1 Background of Reform

1.1 Why Is the Electoral System Important?

Electoral reform occurred in the first phase of the political reforms of the 1990s, and had a decisive impact on Japanese politics in the ensuing years. As was noted earlier, the targets of political reform were far-reaching. In this context, why was electoral reform so significant? Would it not be more important to raise the quality of politicians or the judgment and “moral standard” of the electorate? While somewhat abstract, it is both necessary and worthwhile to begin by considering and clarifying the effects of electoral systems.Footnote 1

The key is that the electoral system regulates the relationship between voters and politicians, which is fundamental for a democratic system. In every political system, there exists a ruling elite that is involved in the day-to-day management of the government, and the masses who are not. A democratic system is distinct in that voters (the masses) select the politicians (the ruling elite) through elections. Thus, the electorate participates, albeit temporarily and indirectly, in the management of the public sphere. Theoretically, voters can exercise significant control of politicians through means other than elections (extra-electoral political participation) and direct involvement in the management of the public sphere (direct democracy).Footnote 2 However, our discussion here will be limited to representative democracy (indirect democracy), which is the most common globally and the most pertinent to contemporary Japan.

Under a representative democracy, the fact that voters elect their representatives gives legitimacy to policy decisions. The electoral system determines the method by which voters choose politicians, and thus defines the relationship between the masses and the ruling elite. It is the most important, foundational rule in democratic systems.

To put it more simply, if the electoral system changes, both voters and politicians are forced to change their behavior. For example, let us compare the cases of an electoral system in which 15 people are elected from one electoral district (multi-member districts: MMDs) and an election system in which only one candidate is elected (single-member districts: SMDs). In contemporary Japan, the former system is used in municipal councils outside ordinance-designated cities (seirei-shitei toshi), while the latter is used for single-member districts in the House of Representatives and House of Councillors.

Under a system in which many representatives are elected from a single district, the threshold for the share of the votes necessary to win election is lower. I will omit the equation for calculating the threshold here, but in the case of a large electoral district with 15 seats, a vote share of slightly more than 6% is sufficient for victory. Since it is possible to win even if one advocates ideas and policies that only appeal to a small portion of the electorate, it is easy for politicians to run as independent candidates or to belong to small parties focused on minor issues. From the perspective of voters whose views differ from those of the majority, there is a decent chance that if they cast their vote for the candidate closest to their own way of thinking—the candidate they truly prefer—that candidate will be elected.

However, this is not the case in a SMD system, where only one person is elected. Since the threshold for guaranteeing victory is 50%, it is difficult to be elected based on niche interests or ideas. Instead, politicians have no choice but to “put aside minor differences for the sake of the common good,” and voters may be forced to make a “second-best choice” by forsaking their favorite (but niche) candidate for one that has a better chance of winning.

1.2 The Electoral System and Policy

Moreover, when the behavioral incentives for voters and politicians change, there is a high likelihood that the government’s decision-making process and policies will also change. The electoral system makes a difference in the formation of a majority in the legislature, and has a great impact on the state of policy.

In the case of an electoral system in which it is possible to use niche ideas to get elected, as with multi-member districts (MMDs), the legislature will be comprised of many small parties and factions with only a handful of members each. These forces, whose raison d’être lies in emphasizing “small differences,” are not going to settle for joining the majority to compromise in the legislature. Since passing legislation naturally requires a majority, it is necessary to incorporate those forces that are reluctant to compromise, but this requires extensive discussion and persuasion. As a result, decision-making takes time, and there is a strong tendency for policies to incorporate the views of small parties. In the case of a SMD system, however, elections generally result in one large party securing a majority of seats. This makes it easier to pass legislation, as the views of small parties can be discarded.

Let us look at the electoral system’s impact on public policy, using economic policy as an example. In the case of a MMD system, depending on what the minor parties are interested in, micro-targeted satisfaction and consent of specific groups are emphasized over macro-level rationality and universality for society as a whole, which can lead to policies that are detailed and customized but also guided by particularistic interests. Typical examples are individual subsidies that vary by industry or region, or pension systems that vary by occupation. By contrast, under an electoral system with high vote thresholds to win seats, as in a SMD system, minority ideas and interests are discarded, and there is a tendency toward policies that favor the macro over the micro, as decided by the leadership of major parties. There will be a stronger orientation toward comprehensive transfers of financial resources to local communities instead of individual subsidies, and the establishment of a common pension system for all citizens.

The significance of differences in electoral systems is so great that it is convenient if we can grasp these differences with a measure that is as easy to understand as possible. Therefore, in comparative political science today, the concept of “proportionality” is used to measure and understand differences in electoral systems. Proportionality is measured by the degree to which the distribution of each party’s seats in the legislature approximates the distribution of the electorate’s votes for parties.Footnote 3

Generally, the seat distribution among parties—and by association, their legislative power—will more closely approximate their vote share if there are more seats per constituency, and if a proportional representation system (where seats are distributed to parties in proportion to how the electorate votes) is used instead of a system with many majoritarian districts (where plurality rules are used to determine winners). The closer the electoral system is to these principles, the more proportional it will be. Conversely, an important factor that reduces proportionality is the over-representation of large parties and the under-representation of small parties relative to their voter support. If a party with the support of about 30% of the electorate wins 60% of the seats, while a party with the support of 30% of the electorate does not win even 10% of the seats, the electoral system’s proportionality is low.

To rephrase the earlier discussion using this concept, in highly proportional electoral systems, policies reflect the interests of electoral and legislative minorities, but in less proportional systems, those of majorities are emphasized. It is fair to say that a highly proportional electoral system leads to policies that focus on the differences and distinctiveness of the people and groups that make up society, while a less proportional system leads to policies that focus on the commonalities of the people and groups that make up society. Policies that emphasize the difference between groups and individuals are called “particularism,” and policies that emphasize commonalities are called “universalism.” From the perspective of various groups, particularism increases the degree of satisfaction with policies, but it is easy for society as a whole to lose its rationality, while it is easy for universalism to ensure the rationality of the whole but also reduce the satisfaction of individual persons and groups.

1.3 The Problem of the Medium-Sized Electoral District System

At the beginning of the 1990s, the problems with the national electoral system were widely recognized. In particular, criticism of the medium-sized electoral district system, used for the House of Representatives was growing.

As mentioned in Chap. 2, the medium-sized electoral district system is an electoral system in which two to six people are elected from a single electoral district. Theoretically, among plurality systems, where winners are decided in order of their individual vote shares, this is classified as a multi-member district (MMD) system. In the case of Japan, there were mostly three to five winners per district, and voters could vote for only one candidate (single non-transferable voting system, or SNTV). Under the combination of MMD and SNTV (MMD-SNTV), it was common for the vote threshold needed to win a seat to fall below 20%. As a result, the balance of power among parties was closer to that of a proportional representation (PR) system, but to win a legislative majority parties needed to elect multiple candidates per constituency. The high proportionality tended to produce policies with a stronger individualistic or targeted orientation, but also to create a hotbed for particularistic politics.

This kind of electoral system had few parallels in the world, and the state of Japanese politics at the time, especially particularistic politics catering to special interests and political corruption, was considered to be greatly influenced by the MMD-SNTV system.

For example, political scientist Yamaguchi Jirō, who was a leading advocate of electoral reform and a critic of the LDP administration, argued the following in a widely read book at the time. The MMD-SNTV electoral system “obscures the significance of the choice between parties by having multiple candidates from the same party stand for election.” He added, “Fierce competition among candidates will unfold. Moreover, dishonest competition using money will be rampant. In this sense, the medium-sized electoral district system [MMD-SNTV] most easily fosters corruption.”Footnote 4

The same kind of argument was also made by American political scientists. Mark Ramseyer and Frances Rosenbluth, in a work originally published in 1993, pointed out that the MMD-SNTV electoral system induced competition among LDP candidates, which “has led the LDP to adopt a more particularistic set of policies than would be likely under a number of conceivable alternative electoral schemes.”Footnote 5 Such studies have produced a consensus among political scientists around the world, and influenced the views of Japanese politics among foreign media and policymakers, and are thought to have bounced back indirectly to debates within Japan.

1.4 Issues That Should Be Addressed

What were the challenges facing Japanese politics during this period? These can be divided into two.

The first was the change in the international political and economic environment. Postwar Japan suffered enormous human and material damage through its defeat in the war, lost all of its overseas colonies, was completely shut out of international organizations and the like, and started from the position of being occupied by the victorious Allied powers, led by the United States. However, both during and after the Occupation, U.S. policy towards Japan during the Cold War was generally tolerant, treating Japan as the junior partner in the alliance relationship while expecting Japan to assume a leadership role among the liberal countries in Asia. With this help, Japan gradually returned to the international community, and after going through a period of rapid economic growth from the late 1950s until the early 1970s, Japan joined the ranks of the advanced industrialized countries as the world’s second-largest economy. During this period, the influence of the United States, which had created the liberal postwar international economic and political order, gradually declined, and from the late 1970s onward the U.S. began to demand an appropriate sharing of responsibilities by other advanced industrialized countries, including Japan. This kind of change materialized in the 1980s in the form of U.S.-Japan economic friction and demands to strengthen Japan’s defense capabilities. However, Japan tended to perceive this as U.S. pressure on Japan (gaiatsu), and repeatedly made minor concessions aimed at temporarily easing the pressure.

This trend became stronger with the end of the Cold War at the end of the 1980s. During the Cold War, Japan invoked the constitution’s war-renouncing clause to keep its defense buildup to a minimum, and enjoyed the benefits of a liberal economy for its manufacturing exports even as it protected its agriculture, distribution, and other industries. But the confrontation with the greater “evil” of the Communist bloc meant that the United States largely tacitly consented to this situation. After the Cold War, however, that “evil” no longer existed. Rather, it was necessary for the advanced industrialized countries to cooperate to prevent the escalation of regional wars and ethnic conflicts, and to support the economic independence of the former Communist-bloc countries.

The problem of the inability of Japanese politics to respond to the need for cooperation among developed countries was made apparent by the Gulf War of 1991. Faced with the Iraq’s invasion of Kuwait the previous year, Japan’s only response was to provide financial assistance, notwithstanding the unanimous adoption of UN resolutions by the five permanent members of the UN Security Council (the United States, Russia, the United Kingdom, France, and China). The main reason was that under the prevailing interpretation of the constitution, Japan could not participate in any overseas activities with a military aspect, even if there was an international agreement. And of course, it cannot be ignored that there was still a strong apprehension among opposition parties and even within the ruling LDP of “being drawn into an American war.”

Other challenges were worsening political corruption and electoral vote disparity (inequality in the value of one vote across districts or malapportionment). Political corruption scandals have not been uncommon in Japanese politics since before World War II. Generally, where political power exists and is involved with financial resources and the exercise of regulatory authority, it cannot be divorced from corruption. However, the extent to which it is large-scale or the extent to which it catches the public’s attention depends on the circumstances.

In the case of postwar Japan, perhaps since the Lockheed bribery scandal in the mid-1970s, voters have cast a more critical eye toward political corruption under the LDP’s long-term dominance. The background for this was likely that, despite the end of high-speed growth and Japan’s becoming a member of the developed world by virtue of its economic size and social stability, there was a sense that the state of Japanese politics fell short. There was also a perception that the LDP was enthusiastic about particularistic, redistributive politics toward rural areas, and that this was the foundation for its hold on power. The increase in political funds and resources necessary not only to compete in legislative elections and LDP leadership elections, but also for the daily provision of services to supporters like those in the personal support groups of politicians (kōenkai), had become known and was the target for criticism. The Recruit scandal of the late 1980s enflamed the criticism that had come to pervade the political atmosphere.

The “vote disparity” or malapportionment issue, i.e., inequality in the value of each voter’s ballot in House of Representatives elections, was also serious. Due to postwar urbanization, the number of voters in metropolitan areas grew rapidly, but the number of seats across regions was not sufficiently reapportioned to reflect this trend. In the 1960s, equality in the value of votes became strictly required in the U.S. House of Representatives and state legislatures. Lawsuits were also filed in Japan from the 1960s onward to address imbalances in the number of seats, including in local assemblies, and a Supreme Court decision in 1976 recognized that equality in the value of votes for the House of Representatives is a constitutional requirement.Footnote 6 In the mid-1970s, a “market standard” or norm had developed whereby a population disparity of up to three times per member (the ratio of voters per legislative seat between the most and least populous districts needed to be less than 3:1) was generally acceptable, and this view was shared by the Diet and the Supreme Court. However, criticism grew that the basis for the three-to-one ratio was unclear and that the remedy should not be limited to increasing the number of seats in the House of Representatives. In essence, dissatisfaction with political corruption also bled into a harsh evaluation of the limited, “self-approved” correction of the vote disparity imbalance by legislators.

2 Toward the Modernization of Politics

2.1 Multiple Directions

Changes in the international political and economic environment, worsening political corruption and the persistent imbalance in votes all influenced the direction of reform. The direction of the reform to be pursued differed according to which of these issues was deemed most important.

This issue illustrated the inadequate response capacity of Japanese politics. Response capacity refers to the ability to confront challenges directly, consider responses appropriately not only for the short term but also for the medium to long term, and to decide on appropriate policies. The Cold War, the U.S.-Japan Alliance, and the LDP’s long-term dominance were established, given conditions, and movements to pursue a different domestic policy line than particularistic politics, or to adapt foreign and economic policies to changes in the international environment, failed to gain momentum. This was not apparent in the 1980s, when Japan was intoxicated by the words “Japan as number one” and the Bubble economy, but shortly after the change in emperor from Shōwa to Heisei, difficult challenges thrust themselves into Japanese politics. The political scientist Sasaki Takeshi, who would later advocate reforms in many areas, published a book in 1987 called What Is Possible in Politics Today? in which he sharply pointed out that Japanese politics was characterized by inertia, and that its response to new challenges like economic friction with the United States was sluggish and inadequate.Footnote 7 This book was highly regarded and received the Yoshino Sakuzō prize, and went through additional printings in a short period after publication. Furthermore, the response to the Gulf War, where policies took a long time to formulate and, when undertaken, were not favorably received internationally, led Ozawa Ichirō and other leading politicians in the LDP to feel the validity of Sasaki’s criticisms.Footnote 8

The key to increasing response capacity was to foster more inter-party competition, or, in other words, a party politics accompanied by a sense of competitive tension due to the possibility of a change of government. The 1955 system, born out of the unification of the Socialist Party (between its Left and Right wings) and the conservative merger (between the Liberals and Democrats), was initially touted as a two-party system of the LDP and the JSP, but in reality the LDP secured twice as many seats as the JSP and secured long-term dominance over Japanese politics. Through this process, the Socialist Party effectively gave up its pursuit of a legislative majority, and came to seek its raison d'être in principled criticism of foreign and security policies. On the other hand, the LDP, which grew accustomed to long-term single-party dominance, found a taste for distributing the “pie” of high-speed economic growth through redistributive, particularistic politics. Voters who were critical of the LDP but dissatisfied with the Socialist Party gave seats to small parties like Komeito, the Democratic Socialist Party, and the Japanese Communist Party, but these actors could not participate in government or implement policies without the cooperation of the LDP or the JSP, so their role was limited.

It was thought, as mentioned previously, that this stagnation in party politics and fixed division of labor was due largely to the multi-member district with single, non-transferable voting (MMD-SNTV) system. Given its relatively high level of proportionality, the MMD-SNTV system makes it difficult for seat shares and power relations between parties to change. Following the LDP’s emergence as the sole majority party after the 1955 conservative merger, dynamic inter-party competition for power was significantly weakened. The high proportionality also led to a multiplicity of parties other than the LDP. As a result, the non-LDP parties had no choice but to propose multi-party coalition government, a situation in which policy disagreements and other conflicts could easily become apparent. Although actors at this time did not necessarily organize the full effects of reform in these terms, based on the perceptions I have described here, abolishing the MMD-SNTV system was explored as a starting point for electoral reform.

From the latter position—addressing political corruption and electoral malapportionment—it was recognized that the biggest problem was the inability of politicians and parties to self-correct. As mentioned earlier, corruption has long been a problem in political history, and in many cases, it has been addressed by strengthening penalties for corruption and changing political finance regulations. If so, it would be important to impose strict sanctions on politicians who have caused problems, such as by strengthening the Public Offices Election Act’s guilt-by-association system (penalizing candidates for illegal behavior on the part of their campaign staff and family members), or strengthening restrictions on the use of political funds, for instance by making political financial records more transparent. Regarding malapportionment, it was possible for the House of Representatives to apportion seats in a way that would greatly reduce the “inequality of one vote.” Vote inequality was harder to address in the House of Councillors, due to that chamber’s staggered elections and its use of proportional representation.

Whether realistic or not, correcting the imbalance in the number of seats would mean a significant increase in the number of seats in urban areas, which would also solve the problem of excessive redistributive policies towards rural areas, and increase the possibility of a government turnover should parties be able to obtain the support of urban voters. Therefore, it was thought that legislative reforms within the framework of the current system should be pursued.

2.2 Choice of Philosophy Based on the “Outline for Political Reform”

What was lacking in Japanese politics: adequate responsiveness or the ability to self-correct? Perceptions on this point became a determining factor in whether or not to embark on electoral reform, and led to decisive choices that permeated the entirety of the political reforms that followed. If politics as a whole lacked responsiveness, it was impossible not to think that the primary problem was not the quality of individual politicians, but rather a structural or institutional problem, including relations between parties. If there was a lack of capacity for self-purification, the first step needed to be to learn from anti-corruption and political finance regulations in other countries, and to carry out a large-scale correction of the imbalance of seats under the MMD-SNTV system.

The key to the choice was the LDP. If the LDP, as the ruling party occupying the number one position in both houses of the Diet, did not approve, no reform could be realized. Unless LDP lawmakers recognized the importance of the issue, reform plans—no matter how sound and meaningful they were—would be nothing more than empty words from outsiders. On this point, Sasaki Takeshi would later argue that the driving force of reform was “a naïve feeling that ‘politics as usual’ cannot go on.”Footnote 9 The key here is what is meant by “politics as usual” and in what sense it “cannot go on.” This was answered succinctly in the LDP’s “Outline for Political Reform” (Seiji Kaikaku Taikō).Footnote 10 This document was released on May 23, 1989, and the Political Reform Committee, which was established on December 27, 1988, played a central role in its drafting. Former Chief Cabinet Secretary Gotōda Masaharu served as its chairman.

The “Outline for Political Reform” has several striking features.

First, while being in principle a response to voters’ distrust of politics stemming from the Recruit scandal and criticism of the LDP, it essentially advocates for electoral reform. While the “Outline” begins by emphasizing “our conviction that political ethics…should be based on individual self-awareness,” it then states, “we are keenly aware that self-regulation alone is insufficient in the face of political structures that necessitate large amounts of political financing, especially the electoral system.” The “Outline” asserts: “Therefore, we must conduct a fundamental review from the standpoint that most of the problems are caused by the defects of the current medium-sized electoral district system [MMD-SNTV].” The “politics as usual” touched on earlier is “politics that costs money,” and “what is wrong” is political activity under the MMD-SNTV system.

Second, the change from MMD-SNTV to an electoral system centered on single-member districts is linked to the organizational modernization of the LDP. Calls to modernize the LDP had been repeatedly heard since the 1960s. Specifically, reformers advocated the dissolution of factions and the transition to a modern party organization, as typified by the Social Democratic Party of Germany and the Labour Party of the United Kingdom.Footnote 11 In its account of party modernization, the “Outline” discusses the problems of factions and policy tribes (zoku). With respect to the selection of parliamentary candidates, it states, “it is necessary to create new rules that open the way to politics to talented people from a wide range of fields, and make it possible to discover and promote newcomers.” It further adds, “In particular, with regard to fundamental reform of the electoral system, the establishment of more stringent criteria for nominating party-approved candidates and more stringent measures to deal with non-approved candidates” must be pursued.

A third feature is that the lack of responsiveness is not discussed. It is clear that although the “Outline for Political Reform” was prepared in the first half of 1989, the international events that would later symbolize that year all occurred after it was published. In other words, the post-Cold War world had still not taken shape when the “Outline” was being prepared, with China’s Tiananmen Square massacre in June and the fall of the Berlin Wall in November. The problem of responsiveness is, ultimately, closely linked to the question of whether to recognize the possibility of a change of government. It is an issue of the ability of party politics as a whole to offer multiple different options in response to new environmental conditions. It may therefore have been difficult to emphasize this in an LDP internal document, even though it was a broad outline.

2.3 Complementary Logic

The framework of the argument, as presented in the “Outline for Political Reform” which proposed electoral reform on the basis of self-purification or self-correction, has the impression of being built upon the experiences and feelings of LDP Diet members, leaving doubts about its deeper logic or reasoning. As mentioned earlier, it is quite possible to argue that if the ability to self-purify is lacking, the problem can be dealt with by tightening the Political Funds Control Act and enacting anti-corruption legislation. Since the unchecked imbalance in seat apportionment also indicates an inability to self-purify, it should be possible to solve this by reducing the number of seats in the House of Representatives and making major changes in the geographical distribution of seats. Of course, according to the views of Yamaguchi Jirō and the American political scientists raised earlier, it is possible to think that the MMD-SNTV system reduced the capacity for self-purification, but it was not until the 1990s that this idea began to attract attention.

However, the world-historical event that was the end of the Cold War soon compensated for this deficiency in logic. The first report of the Eighth Electoral System Advisory Council (Senkyo Seido Shingi-kai), submitted on April 26, 1990, immediately after the fall of the Berlin Wall, shows the germ of this trend. The Electoral System Advisory Council, an advisory body to the prime minister established by law, did not convene often, and this was the first time it had met in 17 years. The chairman was Kobayashi Yosoji, former administrative vice-minister of the Home Ministry, chairman of the Japan Newspaper Publishers Association, and president of the Yomiuri Shimbun. The vice-chairman was constitutional scholar Satō Isao, and political scientist Horie Fukashi was the chairman of its first subcommittee, which dealt with the basics of the electoral system.Footnote 12

The report begins with the sentence, “Today, our country is under pressure to solve an accumulation of domestic problems, and internationally, it is called upon to make positive contributions to world peace and prosperity.” The writers were clearly aware of and sensitive to the Japanese state’s capacity to respond to changes in the international and domestic environments. The statement then goes on to say in concrete terms, “In order for politics to adapt to changing times in our country’s current domestic and international conditions, it is necessary to reflect the will of the people accurately, and, at the same time, to aggregate the will of the people to clarify decision-making in politics and the attribution of responsibility. In addition, for a dynamic and healthy parliamentary democracy, it is necessary to maintain tension in politics through a change of government. In satisfying these demands, when comparing a single-member district system and a proportional representation system, the single-member district system better conforms to these demands.”

This recognition also coincides with the prospectus of the “Political Reform Forum,” (Seiji Kaikaku Fōramu), which held its first meeting on October 9, 1989. The Political Reform Forum was an organization formed by a multi-partisan group of lawmakers, private citizens, and academics, led by the Social and Economic Congress of Japan (Shakai Keizai Kokumin Kaigi), a private economic organization, that became an important force in promoting electoral reform. The prospectus of the Political Reform Forum stressed “a comprehensive review of the entire postwar political system and to carry out fundamental reform with a view towards the twenty-first century,” in order to “address the issues of administrative reform and fiscal consolidation, as well as tax reform, advanced information technology, and international friction.”

The anticipated or hoped-for image of politics presented in these documents has two major characteristics. The first is the existence of several powerful, modern, organized political parties that are deeply rooted in society and supported by a pyramidal structure reaching from a large number of ordinary party members to a small number of senior party executives. The second is that competition among these parties would produce governments that can quickly decide upon policies that are formulated based on the consistent policy goals of the ruling party. It was thought that the responsiveness of the political system as a whole would be bolstered by the combination of these two characteristics. What was imagined was probably the ideal type of late twentieth-century British politics, i.e., the Westminster-style parliamentary system.

On the flip side, this image of politics reflects the recognition that although postwar Japan had adopted a parliamentary system, and political parties came to play a central role, the competition between the LDP—which strongly retained its premodern makeup—and the Socialist Party and other opposition parties—which had abandoned their pursuit of power—lacked tension and responsiveness. This was the starting point of electoral reform. This vision advocated nothing less than the modernization of not only the LDP but of Japanese politics as a whole. Perhaps from late 1989 to early 1990, “politics as usual” came to refer to the pre-modern LDP organization with its factions and policy tribes, as well as the way that competition between parties lacked a sense of tension due to the LDP’s long-time dominance. And it was thought that the lack of response capacity was the reason why the status quo “cannot go on”.

3 The Localization of the Parallel Single-Member District-Proportional Representation Electoral System

3.1 Concrete Choices

Thus, a principle for reform was chosen. However, because electoral reform is a matter directly connected to the vital interests of incumbent politicians, and because there are multiple concrete options, it is remarkably difficult to form a majority to choose a particular alternative. For this reason, there are countries that have decided that the recommendations of experts on electoral reform should be accepted as is. There also exists a view that it is desirable for reform to be supra-partisan so as to avoid criticism that the system is being changed based on partisan interests. Since the electoral system is a set of rules that forms the basis of the electorate’s mandate to the ruling elite and is, in practice, a part of the constitutional structure, it is desirable for reforms to have the understanding and support not only of Diet members, but also of voters. In fact, attempts to introduce a single-member district system by the Hatoyama cabinet in the 1950s and the Tanaka cabinet in the 1970s were both hampered by strong resistance from opposition parties and charges of “partisan interests” from the media. In the case of the electoral reforms that began at the end of the 1980s, the recognition was widely shared that the ruling LDP should not go it alone.

At the time, there were three major options that were considered as feasible alternatives. Their details vary widely, but I will give a brief outline of each system.Footnote 13 Although some of this overlaps with what has already been written, I would like to summarize it without too much repetition.

The first option was the single-member district (SMD) system. As mentioned above, under this system, all seats in a legislature (in this case the House of Representatives) are allotted to electoral districts that each have one seat; voters cast one ballot for a candidate in the electoral district in which they live, and the candidate with the most votes wins the seat. In this system, since only one candidate is elected from each constituency, the field of candidates is narrowed, and the threshold for winning is higher, advantaging large parties with strong party organizations and broad bases of support. Other features of this system are that ballots for candidates who finish second place or lower become wasted votes, in that they do not influence the final seat allocation, and that the distribution of power among parties can fluctuate greatly.

The second option combined SMDs with proportional representation (PR), with completely separate votes for each. In other words, in such a system, the SMD and PR tiers are parallel: the seats are divided between single-member districts and proportional representation, and voters cast one ballot for a candidate in their local SMD, and another for a party in the PR. The counting of votes and the determination of winners are also conducted separately for each tier. The defect of SMDs—that they advantage large parties—is mitigated to some extent by giving small parties the opportunity to win seats through PR, which also reduces the fluctuation in the distribution of power among parties. Another advantage is that since voters are given two votes, it is easier for them to grasp what their own vote means. However, the problem of wasted votes in SMDs is not resolved. The parallel electoral system of SMDs and PR was introduced in Italy and South Korea, which reformed their electoral systems in the late 1980s and early 1990s. In comparative political science, the arrangement is commonly called a mixed-member majoritarian (MMM) system.

The third option was also to combine an SMD-based system and a proportional representation system, but in a different way, with overall seat distribution among parties determined via proportional representation, and votes from the SMDs used only to decide who fills those seats. This system is called a combined constituency-proportional representation system, or a mixed-member proportional (MMP) system, and was adopted by Germany (West Germany) and other countries during the postwar period. In an MMP system, as with an MMM system, there is an SMD ballot and a PR ballot, cast separately for candidates in the constituency and parties in proportional representation. The votes are counted separately. The systems differ from here, however. In an MMP system, the proportional representation results are used to allocate all of the seats in the legislature among the competing parties. Individual winners are then determined using the votes cast in the SMD elections, with candidates who took the most votes in their constituencies awarded the seats secured by their parties. If a party does not have enough winning SMD candidates to fill all of the seats it has won through proportional representation, it selects its additional winners through a list prepared by the party. Conversely, in the event that the number of candidates who are placed first in SMDs exceeds the number of seats won by proportional representation, an “excess seats” rule may be applied, wherein the total number of seats in the legislature is increased to ensure overall proportionality, as determined by the PR vote shares. In terms of the balance of power among parties, the election results will be the same as if seats were awarded entirely by proportional representation; only the identity of individual lawmakers is determined by direct district voting.

If we look at the distribution of seats by party, the proportionality of these three systems is highest in the MMP system of combined constituency voting and proportional representation, and lowest in the pure SMD system. In the context of Japan’s electoral reform, this meant that small parties tended to favor a combined system and large parties the SMD option. In the end, the Electoral System Advisory Council recommended the second option described above: the mixed-member majoritarian (MMM) system of parallel SMDs and PR, which fell somewhere between the two extremes.

3.2 Forming a Majority

The Electoral System Advisory Council is a public advisory body established by law, and its report carried more weight than the conclusions of private advisory bodies or various study groups. However, the council’s report on the electoral system, which was directly connected to the vital interests of politicians and political parties, was not easy to implement in a short period of time. Legally, the Act for the Establishment of the Electoral System Advisory Council does not stipulate any obligation to follow its reports, and it is possible for the prime minister and political parties to simply ignore them. In fact, there were not many precedents of the Advisory Council’s reports being respected and their reforms being implemented.

Thus, the process of majority formation after the report’s advocation of the parallel SMD-PR system began. The concept of the localization of ideas, introduced in Chap. 1, becomes important here. Even if a foundational idea or principle exists, in order for a reform proposal in a given area to gain the support of a majority, it must become a concrete proposal that is acceptable to actors in that area. This is localization. The main actors in electoral reform are politicians from ruling and opposition parties, the mass media, and organizations that have promoted the reforms. In the following, I will quickly describe the process of majority formation, while focusing on the nature of localization.Footnote 14

The prime minister who received the first report of the Electoral System Advisory Council in April 1990 was Kaifu Toshiki. Kaifu expressed his desire to reform the electoral system based on the report, and LDP Secretary-General Ozawa Ichirō adopted the same tack. Ozawa’s proposed schedule had the LDP reaching a definite plan internally by November 1990, and passing a reform bill in the ordinary session of the Diet in 1991. However, the LDP’s joint conference for political reform (a joint meeting of the party’s Political Reform Headquarters and the Electoral System Study Group), which was responsible for the electoral reform bill, objected to both the substance and the schedule of reform. In the summer of 1990, the Gulf War also broke out as Iraq’s Saddam Hussein invaded Kuwait, and the movement to reform the electoral system became increasingly uncertain.

Eventually, on December 25, 1990, the LDP compiled a basic outline for political reform that included, as a pillar, the introduction of a parallel SMD-PR system. Although this was considered to be consistent with both the report of the Electoral System Advisory Council and the previous year’s “Outline for Political Reform,” opposition within the party showed no sign of abating. Support for political reform was emerging little by little, but there did not appear to be momentum in converting intra-party opponents.Footnote 15 In the 1991 Tokyo gubernatorial election, former NHK newscaster Isomura Hisanori, nominated collectively by the LDP, Komeito, and the Democratic Socialist Party, lost to incumbent Suzuki Shun-ichi. Ozawa resigned as secretary-general to take responsibility for the loss, another major blow. At the end of June that same year, the LDP finally reached a party decision and submitted a political reform bill to the Diet in July, but it was forced to abandon efforts to pass it at the end of September. Kaifu tried to pass the legislation by hinting at his determination to dissolve the lower house, but instead, criticism from within the LDP mounted and his cabinet was forced to resign.

The same developments continued under Miyazawa Kiichi, who succeed Kaifu as prime minister. Miyazawa, who did not himself think that electoral reform would be a plus for Japanese politics, nevertheless advocated the political reform legislation based on the intentions of Ozawa and the Takeshita faction (to which Ozawa belonged). And so the opposition within the LDP gathered strength. At the same time, however, no matter how opposed they might have been, lawmakers could not brazenly insist on maintaining the status quo. Thus, by the end of 1992, party policy coalesced around the introduction of a pure SMD system in the House of Representatives that would not be combined with a PR tier. The policy had no chance of securing the approval of opposition parties, however, and since the LDP had lost its majority in the House of Councillors in 1989, it effectively amounted to maintaining the status quo. Kajiyama Seiroku, the LDP secretary-general, pushed this tactic, but Takeshita Noboru, a former prime minister, was seen as being behind it. Ozawa and Hata, who, within the faction, were closer to Kanemaru Shin than to Takeshita, left the Takeshita faction at the end of 1992 and formed the Hata faction.

What moves did the opposition parties make during this period? Among the leading opposition parties, including the Socialist Party, Komeito, and the Democratic Socialist Party, but not the Japanese Communist Party, there existed a belief that it was necessary to reform the electoral system by abolishing the MMD-SNTV system. However, there was also strong support for the new electoral system to be either a combined SMD-PR (MMP) system or a similar additional member system (AMS), which would be more proportional than the parallel SMD-PR (MMM) system.Footnote 16 Some members of the Socialist Party also took the position of supporting the continuation of the MMD-SNTV system. The party could take roughly one seat (out of an average of four) in many electoral districts under the status quo system, and, since the party had long abandoned hope of forming a single-party government, there was little danger of the fierce intra-party competition that would result if it fielded two or more candidates in a district. The MMD-SNTV system was a convenient institution from the perspective of the Socialist Party.

Nevertheless, under the influence of the Japanese Trade Union Confederation (Rengō), which was promoting electoral reform, and the mass media, the Socialist Party and Komeito first jointly proposed a mixed-member proportional system, and then, when the Council for the Promotion of Political Reform (the Minkan Seiji Rinchō) proposed an additional member system in April 1993, they concurred. The opposition parties had strongly criticized the LDP’s money politics and corruption, but when the focus of the debate was linked to the responsiveness of Japanese politics, they had no choice but to support an electoral system that incorporated elements of a small constituency system, at least partially. Additionally, since organized labor, including the Socialist Party-affiliated Sōhyō and the Democratic Socialist Party-affiliated Dōmei, had united to form Rengō in 1989, even leftist parties had come to think it desirable for the opposition to unite and try to take power. As long as the debate developed on this kind of playing field, it is difficult to imagine that reform would have adopted a highly proportional electoral system such as the mixed-member proportional or additional member systems.

3.3 Kanemaru’s Arrest as the Decisive Step

In March 1993, Kanemaru Shin, then serving as the LDP’s vice president, who as a senior leader of the Takeshita faction was the mentor of Ozawa Ichirō and Hata Tsutomu, was arrested by the Special Investigations Department of the Tokyo Public Prosecutors Office. The charges were tax evasion by concealing income through the purchase of discount bonds issued by the Nippon Credit Bank. The funds for the purchase of the bonds were believed to have included a secret donation from Tokyo Sagawa Kyūbin, a logistics company. Although this secret donation had come to light the previous year, the Public Prosecutors Office had imposed only a slap on the wrist, for which it had been criticized by the public. This time, it accused Kanemaru of income tax evasion of ¥400 million [approximately $3 million in current USD]. It was also clear to all that Kanemaru had violated the Political Funds Control Act in amassing a personal fortune.

Kanemaru's arrest intensified voter anger against political corruption. As political scientist Taniguchi Masaki said, “Kanemaru’s arrest was a turning point that rapidly increased pressure from the public for carrying out political reform…Political reform began to tighten around party politics like an iron ring, and the escape routes for each party closed rapidly.”Footnote 17 At the same time, for Ozawa and other politicians close to Kanemaru, the event was likely to give the impression that they were equally corrupt.

There is reason to argue that this case was the result of inadequate political finance regulations and that it was strange that it should be linked to electoral reform. However, it is not uncommon in the process of an idea becoming predominant for there to be a great leap, or a phenomenon whereby other options are discarded without sufficient consideration, and by which a majority comes to accept an idea as “incontrovertible justice.” For example, Article 9 of the Constitution of Japan was originally conceived as the embodiment of the prevention of aggression, based on the calamity of World War II and the 1928 Kellogg-Briand Pact. The Kellogg-Briand Pact does not reject military force, and it originally included a wide variety of measures to prevent a country from invading others. However, in the case of the pacifism of Article 9 of the Constitution, the interpretation of Article 9 as absolute pacifism that renounces all military power appeared soon after the war, and it became widely accepted by the public because it gained significance as a symbol of a reborn Japan.Footnote 18

A document called “The Main Points of Political Reform” summarizing the key issues of contention presented to the executive council of the House of Representatives’ Special Investigatory Committee on Political Reform on May 26, 1993, included the following aims: “the realization of politics that can make responsible decisions in the environment of a new era” and “the establishment of a political system with tension and a high possibility of a change in government.” This was a concrete manifestation of the fact that political reform and the introduction of a new electoral system had become equivalent, making electoral reform “incontrovertibly just” and narrowing the “escape route” for other options.Footnote 19 The same applied to the LDP, which called for an SMD system but had no intention of achieving it. If the low proportionality of an electoral system can be used as a proxy for responsiveness, it certainly makes sense to unify around an SMD system. However, since this was a proposal aimed at preventing the realization of electoral reform, there was no choice but to view it as an “escape route.”

What happened next is well known. Miyazawa, facing strong countervailing winds, attempted to make electoral reform a higher priority, but this just gave the impression of adding a new coat of varnish to the effort. Eventually, Miyazawa was forced by Kajiyama, the LDP secretary-general, to abandon the goal of passing the bill in the Diet session scheduled to end on June 20, 1993. In response, Ozawa, Hata, and other supporters of electoral reform in the LDP voted in favor of a no-confidence motion against the Miyazawa cabinet on June 18 and left the party, forming the Japan Renewal Party on June 23. For Ozawa and his colleagues, gambling on the realization of electoral reform was probably the only way for them to continue to carry the mantle of reform after Kanemaru’s arrest. Although the fractured LDP still remained the largest party after the subsequent dissolution of the Diet and general election, it fell far short of a simple majority and, for the first time since its formation in 1955, it lost its position as the ruling party.

Under the non-LDP coalition government of Hosokawa Morihiro that formed in August 1993, the introduction of a parallel system of SMDs and PR was, at last, effectively decided at a meeting between Hosokawa and LDP leader Kōno Yōhei on January 28, 1994. Despite some complications, including the rebellion in the upper house of some members of the Socialist Party, which had become part of the ruling coalition, the reform bill was passed into law the following day on January 29. The view also exists that the pursuit of other options was still possible under the Hosokawa government. For example, political scientist Okazaki Seiki points out that during deliberations in the Diet, Prime Minister Hosokawa gradually warmed to the position that reforms would facilitate changes of government (government choice theory), which led to a compromise with the LDP in the form of increasing the share of single-member districts.Footnote 20 It is possible that Hosokawa personally underwent such a change in beliefs. However, even had Hosokawa hypothetically insisted upon a position other than the government choice theory, he would have been in the minority within the ruling coalition, and he would not have been able to sustain that position given his emphasis on cooperation with Ozawa and other former LDP members. Of course, the selection of a general framework for reform was already decided at the point that the Hosokawa government was formed. Although the formation of a final agreement among the major parties had to wait until after the establishment of the Hosokawa cabinet, the actual choice had been made by the end of the Miyazawa cabinet.

3.4 Did Voters Agree?

How did voters react to these developments? As mentioned in Chap. 1, one assessment of electoral reform is that it was a reflection of the interests of certain politicians like Ozawa Ichirō that diverged from the will of the electorate. If voters’ intentions, as reflected in public opinion polls, do not mesh with the process of institutional reform as described above, then this point will certainly be valid. Let us examine this point using Asahi Shimbun’s article database “Kikuzō II Visual.”

Expectations for political reform were virtually never discussed until the Recruit scandal came to the attention of the public. A search of the database using the keyword “political reform” showed that articles in 1987 and the first half of 1988 largely dealt with foreign countries, such as Perestroika in the Soviet Union. During this period, it was not even included as an option among important policy issues presented to respondents in public opinion surveys. Even after the Recruit scandal came to light, there were still only a few articles in the second half of 1988. It was only in the November 23 morning edition that an article finally appeared with the headline, “Advisory Body Established to Review Electoral System and Political Funds—Prime Minister Intends to Promote Ethics.” The advisory body was finally established as the Eighth Electoral System Advisory Council. In the morning edition of November 26, an article appeared under the headline, “Prime Minister Takeshita demands investigation of ‘political reform’ in meeting with Gōtoda.” The article was connected to the LDP’s political reform committee. The Asahi Shimbun, in its editorial on November 27, fundamentally supported these moves, arguing, “the first round of political reform should be implemented by the end of the year.”

The public’s interest also began to increase during this period, perhaps due to these media reports. The results of an opinion poll reported in December 10, 1988, showed a significant drop in the Takeshita cabinet’s approval rating, with the primary causes being the introduction of the consumption tax along with concern about “political ethics” aroused by the Recruit scandal. However, even if there was a problem with politicians’ ethics and it was important that this was established as a factual matter, the fact that there were no ideas on hand for concrete proposals to remedy it was a bigger problem. Whether due to Takeshita (the prime minister) or to Gotōda (former chief cabinet secretary), political ethics and electoral reform were already linked issues among senior LDP politicians. However, it is doubtful whether voters were aware of this connection at the time. The electorate’s interest was probably closer to an abstract recognition of a lack in government capacity for self-purification.

However, electoral reform and the promotion of political ethics gradually came to stand shoulder to shoulder as linked issues. According to an Asahi Shimbun poll conducted in mid-March 1989, “the promotion of political ethics” was selected by 29% of respondents and “review of the electoral system” by 24% of respondents as the most important issues for political reform. This trend temporarily weakened after Takeshita’s resignation from the premiership, the reversal of LDP’s defeat in the upper house elections of July 1989, and the breakthrough of the Socialist Party in the February 1990 general election. This was because voters saw that political reform had not advanced under the administrations of Uno Sōsuke and Kaifu Toshiki, who followed Takeshita, and they began to demand “greater political morality” and stricter regulation of political finances within the confines of the status quo electoral system.

The situation changed further when Miyazawa Kiichi assumed the prime minister’s post in November 1991, and from the start of the ordinary Diet session in March 1992. In a public opinion poll conducted in early March, “make fundamental reforms, including to the electoral system” was the top choice for 42% of respondents as a concrete means for advancing political reform. Thereafter, it was no longer uncommon to find numbers close to the “promotion of political ethics.” This trend strengthened with the revelation of the Tokyo Sagawa Kyūbin scandal in 1992, which brought to light then-prime minister Takeshita’s ties to organized crime, and the arrest of Kanemaru in connection with the same malfeasance in 1993.

Indeed, in an opinion poll near the end of the Miyazawa administration, in late April 1993, “strengthening penalties to prevent political corruption” topped the list of most-desired political reforms at 49%, while “reform of the electoral system” was selected by only 16% of respondents. However, this seems to have been an expression of voters’ awareness that institutional reform under the Miyazawa administration was unlikely, and their prioritization of what could be done immediately. According to a poll conducted at the end of May, 57% expressed the opinion that electoral reform should be enacted during that Diet session. Again, at that time, the view that it was necessary to abolish the MMD-SNTV system in order to prevent political corruption and improve political ethics was shared widely and attracted much attention.

The change in public opinion following the inauguration of the Hosokawa government corroborates this. In a poll conducted in early September 1993, although the most important reform cited by respondents was “transparency in the flow of political finances,” support for institutional change to a parallel SMD-PR system reached 53%. Furthermore, immediately after the Four Acts on Political Reform (Seiji Kaikaku Yon-hō) passed, after some twists and turns, in late January 1994, a total of 66% of respondents said that they “greatly approve” or “somewhat approve” of reform as a whole, and 39% of voters responded that the new electoral system was “good,” far more than the 23% who said it was “not good.” There was strong interest in political funds, and this matter was not ignored in the four reform laws, which also established the party subsidy system (state subventions for political party activities) and strengthened political finance regulations. Collectively, it would not be reasonable to view institutional reforms as being bereft of voter support or a reflection of the self-interest of a few politicians. However, it is difficult to determine, as far as poll results are concerned, whether voters accepted electoral reform as a solution to improving the responsiveness of Japanese politics, given some differences and ambiguity in the phrasing of questions.

4 What Happened?

4.1 Expected Consequences

The parallel SMD-PR system as introduced in the 1996 House of Representatives election had a total of 500 seats, of which 300 were elected from single-member districts and 200 from the proportional representation tier, which was divided into 11 regional blocs. The fixed number of seats has been reduced several times since, changing to 480 (300 SMDs, 180 PR) in 2000, to 475 (295 SMDs, 180 PR) in 2014, and again in 2017 to 465 (289 SMDs, 176 PR). However, the ratio of constituency seats to the total number of seats has fluctuated only between 60% and 62.5%, so the fundamental characteristics of the electoral system have remained constant. That voters have two votes, one in the single-member district and another in the regional proportional representation tier, has remained unchanged.

From the time its introduction was proposed, this system has been considered to be a SMD-centered electoral system. For example, the first report of the Eighth Electoral System Advisory Council, mentioned earlier, states the following (a slightly lengthy quotation):

In order for politics to adapt to a time of changes in our country’s domestic and international conditions, it is necessary to reflect the will of the people accurately, and, at the same time, to aggregate the will of the people to clarify decision-making in politics and the attribution of responsibility. In addition, for a dynamic and healthy parliamentary democracy, it is necessary to maintain tension in politics through a change of government. In satisfying these demands, when comparing a single-member district system and a proportional representation system, the single-member district system better conforms to these demands. However, since both the single-member district system and the proportional representation system have their own problems, as mentioned above, it is considered appropriate to use a combination of the two systems.

This Council believes that it is appropriate to adopt a system of parallel single-seat districts and proportional representation, taking into consideration that the aggregation of the popular will, the clarification of political decision-making and the attribution of responsibility, and the possibility of a change of government should be emphasized; the importance of also taking into consideration that minority views should be reflected in national government; and the desirability of a system that is easy to understand.

The understanding that can clearly be seen here is that while an SMD system was deemed desirable in principle, its weakness was the insufficient reflection of minority views, and so a PR system was used in parallel to supplement it. As is clear from the previous examination of the process of majority formation, since the LDP, which is the leading party, had no possibility of winning support for reforms from small parties if a pure SMD system was used, the PR tier was used as a “safety margin” so to speak. This was localization. Furthermore, it was assumed that competition for the government between two large parties would emerge, and that overall victory would be determined primarily by developments in the single-member districts, with the party that emerged victorious establishing a single-party government. Voters were also expected to cast their ballots on the decisive and effective sole criterion of whether the candidate belonged to the party they wanted to entrust with power.

If inter-party competition becomes a struggle for power between two major parties, the organizational structure within the parties should also change to reflect this. The recognition that the MMD-SNTV system gave rise to factions and policy tribes, and that competition between candidates from the same party led to particularistic politics, has already been seen in the “Outline for Political Reform.” The understanding that modern European political parties, as in the United Kingdom and Germany, were models for party organization that should be emulated was common—not only among political scientists, but also among politicians who were the affected actors. Even under the 1955 system, the LDP repeatedly aimed to modernize by dissolving its factions.

Therefore, in May 1993, the Council for the Promotion of Political Reform issued a “Proposal for a New Form of Political Party,” and stated after electoral reform that, “It is necessary to clarify the party’s decision-making structure and where responsibility lies.” It further added, “We must clarify what is to be discussed, by what procedures and at what stage consensus will be sought, as well as to make greater efforts than before to achieve both freedom of political activity and the discipline of political parties.” It was hypothesized that centralization in party executives would proceed further, and that party organization would be managed top-down as a matter of party policy.Footnote 21 However, considering that the main focus of attention was on the shape of inter-party competition (two-party or multi-party) and perhaps also the level of international research on electoral systems at the time, it was unclear what causal mechanism would give birth to the proposed party structure and what the incentive structure would be for individual members.

4.2 Development as Expected

The resulting developments in each party were also based on this understanding. Regarding competition among parties, the movements of the second party (the largest non-LDP party) were entirely different from what had come before. The New Frontier Party (NFP), which formed in 1994 around the ruling parties that had supported the Hosokawa cabinet (apart from the Socialist Party and the New Party Sakigake), took 31.1% of the seats after the 1996 general election, recording the second-largest figure for a second party in the House of Representatives in the postwar era, after the Socialist Party in 1958 (35.5%). After the NFP dissolved, the Democratic Party of Japan (DPJ) rose to prominence, finally achieving a record-high seat share for a postwar second party at 36.9% after the general election of 2003. The DPJ went on to become the majority party in 2009, but the LDP returned to power in a coalition government with Komeito in 2012. It seemed that the goal of realizing regular changes in government through inter-party competition among parties had been achieved, albeit briefly.

The electorate’s voting behavior was also less clear in the early years of the new electoral system, but voting based on party labels gradually became mainstream. Voters today decide whom to vote for based not on their evaluation of individual candidates in their district but on their assessment of the party that fielded the candidate. The fact that terms like the 2005 general election’s “Koizumi Children,” 2009’s “Ozawa Girls,” and also 2012’s newly elected “Abe Children” began to appear in the mass media, indicates the arrival of many newly elected lawmakers who were elected based not on the evaluation of the candidates but of the party. On the flip side, these derisive expressions offer a glimpse of nostalgia for the era of the MMD-SNTV system when the emphasis was on individual candidates, and for the political reporting style—closely following the political situation—that unfolded from it. In addition, as American political scientist Daniel Smith has pointed out, changes in the electorate’s voting behavior weakened the tendency of political parties to field candidates with high name recognition, leading to a decline in the number of hereditary lawmakers.Footnote 22

The major parties have also undergone significant changes in their organizational structure. This is the predicted centralization of power. The LDP during the age of the MMD-SNTV system had the following organizational characteristics: elections depended on the personal support groups of politicians (kōenkai), policy was made from the bottom up by the Policy Affairs Research Council and its subcommittees, and career paths, including promotion in the party and entry into the cabinet, were based on faction size and number of re-elections. To a considerable extent, it was possible to be elected and to pursue legislative activities without the help of the party’s executive (party president or secretary-general), and it was not unusual for party executives to be unable to control members of their own party.

However, this changed completely with the reform of the electoral system. Winning elections required the party’s official endorsement, and the introduction of the party subsidy system and other changes to the flow of political funds meant that legislators came to rely heavily on the party executive. The party executive now had the power of “life and death” over individual lawmakers, and could impose sanctions on members who disobeyed the wishes of the executive, whether by giving them the cold shoulder or even in some cases removing them from the list of endorsed candidates. The clearest example of this is the 2005 general election during the Koizumi administration, when LDP candidates who opposed the Koizumi-backed postal privatization bill had their formal endorsements revoked.

This centralization of power increased the importance of party leaders in inter-party competition, and also created a structure in which the evaluation of party policies became directly linked to the party leader. The introduction of party leadership debates in 1999 and the introduction of party election manifestoes, which outlined the party’s basic policy platforms, in 2003, reflect this change. At the same time, this has also led to situations of crisis-like disorder for major parties when party leaders advocate policies that are half-baked or have a low possibility of being implemented, or when the executive cannot control dissent within the party. The LDP from 2006 to 2009, and the DPJ from 2009, are both examples of how disorder can lead to a loss of power.

In terms of policy, it was also thought that as the proportionality of the electoral system decreased with the replacement of the MMD-SNTV system with the parallel SMD-PR system, a shift from micro-oriented, particularistic policy to macro-oriented, universalistic policy would occur. Many scholars and commentators, especially those overseas, have observed that such a change is actually taking place, coupled with the centralization of power within the ruling party.

For example, Margarita Estevez-Abe, a comparative political scientist active in the United States, writes that since the late 1990s, there has been a major shift towards universalism in Japan’s social security policies, which she attributes to electoral reform. A typical example is the introduction of a long-term care insurance system, in which all citizens pay premiums and receive services based on the same calculation standards. The previous insurance and pension schemes were highly particularistic, with different systems for employees depending on where they worked and distinct benefit levels depending on family composition. In the case of the Long-Term Care Insurance System, there are only slight differences in the calculation of insurance premiums and long-term care certification (for persons with disabilities) depending on the municipality where one resides, and differences in employment and family structure are not taken into consideration. This is a fundamentally different concept from the traditional Japanese social security system.Footnote 23

Frances Rosenbluth and Michael Thies, also American scholars of comparative politics, argue that the “pipeline of special interests” that provided generous protection for farmers and small businesses was severed by electoral reform, and that policy changes were underway to unify pensions that had been divided into the National Pension Plan, the Employees’ Pension, and the Mutual Aid Pension, based on job category.Footnote 24 The child allowance sought by the DPJ administration, which did not feature income-based limits, and the Trans-Pacific Partnership (TPP) concluded under the second Abe administration were also policies that have this universalistic orientation. Amy Catalinac, another political scientist conducting research in the United States, points out that the number of politicians who are interested in policy areas like foreign and security affairs, which used to be neglected because they could not be used to gather votes, have increased due to electoral reform.Footnote 25 As the effectiveness of evaluations of individual candidates for attracting votes has declined, there is now more room for politicians to deal with policies that are not directly connected to their reputations with voters in their constituencies.

4.3 Some Important Unanticipated Consequences

However, while the introduction of the single-member district system produced the anticipated effects, we should say that the impact of the parallel PR tier has been underestimated. At the time that electoral reform was pursued, there were still few examples of parallel systems in the world, and research on their effects had not been fully developed. Therefore, it was thought that a parallel electoral system would display the various characteristics of single-member districts and proportional representation separately. In reality, however, this was not the case. The existence of proportional representation had a complex impact on the behavior of parties and voters, going beyond simply “reflecting minority opinions in national politics.”

One effect is that the parallel system has not led to the decline of smaller parties but has instead encouraged the splintering of large parties. A large party that is seeking a simple majority in the House of Representatives naturally has a variety of views coexisting within it, and it is easy for splinter groups to emerge. The existence of proportional representation means that members of the intra-party minority and their supporters have room to defect from the party if they are unable to reach a compromise. The Democratic Party of Japan (DPJ; Minshutō), after taking power in 2009, split over whether to support or oppose a consumption tax increase. Its successor, the Democratic Party (DP, Minshintō), grew increasingly fractious over its stance towards the 2015 security laws and ultimately split over the issue of whether to merge with the “Party of Hope” (Kibō-no-tō) founded by Koike Yuriko. Each provides an example of this tendency. The Constitutional Democratic Party of Japan (CDP), which became the leading opposition party after breaking with the DP, has not tried to unify non-LDP forces into a single party, which would be the most effective strategy under a single-member district system. This is likely due to the recognition that, on the one hand, it would be more highly regarded by voters as a party that held on to its principles, and, on the other hand, its recognition that maintaining a major party would be difficult, based on the experiences of the DPJ and the DP.

Another impact is that small parties aiming to win seats through proportional representation continue to field candidates in single-member districts in order to engage in so-called “vote digging.” While small party SMD candidates may have no chance of winning, their campaign activities may nevertheless mobilize and persuade voters to choose their party in the PR ballot. This tactic changes the structure of competition among major parties. A typical example is the behavior of the Japanese Communist Party (JCP). Until the general election of 2017, the JCP had fielded candidates in many constituencies even though it had won a grand total of only four seats in SMDs under the current electoral system. This has clearly had a negative impact on major parties that are positioned between the LDP (center-right) and the JCP (far left-wing). These parties lose many votes on the ideological left that they would have gained if the JCP had not fielded candidates, with the result that these parties’ candidates lose out to LDP candidates. Although the JCP has begun to withdraw candidates in lower house and upper house SMDs, there is no doubt that, at least until recent years, its candidate nominations have benefited the LDP. It is possible that the same thing could happen to the LDP if a party to the LDP’s right were to emerge, but this phenomenon has been rare. The fielding of Japan Innovation Party (Nippon Ishin-no-kai) candidates outside of Osaka could potentially have the same effect. This kind of effect is called a “contamination effect” or “linkage effect” in political science.Footnote 26

Even in terms of policy shifts, it is difficult to say that a macro-level orientation and universalism have been thoroughly implemented. Certainly, the reorganization of costly and overly complex institutions, particularly in social security policies, has proceeded, and structural reforms have been carried out in areas such as agriculture and distribution, in which industrial protection was conspicuous in the past. This was the realization of policies that experts and some bureaucrats had long considered rational. However, particularistic policies that apply different rules to different industries and business categories are still not uncommon, such as the consumption tax rebate system introduced when the consumption tax rate was raised in October 2019. Postal privatization was intended to reorganize business divisions that did not conform with international standards, such as the postal savings system, and those that were unprofitable, such as the post offices and mail delivery networks. This is one example where macro-rationality was prioritized. However, the frameworks of both the postal savings and post office networks have not changed greatly from the pre-reform period, due to political decisions made in later years. Whereas in the past, policy change was often blocked by industry groups in combination with LDP politicians and the relevant ministries and agencies, today it tends to be influenced by relations between parties within the coalition government.Footnote 27 This is largely due to the fact that, in the end, electoral reform has not resulted in competition between the two major parties and regular changes in government, for which it had aimed.

4.4 Lack of Interest in Linkages

The reform of the electoral system for the House of Representatives was substantial, and while the effects of it being a mixed system, such as the contamination effect, cannot be ignored, it brought about striking changes to both party competition and party organization. However, in relation to the multilevel mixture discussed in Chap. 1, it is also necessary to look at the interconnectedness of the electoral system. Specifically, it is not only the House of Representatives that defines the nature of Japanese politics; the House of Councillors and local politics must also be taken into consideration. Under the Constitution of Japan, the bicameral relationship between the House of Representatives and House of Councillors is almost equal. The formation of a government is possible only by the will of a majority of the House of Representatives, due to its supremacy in nominating the prime minister. However, the House of Councillors is equal to the House of Representatives with regard to the passage of bills other than budget bills, and a censure motion against the prime minister and cabinet ministers has a similar effect to a vote of no confidence in the House of Representatives. This means that for the continuance of government, the will of a majority of the House of Councillors must also be taken into account. Local politics also plays a major role, with local governments having an integrated relationship with the central government and accounting for 70% of total government expenditures. In addition, the importance of local party organizations and lawmaker networks influences the management of political parties at the national level. The way in which politicians are elected in the House of Councillors and local politics, i.e., the electoral systems for the House of Councillors and local governments, should have received attention equal to that of the electoral system of the House of Representatives.Footnote 28

With regard to the electoral system of the House of Councillors, it was recognized that the current situation was problematic. On November 1, 1988, the Study Group on the Institution of the House of Councillors submitted its “Opinion on the State of the House of Councillors and Its Reform” in response to an inquiry from the Speaker of the House.Footnote 29 Additionally, on July 31, 1990, the Eighth Electoral System Advisory Council issued its second report, “Report on the Reform of the Election System for Members of the House of Councillors and Public Subsidies for Political Parties.” However, both reports, in stressing the importance of maintaining the House of Councillors’ “distinctiveness” from the House of Representatives, which was the arena for inter-party competition, proposed an electoral system that would allow it to have a different existence or rationale from the House of Representatives.Footnote 30 Specifically, it proposed to replace the current system, which combined prefecture-based electoral districts with closed-list proportional representation, with a pure prefecture-based districting system (each of the 47 prefectures would be an electoral district), or to convert the PR tier into an open-list system, wherein voters would cast their ballots for candidates instead of parties. In either case, if realized, voting based on the evaluation of candidates would become mainstream, and there is a high likelihood that results would differ from the House of Representatives, where voting is expected to focus mainly on the evaluation of parties.

At the time, there were concerns that the House of Councillors had become a mere “carbon copy” of the House of Representatives, repeating the same deliberations with the same substance with members of similar quality. As such, it was perhaps unavoidable that uniqueness and differentiation would become focal points. Moreover, these proposals are not incomprehensible or illogical if differentiation from the House of Representatives is seen as the main goal. The change from a closed list to an open list PR system would have had only a limited effect, due to the small number of seats available in the PR tier. This system was actually introduced in 2001, and although there appears to have been no major impact since then, it was probably difficult to foresee this before its implementation.

However, from this book’s stance that we should take into consideration the effects of the multilevel mixture, it is hard to shake the impression that all of these arguments’ starting points were misguided. If the House of Councillors was to be reformed, it was necessary to make a choice. One option was to differentiate the two houses by making their electoral systems different while changing their power relationship from one of near-equality. The other was to maintain the equality of power, while making their electoral systems more homogenous and less prone to differences in partisan composition and opinion between the two houses. There is no evidence that serious consideration was given to the problems that could arise when two equal houses form different majorities under different electoral systems. The only reform proposal that seemed to consider this point was the opinion paper of the “Advisory Panel to consider the future of the House of Councillors” in 2000, but perhaps because the proposal included constitutional revision, it was never taken up head-on.Footnote 31

From the report of the Study Group on the Institution of the House of Councillors and other documents, it seemed that the intention was to make the upper house a “chamber of common sense” or “house of wisdom,” in which majority formation would be made more fluid by, for example, relaxing restrictions on compelling members to follow their party. However, the means to achieve this goal were left to each party, and no concrete path was shown. It is true that there was still little research at the time on the effects of bicameralism on the policymaking process, but we cannot deny the possibility that members of the House of Representatives decided they should not intervene in the reform of the House of Councillors.

And in fact, due to time constraints and lack of interest on the part of the actors involved, these recommendations were only partially realized. While the framework of the House of Councillors’ electoral system and the equality of the two houses was left completely untouched, the effects of the electoral reform of the House of Representatives began to show. This has had a major impact on Japanese politics from the late 1990s onward, including the organization of coalition governments and the appearance of the so-called “twisted Diets,” wherein different parties comprise majorities in each chamber.

Local election systems have not gone unnoticed in political reforms since the 1990s. For example, the “Basic Policy for Political Reform,” adopted at the inaugural meeting of the Council for the Promotion of Political Reform on April 20, 1992, stated: “We will also give full consideration to the shape of the local election system and the local assembly system, which have not been sufficiently studied so far. In particular, in the event that new electoral systems are introduced in both the House of Representatives and the House of Councillors, we will consider how the electoral systems for prefectures and municipalities should be organized, as well as the clarification of the roles and responsibilities of Diet members and local representatives.”Footnote 32 However, this item was placed at the end of a total of six “basic policies,” and subsequent reforms related to local politics were overwhelmingly focused on reforming the relationship between the central government and local governments, i.e., decentralization. There may have been the recognition that local political systems should be decided by the localities themselves, notwithstanding that the Constitution and the Local Autonomy Act stipulate otherwise.

In theory, if local decentralization is to proceed and the autonomy of local government is to be increased, the question of what kind of governance structure is consistent with that decentralization should be considered in parallel. However, this did not happen, and power and financial resources were delegated to local governments before the rationalization of government operations or inter-party competition and before checks and balances to produce a sense of tension among powerholders could be established. Moreover, despite decentralization reforms, as long as Japan is a unitary, not federal, state, collaboration between the central and local governments and linkages between national and local politics must remain. It is difficult to say that sufficient consideration has been given to this issue either. Today, disagreements between the central and local governments over the location of U.S. military bases in Japan and the restart of nuclear power plants are conspicuous. It is undeniable that the political reform process failed to consider adequately how local government policies should be decided and whether political positions or stances should have any meaning whatsoever.