The Aftermath

S-1 Examination

The passage of the national public-service law in 1947 and its substantial revision in 1948 helped to establish the basic legal and institutional framework for the modernization and democratization of the Japanese-government bureaucracy. However, SCAP knew all along that a comprehensive civil-service law and a strong central personnel agency would be powerless to overcome the inertia, or, worse yet, the resistance of the old-line bureaucrats. What was needed, therefore, was

[53] Jinji-in, Jinji gyosei nijunen , pp. 37-39; Asai, Shinpan kokka komuinho seigi , pp. 6-11.

a mechanism through which to identify and weed out the most unfit and to infuse the bureaucracy with fresh blood. To accomplish this goal, article 9 of the supplementary provisions of the national public-service law required the holding of an examination to fill all administrative positions at the level of section chief and above.^[54]

In the course of preparing for the examination, SCAP's Civil Service Division decided that "the examination should have none of the aspects of a purge," even though that "would probably make it difficult to attain the objectives of the examination in full measure." Underlying this decision were two considerations: first, "many of the more recalcitrant and unadaptable officials had [already] been eased out under pressure of the various SCAP sections"; second, it was deemed most important "to preserve the integrity and objectivity of the examination." The civil-service division also decided to make the examination "open and competitive," rather than qualifying. In other words, all incumbents of designated positions would be required to compete with other candidates if they wished to be reappointed. Insofar as designating positions subject to examination was concerned, the key criterion was whether or not their duties were primarily administrative. The positions of administrative vice-minister, bureau chief, division chief, and section chief in the regular administrative agencies were included, in accordance with the law. A total of 2,621 positions were designated as subject to examination, and they were in turn grouped into 60 occupational areas ranging from general administration to police.^[55]

Each designated position was not only allocated to an occupational group but also placed in one of four administrative levels. Level 1 encompassed administrative vice-ministers and comparable positions (such as directors general of agencies) and level 4 encompassed section chiefs and their equivalents. The examination was designed to measure four essential characteristics: "(a) administrative aptitude, (b) technical competence, (c) satisfactory character, and (d) physical fitness." Administrative knowledge and aptitude were tested through two procedures. First, each candidate was required to take a written examination

[54] Maynard N. Shirven and Joseph L. Speicher, "Examination of Japan's Upper Bureaucracy," Personnel Administration 14, no, 4 (July 1951): 48-50. When they wrote this article, the authors were on the staff of the SCAP General Headquarters. Shirven was Chief of the Civil Service Division, the Government Section, and Speicher was Chief of the Examinations Branch of the same Division.

[55] Ibid., pp. 50-52.

consisting of "90 multiple-choice items relating to organization, personnel administration, financial management, principles of supervision, interpretation of data, and other facets of the practical work of public administrators." Second, in addition to the written test, a candidate was required to meet "a qualifying standard of practical experience." A candidate for the position of administrative vice-minister, for example, "was required to have served at least one year as a bureau chief or in a position of similar responsibility."^[56]

The decision to include a test of technical competence was based on SCAP's perception that "one of the principal weaknesses of the Japanese civil-service system was the preponderance of officials with only legal training and background in positions where technical knowledge was essential." Candidates for positions in 45 of the 60 occupational groups were required to take tests consisting of "80 multiple-choice items constructed from basic reference material and from experience in the occupation." These tests were said to have been "rather elementary in nature." In addition, the candidates had to meet the experience requirement in the occupational area of their choice.^[57]

Determination of "satisfactory character" was the responsibility of the National Personnel Authority's bureau of investigation. First, the candidate's own statements relating to his experience, education, and personal history contained in his application were verified. Second, an attempt was made to evaluate his personality traits by obtaining ratings on such qualities as "ability to train subordinates, originality, and planning ability" from at least three persons familiar with the candidate's previous experience. On the basis of these data, each candidate was given an overall score of either pass or fail. The final component of the examination was designed "to eliminate those candidates who were physically unfit for the performance of the duties of the designated positions." Inasmuch as incumbent employees of the national and local governments and other public bodies were subjected to periodic medical examinations, only candidates from outside the government were required to take the physical examination.^[58]

Unhappy about the imposed nature of the examination and uncertain about its results, the bureaucrats tried to force its cancellation by postponing the submission of applications until the last minute. The

[56] Ibid., pp. 52-53.

[57] Ibid., p. 53.

[58] Ibid., pp. 54-55.

"war of nerves," however, was won by SCAP, and before the deadline was past "a total of 12,000 applications from 8,076 individuals had been filed for the 2,621 designated positions." After preliminary screening, 7,815 candidates were allowed to take the examination. The administrative aptitude test was given on 15 January 1950, and the technical tests for various occupational groups were given from 20 to 22 January. Because no time limit was imposed and candidates were permitted to smoke and were served hot tea, the examination was dubbed a "paradise examination." It was also called S-1, signifying that it was the first examination for supervisory personnel in postwar Japan.^[59]

Did the examination accomplish its purpose? Directly, it helped to eliminate about 21 percent of incumbents. At level 1 (administrative vice-minister), however, the attrition rate was only 12 percent. "Indirectly, it is estimated that an additional 10-20 percent may have been eliminated. For example, when the examination was announced, a number of incumbents either resigned or did not apply for the examination, while others secured transfers to non-designated positions." One reason for the remarkably high rate of retention of incumbents was that they were given preferential treatment: once they passed the examination, they "were likely to be reappointed regardless of their position on the eligibility list"—that is, regardless of their actual scores in the examination.^[60]

Unbeknownst to SCAP, the Japanese bureaucrats also resorted to the "self-defensive measure" of asking other qualified candidates to withdraw their names from the eligibility list.^[61] So widespread and effective was this practice, in fact, that of the 175 persons who were appointed to levels I and 2 positions following the examination, only one was a "civilian"—that is, a person who had not previously served as a government official.^[62] Not surprisingly, the examination served to tarnish the image of SCAP even among those who survived the ordeal. The administrative-aptitude test, for example, was perceived as being too simplistic and of questionable validity. Recalling his experience two decades later, one Japanese participant wrote that senior civil servants

[59] Ibid., p. 55; Jinji-in, Jinji gyosei nijunen , p. 119. For a detailed analysis of the S-1 examination, see Miyake, Gyoseigaku , pp. 112-28. He provides examples of questions along with correct answers.

[60] Shirven and Speicher, "Examination of Japan's Upper Bureaucracy," pp. 56-57.

[61] Kaneji Hiroshi, "Jinji-in no omoide" [Recollections on the National Personnel Authority] in Jinji-in, Jinji gyosei nijunen , p. 53.

[62] Hata, Kanryo no kenkyu , p. 106.

taking the test had been appalled by SCAP's low opinion of their ability reflected in the "simpleminded and mechanical" questions in the test.^[63]

Post-Occupation Developments, 1952-87

Notwithstanding the lofty goals pursued by the Occupation, its methods of operation were autocratic and frequently high-handed, generating resentments on the part of the Japanese people, particularly those in the bureaucracy and in the Diet. The formal termination of the Occupation in April 1952, therefore, paved the way for a reappraisal of the various reform measures implemented under American tutelage. The sentiment for revision of the national public-service law was quite strong, because it was perceived as an alien imposition not suited to Japanese needs. The National Personnel Authority had become an object of intense resentment, not only because of its unprecedented powers but also because it had served as a tool of SCAP's civil-service division. Proposals for reform, even abolition, of the authority were justified on various grounds.^[64]

The government and the ruling party argued that the authority's independence and powers undermined the cabinet-responsible system on which the Japanese government was based. They further pointed to what they described as deleterious effects of the authority's annual recommendations on compensation: the latter allegedly provided fuel for organized labor's spring offensive for wage increases and influenced wage levels in the private sector. The opposition parties and labor organizations, on the other hand, advocated reform of the authority on the ground that it was not fulfilling its original goal of safeguarding the interests of public servants. The main function of the annual recommendations on compensation, they argued, was to legitimize the government's policy of holding wages and salaries to a minimum.

Another consideration that underscored the need for revision of the law was its incompatibility with Convention Number 87 of the International Labor Organization (ILO), which the Japanese government was interested in ratifying. The ILO treaty called for a guarantee by its contracting parties of their citizens' right to establish and join

[63] Hayashi Shuzo, "Shiken jigoku iroiro" [Aspects of an Examination Hell], Jinji-in geppo 236 (Oct. 1970): 4. See also "Zadankai: Jinji-in no omoide arekore," p. 13.

[64] Asai, Shinpan kokka komuinho seigi , pp. 15-16. For an account of the resentment the authority faced and how it overcame attempts to emasculate it in the waning days of the Occupation, see Tomioka Shozo, "'Hoittoni keikoku shokan' o sukupu" [Getting a Scoop on the "Whitney Warning Letter"], Kankai [The Official World], July 1986, pp. 188-93.

labor unions, the unions' autonomy, their right to establish and join federations and international organizations, and other labor-related rights.^[65]

None of the nine formal proposals for the revision of the national public service law submitted to the Diet between May 1952 and December 1963, however, succeeded in mustering sufficient support. By the time a bill for revision of the law was submitted to the Diet for the tenth time in January 1965 by the first Sato government, a sense of urgency had been added: the powerful labor federation Sohyo (Nihon Rodo Kumiai Sohyogikai, or the General Council of Japan Labor Unions) and a union representing locomotive engineers of the national railroads had filed a complaint against the Japanese government with the ILO, and there were signs that other labor unions might follow suit. The ILO had already initiated an investigation into the complaint. Against such a backdrop the Sato government cited the need to amend the provisions in the national public-service law pertaining to the rights of public servants to organize so that Japan might ratify ILO Convention Number 87.

Additionally, the government stressed the need to establish a system of responsibility in public-personnel administration, proposing to set up a new personnel agency within the office of the prime minister. Instead of replacing the National Personnel Authority, however, the new agency, to be called the Bureau of Personnel (Jinji-kyoku), would simply take over some of the functions previously performed by the authority and by the Finance Ministry. Specifically, the bureau would be given jurisdiction over the efficiency, welfare, and service of public servants, their retirement allowances, the compensation of special-service public servants, the coordination of personnel management among the various ministries and agencies, and the collection and analysis of statistical data regarding personnel administration. The bill embodying the preceding proposals was approved by both houses of the Diet on 17 May 1965. On the same day the Diet also passed a bill empowering the government to ratify ILO Convention Number 87.

In sum, the period of "trial" for the National Personnel Authority came to an end in 1965 with its powers only marginally altered.^[66] The

[65] Asai, Shinpan komuinho seigi , p. 16; Jinji-in, Jinji gyosei nijunen , pp. 39-40.

[66] Tsuji Kiyoaki, a distinguished student of Japanese public administration, characterized the thirteen years following the end of the American Occupation as the "period of trial for the National Personnel Authority" (Jinji-in no kuro no jidal ). See "Zadankai: Korekara no jinji gyosei" [Roundtable Discussion: Public Personnel Administration in the Days Ahead], Jinji-in geppo 274(Dec. 1973): 4.

civil-service reform spearheaded by the Occupation thus overcame the first significant challenge to its viability. Although numerous changes were introduced in subsequent years, none can be characterized as major. The more noteworthy changes pertained to (1) civil-service examinations, (2) the expansion of postentry training programs, (3) the introduction of a mandatory retirement system, and (4) reorganization of government agencies.

The civil-service-examination system has undergone change three times in the postoccupation period. Until 1956, higher' civil servants were recruited through the examination to recruit sixth-grade national civil servants (6-kyu shoku kokka komuin saiyo shiken ). But in 1957, in conjunction with the revision of the grade system from a fifteen-grade to an eight-grade system, the sixth-grade civil-service examination was renamed the higher civil-service examination (kokka komuin saiyo jokyu shiken ). The erstwhile fifth-grade civil-service examination became the intermediate civil-service examination (kokka komuin saiyo jukyu shiken ). Along with the lower civil-service examination (kokka komuin saiyo shokyu shiken ), the civil-service-examination system now consisted of three distinct levels.^[67]

Then in 1960, the higher civil-service examination was subdivided into types A and B (koshu and otsushu ). Officially, however, this change was characterized as the renaming of the previous higher civil-service examination (HCSE) as the type-B HCSE, accompanied by the introduction of an entirely new examination, the type-A HCSE. The stated rationale for the change was that the new type-A examination would make the civil service more competitive with the private sector, for it would upgrade the initial rank of new employees (from grade 7 to grade 6) and improve their prospects for promotion to managerial positions.^[68] As we shall see shortly, however, what really happened was the creation of two distinct classes of higher civil servants: those in the elite track (type-A-examination graduates) and those in the quasi-elite track (type-B-examination graduates).

The latest change in the examination system occurred in 1985, when the type-A HCSE was renamed the type-I civil-service examination (isshu shiken ), the type-B HCSE was abolished, the intermediate examination was, in effect, replaced by the type-II civil-service examination (nishu shiken ), and the lower examination was renamed the

[67] Jinji-in, Jinji gyosei nijunen , p. 136.

[68] Ibid., p. 137.

type-Ill civil-service examination (sanshu shiken ). Officially, however, the type-II examination was described as an entirely new examination rather than a successor to the intermediate examination. It was, in fact, characterized as an upgraded version of the intermediate examination, which in the preceding years had become a vehicle for recruiting not high-school graduates but primarily university graduates. This reflected in part the increase in the number of university graduates in the general population and in part the competitiveness of the higher civil-service examinations. Whether the change will lead to the recruitment of candidates for middle-level management who are better qualified than before, however, remains to be seen.

On the other hand, the educational level of the applicants for the type-II examination in the first two years of its existence was slightly higher than that of applicants for the intermediate examination. Whereas in 1984, the last year in which the intermediate examination was given, only 1 percent of applicants had done some graduate work, the proportion increased to 1.6 in 1985 and to 1.8 in 1986. Among the successful candidates, the increase was more substantial: from 1.8 percent in 1984 to 4.2 percent in 1985 and to 3.8 percent in 1986. So far as university graduates were concerned, however, the picture changed very little. Although the proportion of university graduates increased slightly among all applicants (80.5 percent in 1984, 82.1 percent in 1985, and 81.5 percent in 1986), it remained all but identical among the successful candidates (91.3 percent in 1984, 91.5 percent in 1985, and 91.3 percent in 1986). Overall, the proportion of the applicants with at least some university education increased from 81.5 percent in 1984 to 83.3 percent in 1986, and that of the successful applicants with the same level of educational attainments increased from 93.1 percent in 1984 to 95.1 percent in 1986. There was also notable change in the type of universities attended by successful applicants. The proportion of those who had attended national universities increased from 31.8 percent in 1984 to 38.7 percent in 1986, whereas that of those who had attended private universities decreased from 57.2 percent to 52.2 percent in the same period.^[69] Although this may be viewed as an indirect indicator of the greater competitiveness of the type-II examination as compared with that of its predecessor, it also suggests a step backward from the standpoint of making the civil service more democratic.

[69] Jinji-in, Komuin hakusho [White Paper on Civil Servants], 1987 (Tokyo: Okurasho, Insatsukyoku, 1987), p. 36.

Another development related to the civil-service examinations is the gradual lifting of restrictions on women, a development that is plainly conducive to democratization. In 1979 women were allowed for the first time to take entrance examinations to state-run schools that train air-traffic controllers, maritime-security officers, and meteorologists. There were 479 women applicants, of whom 36 were successful.^[70] In the following year, women were allowed to take civil-service examinations for "national tax specialists" (kokuzei senmonkan ) and guards in the imperial palace. There were 573 women applicants for the former, of whom 32 were successful, and 91 women applicants for the latter, of whom 2 were successful.^[71] Restrictions were further relaxed in 1981, allowing women to become border guards (nyukoku keibikan ) and prison guards (keimukan ). Three women (out of 163 female applicants) passed the examination for the former, and 54 women (out of 466 female applicants) passed the examination for the latter.^[72]

Another noteworthy trend is a phenomenal increase in the number and variety of training programs for civil servants. Inasmuch as we shall examine the civil-servants' postentry training at length in chapter 6, it will suffice here to note a few statistics. Whereas in the late 1940s there existed only a dozen or so training programs, in which less than 1 percent of the civil servants participated, by the late 1960s the number of training programs had exceeded 4,100 and nearly 17 percent of the civil servants participated in them. By the late 1970s, nearly one out of every four civil servants was participating in some 6,700 programs. Finally, in the mid-1980s nearly four of every ten civil servants were participating in 9,200 programs. These programs encompassed a wide array of organizations, the entire spectrum of civil servants up to the section-chief level, and the whole gamut of subjects.^[73]

In 1985, a mandatory retirement system for civil servants went into effect. With a few exceptions, the retirement age was set at sixty. Civil servants whose duties and responsibilities are deemed to be "special" or who cannot be easily replaced, however, would be allowed to extend their retirement age up to five years. Physicians and dentists were prime examples of such persons.^[74] The new system, however, was not

[70] Ibid., 1980, p. 25.

[71] Ibid., 1981, p. 25.

[72] Ibid., 1982, p. 26.

[73] Jinji-in, Nenji hokokusho [Annual Report], 1948-49 through 1986 (Tokyo: Okurasho, Insatsukyoku, 1950-87).

[74] Jinji-in, Komuin hakusho , 1982, pp. 44-47; ibid., 1986, pp. 51-52.

expected to affect the retirement practices of higher civil servants, who retire in their fifties. We shall examine this issue in detail in chapter 8.

Finally, we shall note briefly salient changes in the structure of the executive branch of the Japanese government. To list the most noteworthy of them, we must mention the establishment of the Economic Planning Agency in 1955, the Science and Technology Agency in 1956, the Ministry of Home Affairs (Jichisho, initially translated as the Ministry of Autonomy) in 1960, the Environment Agency in 1971, the Okinawa Development Agency in 1972, the National Land Agency in 1974, and the Management and Coordination Agency in 1984. In most of these cases, existing organizations were renamed, upgraded, or restructured. To cite two examples, in 1960 the Autonomy Agency (Jichicho) was upgraded to the Ministry of Home Affairs, and in 1984 the Administrative Management Agency was in effect replaced by the Management and Coordination Agency (Somucho), which also assumed some of the functions that had previously been performed by the prime minister's office.^[75]

Even though the goal of "administrative reform" (gyosei kaikaku ) has received much attention in the postoccupation period, actual results attained have been meagre. The number of "regular service" national civil servants (ippan shoku kokka komuin ) has ranged from 643,057 in 1955 to 867,172 in 1981. In 1986, it declined to 847,004.^[76] Although a significant streamlining of the government structure would require a thoroughgoing reappraisal and a substantial reduction of the powers of the central-government ministries and agencies, the entrenched powers of the bureaucracy have thus far forestalled such an approach, necessitating the adoption of a "uniform reduction method" whereby all ministries and agencies are required to eliminate the same number of bureaus, sections, or other units.^[77] A recent example of such an approach is Prime Minister Takeshita Noboru's plan to seek a transfer

[75] Gyosei Kanricho, Gyosei kanri no genjo: Gyosei kaikaku no doko [The Present Condition of Administrative Management: Trends in Administrative Reform] (Tokyo: Okurasho, Insatsukyoku, 1984), pp. 56-57.

[76] For these and related statistics, see Jinji-in, Jinji gyosei sanjunen no ayumi [The Path of Thirty Years of Public Personnel Administration] (Tokyo: Okurasho, In-satsukyoku, 1978), p. 544; idem, Komuin hakusho , 1978 through 1987.

[77] For a critical assessment of this issue, see Yoshinori Ide, "Administrative Reform and Innovation: the Japanese Case," International Social Science Journal 21, no. 1 (1969): 56-67; Naka Mamoru, "Kongetsu no shoten: Nakasone gyokaku no sannenban o sokatsu suru" [Focus of the Month: Summarizing Three and a Half Years of Nakasone's Administrative Reform], Kankai , Aug. 1986, pp. 82-93.

of one unit from each national-government ministry or agency to the local level.^[78]

The poor record of the Japanese government in administrative reform should caution against minimizing the results of the Occupation reforms. Let us attempt to assess the latter in an impressionistic way.