Part I

 

Theoretical Framework

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

CHAPTER 2

POLICY LEGITIMACY AS A DETERMINANT OF POLICY OUTPUTS

 

Policy Dilemmas and Their Solutions

            Reelection is one of politicians’ dominant goals.[1]  Political parties are assemblies of politicians with common objectives容.g., control of government (Downs, 1957), policy goals, and a vision of good society預nd seek to achieve the goals within a set of institutional, structural constraints (North, 1990).

            Despite the temptation to serve particularized interests to win votes and campaign funds, politicians as a whole face the task of attending to the general needs of society at large, and general and particular interests can conflict with each other.  Policy needs may arise, for instance, of keeping the national economy in a sound state by balancing national revenue and expenditure.  Constituents may understand the need to cut spending or increase revenue or both.  But it is no easy task to solve the specific question of who should bear the burden.

The interests of a party as a whole and the needs of its individual members may conflict with each other, as a party represents a larger number and wider range of constituents than each of its members.  A policy that brings more seats to the party may cause reductions in votes in some politicians’ districts.  The party may suffer an electoral setback as a result of its inability to override some individual politicians’ opposition to a policy that, if implemented, would increase the party痴 seats.  Some policies may benefit some politicians at the cost of others.  Some policies may thinly spread benefits among a large number of member politicians.  It is not always easy to aggregate individual members’ diverse interests into a coherent party policy.  But the party will be blamed for its incapability to govern if it does not accomplish the task.

            Politicians may also face conflicts between short- and long-term interests.  Making trade-offs can be problematic.  A party in power may be disposed toward short-term popularity.  Politicians face elections every few years, and the fear of losing the next election may make it difficult for the ruling party to make a policy that will produce long-term benefits but impose short-term costs.  Fiscal austerity programs to bring inflation under control or reduce budget deficits exemplify this type of policy.  But if the party aims for longer terms of tenure, it must also take into account the long-term costs and benefits of a policy.  A policy that woos constituents in the short run may not be in its best long-term interest, as the public may eventually lose trust in the party痴 governing abilities.

            Party leaders and senior politicians謡ho tend to have greater electoral security than backbenchers洋ay have the potential to serve as a protective shield against particularistic politics.[2]  Party leaders are also responsible for guiding the party toward gaining control of government and must heed the electoral power of the party as a whole by making coherent policy programs and serving also general interests.  But despite such potential, backbenchers’ electoral needs may still hinder the leaders from pushing a policy opposed by particular interests.  Certainly, backbenchers may sometimes have to follow the leaders’ policy decision they oppose, because they derive from party leaders such political resources as party endorsement, campaign finance, posts (in the party, Diet, and cabinet) and their political influence is contingent upon the well-being of the party as a whole.  But their dependence is mutual; the leaders also depend on backbenchers’ support for their own power within the party and for legislation of their policies.  Further, leaders’ control of government and pursuit of their other goals rest on how backbenchers or the party as a whole does in elections.  Thus, when backbenchers oppose a policy strongly enough, the leaders have a good reason to take the electoral implications of their opposition seriously.  The result would be the same old particularistic politics.

            Nevertheless, politicians sometimes manage to overcome difficult conflicts despite the magnitude of stakes involved for politicians or constituent groups and the complexity of interest coordination.  And they do so apparently without necessarily satisfying the competing interests.  For instance, although many Liberal Democratic Party (LDP) politicians opposed Prime Minister Takeshita痴 new tax and their electoral concerns were not necessarily alleviated, they eventually complied with his tax reform initiative.  Similarly, the opposition Clean Government Party (CGP) and Democratic Socialist Party (DSP) opposed the tax, but agreed to let it pass the Diet while they publicly remained opposed.  In the Hosokawa administration痴 electoral reform, too, many coalition and opposition politicians用articularly of the Social Democratic Party (SDPJ) and LDP熔pposed the institution of the mixed SMD-PR system, but voted along with their parties to approve it in the end.  These suggest that a policy may still obtain approval even when the self-interest of relevant actors does not easily justify it.

            How can politicians overcome their proclivity toward particularism and make electorally unpopular or contested policies?  I propose two answers.  First, politicians tactically seek to minimize the repercussions of an unpopular or contested policy.  They exploit available strategies and devices to protect their interests, including (1) the modifications of the distribution and magnitude of the costs and benefits of a policy; (2) the maintenance of the koenkai as vote mobilization machines; (3) the exploitation of the relative absence of issue-voting in Japanese elections; (4) the use of party control as justification for complying with an unpopular policy decision; (5) the active use of foreign pressure as justification; and (6) the delegation of politically delicate decisions to regulatory commissions, ad hoc commissions, and the like that may protect politicians from blame (Pierson and Weaver, 1993).  When effectively employed, these strategies can mitigate politicians’ fear of electoral repercussions to some extent.  (The first five strategies will be explained shortly.)

            These strategies, however, do not guarantee the elimination of the negative effects of an unpopular or contested policy nor the removal of politicians’ fear of electoral retribution.  They may not be sufficiently potent or appropriate to mitigate a particular policy痴 negative impacts.  The magnitude of policy opposition may exceed the capacity of the strategies.  Politicians may also misuse strategies.  If politicians could so skillfully counteract the negative consequences of their actions, they might not lose elections.  If they could competently calculate a policy痴 electoral impacts, they might not seek an electorally risky policy in the first place.  Also, if the distribution of costs and benefits were the major determinant of legislative success and failure, politicians would have great difficulty legislating a policy that runs counter to narrow constituency interests.

            The second reason why politicians can sometimes break free of particularism is that self-interest is not the only controlling factor in politics, and politicians can also appeal to or contrive the legitimacy of their policy which can stand independent of self-interest.  That is, if a governing party successfully engineers legitimacy for an opposed policy, it may be able to have the policy approved while lessening its electoral repercussions.[3] 

            One of the keys to politicians’ behavior is whether and how well they can explain and justify their actions to constituents and voters (Kingdon, 1981, pp. 47-54).  If they can evade electoral punishment, they are more or less free to pursue an unpopular policy or whatever other goals they may have.  Policy legitimacy gains importance because it provides politicians with such justification to explain a contested policy decision even when it does not conform to actors’ immediate interests.  It can lead competing actors擁nterest groups, voters, and politicians葉o approve a policy without fundamentally resolving conflicts of interest that might be difficult to settle by solely changing the distribution and magnitude of the policy痴 costs and benefits.  It thus renders the costs of such a policy justifiable or tolerable to the actors.[4]

            I now turn to politicians’ strategies and then explain policy legitimacy.

Modification of the distribution of costs and benefits and compensation

            Politicians can mitigate the negative effects of a loss-imposing policy by changing the distribution and magnitude of the policy痴 costs and benefits to increase benefits and/or reduce costs to its opponents whose support is essential to policy approval.  The purpose of this modification strategy is to 杜old a policy so that it conforms better to legislators’ and citizens’ preferences” (Arnold, 1990, p. 108).  The modifications may take the form of the creation of special provisos in a policy so as to reduce its burden to those adversely affected.

A case in point is many tax exemptions the Nakasone administration created in its sales tax proposal (1987) when faced with opposition from businesses and industries.  Nakasone痴 and Takeshita痴 tax proposals (1987 and 1988) also included income and corporate tax reductions preceding the introduction of a new consumption tax to curb public opposition.  Takeshita and Prime Minister Murayama (1994) also reduced the magnitude of their tax increases.

            A governing party uses the modification strategy also toward the opposition parties to mitigate their policy opposition.  Under LDP rule (1955-1993), the party typically gave the opposition parties small policy concessions for which they could claim credit before their constituents.  The opposition parties, in exchange, would provide cooperation on the LDP痴 policy.  The Takeshita administration, for instance, gained the opposition CGP痴 and DSP痴 cooperation on consumption tax legislation by meeting their demands on the particulars of the tax implementation and for an expansion of welfare programs.

            A governing party can also seek to win opposition parties’ cooperation on a bill in exchange for cooperation on another bill they wish to have approved.  The aim of logrolling is to 菟ersuade legislators to vote against their true interests...in exchange for leaders’ or other legislators’ assistance on matters of greater importance to these legislators” (Arnold, 1990, p. 92).  In the days of its one-party rule, the LDP would typically 鍍ake hostage” of bills that were important to the SDPJ (e.g., those on pay increases for public employees) and would demand the SDPJ痴 cooperation on the LDP痴 own bill.[5]  (The opposition parties could also use this strategy to stall the LDP痴 bill by refusing deliberations on other bills that were also important to the LDP.)

The koenkai

            Japanese politicians maintain in their districts a network of personal supporters, the koenkai (support associations).[6]  While the primary function of this vote-mobilizing apparatus is to ensure entrance into politics and reelection, it also serves as a buffer against the electoral retribution for an unpopular policy. 

In the politics between politicians and koenkai members, policy positions用articularly those with national implications預re not likely to be an issue, because koenkai members’ support is not contingent upon politicians’ positions on national policies, but upon their personality and ability to bring benefits to the supporters in the form of constituency service and government resources.[7]  Therefore, individual politicians’ compliance or incompliance with their party leaders’ unpopular policy decision not regarding their districts will not likely have bearing on koenkai members’ support.  Politicians can thus count on koenkai votes and mitigate an unpopular policy痴 negative electoral impacts.[8] 

Exploitation of the relative absence of issue-voting

            The relative absence of issue-voting is characteristic of Japanese elections in general.  This makes it somewhat easier for politicians to avoid electoral retribution for making an unpopular policy.[9] 

Most Japanese voters do not have any position on issues.  Even when they do, they are unable to identify a party that represents a view closest to their own or do not vote for the party that they feel is closest (Kohei, Miyake, and Watanuki, 1991).[10]  Richardson (1986, 1988) similarly shows that around 70 percent of the electorate are consistent voters who regularly vote for the same party with or without a stable party identification, that about 20 percent of the entire electorate or 30 percent of the consistent voters are habitual voters who do not often have a substantial issue component, and that issue proximity to a party has a much weaker influence on the vote than party identification, party images, and party-focused habitual voting. 

The small impact of issues on voting suggests that politicians can relatively safely comply with their party痴 unpopular decisions without significant electoral retribution, as long as they do not directly affect the interests of their constituents.  This should hold true especially if voters have difficulty linking individual politicians’ actions to a specific government policy and to its effects.[11]  But some policies do affect the well-being of voters.  Such was the case with the Japanese politicians’ five attempts at tax increases.  In such a case, politicians have good reason to be concerned about electoral retribution.  (Politicians’ electoral concerns indeed caused the withdrawal of the tax proposals in 1979 and 1987.) 

Party Control

            Japanese politicians can exploit strong party discipline as justification for their inability to block a party policy opposed by their constituents.  In Japan, cross-partisan voting in the Diet is rare.[12]  Constituent groups are cognizant of party control.  This provides politicians with a means to mitigate electoral retribution for their compliance with their party痴 unpopular policy decision. 

            Japanese politicians and parties collude and stage scenes visible to the public in which backbenchers vehemently protest a decision made by their party leadership, while knowing that it is irreversible.  They arrange these scenes to give individual politicians the opportunity to show their constituent groups that they did what they could to veto a party policy, but in vain.  In the LDP, its Policy Affairs Research Council and Executive Council provide forums for politicians pushed by discontented constituents to register their opposition in an apparently ritualistic effort to show loyalty to constituents.[13] 

            Similar ritualistic scenes have also been arranged between ruling and opposition parties in situations where the opposition parties agreed on the need for a ruling party policy or their minority status made the shelving of a policy difficult, but their constituents’ opposition did not allow them to support it or let it pass the Diet outright (Interview with a former prime minister, May 20, 1994).  In such a situation, the opposition parties may deliberately set up a scene in which they boycott Diet deliberations and votes in formalistic protest against a policy.  Their aim is to demonstrate to constituents that they did everything to block the policy, but in vain.[14]

Foreign Pressure

            Just as Japanese politicians exploit party control, they can exploit foreign pressure to seek a policy which domestic electoral pressure might make difficult in the absence of external pressure. 

As Calder (1988) points out, the impetus to change in Japan痴 foreign economic policy is typically provided by external pressure and that the country has been more forthcoming with specific policy changes, when faced with foreign pressure.  While foreign pressure constrains Japanese politicians’ actions, they can also exploit it to promote their agenda that might otherwise be politically infeasible (see also Schoppa, 1993).  It gives them justification for making an unpopular policy decision. 

            Consider Japan痴 import liberalization of beef and oranges (1988) and rice (1993).  Foreign governments strongly demanded liberalization, but Japanese farmers opposed it on both occasions.  All Japanese political parties also opposed and sought to resist it, since they all drew electoral support from farmers.  They should, then, have had difficulty meeting the foreign demands.  But the Japanese government eventually agreed to liberalization in the face of farmers’ opposition.  Nevertheless, no political party suffered electoral retribution in the elections that followed.[15] 

            Behind the exploitation of foreign pressure are the same mechanisms at work in politicians’ manipulation of party control; the presence of strong foreign pressure can make recalcitrant constituents resigned to politicians’ inability to defy foreign demands, and the constituents spare the politicians penalty for an unpopular policy decision.  The effectiveness of this tactic is often complemented by the use of other strategies.  In making a domestically unpopular policy move, Japanese politicians typically employ the strategies of modification and compensation to lessen domestic opposition.

 

Politicians may simultaneously employ these strategies in any combination.  Multiple strategies are likely to be employed when a policy entails large visible costs to a number and variety of citizens and constituent groups.  The Takeshita administration, for instance, took a variety of measures to mitigate opposition to his new tax among the public, small-sized and medium businesses, and opposition parties.  Also included were LDP backbenchers’ performance of opposition against party leadership and the CGP痴 and DSP痴 ritualistic opposition.

            These strategies can help politicians resolve dilemmas between their electoral and policy needs to a certain degree.  If the strategies or other circumstances help significantly reduce the negative repercussions of a contested policy, the possibility increases that politicians will have the policy approved.  But the effects of such strategies may not be sufficient to mitigate policy opposition or politicians’ fear of electoral retribution.  When such is the case, the success or failure of a policy attempt will depend on whether the policy has legitimacy sufficient to provide politicians with a means to override opposition by their constituents and other relevant actors.

 

Social Foundations of Politics

 

            Normative factors such as standards of appropriateness and legitimacy can figure in politics and self-interest is not rigidly controlling, because while political actors act in a self-regarding manner much of the time, their pursuit of self-interest takes place in a society where they share a conception of norms and rules for appropriate behavior, and because of the limits of human rationality, they accept some socially transmitted behaviors and norms without independently evaluating their contribution to their personal self-interest (Simon, 1990, 1995).[16]  And because of individuals’ divergent subjective mental constructs and limited cognitive abilities, the complexity of the environment, and incomplete information, factors such as beliefs, norms, ideology, and culture help produce behavior and outcomes divergent across actors and places and distinct from what is predicted by the rationality assumption (North, 1990).

When a solution to the coordination problem that would satisfy everybody痴 interests is unattainable, policy makers need to rest justification for their policy decision upon something other than individuals’ self-interest.  Aside from the use of force and coercion, 殿ppropriateness” or 斗egitimateness” is often a criterion by which to make decisions on issues irresolvable on the basis of the coordination of competing interests.[17]  Meeting the standards of appropriateness does not guarantee conflict resolution.  But when individuals come to believe that some problems call for loss-imposing solutions and that government and politicians are sometimes forced to make such decisions, the standards can be a powerful tool to justify individuals’ losses.  In this limited sense, a fundamental logic of politics is the 斗ogic of appropriateness,” as well as that of who gets what, when, and how (March and Olsen, 1989, p. 38). 

Such normative factors affect political behavior and outcomes by delineating the parameters of sets of choices available and by prescribing what is appropriate or inappropriate behavior.  They impose upon politicians standards for appropriate ways of decision making on an issue salient to many segments of society.  Policy makers can build legitimacy for their policy to make its implementation possible by following appropriate procedures and taking proper measures that would render the policy justifiable or bearable to the affected actors.

Policy Legitimacy

            Policy legitimacy is a degree of support, acceptance, or tolerance accorded by relevant actors to a particular policy.  When a policy is loss-imposing and faces political actors’ opposition, policy legitimacy is required to overcome the forces of the actors’ self-interest and to gain approval for the policy.  It is a continual variable ranging from active support to passive acceptance, to compliance with explicit objection, to active opposition.  The legitimacy of a policy is enhanced when actors perceive that it is supported by good and proper ideas and made by decision makers with a popular mandate, and the policy decision is reached in an appropriate way.  In other words, in building legitimacy, policy makers communicate to actors how appropriate a policy is and how legitimate they are in asking constituents to bear the policy痴 costs.[18]  A successful building of legitimacy can lead actors to give consideration to the need of a policy even when it is not in their immediate interests.  

Policy legitimacy rarely poses a problem to policy makers, if a policy is neither contested nor electorally unpopular due, for instance, to a lack of publicity.  By this definition, not all approved policies need legitimacy.  But it can become a crucial factor in deciding the fate of a policy when it is contested.  A lack of legitimacy robs policy advocates of justification for pushing a policy and provides the opposition with justification to oppose it. 

Three Components of Policy Legitimacy

            Three components of policy legitimacy are important: the idea, democratic, and decision-norm components.  When policy advocates’ attempts to legislate a contested policy are bolstered by high levels of legitimacy in the three components, the likelihood increases that the policy will obtain approval necessary for implementation.  In contrast, when a policy does not enjoy legitimacy, its advocates will have difficulty persuading the opposition to approve the policy, and its approval will depend more on a match between the policy and actors’ interests.

(1) The Idea Component

            The idea component concerns the substantive ideas of a policy, including dominant values and beliefs, knowledge about cause-and-effect relationships, worldviews, ideologies, or a conception of collective goods, justice, and appropriateness (on the role of ideas in politics, see Goldstein, 1988; Goldstein and Keohane, 1993; Kingdon, 1988; Reich, 1988).  When a policy infringes on actors’ interests and meets with their opposition, its ideas must be appealing to the actors on their merit alone to gain legitimacy.  Policy ideas must be compelling to justify the actors’ letting legitimacy counterbalance the dictates of self-interest.  Put differently, policy makers seek to build support for a policy through persuasion that appeals to the policy痴 normative and moral attributes or arguments and scientific evidence that demonstrate its correctness and effectiveness as a tool to solve a problem (Kingdon, 1988).

Ideas matter also because they constrain policy choices by delineating the contours of viable alternatives.[19]  Scientific or commonsensical knowledge of causal relationships, for instance, informs policy makers of how they should go about solving a problem and which policy tool to choose.  Further, individuals’ particular ideas are a factor in explaining which option they will choose over all the other equally feasible options (Ferejohn, 1991; North, 1990).[20]

            Japan痴 postwar pacifism and primacy of economic growth typify values and beliefs that have informed a country痴 security and economic policy: the primacy of U.S.-Japanese relations and diplomatic dispute resolution through international organizations; reliance for defense on the U.S. military forces; and allocation of national resources to economic growth and exports.  The national consensus on these two principles was historically shaped by the Japanese people痴 keen sense of international vulnerability, the devastation of World War II and the resulting exigency for economic reconstruction, and the negative lesson learned on the use of force as a means of conflict resolution.  The consensus has made the making of a policy deviating from these principles difficult.[21] 

            As another example, the electoral reform of the early 1990s was a case in which a newly emerging public attitude against corruption made the option of 渡o reform” politically unattractive to the parties.  The adoption of a new system that combined single-member districts and proportional representation was also a result of the particular idea at the time that single-member districts would reduce money politics and corruption (Chapter 5).

 

 (2) The Democratic Component

            Democratic principles expect governments’ rule to be in large agreement with the popular will that originally elected them.  But when they need to make unpopular policy decisions, at issue will be whether they can obtain public consent to the decisions, no matter how reluctant it may be.  The democratic component of policy legitimacy concerns to what degree an administration is entrusted or empowered to make a policy going against the immediate interests of constituents.  Put differently, it is about the degree of freedom an administration has in imposing losses on constituents.  A greater electoral mandate may allow an administration to make an unpopular or contested policy more easily or with less electoral repercussion.  This component, thus, increases what Pierson and Weaver call governments’ 斗oss-imposing capabilities” (Pierson and Weaver, 1993).

            Legitimacy is partly a function of popular mandate.  The democratic component of policy legitimacy is determined by the direction and magnitude of public opinion expressed in election and opinion poll results.[22]  First, in Japan, the outcomes of the last election for the Diet (particularly, lower house elections) influence this component, because they show the degree of the mandate the ruling party obtained when its government was formed and are a measure of electoral support of its policies in general.[23]

Second, a governing party can claim the legitimacy of a contested policy, if it wins national or local elections that take place during deliberations on the policy.  These elections include lower house elections called for by prime ministers’ dissolution of the house or by the end of the four-year term, upper house elections that are held every three years, by-elections, and elections for local legislatures and governors.  Conversely, the opposition can claim a lack of legitimacy if the governing party loses them.  Thus, these elections function as referenda on a policy.  For instance, in 1987, the LDP痴 losses in the nationwide local elections and an upper house by-election were interpreted to demonstrate public opposition to Prime Minister Nakasone痴 sales tax proposal, and became a factor in forcing him to withdraw the proposal.

A third kind is the results of opinion polls on the administration, the governing party, or a specific policy.  Also included in this category is an 殿tmosphere” or sentiment among the public on an issue which politicians pick up from their constituents and media reports.

            Public opinion provides a powerful weapon in policy competition and can aid or hinder policy attempts, as argumentation that exploits public opinion has a normative appeal.  It provides politicians with a means to explain their actions to constituents and voters.  When an administration or its policy enjoys favorable scores in these elements of the democratic component of legitimacy, the policy has a higher chance of being approved.  When contemplating a contested policy, politicians make trade-offs between their electoral prospects and policy pursuit and assess the political feasibility of seeking or opposing a policy by looking at public reaction during policy deliberations.[24]

(3) The Decision-Norm Component

            The decision-norm component concerns a way of collective decision making that people regard as appropriate.  Given a choice, individuals may seek to promote their self-interest.  But when a policy conflict is irresolvable on the basis of the coordination of competing interests and they know they cannot escape a sacrifice of self-interest, they care about the appropriateness of the contents of a decision and of the way the decision is reached to make the sacrifice justifiable.  For this reason, groups have standard operating procedures which specify procedural requirements to be followed (March and Olsen, 1989). 

Choice of one decision rule or another can be a function of people痴 beliefs and values about the nature of the world, conflict, and its resolution.  In this sense, the choice can be subject to social differences.[25]  In Japan痴 case, the norm of consensual decision making has constituted part of this component of policy legitimacy and has given a distinctive shape to the policy process and outcomes.  Under its influence, in short, policy advocates’ efforts at consensus building help to increase the legitimacy of a contested policy and to alleviate opposition.[26]  On the other hand, a lack of such efforts works to damage policy legitimacy. 

            Japan痴 consensus norm is process-oriented, and its inconsequential nature distinguishes it from consequential rationality which concerns the choice of action given one痴 end (Elster, 1989; Sakamoto, 1995).  Its essential feature is politicians’ aversion to the use of a majority vote in conflict resolution.  It is buttressed by their belief that a majority vote will entail confrontation and impede the orderly conduct of parliamentary affairs.  Thus, their preference is often for decisions by 砥nanimity,” which is, though, admittedly less than perfect unanimity.[27]  They conduct consensus building until they exhaust deliberations, while knowing that some of the opposition痴 claims need to be dismissed in the end.  Then, a group typically announces that a decision has been made by unanimity without taking a vote.  The opposition grudgingly drops its objections.

            The consensus norm exerts several kinds of impact on policy process and outcome.  First, politicians’ concern with the repercussions of norm violation makes it difficult for a majority party to push a contested bill without consensus building efforts with its member legislators, opposition parties, interest groups, and sometimes the public.  Norm violation and resulting protest can potentially halt the functions of the Diet and hinder the passage of bills, and ruling party politicians also fear their negative electoral repercussions.  Thus, the norm can prolong Diet deliberations, thereby limiting the governing party痴 ability to pass bills in a given Diet session which is already constrained by the short length of legislative sessions (Mochizuki, 1982).[28]

Second, the norm sometimes forces the ruling party to be forthcoming in making concessions to the opposition謡hether they be amendments of a bill at issue or other political compensations—, because intransigence calls into question the party痴 willingness to work out agreement.[29]  In passing a controversial bill, the ruling party also takes face-saving measures for the benefit of opposition parties so that they can better explain their inability to block the bill to their constituents. 

Third, by the particular way in which the norm was practiced during the LDP痴 one-party rule, the governing party sought to obtain at least one other party痴 consent (typically, the CGP痴 or DSP痴) in legislating contested a policy, although its majority status should have obviated such a need.  The inclusion of opposition parties in a temporary policy coalition was the LDP痴 effort to avoid public charges of tyranny of the majority which the party received virtually every time it passed a controversial bill by a majority vote.  This tendency began in the late 1970s and became more distinctive in the late 1980s and early 1990s.  Since 1993, in contrast, the emergence of multiparty governments has reduced the need to seek an opposition party痴 cooperation, as a coalition decision already represents a consensus among multiple parties.

 

How Policy Legitimacy Affects Policy Approval

 

            How does policy legitimacy affect the policy process and outcomes in Japan?  What exactly happens when a contested policy has legitimacy or when it does not?  In short, policy legitimacy affects the fate of a contested policy by moving the dynamics of the policy process葉he unity of the ruling party in support of its policy, unity of opposition parties in opposing it, and public opinion葉oward or away from policy approval.[30]  A high level of policy legitimacy induces high ruling party unity, low opposition unity, and public opinion favorable to a policy, and guides the parliamentary dynamics toward policy approval.  A low level of legitimacy, in contrast, causes low ruling party unity, high opposition unity, and unfavorable public opinion, making it hard to obtain policy approval.  These three factors are interrelated: they are likely to vary together much of the time either toward policy approval or rejection.

(1) Structures and Incentives.

            Parties pursue political goals under a particular structure of the game of politics, and different structures provide different sets of opportunities and constraints.  Let me briefly lay out the structure of party competition in Japan by taking the example of party interactions under the LDP痴 one-party rule prior to 1993. 

            The ruling LDP enjoyed a near monopoly of the right to propose policies, being the only party to have access to the policy making resources bureaucrats had to offer.  The opposition parties’ ability to draft policies was limited by the lack of policy expertise and staff and by the law that made it difficult for individual legislators to introduce a bill.[31]  The opposition parties’ strategy for making political gains was limited to earning credit for blocking the government痴 bills and winning as many and as large policy concessions as possible.  As the SDPJ痴 former chairperson notes, the party痴 role was, 登pposition and resistance,” and it was its 途aison d底tre” (Interview, February 18, 1994).  The history of parliamentary politics involving controversial bills in the post-1955 period was one in which the LDP proposed bills and the SDPJ opposed them.  Policy opposition and the winning of concessions were a reasonable strategy for the SDPJ in the structure where the LDP monopolized policy making resources and the SDPJ could not claim credit for supporting LDP policies.

            The middle-of-the-road CGP and DSP developed strategies slightly different from the SDPJ痴.  In the structure of competition between the major LDP and SDPJ, it was not a profitable strategy for the smaller CGP and DSP to merely ally themselves with either of the two major parties.  If the two small parties supported LDP policies, the credit for their salutary effects would go to the LDP.  If they aligned themselves with the SDPJ in successful opposition to LDP policies, the credit of blocking the policies would go to the SDPJ.[32]  Because of the CGP痴 and DSP痴 perception that their parties could not have as large an impact on the public痴 mind as the LDP and SDPJ, the smaller parties chose the strategy of distinguishing their positions from both the LDP痴 and the SDPJ痴 to make their existence stand out as much as possible and swinging back and forth between the two large parties’ positions to win political gains and wield a casting vote.  This strategy was sensible in light of the fact that the consensus norm made it difficult for the ruling LDP to pass its bills unilaterally.

(2) Policy Legitimacy痴 Effects.

            First, policy legitimacy makes possible a high level of ruling party unity in supporting a contested policy.  The party can mitigate the electoral concerns of backbenchers謡ho are vulnerable to electoral pressures預nd reduce their policy opposition within the party by exploiting strategies and devices (e.g., modifications of the distribution of a policy痴 costs and benefits, party control, the koenkai).  But to the extent that the effectiveness of their strategies is limited, party unity will depend on how effectively party leaders engage in consensus building efforts and persuade the backbenchers to drop their opposition.  If successfully engaged, the efforts will equip the backbenchers with a means to explain to their constituents their compliance with a policy opposed by the constituents.  In contrast, an ineffective consensus building effort will leave backbenchers with no such justification and will likely result in a failure to maintain party unity.[33]  Strong ruling party unity can affect the opposition parties’ policy positions and strategies and weaken their unity, giving the ruling party leverage, since the ruling party could now, at least numerically, force its bill through the Diet with its majority vote.[34]

            Second, policy legitimacy makes it difficult for opposition parties謡hich wish to get credit for defeating the ruling party痴 policy or gain policy concessions葉o persist in opposition to a policy and weakens opposition unity.  Side-payments and policy concessions to the opposition parties can make somewhat easier the ruling party痴 task of breaking opposition unity and obtaining their acquiescence.  But again, the alleviation of policy opposition depends upon whether and how successfully the ruling party engages in a consensus building effort with them and convinces them that the implementation of a contested policy is unavoidable. 

When a ruling party policy is buttressed by proper ideas and a popular mandate and is made through appropriate procedures, the opposition parties may lose justifiable reasons to continue to block the policy, except for the cause of protecting their constituents’ interests.  If they still persist in opposition, they may be regarded as unreasonable or irresponsible.  The public criticism that Japan痴 opposition parties oppose government policy for the sake of opposition and are incompetent derives from this logic.  In such a case, the ruling party will be better justified to unilaterally force its bill through the Diet, or such a move will less likely be regarded as illegitimate.  If the passage of a bill becomes likely in this manner, opposition unity will weaken, since some opposition parties may now move to win policy concessions from the ruling party in exchange for cooperation on the bill.  Other opposition parties also may follow suit and move toward tolerance or acceptance of a bill, now that they may otherwise lose the policy battle with no policy gains (Interviews with a former SDPJ secretary general and the DSP痴 policy staff member, March 14, 24, June 13, 1994).

            A consensus building effort by the ruling party and resulting policy legitimacy also make it easier for opposition parties to gracefully accept the defeat of their policy opposition, as they provide justification to drop opposition.  To give the opposition parties a means to explain their legislative failure to constituent groups, the ruling party also arranges face-saving measures, such as providing a chance for them to stage ritualistic protest against its policy and/or policy concessions and compensations.  Meanwhile, a failure to engage in consensus building and legitimize a policy will foster opposition unity, since the opposition parties can condemn the ruling party for not subjecting the policy for proper deliberations.

            Third, policy legitimacy leads the public to resign itself to a loss-imposing policy as unavoidable and makes it easier for it to swallow its discontent.  This shift in public opinion will be followed by the weakening of opposition unity and the increase of ruling party unity, since it reduces the opposition痴 incentive and justification to sustain opposition.[35] 

            These functions of policy legitimacy make possible the resolution of a conflict that might otherwise be difficult to resolve, as they make it easier for various actors to accept the costs of a policy to their immediate interests.  A lack of policy legitimacy, in contrast, provides opposition parties with legitimate reasons to oppose a contested policy.  It also robs ruling party politicians of justification to comply with the policy and impairs their party痴 unity.  Strong opposition unity leads ruling party politicians to question the political feasibility of pushing their policy and sways their determination, and this increases the opposition痴 ability to block the policy.[36]  Also, when the opposition parties are united in opposition, the consensus norm deters the ruling party from unilaterally forcing its policy through the Diet.  A lack of legitimacy is further likely to turn the public against a policy, and this situation will be capitalized upon by the opposition to the ruling party痴 disadvantage.  As a result, interparty agreement becomes difficult to emerge.

            In sum, policy approval is facilitated when a policy achieves legitimacy in the idea, democratic, and decision-norm components.  When a policy gains legitimacy, the parliamentary dynamics羊uling party unity, opposition unity, and public opinion耀hift favorably toward its approval.

 

Several remarks are in order before concluding.  First, I have explained policy legitimacy as a factor in determining the legislative fate of a contested policy and have shown how it affects policy attempts in Japan.  Standards of policy legitimacy similar to Japan痴 decision making norm may also be found in other societies.  For instance, George (1980) suggests that, in order to build a national consensus on his foreign policy, a U.S. president must meet such procedural requirements as consultation with members of Congress, the avoidance of secrecy, the pursuit of national interest, and bipartisanship.  These parallel Japan痴 norm, to some degree, and can be seen as a variant of consensualism.  However, the way similar norms are interpreted, applied, and enforced can vary across societies, and the variation can generate also divergent political outcomes (North, 1990).

            The political processes and outcomes Japan痴 norm creates may be distinct from what other decision rules elsewhere entail.  In the United States, coalition building is important for the fate of a bill, and coalition leaders’ effort focuses on strategies to create a majority.  But in Japan, a governing party cannot ensure legislation simply by securing a majority vote.  Strong partisan voting in Japan should have obviated the need for the majority LDP to make efforts at interparty coalition building.  But that was not the case; LDP bills were often shelved due to a minority痴 opposition.

Different decision rules produce different outcomes (Mueller, 1989; Scharpf, 1989).  One cannot explain legislative outcomes in Japan just by counting the number of legislators who would vote for or against a bill who have in turn calculated the number of votes they would win or lose in elections as a result of their voting on the bill in one way or the other預 method that may be slightly more plausible in analyzing U.S. Congress.

            Second, the building of policy legitimacy helps facilitate policy approval, because it also opens up room for political leadership and persuasion to play in politics.  Policy makers build policy legitimacy to carry out a contested policy that may otherwise be difficult to implement because of a conflict of interest.  Building legitimacy, then, should require that policy advocates either make relevant actors forgo their private interests or alter their preferences.  Strong and skillful leadership and persuasion are a necessity in such an effort.  Actors’ preferences develop and change within politics and society through education, indoctrination, and experience (March and Olsen, 1989).  Political leaders can transform actors’ beliefs and preferences by persuading them into new beliefs and commitments.

            The role of political leadership and persuasion does not diminish even in game-theoretical frameworks where the existence of multiple equilibria and the unraveling of cooperative tit-for-tat strategies make cooperation among self-interested rational actors difficult.  It has been said that although the strategy of tit-for-tat could solve prisoners’ dilemma situations and cooperation can obtain, so can other non-cooperative outcomes.  Tit-for-tat-based cooperation also can unravel, and tit-for-tat cannot logically be a viable strategy, if actors believe that everyone else is also rational and will defect in the final round, then, in the second-to-last round, then, in the third-to-last round.  But Miller (1990) explains that a small amount of uncertainty on the part of actors about the other person痴 possible commitment to tit-for-tat play can make a tit-for-tat strategy rational, since the unraveling problem depends on each actor痴 belief that everyone else is rational and will defect in the final round.  Therefore, it is, Miller argues, in the creation and encouragement of mutually reinforcing psychological expectations or myth of cooperation that the role of political leadership lies.  Thereby, leadership can make one feasible strategy (cooperation), instead of another feasible strategy (noncooperation), a salient and viable strategy by breeding such expectations of cooperation.

            Further, political actors are not attentive to the costs and benefits of all issues due to resource and capability constraints on their attempt to acquire such information.  If so, a policy advocate can gather support for her policy by persuading inattentive or opposing politicians and constituents of the merits of the policy while de-emphasizing its costs.  Creating or activating favorable preferences can increase the chance of policy approval.  A policy advocate may take advantage of 吐raming effects”[37] and explain her policy in such a way to make it appear more attractive.

            Third, with all its potentially positive effects on the legislation of a policy, policy advocates do not always orchestrate a consensus building effort.  For one thing, politicians do not always know whether and how much consensus building effort is necessary until they actually face opposition.  For another, policy advocates consider resource efficiency.  Consensus building protracts policy deliberations and adds to the time constraint imposed by the short sessions of the Diet (Mochizuki, 1982, pp. 55-66).  Policy advocates also wish to avoid making concessions and compensation to the opposition which may harm the integrity of their policy.

            Meanwhile, while policy legitimacy can increase the chance of policy approval, it does not mean that politicians can engineer legitimacy for any policy.  Although they sometimes concern themselves with the appropriateness and correctness of decisions and actions, they are also as much concerned with their self-interest.  It will be difficult for a policy deviating wildly from the interests of relevant actors to obtain legitimacy sufficient to justify such an deviation.

            Fourth, by stressing policy legitimacy and its normative properties, I do not wish to argue that political actors are irrational when they let normative forces affect policy outputs.  Politicians not only are constrained by policy legitimacy, but can also strategically exploit it to advance their goals.  The building of legitimacy can provide policy makers with a means to carry out an unpopular policy while saving uncertain electoral calculations and minimizing the costs of miscalculations.

At the same time, however, political actors are not rational in the way some rational-choice scholars depict them.[38]  As the tax increase cases reviewed later in this study show, policy makers make mistakes and act to correct them upon learning.  The Ohira, Nakasone, and Hosokawa administrations misjudged the political feasibility of their tax increases and suffered electoral losses or other negative consequences because of their misjudgment.  The Takeshita and Murayama administrations learned from their predecessors’ mistakes, corrected them, and accomplished their policy goals.  This demonstrates not only the limits of rational choice theory, but also the indeterminacy of policy legitimacy and its normative properties in influencing the success and failure of policy attempts.  For the cases also show that policy makers do not always succeed in building or even recognizing a need to build legitimacy for their policies. 

Ambiguities and uncertainties associated with electoral calculations and with norms are a cause of this sort of indeterminacy.  Such indeterminacy in the effects of both rational and normative factors as well as the non-trivial influence of multiple factors complicate analysis of policy making.  Tsebelis (1990, pp. 32-33) notes that rational choice explanations may be less applicable when actors’ goals are fuzzy or rules are fluid and imprecise.  But indeterminacy is caused not only by such goals and rules, but also by the fuzziness and uncertainty of social norms and rules.  Norms have ambiguity in their interpretation, prescription, and application and can be situation-specific (Majeski, 1990).  They give actors equivocal cues to their action.  Politicians may be unsure about under what conditions a norm will be invoked, what it prescribes or proscribes, or even which norm of all the others will be invoked.  If we successfully probe the independent effects of rational and normative factors and their interactions, we will have a much better grasp of political behavior and outcome. 

While much of policy process and outcome in Japan still remains indeterminate, my analytical framework of policy legitimacy provides guidance to understanding when and how politicians manage to make a contested policy or how the public approves it. 


[1] Politicians certainly pursue multiple goals and make trade-offs among the goals.  But the multiplicity of goals is problematic because there is no easy solution to the question of how they actually go about making trade-offs.  As Arnold (1990, p. 5, footnote 4) points out, Fenno (1973) avoids the problem of trade-offs by principally creating three models with one of the three goals each.  Arnold similarly finesses the problem.  I, too, do not have a solution.

[2] Party leaders’ incentives and opportunities may diverge from backbenchers’ in more than the degree of their electoral concerns.  For instance, Weatherford (1993) writes:

[T]heir greater electoral security and their leadership status lead them to give additional weight to two other goals beyond the concern for re-election.  Pursuing personal political power within the Congress is motivated by the calculation that future success depends on a reputation for winning in the past; the risk aversion this goal fosters will make legislative leaders for whom this is the primary goal especially sensitive to the danger of losing a vote, and they will 鍍rade off policy content for prospects of victory.”  Striving to promote their conception of good public policy will motivate legislative leaders to reject not only the strategy of always attempting to anticipate reactions so as to be on the winning side, but also its opposite, that of intransigent position-taking to please the constituents back home.  The concern with good public policy will lead legislative leaders towards weighing the substantive merits of different policy proposals in light of their prospects for solving the immediate problem and correcting the conditions that caused it (pp. 4-5).

[3] The building of policy legitimacy and the exploitation of strategies do not need to be mutually exclusive and can simultaneously assist politicians’ policy attempts.

[4] The logic of my contention is somewhat similar to Stern (1995)痴 argument that political leaders mobilize support for collective efforts such as war by making emotional appeals and preempting or overriding the self-interested calculations of individuals.

[5] The trading of support does not have to concern the contents of a policy.  For instance, the CGP supported the ruling LDP痴 foreign policy initiatives during the Gulf Crisis in return for the LDP痴 support for a CGP candidate in the 1991 Tokyo gubernatorial election (Sasaki, 1992, pp. 76-77).  In a different vein, the LDP also decided not to pass the Peace-Keeping Operations bill (which the SDPJ opposed) in the upper house during the extraordinary session of 1991 in exchange for the SDPJ痴 giving up its demand for a Diet interrogation of a scandal involving the LDP (Sasaki, 1992, pp. 178-179).  An extreme example of trading was an exchange of cash for political concessions among political parties or individual politicians.  During LDP rule, the LDP allegedly paid cash to opposition members to 都mooth” house management (Hirose, 1989, pp. 47-53).

                [6] There is no official figure on the number of Japanese citizens having koenkai membership.  In a poll conducted in 1979, 20 percent of respondents answered that they belonged to one koenkai or another (Miyake, 1995, p. 11).  In another account, LDP politicians tried to collect two to three times as many koenkai members as the number of votes they needed to get reelected under the old MMD system (Hirose, 1989, pp. 38-39).  On the koenkai, see, for instance, Hirose (1989), pp. 36-43; Ishikawa and Hirose (1989), pp. 123-169.  Though the koenkai has always occupied the center of candidates’ electoral strategy, the koenkai vote may decrease in relative importance in the future, since victorious candidates in the new single-member districts (as compared to the old multi-member districts) must now collect significantly more votes than can be covered by koenkai votes.

[7] The koenkai is composed of broadly two tiers of supporters.  One is those tied to politicians by personal relationships, such as relatives, friends, school alumni, and neighbors and other residents in their districts, of which exchanges of personal favors and constituency services are dominant characteristics.  The other is supporters who mobilize votes for politicians in exchange for government projects and subsidies politicians bring to their districts.  This group includes business owners and firms in the areas of agriculture, commerce, and construction, as well as governors, mayors, and members of local legislatures.  Meanwhile, Miyake (1995, pp. 22-23) shows that Japanese voters have become concerned with their own or local material interests, rather than national issues.  He reports that, in 1983, 55 percent of respondents answered that they would vote for a candidate who attended to local interests, and only 24 percent would vote for a candidate who heeded national interests.

[8] But candidates do not expect to win seats only with koenkai votes.  They also cannot expect all their koenkai members to vote for them in every election.  Also, when a policy has local or group bearings, politicians’ issue positions can have impact on koenkai support.

[9] Masumi Ishikawa, a prominent political reporter, said, 摘ven when backbenchers oppose their party痴 issue position, they can vote along the party line without suffering severe electoral punishment, because individual politicians’ voting one way or the other in the Diet does not affect their electoral prospects much, due to the unimportance of issue-voting” (Interview, January 27, 1994).  An SDPJ leader during the 1960 controversy over the revision of the U.S.-Japanese Security Treaty made a similar point, 典he LDP keeps voters attracted to the party by providing material benefits.  So national policy issues don稚 affect how voters vote, particularly in foreign policy, since they don稚 have an [economic] interest in it.  A good illustration is the Security Treaty controversy and the outcomes of the following general election.  Although the LDP痴 revision of the treaty invited very widespread, intense opposition from opposition parties, students, and activists and the LDP forcibly rammed its ratification through the Diet in the face of public protest, the party won the election in the same year (Interview, April 18, 1994).  These observations may exaggerate the degree of politicians’ freedom of action in seeking an unpopular policy.  But the limited influence of issues seems to be common to many democracies (Campbell, Converse, Miller, and Stokes, 1960; Niemi and Weisberg, 1976).  For an opposite view, see Fiorina (1981).

                [10] Analyzing a survey conducted in 1976, Kohei et al. (1991) found that “27 percent of the respondents failed to take any position on these issues and 57 percent were unable to name any party that represented their opinion on the issues” (pp. 283-284).  They write, “43 percent were able to name a party that they felt was closest to their position on a given issue, but only 24 percent mentioned the party for which they voted....If over a quarter of the voters have no positions on the issues, over half are unable to identify any party that represents their views, and over three-quarters fail to link their issue preferences with the party they voted for, then it seems unlikely that a study of issues will contribute greatly to an explanation of voting outcomes.”  Meanwhile, post-1993 Japan has seen a marked increase in the percentage of voters who do not identify with any party.  These non-identifiers now account for more than half of the entire population of eligible voters (e.g., Yomiuri Shimbun, January 7, 1995).  This is further accompanied by a high level of distrust, among both the non-identifiers and party supporters, in the ability of the political parties to manage the country and in efficacy of government.

[11] Issues may still indirectly influence the electoral vote through their impact on party images and party identification.  Fiorina (1981, p. 200) argues that the voter痴 retrospective judgments about policy instruments and outcomes have direct effects on the formation of her future expectations and on her party identification and, through these effects, indirect impacts upon her vote.

[12] The sources of party control derive from parties’ (or intraparty factions’) provision of such political benefits as candidate nomination, campaign funds, position assignments, and political influence, and from Japanese politicians’ common belief that compliance with their parties is a general principle of party politics.  There has been some change in the degree of party control since 1993 when party realignment changed the relationship between party leadership and backbenchers, and some politicians started acting slightly more freely from party discipline.  But partisan voting is still the norm.

                [13] The Japanese customarily call this ritualistic practice gasu-nuki (the releasing of gas).  On such occasions in party meetings behind closed doors, LDP politicians would protest loud, facing the doors outside which their constituent group representatives waited.  Interview with a former LDP secretary general, February 23, 1994.

[14] In my interviews, many politicians of all parties acknowledged the effectiveness of this performance.  A former LDP secretary general explained, 鄭ll of us know opposition and protest won稚 change the result and that it is a futile ritual.  But everybody still does it, and everybody hopes and expects that his constituents will understand his effort and not punish him in elections.  And the constituents know that politicians do the performance knowing it痴 not going to change anything much.  But the constituents will still understand registering opposition is the best backbenchers can do, and will resign themselves, saying, 惣ou tried that hard for us.  It can稚 be helped.樗  Another LDP politician, a former director general of the Defense Agency, said, 鄭s long as we try and register opposition, we will not lose face with our constituents.”  Interviews, February 18, 22, 23, March 8, 30, May 20.

[15] In the 1989 upper house election that followed the beef and orange import liberalization, the LDP did suffer a major setback.  But studies show that the liberalization did not have a significant impact on the electoral outcomes and, instead, the new consumption tax of 1988 had crucial influence (Kabashima, 1992; Miyake, 1992, 1995).  Other recent cases of policy changes made possible by external pressure include the government efforts to increase the sales of foreign semiconductors in Japan (1986); the implementation of measures to reduce structural barriers in Japan痴 markets to the operation of foreign firms and the sales of foreign products (1990); and financial contributions to Japan痴 allies’ efforts in the Gulf Crisis (1990-1992), and the legislation of its Self-Defense Forces’ participation in U.N. peace-keeping operations.

[16] The realism of the neoclassical assumption of universal rationality and its utility for political analysis vary depending on particular situations, structures of interactions, the nature of issues, and the possibility of learning, among others (Tsebelis, 1990; Reed and Sakamoto, 1996).

[17] March and Olsen (1989) write, 菟olitical institutions and the individuals in them need to communicate to their observers that the decisions they make are legitimate....by showing that the decisions accomplish appropriate objectives or by showing that they are made in appropriate ways....So, political actors establish that they are good decision makers by making decisions in a way that symbolizes the qualities that are valued (p. 49).

[18] Although policy legitimacy is postulated as a force that can counterbalance the forces of self-interest, it is not entirely free from strategic considerations.  For actors’ perception of legitimacy cannot help but be affected partly by a match between a policy and their interests.  Also, policy makers can strategically seek to contrive legitimacy to advance their own interests, although this sort of action is still constrained by people痴 conceptions of appropriateness.

[19] Policy ideas gain importance as guidance to politicians’ actions, also when they face uncertainties about constituents’ preferences and the electoral impacts of policy instruments and thus when electoral incentives provide a poor guide to their decisions and actions.  Ideas also influence actors’ perception of the nature of a problem or even its existence.

[20] Ferejohn (1991) notes the role the study of intersubjective understandings can play in solving rational choice theory痴 problem with multiple equilibria.

                [21] Although these principles remain effective, the early 1990s saw some change due to the relative rise of Japanese power and the end of the Cold War.  The change made possible, for instance, the legalization of Japan痴 Self-Defense Forces (SDFs)’ involvement in U.N. peace-keeping operations.  The Socialist Party痴 policy change on the SDFs and the U.S.-Japanese Security Treaty in 1994 also reflect the change.

[22] Public opinion affects policy processes and outcomes both through intervention by politicians’ electoral concerns and through its own normative appeal.  Politicians certainly exploit public opinion to advance their interests.  But without its independent normative appeals, they would not be able to exploit it.  Also, there is no clear definition of public opinion in politicians’ use of the term, and its definition and measurement are subject to multiple interpretations and exploitation.

[23] The influence of different types of elections on policy legitimacy varies depending upon a particular situation.  But lower house elections are generally the most important.  They are more directly linked to policy deliberations in the Diet, because both the ruling and opposition parties use the elections as a weapon in policy competition.  The lower house can pass a non-confidence motion against or reject a confidence motion for a cabinet.  When either of these happens, the cabinet must either resign en masse or dissolve the house and call for a general election (The Japanese Constitution, Article 69).  Opposition parties threaten to submit a non-confidence motion against a ruling party cabinet in opposition to its bill or house management.  The prime minister can also use his power to dissolve the house as a threat to opposition parties and even his own party痴 legislators, when he wishes to induce their support for or weaken their opposition to his bills (The Japanese Constitution, Article 7).

[24] There are caveats.  Public opinion can exert influence on the fate of a policy only under certain conditions.  First, a policy needs to be visible enough to allow the public to be informed and form an opinion.  Second, there need to be actors預rticulate politicians, the media, or interest groups謡ho champion the preferences of the public.  That is, public opinion may have little influence unless powerful actors exploit it or a threat of electoral retribution is imminent.  The Kaifu and Miyazawa administrations’ failed electoral reform attempts (1991, 1993) exemplify cases where public opinion strongly supported a particular policy, but did not exert influence on the legislative outcomes; the reform bills were shelved because a majority of politicians and parties did not want reform.  Third, because of ambiguity and open-endedness in its interpretation, the same public opinion can be invoked by both sides of a policy conflict.

[25] Mueller (1989, ch. 6) explains that different assumptions about the nature of politics and issues lead some scholars to claim the desirability of majority rule and others to advocate unanimity rule.  It is natural to suspect that people痴 preferences for rules likewise vary with their beliefs and values.  As North argues, culture 電efines the way individuals process and utilize information” (1990, p. 42).  Different societies may have different theories and ideas about a problem, different resources, and different solutions to the problem, and culture is part of what embodies these differences.  Societal divergence stems from the path-dependent nature of the social and political development (North, 1990, ch. 11).  Social reality is a result of past events and choices.  When two societies take different developmental tracks, the particular courses can eliminate other alternatives, and their initial divergencies tend not to disappear and can be reinforced.

[26] March and Olsen (1989, p. 50) write that one of the things that choice in institutions is orchestrated to assure its audience is that the 田hoice is sensitive to the concerns of relevant people, that the right interests have been heard in the process.”

[27] Unanimity rule is used in many arenas of Japanese politics.  It is used, albeit with occasional exceptions, in decisions made in conferences of committee directors and the House Management Committee of both houses of the Diet regarding the scheduling of deliberations and votes, and other parliamentary affairs.  It is also the principal decision making rule in cabinet meetings and the LDP痴 Executive Council and committees of the Policy Affairs Research Council, as well as the SDPJ痴 top decision making organs such as the Central Executive Committee, Secretariat, and Diet Strategy Committee.  Unanimity rule is nowhere officially stipulated.  But Japanese politicians subscribe to it and rarely use majority rule to resolve disputes.  But despite their emphasis on unanimity rule, they have no general or specific rule for the size of a majority required for a unanimous decision.  When asked, a former secretary general of the SDPJ told me, 的 don稚 know what kind of majority you need.  It depends.  But I would try to get the consent of 70% of participants before declaring a unanimous decision.” Interview, June 13, 1994.  It should be noted that my use of the terms majority and consensus differs from Lijphart (1984)痴.

One could still stipulate a preference ordering for processes independently of outcomes and conceptualize Japanese consensualism as 途ational” in terms of the correspondence between goals and means.  But this preference ordering would be a non-universal, local one and would itself need to be investigated.  Also, one could set forth a rational explanation of the norm that it is actually the optimization of collective benefits (e.g., the norm prevents tyranny of the majority or unnecessary conflict in the Diet).  But this is essentially distinct from rational choice explanations based upon self-interest, and raises the question, 展hose utility do they maximize?”  Elster (1989, pp. 130-151) cogently argues against the collective rationality argument by noting that not all norms are Pareto-improvements; that some norms that have good consequences are actually not followed; and that the explanation still does not explicate the existence and maintenance of norms.

[28] Although the LDP maintained a majority for its 38-year rule except 1983-86, the legislative process under its one-party dominance was characterized by opposition parties’ great ability to delay Diet deliberations and a not so high success rate for the LDP government in having its bills approved.  Due to the use of unanimity rule in Diet committees, the opposition parties stalled deliberations easily and frequently.  And bills that do not get voted on within the session they are introduced will be shelved unless parliamentary parties formally agree to continue their deliberations after the session.  The scrapping of government bills resulted often from the opposition parties’ attempt to delay deliberations until it became unlikely that the bills would pass in the session.  In such a situation, the LDP was unable to resume deliberations despite its absolute majority and which led the party to make political concessions to the opposition parties to make them agree to resume them.

[29] This compensation requirement is one factor that often makes results of conflict resolution equivocal.  Such settlements result from parties’ concessions to each other痴 demands and have the characteristics of being the least unacceptable to each party.  They are ambiguous in their meanings and subject to multiple interpretations.

[30] These three factors are the yardsticks Japanese politicians use in assessing the situation surrounding a governing party痴 attempt to legislate a policy and deciding and modifying their policy positions and strategies.  In my interviews, many legislators, their secretaries and staff members, and bureaucrats constantly referred to the three as important factors in deciding the fate of a controversial policy.  Their importance is illustrated also by the media痴 frequent reference to them in political reports.

[31] The submission of a bill in the lower house requires twenty co-sponsoring legislators and fifty co-sponsors in the case of a bill that requires a budget (Iwai, 1989, pp. 64-66).

                [32] Consider the case of the successful alliance of the CGP, DSP, and SDPJ against the Nakasone administration痴 sales tax bills in 1987.  Although their opposition scrapped the bills, only the SDPJ gained in the 1987 nationwide local elections that took place right before the administration gave up the bills.  This led the CGP and DSP to keep distance from the SDPJ in the LDP痴 next attempt at a consumption tax and to tacitly cooperate with the LDP.

[33] For most bills, governing party leaders have garnered their backbenchers’ approval by the time bills are introduced to the Diet.  In Japan, all bills are subjected to examination and approval by the ruling party before they are submitted to the cabinet and then to the Diet.  Under LDP rule, deliberations in the divisions of its Policy Affairs Research Council and the Executive Council are steps toward achieving intraparty approval.  This system ensured most of the time that bills had the support of ruling party politicians by the time of their introduction to the Diet.  But intraparty approval thus obtained can sometimes be unstable, and intraparty opposition can recur or emerge particularly when opposition parties orchestrate strong opposition, when public opinion turns unfavorably, when the ruling party loses in an election that takes place during Diet deliberations, and when intraparty power struggle arises.  In such cases, party leaders will be forced to abandon their bill or make a renewed effort at consensus building.

[34] A former secretary general of the SDPJ explains: 鄭 high degree of the LDP痴 determination and unity in support of its bill worked to weaken the SDPJ痴 opposition, because it made the scrapping of the bill difficult.  On the other hand, if the LDP was neither determined nor united, the possibility of the bill痴 being scrapped would become high, so the SDPJ would intensify its opposition.  The SDPJ痴 opposition would gain even greater force, when the opposition parties were united against the bill” (Interview, June 13, 1994).

[35] Political parties modify policy positions and Diet strategies, depending partly on the nature of public opinion.  A former vice secretary general of the cabinet (LDP) said: 撤ublic opinion is one of the most important factors determining the fate of a bill.  We don稚 worry about it while drafting laws.  But we do worry about it in the process of Diet deliberations, because opposition parties decide their positions by looking at it and exploit it in attacking our policy” (Interview, June 2, 1994).  A former secretary general of the House of Representatives further pointed out: 徹pposition parties judge from public opinion whether the ruling party would provoke public criticism if it forced its bill through the Diet.  And, if they think they can have the public on their side, they will put up strong opposition to the bill.  If all the opposition parties are opposed, the ruling party cannot bulldoze its bill.  For doing so would put the Diet into chaos and it would stop functioning.  The ruling party could not afford to let that happen” (Interview, June 6, 1994).  The flow of influence between policy legitimacy and public opinion, then, goes in both directions; for public opinion also affects legitimacy, as the former is a component of the latter.

[36] A former SDPJ secretary general said: 徹ur effort to refuse deliberations and votes would be more successful if the opposition was united.  Our unity would make a difference in the pressure we could put on the LDP.  For instance, the conference of directors of the House Management Committee is composed of nine directors and a chairman.  Let痴 say the LDP had five directors and a chairman, the SDPJ two directors, the CGP one, and the DSP one.  We never take a vote at its meetings (unanimity or near-unanimity is the decision making norm).  But if the opposition parties were united, the power balance would be five (LDP) vs. four (SDPJ, DSP, CGP).  In such a case, the LDP could not get its way.  But if the DSP and CGP sided with the LDP, it would be seven (LDP, CGP, DSP) vs. two (SDPJ).  In this case, there was virtually no pressure on the LDP, and it would get its way” (Interview, June 13, 1994).  Also, when opposition parties are united in their opposition, public opinion is likely to be against a ruling party policy, because parties decide positions and strategies by assessing public opinion.  Interview, June 6, 1994.

                [37] Human decisions are influenced by the way their choice problems are represented; that is, different representations of the same choice problem produce different preferences.  This is called framing effects.  See Tversky and Kahneman (1990).

[38] Rational choice scholars might claim that norm-compliant behavior can be accounted for within their theory by accommodating norms and conventions into actors’ utility functions or by taking account of the long-term interests of cooperation.  One could certainly stretch the definition of self-interest to include such conventionally non-self-interest factors as altruism, compliance with norms, and ideology.  But such an inclusive definition of utility functions would deprive rational choice theory of its methodological and theoretical merits用arsimony, logical rigor, testability, falsifiability, and tractability.  We can gain a more meaningful understanding of political behavior and outcomes by treating social forces as those that interact with self-interest but can also exercise an independent influence on behavior (see also Elster, 1989, ch. 3; Kingdon, 1988).  Rational choice scholars might also claim that their theory can explain miscalculations and mistakes by incorporating imperfect information and uncertainty into their models.  But by allowing the possibility of mistakes or by including non-self-interest factors in utility functions, rational choice theory runs the risk of becoming nothing but ad hoc explanations (see Tsebelis, 1990, p. 40).