Japanese Budget Baransu

John Creighton Campbell

All countries draw up budgets to allocate governmental resources among national needs. Since needs always exceed resources, within every government conflict will occur between spending agencies and the organization assigned responsibility for enforcing restraint. Moreover, certain characteristic budgeting behaviors appear to be common everywhere, at least among the richer countries. Aaron Wildavsky, in his The Politics of the Budgetary Process , calls these behaviors "incrementalism"; he sees American budgeting as historical, specialized and fragmented, nonprogrammatic, sequential and repetitive, and conflict minimizing. More recent research confirms that these observations are valid elsewhere.^[1]

Still, even if the broad outlines and procedures of budgetary systems are similar, it would be surprising if they turned out to be identical in detail. Budgeting, after all, plays a major part in "the authoritative allocation of values for a society," and so lies at the heart of a nation's political process. As politics among rich countries differ so must their budgetary systems. What might be the critical determinants of differences among budgetary systems? Those which immediately suggest themselves are differences in national economic conditions, formal or informal political structures, and political culture, particularly elite political culture.^[2]

In the course of interviewing Japanese budget participants for a forthcoming study,^[3] I became aware of a notion often expressed by the word baransu , a Japanese transliteration of the English word "balance." The word did not refer to equalizing governmental revenues and expenditures, as in American budget terminology; it seemed to mean evenhandedness, or equitable treatment. I also had occasion to examine Japanese budgetary statistics and found that (in comparison with similar American data) these indicated a very balanced pattern of allocations over time. That is, compared with the United States budget, the shares of the budget commanded by each ministry did not vary much from year to year.

These differences between Japan and the United States are expressed in Table 1, which shows average deviations of agency budget growth in one year from the growth of the total budget in that year. A zero would mean that in every year the agency's budget grew at precisely the same rate as the total budget. Institutional arrangements and budgetary classifications are different in the two countries, so this table must be used with caution, but it does seem to demonstrate that Japanese budget allocations do not fluctuate very much. It is as if the new revenues that become available each year are shared out more or less evenly so that each ministry gets about the same proportion of the new as it had of the old. The equilibrium of the previous year is maintained; no "unbalance" is introduced.

This paper is primarily an effort to relate this balanced pattern of budgetary outputs with the norm of baransu or evenhandedness observed among those who draw up the Japanese budget. I will first introduce the Japanese budgetary process, and discuss the meaning of the baransu concept, and then demonstrate how balance is manifest in the budgeting behavior of the three major participants in the process. The conclusions will offer some explanations of why balance ideas are so strong in Japan and suggest some possible effects of this behavior. However, it should be understood throughout that the Japanese budgetary system is still quite similar in many (perhaps most) respects to that of the United States and elsewhere. I have chosen to emphasize differences rather than similarities because they are more interesting. Among the several differences, including Japan's extraordinary economic growth and the unique role of the majority political party in budgeting, balance as an element of elite political culture is given special attention here for two reasons. First, since it shows up throughout the process, tracing balance offers an opportunity to survey many other important aspects of Japanese budget politics conveniently. Second, it seems likely that budget balancing is closely related to similar traits in other forms of Japanese organizational behavior.

The Japanese Budgetary Process

Under the constitution, the annual budget is to be prepared and sent to the Diet by the cabinet, but provisions of the Finance Law assign the responsibility of actual preparation to the Ministry of Finance (MOF) and its budget bureau (Shukeikyoku).

The Japanese fiscal year begins on April 1, so fiscal-year 1970 ran from April 1, 1970, through March 31, 1971. The process begins shortly after the beginning of each fiscal year, when the various ministries and other agencies (henceforth referred to as "the ministries") begin formulating their budget requests for the following fiscal year's budget. In recent years, MOF regulations have required each ministry to limit its request to 125 percent of the budget granted in the current year. To fit the demands of its own sections and bureaus within this limitation, each ministry undergoes a mini-budget process of its own in the summer months. The ministry requests, which usually amount to nearly the maximum permitted, are submitted to the MOF about August 31, after brief hearings that secure the general agreement of the appropriate Liberal Democratic Party (LDP) Policy Affairs Research Council (Seimu chosakai^[*] , or Seichokai^[*] ) divisions (bukai ).

During the month of September, ministry officials appear at the Budget Bureau to explain these requests to the budget examiners (shukeikan ), who ask questions about program details but do not actually negotiate over budget figures at this stage. In October and November, the grinding work of budgeting is performed within the Budget Bureau as examiners and their staffs go over the requests in detail, discussing possible options with the responsible vice-director (there are three), the director, and other bureau officials.

At the same time, the Budget and Tax Bureau directors, ministry-level staff specialists, the administrative vice-minister, and other MOF officials discuss the total budget figure and other "macrobudgeting" questions. These deliberations are based primarily on an estimate of tax revenues and considerations of fiscal policy in regulation of the economy. Although final decisions do not come until mid-December, rough estimates of the funds available are in the Budget Bureau director's mind as he formulates the bureau's posture toward expenditure requests.

For about two weeks in early December, the MOF Ministerial Budget Conference meets to ratify, and sometimes modify, the draft budget prepared within the Budget Bureau, which is then called the "Finance Ministry draft." This series of meetings encompasses the entire chain of command from minister to examiner (and sometimes his assistants), along with other MOF bureau directors and staff. At the same time, the Budget Compilation Policy (Yosan hensei hoshin^[*] ) is drawn up within the Budget Bureau and ratified at the ministry level. This brief and highly abstract statement sets forth the basic principles of the budget, and is passed by the

cabinet at about the same time the MOF draft of the budget is released—despite the fact that, in theory, it is supposed to be the major cabinet-set policy to guide the MOF in compiling the budget.

During October and November, the regular organs of the LDP do not intervene in the compilation process going on within the Budget Bureau, although individual Dietmen will often let their preferences be known (for example, by accompanying groups of petitioners to the Finance Ministry). In December the Policy Affairs Research Council's Deliberation Commission (Seisaku shingikai or Seicho^[*] shingikai) draws up the annual LDP Budget Compilation Program (Yosan hensei taiko^[*] ). This program is then passed by the Party Executive Council (Somukai^[*] ) and referred to the cabinet and the MOF shortly before the release of the MOF draft.

The release of the MOF draft—scheduled for mid-December, though in recent years it has often been postponed until after the New Year holiday—begins the "revival negotiations" (fukkatsu sessho^[*] ) period of (since 1962) about a week. It appears that 3 to 5 percent of the total General Account^[4] is allocated at this time in response to ministry appeals on MOF cuts and the final demands of the LDP. The latter, called political demands, are formulated in a brief process of LDP Research Council division hearings and recommendations, which are reviewed and adjusted by the Research Council Deliberation Commission and the Executive Council, and then implemented by the leadership through participation in the penultimate cabinet-minister-level negotiations and a final session between party and government top officials.

The budget, as modified, is called the "government draft." It is ratified by a cabinet meeting called immediately after the negotiations with the party and then sent up to the Diet. The opposition parties use the legislative budget deliberations as an opportunity for criticism and questions about government policies, but the budget itself is not amended. After passage, the MOF oversees the execution of the budget, and at the end of the fiscal year, there is a budget settlement and an independent audit by the Board of Audit. This study, however, is limited to the period that ends with the cabinet passage of the government draft.

Balance

We may approach the somewhat amorphous idea of balance through examining its formal usage. Kono^[*] Kazuyuki, a former vice-minister of finance, said that balance—meaning the comparative balance (hikaku kinko^[*] ) among expenditures—is one of the two criteria used by the Budget Bureau in reviewing budgets.^[5] Expenditures in the same category, or

similar items, should particularly be balanced, but the balance criterion should also be applied more widely. Similarly the economist Kotake Toyoji identified five criteria for budget review; "balance" meant relating expenditure levels to those of similar or related items.^[6] Such explanations indicate the importance of the concept, but do not impart much of its content. When asked to define balance, a high Budget Bureau official said it means a round budget (marui yosan ), one without "bulges" of too much money given to any one program, policy area, or agency.^[7]

To balance a budget item one compares it with other items; it is in balance when it is not receiving too much, or too little, relative to other items. A similar idea in American budgeting is "fair share," a "convergence of expectations on roughly how much the agency is to receive in comparison to others" when the total budget is increased.^[8] What balance is emerges more clearly by considering what balance is not: it is not setting an expenditure level by assessing how important a program is to national priorities; how effective it has been in the past; or even how much political support it has attracted. In particular, it is not a comparison of two similar or otherwise related programs to see which is better—which produces "more bang for the buck" in a cost-benefit analysis. Quite the opposite—balancing represents avoidance of comparisons among programs and their merits by implying that simply because they are similar they should receive the same or equivalent budgets.

What determines whether two items are similar and require balance? Kono^[*] gives three examples. The simplest is that levels of benefits among the various government-supported health insurance systems should be balanced. Second, health insurance in general should be balanced against other social welfare programs like unemployment insurance and public assistance. Finally, compensation in the "postwar settlement" to groups like veterans, war-bereaved families, or those who lost their homes or overseas property should be equitable.^[9] Of course, the question of what should be balanced—for example, individual benefits to members of these groups or the budget allocation for the group as a whole—is left open. From an administrative point of view, routine expenses will nearly always be balanced throughout the government, so a ministry will be assigned automobiles by its number of employees or maintenance expenses by square feet of office space. At the opposite extreme, some people hold that entire policy areas, such as social welfare and defense, require balance.

Since each of these expenditure items is assigned to a particular administrative organ, balance may also be seen as equitable sharing among real organizations, whether sections or ministries.

Clearly, balance is not a tightly defined administrative rule. It is best understood in the first place as a value widely shared within the Japanese elite political culture, one that is well understood in principle even if argued about in application. Balance in this sense is an end, not a means. We might say that budget participants simply feel more comfortable when the budget seems to be balanced. However, the strength of this value also gives it two important instrumental functions: in Wildavsky's terms, it is a helpful "aid to calculation" which eases the massive decision-making burden on budgetary participants;^[10] and it becomes a major strategic resource for all participants. Finally, the relative strength of this balance value means that other values—efficiency, responsiveness, perhaps even effectiveness—must be relatively weak. Thus, when the Japanese budget system is evaluated in terms of these other values, balance will be found to have dysfunctional aspects.

Balance for the Ministries

The ministries (plus some agencies) are the most salient level of administration in Japan; bureaus may be powerful but lack the coherence and independence typical in the United States. Ministries are large and complicated enough that even internal balance considerations are significant. They appear particularly when budget requests are decided. Since the request for each ministry must not exceed 125 percent of the budget actually received in the previous year, constituent bureaus cannot be permitted simply to ask what they wish. Either a similar percentage limitation must be imposed on each bureau within a ministry, or a miniature budget review process of cutting down bureau requests will be required. Although one informant said that the Ministry of Education followed the first practice (an extreme of balancing, and very stable), the latter is more common; most ministries impose either a softer limitation or none at all.

However, even under the latter pattern, the ministry-level accounting section (kaikeika ) responsible for budgeting will often go to great lengths to maintain balanced treatment of bureaus. For example, a study of the Ministry of Agriculture and Forestry budget process in the early 1960s turned up the following set of guidelines for compilation:

Except for disaster reclamation projects, each public works item is adjusted so that the budgets of the Agricultural Land Bureau, the Forestry Agency and the Fisheries Agency have equivalent growth rates from the initial budgets of the

previous year. Disaster reclamation projects are adjusted restrictively so that the same rate of progress will apply to all.^[11]

A mechanical formula like this obviates the need for the ministry's central staff to make value choices among programs, and therefore tends to dampen conflict among the bureaus. Agriculture is a ministry with relatively powerful bureaus. In others, the ministry-level staff (particularly the administrative vice-minister and deputy vice-minister) may have more influence over the bureaus and impose its own priorities on budget requests. Proportionally larger budgets may be requested for those programs that the ministry leadership thinks are most important or will look most attractive to the MOF.^[12] The composition of budget requests from such ministries may therefore fluctuate more from year to year than those of more balanced ministries where the bureaus are more powerful.

Once the requests have been settled internally, the ministries turn to the task of obtaining as many funds as possible from the MOF.^[13] Over the years they have developed many strategies for this. In the main, perhaps to a surprising extent, the repertoire of Japanese ministerial budgeting strategies closely resembles what is found in the United States and elsewhere; included are tactics using clientele groups, "camel's nose" techniques, substitutions and intricate accounting devices, the occasional lucky "crisis," and so forth.^[14] Notable, however, is the extent to which the ministerial budgeter relies on the value of balance in arguing before the MOF. That is, rather than justifying a program or policy in terms of intrinsic merit or national need, he will claim that their item is entitled to the "same" level of effort as some other item.

On the most straightforward level, programs that are very similar in objectives and target groups will be balanced almost automatically. When benefits for one of the government-financed health insurance plans are raised, the Ministry of Health and Welfare will not have a difficult time in appealing for increases in the others. Similarly, increased support for national universities will occasion an appeal to the MOF to raise the government's contribution to private universities. When extended further, the demand for balance becomes a little more tenuous; an increase in college-level funds brings arguments for raising secondary-school support (or vice versa).^[15]

If the ministry hopes to receive more funds than can be justified by balancing with current expenditure levels of similar programs, it will have to go farther afield. For example, a Ministry of Education official, commenting on the budget for per-student allocations to private universities, said: "The MOF is usually tough on this item and wants to give us just an increase of 5 percent or so to cover cost increases. Therefore, we usually compare the proportions of the aid given today with prewar figures [for the same program] to show that this is much lower now."

Rather more common than citing prewar figures is the "keeping up with the Joneses" ploy—low Japanese expenditures in a given field are compared with those overseas. The references chosen vary with the program being emphasized: scientists seeking more research funds look toward American government-financed research and development programs; the Ministry of Health and Welfare points out the higher levels of social security payments in England and Scandinavia; the Ministry of Agriculture and Forestry defends its rice-price support system by noting that all industrialized countries subsidize agriculture.^[16] Here, rather than seeking balance among specific programs, policy areas, or organizations, the appeal is to a fundamental sense that Japan as a nation should be in balance with its industrialized peers. Undoubtedly, the effectiveness of such arguments is due in large measure to Japan's century-old tendency toward using Western nations as both sources of innovation and bench marks of achievement.

When dealing with broader policy areas, appeals to balance are not as inherently plausible as they are for smaller programs with near-equivalent counterparts. Attempts by the Ministry of Health and Welfare to defend its budget provide an example. In the mid-1950s, it held that giving larger increases to defense than to social welfare would constitute "remilitarization," but when the defense growth rate slowed enough to make this argument unprofitable, welfare spokesmen began, instead, to compare their programs with the public works sector (on the general grounds that people are more important than things). It is interesting that the MOF did not respond to this ingenious rationale by arguing that public works and welfare are apples and oranges, essentially impossible to compare; rather, in 1961, the MOF manipulated its budget categories, adding enough programs usually counted as public works into the welfare total to balance the two sectors, at least rhetorically.^[17]

If a ministry cannot make a strong case for a balance in spending for its own policy area with that in another specific field, it invokes the more general notion of budget "share." Unless it can be proved that an activity has declined in importance during the preceding year—and the MOF is rarely venturesome enough to make such statements publicly—the ministry will argue that there is no reason it should not receive the same share of the budget (or perhaps of GNP) that it had previously. In other words, the item should have the same growth rate as the total budget. Such appeals are obviously not useful to ministries like welfare, which are trying to enlarge their budget shares, but for a ministry whose programs are under attack, "fair share" offers an attractive defense.

The most obvious example is the Ministry of Agriculture and Forestry, which despite strong support from the LDP has had to face a long-term decline in the importance of agriculture in Japan's economy. In the mid-1950s, when the total budget ceiling was held constant, this ministry's budget allocation was actually cut for three consecutive years. When growth resumed, ministry spokesmen tried to maintain the "10 percent principle"—that one-tenth of the budget should go to agriculture. Ikeda Hayato, newly installed as prime minister in time for the 1961 budget process, tried to initiate a gradual cutback in the proportion of national resources going to agriculture, but met with vehement protests from the agricultural ministry and its supporters in the LDP; in the end, an enormous grant to the rice-price control system had to be made during the appeals negotiations.^[18] The growth rate in agricultural spending again dropped slightly below that of the budget as a whole in the 1965 MOF draft, and the protest from the same quarters was immediate.^[19] Such "share-consciousness" gave agriculture and its supporters a reputation for caring little about programs, for desiring only to grab as much of the budget as they could regardless of how the money would be spent.^[20] Rather similar was the response of the Ministry of Home Affairs to a cut in local taxes scheduled for 1960 (a time when the financial position of local governments, which this ministry represents, was relatively strong): "Losing this tax revenue was like a man getting hit by a car. Even if the victim is wealthy, he still deserves a solatium."^[21]

The Defense Agency might provide a good case study of the helpfulness of balance ideas in a ministry's budget strategy. This agency finds it difficult to formulate objectively convincing budget proposals because it seems unlikely that the self-defense forces will be at war in the near future, and in any case they are limited to defending the Japanese main islands against conventional attack.^[22] The military requirements for this task do

not rise very rapidly. Furthermore, political pressures prevent the agency from postulating even a hypothetical enemy, so its planning must be vague, and officials cannot use the familiar Pentagon ploy of racing to keep ahead of real or imagined Soviet military advances. Therefore, rather than basing budget demands on some measure of the exterior threat to the nation, supporters of more defense spending turn attention to the proportion of the national budget or GNP going to their sector. Stressing the point that this percentage has declined over time (until 1969) helps to obscure the rapid rise of the defense budget in absolute terms—because of high GNP growth, it has doubled about every six years.

This "fair share" concept is useful for keeping the defense budget growing at about the same rate as the national budget, but it does not provide an argument for major increases. Hence, defense industrialists and others seeking a large increase in military spending use another balance technique: the proportion of GNP going to defense in Japan compared with the much higher figure for the United States and other advanced countries. Such appeals—to raise Japan's defense GNP share from around 1 percent to 2 percent—are heard far more often than, say, arguments that a new weapons system is vital or that the mission of the forces should be extended.^[23] After an interview in which this percentage of GNP criterion had been mentioned several times, a Defense Agency official was asked why—he conceded that it had little real meaning but "is the only objective indicator available."

The Defense Agency example displays all three functions of balance. Deciding a "proper" level for defense expenditures through more "objective" indicators would be most difficult, and so balancing (as a share of the budget, or as a comparison with other countries) offers a far simpler means of calculating budget requests. Just as important, however, is the underlying agreement within the budget system that balance is an important value in itself. Appealing to this norm thus became a helpful budget strategy for the agency.

Balance for the Liberal Democratic Party

Japan is probably unique, at least among capitalist nations, in the degree to which a political party organization directly penetrates the budget process. One or another segment of the LDP structure plays a part in nearly all budgetary decisions from the lowest to the highest levels. Still, the Party's real influence (in the sense of actually modifying budget outcomes) is not as extensive as might be imagined. This is largely because

the LDP's internal processes are so influenced by balance considerations.

Though many have viewed the LDP as "a loose coalition of factions,"^[24] it may also be seen as a coalition of many other groupings, which might be regional, interest based, or policy oriented. The devices that maintain the coalition vary depending on the salient issues of the moment, but many of them boil down to the principle enunciated by Dietman Fukunaga Kenji with respect to geographical balance in the Executive Council: "If all regions are represented, then no regions are overlooked."^[25] To the extent the Party adheres to this principle, its capacity to favor one region or interest over another is, inhibited; choosing among competing alternatives becomes more difficult. Such choices are the essence of budgeting.

There are a number of points for LDP entry into the budget process: individual dietmen seek projects on behalf of constituents; members band together in interest-group-sponsored "Dietmen's Leagues"; influential politicians obtain favors through personal or factional connections; and Party leaders participate in informal high-level discussion on top macro-budgeting issues like the size of the total budget. An important influence, though one hard to measure, is the review of ministerial requests by the divisions of the party's Policy Affairs Research Council in late summer. Although the requests themselves are not often modified significantly after these sessions, the fact that they are discussed, and in effect approved, by the division must be on the minds of the ministry officials who draw them up. Divisional views are also passed along to budget examiners by ministry officials. These and other fragmented interventions may be influential in one segment of the budget, but none has the potential for becoming a comprehensive, fully considered, official, and "rational" party plan. Our attention, therefore, focuses on formal intraparty budget processes, since these are (or at least once were) designed to impose the LDP's will on the budget as a whole.

There is an ideal view of how the party should participate in budgeting, endorsed by many politicians, officials, analysts, and newspaper editorial writers. They agree that although democracy allows an important voice for a majority party in budgeting, it should not intervene in the petty details of administration. The LDP should decide broad priorities for the budget each year, and identify policy innovations and new opportunities for governmental action. Procedurally, this ideal might mean that each division should critically examine detailed programs and plans, recommending which should be initiated or increased and which should be cut

back or abolished. These divisional recommendations should be discussed objectively and from a "national" point of view by the Research Council's Deliberation Commission and the Executive Council. Their report on the highest priorities for the year would then be sent to the cabinet and the MOF early enough to guide the process of budget compilation.

At the creation of the LDP in 1955, many thought that this ideal of party participation in budgeting would be implemented. Business leaders hoped that party unification would bring a strong administration, able to restrain the mongering of popular programs typical of the earlier multi-party system; even some MOF officials predicted that more rational and priority-based policies would replace logrolling of politically favored items through Diet amendments.^[26] However, participants were uncertain about what relationship of party to budget would emerge. Guesses ranged from a party takeover of the entire process to complete MOF independence, and even included a proposal to move the Budget Bureau from MOF to cabinet control. The first budget made up under the new party—that for 1956—proceeded through close consultations and cooperation between LDP and MOF leaders, and the Research Council's Deliberation Commission passed a very specific, yet rational and responsible set of budget recommendations. However, this cooperation broke down because pressure from rank-and-file Dietmen, demanding a few more yen for favored projects, could not be controlled by the party leadership. In the end, the leadership was forced to demand additional budgetary "intraparty adjustment expenses" to quiet down their followers, thereby violating earlier promises to the MOF and severely damaging the credibility of the LDP as a budgeting agency.

Budgeting means cutting; requests always add up to more than the resources available, and the purpose of budgetary mechanisms is to reduce these requests to manageable size. Because it controls the cabinet and the Diet, if the LDP within itself could have drawn up an implementable budget plan, it would have absolutely dominated the budget process. Alternatively, even granting that the enormous informational requirements and decision costs of item-by-item budgeting might make LDP production of a fully detailed draft impossible, concrete statements of priorities plus a realistic selection of new programs were well within the party's technical capabilities. Such a program was attempted in 1956 (and a few times afterwards), but floundered on the party's inability to control itself. In 1956 and repeatedly thereafter, the LDP demonstrated that it could not aggregate the interests and opinions of its constituent parts effectively and then enforce its final decisions. This fundamental problem is seen clearly in the LDP budget process as it appeared toward the end of the 1960s.

The building blocks of the party's official Budget Compilation Program

are the reports of the Research Council's fifteen divisions. These divisions correspond generally to the ministries, and range in size from 44 members in the foreign relations division to 155 in agriculture and forestry.^[27] Many members serve at their own request. Division-ministry relations are usually very close because of a number of factors: the many ministry ex-officials influential in the divisions; the ability of officials to implement a Dietman's policy ideas and aid his constituents; the importance of LDP contacts for an official's future career both inside and outside his ministry; and simply the ties that grow naturally from long association.

It is sometimes fruitful to regard the LDP organization as the real legislature of Japan; from this point of view, the Research Council divisions are equivalent to American legislative (not appropriations) committees. They are intimately concerned with the affairs of their corresponding ministries and ordinarily support their interests, inside and outside the party, against the competing claims of other agencies. According to ministry and party informants, divisions seldom oppose any portion of a budget request, and independent new suggestions from them are also rare. They commonly wish to see even more spent than the ministry has requested (though the 125 percent limitation normally prevents these desires from being included in requests), particularly for items with great local-constituency, interest-group, or popular appeal. During the budget process, divisions will often help coordinate tactics among interest group representatives, supporting Dietmen and ministry officials, and generally act as "cheering sections" for their ministries.

There are exceptions to this pattern. Divisions have been known to disagree sharply with their ministries on small or large points. Still, even in these cases, all divisions always call for more expenditures within their policy sectors. Statements such as "our policies x are important, but expenditures here should be deferred this year because policies y are still more critical" are never encountered. Therefore, if a comprehensive plan is to be prepared, the ministry-by-ministry reports must be aggregated by an upper-level body in the LDP—the Research Council's Deliberation Commission. This commission is made up of fifteen to twenty members with "impressive credentials";^[28] it meets several times in early December to evaluate written and oral divisional recommendations and to draw up a draft of the Party's Budget Compilation Program. This program is then ratified—sometimes after a few deletions, additions, and changes—by the Executive Council.

An examination of the Budget Compilation Program for the 1970 budget will give an idea of the character of these documents. It is divided

into two unequal parts; the first, called "basic policy" (kihon hoshin^[*] ), has seven items.^[29] Six are unexceptional slogans calling for growth without overstimulation, tax cuts, elimination of supplementary budgets, slowing inflation, administrative savings, and healthy local administration. Number 6 lists expenditure priorities: "emphasis" on social capital, comprehensive agricultural policies, promotion of small business, environmental problems and traffic safety, and reductions in nonessential items. The second part, "important policies" (juten^[*] jisaku ), is over ten times the length of the first part and has fourteen major headings (for example, "The Renovation of Education and the Promotion of Science and Technology"). Each is followed by several items, which in turn have subitems; the total number of entries not further subdivided is 116. The translation of one will give the flavor of this document. Under the heading "Promotion of Agriculture, Forestry, and Fishing," one of the three subitems under item 3, "Promotion of Fishing," reads:

Along with accelerating fishing-harbor construction, seashore-preservation works, and repair and construction of roads to fishing harbors, the establishment of large-scale underwater rocky areas for fishing, the development of new fishing grounds and structural reform works for coastal fishing will be systematically promoted.

Not all are quite this broad, but it will be appreciated that the 116 entries can cover a lot of ground. Indeed, most governmental programs are included somewhere in this document.

Thus the Budget Compilation Program is comprehensive, and because it is the product of detailed hearings followed by due deliberation and ratification by official party bodies, it may also be called exhaustive and authoritative. On the other hand, is it practical and influential? No, for at least three reasons: (1) No monetary figures are attached to the recommendations. For the most part, the document simply urges that each program be "stressed," "completed," or "strengthened." (2) Final passage of the program comes only two to four days before the MOF draft budget is released, too late to guide the compilers.^[30] (3) There is virtually no indication of priority among the programs recommended. The fourteen headings under "important policies" provide a shopping list in almost arbitrary order. It is true that six expenditure programs are singled out for attention in item 6 of the "basic policy" section, but this is not at all a plausible list of the party's priorities for the 1970 budget, since other evidence indicates that no unusual attention was given to social capital or small business that year, and the themes of pollution and traffic safety

were hardly mentioned again by LDP budget participants (agriculture did get extra attention in 1970, but this item is included in the "basic policy" list every year). And needless to say, the Party made no efforts toward reducing expenditures on nonessential items; quite the opposite.

If it is true that the LDP potentially could control the budget process through this program, why has it not done so? And if the document has no real effect, why is it still produced so laboriously? The answers to both questions have to do with "balance." Everything is included in the Compilation Program simply because then nothing need be left out, and each division's desires are represented. The policies are not listed by priority because this would represent a judgment by the Deliberation Commission or Executive Council that one division's recommendations are inferior to another's. Money figures are not included partly because, if they were cut, there would be opposition from the divisions, and if they were not cut the total would be embarrassingly large, well beyond any possible budget. (It should also be noted here that the LDP does not have the staff capability to prepare a specific "alternate budget," and that the MOF has pursued a consistent strategy of inhibiting other participants from discussing actual amounts of allocations.) Commission and council members, and LDP Dietmen generally, realize that the practice of making the program vague, all-encompassing, and late prevents it from having any influence, but have implicitly decided (after several attempts) that trying to reform the process is useless.

However, survival of the Budget Compilation Program should not be regarded as merely vestigial. It does have real functions. One is public relations: the document is given wide circulation and is designed to show both the general public (note the pollution and traffic safety items) and specific interest groups (agriculture, small business, local governments, and almost all others) that the party has their interests at heart. Perhaps more importantly, many Dietmen actively participate in this process and have their views taken into account in the final document. Finally, blame for omission of any recommendation in the final budget can be laid to the MOF, not the party, thereby mitigating resentment against the leadership. The "balance" of this symbolic output therefore functions to maintain satisfaction among party members and clientele.

However, not all products of formal LDP processes are merely symbolic in impact. At the later "revival negotiations" stage, under much more hurried conditions and without gestures toward public opinion, a truly influential set of recommendations is brought forth. This process begins shortly after the release of the MOF draft budget, when divisional members talk over the MOF cuts and possible responses with ministry officials. The divisions also receive copies of ministerial request forms, which they label with A, B, or C to indicate priority. Then, in a hectic, one-day series of hearings, the Research Council's Deliberation Commission listens to spokesmen of all the divisions (plus many related special

committees and investigatory commissions) and marks these forms with its own symbols. For the most part, divisional recommendations are marked with equivalent symbols by the commission, preserving balance. However, some are marked with the marusei (a circle containing the character sei , which participants say refers to either seiji , politics, or seisaku , policy). This symbol indicates an item of party concern that should be taken up at the political level—either the cabinet minister negotiations (at which the Policy Affairs Research Council chairman is present) or the final top-level LDP-MOF talks. In 1970, about one-fifth of the items marked "A" by the divisions also received the marusei label. They represent real choices, true budgeting behavior, rather than the more usual evenhanded sharing out of benefits.

To support this point we may examine the contents of the thirty-nine marusei recommendations for 1970.^[31] Their distribution by ministry was very uneven: twelve went to the Ministry of Agriculture and Forestry, nine to the Transport Ministry, and although the Ministry of Construction received only four, they covered 80 percent of its budget. The remainder was scattered much more thinly among the other ministries. Over half were for various types of public works. This pattern reflects the LDP's interest in programs with great appeal to local constituencies, particularly in rural areas. Notably absent were programs that drew much broader, but more diffuse, public interest—for example, pollution and traffic safety, emphasized so strongly in the earlier Budget Compilation Program, were completely ignored.

Indeed, these revival negotiations recommendations are not publicized by the Party and are written in a budget shorthand nearly unintelligible to outsiders. Few in number and quite specific, the marusei items are taken seriously by the ministries, both the rank-and-file leaders of the Party, and MOF budgeters. The LDP here sacrifices balance for effectiveness, symbolic image-creation for a real role in budgetary decision-making. Not that balance is completely neglected: virtually all divisional recommendations are formally endorsed somewhere in the Deliberation Commission's report, and when two divisions propose contradictory items, as commonly occurs when ministries battle over which will control a new activity, the commission may well grant the marusei impartially to both. At least eight of the thirty-nine marusei items for 1970 represented such overlaps, and two remained even after the report had been tidied up by the Executive Council on the following day.

The last point reflects on a claim often put forward by defenders of the

LDP's role in the Japanese policy process, that the Party can effectively intervene when two ministries are at loggerheads, or on behalf of new activities that fall between the jurisdictions of two or more ministries. At least in the budget process, which is the major stream of domestic policy-making, such conflict resolution seems to occur rather rarely. Much more commonly, divisions back up their ministries, and the two aggregating bodies will lack the capacity to violate balance by choosing one over the other or by imposing an independent solution.^[32]

More generally, party spokesmen often talk of party "dynamism" and flexibility, contrasted to the stodgy conservatism of the bureaucracy. This comment by an LDP leader's staff director in an interview is typical:

Bureaucrats are awfully good at riding along on an old policy, like a locomotive driver, but they can't change direction. This is because of organization . . . because of organizational ties, they can't start something new. The party, on the other hand, doesn't have to worry about the organizations fastened to [old programs] and instead can ask, what is the problem for the people? What did they ask for in the last election? So the party can become the engine to change policy.

Such statements are not devoid of insight; the party can sometimes free itself from the domination of pressures from below. However, its ideas usually originate in the ministries, and it rarely presses for programs against ministry interests. Its strongest positions are taken on those programs with the closest ties to the votes of specific groups (farmers, pensioners, local officials) and do not change very much over time.

Furthermore, the negative and inhibiting effects of party influence in budgeting must not be ignored. From the MOF point of view, it is party support for wasteful or "backward-looking" programs—such as agricultural price supports and deficit-ridden local railway lines—that prevents a shift of resources into newer and more pressing fields. In its 1968 "break fiscal rigidity movement" (dakai zaisei kochokuka^[*] undo^[*] ), the MOF singled out LDP protectionism as a major contributor to budget inflexibility. As long as the leadership of the LDP remains responsive to its lower level organs, and they, in turn, are so closely tied to ministries and other clientele groups, intervention by the party will not disturb the long-term balance of Japanese allocation patterns.

Balance for the Ministry of Finance

The Ministry of Finance stands at the center of the Japanese budgetary

process, and its budget bureau is the only full-time participant.^[33] The MOF's authority to prepare the budget (yosan hensei ken ) is jealously guarded against interference from outsiders. With its near monopoly of information on the budget as a whole and its specialized skills, the MOF is in a strong position to implement its own policies if it desires to do so. Balance ideas influence MOF budgeting behavior in all three ways: as an aid to calculation, a strategy, and a value in itself. These are intermingled in practice, but may be separated for analytical purposes.

The most obvious case of balance as an aid to calculation, often cited by officials, is when concrete relationships between two programs dictate that a rise in expenditures for one requires a similar hike in another. The amount for the second program thereby might be set almost automatically. Balancing routine administrative expenses is also relatively straightforward: examples include salaries, unit prices for materials, building maintenance, and numbers of government cars per hundred employees—all established on a government-wide basis so that ministries will not be treated "unfairly." However, balancing administrative expenses often extends beyond those that are obviously routine, as in provision of computers or allotments for travel, where it could be argued that needs of ministries vary widely. Reflecting on this problem in an interview, a budget examiner recalled:

Travel expenses were needed for an international economic conference in Bangkok; here, we balance so that, for example, the Ministry of Construction will not have a full delegation and the Ministry of Agriculture and Forestry only half that number. This is checked with great care . . . it is not a matter of policy questions . . . we really pay attention to this extremely routine work of maintaining balance. This is traditionally what has controlled most thinking about budgeting in Japan.

Though it might be that one ministry really should have sent more participants than the other, such a judgment would require the examiner to investigate the meeting's substance and would irritate ministry officials. It is easier to be equitable.

Such considerations extend beyond administrative to policy expenses. Particularly when expenditures within a single budget category are allocated to different ministries or among different programs, there is a tendency for them to grow at similar rates. A budget officer from a

ministry concerned with public works commented: "If the total budget for public works goes up about 10 percent, the public works budget of each individual ministry should also get a 10 percent hike. Unless there is some special reason, they won't sharply cut or raise only our ministry's public works budget [share]." Partly to maintain balance, public works are handled by a specialized budget examiner. This is not true of the "promotion of science and technology" budget, but as a former high budget official noted, similar considerations prevail:

To some extent, a standard is set [medo o tsukeru ] for promotion of science and technology, for example, that it should go up 15 percent next year. The various examiners then keep this in mind while reviewing their ministry's budget. The system is used because of fear of unbalance—for example, among the three Budget Bureau vice-directors, one might be enthusiastic about science, so this budget [for the ministries in his purview] would go up 20 percent, while another with no interest would keep it down to 10 percent.

Again, if a standard growth rate is set for all elements of a category, the need to evaluate and compare individual proposals is largely obviated.

Even when programs are not formally related by budgeting categories, any shared characteristics may be taken as a justification for balancing. The example cited by Kono^[*] Kazuyuki is the "postwar settlement."^[34] The repatriates' case in 1967 was extremely difficult for the MOF, not only because of intense political pressures, but also because property records were chaotic and there were no obvious guidelines for deciding the amount of compensation. Accordingly, the MOF held to a position that the grant should be balanced with earlier payments to equivalent claimants, with the precedents of these cases cited as part of MOF proposals on the amount of payments to individuals or households, the total burden on the budget, methods of payment, and a formula to adjust for inflation.^[35]

As well as simplifying calculation, such appeals to balance are important strategically. Prime Minister Sato^[*] said of the repatriates' case that "since solving the overseas property problem puts the lid on the 'postwar settlement,' I hope it will be kept in balance with other compensation matters."^[36] This hope was motivated by a fear that if the repatriates got more than the other groups, these would then demand an equal or better supplement, and the "settlement" would become unsettled.

Conversely, the earlier precedents provided the MOF with its best arguments for opposing the much larger demands from the LDP. The MOF frequently will draw a connection between a program in dispute and some settled matter to claim that the two should be kept in balance.

For example, when the Ministry of Health and Welfare asked for a 21 percent hike in public assistance payments for 1963, the MOF replied that this increase should be kept in balance with that in public employee salaries (7.2 percent that year) because both were supposed to meet increases in the cost of living. Such gambits can become rather tenuous. In 1967, this ministry had set its sights lower, asking just a 5.5 percent raise (the same as the previous year), but the MOF offered only 4.5 percent on grounds that the public employees salary raise had dropped off from 7.2 to 6.9 percent over the two years so welfare payment growth should slow down too.^[37] The particular logic of each argument is not important; the point is that notions of balance are often used tactically by the MOF in combating demands for individual program increases.

Probably more significant from a strategic perspective is the negative, inhibiting effect of balance on MOF behavior. Its officials are well aware, as a Budget Bureau director put it, that "there is nothing a ministry dislikes more than a 'share-down' in its budget." This means, for one thing, that the MOF must be extremely careful about trying to cut individual budgets, even relatively. The record shows that nearly all the MOF's antispending campaigns have been devoted to across-the-board cuts—equal percentage reductions in certain categories of administrative expenses or subsidies (hojokin )—even though MOF officials believe that waste is much more prevalent in some ministries than others and know from experience that such campaigns have virtually no effect.^[38] More subtly, even giving an unusually large increase to a certain program of high priority can be dangerous because supporters of other, somehow similar programs will feel that balance has been violated. Since an allocation once offered can never be reduced, there is risk of a general "level-up" that will increase total expenditures significantly. For the MOF, maintaining balance is usually seen as the safest course.

A final interesting example of a positive use of balance as a strategy may be offered. Throughout the 1960s the MOF was plagued by the rapid expansion of two programs with strong political backing: subsidies for rice prices and the government contribution to health insurance. From 1960 to 1969, the deficit in the Food Control Special Account, made up from the General Account (that is, ordinary tax revenues), rose from only 0.7 percent of the total budget to 4.5 percent; in the same period, health insurance costs rose from 3.6 to 7.1 percent of the total budget. At least until the end of the period studied, attempts to restrain these programs were ineffective. However, when the organizational budget breakdown is examined, the Ministry of Agriculture and Forestry share rose only from 9.4 to 10.5 percent in those years, and the Ministry of Health and Welfare

from 10.1 to 13.4 percent.^[39] Since both programs are included within the budgets of their respective ministries, this means that the remainder of Ministry of Agriculture and Forestry programs actually declined in budget share from 8.7 to 6.0 percent; for the Ministry of Health and Welfare's remaining programs, the decline was from 6.5 to 6.2 percent (smaller, but perhaps more significant in light of the consensus that Japanese welfare programs should be "catching up" with the West during the 1960s). Operationally, as explained by an agricultural official, "it is a fact that the Foodstuffs Control item has expanded, and this has 'oppressed' [appaku ] the budgets for other agricultural programs. [In reviewing our budget] the examiner will sometimes say that the rise in this item makes the rest of the budget 'difficult.'" This is balance again—a view that the budgets for these ministries or these policy areas should occupy some vaguely defined "fair share" of the total. If one component rises swiftly, the others must grow more slowly. In effect, each ministry (or policy area) has a framework (waku ) which it should not exceed. Indeed, the Ministry of Agriculture and Forestry counterstrategy here has been to say that rice supports are a separate problem from the rest of agriculture, not controllable by the ministry, and should therefore have a separate framework (betsuwaku ). A similar, more general use of this strategy is the frequent MOF demand that any ministry seeking new programs should first eliminate or cut back older programs by a similar amount.^[40]

It is logical to ask about the reverse case—a ministry that administers a rapidly diminishing program. Does the MOF compensate for its losses? One such case is the Ministry of Labor; its largest program—unemployment compensation—has fallen off with the growing demand for labor produced by economic development. Statistics indicate that the labor ministry has not kept pace with other ministries: its budget share slipped from 2.4 percent in 1960 (3.0 in 1955) to 1.7 percent in 1969. On the other hand, interviewed officials claimed, and other observers agreed, that new programs proposed by the ministry were approved by the MOF rather more frequently than usual. In part, this relatively soft treatment may be due to the Ministry of Labor's skill at justifying its requests, but quite possibly it is also a function of the MOF's regard for balance as a value in itself, beyond its sheer utility as simplifier or strategy.

The point that the MOF sees balance as a value for its own sake, more than just an aid or strategy, is sustained by examining its view of the LDP's role in budgeting. While conceding that a majority party in a democracy should have a voice in budgeting, the "finance bureaucrats"

tend to be disdainful of politicians, interest groups, and other bureaucrats who seem subservient to them. An examiner was reputed to have said to an official from the Ministry of Transportation: "You are no better than your bosses—the cab drivers."^[41] Political intervention in budgeting disturb balance, and the MOF will at times use its own influence as a counterweight to such distortions. It may give less rigorous examination to requests from entire ministries with relatively weak political backing, such as labor or education, or to individual programs within one ministry since the LDP is rarely equally supportive of all ministry activities. In such cases, the MOF makes common cause with spending-ministry officials who are themselves concerned about maintaining internal balance against political pressure. A Ministry of Construction official said:

Interest groups and Dietmen tend to be strong on highway and river projects; in the end, their weight falls here, first in pressure on us and then on the MOF . . . since we know this will happen, at the beginning we really push areas like urban development and housing. The highway and river budgets will go up regardless so we work on the others. If we don't do this and just ask for everything equally, the budgets for housing, sewers, parks, and so forth will be very small. Our strategy is: because it's hard to get, do it first—the Dietmen will take care of the rest.

Finance Minister Fukuda made a remark in 1966 that indicates the MOF's willingness to collaborate in this strategy. He observed that he had to be sure to get sufficient funds for housing into the MOF draft budget because, unlike rivers and highways, it would never get enough political support in the appeals negotiations.^[42] The MOF also assists by keeping negotiations with the ministries private, allowing ministry spokesmen to continue making public statements that they are backing up group or party demands to the hilt while actually pushing for less popular programs.

Further evidence of MOF concern for balance is provided by a look at the course of its reforms in the budget process itself. One example is the imposition of a limitation on ministry budgetary requests. The limitation began in 1961 at 150 percent of the previous year's actual budget, was dropped to 130 percent in 1965, and finally to 125 percent in 1968. Such limitations encourage a balanced budget since no ministry can request much larger amounts on grounds of high policy priority, and this in turn diminishes the possibility of a shift in expenditure patterns. The fact that late requests over this limitation have sometimes been allowed merely softens the impact of this reform.

Two other examples both occurred as part of the 1968 "break fiscal rigidity movement," the major occasion when the MOF took the offensive against pressures for more spending. One was called "comprehensive

budgeting" (sogo^[*] yosan shugi ), meaning the elimination of supplementary budgets. Each year, one to three supplements had been passed, amounting to from 2.5 to over 10 percent of the total final budget. It had become customary to devote most of these supplements, beyond the relatively nondiscretionary amounts needed for natural disasters occurring during the year and the mandated local allocation, to two items: salary increases for government employees, decided by the cabinet after recommendations from the National Personnel Authority in midsummer; and the deficit in the Foodstuffs Control Special Account, which depended on decisions, reached by a complicated political process, on the prices for rice paid to producers and asked from consumers. Also, the total amount of the supplements had come to take up most of the natural revenue increase (shizen zoshu^[*] ), the surplus created by underestimating economic growth and tax revenues in the original budget. This meant that the decisions on salaries and rice prices, which were supposed to be based on "objective" factors, in fact were heavily influenced by the amount of money actually available—if the economy was booming, tax revenues would be strong, and these two items would tend to be large. Not only is this behavior procyclical in fiscal policy terms, but from the MOF's point of view, it allowed these two expenditures to grow more quickly than they should. Under "comprehensive budgeting," provision for the rice subsidy and salaries (as well as disasters) was to be made in the initial budget, covered by a "full" revenue estimate, so these decisions could be brought back into the competition of the regular budget process and thereby kept within overall balance patterns.^[43]

Finally, the MOF also proposed to modify the revival negotiations process in 1968. Up to that time, the funds added to various items during this period had come from "hidden resources" (kakushi zaigen ), which were secreted away in accounts under MOF control so only MOF officials knew the amount available. In the mid-1960s, this had amounted to some 3 to 4 percent of the total budget. Partly because the mystery of the figures had encouraged the ministries and party to continue asking for more money, the MOF substituted for this system two devices known as "public resources" (kokai^[*] zaigen ). One, called "secretariat adjustment expenses" (kanbo^[*] choseihi^[*] ), was a sum in a range around ¥10 billion (about $28 million) each for the larger ministries, from which the staff of each ministry was allowed to make allocations among the new or established programs it thought most deserving. The other was a larger amount (¥50 billion or more) reserved for "policy-level" decisions—those taken at the level of cabinet minister or MOF-LDP negotiations, and identified in the MOF draft as "adjustment expenses" (choseihi ). These innovations meant that the MOF was formally renouncing the possibility of making these last decisions itself on programmatic or "merit" grounds in favor of simply

dividing up the remaining resources equitably, allowing the ministries and then the party to decide how to spend them—clearly a victory for balance.

Actually, this reform is not much of a watershed in itself because the degree of its implementation in 1969 and 1970 was somewhat questionable. The reform also seems little more than an institutionalization of a tendency already quite powerful within the MOF; although evidence is hard to gather, it appears that a high proportion of budgetary decisions formally in Budget Bureau hands had already been heavily influenced by ministry opinions. Direct or indirect queries of a ministry's priorities among programs were common in both the September budget request explanations and in later calls from the examiner to the ministry for further information. The most difficult questions were routinely postponed until the revival negotiations where, under either the old or new system, ministry voices are strong. The Ministry of Health and Welfare's budget requests for 1970, which included proposals for forty-eight new items, illustrate the postponement tactic. Of the new items, only one was clearly approved in the MOF draft prepared independently by the Budget Bureau, but at least nineteen more had been approved by the end of the revival negotiations.^[44]

This sort of evidence, though fragmentary, indicates that the MOF has in effect been giving up some of its power to decide on individual programs, and the reforms in the budget process noted above both contribute to and symbolize this trend. The point may be made more clearly by contrasting two ideal-type budget processes. In the first, which might be termed "strong MOF," ministries ask for everything they desire; the MOF evaluates their requests and allocates most of the available funds among them as it sees fit. The ministries (and the LDP) then ask again for additional funds for some programs, and the MOF responds by adding on small increments as justified and necessary. The second ideal type could be called "weak MOF." Here, ministries decide priorities among programs before making requests (to stay within the 125 percent limitation); the MOF makes only mechanical decisions on routine items, postponing the difficult decisions. In the revival negotiations, each ministry (and the party) is allocated a lump sum by formula from which it makes its own decisions about programs. Neither ideal type ever existed in pure form, but the process in the mid-1950s more closely resembled the "strong MOF" pattern whereas, by the late 1960s, the "weak MOF" type was not far from reality.

Why did the Ministry of Finance give up so much of its power, for the most part voluntarily? I will attempt a brief and speculative explanation. Budgeting decisions may be divided into two categories: "microbudget-

ing," individual consideration of each detailed program; and "macrobudgeting," issues like the size of the total budget, the amount of deficit financing, revenue estimating, and economic regulation through fiscal policy. Traditionally, the MOF had considered itself dominant in both spheres, but three trends in the 1955–1970 period led to a partial redefinition of its mission: (1) in common with other advanced industrial nations, there was increased concern for governmental regulation of the economy, the responsibility for which lay with the MOF; (2) with economic development, governmental programs became more numerous and complicated, and thus informed judgments demanded more expertise in ever more specialized policy areas; and (3) the Liberal Democratic Party was established and grew in confidence and power. The last point is important because the LDP, responding to the interest groups and spending ministries that make up its clientele, has been far more interested in specific microbudgeting questions than in macrobudgeting. The Party did not hesitate to intervene to obtain more funds for favored programs, and these interventions began to impinge on the MOF's autonomy in macrobudgeting decisions and even internal organizational matters. Facing a choice among priorities, the MOF sought to protect its near monopoly over macrobudgeting decisions and its organizational boundaries by allowing control over individual program spending to slip away to the experts in the spending ministries and the LDP politicians.

If this line of analysis is correct, it would follow that most conflict would develop in areas of overlap between microbudgeting and macrobudgeting, where individual programs have an impact on overall budget problems. This has indeed been the case. Most LDP-MOF battles have occurred over items like agricultural price supports, medical insurance, and railroad deficits, when fixed costs have threatened to push up the size of the total budget or inhibit fiscal flexibility (as in the "break fiscal rigidity movement").^[45]

How does this shift relate to "balance"? To oversimplify, a major problem in any budgeting system is how to decide on sectoral allocations—the relative amounts to be given to policy areas like education, welfare, defense, and public works—or to each ministry. If the budgeting authority is itself making decisions on each individual program, sectoral allocations will emerge almost automatically from a bottom-up additive process. In Japan, such individual decisions have largely been given over to the fragmented spending ministries and the party. It is still possible for someone to make broad policy judgments—that education, or welfare, or the build-up of social capital deserves greater emphasis this year, and therefore its growth rate should be higher than for other sectors. But to the extent that no participant is willing or able to make such broad policy

choices, little alternative remains but to confirm the existing pattern of allocations and have each ministry and sector receive about the same budget share as it had previously. This is "balance." And although I must again emphasize that all budgetary systems are terribly complex, that they resemble one another more than they differ, and that all the tendencies illustrated here may be found in every system, I conclude that "balance" is particularly significant in Japan.

Conclusions

We may now speak of cause and effect. Why should balance be so important in Japanese allocation patterns and elite political culture? And what difference does it make to how the business of the nation is conducted?

An attractive and immediately available explanation of why balance ideas are so prevalent in Japan is that of culture, perhaps even the Japanese "national character." That is, balance appears to correspond to many traits that anthropologists and others consider as typically Japanese. For example, Chie Nakane notes that in Japan "democracy" means that "any decision should be made on the basis of a consensus which includes those located lower in the hierarchy . . . it should leave no one frustrated or dissatisfied."^[46] A dislike of open conflict is also frequently noted. A case might thus be made that the Japanese budgetary system inevitably tends toward making decisions through relatively impersonal and mechanistic administrative devices that reward all equitably.

Such an argument has merit and is difficult to contradict. Still, I would prefer to emphasize the equally valid point that all budgetary systems probably have strong tendencies toward decision-making patterns that avoid value judgments. This is both because value judgments among programs are technically and intellectually very difficult and because making such judgments opens the decision-maker to attack from the disadvantaged parties. Therefore, budgeters will attempt to routinize their decisions—and balance is a sensible method of routinization—so long as they are not prevented from doing so.

If, then, balance or something like it is the "normal" mode of budgetary decision-making, what must be explained is not stability in allocation patterns from year to year, but deviations from stability. What factors might influence budgeters to change the decisions they made last year? One common situation is when some programs have to be increased because of inflation or other unavoidable cost increases, but resources are not sufficient to provide equal hikes for all programs. This has often been true in American budgeting, but in Japan rapid economic growth has

produced enough "natural revenue increase" each year to take care of cost increases with some left over.^[47] Budgeting is more comfortable.

Another and probably more important factor is illustrated by Table 2, which lists changes in budget share over a ten-year period (1961–1970) for the standard functional categories of the Japanese and American budgets. It is readily apparent, first of all, that the American shares have changed

more than the Japanese shares; in part this is because of the greater year-to-year stability shown in Table 1. But attention should also be directed to the items showing the greatest changes in the two countries. For Japan, these are in large measure items that are nondiscretionary, not part of decision-making on expenditures: "special foreign obligations" are mostly determined by old treaties with countries receiving war reparations; the increase in "national debt" expenditures is a product of the advent of deficit financing in 1965 (a decision on revenue mix, not allocations); and much of the change in "manufacturing and economy" is the increased deficit in the rice-price-supports program, supported by political pressures (leading to a "nondecision") but caused primarily by growing harvests and changing consumption habits, which turned Japan from a rice-deficit to a rice-surplus country.

The "unbalanced" American items have a different character. The enormous increase for "community development and housing" is a figment of accounting peculiarities and should be ignored, but those for the other three—"space research and technology," "education and manpower," and "health"—may be legitimately compared with the Japanese figures. All are the direct result of clear presidential decisions: Kennedy's to reach the moon; Johnson's—in response to the civil rights movement and urban riots—to initiate a "war on poverty." It is clear that the president has had a major impact on American allocation patterns, while no one has had a comparable impact on the Japanese patterns.

Two points may be offered by way of partial explanation of this difference. The first is that Japanese prime ministers have tended not to hitch their political fortunes to domestic issues, at least those requiring money. The chief exception is Ikeda and his income doubling plan, but his policies had more impact on macrobudgeting matters than on allocations (Tanaka is another exception, but his premiership lies outside the period of this study). The second, more structural point is that the prime minister does not occupy anything like the president's key position within the budgetary process. He may conciliate, mediate, and occasionally adjudicate, but he rarely intervenes actively, and his influence over the Ministry of Finance does not compare with the loyalty of the Bureau of the Budget to the president.

A complete discussion of the reasons for such passive prime ministerial behavior in this period would have to include the relative lack of newly formed pressure groups or social movements calling for more spending, the stability of LDP control of the government, a consensus within the elite on national goals, a historically nurtured respect for the bureaucracy, and Japanese styles of leadership behavior in other settings. These are beyond our scope. It is appropriate to point out, however, that the prime minister himself is not unaffected by considerations of balance. He holds his office because he is president of the majority party and, as any study of the LDP will show, maintaining this position requires keeping a majority of

Dietmen satisfied, and balancing off interests of factions and other groupings within the party. The prime minister has most often found that the benefits of strong leadership in domestic policy matters are outweighed by the costs of such a position. Similar calculations apply to the other institutions in a structural position to influence overall allocation patterns, and therefore, none has intervened. Budgeting in Japan has been left to the budgeters. This may well be the primary reason for its expenditure patterns remaining so stable.

What have been the effects of balance and stability? Supporters of programs that are not enthusiastically backed by the LDP, such as public housing, welfare, and some education programs, have clearly benefited by the availability of appeals to fair share. Balance values have served as a counterweight to political pressure. More generally, conflict has been lessened and harmony promoted within the budgetary system. But perhaps the major positive effect of balance has been the avoidance of "boom or bust" financing patterns, which can severely damage the morale and effectiveness of an agency. Under the Japanese pattern, most program decisions are left in the hands of ministry experts. By extrapolating from past budgets, they are able to make accurate estimates of the resources to be available in the future, and thus can plan their activities realistically.

However, at least from the vantage point of a distant observer, the costs of maintaining balance over the long run would appear to be more significant. The annual budget is an expression of the government's policies—responses to the needs of the society. It is inevitable that these needs must change as the society changes. Japanese society has changed during fifteen years of extraordinary economic growth, and while the shape of the budget in the mid-1950s may have been quite appropriate for national needs and goals at that time, by the same token it seems improbable that the same mix of policies would be optimum in the 1970s. Balance tends to prolong the life of programs that are wasteful or no longer required, and to inhibit the shift of resources into higher priority areas and the initiation of large new programs, particularly if they threaten the jurisdictions and budgets of the existing ministries. The responsiveness of the government to its environment is severly limited. One would expect that if balance continues to dominate budgeting for too long, strains and dissatisfactions would build up to a point where they would have to be relieved by rather sudden and large budgetary adjustments. Such a crisis would allow a clearer assessment of the delicate relationships and mutual expectations that make up the Japanese budgetary system.