REVIEW: MITI and the Japanese Miracle, by Chalmers Johnson

MITI and the Japanese Miracle: The Growth of Industrial Policy, 1925-1975, Chalmers Johnson (Stanford University Press, 1982).

I've been interested in East Asian economic planning bureaucracies ever since reading Joe Studwell's How Asia Works (briefly glossed in my review of Flying Blind). But even among those elite organizations, Japan's Ministry of International Trade and Industry (MITI) stands out. For starters, Japanese people watch soap operas about the lives of the bureaucrats, and they're apparently really popular! Not just TV dramas; huge numbers of popular paperback novels are churned out about the men (almost entirely men) who decide what the optimal level of steel production for next year will be. As I understand it, these books are mostly not about economics, and not even about savage interoffice warfare and intraoffice politics, but rather focus on the bureaucrats themselves and their dashing conduct, quick wit, and passionate romances... How did this happen?

It all becomes clearer when you learn that when the Meiji period got rolling, Japan's rulers had a problem: namely, a vast, unruly army of now-unemployed warrior aristocrats. Samurai demobilization was the hot political problem of the 1870s, and the solution was, well…in many cases it was to give the ex-samurai a sinecure as an economic planning bureaucrat. Since positions in the bureaucracy were often quasi-hereditary,¹ what this means is that in some sense the samurai never really went away, they just hung up their swords — frequently literally hung them up on the walls of their offices — and started attacking the problem of optimal industrial allocation with all the focus and fury that they'd once unleashed on each other. According to Johnson, to this day the internal jargon of many Japanese government agencies is clearly and directly descended from the dialects and battle-codes of the samurai clans that seeded them.

This book is about one such organization, MITI, whose responsibilities originally were limited to wartime rationing and grew to encompass, depending who you ask, the entire functioning of the Japanese government. Because this is the buried lede and the true subject of this book: you thought you were here to read about development economics and a successful implementation of the ideas of Friedrich List, but you’re actually here to read about how the entire modern Japanese political system is a sham. This suggestion is less outrageous than it may sound at first blush. By this point most are familiar with the concept of “managed democracy,” wherein there are notionally competitive popular elections, culminating in the selection of a prime minister or president who’s notionally in charge, but in reality some other locus of power secretly runs things behind the scenes.

There are many flavors of managed democracy. The classic one is the “single-party democracy,” which arises when for whatever reason an electoral constituency becomes uncompetitive and returns the same party to power again and again. Traditional democratic theory holds that in this situation the party will split, or a new party will form which triangulates the electorate in just such a way that the elections are competitive again. But sometimes the dominant party is disciplined enough to prevent schisms and to crush potential rivals before they get started. The key insight is that there’s a natural tipping-point where anybody seeking political change will get a better return from working inside the party than from challenging it. This leads to an interesting situation where political competition remains, but moves up a level in abstraction. Now the only contests that matter are the ones between rival factions of party insiders, or powerful interest groups within the party. The system is still competitive, but it is no longer democratic. This story ought to be familiar to inhabitants of Russia, South Africa, or California.

The trouble with single-party democracies is that it’s pretty clear to everybody what’s going on. Yes, there are still elections happening, there may even be fair² elections happening, and inevitably there are journalists who will point to those elections as evidence of the totally-democratic nature of the regime, but nobody is really fooled. The single-party state has a PR problem, and one solution to it is a more postmodern form of managed democracy, the “surface democracy.”

Surface democracies are wildly, raucously competitive. Two or more parties wage an all-out cinematic slugfest over hot-button issues with big, beautiful ratings. There may be a kaleidoscopic cast of quixotic minor parties with unusual obsessions filling the role of comic relief, usually only lasting for a season or two of the hit show Democracy. The spectacle is gripping, everybody is awed by how high the stakes are and agonizes over how to cast their precious vote. Meanwhile, in a bland gray building far away from the action, all of the real decisions are being made by some entirely separate organ of government that rolls onwards largely unaffected by the show.

Losers and haters are perpetually accusing the United States of being a surface democracy. Enemies of the state ranging from Ralph Nader to Vladimir Putin are constantly banging on about it, but this is a Patriotic Substack and we would obviously never countenance such insinuations about our noble republic. So there’s absolutely no chance it’s even the slightest bit true of the US, but…what about Japan?

Well, awkwardly enough, it turns out that the central drama of preindustrial Japanese history was the growing power of unofficial rulers (the shoguns) who ran the country in reality while the official rulers (the emperors) gradually devolved into puppets and figureheads. A “surface monarchy,” if you will. Of course that all ended with the Meiji Restoration of 1868 (c’mon, it says “restoration” right there in the name) which returned the emperor to being fully in charge…which is why when the Japanese declared war on America in 1941, neither the Emperor Hirohito nor the parliament was even consulted. Hang on a minute!

In fact, yes, prewar Japan may have been reigned over by a monarch, but it was ruled by the deep state — especially the career military general staff and the economic planning bureaucracies. I know it’s hard to believe that drab agencies regulating coal and steel production were able to go toe-to-toe with General Tojo, but just imagine that they were all being staffed by fanatical clans of demobilized samurai or something crazy like that. When MacArthur rolled in with the occupation forces, he had a goal of creating total discontinuity with Japan’s past and utterly bulldozing the government. But a guy needs to pick his battles, and so he obviously focused on getting rid of all those nasty generals and admirals he’d just spent years fighting. The harmless paper-pushers, on the other hand, how much trouble could they be? Maybe they could even help organize the place.

The chapter about the post-war occupation is one of the deadpan funniest in Johnson’s book. The American occupiers are genuinely trying to create a liberal democracy out of the ashes, but have no idea that the friendly, helpful bureaucrats they’ve enlisted in this quest were the secret rulers of the regime they'd just conquered. The stats bear this out — of all the officials who controlled Japan’s wartime industry, only a few dozen were ever purged by the Americans. The most striking example of continuity has to be Nobusuke Kishi,³ but there were countless others like him. These were the men charged with translating the occupiers' desires into policy, reconstructing Japanese society, and finally drafting a new constitution. Then eventually the Americans sailed off, and the bureaucrats smiled and waved, and went back to ruling as they'd done for hundreds of years, behind the scenes.

Okay, but how well does that version of history line up with the reality of Japanese government in the second half of the 20th century? Johnson brings a lot of evidence to back up his claim that Japan is still secretly ruled by the bureaucracies, chief among them MITI. He points out, for example, that hardly any bills proposed by individual legislators and representatives go anywhere, while bills proposed by MITI itself are almost always instantly approved by the parliament. But MITI’s authority isn’t limited to the government, it's pretty clear that they control the entire private sector too. That might seem tautological — if MITI’s will always becomes law, then they can unilaterally impose new regulations or mandates that can destroy any company, with zero recourse, so everybody will naturally do what MITI says. But it’s subtler than that — the real mechanism is tangled up in MITI’s dynastic and succession customs.

Remember, this may look like an economic planning bureaucracy, but it’s actually a secret samurai clan. So they’re constantly doing the kinds of stuff that any good feudal nobility does. For instance, the economic planning bureaucrats frequently cement their treaties by marrying off their sister/daughter/niece to a mentor or to a protegé. They also sometimes legally adopt each other, ancient Roman-style. Naturally they also have an extremely complicated set of rules governing their internal hierarchy, rights of deference, etc. But remember, this isn’t just a secret samurai clan, it’s also a government agency! Agencies have rules too — explicit rules written down in binders, rules governing promotion and succession and all the rest. Sometimes, the official rules and the secret rules conflict, butt against each other, and out of that friction something beautiful emerges.

The highest rank in MITI is “Vice-Minister” (the “Minister” is one of those elected political guys who don't actually matter). But it's also the case that somebody who's been at MITI longer or who's older than you (these are actually the same thing, because everybody joins at the same age) is strictly superior to you in seniority. But that can create a paradox! What happens if a young guy becomes Vice-Minister? He would then be more senior than his older colleagues by virtue of office, but they would be more senior by virtue of tenure, and that would mean either an official rule or a secret rule being broken. To resolve this impossible conflict, the instant a new Vice-Minister is selected, everybody who's been in the bureaucracy longer than him resigns immediately, so that his absolute seniority is unambiguous and unquestionable. And then...the first act of the new Vice-Minister is to give everybody who fell on their swords powerful jobs as executives and board members of the biggest Japanese corporations. The entire process is called amakudari, which means “descent from heaven.”

Amakudari is really a win-win-win-win: the new Vice-Minister has unchallenged power within the agency and a whole host of new friends in the private sector, the guys who resigned all have cushy new jobs that come with better pay and perks, the companies that are descended upon now have an employee with great connections to the agency that controls their fates, and MITI as a gestalt entity can spread its tentacles throughout the economy, aided by cadres of alumni who think its way and help translate policy into reality.

I joked before about refusing to tolerate speculation about the US being a surface democracy like Japan, but joking aside I think even the staunchest defender of the reality of popular rule would concede that things have moved in that direction on the margin. Compare the power of agency rulemaking, federal law enforcement, spy agencies, or ostensibly independent NGOs now to where they were even 10 years ago. It would be a stretch to say that the electorate didn’t have influence over the American state, but can they really be said to rule it? Regardless of exactly where you come down on that question, it’s probably safe to say that you'd give a different answer today than you would have twenty, fifty, or a hundred years ago. Moreover, the movement has been fairly monotonic in the direction of less direct popular control over the government. And in fact this phenomenon is not unique to the United States, but reappears in country after country.

Is there something deeper at work here? There’s a theory, popular among the sorts of people who staff the technocracy, that this is all a perfectly innocent outgrowth of modern states being more complex and demanding to run. The thinking goes that it was fine to leave the government in the hands of yeoman farmers and urban proles a century ago, when the government didn’t do very much, but today the technical details of governance are beyond any but the most specialized professionals, so we need to leave it all to them.

I think this explanation has something going for it, I admire the structure of its argument, but it also can’t be the whole story. For starters, it treats the scope and nature of the state’s responsibilities as a fixed law of nature. Another way to frame this objection is that you can easily take the story I just told and reverse the causality — the common people used to rule, and so they created a government simple enough for them to understand and command; whereas today unelected legions of technocrats rule, and so they’ve created a government that plays to their strengths. There’s no a priori reason to prefer one of these explanations over the other. There needs to be a higher principle, a superseding reason that results in selecting one compatible ruler-state dyad over another. I think there is such a principle, we just have to get darker and more cynical.

Let me tell you a parable about the origins of democracy. It isn’t actually true, but as with Nietzsche’s genealogies it isn’t supposed to be true, it’s supposed to be revealing. Once upon a time a country was ruled by a king, and inevitably whenever the old king died there was a huge and bloody civil war. Eventually, after the dust settled, one of the armies would be victorious and the other defeated, and the general of the victorious army would become the new king.

Then one day, somebody came up with a daring suggestion: what if instead of actually fighting a civil war, they instead had a pretend civil war. The two contenders for the throne would arm-wrestle, and everybody would treat the winner as if he had actually won the civil war, and thus many lives would be saved. Everybody applauded this idea, unfortunately the first time it was tried the loser of the arm-wrestling contest decided to try his luck anyways, broke the deal, started the civil war, and won. The problem with this approach is that it’s “unstable,” because one’s ability to win an arm-wrestle is only loosely correlated with one’s ability to win a hypothetical civil war. The rule-by-arm-wrestle system can work so long as nobody challenges it, but as soon as somebody does, it’s prone to collapse.

Then somebody else observed that in the last few civil wars, the side with the bigger army always won, and proposed that instead of settling the succession on the battlefield, the two sides simply count up the number of soldiers they would be able to muster, and the side with the largest hypothetical army would win without the war being fought. Note how different this situation is from the previous proposal! This time, the defeated party of the fake, simulated war has good reason not to be a sore loser, because he’s just seen that if the matter really came to blows, he’d probably lose. The solution is “stable” in this sense, all sides are incentivized to accept the outcome. And thus democracy was born.

I like this as a pragmatic argument for a loosely democratic system. It has nothing to do with the moral case for popular sovereignty, or whether it is right and just for the governed to have a say in government, it’s simply about avoiding violent instability by giving everybody a sneak peek at how the putative civil war might turn out, then all agreeing to not have it. But this theory has another selling-point, which is that it also tells us why democracy arose when it did, and why it may now be on the way out. If the principle is that governments will tend towards a form and structure and rule of succession that’s closely tied to their ability to fend off challengers, the that suggests that the most common form of government will depend heavily on what the dominant military technology and strategy of its era happens to be.

For example: in the early Middle Ages, wars were fought by a much smaller number of people, and success in warfare was more dependent on the actions of an elite group of professional soldier-aristocrats. And sure enough, political power was also concentrated in the hands of this much smaller group, because in the event that somebody decided to contest the state, it was the opinion of this group that mattered, not the opinions of everybody.

Sometime in the nineteenth century, the “meta” for total warfare changed dramatically. The combination of mass production, replaceable parts in machinery, and new weaponry that was deadly even in the hands of the untrained masses, all meant that suddenly the pure, arithmetic quantity of men under arms on each side became a much more potent factor in the military calculus. Is it any wonder that a little while later, democracy began to spread like wildfire around the globe? Mass suffrage and mass conscription are inextricably bound with one another. The people have generally ruled in our lifetimes, but only because a little while before (these things always operate on a lag) wars were decided by masses of conscripts with rifles.

There’s no rule that says this connection between military success and popular support has to hold true forever, and in fact it probably won’t. You can imagine this going a few different ways. Perhaps the conflicts of the future will be settled by vast swarms of autonomous killer robots, and the winner will be whoever can produce the best robots the fastest. This world might be conducive to rule by industrial conglomerates and robber-barons, a return to the great age of oligarchy, but with a less aristocratic, more plutocratic spin. If we look to the past, there was a class of societies whose militaries had an extreme ratio of capital intensity to labor intensity — the Mediterranean merchant republics with their fleets and their mercenary armies of condottieri. If future wars are settled by robots, we may find ourselves bowing to a new, doubtless very different, doge.

There’s another possible world, where control of information becomes supreme. You can think of this world as being an intensification of our current one, with an arms race of ever more sophisticated techniques for swaying the masses. Surface democracy spins out of control as an ecosystem of competing psychological operations vie to program or reprogram or deprogram swarms of bewildered and unsuspecting voters, alternatingly using them as betting chips and battering rams. This is a world ruled by the meme lords — brutally efficient teams of spin doctors, influencers, AIs, and the occasional legacy media organization. Like I said, pretty much just an intensified version of our current world.

My guess, however, is that neither of these worlds will come to pass, but instead a third one. The history of military technology is a history of the ancient contest between offensive technologies and defensive technologies, with both sides having held the crown at various points. We may be about to see the balance shift decisively in favor of offensive technologies, with extreme political consequences. Arguably we’ve been in that world ever since the invention of the atom bomb, but WMDs haven’t affected this strategic calculus as much as you might guess, due to all the issues surrounding their use (to be clear, this is a good thing).

Technology marches on, however, and I believe there’s a chance that it’s about to deliver us into a new golden age of assassination.⁴ Between miniaturized drones with onboard target recognition, bioengineered plagues designed to target exactly one person, and a host of more creative ideas that I don’t even want to write about for fear of summoning them into existence, it may soon become very dangerous to be a public figure with any enemies — that is to say, dangerous to be a public figure at all. What kind of men will rule such a world, where your reign could end the moment somebody discovers it?

Two kinds of men: men with nothing to lose, and men that you will never find. This world of ever-present threat to those with power is a world eerily well adapted to governance by grey, faceless men in grey, faceless buildings. A world of conspiracies hatched in unobtrusive exurban office parks, of directives concealed within stacks of paperwork, where the primary goal of power is to hide itself from view. In other words it’s the world that MITI already inhabits. As in so many things, the future is here, it’s just unevenly distributed.

1. And you thought civil service protections in the US were bad!

2. What “fairness” means when there is only one party, everybody knows there is one party, and everybody knows how to get on the party’s good side, is sort of an interesting philosophical question. Not all orders need to be explicit.

3. Briefly: Kishi was a descendant of samurai (of course) who became an economic planning bureaucrat (of course) and then the dictator of the Japanese puppet state in Manchuria in the 1930s. During his reign he tried out a lot of the industrial policy ideas that would later fuel the Japanese postwar boom…and also brutalized the population to such an extent that even other Imperial Japanese colonial administrators thought he was excessive. Later he signed the declaration of war against the United States (he was an economic planning bureaucrat, after all), and was briefly imprisoned as a war criminal after the Japanese surrender. Within a few years, however, he was back out, and running the country as prime minister. His brother was also prime minister. Oh…and his grandson was a guy you might have heard of, a guy named “Shinzo Abe.”

4. Japan had a high-profile and socially traumatizing assassination just recently. I find it noteworthy that Abe was killed when he wasn’t Prime Minister anymore, but was perhaps more influential than ever as a deep state power player.