Chapter Two
Japanese Bureaucracy During the Prewar Era

To a striking degree, the characteristics of Japanese government bureaucracy were formed in the prewar period. A brief survey of its historical background will, therefore, enable us to appreciate the extent of continuity and change between the prewar and postwar patterns of Japanese bureaucracy. We shall dwell upon the circumstances surrounding the emergence of modern bureaucracy in the Meiji era, its power and prestige, and the outstanding aspects of its structure and behavior, notably, stratification, elitism, legalism, security of tenure, and retirement practices.

Emergence

If by bureaucracy we mean a pattern of organizing and managing human affairs on the basis, inter alia, of expertise demonstrated by education or experience, bureaucracy did not emerge in Japan until after the Meiji Restoration in the latter half of the nineteenth century.^[1] To be

[1] Mori Hiroshi and Yazawa Shujiro, Kanryosei no shihai [Rule by Bureaucracy] (Tokyo: Yuhikaku, 1981), p. 162; Oe Shinobu, "Nihon ni okeru kanryosei no kiseki" [The Locus of the Bureaucratic System in Japan], Hogaku seminaà zokan, sogo tokushu shirizu, 9: Naikaku to kanryo [Legal Studies Seminar Extra Issue, Comprehensive Special Series, 9: The Cabinet and Bureaucrats], Mar. 1979, p. 105.

sure, Japan did borrow the idea of civil-service examinations from T'ang China in the seventh century, but, unlike the situation in China, where it flourished, the idea was never fully implemented.^[2]

During the Tokugawa period (1603-1868), structural arrangements for administration that existed both in the central government and in the fiefs were patrimonial rather than bureaucratic. The criterion of recruitment to key offices was primarily ascriptive, such as membership in the shogun's immediate vassalage in Edo and the possession of a specified feudal family rank in the fiefs. In its waning days, however, the Tokugawa shogunate was compelled by the crisis precipitated by the arrival of Commodore Matthew C. Perry's four "black ships" to recruit officials on the basis of "demonstrable merit." This was done, it should be stressed, only within the elitist boundaries. Nonetheless, the recruitment of "men of relatively low social status" to important positions in Edo and, particularly, the "appointment of subjects of individual fiefs to offices in the shogunal government" signaled the beginning of the end of the traditional feudal hierarchy.^[3]

If the Meiji Restoration opened the way for a fundamental restructuring of Japan's political system, it did not usher in a modern bureaucratic system immediately. On the contrary, the first two decades of the Meiji era saw the implementation of a spoils system: key government positions were doled out to those who played the leading role in the Restoration—the lower samurai from the fiefs of Satsuma, Choshu, Tosa, and Hizen as well as court nobles who had collaborated. An interesting aspect of the spoils system pertained to its use as a device for coopting the opponents of the new regime. Some leaders of the

[2] Spaulding, Imperial Japan's Higher Civil Service Examinations , pp. 9-19. For a study that underscores the early development of bureaucracy in Japan, arguing that Japanese bureaucracy had reached maturity as early as the late seventh and early eighth centuries, see Nomura Tadao, Nihon kanryo no genzo [The Original Portrait of Japanese Bureaucrats] (Tokyo: PHP Kenkyujo, 1983). T. J. Pempel writes: "Bureaucratic strands wind through Japan's history from at least the Nara Period (710-84), but the establishment of a modern bureaucracy dates from the opening of Japan by the West in the middle of the nineteenth century." See his "Organizing for Efficiency," p. 78.

[3] Masamichi Inoki, "The Civil Bureaucracy" in Robert E. Ward and Dankwart A. Rustow, eds., Political Modernization in Japan and Turkey (Princeton: Princeton University Press, 1964), pp. 283-87. See also Bernard S. Silverman, "The Bureaucracy and Economic Development in Japan," Asian Survey 5, no. 11 (Nov. 1965), pp. 529-37. Even Nomura's study shows that ascriptive criteria eclipsed those of achievement in the recruitment and promotion of officials in the eighth and ninth centuries. Nomura, Nihon kanryo no genzo , esp. pp. 63-91.

opposition clamoring for "freedom and civil rights" (jiyu minken ) were coopted into the government.^[4]

As the forces of democracy grew in strength, however, the Meiji oligarchs were compelled to make concessions, including a commitment to establish a parliament by the year 1890. The need to cope with opposition politicians and to guard against the possibility of the abuse of "free appointment" privileges by party politicians upon winning power provided the oligarchs with a strong incentive to institutionalize the merit principle in the recruitment of officials. The most compelling pressure of all, however, emanated from the outside. The revision of the humiliating unequal treaties with Western powers embodying the principle of extraterritoriality necessitated the establishment of a modern judicial system. This in turn called for the enactment of a constitution as well as administrative, civil, and criminal laws along Western lines. All this would also require the adoption of a credible system of recruiting officials to man the judicial apparatus.^[5]

To fashion a modern system of government equipped to handle all these problems, the Meiji oligarchs not only invited Western experts to Japan but also dispatched delegations to Europe on extended learning tours. Those who went abroad on such missions, of whom Ito Hirobumi was the most prominent, appear to have been deeply impressed by what they saw and heard in Berlin and Vienna. From such Prussian and Austrian mentors as Rudolf von Gneist, Albert Mosse, and Lorenz von Stein they learned the rudimentary principles of limited constitutionalism, parliamentary government, and a civil-service system. Regarding the latter, two things struck the Japanese most: the principle of imperial prerogative in appointment and the principle of recruitment based on educational attainments. Those with the proper

[4] Wada Zen'ichi, "Bunkan nin'yo seido no rekishi" [The History of the Appointment System for Civil Officials], I, Jinji-in geppo [Monthly Bulletin of the National Personnel Authority] 95(Jan. 1959): 10. According to data compiled by proponents of jiyu minken in 1874, 65.7 percent of chokunin officials (bureau chiefs and above) and 37.6 percent of sonin officials came from the fiefs of Satsuma, Choshu, Tosa, and Hizen. Satsuma and Choshu together accounted for 44.8 percent of the chokunin and 27.8 percent of the sonin officials. Fukumoto Kunio, Kanryo [Bureaucrats] (Tokyo: Kobundo, 1959), pp. 81-82. For definitions of chokunin and sonin officials, see table 1 and the accompanying text later in this chapter.

[5] Spaulding, Imperial Japan's Higher Civil Service Examinations , esp. pp. 4-5, 34. For a Marxian analysis of the relationship between the jiyu minken movement and bureaucracy, see Yamanaka Einosuke, Nihon kindai kokka no keisei to kanryosei [The Bureaucracy and the Formation of the Modern Japanese State] (Tokyo: Kobundo, 1974).

educational credentials, the Japanese were told, must first pass appropriate examinations before being appointed.^[6]

The trials, tribulations, and politics that accompanied the adoption of coherent civil-service examinations for administrative, judicial, and diplomatic personnel need not detain us here, for they have already been analyzed in Spaulding's definitive study. What needs to be stressed, however, is that although the German model clearly exerted major influence on the thinking of Japanese leaders, the system they ultimately adopted was far from a carbon copy of the Prussian prototype. In terms of the sequence of the key steps involved, the first significant step was the establishment of law schools in the 1870s—the French law course in the Justice Ministry and the Anglo-American law course in the Education Ministry. Next came the establishment of the bar examination in 1876, followed by the adoption of the Judicial Appointment Rule in 1884.^[7]

Then, in July 1887, the first general-examination ordinance was promulgated, extending examinations from the judiciary to all parts of the government. The new system set up two levels of civil-service examinations: higher examinations for sonin officials and ordinary ones for hannin officials. It retained, however, the privileged status of Tokyo Imperial University (Todai ) graduates. Just as they were exempt from the bar and judicial examinations, so they would continue to be exempt from the higher civil-service examinations. Although the exemption under the 1887 ordinance would be extended to Todai graduates in law and letters only, that in no way diminished the privileged status of all Todai graduates. Todai men in the other fields, such as agriculture, engineering, and medicine, would be eligible for appointment to technical posts in government without taking examinations along with all other technical personnel. The requirement of a three-year training period following initial appointment was an echo of the German model. In the revamping of the examination system in 1893, however, the Todai exemption was abolished in the administrative examinations, though it was retained in the judicial. In the same year, separate examinations for diplomatic and consular personnel were established. Then, during the first two decades of the twentieth century a struggle was waged to ensure legal equality among all

[6] Spaulding, Imperial Japan , pp. 46-50.

[7] Ibid., pp. 33-72.

examination candidates, culminating in the abolition of exemptions in toto; a unification of examination systems was also accomplished, bringing under the single umbrella of higher examinations the three disparate fields of law, administration, and diplomacy.^[8]

Power And Prestige

According to the distinguished Japanese political scientist Royama Masamichi, the Meiji Restoration helped to transform Japan from a "feudal police state" (hokenteki keisatsu kokka ) into a "centralized bureaucratic state" (chuo shukenteki kanryo kokka ). Although the Meiji government enshrined the doctrine of imperial sovereignty, he writes, it was in reality controlled by a coalition of hanbatsu (domain cliques) and court nobles. More important, it replaced the feudal warrior system with its twin functional equivalents, the military and the bureaucracy.^[9]

Bureaucracy, in other words, was part of the ruling elite in prewar Japan. Its power and prestige were greatly augmented by the explicit linking of bureaucratic authority and imperial prerogatives. Article 10 of the Meiji constitution of 1889 stipulated that "the Emperor determines the organization of the different branches of the administration and salaries of all civil and military officers, and appoints and dismisses the same."^[10] This implied that officials were appointed by and accountable to the Emperor only; hence they were the Emperor's officials (tenno no kanri ) rather than civil servants. So long as the prestige of the Emperor remained supreme and the myth of imperial rule was kept alive, therefore, his officials could exercise virtually unfettered authority in his name.

The position of the bureaucracy vis-à-vis political parties and the Imperial Diet was bolstered by the constitution and practice alike. Pursuant to the advice of their Prussian and Austrian mentors, the Meiji oligarchs had taken pains to establish the framework of the bureaucratic

[8] Ibid., pp. 73-178; Watanabe Yasuo, "Nihon no komuinsei" [The Civil Service System of Japan] in Tsuji Kiyoaki, ed., Gyoseigaku koza, dai 2-kan: Gyosei no rekishi [Lectures on Public Administration, vol. 2: The History of Public Administration] (Tokyo: Tokyo Daigaku Shuppankai, 1978), pp. 114-16.

[9] Royama Masamichi, "Kindai kanri seido no hattatsu" [The Development of a Modern Bureaucratic System] in Royama Masamichi, Gyoseigaku kenkyu ronbunshu [Collection of Research Papers on Public Administration] (Tokyo: Keiso Shobo, 1965), p. 229.

[10] Spaulding, Imperial Japan , p. 83; Wada, "Bunkan nin'yo seido no rekishi," I, p. 13.

system prior to the formation of the Diet. This meant that all the rules regarding the structure and functioning of the bureaucracy were embodied in imperial ordinances rather than in statutes. This practice was continued even after the Diet came into being.^[11]

Although the Diet did emerge as a significant political force, particularly during the Taisho era (1912-26), it never attained sufficient power to control the executive branch. As the instrument and embodiment of executive power, the bureaucracy was thus assured a dominant role in the political system. Further enhancing bureaucratic power was the entry of bureaucrats into political parties, eventually constituting one of the two key groups in ruling parties, together with the so-called "pure politicians."^[12] As Spaulding suggests, the bureaucracy was never a pliant tool of the cabinet, its putative master. Virtually all cabinets, including military cabinets "at the zenith of their power," found it necessary to govern the country "by informal coalitions or working agreements with career civilian bureaucrats, chiefly examination men."^[13] A striking index of the power of the bureaucracy is that, of the thirty men who served as prime ministers in prewar Japan, only two lacked a bureaucratic background, either civil or military. The exceptions were Prince Saionji Kinmochi (1906-08 and 1911-12) and Prince Konoe Fumimaro (1937 and 1940-41).^[14]

Generally speaking, power and prestige go hand in hand; hence, despite the absence of survey data, one may surmise that bureaucrats in prewar Japan, symbols as well as wielders of awesome power, enjoyed high prestige. One indication of their prestige was the consistently keen competition for the higher civil-service examinations, the gateway to the upper ranks of the bureaucracy. In the period 1928-43, the failure rate of applicants for the administrative section of the higher civil-service examinations averaged 90 percent. Moreover, the overwhelming proportion of the successful candidates came from Tokyo Imperial Univer-

[11] Inoki, "The Civil Bureaucracy," pp. 291-92.

[12] Robert A. Scalapino, "Elections and Political Modernization in Prewar Japan" in Robert E. Ward, ed., Political Development in Modern Japan (Princeton: Princeton University Press, 1968), pp. 249-91; Robert A. Scalapino, Democracy and the Party Movement in Prewar Japan (Berkeley: University of California Press, 1962).

[13] Robert M. Spaulding, Jr., "The Bureaucracy as a Political Force, 1920-45" in James W. Morley, ed., Dilemmas of Growth in Prewar Japan (Princeton: Princeton University Press, 1971), p. 76.

[14] Inoki, "The Civil Bureaucracy," p. 293; Bessatsu kokkai benran: Shiryo soshuhen [Separate-Volume Supplement to the National Diet Handbook: Complete Collection of Materials] (Tokyo: Nihon Seikei Shinbun Shuppanbu, 1975), pp. 6-7.

sity, the most prestigious of all institutions of higher learning in prewar Japan.^[15]

It is worth noting that bureaucrats may have been more feared than respected. Their arrogance was legendary, and tales of the abuse of bureaucratic power were legion.^[16] The well-known phrase, kanson minpi (officials revered, citizens despised), not only summed up a major theme in the political culture of prewar Japan; it also epitomized the reality, of bureaucratic dominance in Japanese society.

Structure And Behavior

Stratification

The prewar Japanese bureaucracy was a veritable caste system. It embraced distinct classes of persons, and the status differences and social distances between the classes were wide. The most basic distinction was between officials (kanri ) and nonofficials (hikanri ). The officials, who were appointed either by the Emperor or by the government to which authority had been delegated, owed an unlimited loyalty to the Emperor and were regulated by public law. The nonofficials, on the other hand, lay outside the scope of the Emperor's appointment authority and were subject to the rules of employment contract in private law.^[17]

The officials were divided into four distinct classes: (1) shinninkan , (2) chokuninkan , (3) soninkan , and (4) hanninkan . The main criterion of this classification was the distance from the Emperor or, alternatively, the mode of appointment. Shinnin officials were awarded their letters of appointment, signed by the Emperor and countersigned by the prime minister, in a palace ceremony called the shinnin shiki and attended by the Emperor. Chokunin officials, too, received letters of appointment bearing the signatures of both the Emperor and the prime minister but without a palace ceremony. Sonin officials, on the other hand, were appointed by the prime minister acting on behalf of the cabinet and on

[15] Spaulding, Imperial Japan , pp. 265, 268-69. Some relevant statistics will be presented later in this chapter (see table 2).

[16] See, for example, Fukumoto, Kanryo , pp. 80-95; Mori and Yazawa, Kanryosei no shihai , pp. 162-65 and passim.

[17] Watanabe, "Nihon no komuinsei," pp. 112; Miyake Taro, Gyoseigaku to gyosei kanri [Public Administration and Public Management] (Tokyo: Sakai Shoten, 1974), pp. 78-88.

authority delegated by the Emperor. Finally, hannin officials were appointed by the individual ministers exercising delegated authority. The first three classes of officials were collectively known as kotokan (higher officials). There were nine grades among the kotokan and four among the hanninkan .^[18] Table I displays the ranks and grades of the higher officials, together with their corresponding positions. The barrier between the higher and ordinary officials was so great as to be all but insurmountable; they could be compared with commissioned and noncommissioned officers in the armed forces.^[19]

Because they were outside the realm of public law, the nonofficials (hikanri ) were subject to the vagaries of the changing needs and budgetary resources of individual ministries. Neither their qualifications nor their duties were precisely defined. Nonetheless, there were generally three types of nonofficials: (1) koin , (2) yonin , and (3) shokutaku . The koin performed routine clerical tasks, but many of them aspired for promotion to the hannin rank eventually. The yonin engaged in manual

[18] Watanabe, "Nihon no komuinsei," pp. 112-13; Miyake, Gyoseigaku to gyosei kanri , pp. 79-82; Murobushi Tetsuro, Kokyu kanryo [Higher Civil Servants] (Tokyo: Sekai Shoin, 1983), pp. 28-29.

[19] For a comparison of the kotokan-hanninkan dichotomy to the distinction between commissioned and noncommissioned officers in the military, see Spaulding, "The Bureaucracy as a Political Force," pp. 37-38 and Johnson, MITI and the Japanese Miracle , pp. 57-58. For an inside account of the privileges of kotokan , see Hayashi, Nihon kanryo kenkoku ron , pp. 5-10.

labor. Finally, the shokutaku typically had special ties with a cabinet minister or a parliamentary vice-minister, serving him for the duration of their terms of office. Many shokutaku were retired officials and persons engaged in specialized research activities for the government.^[20]

We have already seen that civil-service examinations were the principal means by which civil officials were recruited. Those who passed the higher civil-service examinations were eligible for appointment as soninkan trainee (shiho ), whereas those who passed the ordinary civil-service examinations could become hanninkan apprentice (minarai ). In neither case was appointment automatic. Passing these examinations simply certified that the successful candidates were qualified for government service. The actual hiring was done by each government ministry or agency, typically following an interview. In most cases the candidates themselves had to take the initiative. From 1887 to 1893, graduates of Tokyo Imperial University were exempt from the higher civil-service examinations. In a sense, graduation from Todai could be viewed as the chief means of recruitment into the higher civil service, whereas passing the examination was the auxiliary means. For Todai was intended from the outset to serve as a training school for the core members of the bureaucracy.^[21]

Another important mode of civil-service recruitment was senko (evaluation), which was applied to teachers, engineers, physicians, and other technical personnel. Until 1899, moreover, all chokunin officials, which included not only section chiefs, bureau chiefs, and vice-ministers but also the chief of the cabinet secretariat, the chief of the cabinet legislation bureau, and prefectural governors, were subject to "free appointment." The scope of discretionary appointment, however, steadily dwindled in subsequent years. Finally, the significance of the ordinary civil-service examinations was considerably diluted in 1913, when all graduates of middle schools were granted exemption; previously, only public-middle-school graduates had been exempt. In addition, all koin with five years or more of experience became eligible for hannin positions.^[22]

Behind the veil of seeming complexity stood one simple fact: the prewar Japanese bureaucracy contained distinctly unequal strata. At the bottom of the pecking order were "nonofficials." Then came ordinary

[20] Watanabe, "Nihon no kanryosei," pp. 113-14.

[21] Ibid., pp. 114-15; Wada, "Bunkan nin'yo seido no rekishi," I, pp. 11-13.

[22] Watanabe, "Nihon no komuinsei," pp. 115-16; Wada, "Bunkan nin'yo seido no rekishi," II, Jinji-in geppo 96(Feb. 1959): 13-15.

or hannin officials. Above them all reigned higher officials or kotokan . Although the wall separating the ordinary and higher officials was by no means impenetrable, it was nonetheless high. After the abolition of the "imperial university privilege" (teidai tokken ) in 1893, nearly all aspirants for elite administrative positions had to scale the heights of the higher civil-service examinations.

Elitism

Another striking characteristic of the prewar Japanese bureaucracy was its domination by graduates of Tokyo Imperial University. Of the ninety-seven men who were appointed as higher-civil-service trainees (administrative) in the 1888-91 period, all but nine were Todai graduates. More important, in two of the four years involved, not a single examination candidate was appointed. In fact, administrative examinations were not even held in 1891, a situation that was repeated in the following year. Inasmuch as "the function of the examinations was simply to fill positions not claimed by Todai graduates," no examinations were necessary in those years, because of the availability of a sufficient number of Todai men. The government's decision in 1893 to abolish the Imperial University exemption stemmed not simply from its desire to accommodate the critics of the privilege but also from the practical need to cope with growing numbers of Todai graduates.^[23]

However, the preponderance of Todai men in the upper echelons of the bureaucracy continued unabated. The principal reason for this was that Todai men dominated the administrative section of the higher civil-service examinations in both absolute and proportionate terms. Whereas the overall success rates (ratios of successful to total candidates) were 25 percent in 1906, 7 percent in 1941, 14 percent in 1942, and 15 percent in 1943, those for Todai men were 50 percent, 16 percent, 29 percent, and 25 percent in those years. About 47 percent of all successful candidates in the 1941-43 period came from Todai, whereas about 10 percent was supplied by Kyoto Imperial University.^[24] If we look at longer time periods, as we do in table 2, the preponderance of Todai men in the examinations becomes even more pronounced. Roughly six out of ten successes came from Todai. The two top

[23] Spaulding, Imperial Japan , pp. 90-99. The quotation is from p. 91.

[24] These statistics were either adapted or estimated from ibid., pp. 268-69 and 277 (tables 48, 49, and 54).

universities, Todai and Kyodai, together accounted for seven out of every ten successes.

Inoki Masamichi's analysis of the academic background of 1,377 higher civil servants (excluding judicial officials) whose biographical sketches appeared in the 1937 edition of Jinji koshinroku (Who's Who) revealed that 73.1 percent were Todai graduates and 9.0 percent were Kyodai graduates.^[25] This is consistent with the data on the successful candidates in the higher examinations discussed above. In comparative terms, the only other country that can match this record is the United Kingdom. According to R. K. Kelsall, the academic background of British senior civil servants (those above the rank of assistant secretary) in 1939 was as follows: of those who entered the civil service through open competition, 51.3 percent had graduated from Oxford and 25.7 percent from Cambridge. When those who entered the civil service through other routes are added, the overall proportion of Oxford men becomes 41.3 percent, and that of Cambridge men 20.1 percent. Significantly, 19 percent of the 1939 senior civil servants included in Kelsall's analysis lacked university education altogether.^[26]

If, as suggested above, the predominance of Todai men in the upper ranks of the bureaucracy was a function of their preponderance among the successful candidates in the higher civil-service examinations, then

[25] Inoki, "The Civil Bureaucracy," p. 296.

[26] R. K. Kelsall, Higher Civil Servants in Britain (London: Routledge and Kegan Paul, 1955), p. 138. table 22. The size of Kelsall's sample was 179.

what accounted for this latter phenomenon? A simplistic explanation might be that, assuming that the examinations were valid and reliable, Todai men were better prepared for them by virtue of their superior intellectual ability and academic training. Given its high prestige, Todai attracted both high-caliber students and distinguished faculty. It is also widely believed that Todai men enjoyed an edge in competition thanks to the near monopoly of the posts of higher examiners by Todai professors.^[27]

According to Spaulding, the proportion of Todai professors among administrative examiners in four different time periods was as follows: 1894-1905: 46.0 percent; 1906-17: 66.0 percent; 1918-28: 44.6 percent; and 1929-41: 41.2 percent.^[28] With the exception of 1906-17, Todai professors were in the minority among the administrative high examiners. In terms of the university background of all administrative high examiners during the four periods, however, the picture changes substantially, with Todai graduates accounting for 72.5 percent, 95.9 percent, 85.7 percent, and 67.5 percent, respectively.^[29] Although Spaulding interprets these data as disproving the view that the performance of Todai men in the higher examinations was related to the control of the latter by the professors and alumni of Todai, it seems nonetheless clear that Todai men did enjoy advantages in more ways than one. Their previous exposure to lectures and examinations by such a high proportion of the examiners was bound to be beneficial not only in the written but also in the oral phase of the higher examinations.

The preponderance of Todai graduates in the higher civil service also bespoke its elitist character in terms of social origins. Although admission to Todai was open to all on the basis of merit, in practice only a very small number of students from poor families could attend it on "scholarships provided by former feudal lords and other rich people." Because "one had to be able to afford six years of schooling—three of higher, or preparatory, school and three of university work," most students at Todai were in fact "the sons of upper and upper-middle-class families (of civil bureaucrats, military officers, landlords, rich farmers, businessmen, and industrialists)."^[30] Of the 1,377 higher civil servants in 1937 analyzed by Inoki, "358 (26.4 percent) were of

[27] Sato Tomoyuki et al., Todaibatsu [University of Tokyo Clique] (Tokyo: Eru Shuppansha, 1972), p. 196.

[28] Spaulding, Imperial Japan , p. 249.

[29] Ibid., p. 251.

[30] Inoki, "The Civil Bureaucracy," pp. 295-96.

samurai (shizoku ) origin and 17 (1.2 percent) were of noble (kazoku ) origin." This means that "72.4 percent of the civil servants were commoners (heimin )."^[31]

Table 3 provides us with a few more details regarding the social origins of prewar Japanese bureaucrats. A major drawback of the data displayed in the table, however, is that they do not specify the time period to which they pertain. The table nonetheless reveals that the single largest category of fathers' occupation is government officials, indicating a high degree of occupational succession. If we discount ordinary officials, however, the proportion of government officials drops to 21.2 percent, which is below that of "agriculture and forestry" (noringyo ). This broad category probably subsumes landlords and rich

[31] Ibid., p. 296.

farmers. No matter how one defines elitist occupations, they appear to be disproportionately represented.

Legalism

Closely related to the phenomenon of Todai dominance in the higher civil service was its markedly legalistic orientation. This was manifested in two related ways: the preponderance of law graduates among the higher civil servants and their preoccupation with formal rules. The popular or journalistic term for this phenonomenon was hoka banno (the omnipotence of law). The emergence of hoka banno was due to a number of circumstances. First and foremost, the decision of the Meiji leaders to emulate the Prussian model and to require legal training as the principal qualification for civil-service appointment was a key contributing factor. The preferential treatment given to Todai law graduates in the early days of the civil service was another. In fact, within Todai itself, the faculty of law (hogakubu ) enjoyed the status of primus inter pares: in the 1880s its dean served concurrently as the president of Todai. He was also empowered to supervise all other law schools that had come into being in Tokyo, of which there were five.^[32]

The subjects in which candidates were tested in the higher civil-service examinations were predominantly law-related. The administrative examinations from 1894 to 1928 had six required subjects: constitutional law, administrative law, criminal law, civil law, international law, and economics. In addition, candidates were required to select one subject from a set of four electives: criminal procedure, civil procedure, commercial law, and finance.^[33] From 1929 to 1941, criminal law and international law were dropped from the list of compulsory subjects; instead, they were added to the greatly expanded list of elective subjects. The other subjects that were newly added were political science, political history, Japanese history, economic history, agricultural policy, industrial policy, social policy, ethics, logic, sociology, psychology, philosophy, and Japanese and Chinese literature. Candidates were required to choose three subjects from the preceding

[32] "Tokyo Daigaku no hyakunen" Henshu Iinkai, Tokyo Daigaku no hyakunen, 1877-1977 [Hundred Years of the University of Tokyo, 1877-1977] (Tokyo: Tokyo Daigaku Shuppankai, 1977), pp. 85-93. From 1886 to 1919 the Faculty of Law was called the College of Law (Hoka Daigaku ).

[33] Spaulding, Imperial Japan , p. 210.

list.^[34] The 1929 reorganization, then, made it possible for administrative candidates to take only three law-related subjects, instead of five.

The 1937 data examined by Inoki indicate that about 47 percent of the higher civil servants were graduates of Todai's law faculty. If graduates of other law faculties were added, such as those of Kyoto, Tohoku, Kyushu, and Chuo universities, the proportion of law graduates would increase markedly. Of the Todai men in the sample, over 63 percent were products of its law faculty.^[35] Another study, by Hata Ikuhiko, found that a total of 5,969 Todai men passed the administrative higher examinations between 1894 and 1947 and that 94.7 percent of them (N = 5,653) were graduates of its law faculty.^[36]

The legalistic approach to administration spawned and perpetuated by the predominance of law graduates in Japanese bureaucracy was exemplified by extreme stress upon rules and precedents: nearly all important matters of administration, and sometimes even trivial ones, needed to be grounded in law—statutes, ordinances, regulations, or precedents. In the words of Milton J. Esman:

Regardless of the emergency nature of specific situations or social necessity, [the official] construes the absence of specific legal authority as full justification for failure to act; and regardless of common sense, he follows regulations m the letter. This dependence on legality resolves all new operating problems into legal problems which can be met only by the issuance or amendment of regulations. The code of regulations thus increases in bulk and complexity. To those trained to understand them, the provisions of the code become surrounded with an inviolability which frequently overrides in importance the situation they are designed to settle.^[37]

Esman argues that an explanation of the prevalence of legalism and other aspects of prewar Japanese administration in terms of German and Prussian origins is inadequate. For "while many features of German administration were successfully transplanted, others never took root." As examples of the procedures the Japanese failed to emulate, Esman cites the "in-service training of young officials" and "Germany's highly integrated budget system and its flexible central departmental staff organization."^[38] It should be pointed out, however, that the Japanese

[34] Ibid., p. 213.

[35] Inoki, "The Civil Bureaucracy," pp. 296-97.

[36] Hata, Kanryo no kenkyu , p. 17.

[37] Milton J. Esman, "Japanese Administration—A Comparative View," Public Administration Review 7, no. 2 (Spring 1947): 101.

[38] Ibid., p. 109.

did embrace the idea of postentry training, even though it was never fully implemented.

In any event, Esman believes that cultural and societal variables offer better clues to Japanese practices. Whereas, in the United States, the tradition of individual rights, equality, limited government, and utilitarianism helped to foster a distinctly pragmatic approach to administration, such a tradition was conspicuously absent in Japan. On the contrary, Japanese society valued the collectivity above the individual, hierarchy above equality, loyalty to the emperor and the state, and reverence for established authority and traditional ways of doing things. Furthermore, legalism served the interests of the ruling elite in the Meiji period and beyond. According to Esman:

The concept of Rechtsstaat . . . . was adopted [from Prussia] primarily to convince the western world that because all acts of Japanese officials were regulated by a strict rule of law, extraterritoriality could safely be withdrawn. Legalism, however, had other important uses for Japanese officialdom. It provided them a detailed and intricate mystery which only initiates into the legal priesthood could properly understand. By making government a complex web of administrative regulations, by thus establishing the indispensability of legal learning, the Samurai and their successors have protected their monopoly of higher administrative positions and preserved the power of the bureaucracy against possible assault by the military, the nobility, the financial clique, or the political parties.^[39]

Although quite plausible, these ideas are necessarily speculative; credible evidence with which to test their validity is not readily available. Nonetheless, the salience of legalism in the prewar Japanese bureaucracy is beyond challenge, and its deleterious effects are widely recognized.

Jimukan Vs. Gikan

A corollary of the ascendancy of legalism was the subordination of technical specialists (gikan ) to administrative generalists (jimukan ).^[40] We have previously noted that the gikan were recruited without

[39] Ibid., pp. 109-11. The quotation is from p. 111. It should be stressed that the "successors" to the samurai mentioned by Esman were predominantly commoners (72.4 percent in Inoki's 1937 sample). Only 26.4 percent of the 1937 sample was of samurai (shizoku ) origin. see n. 30, above.

[40] In prewar Japan, the term jimukan was used in two distinctly different senses. First, it referred to all non-gikan . Second, it encompassed soninkan with grades 4 through 6. See table 1.

examination through screening (senko ). Although they acquired sonin rank and therefore were classified as higher officials (kotokan ), only a small fraction of them was given managerial or supervisory positions. Chiefs of sections, divisions, and bureaus dealing with complex technical matters were typically jimukan trained in law but totally lacking in substantive expertise.

Moreover, as table 4 shows, there was a sizable gap between the two types of officials in the speed of promotion. In the Ministry of Home Affairs, gikan trailed behind jimukan every step of the way. It took gikan eleven more years than jimukan to reach grade 1. Those who made it, however, were the lucky ones, for the vast majority did not. In the technically oriented Ministry of Posts and Communications, gikan had an edge over jimukan in the early years of their career and then began to lose ground. The overall disadvantage of six years and three months, nonetheless, was considerably better than the lag of eleven years and four months in the Home Affairs Ministry. When all six ministries are considered, jimukan enjoy an edge in all stages of promotion, accumulating a net advantage of over seven years.

This anomalous situation bred inefficiency and conflict and had demoralizing effects on technical personnel, many of whom were graduates of Todai and other prestigious universities. Because there were more gikan than jimukan in the prewar higher civil service (about 53 percent to 47 percent), the situation in effect amounted to minority rule—the triumph of dilettantism over expertise.^[41]

Sectionalism

Nakane Chie has coined the term tate shakai (vertical society) to underscore the precedence of vertical over horizontal relationships—or ba (frame) over shikaku (attribute)—in Japan.^[42] In organizational terms, this idea implies the development of a strong sense of identification with the organization and intense feelings of solidarity among all members of the organization. Two personnel practices helped to bolster such a tendency: decentralized recruitment and lifetime employment. The former meant that although the civil-service examinations were centrally administered, the actual recruitment and appointment of qualified candidates was the responsibility of each ministry. Hence one became an official, not of the Japanese imperial government as such, but of the Home Ministry, the Finance Ministry, and so on. The custom of lifetime employment meant that one was expected to stay in the same organization until retirement. Temporary assignments outside one's own ministry were more frequent than interministerial transfers on a permanent basis. Within the same ministry, however, officials were rotated between the headquarters and the field as well as between different bureaus, divisions, and sections.^[43]

[41] Lt. Col. Hugh H. MacDonald and Lt. Milton J. Esman, "The Japanese Civil Service," Public Personnel Review 7. no. 4 (Oct. 1946): 218-19; Watanabe, "Nihon no komuinsei," p. 130. Of the 186 gikan covered by table 4, 145 (78 percent) were Todai graduates and 20 (10.8 percent) were graduates of Kyodai. All of the remainder had gone to either Kyushu or Tohoku University. On the other hand, 142 of the 152 jimukan (94 percent) were Todai men. The remainder were all Kyodai graduates. See Jinji-in, Kyuyokyoku, "Kyu kanri seidoka ni okeru kotokan no keireki chosa no kekka gaiyo" [A Summary of the Results of an Investigation into the Background of Higher Officials Under the Old Bureaucratic System], Kikan jinji gyosei [Public Personnel Administration Quarterly] 25 (Aug. 1983): 96.

[42] Nakane Chie, Tate shakai no ningen kankei: Tan'itsu shakai no riron [Human Relations in a Vertical Society: The Theory of a Homogeneous Society] (Tokyo: Kodansha, 1967), pp. 26-67 and passim.

[43] The administrative officials in Jinji-in's study, on which table 4 was based, experienced between eleven and thirteen changes of assignment; the average duration of each assignment was less than two years. Jinji-in, Kyuyokyoku, "Kyu kanri seidoka ni okeru kotokan no keireki chosa no kekka gaiyo," p. 108.

Takahashi Hideki's analysis of the career patterns of 145 officials who entered the Finance Ministry after passing the higher civil-service examinations during the Taisho era (1912-26) shows that those who eventually achieved the positions of bureau chief or administrative vice-minister typically served as financial clerks (zaimu shoki ) in overseas diplomatic or consular missions within the first year or two of appointment and then as heads of local tax offices (zeimushocho ) before being assigned to the headquarters. After serving in a number of posts, they would reach the post of section chief in their fourteenth year of service.^[44]

Such practices, reinforced by the cultural norms of tate shakai , bred jurisdictional rivalries between ministries that were dysfunctional for the government as a whole. Jealous guarding of turfs even occurred within the same ministry in some cases.^[45]

Security Of Tenure

Civil servants in prewar Japan enjoyed security of tenure, a sine qua non of rational-legal bureaucracy as stipulated by Max Weber.^[46] As was true in the Weberian model, this was a relative kind of security rather than an absolute one. In Weber's words, "where legal guarantees against arbitrary dismissal or transfer are developed, they merely serve to guarantee a strictly objective discharge or specific duties free from all personal considerations."^[47]

An imperial ordinance on the status of civil officials (Bunkan Bungenrei ) promulgated in 1899 enumerated three grounds on which civil officials might be discharged: (1) inability to perform duties due to injury, illness, or enfeeblement of mind and body; (2) request by an official invoking illness or other personal circumstances; and (3)

[44] Takahashi Hideki, "Kobun seidoka no shoshin jittai no bunseki" [Analysis of Promotion Patterns Under the (Prewar) Higher Examination System] I, Jinji-in geppo 342(Aug. 1979): 23-24.

[45] For an elaboration of this point, see Albert M. Craig, "Functional and Dysfunctional Aspects of Government Bureaucracy" in Vogel, ed., Modern Japanese Organization and Decision-Making , pp. 15-17. See also Spaulding, "The Bureaucracy as a Political Force," pp. 56-60. Spaulding points out that the adverse effects of sectionalism were somewhat mitigated by "informal collaboration among bureaucrats of different ministries," which "appears to have been very common.' Ibid., p. 58.

[46] Hans Gerth and C. Wright Mills, trans. and eds., From Max Weber: Essays in Sociology (New York: Oxford University Press, 1946), pp. 202-3; Max Weber, The Theory of Social and Economic Organization , trans. A.M. Henderson and Talcott Parsons (New York: Oxford University Press, 1947), pp. 333-34.

[47] Gerth and Mills, From Max Weber , p. 202.

reduction in force occasioned by reorganization. The same ordinance specified that officials might be furloughed for any of the following reasons: (1) recommendation of disciplinary action by the disciplinary committee; (2) being charged with a crime; (3) reduction in force occasioned by reorganization; and (4) administrative necessity in the ministry or agency concerned. Because this last-mentioned ground lent itself to abuse by party politicians, it was tightened up in 1932 by requiring review of any proposed furlough on grounds of administrative necessity by the committee on the status of officials. The second ground for furlough was also revised to read "being indicted in a criminal case."^[48]

Early Retirement

Although civil servants enjoyed a relatively high degree of job security and were in effect guaranteed lifetime employment, their actual careers were fairly short. According to MacDonald and Esman, most higher officials were expected to retire in their late forties. In as much as the pension scheme allowed civil officials "to retire after 17 years of service on one third of their base pay," those who retired in their late forties or even in their early fifties found it necessary to supplement their income somehow.^[49] Some found their "second careers" in private business, and others "received Imperial nominations to the old House of Peers."^[50]

[48] Wada, "Bunkan nin'yo seido no rekishi," II, pp. 10-11; ibid., III, Jinji-in geppo 97 (Mar. 1959): 8-9.

[49] MacDonald and Esman, "The Japanese Civil Service," pp. 222-23. The quotation is from p. 222. Chalmers Johnson writes, however, that the "prewar civilian bureaucracy did not retire as early as contemporary officials do." See his article "The Reemployment of Retired Government Bureaucrats in Japanese Big Business," Asian Survey 14, no. 11 (Nov. 1974): 958. In the absence of any statistical evidence, it is difficult to reconcile the conflicting views.

[50] Johnson, ibid., p. 958. Kusayanagi Daizo writes that prewar pensions were generally sufficient to support retired civil officials. See his book, Kanryo okoku ron [On the Kingdom of Bureaucrats] (Tokyo: Bungei Shunju, 1975), p. 78. Strictly speaking, whether pensions were adequate or not is a relative matter. From 1931 to 1945, for example, the basic salary of administrative vice-minister was 5,796 yen, and that of bureau chief was 4,572 yen. The amount of annual pension for a former vice-minister, therefore, would be 1,932 yen, and that for a former bureau chief 1,524 yen. This would mean that the retired vice-minister would still receive more than the highest-paid hannin official in his midcareer. The pension for a retired bureau chief, on the other hand, would be on a par with the salary of the second-highest-paid hannin official (1,540 yen) and a sonin official with four or five years of service. See the salary schedule for 1931 officials in Watanabe, "Nihon no komuinsei," p. 129. Between 1894 and 1935, a total of 146 retired bureaucrats served as cabinet ministers, 152 as members of the House of Peers, 295 as members of the House of Representatives, and 24 as elected prefectural governors. Fifteen others served as advisers to the Privy Council. See Hata, Kanryo no kenkyu , p. 25.

A key factor that helps to explain the phenomenon of early retirement was the unwritten rule within the bureaucracy decreeing that all members of the same entering class in a given ministry should resign as soon as one of them attained the position of administrative vice-minister. This custom served the dual function of sparing the vice-minister the embarrassment of having to issue orders to members of his peer group and of clearing the way for advancement of younger officials to senior posts. The former is an important consideration in Japanese culture, where a strong sense of hierarchy is counterbalanced by an equally strong sense of equality among the cohort. Having passed the higher civil-service examination and entered the ministry in the same year, the cohort typically forged extremely close personal ties throughout their government careers.

Conclusion

In sum, the emergence of a modern government bureaucracy in prewar Japan resulted from both internal and external factors. The decision of the Meiji oligarchs to adopt the trappings of modern government, including a constitution, a parliament, a cabinet, and a merit bureaucracy, reflected their desire not only to appease the internal forces championing the cause of democracy but also to convince foreign powers that Japan merited respect and, above all, equal treatment.

Both the constitutional framework and the structure and principles of bureaucracy that the Japanese ultimately adopted were patterned after the German model. Enshrining the myth of imperial supremacy, the new system, in effect, bestowed awesome powers on those who presumed to act as the emperor's agents, which included the bureaucrats. The single most important feature of the bureaucracy was the institutionalization of the merit principle: civil-service examinations became firmly established as the principal means of selecting the most qualified candidates for the exercise of bureaucratic authority.

The Japanese also borrowed from their German model a system of rigid stratification, which separated officials from nonofficials and divided officials into two distinctly unequal status groups, kotokan and hanninkan . The former was in turn divided into three classes and nine grades. The veritable caste system all this spawned necessarily entailed a pronounced degree of elitism. Such elitism, it should be stressed, was based not on ascription but on achievement. A major factor contributing to and sustaining this tendency was the domination of the higher

civil service, both in number and in influence, by graduates of Tokyo Imperial University, an institution designed originally as a training school for government officials.

Another notable feature of the prewar Japanese bureaucracy was legalism, which manifested itself in two ways: the preponderance of law graduates in the higher civil service and the preeminence of laws, rules, regulations, and precedents in public administration. A corollary of these twin phenomena was the ascendancy of administrative officials trained in law over technical specialists. Although a numerical minority, the former controlled most of the levers of power, outperforming the latter in promotion by a substantial margin.

The propensity of bureaucratic organizations to guard their bailiwicks against encroachment by their rivals is a universal phenomenon. However, it was reinforced in Japan by the cultural norms of a "vertical society" and by institutional practices; the latter included the decentralized hiring of civil servants, the custom of lifetime employment, and the infrequency of interministerial transfers. Like their counterparts elsewhere, Japanese civil servants enjoyed a high degree of job security. However, its benefits were somewhat diluted by the practice of early retirement; "lifetime" employment meant considerably less than what it implied, for, according to MacDonald and Esman, retirement typically occurred in the late forties, the prime of the bureaucrat's life. This, too, was a function, in part, of cultural norms, notably a strong sense of equality among the cohort and the resultant difficulty of maintaining hierarchical relationships among them. No less important was the function the practice performed for the organization as a whole: facilitating frequent turnovers at its upper echelons, where, because of the structural imperative of a hierarchical organization, positions become progressively scarce.

The Japanese experience demonstrated the inherent limits of foreign institutional models. Although they perform the indispensable function of guiding the uninitiated groping in the dark, their actual impact in the adopted country is contingent upon the dynamics of politics and, above all, their interaction with cultural norms.