by economies of scale and by learning-by-doing effects. The latter basically means that those armies of Chinese engineers are figuring out a million tiny ways to produce solar panels and batteries more cheaply, day in and day out.

So in general, it seems to me that China is basically poised to dominate three types of industries:

• New technologies that start off in niche markets

• Primary industries and simple manufacturing industries that require massive resource inputs to eke out a profit

• Mid-tech industries where lots of little innovations add up

Interestingly, all of these advantages seem to come together in the EV industry. Making cars and car parts is hard, but not rocket science. It’s a very high-throughput industry where both scale and continuous improvement are important. And because the fundamental technologies we use to make cars are shifting so rapidly, the space seems ripe for disruption.

So I wouldn’t bet against China in the global EV race. But there are plenty of other areas where one or more of China’s key advantages can be brought to bear as well. Really, any manufacturing industry that doesn’t depend on either low labor costs, a high-tech specialized supplier network, continuous R&D, or a vast accumulation of tacit knowledge in a highly specialized technology has a decent chance of falling in one of these categories. China isn’t really the “make-everything country”, but it’s not far off from that, either.

In future posts, I’ll try to pull together some ideas about how the developed democracies can respond to those advantages.

---