Functional and Dysfunctional Aspects of Government Bureaucracy

Albert M. Craig

There is no language in the world in which the word "bureaucrat" means something good. In Japanese the word kanryo^[*] has exceedingly unpleasant overtones. It suggests pettiness, self-importance, narrowness, and formalism—attributes, in the Confucian phrase, of a "small man." Japanese historians tend to view the modern Japanese bureaucrat as the willing tool of all of the villains of history from the absolutist leaders of the early Meiji to the undemocratic party leaders of the Taisho, the militarists of the thirties, and the conservative party politicians of today. In the historical literature perhaps the only figure that rivals the bureaucrat as a subject of opprobrium is the "parasitic landlord." Some critics even suggest that the bureaucracy has dominated policy formation, bringing to heel the elected political leaders who are nominally in charge of policy.

One may object that both "bureaucrat" (kanryo ) and "official" (yakunin ) are pejorative terms. Kanri , which also may be rendered as "official," sounds much better. It has a dignified ring; it suggests capability, steadfastness of purpose, and trustworthiness. It was the term used in the Meiji constitution to speak of "the officials of the emperor." But it is not the term that enters the mind of the average Japanese when he thinks of Japanese officialdom.

Even more critical of bureaucracy than the writings of postwar historians is a second large literature written by ex-bureaucrats and newspaper reporters to expose the foibles, failings, formalism, and frustrations of bureaucratic life in Japan. This genre of books elucidates what is wrong with Japanese bureaucracy. It gives firsthand illustrations of

the dysfunctions that can be found in government administration. If Japanese bureaucracy did not work, this literature would tell why.^[1]

During the past few years I have been studying early Meiji government and bureaucracy. In connection with this study, hoping in some measure to triangulate back from the present, I have read most of this exposé literature, and I have also spent several months visiting Japanese government offices—at the local, prefectural, and national levels—and talking with government officials. This paper is an attempt to reconcile what I have read with what I have seen and heard.

I would stress in advance that I am not out to whitewash Japanese bureaucracy. It does have dysfunctions. Japanese officials all agree that they exist. There is some truth in the claim by Japanese scholars that Westerners, seeking to explain Japan's "economic miracle," sometimes adopt an overly rosy view of the Japanese ability to cooperate and work harmoniously for commonly held goals. Even so, in comparison with most bureaucracies in the world, Japanese bureaucracy does get things done. The question must be posed: given its dysfunctions, why does it work as well as it does?

An Elite Bureaucracy

Weber argued for the legal character of bureaucracy. His characterization fits Japan rather well. Whether in the Ministry of Labor or the Ministry of Finance, the top career bureaucrats tend to be law school graduates. Their life as officials is governed by regulations and ordinances, which they observe and obey. Their work consists of drafting laws, planning the projects that will be embodied in bills, and enforcing laws already passed.

But many other matters also affect how bureaucracy functions. One is the prestige that the official enjoys in his society and the way it helps or hinders recruitment. In the Tokugawa era (1600–1868) officials were ranking samurai, the elite within the military aristocracy. After the Restoration, government officials were, perhaps more than the army, the legitimate heirs of the samurai. For a time after the Restoration even local post-office officials wore swords. In 1876 one government leader wrote: "Carefully observing today's situation, the peasants, the merchants, the samurai, all under heaven are dissatisfied . . . only officials are pleased with their lot."^[2] There was ample reason for their satisfaction: they were

[1] The best scholarly work on Japanese bureaucracy is Tsuji Kiyoaki, Nihon kanryosei^[*] no kenkyu^[*] [A Study of the Japanese Bureaucratic System] (Tokyo, 1969). Representative works in the genre of exposé literature are Imai Kazuo, Kanryo^[*] , sono seitai to uchimaku [Bureaucracy, Its Mode of Life and Inner Workings] (Tokyo, 1953); Masuda Yoneji, Oyakunin [Honorable Official] (Tokyo, 1956); Mainichi shimbunsha, ed., Kanryo Nippon [Bureaucratic Japan] (Tokyo, 1956); Tokyo shimbunsha, ed., Kancho^[*] monogatari [A Tale of Government Offices] (Tokyo, 1962); and Hori Makoto, Oshoku [Corruption] (Tokyo, 1957). Of the latter works, the volume by Imai is by far the best. He exaggerates so that his book will sell, but he writes with insight, humor, and wit.

[2] Kido Koin^[*] nikki [The Kido Koin Diaries] (Tokyo, 1933), III, 433.

the officials of the emperor and partook of his glory. Higher officials, from a slightly later period, had fancy uniforms with shiny buttons, braids, and swords to wear on ceremonial occasions. Their pay was high. Even a lower official (hanninkan ), Ukai Nobushige has stated, could afford to visit geisha.^[3] It was not until the 1920s that the salaries of company managers pulled ahead of those of government officials at the same educational level.

After World War II the pay of bureaucrats declined relative to other sectors of society. They were redefined in the 1947 constitution as "servants of the people" (kokumin no koboku^[*] ). They were stripped of their symbolic perquisites. Decorations, for example, were abolished for a time. And, reflecting the decline in their livelihood and morale, incidents of corruption or bribery, exceedingly rare before the war, began to be reported. Yet many of the best graduates of the Law Faculty of Tokyo University continued to enter the bureaucracy. Perhaps only in France and Germany did officials have a comparable sense of themselves as an academic elite.

Imai Kazuo, a former bureau chief in the Ministry of Finance, recounts an occasion when a student, armed with a letter of introduction, called on a bureau chief in a central ministry in the hope of obtaining a job. He was asked about his university record and then dismissed with: "You haven't got a chance with only that number of As. I had seventeen. The famous Professor So-and-so at the university had only fifteen, a considerably poorer record than mine." The student left in chagrin. When first told of this conversation, Imai remarked on how true to life it was.^[4] The same consciousness is reflected in the more recent words of a ten-year MITI bureaucrat who asserted: "We [graduates] of the [Tokyo University] Law Faculty are the most Japanese of Japanese" since Japan is essentially an "examination society."^[5]

Within the bureaucracy there was and is a sharp cleavage between the elite of career officials, who have passed the special higher examination, and lower-level bureaucrats. Distinguishing between these two levels enables one to resolve many contradictory statements about Japanese bureaucracy. I arrive at a bimodal view: of exceedingly able men directing the work of the central ministries, but much less able men in the offices of local government; of graduates from the two or three best universities versus those from lesser schools and, at a lower level yet, high school graduates; of generalists versus specialists, technicians, and clerks; of relative freedom in creative work versus mechanical, rote tasks; of the satisfaction of power versus the frustrations of a narrow routine. Of course, the bimodality is not simply between the central ministries and local government. Those who will reach the highest appointive positions in the

[3] Ukai Nobushige, personal communication. Imai makes the same point, Kanryo^[*] , p. 25.

[4] Imai, Kanryo , pp. 47–48.

[5] Kawamoto Taro^[*] , "Sore demo Todai^[*] hogakubu^[*] wa Nihon o ugokasu" [Even Then, It is the Law Faculty of Tokyo University that Moves Japan], Bungei shunju^[*] , September 1972, p. 179.

central government will spend some time in prefectural-level offices. There are planning and executive posts at the prefectural level that are highly demanding, and many of the sections of the central ministries are filled with clerks, assistants, researchers, and temporary employees performing the most routine kind of work.

The elite bureaucrats move frequently from post to post and often occupy certain positions which are, as it were, on a special track, reserved for the most promising careerists. In the late Meiji period, a Tokyo University graduate might enter the Home Ministry, then become the assistant chief of police in a small city, move back to a different bureau in the ministry, then to a prefectural government post, and so on. Today, a Finance Ministry official may have a similar career pattern—moving from the ministry to a brief appointment in a prefectural tax office, then back to the ministry, then again to a prefectural office, and then back again to the ministry. Even after the Occupation separation—along American lines—of central and prefectural government, this type of transfer of officials from the central to the state level still continues.

The special aura of power that still surrounds the elite administrators can be explained partially in terms of their postcareer pattern. There is an unwritten rule in the bureaucracy that university graduates who entered in any given year cannot serve under men who entered at the same time or later. Since bureaucracy is pyramidal—the number of posts at each higher rank decreases—this rule means that those of each class passed over for promotion must be retired. This resembles the system in the armed forces of the United States. Some of those sloughed off in this fashion are kept in service for a while longer by postings to high positions in prefectural governments. But even the few who become bureau chiefs or vice-ministers, the highest posts in the system, are usually retired in their early or middle fifties. These men, the elite of the elite, often receive their rewards for loyal service after retirement. Some become the directors of corporations, receiving fees considerably greater than their former salaries. Some are appointed as members of semi-official advisory commissions (shingikai ). A few begin a political career and make it to the Diet, and a handful achieve the eminence of a cabinet post. Because such postretirement careers are the pattern in Japan (some ministries are better situated for this than others), even the junior bureaucrat who has passed the higher examination is viewed as a possible future member of the select company that defines Japan's national goals.^[6] This contributes to his status,

[6] In the Kancho^[*] monogatari , the commissions and other gaikaku dantai are compared to Ubasuteyama, the folktale mountain where the old were abandoned when they could no longer contribute to the life of the community. See Kancho monogatari , p. 45. Sometimes the financial power of the ministry is employed to assure a good reception for its retirees. In one instance the Education Faculty at a Japanese national university was asked by the Ministry of Education to take on as a professor an official of a certain rank who had been passed over for promotion in the ministry. The faculty council argued about the matter for some time. Eventually they accepted the official and gave him the appointment. As a trade-off theyreceived a large, handsome new building. Imai writes of another instance in which a post as section head opened up in a prefecture. A man, able, experienced, and first in line in the prefectural officialdom, was chosen for the job. But when permission was asked of the ministry concerned, the man was rejected. (The officials who handle the prefectural affairs of the central ministries are appointed only with the approval of the ministry concerned.) The governor took the unusual step of going to Tokyo and asking the ministry to let him have the man he wanted. He was, however, turned down and had to accept a man sent out from the ministry. The man from the ministry was older, less able, and less experienced. Imai writes that for the ministry to send this official who was old, lacking in ability, but without glaring failures in his record, to an honorable post in the countryside was a way of getting rid of deadwood (funzumari o dakai suru ) in the ministry. When he arrived to take up the post of section head in the prefecture, he was not well received. The new appointee, however, was exultant at having attained the post of section head, and immediately ordered his underlings to procure a car for his personal use. There was no car in the budget, and his juniors reported back that it was impossible. The new section head berated them as incompetents and said he himself would get it from the ministry. His juniors felt that at a time of budgetary cuts the frugal ministry would not waste money on such a sad figure of a section chief (chinpira kacho^[*] ). But he surprised them by coming back with an allocation, which he used to buy the car. This established his reputation in the prefectural bureaucracy as a real somebody. Even though it was a complete waste of money, having forced the prefecture to accept the man, the ministry had to pay this price to get him established. Imai, Kanryo^[*] , pp. 188–190.

even though his salary may be less than his opposite number in Mitsubishi.

Because of the variety of career patterns, even among the higher officialdom, there is no unanimity about the merit of life in the bureaucracy. One official of the prewar Home Ministry privately compared the life of the official to "soaking in lukewarm water." It lacks the satisfaction of a good hot bath, yet "as long as one remains in it, one will not catch cold. But if one loses his patience and jumps out, he will invariably come down with a cold."^[7] On the other extreme the editorialist Akimoto Hideo once rode on the bullet train with a budget examiner, one of the most powerful officials of the Finance Ministry. The official suddenly told him to look out the window. "See that bridge over there? I built it." A few minutes later he pointed out: "See that road? The wide one. That's my work too." Then after a while he said again: "See that windbreak forest over there? That was one of my most difficult jobs." Akimoto noted the expression of consummate satisfaction on the face of the official.^[8] Between these two extremes are the observations of a young career official in the Foreign Ministry:

We climb the hierarchy with exactly the same speed. The question is which post one is assigned to. This is not an absolute indicator, as the Foreign Ministry has the policy of assigning career officers alternately between good (Europe and the United States) and not so desirable (Africa, some of the Asian countries, and South America) posts. As we climb with the same speed, there is no real competition, although almost all of us, having gone through a very competitive educational system, are competitive in a good sense. That is, everybody is a genuinely hard worker, disliking easy posts. Usually we are intelligent, but, surprisingly, some of us are not so intelligent. Because we are given reasonably responsible and important

[7] Imai, Kanryo , pp. 36–37.

[8] Kawamoto, "Todai^[*] hogakubu^[*] ," p. 171.

jobs while in junior grade, the stupid ones become conspicuous at once. They are not, however, punished, at least in salary or grade. This kind of promotion continues until we become counselors. From then on it is luck.

We share the same interests both in personal matters and professionally. Over dinner or cocktails we talk a lot about foreign affairs and are quite willing to be critical of the government, though we aren't to people who might quote us. The atmosphere of the office, either in Tokyo or abroad, depends a great deal on the personality of the men, particularly of the chief. Generally speaking, discussion is free. We are invited to dissent. The level of discussion is usually high. For example, I am in charge of such and such a matter and am expected to know the facts about it. Facts are important. Treaty clauses, technical details of strategic thinking, dates, figures, etc., are considered more important than sheer speculation, although speculation, too, is expected of us. We often say that if an officer is incapable of becoming an expert in six months concerning the subject matter of his assignment, he shouldn't be taken seriously. The majority of the officers accomplish this mark with ease.^[9]

When one talks with the higher officials of other ministries, the content of their work may vary, but the sense of competence and self-confidence they project is much the same. In this respect they are not unlike Japanese business leaders.

Is this elitist consciousness appropriate in a democratic polity? Many Japanese say that it is not, arguing that bureaucrats tend to look down on civilians from their position of superiority. This is summed up in the old maxim, "officials are honored and the people are despised" (kanson minpi ). It is not that the populace distrusts the bureaucracy—though some on the left do. Most recognize its competence and give it a measure of respect. It is not too much to say, in fact, that many in the opposition parties dislike it because of its competence. But all complain of the arrogance of officials. The resentment against this arrogance may be the most seriously dysfunctional aspect of the whole system.

Masuda Yoneji, a onetime official in the Ministry of Labor, wrote that after he became section chief, visitors to the ministry would bow deeply and speak to him in exaggeratedly polite language as if he were "a member of a different race." When he went to inspect a textile mill in Aichi, he was made to feel a degree of superiority and satisfaction as never before in his life. He was met on the train platform by the section chief, a ten-year official, and his two aides from the local government office. They treated him politely, bowed, carried his briefcase to the waiting car, and escorted him to the local inn. After he bathed, he was visited by the works manager and section head in charge of labor affairs of the factory he was to inspect. They called in geisha, plied him with drinks, and in general treated him with such a combination of respect and deference (English does not do justice to the expressions used) that for the first time in his life he felt the "indescribable pleasure" of embodying in himself "the power of

[9] Personal communication. I have been asked not to cite the interviewee by name.

the state."^[10] Imai, after describing instances of the same kind of obsequious behavior in even more loving detail, offers an analysis of it:

In the course of such experiences—like starting to smoke or taking up the habit of an evening drink—one at first feels a little ticklish, but before long it begins to feel good, and one feels that something is missing if he isn't treated in this way. Next one feels an inward dissatisfaction, and in the final stage, one advances to the symptom of indignation. It is exactly the same as addiction to a drug. . . . The more one is a petty, self-righteous type (ki no chiisai zennin gata ), the more likely he is to fall prey to this disease of "bureaucratic mentality" and the worse will be its ravages. . . . And almost no one is entirely immune.^[11]

Nor would it exaggerate to say that there are some who look forward to catching the disease. In 1972 a student who had passed the entrance examination for Tokyo University was interviewed on television and asked about his future. He replied, and his face was radiant as he spoke: "I've been accepted by the Law Faculty. In the future I'll pass the exam of the Ministry of Justice and become a judge. As a judge I'll be a god. I'll have the power of life and death over men. I'll be able to move society in any direction I please."^[12] In this response there is an element of brashness and an element of black humor, together with a large measure of overoptimistic anticipation. Yet it is not wholly off the mark. The entering freshman in the Law Faculty of Tokyo University, if he perseveres, stands a good chance of becoming a judge. Judges in Japan are officials of a kind. The path to becoming a judge is separate but not different from that into the Ministry of Justice. The kind of authority exercised by a judge and the unassailable character of his position is, like that of other officials, based on a legal monopoly of certain powers. The potential of those officials who fell prey to the "disease of bureaucratic mentality" to act arrogantly stemmed from a system that, until recently, had few political checks on the use of this power.

Since Japanese officialdom is characterized by bimodality, its relations to the clientele it serves are also of two kinds. Higher officials deal mainly with other officials and politicians, or with corporations and other large organizations. They rarely come into contact with the average citizen. The attitude of the higher officials toward, say, corporations is determined by a combination of law, custom, and politics. As in most countries, most laws are not strictly enforced most of the time; they function as guidelines or limits. Bureaucrats could throw the economy into chaos if they began to implement strictly all of the laws for which they are responsible, just as police can start a crime wave by enforcing all of the laws in the criminal code. Of course they do not. Yet because of the unused discretionary power of the bureaucracy, companies large and small take great pains to avoid

[10] Masuda, Oyakunin , pp. 35–36.

[11] Imai, Kanryo^[*] , pp. 52–53.

[12] Kawamoto, "Todai^[*] hogakubu^[*] ," p. 171.

falling into its bad graces. Businessmen employ a precautionary politeness in dealing with officials. They assume an inferior status and treat even low-ranking officials with an exaggerated deference—as if they were ministers of cabinet rank. This has been brought to a high art. It is the safe way of manipulating officials. It is seen as a low price to pay for getting along with a bureaucracy that is largely honest and supportive toward business. But it is resented nonetheless, for businessmen feel that it is they and not the bureaucracy that are the propulsive force in today's Japan.

The attitude of the local bureaucracy toward its civilian clientele is perhaps the more serious problem. Whether a Japanese is going to the district office for a copy of his birth registration certificate or to the local tax office, he goes in some degree as a supplicant. If he does not act as one, he can, depending on the circumstances, be made to wait or told to come back another day. The official is doing him a favor. The official tends to see the power of the office as his personal power. The resentment engendered by this kind of treatment is widespread and affects the Japanese view of government. This is far from being as uniquely Japanese as the Japanese imagine. In good measure it is the behavior of the petit fonctionnaire anywhere.^[13] It is especially characteristic of countries where democracy is recent and is predated by a bureaucratic tradition. Yet perhaps it is more visible in Japan where degrees of politeness can be precisely measured by the depth of bows and the incremental use of honorific language.

In Japanese society at present there are some trends that run counter to the tradition of bureaucratic arrogance. Local government in the early postwar years was predominantly in the hands of conservative politicians who, appealing to all of the voters, were not formally identified with a party. Recent local elections, however, especially those in the larger cities, have become politicized; the power of the conservatives has been challenged. To attract voters, many administrations have adopted a new service orientation toward the citizenry. "Progressive" politicians, such as the mayor of Kyoto, have established "citizen consultation offices" (shimin sodanshitsu^[*] ) to handle complaints about broken roads, garbage, or whatever. Conservative mayors have reacted by setting up similar offices. The mayor of Matsudo in Chiba Prefecture called his the "do-it-right-away section" (sugu yaru ka ). Through such small changes the attitudes of officials are being democratized little by little. The outbreak of "residents' movements" during the early seventies in protest against pollution is a parallel development. As these movements turn to the electoral process,

[13] Bureaucratic subordinates may suffer along with the public. Imai tells the story of an official who after thirty years became head of a small government office. He lined his staff up before New Year's and after lecturing them on the proper attitude toward their work said: "It is most regrettable that recently the number of officials who do not understand propriety have increased. At the end of last year there were even some who brought me year-end presents bought at a department store special sale." Imai, Kanryo^[*] , p. 61.

local government officials move further in the direction of accommodating public interests.

Vertical Relationships and Cliques

Japanese officials are Japanese and bring with them into bureaucracy patterns of behavior characteristic of their society. If the laws, ordinances, and regulations constitute the formal system of the bureaucracy, then these patterns of social relations that affect official behavior may be called the informal system. Two aspects of the informal system are so important to the way Japanese government functions that they merit special comment. One is the vertical personal tie; the dysfunction associated with this is cliquism. The other is the solidarity of the office, which gives rise to interoffice jealousies and struggles over jurisdictions.

Japan's premodern society was organized about the vertical personal tie. Not only did this "feudal" bond define the relationship between lords and their samurai vassals, but it encompassed other relationships as well: that of master artisans and their apprentices, of merchants and their clerks, and so on. Within the domain governments, personal ties also formed between senior and junior officials. These became the basis for the political cliques which played a constant role in the politics of that age.

One view of government in premodern Japan holds that it was essentially feudal, however encrusted with laws and bureaucratic practices. If one follows this interpretation, then the transition to the more legalistic bureaucracy of the Meiji era becomes a sort of quantum jump. A second view argues that the late Tokugawa government was essentially bureaucratic, though with a great variety of lingering feudal survivals. This view permits a more evolutionary interpretation of the formation of the modern Japanese state. Yet even this second view must recognize that personal ties remained important after the opening of the modern era. When an early Meiji leader moved from one post to another, he took with him a number of his "men." By mid-Meiji this practice was discouraged as disruptive of order within the ministries of government. Yet as Roger Hackett has shown in his analysis of the "Yamagata machine," personal ties only slowly diminished in importance. Cliques were found in most ministries as well as in the services during the troubled years of the thirties, and, in some measure, still exist in most ministries today.

When a new class of university graduates enters a ministry, they do not join a clique immediately. Rather, they spend ten years or so in one or another post learning the work of the ministry. During this period they establish working relationships with their seniors. Seniors want able juniors. By the time a junior becomes a section chief, he will probably have established particularly good relationships with one or two senior officials. When the senior official becomes a bureau chief or vice-minister, he may recommend those juniors who are close to him for key positions. And after

the senior official has retired, if he should enter the Diet or the cabinet, or join a government commission, these relationships may become even more important. Perhaps the essence of the personal clique is that it creates an unofficial link between the bureaucracy and the world of politics.

Viewed in qualitative terms, what is most striking about the personal tie is the easiness of the relationship between junior and senior. This easiness sustains an unimpeded two-way flow of communications. This is a critical consideration for Japanese bureaucracy and contrasts with the situation in some other countries. Michel Crozier's study of French bureaucracy, for example, found that egalitarian sentiments were so strong in France as to produce "strata isolation" and a breakdown of communications between different levels in its bureaucratic hierarchy.^[14] In India the haughtiness or abruptness of senior officials often makes them unwilling to listen to what their juniors have to say—leading to a serious lack of feedback within the Indian bureaucracy. In Japan, however, the weakness of the sort of egalitarian sentiment that would require the self to hold those who are not equals at arm's length makes vertical communications less difficult. Language reflects the absence of barriers. Once their relationship is established, the senior will speak to the junior in familiar terms (calling him So-and-so kun ). This suggests a paternalistic warmth and kindly concern for the future of the junior, his training, and his welfare in general. The junior accepts this, is respectful, and works hard. At moments he may even show a kind of amae (active dependency) toward his superior. Of course, the emotional content in such relationships is not usually apparent. There are circumstances when it can be expressed, but most call for businesslike formality and impersonality.

Another related factor contributing to the ease of the vertical personal relationship in Japan is the relatively easy acceptance of status distinctions. When I have suggested this explanation to Japanese friends, they have usually disagreed with me, arguing that they resent status distinctions as much as anyone. Yet I feel that what they find unacceptable is not status per se, but certain kinds of behavior by persons in high-status positions. To put this in a comparative context, there is a well-known study in which a sociologist compared human relations on a British ship and an American ship. He found that the American ship had a more elaborate set of rules regulating the behavior of the ship's personnel. His explanation was that the American sense of egalitarianism is such that more rules are required to produce the order that the British obtained with less effort due to their ingrained sense of class. In terms of this comparison the Japanese are more like the British—though in Japan it is distinctions of status, rather than class, that are recognized and respected. A Japanese ship, I would guess, would require even fewer rules regarding the proper

[14] Michel Crozier, The Bureaucratic Phenomenon (Chicago: University of Chicago Press, 1964), pp. 220–224.

behavior due seniors than the British ship. At the same time the Japanese ship would probably go beyond either the American or British ship in drawing up a uniform system of status rankings. The use of court ranks gave all Meiji officials fixed positions in a unified hierarchy of statuses from the late 1860s. When court ranks were dropped, a standardized nomenclature for positions was substituted. When this was discarded during the Occupation period, officials frequently complained of not knowing the meaning of the names of positions on the meishi (name cards) of other officials.^[15]

Alongside the personal cliques within Japanese bureaucracy, overlapping them and often confused with them, is a second kind of grouping that is also called a clique in Japanese. I call these the official cliques. During the Meiji and Taisho eras there were the Satsuma and Choshu^[*] domain cliques (hanbatsu ), and from the middle of the Meiji era until the present there were school cliques (gakubatsu ), particularly that of the Law Faculty of Tokyo University.

The essential distinction that I would draw between the official clique and the personal clique is that while the official clique may contain networks of personal ties, it has an objective base external to the bureaucracy. The personal clique does not. The balanced recruitment of officials from Satsuma and Choshu was not solely a matter of balances between offices, but was part of a larger equilibrium on which the new state was built. It reflected the fact that these two domains had the balance of military strength in the country and controlled access to the emperor. Once in the government, Choshu officials usually joined personal cliques under other officials from the same domain. But they also recruited able men from other domains into their own personal cliques. Okuma^[*] , for example, was first attached to the personal clique of Kido of Choshu and then to that of Okubo^[*] of Satsuma. He had Choshu men in his own personal following even while he was a follower of Okubo. Even in the early Meiji years personal cliques were not congruent with official cliques.

The distinction was even clearer in the case of school cliques. Tokyo University was established as the training ground for the bureaucracy. The "Tokyo University clique" was always more the name for a system of preferential recruitment than it was a functional bureaucratic clique. The

[15] Imai, Kanryo^[*] , p. 65. Vertical relationships are not always as lovely (or "old-fashioned") as I have described them. The superior may be arrogant or cold. He may be a whipcracker, feared by his subordinates, but unable to evoke their loyalty. See Masuda, Oyakunin , pp. 38–44, for examples. A superior may refuse to warm up to a junior who went to the wrong school or who lacks ability. Conversely, it may be the modern junior who rejects the assumptions of paternalism on the part of the senior, especially when the paternalism is used as a thin cover for exploitation. The polite language of such a junior may be devoid of any human feeling, or may merely conceal his contempt. Masuda also discusses officials who are liked, but ineffective (ibid., pp. 45–61) or are too brilliant to be effective (ibid., p. 61). On the latter point, Noda Kazuo has commented that leadership in Japan requires warmth rather than sharpness.

cliques in the Foreign Ministry, which during the 1930s stood in opposition to one another regarding government policy, were all, basically, Tokyo University cliques. Even today the distinction between personal and official cliques continues. Of course, it goes without saying that "old boy" ties did not hurt. A graduate of the Tokyo University Law Faculty entering a ministry was a kohai^[*] (later graduate—junior) vis-à-vis senpai (former graduates—seniors) already established in the ministry. They had a common ground on which to build a relationship. There was a presumption that the junior would be talented, whereas an "outsider" would have to demonstrate his ability. If matters went poorly, a man not from Tokyo University might still feel himself an outsider ten years after entering the ministry.

Because of the historical resentment against the Sat-Cho^[*] cliques, because of the resentment against Tokyo University, because official cliques are confused with personal cliques and personal cliques are seen as running counter to the public interest, cliques are much talked about as an evil in Japan. Critics argue that cliques prevent the recognition of talent. They see the bureaucracy as a system in which preferential recruitment leads to preferential promotion and then to differential rewards. Others see in the ethic of personal loyalty a negative survival of Japan's feudal past. Matsumoto Seicho^[*] , a socialist and long-time critic of Japanese bureaucracy, could write a plausible detective novel about a government office in which the assistant section chief shoulders the blame to cover up a crime committed by his superiors.

Bureaucratic cliques, however, are not simply feudal survivals. They are tolerated because they have positive functions as well. One is to bypass an occasional blockage up-and-down the organization by forming a conduit for communications outside of channels. Another important function is lateral communication with a high level of trust. The coordination of work on a horizontal axis is often difficult within Japanese bureaucracy since vertical ties are so important that they always take precedence. But if the chief of one bureau was formerly the section chief in another bureau, he may well know its present section chief and will certainly know some of its permanent lower officials. This may enable him to reach them directly without going through the other bureau chief, especially if he is close to the present section chief. His effectiveness as a bureaucrat depends in some measure on the cultivation of such personal ties. (One of the arguments constantly heard within organizations in Japan against the recruitment of persons not fresh out of university is that they will not be able to function effectively since they will be strangers in the organization.)

Personal cliques are limited to the confines of a single ministry since officials are not transferred from one ministry to another. Consequently personal cliques are of no use in coordinating work between ministries. But personal friendships or the official clique are sometimes of use. If an

acquaintance from the same class at Tokyo University is in another ministry, a short telephone call may accomplish what would take days or weeks if sent up, over, and down, through channels. Phone calls, of course, are made constantly even when personal ties are lacking, but jurisdictional sensitivities are so acute that they are often ineffective. Thus, when a bureau in, say, the Ministry of Agriculture and Forestry has business with a bureau in the Ministry of Finance, the bureau chief may ask his subordinates if they know anyone in that bureau. Or officials in different ministries who did not know each other in school may establish a working relationship on the ground of having attended the same school. In a variety of ways, the informal system within the bureaucracy complements and is necessary to the formal system.^[16]

Solidarity and Jurisdictional Jealousies

A second major characteristic of Japanese social relations is work-group solidarity—the sense of common purpose, cohesion, and belonging that unites a section, bureau, or ministry. The dysfunction associated with this is a we-versus-them attitude toward other offices or ministries that often makes cooperation difficult. There are various terms in the literature which describe this; each involves a slightly different context and perspective. If the focus is on the powers of an office and the extent of its authority, then "the roped off area" (nawabari ) or "the limit of authority" (kengen ) may be used. If there is a struggle between two offices over a conflicting jurisdiction, then kengen arasoi may be used. If the autonomy of the ministries is under discussion, then the term is kakkyosei or "sectional independence," a term used in earlier times to refer to the independent power of regional feudal domains.

The keen awareness of jurisdictional limits is not new in Japan, and certainly not unique to Japan, though office solidarity may make it stronger in Japan than elsewhere. Even in the Edo period in the regulations of the bakufu exchequer (kanjo bugyo^[*] ), great care was taken to spell out the specific functions and the limits of competence of its various suboffices. An English engineer working for the Ministry of Public Works during the early Meiji period described the same scrupulous concern for jurisdictions—almost as if they resembled the boundaries of fiefs—in the following words: "The Vice-Minister of Public Works was himself in no

[16] Another aspect of the society of bureaucrats, if not of bureaucratic society, is marriage. The wives of bureaucrats like the wives of businessmen are not supposed to intrude into the world of the office, and in fact they usually do not. Still, everyone is aware of marriage ties. The foreigner in Japan is always struck by how small and interconnected the Japanese establishment is. Businessmen and bankers went to the same universities as the higher officials. The majority of students at these universities are born outside of the magic circle of power, but after graduating and entering an office they tend to marry the daughters of the elite of the previous generation. The establishment almost appears to perpetuate itself by matrilineal adoption.

way connected with the project, owing to the curious mutual jealousy of departments and consequent duplication of administrative arrangements constantly to be remarked in Japan. . . . "^[17] Even as the bureaucracy became more modern, these intraministry jealousies did not disappear. Before World War II an official in the Ministry of Finance was sent to another section of the same ministry to borrow some statistical materials, which that section had collected from all of the prefectures in Japan, but which it no longer was using. The request was turned down by the chief of the other section, whose last words were: "If the materials are so necessary, why doesn't your section notify the prefectures and collect them separately?" The conclusion of the official, albeit slightly overstated, was that the jurisdictions of Japanese offices are, like the person of the emperor, "sacred and inviolable" (shinsei ni shite okasu bekarazu ).^[18]

Between ministries, competition and jealousies are even more frequent. Ministries have varying levels of prestige, and the Finance Ministry is the most powerful and the most prestigious because of its authority to approve or disapprove of not only the general budgets of other ministries but also of their specific projects. Other ministries resent their dependence on it. When meetings must be held between officials of different ministries, the question of who should come to whom is exceedingly delicate. Also, the question of whose approval to get first is a touchy matter. If one office feels slighted or taken for granted, then it may make trouble. During World War II, Ayukawa Gisuke was put in charge of Manchurian heavy industry, taking it over with a "force that would stop a bird in flight." Yet when he came to Tokyo with his yearly plan, he had to repeat the same explanation over and over tens of times, visiting each of the ministries concerned. He also had to proceed in the proper order^[19] and call on them according to the convenience of their bureau chiefs. When busy, they did not give him an appointment on the hour and day requested. The entire process required more than a month. That so powerful and dynamic a figure as Ayukawa was subjected to such treatment in wartime, one author writes, makes one realize how staunchly the splendid bureaucratic tradition of "protecting one's territory" is upheld.^[20]

The exposé literature abounds with examples, such as those given above, of the dysfunctions inherent in jurisdictional jealousies. Yet even such prickly antagonism toward organizations other than one's own has a silver lining—the strong positive identification with one's own office and with its goals. When there is work to be done, Japanese officials work long hours; I have walked through the corridors of several ministries at six or seven in the evening and have been surprised at the number of officials still at their

[17] E. G. Holtham, Eight Years in Japan (London, 1883), p. 271.

[18] Imai, Kanryo^[*] , p. 142.

[19] Ministries were ranked in 1885, and those formed after that time were ranked in order of formation.

[20] Imai, Kanryo , pp. 161–162.

desks. They feel their work is important and are proud of it. They feel that they are a vital part of an enterprise. And their awareness of belonging to the little society of the office is reinforced, as Nakane Chie has described it, by drinking or playing mahjong together after hours, or by office outings to the sea or a hotspring once or twice a year.

How then is this office solidarity reconciled with cliques? The answer, first of all, is that the office in which junior officials, clerks, typists, and researchers work together—or even the section with much coming and going among several closely grouped rooms—is below the level of cliques. The section chief may be particularly close to one or another senior, but the personal clique stops with him. (The assistant section chief will also give a special kind of personal allegiance to his boss, but, because of his limited career mobility, he is not usually considered a clique member.) Among the higher posts in the ministry there is also a kind of solidarity or esprit de corps. Arising from the face-to-face contacts of the higher officials, their sense of mission, and their elite consciousness, this sense of solidarity is what binds the ministry together. Ministerial solidarity is always in a delicate balance with personal vertical ties. As the higher officials move about from post to post, they must work now with one man and then with another. Particular vertical ties are never permitted to become so strong that they interfere with work relations. To the extent such ties are seen as related to official cliques, they are criticized—as the intrusion of particularism where only merit should count. To the extent they are seen as personal cliques, they are more often interpreted positively as good senior-junior working relationships of the kind that contribute to the accomplishment of the ministry's task.

Decision-Making

Much of the writing on decision-making in Japan has focused on the defects of the ringi system—the procedure by which papers are passed up and down the line, from section to section and bureau to bureau, until they have gathered the required number of seals of approval. In the paragraphs that follow I would like to suggest that this emphasis is more an obstacle than an aid to our understanding. It makes Japanese office procedures mysteriously different from our own. Before turning to the internal workings of bureaucracy, however, it is necessary to touch briefly on the political context of Japanese bureaucracy.

Weber said that above every bureaucracy is a political system that sets the goals. This is clearly true in both Japan and the United States: the Diet and Congress pass laws; the respective administrations see that the laws are carried out. Yet there are significant procedural differences. In the United States bills may be drafted by congressmen (Taft-Hartley, Burke-Hartke) or by the president's White House staff, as well as in the various departments of the federal government; in Japan the overwhelm-

ing majority of bills are drafted by bureaucrats in the ministries. If the drafting of bills is a political function, then in Weber's terms, part of the Japanese political system is found within the higher bureaucracy; if it is not, then a part of the American bureaucracy is located within the political system, above the departments of government. It is possible to resolve the two alternatives by saying that the Japanese bureaucracy performs staff functions for the political leaders of a kind not done by department bureaucrats in the United States. But "staff functions" is a vague term. If it means the actual drafting of laws, then does it not betoken at least a partial role in the setting of goals? In any case, it is clear that the central ministries of state in Japan are considerably more powerful than the equivalent American departments.

A second major difference is the degree of insulation from politics. The American department is open at the bottom, open at the middle to men with experience, and at the top it is infiltrated by political appointees down to the assistant secretary level (the equivalent, perhaps, of the Japanese bureau chief). In contrast, the Japanese ministry is a closed body. It recruits only at the bottom and takes only virgin labor—fresh out of school and plastic. At the top of each ministry there are three nonbureaucratic posts: the minister who is in the cabinet and two "political vice-ministers" from the Diet. These are powerful figures, yet by and large these men deal with the ministry through the bureaucratic vice-minister. They do not penetrate the ministry; the ministry maintains its insulated corporate unity beneath them; it is not politicized.

One reason for the weak politicization of the bureaucracy was its prewar tradition of neutrality; the laws that framed it were designed to keep it separate from politics. But perhaps more important was the fact that throughout the postwar period the Japanese bureaucracy has had a single master, which it has been able to serve with a rare singleness of mind. Had power alternated between parties with different programs, it would have been necessary to develop mechanisms within the bureaucracy to handle the shift of policy. Whatever form these might have taken, the consequence inevitably would have been the politicization of the upper levels of officials in the ministries. Since this has not been necessary, the result has been a bureaucracy that is both conservative and apolitical and that has worked in harmony with a series of conservative administrations.^[21]

[21] The question of the influence of ideology on bureaucratic behavior, raised by Blau, is not easy to answer in the case of Japan. It is easier to prove homogeneity than conservatism, especially since the latter term has different meanings in Japan than in the United States. Imai (Kanryo^[*] , p. 96) writes that drafts are considered at innumerable kaigi and are revised over and over again, but the discussion of them never leaves a certain framework since everyone at the kaigi has the same education and background of experience. Kawamoto, ("Todai^[*] hogakubu^[*] ," p. 170) buttresses this contention with figures from Kubota Akira: 79 percent of Japan's postwar higher bureaucracy were from Tokyo University, 47.3 percent of England's from Oxford and Cambridge, 11.2 percent of America's from Harvard. Masuda (Oyakunin , pp. 153–154, 162–167), on the other hand, talks of the recruitment of radicalgraduates into the Ministry of Labor in the early postwar era, of his own position as a section chief of a reactionary government, and of how the radicals sobered up after some time in service.

Japanese bureaucrats would agree with the description of the bureaucracy as apolitical. They emphasize that Japanese bureaucracy is not racked by the dissensions of party strife. They see themselves as operating above politics in the national interest. Occasionally a bureaucrat will suggest that the factions within the Liberal Democratic Party (LDP) are the equivalents of parties in other countries. After a shuffling of cabinet portfolios, a new minister may put in "his man" as vice-minister or as bureau chief in a key bureau. Yet his man will be a career official whom he knows, who can work with those already in the ministry, and who has the seniority and experience needed for the appointment. Such an appointment usually has few implications for policy. And often dramatic changes occur in the cabinet and in national policy without any corresponding shifts of personnel within the ministries. When Tanaka became prime minister, for example, Japan's China policy turned completely about. However, hard-line officials in the Foreign Ministry with ties to the Fukuda clique did not lose their positions. They continued to handle China policy. But the policy they made and the speeches they gave in defense of it were markedly different in content.

Bureaucrats, however, protest the label of "conservative," arguing that they have a sphere of autonomy and do not look simply to the LDP. In drafting a bill, for example, they often check with the socialists as well as with the LDP. Officials also point out that, from time to time, the bureaucracy is successful in opposing the policy directives of LDP organs. Even the socialists, bureaucrats argue, recognize their fairness and their concern for the national interest. I feel that the preceding points are not untrue, but need some qualification. Token coordination with opposition parties is a way of helping the LDP avoid the charge of "tyranny of the majority" on the floor of the Diet. Small changes in a bill that is not controversial may even enable the government to obtain socialist support for the measure. The socialists' recognition of the fairness of the bureaucracy—to the extent that this is recognized—is a recognition that the bureaucracy is as national or as impartial as they could expect given LDP control of the government . Bureaucratic opposition to LDP policies is, in fact, very small. Consider the case cited by one official in which the Ministry of Agriculture and Forestry resisted the demands of the Agricultural and Forestry Division (Norinbukai^[*] ) of the LDP Policy Affairs Research Council (Seimu chosakai^[*] ) for higher prices for agricultural products. Such prices are a subsidy to farmers who vote conservative. The successful resistance by the ministry cannot be interpreted simply as bureaucratic independence of the LDP. It must be seen as a minor defeat for LDP farm interests in order to achieve a balance with growing LDP

urban interests. Too much pandering to the farmer, a declining percentage of the population, would weaken the LDP in the cities.

Qualifications such as these suggest that it is not unfair to apply the label of conservative to the higher bureaucracy. Most, though not all, of its members make no bones in private about their support for the LDP. They see the LDP as the party closest to the national interest, and they are willing and efficient in carrying out its policies.

Can one, then, not go a step further and argue that the bureaucracy has been permitted to maintain its substantial autonomy and powers because of its willingness to act for the LDP? This question can be examined in terms of where the initiative comes from when a bill is drafted. Is it from the bureaucracy, from the party by way of the divisions of the Policy Affairs Research Council, or from the commissions made up of businessmen, professors, Diet members, and retired bureaucrats? The answer varies. For a politically neutral matter—a law intended to aid the implementation of other laws—the initiative may come wholly from within the bureaucracy. For a political sensitive matter, such as the violence in the universities during the late sixties, the LDP will define the general sense of what is to be done, and the Central Educational Commission will provide specific policy guidelines. Since neither organization has an adequate staff, the Ministry of Education will draft the actual bill. In this situation the politicians feel that they are in control and are using the bureaucracy to handle the technical details. They are confident that the product will be what they desire. The bureaucrats feel that their input into party policy is greater than party influence on their drafts. Each side appears satisfied with the relationship. If there is delusion, it is on the side of the bureaucracy, which feels that its autonomy and powers are properly its own and the result of its ability.

Discussion of the decision-making process within the bureaucracy itself needs several prefatory remarks. First, most decisions in the bureaucracy are concerned with carrying out laws already passed. The actual enforcement or implementation of the provisions is often handled by prefectural governments. For example, agricultural extension services, which in the United States would be handled by local offices of the federal government, are in Japan "delegated" by the Ministry of Agriculture and Forestry to the prefectures. This was the Japanese prewar practice, and it continued even after the Occupation reforms. Central government control over prefectural expenditures permits a very tight rein on the way in which the delegated functions will be carried out. Decisions regarding the interpretation of laws, for example, are referred back to the concerned ministry.

Second, the majority of matters in Japanese organizations, as elsewhere, are settled by the single decision (senketsu ) of the appropriate official or office. This is, after all, what bureaucracy is about: the division of a complex task into simpler parts and the assignment of specific functions or

jobs to specific offices. Most decisions occur within parameters already defined. They clearly fall within the jurisdiction of a single office, and neither infringe upon the authority nor offend the sensibilities of other offices. It would be silly to refer such decisions to complicated procedures. At most the matter will be passed to the official's superior for an automatic stamp of approval.

Third, both in business and government in Japan the decision-making process is fairly formalized. There are tables of organization and instructions saying what office shall deal with what kinds of matters. There are charts or lists saying how far up in the organization different matters must be taken for final approval. At times this will be indicated by checks in the appropriate space at the top of a document. At times regulations even spell out what kind of decision-making process shall handle what kind of matter.^[22]

Let us, however, consider a complex decision, the draft of a new law that has political ramifications and affects the work of other ministries. Once initiated from above, it may go through the following stages:

1. Drafting the bill—done in the section (ka ) that has responsibility for the matter. It will involve informal discussions within the section, and may involve querying other offices about what would be acceptable to them.

2. Consultation and meetings—informal talks and formal meetings to revise the draft and to gather support and approval for it.

3. Circulation—passing the completed document up and down through the concerned offices for seals of approval by the office chiefs until the draft reaches the office of the vice-minister. This may overlap the previous stage.

4. Submission—giving the document to the minister for consideration by the cabinet or submission to the Diet.

In drafting the bill, the main difference between the United States and Japan has already been referred to: in Japan it is usually done in the bureaucracy, in the United States it may be done in a Senate committee or in the White House. Yet even when an American bill is drafted in a department, there is a second difference: the American bill may well be drafted by the staff of a high-level official, whereas in Japan the vice-minister will give the job to the bureau chief concerned, who will pass it to the appropriate section where younger career officers will do the drafting. The difference in the two procedures reflects a difference in

[22] At least in regard to formal procedures, there are striking similarities between government and company bureaucracies. Compare, for example, the corporation forms and rules for ringi contained in Nihon keiei seisaku gakkai, ed., Keiei shiryoshu^[*] taisei [Collected materials on management] (Tokyo, 1968), IV, with those for government offices given by Tsuji Kiyoaki in "Decision-Making in the Japanese Government: A Study of Ringisei ," in Robert Ward, ed., Political Development in Modern Japan (Princeton: Princeton University Press, 1968). Also see chap. 9 in M. Y. Yoshino, Japan's Managerial System (Cambridge, Mass.: MIT Press, 1968), for an interesting discussion of the varieties of decision-making in Japanese corporations.

organizational structure. Japanese bureaucracy is by and large a line organization. It is a vertical command structure of offices beneath offices beneath offices. It has few staff officials—that is to say, few officials who are attached laterally to high offices and who themselves have no subordinates.

Some coordination with other offices may take place even during the preparation of the original draft of a bill. If the measure involves two sections or two bureaus, the junior officials preparing the draft may sound out their opposites in the other concerned office. Most consultations with other offices, however, occur after the draft is completed and approved by the section chief. There is no predetermined order of consultation; this varies with the measure. Nor is it absolutely fixed which consultations will be formal and which informal. Informal consultations are sometimes called nemawashi . The primary reference of this term is to the preparations made in advance of transplanting a large tree: digging a trench around it, cutting the thick lateral roots, and bending the smaller roots circularly around the earth clump that will be moved. As applied to bureaucratic practice, it refers to the "spadework" involved in gathering together backing for a measure. The other form of consultation is the meeting (kaigi ). In preparing a measure for the Diet, the document may first be brought before the weekly meeting of section chiefs of the ministry. Next it may be cleared with the appropriate division of the LDP Policy Affairs Research Council. All drafts of laws must also be cleared by the legal bureau of the cabinet (Naikaku hoseikyoku^[*] ), which checks the compatibility of the proposed law with existing laws. One official from each ministry serves in this bureau; officials consider this a good assignment. If the bill involves money, it must be approved by the Ministry of Finance; this requires another meeting. At some point the draft may be considered at a meeting of the vice-minister and bureau chiefs of the concerned ministry, or a meeting of the vice-ministers of all of the ministries.

If there is a difference between the process of consultation in Japan and that of the United States, it is one of emphasis. In the United States the tendency is to call a meeting of the concerned offices and thrash out any differences in the open. Face-to-face disagreements are not avoided. Open confrontation is the rule rather than the exception. This is not to say, of course, that a good deal of "buttonholing" (the American version of nemawashi ) will not precede the meeting. In Japan, on the other hand, the tendency is to settle differences beforehand through informal talks. This avoids open confrontation. It avoids a situation in which one man wins and another, who has openly revealed his position, is marked as a loser when the decision goes against him. Organizational solidarity will not tolerate procedures that create winners and losers. At times the building of a consensus requires compromises that actually affect the decision reached. At times the compromises are merely token gestures to placate someone who otherwise would be seen as a loser and make it possible for him to join

in a nominal consensus. In Japan as elsewhere there are many either-or situations. Consensus cannot miraculously do away with hard decisions. Whoever wins acceptance for his policy and gains the authority to implement the new policy is left in a strong position. In this respect Japan is not as different from other countries as some writers have suggested.

Because of the importance of informal discussions, and because open confrontation is usually avoided at meetings, the function of the meeting can be reduced to the formal approval of a settlement already reached during informal talks. In this connection Imai Kazuo tells the story of a Tokyo middle-school baseball team, which, each inning before taking up its position on the field, would huddle at the pitcher's mound, put their arms on each other's shoulders, and listen to words from their captain. A visiting team from the countryside was lost in admiration at this sophisticated custom and decided to imitate it. Before the last inning they formed a circle around the mound. But the captain of the team, having no strategic message to impart, merely said: "Let's all eat watermelon after the game." Imai continues: "I will probably be criticized for comparing the kaigi (meeting) in governmental offices with this watermelon kaigi . Yet such kaigi are not altogether lacking."^[23]

Meetings of bureau chiefs or section chiefs are held with such frequency that many may lack issues to discuss, or they may deal chiefly with matters of jurisdictions. The potential for a drift toward formalism in any routine procedure is always strong. Yet even Imai's criticism stopped at "not altogether lacking." I would stress, based on conversations with officials, the centrality of the meeting in the decision-making process. Discussions at meetings are often substantial. Drafts are revised and rerevised. At times an office opposed to a measure will openly denounce it at the start of a meeting as a tactic to force others to compromise—since to override his objection would make him a loser. (Of course, such a tactic is used infrequently. Potential conflict situations are usually resolved beforehand.)

Nakane Chie has written that Japan is a land of interminable meetings, dragged out in the name of democracy—which in Japan really means harmony.^[24] This may be true of faculty meetings in Japanese universities where ideological differences make informal agreements difficult. It is also true of Diet sessions where face-to-face confrontation is institutionalized and the opposition can use the value of harmony against the "tyranny of the majority." But it does not characterize meetings within the bureaucracy where differences are concerned with means, not ends.

There is also a difference between Japan and the United States in regard to stage three, the circulation of documents. In the United States the decision taken at the meeting, or after a series of consultations and

[23] Imai, Kanryo^[*] , pp. 105–106.

[24] Chie Nakane, Japanese Society (Berkeley and Los Angeles: University of California Press, 1970), p. 145.

meetings, is final. All that remains is to inform other concerned persons of the outcome. But in Japan, though agreement may be reached at the meeting, the final formal approval occurs when the document is passed from office to office and stamped with the seals of the responsible officials. Some writings have gone so far as to argue that this paper-passing process (the ringi system) is the decision-making process in Japan. The reason for this undue emphasis on one part of the total process is that, while informal consultation is hard to see and meetings so ordinary that they are hardly worth mentioning, the ringi process is highly visible and peculiarly Japanese. It is also because, for unimportant matters , the draft may be circulated for approval without prior consultation or meetings. But for significant matters, the primary function of the ringi system is to circulate information about decisions already taken. As such, it strongly resembles H. A. Simon's "referral and clearance system."

Viewing then the entire process of decision-making, and bearing in mind that reality is more variable than the paradigm offered above, one can say that the functions that are handled by the "meeting" in the United States are in Japan distributed over the informal consultation, the formal meeting, and to a small extent the ringi system. These are differences of emphasis arising from differences in bureaucratic organization and social behavior, but overall the process is less dissimilar than it is usually made out to be.

Strengths and Weaknesses in the Decision-Making Process

The Japanese pattern of line organization has one obvious advantage over the American line-staff pattern: young officials are given important work and vital experience almost from the time they enter the ministry. In contrast, the American college graduate who has just entered the bureaucracy may have to wait ten or fifteen years before he is promoted to a level where meaningful work is done. Japanese officials take for granted the correctness of their procedures. When queried, senior officials reply that the young are energetic and innovative while older officials have the wisdom needed to tone down the final product. This attitude is not new in Japan. In On Serving One's Lord (Jikun teiko^[*] ) written in the 1730s, Kani Yosai^[*] , discussing the division of labor between lower and upper officials, argued that "rough drafts (aragoshirae ) should be made by lower officials and then polished by higher officials." Ishii Shiro^[*] commented that this is what is meant today by giving young officials "something to live for" (ikigai o ataeru ).^[25]

[25] Professor Ishii's comment occurred during a discussion of the passage quoted. Kani Yosai, "Jikun teiko" [On Serving the Lord], Nihon keizai taiten [Compendium of Japanese economic writings] (Tokyo, 1928), XIV, 189.

The exposé literature on Japanese bureaucracy, however, often views the same phenomenon in a much less favorable light. It points out the undue influence of young or lower officials on policy decisions as one of the major dysfunctions of the system. Such an influence is termed "politics by subordinate officials" (zokuryo^[*] seiji or zokkan seiji ). In an extreme instance, even the word used to describe the overthrow of lords by their vassals (gekokujo^[*] ) in feudal times has been applied to it. As evidence of this phenomenon, two kinds of examples are cited.

The first is the influence exercised by subordinate officers on the decisions that led to the Pacific War. In approaching this question one should note first that the sections concerned—such as the Operations Sections of the Army General Staff, the Military Affairs Section of the Army Ministry, the First Section of the Naval Affairs Bureau—were the key sections in the services. They contained the elite of the service elites, men comparable in every way (except breadth of vision) with the higher bureaucracy of the nonservice ministries. It was their job to draft policy and strategy for the services; they did not usurp this function. Second, the "subordinate officers" were not that young. Lieutenants and captains in the sections were young, but the majors and colonels who served as section chiefs were in their middle or late thirties. Third, many of these men had been military attachés in Germany and favored a strong pro-German policy. This operated as an ideological factor setting them off to some extent from the senior officers. This degree of ideological difference is unusual in Japanese bureaucracy. Even in these anomalous circumstances the middle-level officers could make their influence felt only because the generals and admirals themselves were divided on questions of policy, and because of a political situation in which strategic decisions displaced political decisions as generals and admirals replaced political leaders in the cabinet.

A recent visitor to the United States who had been a junior officer (not a section chief) in the Japanese Supreme Command from 1939 to 1945 was asked whether he had encountered the phenomenon of the "inferior overpowering the superior." He appeared not to understand the question. The next day he was asked who drafted the plans for the General Staff, and he replied that it was done by the younger officers. He was then asked whether the generals always agreed with the plans. He replied that sometimes there were difficulties, but he was unwilling to use the label of gekokujo . He described a process of testing, discussion, and compromise, with the superiors capable at any one point, if not at all points, of having their own way against the proposals of their junior staff. He also stated that disagreements were the exception and not the rule.

In short, the kind of influence that "juniors" had during the late thirties came from line organization (even the General Staff had a kind of line organization) in which policy was drafted in the section. But this type of organization could have the baleful consequences it had only because of a

combination of exceptional circumstances. And even then, the influence of juniors should not be overestimated.

A second kind of example cited to illustrate the power of lower officials often centers on the figure of the assistant section chief (kacho^[*] hosa ). This official may be a careerist on his way up, but is more likely to be a clerk who has not passed the higher examination. He may have a fairly long tenure in the post while careerist section chiefs come and go. Masuda Yoneji describes an incident which occurred during his years in the Ministry of Labor:

After two or three years as a section head I gradually became accustomed to the life of an official and learned the mores of the office. So I began, according to my own lights, to make small changes on documents to which I had previously given a blind seal. At this the administrative assistant came in bearing the sets of ledgers pertaining to drafts of documents, overtime, etc., and said: "Section chief. From now on please handle all of these yourself. I can no longer bear the responsibility for handling the papers of the section." With these words he piled up a mountainous stack of ledgers on my desk. I was dumbfounded by this sudden coup d'état on the part of this man on whom I had so relied. "Now, now there, don't say that. Won't you please take care of them," I replied, making a total, unconditional surrender. Afterwards I again fell into the pattern of putting a blind seal on whatever documents he gave me. With that his loyalty to me became even stronger than it had been before.^[26]

Masuda had earlier described how his assistant covered up for him when he was absent from the office on a late social lunch, protecting him from the anger of the bureau chief. Masuda also described his assistant, an older, prewar type of official, as "sitting upright in his chair" and "running the various affairs of the office with the authority and dignity of an elder (karo^[*] )."^[27] Masuda continued:

Of course what must be taken into consideration is that from this time I became unable to examine carefully all of the documents that came through the office, which was my primary responsibility in the statistical investigation section. I came to understand from my later experience that this sort of maneuver was, for good or bad, a special characteristic of the lower officials (zokkan ).^[28]

Another example of the same phenomenon was the case of "Emperor So-and-so," a permanent lower official in MITI, who was said to have the power to block a request, even when the minister favored it.

I would not deny the existence of the kind of subordinate power that these examples illustrate. But I would argue that it is office power, not policy-planning power. It is the kind of power held in the United States by

[26] Masuda, Oyakunin , pp. 29–30. Imai (Kanryo^[*] , pp. 220–221) tells a similar story of a confrontation between a brilliant and experienced minister, who had reduced his ministry to abject submission, and a clever bureau chief, who got the best of him with a variation of the same stratagem.

[27] Masuda, Oyakunin , p. 26.

[28] Ibid., p. 30.

certain master sergeants on army bases, by certain lower officials in government offices, or by long-term secretaries in university department offices. It is based on a strategic position in the organization and a technical mastery of official forms and the petty details of administrative routine that higher officials do not want to be bothered with. When such a person retires, higher officials are momentarily pleased at the prospect of regaining control of the nether reaches of their own organization, and then they immediately set about finding the same type of person as a replacement.

Thus my conclusions are as follows. First, there is nothing uniquely Japanese about office power and not too much can be made of it. Second, young career officials in Japan do have a special kind of influence. Only a handful of such men are admitted to a given ministry each year.^[29] The key sections in a ministry are few in number. From the time they enter, they are apprentice policy-makers, and the system intends that they function as such. They draft bills. There is a saying often quoted by Japanese officials that "the original draft is in a 70 percent position of strength" (gen'an wa shichi bu no tsuyomi ).^[30] This is an index of their influence. But this is far from saying that Japanese bureaucracy is run from below. If one asks junior officials whether they dictate policy, they laugh. The opposite contention, that superiors are sometimes tyrannical and that juniors show a slavish deference toward them (jidai konjo^[*] ) is also heard. Ex-officials who served in the prewar years are particularly fond of describing office martinets—almost as if they were mourning the passing of the type: the section chief who would return a draft without saying what was wrong with it until his underlings produced one that met with his approval, the bureau chief so fearful that his subordinates wet their pants and came down sick when called on the carpet before him,^[31] or the section chief who made small meaningless changes in a draft so as not to incur the displeasure of a bureau chief known to dislike an automatic seal of approval.^[32]

A second major criticism of Japanese decision-making in the exposé literature is that the ringi system has several built-in weaknesses. First, the flow of documents through offices is so great that office chiefs must blindly

[29] In the prewar years about 2,000 took the higher examination each year; only about 250 were successful. The system was reformed during the Occupation. To become an official one must now pass the kokka komuin^[*] saiyo^[*] shiken . This has three levels for graduates of high school, junior college, and university. It is administered by the Jinjiin (National Personnel Authority). I have been told by officials that the university level exam, the jokyu^[*] shiken , has many more takers than in the prewar era, but that the number accepted by the ministries is only slightly larger. See Kancho^[*] monogatari , p. 126, for a description of the new examination system. See Robert M. Spaulding, Imperial Japan's Higher Civil Service Examinations (Princeton: Princeton University Press, 1967), for a detailed analysis of the prewar system.

[30] This is quoted by Dietmember Kawasaki in the course of a discussion with Tsuji, Imai, Sumimoto, and Ukai, printed in Jinji gyosei^[*] [Personnel Administration], 4.10 (September 1953).

[31] Masuda, Oyakunin , p. 27.

[32] Imai, Kanryo^[*] , pp. 100–101.

put their seals on papers that they have not read and do not know the contents of. Second, the time required for the papers to pass across the desks of the concerned office chiefs is so great and the potential capacity of any official to sabotage or critically delay the decision by letting the document sit on his desk is such that the system is slow, clumsy, and inefficient. Third, since a measure is drafted at one level, approved at another, and signed at all levels in between, the locus of responsibility is vague.

No one disagrees with the charge that the volume of paper flowing through offices is excessive. Imai writes that as the head of a small local government office he once affixed his seal eight thousand times in a single day, that this was the record for the year, though he was not trying to set a record, and that it took all of his energies to keep on hitting the right spot.^[33] Judging from the writings of ex-bureaucrats, there is even a kind of mystique built up above the seal. One official bragged that "a person's character is revealed in a single stamp of the seal" (hanko hitotsu ni mo jinkaku ga deru ).^[34] Another official who forgot his seal two days running was scolded by his chief: "The seal of an official is like the rifle of a soldier. Is there a soldier who would go to the battlefield two days in a row without his rifle."^[35] On the other hand, most officials I have talked to say that the problem of the "blind seal" is exaggerated. A great volume of paper flows over the average desk. Most of it gets a blind seal. Sometimes an assistant does the actual stamping. Mistakes do occur. But important papers are picked out and read.

Regarding the problem of delays, the Japanese literature stresses the capacity of an official, who was not invited to the meeting where a draft was approved or who feels that the authority of his office has been slighted, to pigeonhole a document during the circulation (ringi ) process. Such an official can hold up (nigiri komu ) the draft bill by saying, "let me research it a bit" or "let me think about it some more," and when pressed can reply, "just now I'm very busy" or "a little more time." Imai compares the affixing of the seal to the procedure at a customs barrier or the authority to issue a visa: the greater the number of barriers controlled, "the greater the opportunities to inflict maliciously one's power on (ijiwaru ) others."^[36] My impression, however, is that the slowness of Japanese decision-making (if it is slower than in other countries) arises from the need for a measure of consensus in the process of consultation and meetings rather than from passive resistance during the circulation of the document. If an important matter has been resolved, if it has had a settlement (matome ) at each level

[33] Ibid., pp. 127–128. Can we believe one stamp of the seal every 3.6 seconds with no time off for tea?

[34] Kaiko^[*] Ken, "Aru tocho^[*] shokuin no ichi nichi" [One Day in the Life of a City Official], Shukan^[*] asahi [The weekly Asahi], September 16, 1964, p. 32.

[35] Imai, Kanryo^[*] , p. 126.

[36] Ibid., pp. 125–126.

up the hierarchy, it is extremely difficult to block. The necessary seals can be obtained by orders from above, by energizing the personal clique, or the responsible official can hand-carry the document around from desk to desk, in each case bowing and waiting for the official to affix his seal. Few officials can hold out under such pressure. Even Imai suggests that on occasions an obstinate official can be "bowled over with adulation" (ogami taosareta ).^[37]

The problem of responsibility or the lack of it in Japanese bureaucracy is more complex. Only a portion of the problem is involved with the ringi system. Imai writes that no one talks about responsibility as much as the bureaucrat. It is a favorite word, yet what he really means by it is the power of his office, not his duty.^[38] Certainly in Japan the value of work does not lie solely in the work itself. Rather, work validates one's membership in the work group, work justifies office solidarity. Since any bureaucracy is hierarchical, responsibility is heaviest at the top. But in Japan the tendency to vest all responsibility in the head of the organization is carried to an extreme that appears ridiculous in Western eyes (and in some Japanese eyes as well). If a train crashes and lives are lost, be it sabotage or the negligence of a switchman, the director of the national railways may offer his resignation. When Ambassador Reischauer was stabbed, the director-general of Public Safety offered to resign, though this was an attempt at what the Japanese term an "instant assassination" for which the director general had not the faintest responsibility by Western standards. In part, this notion of responsibility was influenced by German legal concepts adopted during the Meiji period, according to which total responsibility was vested in each minister. But it also reflects the fact that organizational solidarity in Japan is symbolized by loyalty to its head. To justify this loyalty and maintain solidarity, the head assumes responsibility for incidents highly visible to the outside society.

The same sweeping responsibility was accepted by Tanaka Kakuei when he was appointed to head the Ministry of Finance. His words to the assembled bureaucrats of that ministry were: "You are the elite of the elite. No one can match your intelligence. Accordingly I will leave the thinking to you. And I, Kakuei, am ready to bear full responsibility for the consequences of your thought."^[39] One of the Meiji leaders used almost exactly the same words when he took over a ministry during the 1870s. Tanaka's speech was calculated to elicit loyalty and hard work. Ex-officials often criticize the don't-rock-the-boat (kotonakare shugi ) mentality of Japanese bureaucracy. Tanaka was telling his ministry that no one would be punished for new initiatives and creativity. Perhaps he was also telling them that, though not a graduate-bureaucrat himself, he knew how to handle bureaucrats.

[37] Ibid., p. 150.

[38] Ibid., pp. 111–113.

[39] Kawamoto, "Todai^[*] hogakubu^[*] ," p. 180.

Within the ministry the same solidarity is reflected in the diffuseness or collective nature of responsibility. In the exposé literature the problem is viewed as a consequence of the separation of work and authority. The lower official feels only a limited responsibility since he is merely submitting a draft. Higher officials feel only a nominal responsibility since they receive a completed document already bearing a number of seals of "approval."

I do not disagree with the description, but it needs several qualifications. First, this problem is not unknown elsewhere; the United States, too, has the proverbial official who never signed what he wrote, or wrote what he signed. Japanese line organization merely increases the magnitude of the problem. Second, though some officials assert that responsibility is vague and collective responsibility is no responsibility, others protest that responsibility is not so vague. They contend that whoever may prepare the draft, it is the section chief who bears the primary responsibility and his bureau chief who takes a share of it. Seals by other offices only peripherally concerned with the matter mean the document has been seen and not objected to. Third, to the extent that responsibility is collective, it rests to that extent on the consensus formed during consultations and meetings. The final formal document with its collection of seals may symbolize the consensus, but it is subsequent to it. As long as consensus is seen as necessary, even doing away with the circulation of documents would not produce a different kind of decision-making.

Conclusion

It is not easy to evaluate the performance of a bureaucracy. In the case of a private company, the profits from the sale of its product provide a yardstick by which to measure the efficiency of its parts. But government bureaucracy has no production schedules, no sales, and no profits, and it is difficult even to define its product. Japan's postwar goal was economic growth. The bureaucracy effectively translated this goal into policies highly supportive toward business. And perhaps just as important for growth was that the bureaucracy interfered very little in the decentralized decision-making of private companies. Imai writes that officials know little of the real world and for them to draw up laws is "like a tailor making a suit without taking measurements."^[40] The combination of supportive policies and noninterference created a maximal environment for growth. But a by-product of the policy was instances of chemical poisoning without a parallel in the world.

It is difficult, too, to compare one bureaucracy with another for they vary in their organizations, functions, and positions in their polities. It requires a formidable exercise of imagination to envision a Japanese

[40] Imai, Kanryo^[*] , p. 111.

Henry Kissinger usurping the policy-making functions of the Foreign Ministry. Japanese diplomats smile and dismiss the possibility. Yet the powers and the insulated, self-contained nature of the ministries in Japan, which Japanese officials take for granted, may have more to do with the Japanese pattern of one-party government than with the legal position of the bureaucracy or with its intrinsic quality. If Japan's parliamentary system continues in good health, and if parties or coalitions of parties should begin to alternate in power, then the present bureaucracy could change quite suddenly. A two-platoon system could develop within the top ranks of the central ministries, or the parties could develop within their organizations the equivalent of White House staffs as a countervailing force against the know-how of the ministries. The embryos of such staffs already exist. If such changes occurred, the bureaucracy might be much less efficient since it would no longer be carrying out policies that it had a hand in drafting. In any case, it is clear that much of the autonomy of the existing bureaucracy—which present-day Japanese officials see as an eternal verity—is in fact precariously dependent on political factors external to the bureaucracy itself.

Turning from the external situation to the internal workings of the bureaucracy, I would argue for a mixed analysis of the formal system and the informal patterns found in the larger society. The solidarity of the office, for example, forms almost spontaneously in response to the psychological need to belong to a warm, interdependent social group. It is a need that exists prior to the bureaucracy. The office group resembles the solidarity of nonbureaucratic groups in Japan. It arises wholly apart from the official purposes of the office organization. The particular character of vertical ties among Japanese bureaucrats can also be explained in terms of patterns of human relationships learned before they became officials. Yet once brought into the office, these general social proclivities—which can assume a variety of forms—are molded onto the formal structure and are adapted to serve the official ends of the organization. The personal clique becomes a vertical action group. The office community becomes a work team. The energies of the informal system are thus harnessed to the service of the official system.

Harnessing such energies, however, does not mean that they have been wholly converted into a positive force. The favoritism and discrimination associated with the personal clique remain dysfunctional. The office community uses the jurisdictional powers it is given by the official system to reinforce its natural animosities toward outgroups, producing frictions that interfere with the work of the organization as a whole.

This suggests that all bureaucracies are necessarily imperfect since the requirements of perfection are mutually contradictory. If vertical communication is easy, then horizontal communication is likely to be difficult. The ease of vertical relationships in Japan, in contrast to France or India, is a plus factor; yet it is just this ease that nurtures the personal ties that

lead to the formation of cliques. To eliminate cliques would require a change in the nature of personal ties, which would doubtless lead to a new set of dysfunctions. Solidarity is good, but poor cooperation between offices is not. Rules are necessary if the workings of an office are to be rational and predictable, yet to the extent an office is bound by rules there is the opposite danger of formalism. Men of ability are necessary for good government, yet if men of great talents are recruited, elitism is hard to avoid.

I do not mean to imply that dysfunctions should be welcomed. Some dysfunctions such as corruption have an entirely negative input. In many countries the dysfunctions of bureaucracy are so severe that the tasks set by political leaders are not accomplished. But I would suggest that particular dysfunctions should be judged in the larger context of the end results produced by the bureaucracy. If the bureaucracy gets the job done, then some entropy in the system can be tolerated. Japanese bureaucracy is not the model of efficiency that some who talk of Japan, Incorporated, would have, nor is it the neutral body pursuing national interests that the bureaucrats themselves often describe. But neither is it the jumble of dysfunctions portrayed in the writings of its critics. Japanese officials work long hours when necessary; in terms of any comparative standard they are honest; and they accomplish most of the tasks set for them with a reasonable competence.