Why Japanese companies do so many different things
The internal logic of the world’s strangest corporations
Consider Toto.
If you spend much time in American public bathrooms, or rather if you’re simply a particularly attentive patron of American public bathrooms, you’ll probably have noticed Toto’s toilets at some point or another: they’re distinguished by a quite memorable serif-font “TOTO” logo. Toto toilets aren’t quite dominant in American bathrooms, since they have healthy competition from our homegrown toilet champions American Standard and Kohler—though Toto is doing better and better as Americans start to fall in love with the bidet-toilet—but globally Toto is the world’s largest manufacturer of toilets and bidets. And in its home country of Japan, Toto is simply everywhere: 80 percent of Japanese homes contain a Toto bidet-toilet.
And if you’re a longtime Toto shareholder—maybe an investor with a particular interest in bathroom fixtures—this has been a wonderfully lucrative year for you. Toto’s stock is up 60 percent year to date; in just the last few weeks, it’s risen by 30 percent. Toto is doing better than ever: its net profit, in the first quarter of 2026, was up 230 percent year over year.
But Toto’s remarkable year doesn’t have much to do with toilets or bidets. Toto might have been founded in the 1910s to “provide a healthy and civilized way of life” through affordable toilets, and in the decades since might have become the global leader in the bathroom game. But Toto also does a lot of other things. Toto manufactures not just bidets and toilets but also bathroom tiles, prefabricated bathroom modules, faucets, modular kitchens, photocatalytic coatings for buildings, and assistive equipment for the elderly. And, most importantly, Toto has a very lucrative sideline in the fabrication of memory chips.
Since 1988, in a once-obscure corner of the company called the “advanced ceramics division,” Toto has been producing a very particular component called the electrostatic chuck, or the “e-chuck.” The e-chuck is a sort of high-precision ceramic plate, about the size of a steering wheel, that uses electrostatic force to hold a silicon wafer perfectly flat and thermally stable while memory chips are etched into it with bombardments of plasma. Making these components is extraordinarily difficult, since the ceramic body needs to have near-zero particle generation and be polished to submicron flatness: and this means that there are only a few companies in the world that are capable of manufacturing e-chucks reliably. Almost all of them—Shinko Electric, NGK, Toto, Kyocera, Sumitomo Osaka Cement, Niterra—are based in Japan.
For most of its history, the advanced ceramics division was a rounding error on Toto’s balance sheet: the money maker, as it had been since the 1910s, was the toilet and bidet business. But we’re in a new era. Demand for AI is exploding, meaning that demand for the high-bandwidth memory that AI data centers require is exploding, meaning that demand for memory chips is exploding, meaning that demand for e-chucks is exploding. And so Toto’s advanced ceramics division is suddenly the company’s largest business, generating the majority of its operating profit. Toto’s leadership, suddenly awash in AI-driven revenue, announced that they would double down by investing hundreds of millions in expanded electrostatic chuck production: the toilet company had become, quite unexpectedly, a supplier to the semiconductor supply chain.
Why Japanese companies do so many different things
The internal logic of the world’s strangest corporations
Consider Toto.
If you spend much time in American public bathrooms, or rather if you’re simply a particularly attentive patron of American public bathrooms, you’ll probably have noticed Toto’s toilets at some point or another: they’re distinguished by a quite memorable serif-font “TOTO” logo. Toto toilets aren’t quite dominant in American bathrooms, since they have healthy competition from our homegrown toilet champions American Standard and Kohler—though Toto is doing better and better as Americans start to fall in love with the bidet-toilet—but globally Toto is the world’s largest manufacturer of toilets and bidets. And in its home country of Japan, Toto is simply everywhere: 80 percent of Japanese homes contain a Toto bidet-toilet.
The Toto story is a fun and interesting illustration of corporate diversification and how strange bets can pay off. But that type of diversification—a toilet company that also produces photocatalytic coating and high-precision components for semiconductors—isn’t really unique to Toto. Practically every company in Japan seems to do a thousand very different things.
Consider, for example, Kyocera, another one of the e-chuck makers. Kyocera was founded in 1959 as a producer of ceramic insulators for cathode-ray tubes; today it manufactures not only industrial ceramics but also printers, smartphones, ballpoint pens, kitchen knives, solar PV modules, lens components, industrial cutting tools, automotive camera modules, electronics components, semiconductor packaging, biocompatible tooth and joint replacements, UV-LED curing systems, LCD systems, medical products, and lab-grown gemstones. Or another e-chuck maker. Sumitomo Osaka Cement, as you might have been able to deduce from the name, produces cement and ready-mixed concrete; but it also produces optical components, measuring instruments, industrial ceramics, artificial marine reefs, cosmetics and nanoparticle materials.
And this degree of diversification extends to many of Japan’s most famous companies. Yamaha, for example, manufactures pianos, motorcycles, guitars, drums, boats, snowmobiles, ATVs, audio equipment, golf clubs, tennis rackets, home appliances, specialty metals, molding and bonding equipment for semiconductors, and industrial robots. Hitachi makes nuclear reactors, power grids, railway systems, elevators, semiconductor manufacturing equipment, medical imaging devices, data storage, IT consulting, and industrial machinery. Ajinomoto makes MSG, frozen gyoza, instant coffee, amino acids for animal feed, IV drips, oligonucleotide drugs, and the insulating film layered inside high-end CPUs. Even a company as simple as Oji, Japan’s largest paper mill, has been drawn into the production of disposable diapers, functional films, adhesives, cellulose nanofibers, and wood-based EUV photoresists; and it also operates a hotel, an airport catering business, a concert hall, and an insurance agency.
All of which is to say: Japanese companies do a lot of things.
There are, of course, other countries with companies that “do lots of things”: much of Indian economic life, for example, is defined by the sprawling activities of a few large business clans—the Adanis, the Ambanis, the Tatas, the Birlas. But India is a relatively poor country with a low level of economic specialization, and the sprawling conglomerates that dominate its economy focus on relatively simple things like cement, steel, ports, and telecommunications. Japan, by contrast, is a wealthy, developed society—by one measure, the most economically complex country in the world. What’s striking about Japanese companies is not that they do lots of different things but rather that they do them very well. There are all sorts of high-precision inputs—the e-chuck being just one example—that are produced virtually only by Japanese firms.
This is very different from how most wealthy countries operate. American firms, for example, tend to prioritize focus above all else: it would be bizarre for an American paper mill to also operate a concert hall and an airport catering business, or for American Standard or Kohler to somehow have something to do with semiconductors. Even a country like Germany, which matches Japan in its depth of high-precision firms, has nothing like Japan’s corporate diversification. Only a few large conglomerates, like Siemens, have anything approaching the lateral breadth of the Japanese firm. South Korea—whose economic system was not coincidentally modeled off the Japanese one—does have a few chaebol conglomerates, like Samsung and SK, that truly do as many things as Japanese companies. But these are economy-dominating, state-entangled megafirms, cultivated as national champions by Korean industrial policy. They look nothing like, say, Sumitomo Osaka Cement, which is hugely diversified despite being relatively small. (“Look what they need to mimic a fraction of our power!”)
So why are Japanese companies like this? Why do they do so many different things? And how do they manage to do so all those different things so well?
Here is the answer I want to suggest: Japanese companies excel in lots of very different domains because it’s inherent in how they’re structured. The form of the corporation that we know and love in the United States—specialized, market-oriented, governed by shareholders—is just one form that the corporation can take; but it’s not the only way to coordinate capital and labor in a successful and profitable way. The protean corporations of Japan are best understood as a different species of thing altogether: better at some things, worse at others, but still highly adapted to their particular environment. And the things that they’re very good at turn out to be extraordinarily helpful for all sorts of things in which American companies tend to struggle.
To see why, we need to learn a little bit about the economics of industrial organization.
Companies are bundles of practices
In 1990, two economists—Paul Milgrom and John Roberts, both of Stanford—published a paper called “The Economics of Modern Manufacturing.” You should forgive them for the rather bland title. It was a very interesting, and very influential, paper.
Milgrom and Roberts started out by noting that manufacturing was “undergoing a revolution.” One paradigm of production was getting swapped out for another. In the past, there had been the “Fordist” paradigm: the factories that worked in this paradigm had long assembly lines of standardized goods, large buffer inventories, narrow and repetitive jobs for their workers, and dedicated single-purpose machinery. But that approach was being superseded by a new model: “a vision of a flexible multiproduct firm that emphasizes quality and speedy response to market conditions while utilizing technologically advanced equipment and new forms of organization.” This was the “post-Fordist” vision. In practice, this meant shorter production runs, rapid changeovers between products, smaller and more frequent deliveries from suppliers, workers trained to operate multiple machines and diagnose problems on the fly, and quality control embedded at every stage of the process. It was an entirely different way of producing things.
The question that Milgrom and Roberts wanted to answer was simple: why did all of these changes come as a package? Maybe it made sense for a specific firm to adopt shorter production runs; but why did it also make sense for them to do everything else in the “post-Fordist” category? Why did the changes seem to be so tightly clustered, with firms either having none of these practices or having all of them?
The explanation that Milgrom and Roberts offered was that the practices were complementary. Adopting any one of the “post-Fordist” practices raised the returns to adopting others, such that adopting only one of the practices didn’t make nearly as much sense as adopting the entire set.
Milgrom and Roberts formalized their argument using the mathematics of supermodular functions. But you don’t really need to know anything about math to understand the idea intuitively.
Here’s an illustration. Let’s say you run a factory. You decide that you want your lines to produce fewer defective goods: maybe you want to improve your yield from 95 percent to 98 percent. So you decide to invest in better training for your workers: maybe training now lasts six weeks instead of two weeks. This works, and now your yield is higher; but that change makes other things more attractive too. For example: now that your yield is higher, it makes sense for you to reduce your inventory, since fewer defects mean you no longer need a large buffer of spare parts to replace the bad ones. So now you’ve cut your inventory: but now it makes sense for you to shorten your production runs and switch more frequently between products, since without a mountain of inventory to work through you can afford to change what the line is making. And if you’re switching frequently between products, then it makes sense for you to invest in flexible, reprogrammable machinery instead of dedicated, single-purpose equipment. So one relatively small tweak shifts the entire calculus of what you do.
In short: each practice makes the others more valuable, and each practice is valuable because it’s implemented alongside other complementary practices. Doing just one of these things—investing in flexible machinery, for example—doesn’t really make sense alone. The practice needs to work well with all the other practices that you have.
So the correct way to think about organizational practices, Milgrom and Roberts suggested, was as bundles. A complete bundle of practices was worth more than the sum of its parts; and each part was worth less in isolation than as part of a bundle. So there was a coherent “Fordist” bundle of practices, and a coherent “post-Fordist” bundle. But there wasn’t much in between.
“The Economics of Modern Manufacturing” turned out to be the cornerstone paper for an entire paradigm of thinking about firms and how they work. (Milgrom won the Nobel Prize in 2020, though mainly for his separate work on the theory of auctions; you can watch a delightful video where he’s woken up in the middle of the night by his neighbor and corecipient, Robert Wilson, because he was sleeping and didn’t answer the call from the Nobel committee.)
Most importantly, the Milgrom-Roberts framework gave a strong answer to the question of why firms are the way they are, and why it’s so hard for them to change. A firm that uses one coherent bundle can’t easily move to another: changing one practice without changing the others will typically make the firm strictly worse off.
So if we want to know why Japanese companies have one apparently unusual practice—why they’re so diversified into countless unrelated industries—we can’t really answer the question in isolation. We need to ask which bundle of practices they employ.
And luckily for us, people have looked into this question. The central figure here is the economist Masahiko Aoki, who taught at Stanford alongside Milgrom and Roberts and worked closely with both of them. Through the papers that their collaboration produced—some by Milgrom and Roberts, others by Aoki alone—we can sketch a picture of what the Japanese corporation is, and why it works the way that it does.
Japanese companies are nothing like American ones
The first thing we should note is that Japanese companies do a lot of things differently from Western companies.
The most important of these, by far, is lifetime employment. Japanese firms tend to hire only at the very bottom, plucking new recruits straight from high school or university; they have all of those new recruits start on the same day of the year (the first of April); and they generally expect to keep these employees until they retire. Mass layoffs are essentially unheard of. Even in times of acute distress, a Japanese firm will go to great lengths to find its employees positions at smaller affiliates rather than releasing them onto the labor market. And individual performance isn’t really a huge criterion in someone’s career. Promotions are based largely on seniority; pay differentials between ranks are modest; and bonuses are tied to the performance of the firm.
Because they work for the same company for their life and socialize largely within that firm—nomikai drinking parties with colleagues are part of everyday corporate life—Japanese workers are often deeply attached to their company. Some employees even wear lapel pins to indicate their corporate loyalties. (For a time employees also sang corporate anthems, though that tradition has faded.) There are unions, but they’re organized within the firm: rather than a “national autoworkers’ union” that organizes in both Toyota and Honda, there is a “Toyota union” and a “Honda union” that don’t have much to do with each other.
And this means that Japanese companies strive to avoid financial pressure from outsiders. Relationships with suppliers are longstanding and entrenched: many Japanese companies have been working with the same suppliers for 50 years or longer. Outside investors seeking to interfere in this happy picture will find few avenues for influence. A standard Japanese firm’s board of directors is composed almost exclusively of the firm’s own senior managers; a large fraction of the firm’s equity is held not by outside investors but cross-held by other Japanese firms; and most of the firm’s financing comes from a single “main bank” that provides loans and monitors performance.
And as a result, Japanese companies don’t really try too hard to return profits to shareholders. Earnings are mostly reinvested, and investor dividends are kept low. For a long time, Japanese firms would spend as much entertaining the managers of other firms as they would on dividends to shareholders.
The crucial thing, Aoki suggests, is that we understand all of these distinctive features—lifetime employment, no benefits for individual performance, hostility to outside financing—as reflecting a particular bundle: a “J-firm” bundle, as he calls it, as opposed to the “H-firm” bundle that you encounter in the United States or Europe. The core difference, Aoki says, is that while in the H-mode production is organized vertically, in the J-mode it’s organized horizontally. (H is for hierarchy; J is for Japanese.)
Consider, for example, the famous “Toyota Production System,” the philosophy that determines how Toyota makes its cars. In a Toyota factory, there’s a rope called the andon cord that runs along the assembly line, within the reach of every worker. Anyone who spots a defect—like, say, a misaligned door seal, or a bolt torqued to the wrong specification—can pull the cord and halt production at any time; once they’ve pulled the cord, the workers and team leaders closest to the problem will converge and try to solve it on the spot. In an H-firm factory, by contrast—you can think of a classic Ford plant here—defects are reported to a line manager, who will make a report and send it up the chain of command, and the higher-ups will solve the problem.
The andon method is really the J-mode in miniature. Information flows laterally, authority to act is widely distributed, and the people closest to the problems are the ones who fix it. And one result of the Toyota-style approach is that Japanese automakers have produced fewer defective cars than their American competitors for a very long time.
But Aoki points out that the horizontal coordination embodied by the andon cord doesn’t work without other practices as well. For example: horizontal coordination requires that workers know each other’s jobs, since a worker who spots a problem in one area of the line can only act on it if he understands what that area is supposed to be doing. But in order to understand each other’s jobs, workers cannot be specialized: they have to rotate across different workplace functions to the point where they’re familiar with much of the plant’s operations. In order to rotate across different workplace functions, they need broad training; and it makes no sense to train them broadly if you don’t keep them for a very long time. And if you have generalist workers who are around for a long time, you can’t reward them based on how they do in one role, because then they’d have no desire to leave that role for another role where they might do worse. Instead you have to pay them based on company performance, and promote them based on seniority. And you also have to give them an ironclad commitment not to fire them if economic conditions worsen: if they can get laid off at any moment, why would they invest years of effort in learning all the idiosyncratic things that your firm does?
So now you have a firm that has lots of lifetime employees who can’t be fired, and whose skills are tailored to what your firm needs rather than to a particular occupational category transferable to any employer. That works very well for your company’s employees; but it makes no sense to outsiders. So the system only makes sense if the company is also insulated from outside pressure, whether from organized labor or from organized capital. Thus the other features of Japanese corporations: firm-level unions, insider-dominated boards, and broad hostility to outside capital.
So something as apparently simple as horizontal coordination only makes complete sense once coupled with an entire bundle of different things. That’s why attempts to install the andon cord and other aspects of the Toyota system in American car factories have generally produced mediocre results. American automakers noticed the superiority of Japanese cars a long time ago, and tried to implement the andon cord: but it just didn’t work with how their companies were organized. In 2007, workers at a Toyota plant in Kentucky pulled the andon cord 2,000 times per week; workers at a Ford plant in Michigan pulled it just twice a week. You can’t get all the benefits of a single practice without installing the complete bundle.
And the complete Japanese bundle, I should say, ends up producing something with entirely different objectives and interests than the American bundle. The H-firm exists to make money, or rather to return money to shareholders; but the J-firm, run by its employees and largely indifferent to the interests of shareholders, exists simply to continue existing. That’s why Japanese companies are so protean and willing to change what they do. Nintendo was founded in 1889 as a maker of handmade playing cards; in the 1960s, it was pushed out of the playing cards game by a wave of competition; and it spent several years experimenting with new markets—taxi services and instant rice, though contrary to the rumors not love hotels—before finding its way to video games. Fujifilm, which faced a near-total collapse of photographic film in the 2000s, simply used its expertise in chemical coatings and fine optics to pivot into cosmetics, pharmaceuticals, LCD films, and semiconductor process materials.
And that basic impulse toward survival is why Japanese companies are so insistent on diversification. If you’ve made a commitment to keep people employed for life, then you need to create jobs for them if their current jobs stop making sense: indeed, you might need to keep them employed even if you can’t find anything for them to do. If you’re not very worried about profitability, and have lots of well-trained generalist employees, then it makes perfect sense to reinvest your company’s earnings by expanding into new industries: doing so not only allows your company to survive longer—your company’s portfolio of bets is now more diversified and thus lower-risk—but also ensures that you’re able to keep your surplus workers busy in one way or another.
Bundles are hard to make and hard to unmake
If bundles are self-reinforcing, then the bundle, once established, will be very hard to dislodge. The only way to get from one peak to another is to change many things at once: and that kind of wholesale transformation almost never happens under normal conditions. It only happens during moments of acute crisis which make necessary a wholesale transformation in how things are done.
In the case of the Japanese bundle, that moment of acute crisis was the Second World War.
In the 1920s, the Japanese economy looked, on a structural level, quite American. It was already an industrial society—not quite a leading industrial power, but by far the wealthiest country in Asia—and it already had shipyards, steel mills, stock exchanges, and a growing electrical machinery sector. Heavy industry was dominated by a few family-owned conglomerates, the zaibatsu; but they operated more or less as normal firms, raising capital on public equity markets and operating under shareholder discipline. Workers, for their part, moved freely between firms and organized labor unions.
But in the 1930s and ‘40s, as Japan mobilized for total war in Asia and the Pacific, that system was reworked entirely. Total war required the rapid expansion of arms production, which meant channeling virtually the entirety of economic production into heavy industry. Japan became a “national defense state.” Capital was rerouted out of the equity markets and into the banking system, where it could be allocated under state supervision; firms were instructed to prioritize employees over shareholders in order to maximize production; wages were standardized by seniority to suppress bidding wars for skilled labor and keep workers in place. The economist Yukio Noguchi calls this planned economy “the 1940 system.” The entire point was to orient every aspect of economic life toward maximum production at all costs.
Of course, Japan wasn’t alone in that regard. Every major belligerent in the Second World War adopted some version of a production-oriented planned economy characterized by some kind of financial repression. But in Japan, the 1940 system outlasted the war by decades. Japan was defeated in 1945, and occupied by the American military until 1952; but after an abortive attempt to reorganize Japanese economic life, the Americans decided that the Cold War instead demanded the strengthening and entrenchment of the Japanese system. (This was dubbed the “Reverse Course.”) And the 1940 system, in its essence, survived: and the Japanese firms that emerged in the twentieth century were the result.
And this system, as it turned out, was really good at particular things. Aoki’s key insight was that the J-mode had a comparative advantage in environments of moderate volatility: situations where conditions changed frequently enough that rigid central plans would be outdated before they were executed, but not so radically that only top-down strategic intervention could cope. In an environment of stable, predictable demand, the H-firm did fine; in an environment of extreme disruption, where the whole product line had to be rethought, centralized authority was indispensable, and the H-firm also did fine. But in between—where the challenge was to make constant small adjustments in a changing but recognizable paradigm—the J-firm excelled.
And this was exactly what Japan needed. The postwar challenge was catch-up growth: Japan had to grow fast, and to do that it had to absorb and improve upon technologies that the West had already pioneered. J-mode firms—with their collaborative cultures, deep pools of broadly trained workers, culture of incremental shop-floor refinement, and large reservoirs of patient capital—were perfectly suited to the task. They could throw enormous amounts of patient capital at a problem, spend years refining a process without any imminent expectation of profit, and keep hundreds of broadly trained workers iterating on the shop floor until the quality of the output was world-class. And since profitability was never the primary objective, there was no pressure to abandon a difficult market for an easier one.
And the results truly were remarkable. By the 1960s, Japanese firms had begun to displace American ones in countless industrial sectors, from automaking to television manufacturing. Soon Japanese manufacturing was the envy of the world. Between 1946 and 1986, Japanese real per capita GDP grew tenfold, one of the fastest rates of growth in recorded history.
But catch-up growth, by definition, has to end: at some point you’ve caught up, and the challenge at the frontier is not only to refine what’s already known but to invent what is not known. And paradigm invention is precisely the sharp discontinuity for which the J-mode has no particular gift. Consensus-driven, horizontally coordinated organizations are very good at refining what already exists: but they are very bad at deciding what should exist.
That basic weakness is why Japanese firms are so dominant in some domains and entirely absent in others. Japan excels in automotive manufacturing, machine tools, industrial robotics, optics, and precision materials: domains characterized by incremental refinement. But they have very little to add in software, internet platforms, artificial intelligence, or electric vehicles. The architecture of the Japanese firm is built to perfect a domain through progressive advancement; it’s quite poorly suited to sharp discontinuity.
Consider Sony, which by the 2000s manufactured the world’s best portable music players, its best small cameras, its best mobile displays, and its best lithium-ion batteries: every component of what would become the smartphone. Purely on a material basis one would expect that Sony was the best-positioned company in the world to make the smartphone. But Sony didn’t do it. It was Apple, an H-firm par excellence, that reimagined the entire product category from the top down, largely because Apple was organized to give extraordinary power to a single visionary leader.
By the time that Aoki, Milgrom, and Roberts were writing in the last few years of the twentieth century, the shine of the Japanese model had already begun to fade. Asset prices in Japan had begun to deflate in 1990, inaugurating the country’s “lost decades.” Firms that had balanced against their assets at inflated prices now had more debt than they were worth, and the closely-affiliated banks that had lent them the money were now buried in so much bad debt that marking the loans to market value would have destroyed both the companies and themselves. Bankruptcies and mass layoffs were impossible in the Japanese system: a wave of firings or corporate restructurings would have undermined the entire social settlement that governed Japanese life.
So debts weren’t called: banks and companies simply soldiered on, “zombies” suspended in a state between life and death. Japanese business was no longer the envy of anyone in particular.
The Japanese bundle was exceptionally good, indeed world-historically good, at catch-up growth; but it was very bad at figuring out what to do once it found itself in trouble. Organizational bundles are remarkably resistant to change, even as conditions themselves change.
This is exactly what those who tried to reform Japanese corporate life in the last few decades have discovered.
In the 1990s, Fujitsu and other electronics firms experimented with performance-based pay: the idea had worked well at American firms and seemed like an obvious way to make Japanese workers more productive and thus to get the economy out of its slump. But performance-based pay didn’t cohere at all with the rest of the Japanese system. Team cooperation broke down, because output was measured individually and helping a colleague now hurt you in the rankings; senior engineers stopped mentoring juniors, because mentoring was uncompensated and the mentored juniors became future rivals; and managers struggled to keep their teams from disbanding. By 2001 Fujitsu had abandoned the practice. The episode became so infamous that one former Fujitsu executive wrote a book about it, titled The Downfall of Performance-Based Pay at Fujitsu as Seen by an Insider.
This is exactly what the economics of industrial organization would predict. High-powered individual performance pay makes sense when jobs are narrow, tasks are clearly measurable, and cooperation is inessential; but it makes no sense within the bundle that defines the Japanese firm. The same goes for practically every institution: piecemeal changes to a coherent bundle of organizational practices don’t really work; they only make things work less well. A reform that moves one coordinate but leaves the others in place produces a kind of organizational chimera, an entity that has lost the coherence of its old bundle without gaining the benefits of the new one.
But the Japanese bundle, however antiquated it might seem, still does result in some of the most remarkable companies in the world. The type of deep process knowledge that has accreted within companies like Kyocera and Toto is almost impossible to replicate. The American bundle of practices, with its emphasis on profits, entrepreneurship, and financialized risk, is probably the world’s best at innovation and frontier discovery. But as we are now discovering with the global rush on memory chips and other esoteric parts of the semiconductor supply chain, our entrepreneurial American system only works completely if it’s paired with a very non-entrepreneurial system like the one that we find in Japan.
This Substack is supported by readers like you.
And if you’re a longtime Toto shareholder—maybe an investor with a particular interest in bathroom fixtures—this has been a wonderfully lucrative year for you. Toto’s stock is up 60 percent year to date; in just the last few weeks, it’s risen by 30 percent. Toto is doing better than ever: its net profit, in the first quarter of 2026, was up 230 percent year over year.
But Toto’s remarkable year doesn’t have much to do with toilets or bidets. Toto might have been founded in the 1910s to “provide a healthy and civilized way of life” through affordable toilets, and in the decades since might have become the global leader in the bathroom game. But Toto also does a lot of other things. Toto manufactures not just bidets and toilets but also bathroom tiles, prefabricated bathroom modules, faucets, modular kitchens, photocatalytic coatings for buildings, and assistive equipment for the elderly. And, most importantly, Toto has a very lucrative sideline in the fabrication of memory chips.
Since 1988, in a once-obscure corner of the company called the “advanced ceramics division,” Toto has been producing a very particular component called the electrostatic chuck, or the “e-chuck.” The e-chuck is a sort of high-precision ceramic plate, about the size of a steering wheel, that uses electrostatic force to hold a silicon wafer perfectly flat and thermally stable while memory chips are etched into it with bombardments of plasma. Making these components is extraordinarily difficult, since the ceramic body needs to have near-zero particle generation and be polished to submicron flatness: and this means that there are only a few companies in the world that are capable of manufacturing e-chucks reliably. Almost all of them—Shinko Electric, NGK, Toto, Kyocera, Sumitomo Osaka Cement, Niterra—are based in Japan.
For most of its history, the advanced ceramics division was a rounding error on Toto’s balance sheet: the money maker, as it had been since the 1910s, was the toilet and bidet business. But we’re in a new era. Demand for AI is exploding, meaning that demand for the high-bandwidth memory that AI data centers require is exploding, meaning that demand for memory chips is exploding, meaning that demand for e-chucks is exploding. And so Toto’s advanced ceramics division is suddenly the company’s largest business, generating the majority of its operating profit. Toto’s leadership, suddenly awash in AI-driven revenue, announced that they would double down by investing hundreds of millions in expanded electrostatic chuck production: the toilet company had become, quite unexpectedly, a supplier to the semiconductor supply chain.
The Toto story is a fun and interesting illustration of corporate diversification and how strange bets can pay off. But that type of diversification—a toilet company that also produces photocatalytic coating and high-precision components for semiconductors—isn’t really unique to Toto. Practically every company in Japan seems to do a thousand very different things.
Consider, for example, Kyocera, another one of the e-chuck makers. Kyocera was founded in 1959 as a producer of ceramic insulators for cathode-ray tubes; today it manufactures not only industrial ceramics but also printers, smartphones, ballpoint pens, kitchen knives, solar PV modules, lens components, industrial cutting tools, automotive camera modules, electronics components, semiconductor packaging, biocompatible tooth and joint replacements, UV-LED curing systems, LCD systems, medical products, and lab-grown gemstones. Or another e-chuck maker. Sumitomo Osaka Cement, as you might have been able to deduce from the name, produces cement and ready-mixed concrete; but it also produces optical components, measuring instruments, industrial ceramics, artificial marine reefs, cosmetics and nanoparticle materials.
And this degree of diversification extends to many of Japan’s most famous companies. Yamaha, for example, manufactures pianos, motorcycles, guitars, drums, boats, snowmobiles, ATVs, audio equipment, golf clubs, tennis rackets, home appliances, specialty metals, molding and bonding equipment for semiconductors, and industrial robots. Hitachi makes nuclear reactors, power grids, railway systems, elevators, semiconductor manufacturing equipment, medical imaging devices, data storage, IT consulting, and industrial machinery. Ajinomoto makes MSG, frozen gyoza, instant coffee, amino acids for animal feed, IV drips, oligonucleotide drugs, and the insulating film layered inside high-end CPUs. Even a company as simple as Oji, Japan’s largest paper mill, has been drawn into the production of disposable diapers, functional films, adhesives, cellulose nanofibers, and wood-based EUV photoresists; and it also operates a hotel, an airport catering business, a concert hall, and an insurance agency.
All of which is to say: Japanese companies do a lot of things.
There are, of course, other countries with companies that “do lots of things”: much of Indian economic life, for example, is defined by the sprawling activities of a few large business clans—the Adanis, the Ambanis, the Tatas, the Birlas. But India is a relatively poor country with a low level of economic specialization, and the sprawling conglomerates that dominate its economy focus on relatively simple things like cement, steel, ports, and telecommunications. Japan, by contrast, is a wealthy, developed society—by one measure, the most economically complex country in the world. What’s striking about Japanese companies is not that they do lots of different things but rather that they do them very well. There are all sorts of high-precision inputs—the e-chuck being just one example—that are produced virtually only by Japanese firms.
This is very different from how most wealthy countries operate. American firms, for example, tend to prioritize focus above all else: it would be bizarre for an American paper mill to also operate a concert hall and an airport catering business, or for American Standard or Kohler to somehow have something to do with semiconductors. Even a country like Germany, which matches Japan in its depth of high-precision firms, has nothing like Japan’s corporate diversification. Only a few large conglomerates, like Siemens, have anything approaching the lateral breadth of the Japanese firm. South Korea—whose economic system was not coincidentally modeled off the Japanese one—does have a few chaebol conglomerates, like Samsung and SK, that truly do as many things as Japanese companies. But these are economy-dominating, state-entangled megafirms, cultivated as national champions by Korean industrial policy. They look nothing like, say, Sumitomo Osaka Cement, which is hugely diversified despite being relatively small. (“Look what they need to mimic a fraction of our power!”)
So why are Japanese companies like this? Why do they do so many different things? And how do they manage to do so all those different things so well?
Here is the answer I want to suggest: Japanese companies excel in lots of very different domains because it’s inherent in how they’re structured. The form of the corporation that we know and love in the United States—specialized, market-oriented, governed by shareholders—is just one form that the corporation can take; but it’s not the only way to coordinate capital and labor in a successful and profitable way. The protean corporations of Japan are best understood as a different species of thing altogether: better at some things, worse at others, but still highly adapted to their particular environment. And the things that they’re very good at turn out to be extraordinarily helpful for all sorts of things in which American companies tend to struggle.
To see why, we need to learn a little bit about the economics of industrial organization.
Companies are bundles of practices
In 1990, two economists—Paul Milgrom and John Roberts, both of Stanford—published a paper called “The Economics of Modern Manufacturing.” You should forgive them for the rather bland title. It was a very interesting, and very influential, paper.
Milgrom and Roberts started out by noting that manufacturing was “undergoing a revolution.” One paradigm of production was getting swapped out for another. In the past, there had been the “Fordist” paradigm: the factories that worked in this paradigm had long assembly lines of standardized goods, large buffer inventories, narrow and repetitive jobs for their workers, and dedicated single-purpose machinery. But that approach was being superseded by a new model: “a vision of a flexible multiproduct firm that emphasizes quality and speedy response to market conditions while utilizing technologically advanced equipment and new forms of organization.” This was the “post-Fordist” vision. In practice, this meant shorter production runs, rapid changeovers between products, smaller and more frequent deliveries from suppliers, workers trained to operate multiple machines and diagnose problems on the fly, and quality control embedded at every stage of the process. It was an entirely different way of producing things.
The question that Milgrom and Roberts wanted to answer was simple: why did all of these changes come as a package? Maybe it made sense for a specific firm to adopt shorter production runs; but why did it also make sense for them to do everything else in the “post-Fordist” category? Why did the changes seem to be so tightly clustered, with firms either having none of these practices or having all of them?
The explanation that Milgrom and Roberts offered was that the practices were complementary. Adopting any one of the “post-Fordist” practices raised the returns to adopting others, such that adopting only one of the practices didn’t make nearly as much sense as adopting the entire set.
Milgrom and Roberts formalized their argument using the mathematics of supermodular functions. But you don’t really need to know anything about math to understand the idea intuitively.
Here’s an illustration. Let’s say you run a factory. You decide that you want your lines to produce fewer defective goods: maybe you want to improve your yield from 95 percent to 98 percent. So you decide to invest in better training for your workers: maybe training now lasts six weeks instead of two weeks. This works, and now your yield is higher; but that change makes other things more attractive too. For example: now that your yield is higher, it makes sense for you to reduce your inventory, since fewer defects mean you no longer need a large buffer of spare parts to replace the bad ones. So now you’ve cut your inventory: but now it makes sense for you to shorten your production runs and switch more frequently between products, since without a mountain of inventory to work through you can afford to change what the line is making. And if you’re switching frequently between products, then it makes sense for you to invest in flexible, reprogrammable machinery instead of dedicated, single-purpose equipment. So one relatively small tweak shifts the entire calculus of what you do.
In short: each practice makes the others more valuable, and each practice is valuable because it’s implemented alongside other complementary practices. Doing just one of these things—investing in flexible machinery, for example—doesn’t really make sense alone. The practice needs to work well with all the other practices that you have.
So the correct way to think about organizational practices, Milgrom and Roberts suggested, was as bundles. A complete bundle of practices was worth more than the sum of its parts; and each part was worth less in isolation than as part of a bundle. So there was a coherent “Fordist” bundle of practices, and a coherent “post-Fordist” bundle. But there wasn’t much in between.
“The Economics of Modern Manufacturing” turned out to be the cornerstone paper for an entire paradigm of thinking about firms and how they work. (Milgrom won the Nobel Prize in 2020, though mainly for his separate work on the theory of auctions; you can watch a delightful video where he’s woken up in the middle of the night by his neighbor and corecipient, Robert Wilson, because he was sleeping and didn’t answer the call from the Nobel committee.)
Most importantly, the Milgrom-Roberts framework gave a strong answer to the question of why firms are the way they are, and why it’s so hard for them to change. A firm that uses one coherent bundle can’t easily move to another: changing one practice without changing the others will typically make the firm strictly worse off.
So if we want to know why Japanese companies have one apparently unusual practice—why they’re so diversified into countless unrelated industries—we can’t really answer the question in isolation. We need to ask which bundle of practices they employ.
And luckily for us, people have looked into this question. The central figure here is the economist Masahiko Aoki, who taught at Stanford alongside Milgrom and Roberts and worked closely with both of them. Through the papers that their collaboration produced—some by Milgrom and Roberts, others by Aoki alone—we can sketch a picture of what the Japanese corporation is, and why it works the way that it does.
Japanese companies are nothing like American ones
The first thing we should note is that Japanese companies do a lot of things differently from Western companies.
The most important of these, by far, is lifetime employment. Japanese firms tend to hire only at the very bottom, plucking new recruits straight from high school or university; they have all of those new recruits start on the same day of the year (the first of April); and they generally expect to keep these employees until they retire. Mass layoffs are essentially unheard of. Even in times of acute distress, a Japanese firm will go to great lengths to find its employees positions at smaller affiliates rather than releasing them onto the labor market. And individual performance isn’t really a huge criterion in someone’s career. Promotions are based largely on seniority; pay differentials between ranks are modest; and bonuses are tied to the performance of the firm.
Because they work for the same company for their life and socialize largely within that firm—nomikai drinking parties with colleagues are part of everyday corporate life—Japanese workers are often deeply attached to their company. Some employees even wear lapel pins to indicate their corporate loyalties. (For a time employees also sang corporate anthems, though that tradition has faded.) There are unions, but they’re organized within the firm: rather than a “national autoworkers’ union” that organizes in both Toyota and Honda, there is a “Toyota union” and a “Honda union” that don’t have much to do with each other.
And this means that Japanese companies strive to avoid financial pressure from outsiders. Relationships with suppliers are longstanding and entrenched: many Japanese companies have been working with the same suppliers for 50 years or longer. Outside investors seeking to interfere in this happy picture will find few avenues for influence. A standard Japanese firm’s board of directors is composed almost exclusively of the firm’s own senior managers; a large fraction of the firm’s equity is held not by outside investors but cross-held by other Japanese firms; and most of the firm’s financing comes from a single “main bank” that provides loans and monitors performance.
And as a result, Japanese companies don’t really try too hard to return profits to shareholders. Earnings are mostly reinvested, and investor dividends are kept low. For a long time, Japanese firms would spend as much entertaining the managers of other firms as they would on dividends to shareholders.
The crucial thing, Aoki suggests, is that we understand all of these distinctive features—lifetime employment, no benefits for individual performance, hostility to outside financing—as reflecting a particular bundle: a “J-firm” bundle, as he calls it, as opposed to the “H-firm” bundle that you encounter in the United States or Europe. The core difference, Aoki says, is that while in the H-mode production is organized vertically, in the J-mode it’s organized horizontally. (H is for hierarchy; J is for Japanese.)
Consider, for example, the famous “Toyota Production System,” the philosophy that determines how Toyota makes its cars. In a Toyota factory, there’s a rope called the andon cord that runs along the assembly line, within the reach of every worker. Anyone who spots a defect—like, say, a misaligned door seal, or a bolt torqued to the wrong specification—can pull the cord and halt production at any time; once they’ve pulled the cord, the workers and team leaders closest to the problem will converge and try to solve it on the spot. In an H-firm factory, by contrast—you can think of a classic Ford plant here—defects are reported to a line manager, who will make a report and send it up the chain of command, and the higher-ups will solve the problem.
The andon method is really the J-mode in miniature. Information flows laterally, authority to act is widely distributed, and the people closest to the problems are the ones who fix it. And one result of the Toyota-style approach is that Japanese automakers have produced fewer defective cars than their American competitors for a very long time.
But Aoki points out that the horizontal coordination embodied by the andon cord doesn’t work without other practices as well. For example: horizontal coordination requires that workers know each other’s jobs, since a worker who spots a problem in one area of the line can only act on it if he understands what that area is supposed to be doing. But in order to understand each other’s jobs, workers cannot be specialized: they have to rotate across different workplace functions to the point where they’re familiar with much of the plant’s operations. In order to rotate across different workplace functions, they need broad training; and it makes no sense to train them broadly if you don’t keep them for a very long time. And if you have generalist workers who are around for a long time, you can’t reward them based on how they do in one role, because then they’d have no desire to leave that role for another role where they might do worse. Instead you have to pay them based on company performance, and promote them based on seniority. And you also have to give them an ironclad commitment not to fire them if economic conditions worsen: if they can get laid off at any moment, why would they invest years of effort in learning all the idiosyncratic things that your firm does?
So now you have a firm that has lots of lifetime employees who can’t be fired, and whose skills are tailored to what your firm needs rather than to a particular occupational category transferable to any employer. That works very well for your company’s employees; but it makes no sense to outsiders. So the system only makes sense if the company is also insulated from outside pressure, whether from organized labor or from organized capital. Thus the other features of Japanese corporations: firm-level unions, insider-dominated boards, and broad hostility to outside capital.
So something as apparently simple as horizontal coordination only makes complete sense once coupled with an entire bundle of different things. That’s why attempts to install the andon cord and other aspects of the Toyota system in American car factories have generally produced mediocre results. American automakers noticed the superiority of Japanese cars a long time ago, and tried to implement the andon cord: but it just didn’t work with how their companies were organized. In 2007, workers at a Toyota plant in Kentucky pulled the andon cord 2,000 times per week; workers at a Ford plant in Michigan pulled it just twice a week. You can’t get all the benefits of a single practice without installing the complete bundle.
And the complete Japanese bundle, I should say, ends up producing something with entirely different objectives and interests than the American bundle. The H-firm exists to make money, or rather to return money to shareholders; but the J-firm, run by its employees and largely indifferent to the interests of shareholders, exists simply to continue existing. That’s why Japanese companies are so protean and willing to change what they do. Nintendo was founded in 1889 as a maker of handmade playing cards; in the 1960s, it was pushed out of the playing cards game by a wave of competition; and it spent several years experimenting with new markets—taxi services and instant rice, though contrary to the rumors not love hotels—before finding its way to video games. Fujifilm, which faced a near-total collapse of photographic film in the 2000s, simply used its expertise in chemical coatings and fine optics to pivot into cosmetics, pharmaceuticals, LCD films, and semiconductor process materials.
And that basic impulse toward survival is why Japanese companies are so insistent on diversification. If you’ve made a commitment to keep people employed for life, then you need to create jobs for them if their current jobs stop making sense: indeed, you might need to keep them employed even if you can’t find anything for them to do. If you’re not very worried about profitability, and have lots of well-trained generalist employees, then it makes perfect sense to reinvest your company’s earnings by expanding into new industries: doing so not only allows your company to survive longer—your company’s portfolio of bets is now more diversified and thus lower-risk—but also ensures that you’re able to keep your surplus workers busy in one way or another.
Bundles are hard to make and hard to unmake
If bundles are self-reinforcing, then the bundle, once established, will be very hard to dislodge. The only way to get from one peak to another is to change many things at once: and that kind of wholesale transformation almost never happens under normal conditions. It only happens during moments of acute crisis which make necessary a wholesale transformation in how things are done.
In the case of the Japanese bundle, that moment of acute crisis was the Second World War.
In the 1920s, the Japanese economy looked, on a structural level, quite American. It was already an industrial society—not quite a leading industrial power, but by far the wealthiest country in Asia—and it already had shipyards, steel mills, stock exchanges, and a growing electrical machinery sector. Heavy industry was dominated by a few family-owned conglomerates, the zaibatsu; but they operated more or less as normal firms, raising capital on public equity markets and operating under shareholder discipline. Workers, for their part, moved freely between firms and organized labor unions.
But in the 1930s and ‘40s, as Japan mobilized for total war in Asia and the Pacific, that system was reworked entirely. Total war required the rapid expansion of arms production, which meant channeling virtually the entirety of economic production into heavy industry. Japan became a “national defense state.” Capital was rerouted out of the equity markets and into the banking system, where it could be allocated under state supervision; firms were instructed to prioritize employees over shareholders in order to maximize production; wages were standardized by seniority to suppress bidding wars for skilled labor and keep workers in place. The economist Yukio Noguchi calls this planned economy “the 1940 system.” The entire point was to orient every aspect of economic life toward maximum production at all costs.
Of course, Japan wasn’t alone in that regard. Every major belligerent in the Second World War adopted some version of a production-oriented planned economy characterized by some kind of financial repression. But in Japan, the 1940 system outlasted the war by decades. Japan was defeated in 1945, and occupied by the American military until 1952; but after an abortive attempt to reorganize Japanese economic life, the Americans decided that the Cold War instead demanded the strengthening and entrenchment of the Japanese system. (This was dubbed the “Reverse Course.”) And the 1940 system, in its essence, survived: and the Japanese firms that emerged in the twentieth century were the result.
And this system, as it turned out, was really good at particular things. Aoki’s key insight was that the J-mode had a comparative advantage in environments of moderate volatility: situations where conditions changed frequently enough that rigid central plans would be outdated before they were executed, but not so radically that only top-down strategic intervention could cope. In an environment of stable, predictable demand, the H-firm did fine; in an environment of extreme disruption, where the whole product line had to be rethought, centralized authority was indispensable, and the H-firm also did fine. But in between—where the challenge was to make constant small adjustments in a changing but recognizable paradigm—the J-firm excelled.
And this was exactly what Japan needed. The postwar challenge was catch-up growth: Japan had to grow fast, and to do that it had to absorb and improve upon technologies that the West had already pioneered. J-mode firms—with their collaborative cultures, deep pools of broadly trained workers, culture of incremental shop-floor refinement, and large reservoirs of patient capital—were perfectly suited to the task. They could throw enormous amounts of patient capital at a problem, spend years refining a process without any imminent expectation of profit, and keep hundreds of broadly trained workers iterating on the shop floor until the quality of the output was world-class. And since profitability was never the primary objective, there was no pressure to abandon a difficult market for an easier one.
And the results truly were remarkable. By the 1960s, Japanese firms had begun to displace American ones in countless industrial sectors, from automaking to television manufacturing. Soon Japanese manufacturing was the envy of the world. Between 1946 and 1986, Japanese real per capita GDP grew tenfold, one of the fastest rates of growth in recorded history.
But catch-up growth, by definition, has to end: at some point you’ve caught up, and the challenge at the frontier is not only to refine what’s already known but to invent what is not known. And paradigm invention is precisely the sharp discontinuity for which the J-mode has no particular gift. Consensus-driven, horizontally coordinated organizations are very good at refining what already exists: but they are very bad at deciding what should exist.
That basic weakness is why Japanese firms are so dominant in some domains and entirely absent in others. Japan excels in automotive manufacturing, machine tools, industrial robotics, optics, and precision materials: domains characterized by incremental refinement. But they have very little to add in software, internet platforms, artificial intelligence, or electric vehicles. The architecture of the Japanese firm is built to perfect a domain through progressive advancement; it’s quite poorly suited to sharp discontinuity.
Consider Sony, which by the 2000s manufactured the world’s best portable music players, its best small cameras, its best mobile displays, and its best lithium-ion batteries: every component of what would become the smartphone. Purely on a material basis one would expect that Sony was the best-positioned company in the world to make the smartphone. But Sony didn’t do it. It was Apple, an H-firm par excellence, that reimagined the entire product category from the top down, largely because Apple was organized to give extraordinary power to a single visionary leader.
By the time that Aoki, Milgrom, and Roberts were writing in the last few years of the twentieth century, the shine of the Japanese model had already begun to fade. Asset prices in Japan had begun to deflate in 1990, inaugurating the country’s “lost decades.” Firms that had balanced against their assets at inflated prices now had more debt than they were worth, and the closely-affiliated banks that had lent them the money were now buried in so much bad debt that marking the loans to market value would have destroyed both the companies and themselves. Bankruptcies and mass layoffs were impossible in the Japanese system: a wave of firings or corporate restructurings would have undermined the entire social settlement that governed Japanese life.
So debts weren’t called: banks and companies simply soldiered on, “zombies” suspended in a state between life and death. Japanese business was no longer the envy of anyone in particular.
The Japanese bundle was exceptionally good, indeed world-historically good, at catch-up growth; but it was very bad at figuring out what to do once it found itself in trouble. Organizational bundles are remarkably resistant to change, even as conditions themselves change.
This is exactly what those who tried to reform Japanese corporate life in the last few decades have discovered.
In the 1990s, Fujitsu and other electronics firms experimented with performance-based pay: the idea had worked well at American firms and seemed like an obvious way to make Japanese workers more productive and thus to get the economy out of its slump. But performance-based pay didn’t cohere at all with the rest of the Japanese system. Team cooperation broke down, because output was measured individually and helping a colleague now hurt you in the rankings; senior engineers stopped mentoring juniors, because mentoring was uncompensated and the mentored juniors became future rivals; and managers struggled to keep their teams from disbanding. By 2001 Fujitsu had abandoned the practice. The episode became so infamous that one former Fujitsu executive wrote a book about it, titled The Downfall of Performance-Based Pay at Fujitsu as Seen by an Insider.
This is exactly what the economics of industrial organization would predict. High-powered individual performance pay makes sense when jobs are narrow, tasks are clearly measurable, and cooperation is inessential; but it makes no sense within the bundle that defines the Japanese firm. The same goes for practically every institution: piecemeal changes to a coherent bundle of organizational practices don’t really work; they only make things work less well. A reform that moves one coordinate but leaves the others in place produces a kind of organizational chimera, an entity that has lost the coherence of its old bundle without gaining the benefits of the new one.
But the Japanese bundle, however antiquated it might seem, still does result in some of the most remarkable companies in the world. The type of deep process knowledge that has accreted within companies like Kyocera and Toto is almost impossible to replicate. The American bundle of practices, with its emphasis on profits, entrepreneurship, and financialized risk, is probably the world’s best at innovation and frontier discovery. But as we are now discovering with the global rush on memory chips and other esoteric parts of the semiconductor supply chain, our entrepreneurial American system only works completely if it’s paired with a very non-entrepreneurial system like the one that we find in Japan.