Salient Features Of The Policy-Making Process

Bureaucrats And Politicians: Four Images

In their comparative study of bureaucrats and politicians in Western democracies, Joel D. Aberbach and his colleagues posit four possible "images" of the relationship between the two sets of players in the

political arena. The oldest and simplest of the four is image I: "politicians make policy; civil servants administer. Politicians make decisions; bureaucrats merely implement them."^[1] Although this image was propagated by classical writers on bureaucracy and public administration, it has relatively few adherents today. The boundary line between politics and administration is far from clear-cut in the real world.

Somewhat more credible is image II, which "assumes that both politicians and civil servants participate in making policy, but that they make distinctive contributions." Civil servants contribute "neutral expertise," emphasizing "the technical efficacy of policy," whereas politicians contribute "political sensitivity," focusing on the "responsiveness [of policy] to relevant constituencies."^[2] If this image comes closer to portraying the reality than the preceding one, it still leaves something to be desired as an adequate account of the division of labor between politicians and bureaucrats; it fails, for example, to take into account either the growing expertise of politicians in substantive policy areas or the politicization of bureaucrats.

Images III and IV aim at rectifying this shortcoming. Image III posits that "both bureaucrats and politicians engage in policy making, and both are concerned with politics." There is nonetheless a division of labor between them: "whereas politicians articulate broad, diffuse interests of unorganized individuals, bureaucrats mediate narrow, focused interests of organized clienteles."^[3]

Image IV goes farther than any of the preceding ones in blurring the distinction between the roles of politicians and bureaucrats; it suggests that developments in the last few decades, notably a "politicization" of the bureaucracy and a "bureaucratization" of politics, have helped to produce a "pure hybrid." This image, caution Aberbach and his associates, is highly speculative. Whether such hybrids have actually become the dominant players in Western-style democracies remains to be verified.^[4]

[1] Joel D. Aberbach, Robert D. Putnam, and Bert A. Rockman, Bureaucrats and Politicians in Western Democracies (Cambridge: Harvard University Press, 1981), p. 4. Three other scholars also collaborated in the study: Thomas J. Anton, Samuel J. Eldersveld, and Ronald Inglehart.

[2] Ibid., p. 6.

[3] Ibid., p. 9.

[4] Ibid., pp. 16-19.

The Role Of The Section

Which of the preceding four images best illuminates the Japanese reality? Because we are interested primarily in higher civil servants, we can readily dismiss image I. Only at the lowest rungs of Japanese bureaucracy can a dichotomy between policy (or politics) and administration be sustained; for most civil servants, particularly those in the intermediate and high echelons, participation in policy making in some capacity is taken for granted.

To ascertain which of the remaining three images is most appropriate for Japan, we need to delineate the outstanding features of the policy-making process in the Japanese government. To consider how policy proposals are prepared for submission to the Diet, we must first note the centrality of the section (ka ) of a ministry, which does most of the spadework, including the drafting of a proposal. This does not mean, however, that ideas for a policy necessarily originate at that level. Even when they originate at higher levels, as they frequently do, the burden of preparing the paperwork and conducting the preliminary consultation falls on the section.^[5]

Within the section the task of preparing the draft proposal is typically given to an assistant section chief (kacho hosa ) who is a "noncareer" man. Because "career" civil servants—those on the elite track tend to be rotated frequently, their level of substantive expertise is relatively low. By contrast, "noncareer" civil servants stay on the same job for a prolonged period of time; for most of them the position of assistant section chief may be the ultimate destination. They are therefore living repositories of institutional memory and substantive expertise on whom the section relies for the performance of such key functions as preparing drafts of proposed legislation.^[6]

Ringisei

When a draft is prepared, it takes the form of a formal internal document known as ringisho . The circulation of such documents upward through the chain of command for approval is known as ringisei .^[7] Given its notoriety, ringisei merits further elaboration. Ac-

[5] Murakawa Ichiro, Seisaku kettei katei [Policy-Making Process], Gyosei kiko shirizu [Administrative Organization Series], no. 121 (Tokyo: Kyoikusha, 1979), p. 125.

[6] Ibid., pp. 126-29.

[7] The standard source on ringisei is Tsuji, Shinpan Nihon kanryosei no kenkyu , pp. 155-72; see also Tsuji, "Decision-Making in the Japanese Government: A Study of Ringisei ," in Ward, ed., Political Development in Modern Japan , pp. 457-75; Miyake, Gyoseigaku to gyosei kanri , pp. 315-34. Miyake's book contains an informative discussion of ringisei as practiced in the private sector, based on survey data.

cording to Tsuji Kiyoaki, it was first introduced by private business firms in the Meiji era but has since gained wide acceptance by public and private organizations alike. He notes three distinguishing features of the system. First, ringisho is drafted by a rank-and-file official (mattan no jimukan ) who lacks any decision-making authority. Tsuji suggests that even officials at the rank of subsection chief (kakaricho ) or below are sometimes given the responsibility.^[8]

Second, in principle ringisho is discussed and evaluated separately by the officials of the section, division, and bureau concerned; it does not necessitate a collective discussion in a conference of all relevant administrators. Third, although the head of the organization concerned-for example, the minister in a government ministry—has the sole legal authority to approve ringisho , it is customary for him to ratify the result of prolonged evaluation by his subordinates.^[9]

In Tsuji's view, ringisei has both advantages and disadvantages. On the credit side of the ledger, he lists the opportunity the system provides for participation in decision making by many persons, thus ensuring their cooperation at a later stage and the preservation of a record of decision making. On the debit side, Tsuji sees three problems. First, the system breeds inefficiency, because the process can take a long time. Sometimes delays are caused by the absence of the appropriate officials or by their negligence. At other times delays may stem from deliberate stalling by officials who, although unhappy about the document, find it difficult to propose revisions.^[10]

Another shortcoming Tsuji cites has to do with the diffusion of responsibility. Because officials other than the minister or the head of an organization see their role in the process as laying the groundwork for the final decision, they seldom feel a sense of responsibility for the outcome. As for the minister, endowed with neither expertise nor a staff that can provide an independent assessment, he usually ends up rubber-stamping the document; as a result, his sense of responsibility is equally weak.^[11]

[8] Tsuji, Shinpan Nihon kanryosei no kenkyu , pp. 155-56.

[9] Ibid.

[10] Ibid., pp. 158-59.

[11] Ibid., pp. 159-60.

Finally, Tsuji faults ringisei for stifling the exercise of leadership. When higher civil servants or their political superiors, such as a minister, feel inclined to take the initiative, they often find themselves stymied by the rigidity of ringisei . A policy proposal by a minister, for example, must first go to the lowest echelon and then travel upward through the ringi system, causing the usual delays. Any attempt by a higher official to short-circuit the process, according to Tsuji, triggers shock waves throughout the organization, sometimes resulting in his ouster. Tsuji mentions the case of Kobayashi Ichizo, who was forced to resign as minister of commerce and industry in the prewar period, after his attempts to introduce changes, bypassing the established modus operandi, had alienated his vice-minister, Kishi Nobusuke, and other bureaucrats of the ministry.^[12]

It appears, however, that either Tsuji overemphasizes the rigidity of ringisei or it has undergone significant change over the years. According to Ojimi Yoshihisa, a former administrative vice-minister of the Ministry of International Trade and Industry (MITI):

We consider the ringi document as part of office equipment along with pencils and paper. The ringi document is rarely written at the section-chief level. More commonly it is written by the person in charge of a certain matter below that level. . . . The simplest routine matters are simply written on a standard printed form as a ringi document. And at the other extreme, covering important matters, decisions are made before the ringi is ever written and passed around. Of course, ringi does serve to make the content of the decision dear, to put things on record, and to keep people in various parts of the organization mutually informed.^[13]

Ojimi also informs us that although there is reluctance to revise the ringi document as it moves up the ladder of bureaucratic hierarchy, some administrators do propose revisions from time to time. Moreover, he states that "there are no matters where the power of subordinates is greater than that of superiors. . . . Subordinates give their opinions freely and generously but there are no cases where the actions of the subordinates do not reflect the intentions of their superiors."^[14] From this account of how ringisei works, Ezra F. Vogel concludes that the "ringi system is often an ato ringi system, in which the leaders make the

[12] Ibid., pp. 160-61.

[13] Yoshihisa Ojimi, "A Government Ministry: The Case of the Ministry of International Trade and Industry" in Vogel, ed., Modern Japanese Organization and Decision-Making , pp. 109-10.

[14] Ibid., pp. 110 and 103. The quotation is from p. 103.

major decisions and then encourage lower levels to draft documents in line with the decision."^[15]

More compelling evidence underscoring the instrumental nature of ringisei is provided by Yung H. Park. Based on interviews with nearly two hundred former bureaucrats, politicians, cabinet ministers, and other individuals with first-hand knowledge of the situation, Park demonstrates that the notion that the ringi system gives subordinate officials "tyrannical powers" over superiors is a myth. He quotes, for example, Kobayashi Yukio, a former administrative vice-minister of the Ministry of Education, as saying:

A ringisho is prepared by a [section] only after careful consultations with its bureau director. During the consultations he may give his approval or instructions as to how the proposal on hand should be prepared or modified. . . . Thus, a ringisho is a written version of what transpires between the bureau director and the [section]. Hence it is not necessarily synonymous with the ideas originally entertained by the [section].^[16]

Kobayashi also told Park that a standard operating procedure is for the bureau chief to "consult with and seek approval from his superiors—[the administrative vice-minister] and, especially, the minister." The bureau chief also takes pains to clear the matter with the relevant organs of the Liberal-Democratic party (LDP), namely the division (bukai ) of its Policy Affairs Research Council (PARC, or Seimu Chosakai) having jurisdiction over the ministry.^[17]

Consultations With Other Ministries

While decision making is under way within a ministry, consultations occur not only between the ministry and the LDP but also between the ministry and other ministries and agencies. Because jurisdictional boundaries are seldom clear-cut and because most issues encroach upon the bailiwicks of several ministries and agencies, interministerial consultations and coordination are a sine qua non. All proposals for legislation and ordinances need a stamp of approval from the cabinet legislation

[15] Ezra F. Vogel, "Introduction: Toward More Accurate Concepts," in ibid., p. xvii. "Ato ringi" may be translated as "prearranged ringi ."

[16] Park, Bureaucrats and Ministers in Contemporary Japanese Government , p. 21. In his original text Park uses the word "division" for ka . I have substituted "section" for "division" to maintain consistency.

[17] Ibid., pp. 21-22.

bureau (Naikaku Hoseikyoku), which acts as legal advisor for the executive branch.^[18]

Any proposal for new policy requiring an expenditure of funds must also be cleared with the Ministry of Finance. Whereas consultations with these two agencies leave relatively little room for compromise, those with other ministries and agencies usually involve a symmetry of power. If compromise proves to be elusive, mediation by a third party may become necessary. Occasionally, the ministry asks its PARC counterpart in the LDP to assume the role of mediator. In short, the three most important external organizations in the decision-making process are the cabinet legislation bureau, the Finance Ministry, and the PARC division of the LDP.^[19]

Intraministerial Decision Making

With regard to intraministerial decision making, there are a number of mechanisms for reaching a consensus, such as bureau conference (kyokugi ), leaders' conference (kanbu kaigi ), interbureau liaison conference (kakukyoku renraku kaigi ) and ministry conference (shogi ). Although specific forms used vary from ministry to ministry, the aim of these conferences is to enable responsible officials at various levels to exchange ideas, iron out differences, and form a common outlook and approach.^[20]

In the Finance Ministry, bureau conference is subdivided into regular (kyokugi ) and "important" (juyo kyokugi ). The latter, which serves as the highest decision-making body of each bureau, is attended by the bureau chief, his deputy, section chiefs, and assistant section chiefs. Occasionally, even subsection chiefs and senior members of subsections (shunin ) are invited. To promote free discussion by all participants, one bureau is said to impose fines on those who fail to speak up.^[21] The regular bureau conference is presided over by the assistant bureau chief. The supreme decision-making organ of the Finance Ministry as a whole is leaders' conference. Held every Monday, this "summit conference" is

[18] On the role of the Cabinet Legislation Bureau, consult Kyoikusha, Kaikei Kensa-in, Jinji-in, Naikaku Hoseikyoku , pp. 119-63.

[19] Murakawa, Seisaku kettei katei , pp. 128-29.

[20] Ibid., pp. 132-37; Ojimi, "A Government Ministry," p. 102.

[21] Sakakibara, Nihon o enshutsu suru shin kanryozo , p. 56; Kuribayashi, Okurasho shukeikyoku , p. 53.

attended by the administrative vice-minister, the assistant vice-minister for financial affairs (zaimukan ), chiefs of the seven main bureaus (honsho shichikyoku ), and the chief of the minister's secretariat. Occasionally, the director general of the Tax Administration Agency (Kokuzeicho chokan), the chief of the Mint Bureau, and the chief of the Printing Bureau also join the select group. In addition, section chiefs of the minister's secretariat observe the proceedings. The difficult task of reconciling conflicting positions, however, is performed in advance of a leaders' conference through consultations and negotiations among senior section chiefs representing the various bureaus.^[22]

The Role of the LDP

When the minister officially gives his final approval, the policy proposal formally becomes that of the ministry. After the cabinet legislation bureau and the Finance Ministry have officially endorsed it, the proposal is submitted to the appropriate PARC division of the LDP. There are seventeen PARC divisions, each of which supervises one or more ministries and agencies.^[23] All members of the divisions are LDP Diet members, who are required by party rules to belong to at least one division.^[24] The chairmanships of the divisions usually go to LDP Diet members who have been elected at least three times, most of whom have previously served as directors (riji ) of Diet standing committees, parliamentary vice-ministers, or deputy chairmen of PARC divisions.^[25] According to Yung H. Park:

PARC divisional chairmanship is a key policy-making office. In the area of divisional jurisdiction, the chairman is one of the most powerful actors. . . .

[22] Jin, Okura kanryo , pp. 171-72.

[23] The divisions are (1) Cabinet, (2) Local Administration, (3) National Defense, (4) Justice, (5) Foreign Affairs, (6) Finance, (7) Education, (8) Social Affairs [Shakai Bukai ], (9) Labor, (10) Agriculture and Forestry, (11) Fisheries, (12) Commerce and Industry, (13) Transportation, (14) Communications, (15) Construction, (16) Science and Technology, and (17) Environment. Seiji handobukku [Handbook of Politics], February 1987 (Tokyo: Seiji Koho Sentaà, 1987), pp. 179-82.

[24] Inoguchi Takashi and Iwai Tomoaki, "Zoku giin" no kenkyu [A Study of "Tribal Diet Members"] (Tokyo: Nihon Keizai Shinbunsha, 1987), pp. 103-4. Because all LDP Diet members, except those serving as cabinet ministers and parliamentary vice-ministers, automatically belong to the PARC divisions that are counterparts of the Diet standing committees of which they are members, they are allowed to join two additional divisions. This means that most LDP Diet members belong to three divisions.

[25] Sato Seisaburo and Matsuzaki Tetsuhisa, Jiminto seiken [The Regime of the Liberal Democratic Party] (Tokyo: Kodansha, 1986), pp. 39 and 85-86. According to Inognchi and Iwai, most divisional chairmen have been elected to the Diet four times. See their "Zoku giin" no kenkyu , p. 121.

As far as the ministry under the division's supervision is concerned, he is the first and, most likely, the foremost party figure who must be persuaded in support of its major policy programs. Without his blessings, no agency policy can get party authorization. Hence he plays a major role in agency policy and budgetary processes and even in agency personnel decisions. He is a principal medium through whom the party influences ministry decisions.^[26]

As mentioned earlier, consultations between the ministry and the PARC division usually precede the formal submission of the ministry's legislative proposal, hence the deliberations in the division are sometimes a mere formality. The relatively large size of divisions also serves to inhibit an in-depth evaluation of legislative proposals by them. Some divisions have over a hundred members. As of March 1986, for example, the Agriculture and Forestry Division had 149 members. Substantive decisions are frequently made in informal meetings of Diet members that have become specialists in the subject matter at hand; known as zoku giin (literally, "tribal Diet members"), these include the chairmen of PARC divisions. Officials of the ministries concerned usually participate in such informal meetings.^[27]

Most of the deliberations in formal sessions of PARC divisions take less than one and a half hours. Thanks to painstaking preparatory work, known as nemawashi , the ministry's proposal is approved intact in nine out of ten cases. Because all of the division members also belong to the Diet standing committee having jurisdiction over the same ministry, their participation in this stage of policy making helps them become better prepared for the task of defending the bill against attacks by opposition Diet members in the legislative stage.^[28]

The PARC Deliberation Commission (Seimu Chosakai Shingikai), to which the policy proposal goes next, consists of about thirty members, including the chairman and vice-chairmen of the PARC. Whereas in the preceding stage the responsibility of defending the proposal rested with the bureau chief and section chiefs of the ministry, in the Shingikai phase the ball is carried by the chairman of the PARC division, who sometimes seeks the assistance of bureaucrats when questions become too technical or specific. In rare cases where a division of opinion

[26] Park, Bureaucrats and Ministers , p. 38.

[27] Sato and Matsuzaki, Jiminto seiken , pp. 91-93.

[28] Murakawa, Seisaku kettei katei , pp. 146-51. For an explanation of nemawashi , see Vogel, "Introduction: Toward More Accurate Concepts," pp. xxii-xxiii.

persists, direct negotiations between the PARC chairman and the minister concerned may become necessary.^[29]

The final hurdle within the LDP is the Executive Council (Somukai). One of the three officially designated "decision making organs" of the LDP—the other two being the party conference and the assembly of all Diet members—the Executive Council consists of thirty "executors" (somu ), of whom fifteen are elected by members of the House of Representatives (Shugiin), seven are elected by members of the House of Councillors (Sangiin), and eight are appointed by the president of the LDP. Given their prestige and power, the selection of somu , especially those appointed by the party president, is handled with great care; the need to maintain interfactional balance and other political considerations enter into the equation. Normally, only seasoned politicians, with at least four terms in the House of Representatives (fewer in the case of the upper house) and experience as parliamentary vice-ministers and committee chairmen in the Diet, can hope to win that honor.^[30]

The principal criterion employed by the Executive Council is political. Will the proposed policy benefit the interests of the LDP as a whole? By the time the proposal reaches the council, its need and technical feasibility will have been thoroughly examined and certified, hence the only question that remains to be asked is "will it fly?" Will there be any adverse political fallout? If so, will the anticipated benefits outweigh them? Participating in its deliberations are not only its members but also other leaders of the party—the secretary-general (kanjicho ), the chairman of the PARC, and the chairman of the Diet strategy committee (Kokkai Taisaku Iincho). The job of explaining and, if necessary, defending the proposal at hand is performed by the chairman of the responsible PARC division. Although most of the proposals are approved rather routinely, some become bogged down from time to time. When that happens, there are two options: either remand the case to the PARC or refer it to the weekly meeting of top party leaders (yakuinkai ), which is attended by the LDP president (sosai ), the vice-president, the secretary general, the chairmen of the Executive Council, the PARC,

[29] Murakawa, Seisaku kettei katei , pp. 152-54.

[30] Thayer, How the Conservatives Rule Japan , pp. 237-57. In Feb. 1987 the Executive Council consisted of twenty-eight executors. All eight who were appointees of the LDP President had one of three titles: president, acting president, and vice-president. Six had been elected by LDP members of the House of Councillors. Seiji handobukku , Feb. 1987, p. 179.

and the Diet strategy committee, and other leaders.^[31] Approval by the Executive Council is all but synonymous with approval by the LDP as a whole. The LDP Diet strategy committee deals not with any substantive issues but with those of procedure and strategy. To decide when and how to introduce the bill in the Diet, the committee hears reports on informal consultations with opposition parties as well as explanations from bureaucrats.^[32]

AVM Conference And Cabinet Meetings

The Conference of Administrative Vice-Ministers (Jimujikan Kaigi) takes up the bill after its approval by the LDP Executive Council. Convened twice a week, the AVM conference is attended by all the officials who carry the title of jimu jikan , namely, the highest career officials of the twelve ministries (sho ) and those of eight cabinet-level agencies (cho ). It is also attended by the deputy director general of the cabinet secretariat (Naikaku Kanbo fuku chokan), the director general of the Police Agency (Keisatsucho chokan), and the deputy director general (jikan ) of the cabinet legislation bureau.^[33]

Although the AVM conference is widely perceived as the real decision-making organ in the executive branch, its short duration belies its reputation. The conference, in the form of working breakfasts, usually lasts less than an hour. Like most meetings in Japan, it too is preceded by extensive nemawashi , which obviates the need for further negotiation. According to Tahara Soichiro, who interviewed a number of participant observers, the AVM conference has become a mere formality; one of his informants went so far as to call it a "completely skeletonized" thing (kanzenni keigaika shire iru ).^[34]

[31] Murakawa, Seisaku kettei katei , pp. 154-56. On the role of the leaders' meeting, consult Fukui, Party in Power , pp. 93-95.

[32] Murakawa, Seisaku kettei katei , pp. 157-58. On the role of the Executive Council, consult Fukui, Party in Power , pp. 89-92; on the role of the Diet Strategy Committee, consult Baerwald, Japan's Parliament , pp. 84-86.

[33] Tahara, Nihon no kanryo , pp. 9-13; Murakawa, Seisaku kettei katei , pp. 159-61. The eight cabinet-level agencies are Management and Coordination, Hokkaido Development, Defense, Economic Planning, Science and Technology, Environment, Okinawa Development, and National Land. The Management and Coordination Agency (Somucho ) was established in July 1984 by merging part of the Prime Minister's Office (Sorifu ) with the Administrative Management Agency (Gyosei Kanricho ). See "Kasumigaseki toppu no kao: Jimu jikan" [Faces of the Top Men in Kasumigaseki: Administrative Vice-ministers], Kankai , Sept. 1985, an unpaged pictorial essay.

[34] Tahara, Nihon no kanryo , p. 13.

Officially, the highest decision-making organ in the executive branch is the cabinet meeting (kakugi ). There are three types of cabinet meetings: (1) teirei (ordinary), (2) rinji (extraordinary), and (3) mochimawari (literally, carrying around). Ordinary cabinet meetings are held twice a week, usually on Tuesdays and Fridays, although their frequency may be reduced during the summer. Extraordinary cabinet meetings are convened whenever there is urgent business. Finally, mochimawari kakugi is not really a meeting, for it entails the dispatch of officials to all members of the cabinet to collect their signatures on a document. Two things make such a procedure both necessary and feasible: the lack of time to call an extraordinary meeting and the relative unimportance of the issues involved.^[35]

In addition to all members of the cabinet, who number around twenty, the two deputy directors general of the cabinet secretariat (naikaku kanbo seimu fuku chokan and naikaku kanbo jimu fuku chokan ) and the director general of the cabinet legislation bureau (naikaku hoseikyoku chokan ) also attend cabinet meetings, although the latter three sit apart from cabinet members in a corner of the conference room. By custom, a veil of secrecy surrounds cabinet meetings. Not even "photo opportunities" are allowed in the cabinet room itself. Nor are official minutes kept. A gist of what has transpired, however, is released to the press by the director general of the cabinet secretariat. The rule of unanimity governs the meetings, including mochimawari kakugi .^[36]

Although, as a rule, only those matters that have been finally approved by the LDP Executive Council come before the cabinet meeting, once in a while matters that are still in the LDP pipeline may be taken up to expedite the decision-making process. Such matters are considered in the form of "items outside the agenda" (ankengai ) and given provisional approval.^[37] Given the overlap in the memberships of the cabinet and LDP policy organs, there is only a slight chance that policy proposals that have survived the rigors of LDP scrutiny will generate a fresh controversy and unanticipated problems in the cabinet meeting. In a sense, what happens in that forum is anticlimactic, for all the hard work has already been completed, and all the significant battles have been fought. If the stakes are high, the senior politicians that make

[35] Asahi shinbun , 28 July 1985.

[36] Ibid.; Naikaku Seido Hyakunen-shi Hensan Iinkai, ed., Naikaku hyakunen no ayumi , pp. 22-23; Murakawa, Seisaku kettei katei , pp. 161-63.

[37] Ibid., pp. 162-63.

up the cabinet will already have made their inputs in the preceding phases of the policy process, in their capacities as ministers and as party leaders.