2020.11.18

2020 (Gyo-Tsu) 78

Minshu Vol. 74, No. 8

Judgment concerning the constitutionality of the provisions on the apportionment of seats for members of the House of Councillors (to be elected from constituencies) in Article 14 and Appended Table 3 of the Public Offices Election Act

Case seeking invalidation of an election

Judgment of the Grand Bench, dismissed

Tokyo High Court, Judgment of December 4, 2019

It cannot be said that the disparity between constituencies in terms of the value of votes under the provisions on the apportionment of seats for members of the House of Councillors (to be elected from constituencies) in Article 14 and Appended Table 3 of the Public Offices Election Act amended by Act No. 75 of 2018 indicated the existence of extreme inequality to such an the extent that it could raise a question of unconstitutionality as of the time of the ordinary election of the members of the House of Councillors held on July 21, 2019, and therefore the aforementioned provisions cannot be considered to have violated Article 14, paragraph (1), etc. of the Constitution.

(There are opinions and dissenting opinions.)

Article 14, paragraph (1), Article 15, paragraph (1), Article 15, paragraph (3), Article 43, paragraph (1), and Article 44 of the Constitution and Article 14 and Appended Table 3 of the Public Offices Election Act



The Constitution of Japan

Article 14, paragraph (1)

(1) All of the people are equal under the law and there shall be no discrimination in political, economic or social relations because of race, creed, sex, social status or family origin.

Article 15, paragraphs (1) and (3)

(1) The people have the inalienable right to choose their public officials and to dismiss them.

(3) Universal adult suffrage is guaranteed with regard to the election of public officials.

Article 43, paragraph (1)

(1) Both Houses shall consist of elected members, representative of all the people.

Article 44

The qualifications of members of both Houses and their electors shall be fixed by law. However, there shall be no discrimination because of race, creed, sex, social status, family origin, education, property or income.



Public Offices Election Act

Article 14

(1) Constituencies for members of the House of Councillors to be elected by constituency and the number of members of the House of Councillors to be elected in each constituency are specified in Appended Table 3.

(2) Even in the event of the abolition, creation, division or amalgamation of a prefecture or prefectures effected pursuant to the provisions of Article 6-2, paragraph (1) of the Local Autonomy Act, the constituencies for elections of members of the House of Councillors to be elected by constituency and the number of members of the House of Councillors to be elected in each constituency remain as they were before such event.

The final appeal is dismissed.

The costs of the final appeal shall be borne by the appellants of final appeal.

Concerning the reasons for final appeal stated by the appellants who also act as the counsel for final appeal, YAMAGUCHI Kuniaki, et al.

1. This case is a suit to seek invalidation of an election that was filed by the appellants, who are voters in constituencies in Tokyo Metropolis and Kanagawa Prefecture, arguing as follows: as the provisions on the apportionment of seats for members of the House of Councillors (to be elected from constituencies) in Article 14 and Appended Table 3 of the Public Offices Election Act (hereinafter these provisions, including Appended Table 2 prior to amendment by Act No. 2 of 1994, are referred to as the "provisions on apportionment of seats" throughout the period before and after several amendments) are unconstitutional and thus invalid, elections in the aforementioned constituencies for the ordinary election of members of the House of Councillors that was held on July 21, 2019 (hereinafter referred to as the "Election") based on said provisions are also invalid.

2. The outline of facts lawfully determined by the court of prior instance is as follows.

(1) With regard to election of members of the House of Councillors, the Act on the Election of Members of the House of Councillors (Act No. 11 of 1947) divided 250 members into 100 members to be elected from the national constituency and 150 members to be locally elected and provided that members to be elected from the national constituency are to be elected throughout the area including all prefectures. On the other hand, for members to be locally elected, the same Act specified constituencies and the number of seats in each constituency in the appended table and provided that the members are to be elected from prefecture-based constituencies. Regarding the number of seats in each constituency, considerations were given to ensure that an election takes place for half the members throughout the constituencies in accordance with the provisions of the Constitution that election for half the members of House of Councillors shall take place every three years. Under the policy of apportioning an even number of seats (at least two), an even number of seats, ranging from two to eight, was apportioned to each constituency in proportion to the population of the constituency. The provisions on the apportionment of seats of the Public Offices Election Act, which was enacted in 1950, only inherited the aforementioned provisions on the apportionment of seats of the Act on the Election of Members of the House of Councillors, and no change had been made to the aforementioned provisions on the apportionment of seats until the amendment to the Public Offices Election Act by Act No. 47 of 1994 (hereinafter referred to as the "1994 Amendment"), except for the subsequent addition of two seats to the constituency in Okinawa Prefecture. Incidentally, through amendment to the Public Offices Election Act by Act No. 81 of 1982 (hereinafter referred to as the "1982 Amendment"), 252 members of the House of Councillors were divided into 100 members to be elected by proportional representation, who are to be elected in proportion to the number of votes obtained by each political party, etc., and 152 members to be elected from constituencies, who are to be elected for each prefecture-based constituency. However, these members to be elected from constituencies are nothing more than conventional locally-elected members, though the name was changed. After that, through amendment to the Public Offices Election Act by Act No. 118 of 2000 (hereinafter referred to as the "2000 Amendment"), the total number of seats for members of the House of Councillors was changed to 242, specifically, 96 members to be elected by proportional representation and 146 members to be elected from constituencies.

(2) At the time of the enactment of the Act on the Election of Members of the House of Councillors, the maximum disparity in population per member between constituencies (hereinafter the "maximum disparity between constituencies" at the time of each legislation means this maximum disparity in population) was 2.62 to 1 (hereinafter all figures concerning disparity are approximate figures). However, the maximum disparity continued to gradually expand due to population changes, and at the time of the ordinary election of members of the House of Councillors held in 1992 (hereinafter an ordinary election of members of the House of Councillors is merely referred to as an "ordinary election," and this ordinary election is referred to as the "1992 Election"), the maximum disparity in the number of voters per member between constituencies (hereinafter the "maximum disparity between constituency" at the time of each election means this maximum disparity in the number of voters) reached 6.59 to 1. After that, the maximum disparity between constituencies based on population according to the result of the population census implemented in October 1990 was reduced to 4.81 to 1 owing to a measure to increase the number of seats by two each in four constituencies and decrease the number of seats by four in one constituency and by two each in two constituencies in the 1994 Amendment. After that, the maximum disparity between constituencies at the time of ordinary elections held during the period from 1995 to 2007 hovered around 5 to 1 throughout the period before and after a measure to decrease the number of seats by two each in three constituencies in the 2000 Amendment and a measure to increase the number of seats by two each in two constituencies and decrease the number of seats by two each in two constituencies in the amendment to the Public Offices Election Act by Act No. 52 of 2006 (hereinafter referred to as the "2006 Amendment").

However, in 1979 (Gyo-Tsu) No. 65, the judgment of the Grand Bench of the Supreme Court of April 27, 1983, Minshu Vol. 37, No. 3, at 345 (hereinafter referred to as the "1983 Grand Bench Judgment"), the Grand Bench of this Court indicated a basic framework for making a determination concerning the constitutionality of provisions on the apportionment of seats as mentioned in 3.(1) below. After that, regarding the 1992 Election, this Court held that there existed extreme inequality in the value of votes to such an extent that it could raise a question of unconstitutionality (1994 (Gyo-Tsu) No. 59, the judgment of the Grand Bench of the Supreme Court of September 11, 1996, Minshu Vol 50, No. 8, at 2283). However, regarding two ordinary elections held under the provisions on the apportionment of seats after the 1994 Amendment, this Court held that the elections were not considered to have reached the aforementioned state (1997 (Gyo-Tsu) No. 104, the judgment of the Grand Bench of the Supreme Court of September 2, 1998, Minshu Vol. 52, No. 6, at 1373; 1999 (Gyo-Tsu) No. 241, the judgment of the Grand Bench of the Supreme Court of September 6, 2000, Minshu Vol. 54, No. 7, at 1997). After that, regarding both two ordinary elections held under the provisions on the apportionment of seats after the 2000 Amendment and an ordinary election held in 2007 under the provisions on the apportionment of seats after the 2006 Amendment, the Grand Bench of this Court presented its determination that the relevant provisions on the apportionment of seats cannot be considered to have been unconstitutional in conclusion without making any explicit holding concerning whether the elections had reached the aforementioned state (2003 (Gyo-Tsu) No. 24, the judgment of the Grand Bench of the Supreme Court of January 14, 2004, Minshu Vol. 58, No. 1, at 56; 2005 (Gyo-Tsu) No. 247, the judgment of the Grand Bench of the Supreme Court of October 4, 2006, Minshu Vol. 60, No. 8, at 2696; 2008 (Gyo-Tsu) No. 209, the judgment of the Grand Bench of the Supreme Court of September 30, 2009, Minshu Vol. 63, No. 7, at 1520). However, this Court had come to make virtually stricter determinations concerning the state of disparity from the perspective of equality in the value of votes with the maximum disparity between constituencies being normally around 5 to 1. For example, in the aforementioned judgment of the Grand Bench of the Supreme Court of October 4, 2006, this Court made the following indication: the Diet is expected to make a constant effort to correct inequality in the value of votes, taking into consideration the importance of equality in the value of votes. In the aforementioned judgment of the Grand Bench of the Supreme Court of September 30, 2009, this Court also pointed out as follows: the disparity of the time was in the state where extreme inequality is found in the value of votes and in the situation where efforts to reduce the disparity in the value of votes between constituencies are required; therefore, it is necessary to reform the existing mechanism of the election system itself in order to significantly reduce the maximum disparity.

(3) Regarding an ordinary election held on July 11, 2010 in the situation where the maximum disparity between constituencies was 5.00 to 1 (hereinafter referred to as the "2010 Election"), in the conclusion of 2011 (Gyo-Tsu) No. 51, the judgment of the Grand Bench of the Supreme Court of October 17, 2012, Minshu Vol. 66, No. 10, at 3357 (hereinafter referred to as the "2012 Grand Bench Judgment"), this Court ruled as follows: the provisions on the apportionment of seats as of the time of the same election cannot be considered to have been unconstitutional. However, in light of changes in systems and social circumstances over the years, this Court also stated as follows: the requirement for equality in the value of votes should not be taken more lightly just because the election in question is for members of the House of Councillors; circumstances such as the fact that each prefecture can be defined as a political unit can no longer be regarded as a legitimate reason for leaving great disparity in the value of votes unaddressed for tens of years; with a continuous increase in the disparity in population between prefectures, given limitations regarding the possibility of choosing the option of increasing the total number of seats, it has become extremely difficult to answer the call for greater equality in the value of votes, while maintaining the mechanism designed to use a prefecture as a unit of constituency; upon comprehensive consideration of such circumstances as the fact that despite these calls for improvement, since the 2006 Amendment, no legal amendments have been made before the 2010 Election for improving the situation that possesses such great inequality in the value of votes, the disparity between constituencies in terms of the value of votes shown by the maximum disparity at the time of the 2010 Election indicates the existence of extreme inequality to such an extent that it could raise a question of unconstitutionality. Then, this Court also pointed out as follows: for example, reasonable reform should be made to the current system of setting the number of seats for each prefecture-based constituency, and such legislative measures designed to reform the mechanism of the current election system itself should be taken in order to correct said inequality as soon as possible because such level of inequality raises the issue of unconstitutionality.

(4) After the 2012 Grand Bench Judgment was rendered, the Act Partially Amending the Public Offices Election Act (Act No. 94 of 2012; hereinafter referred to as the "2012 Amendment Act") was enacted on November 16, 2012 and came into effect on November 26, 2012. The content of the same Act involved amendments toward an ordinary election to be held in July 2013, namely, increasing two seats each in two constituencies and decreasing two seats each in two constituencies for members to be elected from constituencies. The Supplementary Provisions of the same Act had the provisions to the effect that review is to be made continuously, aiming to definitely reach a conclusion on the fundamental reform of the election system in preparation for an ordinary election scheduled in 2016.

The first ordinary election under the provisions on the apportionment of seats amended by the 2012 Amendment Act (hereinafter referred to as the "2013 Election") was held on July 21, 2013. The maximum disparity between constituencies at the time of the same election was 4.77 to 1.

(5) In September 2013, the Election System Council was established under the Election System Reform Study Panel at the House of Councillors for the purpose of having deliberations about the reform of the election system for the House of Councillors in preparation for an ordinary election to be held in 2016. At the same Council, in April 2014, a chair's plan was presented as a specific plan for amendment including the reform of the mechanism of the election system, and after that, a revised plan of the same plan was also presented. These plans were basically intended to significantly reduce the maximum disparity between constituencies by merging a certain number of constituencies with a small population per member with adjacent constituencies to reduce the number of seats apportioned to the constituencies and increasing the number of seats in a certain number of constituencies with large populations. The same Council had deliberations over the aforementioned plans and suggestions, etc. submitted by the parliamentary groups of the House of Councillors. However, the parliamentary groups could not reach any agreement. Therefore, on December 26, 2014, a report stating suggestions, etc. presented by the parliamentary groups was submitted to the President of the House of Councillors.

(6) Amid such deliberations, regarding the 2013 Election, this Court held as follows in 2014 (Gyo-Tsu) No. 155 and No. 156, the judgment of the Grand Bench of the Supreme Court of November 26, 2014, Minshu Vol. 68, No. 9, at 1363 (hereinafter referred to as the "2014 Grand Bench Judgment") in line with a determination made in the 2012 Grand Bench Judgment: the abovementioned reapportionment of four seats under the 2012 Amendment Act only resulted in reapportioning seats among some constituencies while maintaining the abovementioned mechanism of the prefecture-based election system; in fact, the maximum disparity between constituencies stayed around the level of 5 to 1 throughout the period before and after said amendment; thus, said reapportionment of four seats should inevitably be held to be insufficient to correct extreme inequality in the value of votes, which could raise a question of unconstitutionality in terms of the disparity in the value of votes; consequently, even after the abovementioned reapportionment of seats was conducted under the same Act, the disparity between constituencies in terms of the value of votes indicated the existence of extreme inequality to such an extent that it could raise a question of unconstitutionality. In addition, this Court also pointed out as follows: it is necessary to take constant steps to discuss and build consensus for a specific proposal for amendment, including making a reasonable change to the current system of setting the number of seats for each prefecture-based constituency, and further take legislative measures to reform the current mechanism of the election system itself as soon as possible, so that the abovementioned level of inequality will be corrected.

(7) The Election System Reform Study Panel had deliberations in response to the submission of the report mentioned in (5) above, but could not reach a conclusion agreed to by all parliamentary groups. Therefore, on May 29, 2015, it was decided that each parliamentary group would work to draft a bill. As a result of discussions held at each parliamentary group, the opinions of the parliamentary groups were basically consolidated into the following plans for amendments including introduction of the merger of constituencies with small populations: [i] increasing two seats each in five constituencies, decreasing two seats each in three constituencies, and creating two merged constituencies from four prefectures (decreasing two seats in each merger) and [ii] increasing two seats each in six constituencies and creating 10 merged constituencies from 20 constituencies (decreasing two seats each in six mergers and decreasing no seat in four mergers). On July 23, 2015, bills to partially amend the Public Offices Election Act based on the aforementioned plans were submitted respectively to the Diet. Regarding constituencies and the number of seats for members to be elected from constituencies, the bill pertaining to the plan for amendment mentioned in [i] above included having Tottori Prefecture and Shimane Prefecture, and Tokushima Prefecture and Kochi Prefecture, respectively, merge into single constituencies with two seats apportioned, as well as decreasing two seats each in three constituencies and increasing two seats in five constituencies. Article 7 of the Supplementary Provisions of the bill provided that review is to be made continuously, aiming to definitely reach a conclusion on the fundamental reform of the election system in preparation for an ordinary election scheduled in 2019, in light of a desirable form of the House of Councillors and in consideration of the correction of the disparity in population per member between constituencies, etc.

The Act Partially Amending the Public Offices Election Act pertaining to the plan for amendment mentioned in [i] above (Act No. 60 of 2015; hereinafter referred to as the "2015 Amendment Act") was enacted on July 28, 2015 and came into effect on November 5, 2015. As a result of the amendment to the Public Offices Election Act by the 2015 Amendment Act (hereinafter referred to as the "2015 Amendment"), the maximum disparity between constituencies based on the population according to the result of the population census implemented in October 2010 was 2.97 to 1.

(8) The first ordinary election under the provisions on the apportionment of seats after the 2015 Amendment was held on July 10, 2016 (hereinafter referred to as the "2016 Election"). The maximum disparity between constituencies at the time of the same election was 3.08 to 1.

In 2017 (Gyo-Tsu) No. 47, the judgment of the Grand Bench of the Supreme Court of September 27, 2017, Minshu Vol. 71, No. 7, at 1139 (hereinafter referred to as the "2017 Grand Bench Judgment"), this Court ruled as follows with regard to the 2015 Amendment Act: the 2015 Amendment Act not only reapportioned seats among some constituencies, but also reformed the mechanism of the election system designed to use a prefecture as a unit of constituency, which was the cause for the prolonged existence of great disparity in the value of votes over a long period of time, by merging less-populated constituencies, which was a measure taken for the first time since the establishment of the House of Councillors; by virtue of this amendment, the maximum disparity between constituencies, which had remained around the level of 5 to 1 for dozens of years, was reduced to 2.97 to 1 (3.08 to 1 at the time of the election); therefore, it can be seen that this amendment endeavored to correct the disparity in line with the purport of the 2012 Grand Bench Judgment and the 2014 Grand Bench Judgment; in addition, the 2015 Amendment Act states in its Supplementary Provisions that review is to be made continuously, aiming to definitely draw a conclusion on the fundamental reform of the election system in preparation for the next ordinary election; we can conclude that the direction and the strong will of the legislative branch toward further correction of the disparity in the value of votes in the future are shown in these provisions, and that the legislative branch is endeavoring not to again create great disparity at the abovementioned level. This Court then concluded as follows: the disparity between constituencies in terms of the value of votes under the provisions on the apportionment of seats at the time of the 2016 Election did not indicate the existence of extreme inequality to such an extent that it could raise a question of unconstitutionality, and the same provisions cannot be held to have been unconstitutional at that point in time.

(9) In the 2016 Election, the voter turnout declined and reached record lows at the time in three of the four merged prefectures (except Shimane Prefecture). In addition, the rate of invalid votes exceeded the national average in these prefectures, and the rate of invalid votes in Kochi Prefecture was the highest in the country. On July 29, 2016, the National Governors' Association adopted the "Resolution on Dissolution of Mergers in Election for the House of Councillors" requesting the immediate dissolution of the merged constituencies, alleging that various adverse effects, including a remarkable decline in the voter turnout, became visible in the 2016 Election. In addition, the National Association of Chairpersons of Prefectural Assemblies, the Japan Association of City Mayors, the National Association of Chairpersons of City Councils, the National Associations of Towns and Villages, and the National Association of Chairpersons of Town and Village Assemblies also passed resolutions toward immediate dissolution of the merged constituencies, etc., and many local assemblies also passed similar resolutions, etc.

After the holding of the 2016 Election, the Council on the Reform of the House of Councillors comprising the representatives of the parliamentary groups of the House of Councillors was established in February 2017, and the "Expert Committee on the Election System" was established under the same Council in April 2017 to conduct intensive investigations on the reform of the election system for the House of Councillors. Regarding the approach to the reform of the election system for the House of Councillors, the Expert Committee on the Election System had deliberations on such matters as the disparity in the value of one vote, framework of the election system and desirable number of seats based thereon, and framework of constituencies. Then, regarding members to be elected from constituencies, the same committee discussed the following options: adopting prefecture-based constituencies in which at least one member is elected from all prefectures; adopting prefecture-based constituencies, including some merged constituencies; and adopting broader constituencies (hereinafter referred to as "block constituencies") in lieu of prefecture-based constituencies. In addition, the same committee also held discussions on a desirable election system, etc. including an option to abolish the current structure consisting of two kinds of members, members to be elected from constituencies and members to be elected by proportional representation. However, opinions on the specific direction of the reform of the election system indicated by the parliamentary groups after going through these discussions had a significant gap in terms of the unit of a constituency, maintenance or abolition of merged constituencies, increase or decrease in the number of seats, etc. In May 2018, the same committee submitted a report on these deliberation results to the Council on the Reform of the House of Councillors.

In June 2018, at the Council on the Reform of the House of Councillors, the Liberal Democratic Party of Japan suggested an option including: using a prefecture as a unit of constituency; increasing the number of seats for members to be elected from constituencies by two and apportioning those seats to Saitama Prefecture while maintaining two merged constituencies from four prefectures based on the 2015 Amendment; and increasing the number of seats for members to be elected by proportional representation by four and introducing a specific ‘frame’ system under which political parties, etc. can specify candidates who should be preferentially elected. After that, deliberations, etc. were held at the meeting of the representatives of parliamentary groups, but there was a gap among the opinions of the parliamentary groups. Therefore, it was decided that each parliamentary group would submit its own bill to the House of Councillors and that the Special Committee on Political Ethics and Election System (hereinafter referred to as the "Special Committee of House of Councillors") would carry forward discussions thereon. In addition to a bill that is in line with the aforementioned suggestions made by the Liberal Democratic Party, bills including elements such as the introduction of a block constituency-based election in lieu of the current election system consisting of election of members to be elected from constituencies and that of members to be elected by proportional representation, were submitted. On July 11, 2018, the Special Committee of House of Councillors determined that the bill partially amending the Public Offices Election Act that is in line with the aforementioned suggestions made by the Liberal Democratic Party should be approved. In doing so, the Special Committee of House of Councillors passed an additional resolution to the effect that "discussions will be continuously held on the reform of the election system for the House of Councillors in the future in light of the role and desirable form of the House of Councillors in accordance with the purport of the Constitution."

A law as stated in the aforementioned bill (Act No. 75 of 2018; hereinafter referred to as the "2018 Amendment Act") was enacted on July 18, 2018 and came into effect on October 25, 2018 (hereinafter the provisions on the apportionment of seats amended by the same Act are referred to as the "Provisions on the Apportionment of Seats"). As a result of the amendment to the Public Offices Election Act by the 2018 Amendment Act (hereinafter referred to as the "2018 Amendment"), the maximum disparity between constituencies based on the population of the Japanese people according to the result of the population census implemented in October 2015 was 2.99 to 1.

(10) The Election was held on July 21, 2019 as the first ordinary election held under the Provisions on the Apportionment of Seats after the 2018 Amendment. The maximum disparity between constituencies was 3.00 to 1 at the time of the Election. In the Election, the voter turnout in Tokushima Prefecture, which was subject to merger of constituencies, was the lowest in the country, and the voter turnout also reached record lows in Tottori Prefecture and Shimane Prefecture, respectively. Moreover, the rate of invalid votes exceeded the national average in all of the four prefectures that became subject to the merger of constituencies, and the rate in Tokushima Prefecture was the highest in the country.

3.(1) The Constitution is considered as requiring equality in the content of the right to vote, in other words, equality in the influence of each voter's vote on the election of a member, or equality in the value of votes. However, the Constitution leaves a decision on the method of designing an election system to the Diet's discretion in order to fairly and effectively reflect the interests and opinions of the people concerning national politics. Therefore, equality in the value of votes does not serve as the only absolute criterion for determining the mechanism of an election system, and it should be realized in harmony with other policy purposes or reasons which can be justifiably taken into account by the Diet. For this reason, as long as a matter that was concretely specified by the Diet is reasonable as the exercise of its discretion, it is not considered to be unconstitutional even if it requires a certain degree of concession to be made in equality in the value of votes.

It is considered that the Constitution adopts the bicameral legislative system and creates a difference in the authority and the term of office, etc. of the members between the House of Representatives and the House of Councillors for the purpose of having these houses exert their own distinctive functions and thereby making the Diet an organization that fairly and effectively represents the people. From such perspective, under the mechanism of the election system for members of the House of Councillors seen in 2.(1) above, members of the House of Councillors are divided into members to be elected from the national constituency (members to be elected by proportional representation after the 1982 Amendment) and members to be locally elected (members to be elected from constituencies after the same Amendment), and an election of the former members is held throughout the area of the whole country (all prefectures) while for the latter members, a prefecture is used as a unit of constituency. It cannot be said that the fact that the Diet set provisions on such mechanism of the election system went beyond the scope of the reasonable exercise of the Diet's discretion at the time of the enactment of the Act on the Election of Members of the House of Councillors in 1947 and the Public Offices Election Act in 1950. However, if extreme inequality in the value of votes arises under the aforementioned mechanism as a result of constant population changes in the midst of severe social and economic changes and failure to take measures for correcting such inequality despite continuation of the inequality for a considerable period of time is determined to go beyond the limits of the Diet's discretion, it is reasonable to consider that the relevant provisions on the apportionment of seats are unconstitutional.

The idea described above is as intended in the cumulative Grand Bench judgments concerning election of members of the House of Councillors (members to be locally elected or members to be elected from constituencies) after the 1983 Grand Bench Judgment, and it is not found necessary to change this idea as the basic framework for determination.

(2) Under the bicameral legislative system, the Constitution grants superiority to the House of Representatives with regard to certain matters but sets the term of office of members of the House of Councillors as six years, a longer period of time than the House of Representatives, and provides that election for half the members shall take place every three years without dissolution (Article 46, etc.). The purport thereof is considered as reflecting the will of the people from a multifaceted and long-term perspective, ensuring the restriction and balance of the authority of the House of Councillors in relationship to the House of Representatives, and ensuring the stability and continuity of the operation of national politics by making the term of office of members of the House of Councillors longer than that of members of the House of Representatives while granting the House of Councillors almost the same authority as the House of Representatives with regard to many matters, including legislation. The issue of what specific election system is to be used to realize the aforementioned purport of the Constitution and achieve harmony with the requirement for equality in the value of votes is left to the reasonable discretion of the Diet, including the issues of how to position the nature and function of the House of Councillors under the bicameral legislative system and differences from the House of Representatives and how to reflect them in the election systems for the House of Councillors and the House of Representatives. It is also considered acceptable as the reasonable exercise of discretion given to the Diet when establishing the mechanism of an election system to adopt different election systems for members of the House of Representatives and members of the House of Councillors to reflect various opinions of the people at all levels of society and thereby have the House of Councillors exert unique functions that differ from those of the House of Representatives.

Moreover, in deciding a specific mechanism of the election system, the idea of taking into account the significance and substance, etc. of each prefecture serving as a political unit as one factor from the perspective of adding the purpose and function of collectively reflecting the will of the residents in certain regions should not be denied in itself. As long as it is in harmony with the requirement for equality in the value of votes, establishment of an election system based on such a factor is not immediately construed to have gone beyond the Diet's reasonable discretion.

(3) The Election was held under the Provisions on the Apportionment of Seats after the amendment by the 2018 Amendment Act, which was enacted after the 2017 Grand Bench Judgment was rendered. The same Act included increasing the total number of seats and increasing two seats in Saitama Prefecture while maintaining two merged constituencies from four prefectures based on the 2015 Amendment.

The merger of constituencies that was introduced through the 2015 Amendment made it possible to reduce the disparity between constituencies while setting an even number of seats for each constituency in light of the constitutional requirement for election for half the members, given limitations regarding the possibility of choosing the option of significantly increasing the total number of seats. However, there were also opinions that point out relevance between the merger and a decline in the voter turnout and a rise in the rate of invalid votes in the subject prefectures and strongly express the desire for the dissolution of the merger. Under such circumstances, after the 2016 Election, the Expert Committee on Election System, established under the Council on the Reform of the House of Councillors, held discussions on the disparity in the value of one vote, framework of the election system, desirable number of seats, and framework of constituencies, etc., and also held broad discussions on the propriety of the merger system and reform plans, such as introduction of block constituencies in lieu of prefecture-based constituencies. However, there was a large gap between parliamentary groups with regard to specific plans for the reform of the election system, and the 2018 Amendment Act was enacted in preparation for the Election without achieving any unanimous conclusion. Because of such background, the content of the same Act does not change the basic mechanism of the conventional election system for members to be elected from constituencies, but it did not only maintain the merger of constituencies despite the existence of the aforementioned strong desire for the dissolution of the merger so as to prevent the disparity reduced by the 2015 Amendment from expanding again but also chose the option of increasing the total number of seats, which had not been conducted for a long period of time, and increased the number of seats in Saitama Prefecture by two, thereby having tried to correct the disparity. As a result, the disparity between constituencies that was reduced from around 5 to 1 to about 3 to 1 through the 2015 Amendment (3.08 to 1 at the time of the 2016 Election) was further reduced, albeit only slightly, to 2.99 to 1 (3.00 to 1 at the time of the Election).

(4) As mentioned in 2.(8) above, the 2016Grand Bench Judgment first pointed out that Article 7 of the Supplementary Provisions of the 2015 Amendment Act provides that review is to be made continuously, aiming to definitely reach a conclusion on the fundamental reform of the election system in preparation for the next ordinary election. Then, the same judgment stated that it can be determined that the 2015 Amendment overcame the prolonged existence of great disparity between constituencies in terms of the value of votes over a long period of time and is aiming to realize further correction of the disparity. The same judgment then held that upon comprehensive consideration of such situations, the disparity between constituencies in terms of the value of votes at the time of the 2016 Election did not indicate the existence of extreme inequality to such an extent that it could raise a question of unconstitutionality. The Election was held under the Provisions on the Apportionment of Seats in the 2018 Amendment Act, which was enacted after the same judgment was rendered, and the disparity in the value of votes in the Election should be evaluated in light of the circumstances held in the same judgment.

We consider the case on this point. As a result, the 2018 Amendment Act, which was enacted after the 2016 Election, only took a measure to increase the number of seats in one constituency by two with regard to members to be elected from constituencies. On the other hand, the same Act did not have provisions like the aforementioned Supplementary Provisions, and an additional resolution to the effect that review is to be made continuously on the reform of the election system for the House of Councillors in line with the purport of the Constitution was made in deliberations concerning the same Act. However, the resolution did not clearly refer to the correction of the disparity between constituencies, etc. An election system that properly reflects the will of the people is the foundation of democracy, and it is hard to find any ground for considering that the requirement for equality in the value of votes may be taken more lightly for an election of members of the House of Councillors. As the 2017 Grand Bench Judgment, etc. also pointed out that equality in the value of votes should be achieved in harmony with the purport of the Constitution as mentioned in (2) above, the legislative branch is being required to hold discussions on measures, etc. necessary for maintaining the reduced disparity while preventing it from expanding again and promote efforts therefor, as well as promoting further correction of the disparity, in anticipation of constant population changes in the future. However, the 2018 Amendment as mentioned above cannot be considered to be showing great progress of such efforts.

However, in light of the background and content, etc. of the 2018 Amendment as mentioned above, the same Amendment maintained the merger of constituencies and corrected the disparity, albeit only slightly, in the situation where a specific reform plan could have not been easily established even after going through various discussions and reviews concerning the election system for the House of Councillors and there was also a strong desire for the dissolution of the merger of constituencies. Therefore, the same Amendment can be considered as giving considerations in order to maintain the direction of the 2015 Amendment Act, which reduced the maximum disparity that had hovered around 5 to 1 over several decades down to the aforementioned level. In addition, taking into account the fact that careful considerations are required for the reform of the election system for the House of Councillors in terms of its nature (for example, it is necessary to take into account the role, etc. which the House of Councillors should play, which are drawn from the mechanism of the bicameral legislative system adopted under the Constitution), realization of the reform must be gradual in part. Based on this, it is impossible to determine that an attitude oriented to the correction of the disparity was lost in the course of the review by the legislative branch.

(5) Upon comprehensive consideration of the aforementioned circumstances, it cannot be said that the disparity between constituencies in terms of the value of votes under the Provisions on the Apportionment of Seats after the 2018 Amendment indicated the existence of extreme inequality to such an extent that it could raise a question of unconstitutionality at the time of the Election. Therefore, the Provisions on the Apportionment of Seats cannot be held to have been unconstitutional at that point in time.

4. Based on the above, the determination of the court of prior instance that the Provisions on the Apportionment of Seats cannot be held to have been unconstitutional at the time of the Election can be accepted. All the arguments made by the appellants are not acceptable.

Accordingly, the Court unanimously decides as set forth in the main text of the judgment, except for the dissenting opinions stated respectively by Justice HAYASHI Keiichi, Justice MIYAZAKI Yuko, and Justice UGA Katsuya. Incidentally, there are the opinions stated respectively by Justice MIURA Mamoru and Justice KUSANO Koichi.

The opinion stated by Justice MIURA Mamoru is as follows.

I agree with the majority opinion conclusion, but I express my opinion as I think that the disparity between constituencies in terms of the value of votes under the Provisions on the Apportionment of Seats indicated the existence of extreme inequality to such an extent that it could raise a question of unconstitutionality.

1. The Grand Bench of this Court has examined the constitutionality of provisions on the apportionment of seats for an election of members of the House of Councillors in light of the relationship between judicial power and legislative power on the premise of a framework for determination comprising the following two steps: [i] whether the disparity between constituencies in terms of the value of votes under the provisions on the apportionment of seats indicates the existence of extreme inequality to such an extent that it could raise a question of unconstitutionality and [ii] if this is indicated by the disparity, whether the provisions on the apportionment of seats are unconstitutional on the grounds that failure to correct the disparity within the period before the relevant election is considered to go beyond the limits of the Diet's discretion. The details of the examination are as held in the 2014 Grand Bench Judgment, etc.

2. With regard to whether the Provisions on the Apportionment of Seats are constitutional or not, I first consider whether the disparity between constituencies in terms of the value of votes under the Provisions on the Apportionment of Seats indicates the existence of extreme inequality to such an extent that it could raise a question of unconstitutionality (hereinafter this is also referred to as an "unconstitutional state") in light of the aforementioned framework for determination.

(1) As mentioned in 3.(2) in the majority opinion, the specific election system that is to be used to realize the purport of the Constitution pertaining to the bicameral legislative system and achieve harmony with the requirement for equality in the value of votes is left to the reasonable discretion of the Diet. It is also considered acceptable as the reasonable exercise of the discretion given to the Diet when establishing the mechanism of the election system to adopt different election systems for members of the House of Representatives and members of the House of Councillors to reflect various opinions of the people at all levels of society and thereby have the House of Councillors exert unique functions that differ from those of the House of Representatives.

On the other hand, looking at changes in the election systems of both houses, system amendments have been conducted in both houses so as to realize an election system putting emphasis on political parties, and both houses have also adopted similar election methods that are based on the combination of election of members in a manner designed to use a prefecture or a segmentalized part thereof as a constituency and election of members to be elected by a proportional representation method designed to use a broader area as a unit of election. As a result, the election systems of both houses have come to have the same nature. In rapidly changing social conditions, the House of Councillors has come to play a greater role in the operation of national politics than ever against the backdrop of the longer term of office of its members.

In such circumstances, the Constitution does not indicate any specific content or direction of the positioning of the character, function, etc. of the House of Councillors and the way of reflecting these elements on its election system, and these matters should be determined based on various political or policy considerations. On the other hand, taking into account that equality in the value of votes is directly required under the Constitution, it is necessary to carefully review and evaluate the reasonableness of an election system.

(2) It is obvious that under the Constitution, the House of Councillors, together with the House of Representatives, is responsible for appropriately reflecting the will of the people on national politics as the highest organ of state power. Therefore, the requirement for equality in the value of votes should not be taken more lightly just because the election in question is for members of the House of Councillors.

In addition, the House of Councillors has unique factors that must be taken into account in apportioning seats. For example, as the Constitution provides that election takes place for half the members of the House of Councillors every three years, it is assumed that an even number of seats is apportioned to each constituency. Even if such factors can be technical constraints on the designing of the election system, it does not give a ground for permitting the existence of much greater disparity in the value of votes compared to the House of Representatives.

In light of the fact that a criterion for demarcation of constituencies to the effect that the disparity between constituencies in terms of population is basically less than 2 to 1 is set for the House of Representatives as an institutional consideration to the requirement for equality in the value of votes, it is also necessary to give sufficient consideration to the requirement for equality in the value of votes in relation to the House of Councilors so that the will of the people is further appropriately reflected.

Based on the idea as described above, I consider the Provisions on the Apportionment of Seats.

(3) A. Concerning the level of the disparity between constituencies

In elections of members of the House of Councillors, the maximum disparity between constituencies have hovered around 5 to 1 over the years. After two judgments of the Grand Bench of this Court to the effect that the disparity between constituencies in terms of the value of votes was in an unconstitutional state, in the 2015 Amendment, merger of constituencies was conducted for some constituencies with a small population for the first time after the foundation of the House of Councillors. Thereby, the maximum disparity between constituencies was reduced to around 3 to 1. In the 2018 Amendment, the option of increasing the total number of seats, which had not been conducted for a long time, was chosen, and the number of seats was increased for one constituency. As a result, the maximum disparity between constituencies was further reduced, albeit only slightly.

However, even though the maximum disparity between constituencies was reduced from around 5 to 1 to around 3 to 1, the reduction does not naturally legitimize such disparity in the value of votes. Actually, Article 7 of the Supplementary Provisions of the 2015 Amendment Act clearly specifies that efforts are to be made for the fundamental reform of the election system in consideration of the correction of the disparity between constituencies, etc. In light of this, the 2017 Grand Bench Judgment also pointed out that the 2015 Amendment can be determined to be aiming to realize further correction of the disparity and determined that the disparity in the value of votes cannot be held to have been in an unconstitutional state. Both the above provisions and the judgment can be understood as being made on the premise that even if the maximum disparity between constituencies was reduced, it would be necessary for the Diet to advance efforts for further correction of the disparity, and that in that sense, there still existed the disparity in the value of votes that should be corrected.

In the first place, the disparity in the value of votes at a level of around 3 to 1 itself must be considered to be still large in light of the fundamental principle of election: one person one vote, and in consideration of the fact that equality in the value of votes is related to the fundamentals of the sovereignty of the people and of parliamentary democratic politics. This can also be considered as the issue of a sense of equality and a sense of fairness in respect to the exercise of rights by the people who are sovereigns.

In addition, at the time of the Election, the disparity between constituencies in terms of the number of voters per member (hereinafter the "disparity between constituencies" at the time of each election means this disparity in the number of voters) exceeded 3 to 1 only in one constituency. However, the disparity between constituencies exceeded 2.9 to 1 in four constituencies, and the total number of voters in those constituencies accounted for about 21.6% of all voters. This should be considered as great disparity that can never be ignored. Comparing this figure with the corresponding figure in the 2016 Election, the disparity between constituencies at the time of the same election exceeded 2.9 to 1 in three constituencies, and the number of voters in those constituencies accounted for about 9.4% of all voters and this rate was just not more than half of that at the time of the Election. Based on this, in terms of the level of the disparity between constituencies of over 2.9 to 1, the disparity in the value of votes rather expanded despite the 2018 Amendment, and it is also expected to further expand in the future.

As mentioned above, the House of Councillors, together with the House of Representatives, is responsible for appropriately reflecting the will of the people in national politics as the highest organ of state power, and a criterion for demarcation of constituencies to the effect that the disparity between constituencies in terms of population is basically less than 2 to 1 is set for the House of Representatives as an institutional consideration to the requirement for equality in the value of votes. In light of these facts, the continuation of such great disparity over several decades with regard to the constitutional requirement for equality in the value of votes at the House of Councillors indicates the existence of obvious inequality that should be corrected, and it thus should be considered to raise a question of unconstitutionality unless there are reasonable circumstances to justify it.

Therefore, I consider whether there are reasonable circumstances to justify the level of the disparity between constituencies at the time of the Election.

B. Concerning a unit of constituency

The mechanism of an election system: setting the number of seats using a prefecture as a unit of constituency, is considered to aim at adding the purpose and function of collectively reflecting the will of residents who constitute a prefecture in light of the fact that each prefecture has been defined as a political unit with its own historical, political, economic, and social significance and substance. It is undeniable that this idea has had proper reasonableness to the extent that a prefecture is a regional unit of administration, etc.

In this regard, the 2017 Grand Bench Judgment ruled that in deciding a specific mechanism of the election system, the idea of taking into account the significance and substance, etc. of each prefecture as one factor should not be denied in itself, and determined that as long as it is in harmony with the requirement for equality in the value of votes, establishment of an election system based on such a factor is not construed to have gone beyond the Diet's reasonable discretion. The issue is whether harmony with equality in the value of votes is kept, and in making a determination concerning it, it is necessary to conduct careful review from the perspective of whether there are reasonable circumstances to justify the continuation of the disparity in the value of votes, which should be corrected.

Incidentally, regarding the number of seats for members to be locally elected in each constituency, the Act on the Election of Members of the House of Councillors, which was enacted in 1947, gave consideration so that election takes place for half the members throughout all constituencies in light of the constitutional requirement for election for half the members. Under the polity of setting an even number of seats and apportioning at least two seats to each constituency, population in each constituency is divided by a value obtained by dividing the total population as of 1946 by the number of seats, 150, and an even number of seats, ranging from two to eight, were apportioned to each constituency in proportion to the size of each prefecture based on the obtained value.

The provisions on the apportionment of seats of the Public Offices Election Act enacted in 1950 inherited the aforementioned apportionment of seats without change. The disparity between prefectures in terms of population greatly expanded due to subsequent population changes, and the apportionment of seats had become non-proportional to population. However, while maintaining the apportionment of two seats to constituencies with a small population on the premise of apportionment of an even number of seats, the Diet had repeatedly made such adjustments as reapportionment of seats in other individual constituencies.

As a result, looking at values obtained by dividing the number of voters per member in each constituency at the time of the Election by the number of voters per member in the whole country, the maximum value was about 1.36 (Miyagi Prefecture), and the obtained values exceeded 1.35 only in this one constituency. On the other hand, the minimum value was about 0.45 (Fukui Prefecture), and the obtained values were less than 0.5 in three constituencies (Fukui Prefecture, Yamanashi Prefecture, and Saga Prefecture), and the obtained values were 0.5 or more but less than 0.65 in six constituencies. Therefore, the current apportionment of seats as a whole can be considered as being biased toward constituencies with less voters. In particular, for three constituencies for which the aforementioned value is less than 0.5, the values are not sufficient even for one seat in the sense of apportionment of seats in proportion to population or the number of voters. Therefore, it is obvious that uniform apportionment of two seats to these constituencies by adding extra one seat is the main factor causing the aforementioned disparity in the value of votes. At the same time, this must also be considered as a factor making it extremely difficult for the Diet to realize further correction of the disparity.

Equality in the value of votes is directly required under the Constitution and is related to the fundamentals of the sovereignty of the people and of parliamentary democratic politics. On the other hand, there is no constitutional requirement for use of a prefecture as a unit of constituency. In addition, while apportionment of an even number of seats to each constituency is assumed as a principle because of the constitutional requirement for election for half the members, it cannot be necessarily considered as a constitutional requirement.

Therefore, in the situation where inequality that should be corrected continues to exist and correction of such inequality is accompanied by extreme difficulties because of maintenance of a basic mechanism designed to use a prefecture as a unit of constituency on the premise of apportionment of two seats even to constituencies with especially small populations, it cannot be said that there are reasonable circumstances to justify the inequality.

C. Concerning the merger of constituencies

The 2015 Amendment introduced an unprecedented method, that is, merger of some constituencies with small populations, and thereby the maximum disparity between constituencies was reduced to a considerable extent. This was the partial reform of the mechanism of the election system designed to use a prefecture as a unit of constituency, as pointed out in the 2017 Grand Bench Judgment. The content thereof was the merger of four constituencies into two constituencies, and demarcation was the same as before for the other 43 constituencies. Therefore, the scope of the reform was very limited. This amendment should be considered as making a partial modification to some constituencies while basically maintaining the aforementioned mechanism of the election system.

As mentioned above, the mechanism of the election system designed to use a prefecture as a unit of constituency is considered to aim at adding the purpose and function of collectively reflecting the will of residents in a prefecture in light of the significance and substance, etc. of the prefecture. The merger of constituencies itself results in denying the premise of the system, the nature of a prefecture as a unit, to that extent. The significance, etc. of a prefecture as a unit is originally not affected by the size of population, and it is hard to say that among many mutually adjacent prefectures, the areas currently merged as constituencies especially have unique significance or substance that is equivalent to a prefecture used as a unit. Therefore, such merger of constituencies seems to be originally inconsistent with the basic idea of the system of collectively reflecting the will of residents in a prefecture in light of the significance and substance, etc. of the prefecture even though it is a policy measure designed to correct the disparity between constituencies in terms of the value of votes.

In addition, in the Election, the number of voters in Fukui Prefecture, where the number of voters was the smallest, was 646,976. However, in the constituency created by merging Tokushima Prefecture and Kochi Prefecture for the election for the House of Councillors, the number of voters in Tokushima Prefecture was 636,739 and that in Kochi Prefecture was 610,498. Then, despite the fact that all three are prefectures and have almost the same number of voters, due to slight differences in the number of voters, they are handled differently from an institutional perspective in terms of the collective reflection of the will of residents. It is also difficult to give reasonable explanation to such handling in relation to the aforementioned basic idea of the system.

These points show the limitations of giving partial treatment on the premise of an inflexible framework, the prefecture, as a method of realizing equality in the value of votes. As prefectures covered by the current merger of constituencies are the four prefectures with the smallest populations in the country, there is a reason that residents in those prefectures consider such measure as discarding areas with a small population. In addition, I can fully understand that many residents strongly desire the dissolution of the merger of constituencies. The expansion of the merger of constituencies to other constituencies is considered to involve considerable difficulties in the Diet's efforts to promote further correction of the disparity.

Based on the above, introduction of such merger of constituencies, which intended to correct the disparity in line with the purport of the aforementioned Grand Bench judgment of this Court, is found to have only partially and provisionally altered the mechanism of the election system, and the aforementioned problem is also recognized. Therefore, even if the merger of constituencies introduced through the 2015 Amendment has been maintained, it cannot be said that there are reasonable circumstances to justify the disparity of around 3 to 1.

D. Concerning the Diet's attitude toward correction of the disparity, etc.

The 2017 Grand Bench Judgment pointed out that the direction and the strong will of the legislative branch toward further correction of the disparity in the value of votes in the future are shown in the provisions of Article 7 of the Supplementary Provisions of the 2015 Amendment Act and that the legislative branch is endeavoring not to again create great disparity at the abovementioned level, and stated that the 2015 Amendment is aiming to realize further correction of the disparity. The same judgment then held that the disparity in the value of votes at the time of the 2016 Election cannot be held to have been in an unconstitutional state. In light of such holding, I consider whether the Diet's attitude, etc. in the 2018 Amendment can be considered as constituting reasonable circumstances to justify the disparity of around 3 to 1.

First of all, as mentioned above, the 2018 Amendment can be considered to be basically the same as legal amendments that had been repeatedly made, specifically, adjusting the number of seats in some constituencies under the framework of the election system basically designed to use a prefecture as a unit of constituency for the election of members to be elected from constituencies (hereinafter referred to as an "election using constituencies"). It is obvious that this amendment cannot be considered as one that changes the basic mechanism of the election system itself.

Moreover, in deliberations at the Diet, the proponent of the bill for the 2018 Amendment Act made an answer to the effect that he/she thinks that measures taken based on this bill fall under the fundamental reform of the election system as provided in the Supplementary Provisions of the 2015 Amendment Act, and he/she then cited the introduction of a specific frame for the election of members to be elected by proportional representation (hereinafter referred to as an "election by proportional representation") as the reason thereof. This specific frame is a system in which political parties, etc. can specify candidates who should be preferentially elected for an election by proportional representation while basically maintaining the current open-list system. The proponent of the bill repeatedly made an answer to the effect that this specific frame is assumed to be utilized for delivering the opinions of residents in regions, that are considered to be minorities in terms of population, such as the prefectures subject to the merger of constituencies, to the national government for the purpose of delivering the opinions of residents in regions in units of a prefecture in light of many people's strong desire for the dissolution of the merger of constituencies. However, election by proportional representation and election using constituencies are provided for as two different types of election in the Public Offices Election Act, and a specific frame is nothing more than a classification of candidates of political parties, etc. in election by proportional representation. A candidate in such specific frame is elected as a result of voting by voters nationwide on the premise of use of a specific frame by a relevant political party, etc., and the will of voters in a specific region is not necessarily consolidated therein.

Therefore, it is obvious that issues related to the reform of the election system, including correction of the disparity, are still left even in light of the 2018 Amendment.

Furthermore, the 2012 Amendment Act and the 2015 Amendment Act had the following provisions in their Supplementary Provisions, respectively: review is to be made with a time limit, aiming to definitely reach a conclusion on the fundamental reform of the election system. On the other hand, the 2018 Amendment Act has no similar provisions. In addition, an additional resolution adopted at the Special Committee of House of Councillors also did not refer to either correction of the disparity or the fundamental reform of the election system, and the content thereof was simply highly abstract and vague. The Diet could not reach a reasonable conclusion with regard to the fundamental reform of the election system though it continued review and deliberation for about six years while setting a time limit twice by itself. Under such situation, the Diet decided not to set any provision that imposes any obligation on itself in the 2018 Amendment Act. This change in the Diet's attitude must be said to be remarkable.

In addition, careful consideration and review are necessary in developing an election system in terms of its nature, and it is also considered as a realistic option pertaining to the Diet's discretion to realize an election system through repeated gradual reforms. However, as it is clear from the background and the content, etc. of the 2018 Amendment, neither specific content nor direction of further correction of the disparity has been set, and further correction of the disparity is generally uncertain. Therefore, the Diet also cannot be evaluated as being on the course of gradual progress toward the correction. Thus, such further correction of the disparity is an issue that goes beyond prevention of re-expansion of the disparity, and even if the maximum disparity between constituencies is slightly reduced as a result of maintenance of the merger of constituencies based on the 2018 Amendment, that fact does not affect the aforementioned determination.

Upon comprehensive consideration of these circumstances, it should be said that the 2018 Amendment cannot be evaluated as one that continuously maintains the direction and the strong will of the legislative branch toward further correction of the disparity indicated in Article 7 of the Supplementary Provisions of the 2015 Amendment Act and that is aiming to realize further correction of the disparity. It is obvious that such attitude of the Diet, etc. does not constitute reasonable circumstances to justify the aforementioned disparity in the value of votes.

Incidentally, I particularly examined the aforementioned points in light of the aforementioned holding in the 2017 Grand Bench Judgment concerning Article 7 of the Supplementary Provisions of the 2015 Amendment Act. However, it can hardly be considered as an objective evaluation of inequality to put emphasis on evaluation concerning such matter that falls under the Diet's discretion in determining whether the disparity in the value of votes is in an unconstitutional state in the absence of provisions indicating the content, direction, etc. of further correction of the disparity in the 2018 Amendment Act, irrespective of the method of evaluating the status of various reviews and efforts made by the Diet during this period. In addition, as I mentioned at the beginning, it is also considered unreasonable even in light of the purport of dividing the framework for determination into two steps based on the relationship between judicial power and legislative power.

(4) As mentioned above, the 2018 Amendment maintained the conventional basic mechanism of the election system and only adjusted the number of seats in some constituencies. Actually, the maximum disparity between constituencies still continues to be at the level of 3 to 1 throughout the period before and after the same Amendment. Such disparity had reached a level that could not be ignored in light of the importance of equality in the value of votes. In addition, no reasonable circumstance to justify such disparity can be found.

Therefore, I must say that the disparity between constituencies in terms of the value of votes under the Provisions on the Apportionment of Seats indicated the existence of extreme inequality to such an extent that it could raise a question of unconstitutionality.

3. Therefore, next, I consider whether it goes beyond the limits of the Diet's discretion that the Diet failed to correct extreme inequality to such an extent that it could raise a question of unconstitutionality during the period up to the Election.

It can be said that the 2012 Grand Bench Judgment and the 2014 Grand Bench Judgment indicated a determination that the disparity between constituencies in terms of the value of votes is in an unconstitutional state and the mechanism of the election system itself needs to be reviewed to solve it, and that this caused the Diet to assume the responsibility of correcting that state. Moreover, provisions to the effect that review is to be made continuously, aiming to form a conclusion on the fundamental reform of the election system, were also established in the Supplementary Provisions of both the 2012 Amendment Act and the 2015 Amendment Act, respectively.

In light of these, the Diet conducted various broad reviews and deliberations with regard to the reform of the election system over several years. However, while such discussions were still ongoing, the 2017 Grand Bench Judgment ruled that the 2015 Amendment can be seen as having endeavored to correct the disparity in line with the purport of the 2012 Grand Bench Judgment and the 2014 Grand Bench Judgment, and indicated a determination that the disparity between constituencies in terms of the value of votes cannot be held to be in an unconstitutional state and that the provisions on the apportionment of seats cannot be held to have been unconstitutional at that point in time. In indicating such determination, this Court neither attached any special clear reservation to the determination nor made any specific indication with regard to the reform of the mechanism of the election system or the necessity of further correction of the disparity.

The Election was held under the Provisions on the Apportionment of Seats after the amendment by the 2018 Amendment Act, which was enacted immediately after the 2017 Grand Bench Judgment. On the premise of the 2017 Grand Bench Judgment as mentioned above, it must be said that it is difficult to find any circumstance based on which the Diet could specifically recognize that the disparity between constituencies in terms of the value of votes under the Provisions on the Apportionment of Seats had reached a level of causing extreme inequality that could raise a question of unconstitutionality before the Election.

Based on the above, failure to correct extreme inequality to such an extent that it could raise a question of unconstitutionality during the period up to the Election cannot be considered to go beyond the limits of the Diet's discretion, and thus, it cannot be said that the Provisions on the Apportionment of Seats were unconstitutional at that point in time.

4. I make an additional remark in the last place. Regarding the election system for members of the House of Councillors, great disparity in the value of votes has continuously existed over the years along with the expanding disparity between prefectures in terms of population caused by concentration of population in urban areas, under the mechanism designed to use a prefecture as a unit of constituency on the premise of apportionment of an even number of seats in light of the constitutional request for election for half the members within the framework of the limited total number of seats.

In light of the cumulative judgments of the Grand Bench of this Court, the Diet had continued reviews and deliberations on the reform of the election system for the House of Councillors and the correction of the disparity in the value of votes over several decades since 2004. The Diet had made such legal amendments as the merger of constituencies with small populations, as well as having reapportioned seats in some constituencies. However, the disparity in the value of votes that should be corrected has yet to be overcome even through these amendments.

Nevertheless, the Diet appears to have not conducted any specific review or deliberation concerning this issue during the period of not less than two years since the 2018 Amendment, even after the Election. Considering this fact along with the fact that no provision indicating the direction, etc. toward further correction of the disparity was established in the 2018 Amendment Act, there is a concern that no specific discussion will be made on this issue.

On the other hand, as mentioned above, taking into account that the disparity in the value of votes is rather expanding in terms of the level of the disparity between constituencies of over 2.9 to 1 despite the 2018 Amendment, if the next and subsequent ordinary elections are held while basically maintaining the Provisions on the Apportionment of Seats, the disparity between constituencies in terms of the value of votes is also expected to further expand.

In light of the fact that the election system that appropriately reflects the will of the people is the fundamentals of the sovereignty of the people and of parliamentary democratic politics and equality in the value of votes is a constitutional requirement: as well as the aforementioned role, etc. of the House of Councillors in national politics, the Diet is required to steadily carry forward the review and consolidation of measures for further correction of the disparity, including alteration of the current system basically designed to use a prefecture as a unit of constituency, so that the will of the people can be more appropriately reflected in anticipation of constant population changes in the future. In this manner, the aforementioned inequality that raises a question of unconstitutionality must be overcome through required legislative measures as soon as possible.

The opinion stated by Justice KUSANO Koichi is as follows.

I agree with the conclusion of the majority opinion, but my opinion somewhat differs from the majority opinion in terms of the reason for the conclusion. Therefore, I would like to clear up my opinion below.

1. The major index that this Court has used so far with regard to the issue of the disparity in the value of votes is the "maximum disparity." Certainly, the maximum disparity is a concise concept, and in light of the fact that both this Court and the legislative branch have carried forward various discussions by using this index, the maximum disparity can be considered as an especially useful index in making a determination concerning constitutionality in consideration of the history of mutual action between this Court and the legislative branch. However, as the maximum disparity is a concept that is focused only on the difference between voters who are granted the largest value of votes and those who are only granted the smallest value of votes, it must be considered as slightly lacking accuracy as an index that is used for discussing the disparity in the allocation of the value of votes in the entire election system, setting aside the case where it is used as an index that is used by voters who are granted the smallest value of votes for highlighting how much disadvantage they are suffering. For this reason, I think of using the Gini coefficient, which is broadly used in statistics as an index for evaluating the disparity in the allocation of profits, as an analytical concept that complements the maximum disparity.

The Gini coefficient pertaining to the value of votes is calculated by the following method.

[i] Prepare a table in which all constituencies are arranged in descending order of the number of members per voter (hereinafter referred to as a "basic table"), and with respect to each constituency, calculate a value by dividing the sum of the numbers of voters from the constituency written at the top of the basic table to the relevant constituency by the total number of voters in all constituencies (hereinafter referred to as the "cumulative voter index").

[ii] With respect to each constituency, calculate a value by dividing the sum of the numbers of members from the constituency written at the top of the basic table to the relevant constituency by the total number of members in all constituencies (hereinafter referred to as the "cumulative member index").

[iii] Create a graph in which the cumulative voter index is put on the horizontal axis and the cumulative member index is put on the vertical axis, and with respect to each constituency, write a point corresponding to the combination of these two values in the graph.

[iv] Connect the aforementioned points with a line in series to obtain a Lorenz curve pertaining to the value of votes. Figure 1 (the URL below) shows a Lorenz curve for the Election, which was prepared by the aforementioned method.

[v] Obtain the ratio of the area of the crescent-shaped part surrounded by the Lorenz curve and a straight line connecting points a and b to the area of a triangle connecting points a, b, and c. The value of this ratio is the Gini coefficient. If the Lorenz curve matches a straight line connecting points a and b in Figure 1, the Gini coefficient is zero. On the other hand, If the Lorenz curve matches a line connecting points a, b, and c in series in Figure 1, the Gini coefficient is 1 (100%). In Figure 1, the Gini coefficient is 0.1422 (14.22%), and therefore, the Gini coefficient for the Election was 14.22% (hereinafter the values of the Gini coefficient and the maximum disparity are round numbers).

The Gini coefficient is an analytical concept suited for the purpose of evaluating the level of the disparity in the value of votes because it is possible to draw a conclusion, "the larger the Gini coefficient, the smaller the total amount of utility that all voters obtain," just using the rules of logic based only on the following two assumptions that can be evaluated as reasonable.

[i] The smaller the number of members per voter in a constituency, the smaller the utility that voters can obtain through holding of the right to vote.

[ii] The smaller the number of members per voter, the gradually higher the rate of reduction in utility (in other words, the larger the number of members per voter, the larger the utility that voters can obtain, but the increase rate of utility is gradual).

In addition, it is possible to determine what plan for improvement of the election system is efficient and what plan for improvement of the election system is inefficient by paying attention to changes in the Gini coefficient (this point is discussed in detail in 3.(3) and 3.(4)).

For the reasons described above, I will use the Gini coefficient as a major index to advance analysis.

2. The Gini coefficient for the 2010 Election and the 2013 Election, for which this Court determined that the disparity in the number of seats for members of the House of Councillors was in an unconstitutional state, was 22.97% and 20.55%, respectively (Figure 2 (the URL below) shows a Lorenz curve for the 2013 Election). In comparison to these values, the Gini coefficient for the Election, 14.22%, is quite low, and it thus can be said that the disparity in the number of seats had been considerably improved in the Election compared to the disparity at the time of the 2010 Election and the 2013 Election.

Incidentally, the apportionment of seats to constituencies at the time of the Election gave the maximum consideration to the disparity in the value of votes as long as the current total number of seats for an election using constituencies, demarcation of constituencies, and apportionment of at least two seats to each constituency are assumed, and therefore, it is unrealistic to desire further improvement of the Gini coefficient because of the following reason: as long as it is construed as impermissible under the Constitution to apportion less than two seats to prefectures for which the number of members per voter is very large (Fukui Prefecture and Saga Prefecture) (the possibility of adopting a different interpretation regarding this point is mentioned in 3.(2)), there is no other choice but to increase the number of seats for prefectures for which the number of members per voter is chronically small (Kanagawa Prefecture, Tokyo Metropolis, Osaka Prefecture, etc.) in order to improve the situation; however, in order to do so, there is no other choice but to reduce the number of seats for prefectures to which four or six seats are apportioned; however, there remains almost no such constituency for which the number of seats per voter is relatively large; if such reduction is forcibly executed, it will result in creating prefectures for which the number of members per voter is further smaller like the result that occurred in Miyagi Prefecture and Niigata Prefecture in the Election, although without fail the Gini coefficient will be slightly improved.

3. In that case, how should the election be changed to alleviate the disparity in the value of votes? In addition, what is an appropriate determination to be made by this Court in light of conceivable plans for improvement? I consider these points below.

(1) First of all, the following two are the only plans for improvement that are realistically conceivable for almost completely overcoming the disparity in the value of votes.

One of them is a method in which prefecture-by-prefecture constituencies are integrated into a large block constituency (hereinafter this plan is referred to as a "large block constituency plan"). For example, a bill submitted by the Japan Innovation Party in 2018 (the total number of seats for members of the House of Councillors is reduced to 218, and election is held by dividing the country into 11 constituencies) is a typical large block constituency plan. Given that this plan for improvement is implemented at the time of the Election, the Gini coefficient is to be reduced to 1.61% (Figure 3 (the URL below) shows a Lorenz curve in that case).

In another plan for improvement, demarcation of constituencies is freely conducted abandoning the idea of using a prefecture as a basic unit for dividing areas (hereinafter this plan is referred to as a "free demarcation plan"). If a thorough free demarcation plan is implemented, it is probably possible to almost completely overcome the disparity in the value of votes in the same manner as a large block constituency plan even if all constituencies are made into single-seat constituencies.

However, at any rate, requiring the Diet to implement either a large block constituency plan or a free demarcation plan poses a problem. First of all, if a large block constituency plan is implemented, election of members of the House of Councillors will be held only under a large-constituency system: more specifically, based on large constituencies consisting of considerably large numbers of seats. It is well known that the single-seat constituency system and large-constituency system both have advantages and disadvantages, and making a big change to the current combination of these two election systems will bring about a significant change to the legislative branch, and eventually to the workings of the State. Secondly, if a free demarcation plan is implemented, it is also doubtful whether voters can have a sense of belonging to a newly established constituency. In addition, a free demarcation plan will also bring about a significant change to the relationship between national politics and local politics, particularly, the workings of political parties that function as ties between these two types of politics.

In consideration of the points described above, the implementation of either a large block constituency plan or a free demarcation plan should be automatically determined by the legislative branch within the scope of discretion granted thereto under the Constitution. If this Court makes a determination that the situation is unconstitutional on the grounds of the legislative branch's failure to implement it, it must be considered as going against the purport of Article 47 of the Constitution that leaves the planning of an election system to the Diet's discretion.

(2)Next, I consider plans for improvement that cannot overcome the disparity in the value of votes but can significantly improve the situation, more specifically, plans that are expected to achieve improvement at the level of reducing the Gini coefficient to half or less of the current value. I think that some plans for improvement are conceivable in theory, but among them, only the following two plans can probably gain support from a certain percentage or more of the people. In the first plan, the current election by proportional representation is abolished (or the number of seats therefor is significantly reduced), and surplus seats arising from the abolition (or significant reduction) are preferentially apportioned to constituencies for which the number of seats per voter is small (hereinafter this plan is referred to as a "proportional-representation constituency abolition plan"). For example, if election by proportional representation is completely abolished and 100 surplus seats arising therefrom are apportioned as mentioned above (based on the number of voters in each constituency in the Election), the Gini coefficient is reduced to 6.19% (Figure 4 (the URL below) shows a Lorenz curve in that case).

In another plan for improvement, constituencies consisting of an odd number of seats, including one, are created on the premise of the current demarcation of constituencies (hereinafter this plan is referred to as an "odd number of seats plan"). If an odd number of seats plan is thoroughly implemented, it can be expected to achieve the reduction of the Gini coefficient that is equivalent to or more than that based on a proportional-representation constituency abolition plan in some cases.

However, for both of the aforementioned two plans for improvement, it must be said that a problem will occur if this Court requires the legislative branch to implement it. First of all, a proportional-representation constituency abolition plan is accompanied by a serious policy determination concerning the abolition of (or significant reduction of the number of seats for) election by proportional representation that has the advantage of large constituencies to the maximum extent. If the judicial branch determines that the legislative branch's failure to implement a proportional-representation constituency abolition plan constitutes an unconstitutional state and forces the legislative branch to implement the plan, it goes against the purport of Article 47 of the Constitution. Secondly, I think that it is possible to implement an odd number of seats plan under the Constitution and that its implementation will have a smaller influence on national politics than any other plans for improvement discussed above. However, it is undeniable that the current system in which an even number of seats is apportioned to all constituencies fits in with the purport of Article 46 of the Constitution best. As long as it is true, I must say that the judicial branch's act of determining that the legislative branch's failure to implement an odd number of seats plan constitutes an unconstitutional state and forcing the legislative branch to implement the plan also goes against the purport of Article 47 of the Constitution.

(3) Then, what are plans that can considerably improve the current situation though they cannot significantly alleviate the disparity in the value of votes? The plan that I can conceive of first is the implementation of the merger of constituencies for which the number of members per voter is large. The merger of constituencies has already been implemented, and increasing constituencies subject to merger can be realized without making significant changes to the current election system, unlike the plans for improvement mentioned above. However, increasing constituencies subject to merger will not cause a remarkable change to the Gini coefficient because of the following reason: an improvement method that will be the most efficient in terms of influence exercised by a change in the number of seats on the Gini coefficient will increase the number of seats for constituencies for which the number of members per voter is small while the most inefficient improvement method is to reduce of the number of seats for constituencies for which the number of members per voter is large (this is a conclusion that is inevitably drawn from the reasonable assumptions mentioned in [i] and [ii] in 1. above); thus, reducing the number of seats for constituencies for which the number of members per voter is large by implementing the merger of such constituencies is nothing less than the implementation of this most inefficient improvement method (another significance of implementation of the merger of constituencies is described in 3.(4)). As an example showing this point, I take up a plan for amendment suggested by the Democratic Party of Japan and Komeito, etc. in 2015 (10 merged constituencies are created by merging 20 prefectures into 10 constituencies of two prefectures each, and thereby, the number of seats for constituencies subject to the merger on the whole is reduced by 12 in total and those 12 seats are apportioned to constituencies for which the number of members per voter is small to increase the number of seats there). This plan for amendment seems to be almost close to the limit as a method designed to use the merger of constituencies in consideration of the actual difficulty of creating merged constituencies. However, assuming that the Election is held based on this plan for amendment, the maximum disparity is certainly reduced to 2.02 to 1 but the Gini coefficient is 12.31% [Figure 5 (the URL below) shows a Lorenz curve in that case], showing no remarkable improvement in terms of the Gini coefficient for the Election (14.22%). In light of the nature of the Gini coefficient as mentioned in 1 above, the total amount of utility that voters can obtain is not increased much even if the number of members in the subject constituencies is reduced by creating merged constituencies.

On the other hand, since long ago, it has been pointed out that the merger of constituencies has negative effects, such as [i] causing adverse effects on a sense of participation in politics among voters in constituencies subject to the merger and [ii] causing residents in a constituency with a smaller population to have a sense of being discriminated against if there is a large population difference between constituencies subject to the merger. In addition, many constituencies subject to merger exist in areas that are struggling to take measures against depopulation. Comparing and weighing these points, it is hardly considered appropriate to determine that the current system is in an unconstitutional state because of failure to increase merged constituencies.

(4) Incidentally, if merged constituencies are increased, it will cause surplus seats, and it thus becomes possible to use those surplus seats to increase the number of seats for constituencies for which the number of members per voter is small. This is the most efficient method for improving the Gini coefficient as discussed in the preceding section (one of the reasons that the Lorenz curve in Figure 1 and that in Figure 5 differ only by 2% in terms of the Gini coefficient is that the same increase in the number of seats as stated in the plan for amendment suggested by the Democratic Party of Japan and Komeito, etc. discussed in (3) was achieved by various methods [including increasing the total number of seats for election using constituencies by two] in the Election with respect to major prefectures for which the number of members per voter is small).

This reveals that an effective method for efficiently improving the Gini coefficient and also achieving such improvement without causing residents in some constituencies to have a sense of alienation or a sense of being discriminated against is to slightly increase the total number of seats and use the increased seats to increase the number of seats in constituencies for which the number of members per voter is small. For example, if the number of seats was increased by two each in Tokyo Metropolis, Kanagawa Prefecture, and Osaka Prefecture at the time of the Election (that is, the number of members covered by the Election was increased by three in total), it per se could reduce the Gini coefficient for the Election to 11.76% and could achieve about 2.5% improvement from the actual value (14.22%). As long as the number of members increased is small, it probably does not go against the purport of Article 43, paragraph (2), etc. of the Constitution to require the Diet to implement such increase.

However, a method using an increase in the total number of seats has the problem of requiring the people to bear a certain burden. Needless to say, increased members will make efforts to improve the welfare of the people, and the Diet will make efforts to avoid an increase in the operating cost of the Diet as far as possible. However, an increase in the number of members is likely to eventually increase the operating cost of the Diet. In light of this point, an abstract reason, existence of the disparity in the value of votes, is insufficient for this Court to make a determination that the situation is unconstitutional on the grounds that the disparity in the value of votes can be alleviated by increasing the number of seats, and it is probably necessary to make clear a specific fact that is sufficient to obtain the public understanding of the bearing of a new burden, in the court.

4. Based on the above consideration, it can be said that the disparity in the value of votes can be efficiently alleviated by a method using a slight increase in the total number of seats. However, I must say that it is difficult to determine that the disparity in the value of votes at the time of the Election was in an unconstitutional state just because this measure has not been taken to a sufficient extent, taking into account that this measure is likely to increase the operating cost of the Diet.

However, it seems that the aforementioned idea can be made into a proposal with practical problem-solving ability if it is modified to a certain extent. The modified idea is as follows: although the current disparity in the value of votes is recognized as prima facie constitutional, upon reconsideration, it will be found to be unconstitutional if a specific and material suspicion that certain people are suffering disadvantages due to the existence of the disparity in the value of votes (hereinafter referred to as a "suspicion of disadvantages") is indicated (hereinafter this idea is referred to as the "conditional constitutionality theory"). The proof of a suspicion of disadvantages required under the conditional constitutionality theory is by no means the strict proof of a causal relationship between the disparity in the value of votes and disadvantages suffered by certain people (originally, such proof is probably impossible). However, it is necessary and sufficient to indicate that there is a significant causal relationship between the disparity in the value of votes and relevant disadvantages even in consideration of other factors that can affect the occurrence of the disadvantages in question.

The conditional constitutionality theory overcomes the problem mentioned in 3.(4). That is, it is probably possible to obtain the understanding of the people with regard to the necessity of correcting the disparity in the value of votes even by assuming additional burden if a suspicion of disadvantages is proven. Furthermore, under the conditional constitutionality theory, about how many members should be increased to eliminate a suspicion of disadvantages and overcome an unconstitutional state becomes clear. Therefore, it becomes possible for this Court to indicate the rough number of members to be increased to overcome an unconstitutional state (explaining this point in detail, for example, if multiple regression analysis using the number of members per certain number of voters [or "certain number of residents"] as one of the explanatory valuables is used as a means of proving a suspicion of disadvantages, the number of members to be increased to eliminate the suspicion of disadvantages can be specified by considering how many seats should be increased in which constituencies to make the regression coefficient of the explanatory valuable cease to be a significant value). This Court would be able to render a judgment indicating the method of overcoming an unconstitutional state in its reasons only after a suspicion of disadvantages is proven.

For the reasons described above, I think that the conditional constitutionality theory is just a stance that this Court should take and that the current disparity in the value of votes cannot be considered as being unconstitutional or in an unconstitutional state because a suspicion of disadvantages has not been proven in this case. Incidentally, conditions where a suspicion of disadvantages arises can be roughly divided into two. One of them is the case where residents in constituencies for which the number of members per voter is small equally suffer disadvantages, and the other is the case where people having certain political conviction suffer disadvantages, irrespective of a constituency in which each of them resides. The former disadvantages probably often arise in relation to an easily measurable event, such as the allocation of a grant or subsidy to a prefecture by the State. Therefore, a suspicion can be proven by traditional statistical techniques (as long as disadvantages actually exist). The latter disadvantages arise, for example, regarding the people who are not committed to a certain political conviction as a result of excessive reflection of such conviction on national politics in the case where many people having the relevant conviction reside in the areas of constituencies for which the number of members per voter is large. I think that a suspicion of disadvantages can also be proven in relation to such disadvantages by devising the methods of verification and analysis (as long as disadvantages actually exist).

The dissenting opinion stated by Justice HAYASHI Keiichi is as follows.

I cannot agree with the majority opinion for the reasons described below and thus determine that the Provisions on the Apportionment of Seats are unconstitutional.

1. First of all, as far as the basic issue of what level of the maximum disparity is permissible is concerned, I mostly maintain my opinion attached to the 2017 Grand Bench Judgment. That is, I demonstrated my determination that the situation where the value of votes of voters in a constituency reaches about three times as much as the value of votes of voters in another constituency is in an unconstitutional state by saying that I "have a reluctance to" clearly evaluate that the situation is constitutional in light of the one-person-one-vote principle and the principle of equality in the value of votes. I determine that the Election was also in an unconstitutional state because there is no significant difference between the 2016 Election and the Election in terms of the maximum disparity.

However, in the previous case, as I highly valued the Diet's response as also pointed out in the majority opinion and came to determine that the election was constitutional in conclusion, I used an understatement by saying that I "have a reluctance to" clearly evaluate that the election was in a constitutional state. However, this time, I cannot evaluate the Diet's response in the same manner, as I describe below, and I can confidently say that the Election was in an unconstitutional state.

2. Behind the majority opinion, there may be an idea that it is not necessary to point out the existence of extreme inequality just because the maximum disparity was around 3 to 1 on the grounds of the uniqueness of the House of Councillors (including the check function that it is expected to fulfill in relation to the House of Representatives and the system of election for half the members, etc.) and in light of the fact that the maximum disparity at the time of the inauguration of the House of Councillors was 2.62 to 1. However, I cannot help thinking that the idea that the disparity of 3 to 1 does not fall under extreme inequality in relation to equality, which is related to the fundamentals of the democratic representation system: specifically equality in the value of votes, is detached from common sense. For example, if the State makes the amount of money paid to residents in a prefecture three times as much as that to residents in another prefecture despite the fact that the purpose of the payments is completely the same, it will immediately lead to the criticism that such gap falls under impermissible inequality, let alone the value of one vote of the people, who are the sovereigns, in a national election, which is the fundamental element of the democratic representation system and secures the legitimacy of the Diet. It is obvious that a determination that such inequality cannot be considered as extreme is not accepted in the world unless there is a special ground under the Constitution. As far as I can see, apportionment of an even number of seats to each constituency, which is one of the factors of great disparity, per se also cannot be considered as a constitutional requirement, and there is no other special ground under the Constitution. In this manner, the disparity of 3 to 1 falls under extreme inequality that is not permissible under the Constitution. Therefore, it should be immediately corrected.

3.(1) On the other hand, the majority opinion states that "realization of the reform must be gradual in part," taking into account that "careful considerations are required" for the reform of the election system for the House of Councillors "in terms of its nature." Certainly, it is true that a certain period of time is required to make an important decision in a democracy, and in particular, overcoming of inequality in the value of votes means elimination of the vested rights of persons who have enjoyed large value of votes. Therefore, it is undeniable that the elimination involves difficult problems from a political perspective. Consequently, constitutionality of a decision may be also evaluated in consideration of the process and direction of making the decision.

I agreed with the majority opinion that the 2016 Election was constitutional in conclusion while evaluating it as having been in an unconstitutional state, just because I highly valued the Diet's efforts in reducing the maximum disparity from about 5 to 1 (this value was determined to be unconstitutional in the 2012 Grand Bench Judgment and the 2014 Grand Bench Judgment) to about 3 to 1 by taking measures, including the first-ever merger of constituencies, and I also comprehensively evaluated the processes and direction of the efforts, in addition to expectations for the fact that the Diet obliged itself to make more in-depth efforts to correct the disparity by providing in the Supplementary Provisions of the 2015 Amendment Act that "taking into consideration correction of the disparity … , etc., review is to be made continuously, aiming to definitely draw a conclusion on the fundamental reform of the election system" (in preparation for the Election).

(2) Regarding the Supplementary Provisions of the 2015 Amendment, it is only necessary to make "review" and reach a "conclusion," and the provisions can also be considered as having only an incidental meaning. However, I think that the Supreme Court purposely highly valued the same Supplementary Provisions while also taking that point into account. That is, the Supreme Court straightforwardly valued the fact that the Diet indicated the "direction and the strong will of the legislative branch toward further correction of the disparity" with a specified time limit, in addition to the actual outcome, despite existence of great political difficulties in reducing the disparity from about 5 to 1 to 3 to 1. If the fact that the maximum disparity was reduced to about 3 to 1 is all that is needed, it was not necessary to especially value this point, and it should be considered that behind the same judgment was the recognition that further in-depth correction of the disparity is necessary.

That is, the 2017 Grand Bench Judgment did not state that the election can be evaluated as constitutional immediately because the maximum disparity was reduced to about 3 to 1 but it evaluated the election as constitutional by "comprehensively" taking into account the direction and the strong will of the Diet toward making efforts for further correction of the disparity by the next ordinary election.

(3) However, as pointed out in the majority opinion, it is obvious that the Diet's efforts for the correction of the disparity after the 2016 Election were poor in the content in consideration of the facts that the Diet had promised to make review on the "fundamental reform" and reach a conclusion as its legal obligation and that the Supreme Court expected such efforts.

Certainly, there is actually a strong opposition to the merger of constituencies mainly in local regions, and in fact, constituencies suited to become subject to another merger cannot be easily found. In consideration of these points, the majority opinion states that the Diet can be considered to have "given consideration so as to maintain the direction" of the correction of the disparity in the 2015 Amendment based on the fact that it increased the number of seats for Saitama Prefecture by two while maintaining the existing merged constituencies and then indicated the evaluation that "it is impossible to determine that an attitude oriented to the correction of the disparity was lost." Then, the majority opinion concludes that on the whole, the Election cannot be considered as indicating the existence of extreme inequality to such an extent that it could raise a question of unconstitutionality.

(4) However, in my eyes, even though it is possible to say that the Diet held broad discussions on various reform plans this time, there are no signs indicating that the Diet came to make an in-dept review to agree on a specific option based on "fundamental reform" probably because it also kept in mind the determination in the 2017 Grand Bench Judgment that the election was constitutional or it took account opposition to the merger of constituencies. I must say that the review resulted only in an insignificant outcome, the 2018 Amendment. If the Supreme Court still determines the election was constitutional based on the result of the 2018 Amendment, for which the outcome of the review at the level of "fundamental reform" is equal to doing nothing from the perspective of correction of the disparity, it may be perceived as substantially giving up comprehensive evaluation of the outcome, including the "direction and the strong will of the legislative branch toward further correction of the disparity" as indicated in the 2017 Grand Bench Judgment, and valuing the maintenance of the disparity of about 3 to 1, thereby accepting this disparity as the "bottom figure" and sending a message that it is only necessary to maintain the status quo and prevent the great re-expansion of the disparity. There is the concern that this will result in stopping the Diet from making efforts for correction of the disparity in the future and make eternal the disparity of as large as 3 to 1.

From such perspective, regretfully, this time, I must consider that I cannot adopt the idea that the election was constitutional in conclusion though it was in an unconstitutional state.

4. For the reasons described above, I determine that the Provisions on the Apportionment of Seats are unconstitutional because the disparity in the value of votes at the time of the Election was in an unconstitutional state from the perspective of the maximum disparity and because the reasonable period for correction thereof is considered to have already passed. Based on this, the invalidity of the Election should originally be declared. However, I would like to take a stance of only making a declaration of unconstitutionality under the doctrine of judgment in consideration of circumstances for the public interest, taking into account the indications of the Diet's struggles and the fact that the declaration of invalidity may create difficult issues, such as how to consider the scope of invalidation.

The dissenting opinion stated by Justice MIYAZAKI Yuko is as follows.

Unlike the majority opinion, I think that the disparity between constituencies in terms of the value of votes under the Provisions on the Apportionment of Seats is in an unconstitutional state and that the Provisions on the Apportionment of Seats are thus unconstitutional. The reasons are as described below.

1. Concerning the framework for determination

I have no objection to the basic framework for determination concerning the constitutionality of the provisions on apportionment of seats in elections for the House of Councillors (members to be locally elected or members to be elected from constituencies), which has been indicated in the 1983 Grand Bench Judgment and subsequent cumulative Grand Bench judgments.

2. Concerning the election system provided by the Provisions on the Apportionment of Seats and its reasonableness

(1) From the perspective of the basic structure of the system, the election system provided by the Provisions on the Apportionment of Seats maintains a system designed to use a prefecture as a unit of constituency (hereinafter referred to as "prefecture-based constituency system) with respect to 43 constituencies that account for 95% of 45 constituencies in total (after introduction of the merger of constituencies) (the number of voters in these 43 constituencies accounts for 97.8% of the total number of voters at the time of the Election). Therefore, said election system can be considered as having the same basic structure as the election system before the 2015 Amendment. Since the first election of members of the House of Councillors to be elected from constituencies in 1947, including the period after the 2015 Amendment, the reason for using a prefecture as a unit of constituency has been explained as follows: consolidation of the will of residents by using a prefecture as a unit has significance because a prefecture is a unit of administration, etc. that has a long history as a regional unit.

With regard to this reason, the 2012 Grand Bench Judgment held as follows: the reason had appropriate reasonableness to the extent that a prefecture is a political unit as a unit of administration, etc.; however, under the situation where the Constitution does not require the use of a prefecture as a unit of constituency and it can rather be found that the inflexible use of a prefecture as a unit of constituency has caused the prolonged great inequality in the value of votes (at that time, the maximum disparity was around 5 to 1) as a result of ever-widening differences in prefectural population, the relevant mechanism itself needs to be reformed. Furthermore, the 2012 Grand Bench Judgment also held as follows: a legislative policy to fix the apportionment of seats for members of the House of Councillors for a longer period of time than for members of the House of Representatives that has been considered permissible in light of the provisions of the Constitution concerning the House of Councillors is no longer regarded as a legitimate reason for leaving great disparity in the value of votes unaddressed for tens of years despite the constant influx of population into urban areas. Based on these holdings, the 2012 Grand Bench Judgment determined that the election held under the election system subject to its determination was in an unconstitutional state. The 2014 Grand Bench Judgment made a decision to almost the same effect. The election systems that were subject to a determination in the 2012 Grand Bench Judgment and the 2014 Grand Bench Judgment were those before the 2015 Amendment, and the maximum disparity was 5.00 to 1 and 4.77 to 1, respectively. Incidentally, the following had already been pointed out in a report submitted by the Expert Committee on the Election System established under an advisory body to the President of the House of Councillors, the Council on the Reform of the House of Councillors, in October 2005, seven years before the 2012 Grand Bench Judgment: as long as the mechanism of an election system is designed to use a prefecture as a unit with respect to all constituencies, even if correction of the disparity is promoted through a measure to transfer seats of constituencies, it is considerably difficult to keep the disparity within 4 to 1.

The Election was the second election held after the adoption of the system of the merger of constituencies based on the 2015 Amendment that was made in response to the determinations in the 2012 Grand Bench Judgment and the 2014 Grand Bench Judgment. The maximum disparity in the first election was 3.08 to 1, and that in the second election, the Election, was 3.00 to 1. Before that, the maximum disparity had continued to significantly exceed 4 to 1, hovering around 5 or more to 1. Therefore, introduction of the merger of constituencies had the effect of reducing the maximum disparity to around 3 to 1.

(2) In separate opinions attached to the judgments of this Court in the past, over 10 justices expressed the opinion that the disparity of 2 or more to 1 or over 2 to 1 is extremely inequal from the perspective of equality in the value of votes. These opinions seem to have a commonality in their purport that inequality in the value of votes of 2 or more to 1 or over 2 to 1 is impermissible in light of social common sense in democratic societies. I also agree with that purport. In addition, in advanced countries' efforts to realize equality in the value of votes (as deviation seems to be more often used as an index than the maximum disparity, if converting deviation into the maximum disparity), a legal system is often created so as to ensure that inequality exceeding 2 to 1 does not arise. This fact is considered to indicate that it is also internationally accepted as the social common sense of citizens in democratic society that the value, 2 to 1, serves as an indication of extreme inequality. By contrast, the maximum disparity in the Election was 3.00 to 1 and far exceeded 2 to 1, and therefore, it indicates extreme inequality. I think that it is undeniable that the situation where the value of votes of voters in a constituency is one-third the value of votes of voters in a neighboring constituency indicates not merely inequality but extreme inequality under the rule of democracy dominated by the extremely strict rule of majority: at an assembly, a resolution is passed by a majority vote even if the difference between approval and disapproval is only one vote while a resolution is rejected even if the number of approving votes is only one vote shy of the majority.

(3) Extreme inequality indicated by the maximum disparity of 3.00 to 1 in the Election can also be explained from the perspective of comparison with the House of Representatives.

Until the reform of the election system for the House of Representatives in 1994, by which the single-seat constituency system was introduced to the House of Representatives, the election systems for the House of Representatives and the House of Councillors had adopted different methods. Since 1994, both houses have adopted similar election methods, and as a result, their election methods have been of the same nature, as also pointed out in the 2012 Grand Bench Judgment and the 2014 Grand Bench Judgment. In particular, election of members of the House of Representatives to be elected from single-seat constituencies and election of members of the House of Councillors to be elected from constituencies are similar to each other in that the apportionment of seats that serves as the basis for them is conducted on a prefecture-by-prefecture basis, and this point is a cause of the disparity in the value of votes for both houses.

Incidentally, both the 2012 Grand Bench Judgment and the 2014 Grand Bench Judgment held that the requirement for equality in the value of votes should not be taken more lightly just because the election in question is for members of the House of Councillors. Even at the time of these judgments, the criterion, less than 2 to 1, had already been provided in law with regard to the disparity in population in the election of members of the House of Representatives to be elected from single-seat constituencies. However, a law that specifically provides for the time and method of realizing the disparity in population of less than 2 to 1 for the election of members of the House of Representatives to be elected from single-seat constituencies was enacted through the amendments to the Act for Establishment of the Council on the House of Representatives Electoral District and the Public Offices Election Act, which were made in May 2016 and June 2017 during the period between the 2015 Amendment and the 2018 Amendment. Then, the maximum disparity between constituencies was lower than 2 to 1, specifically, 1.979 to 1, in the election for the House of Representatives subsequently held in October 2017. Furthermore, as a result of the aforementioned legal amendments, the maximum disparity is expected to be further reduced when the demarcation of constituencies based on the result of the 2020 population census comes into effect.

Comparing this situation with the situation of the House of Councillors, it is clear that the requirement for equality in the value of votes has continuously been taken far more lightly for the House of Concillors compared to the House of Representatives over a long period of time, despite similarity between those houses as mentioned above. Needless to say, even though the election systems for both houses are similar to each other, they are not completely the same. Therefore, it cannot be immediately said that the same maximum disparity must be realized for both houses. However, the maximum disparity of 3.00 to 1 in the Election can be considered to indicate extreme inequality even in comparison with the value that is being specifically achieved for the House of Representatives, the maximum disparity of less than 2 to 1, from the perspective of the 2012 Grand Bench Judgment to the following effect: equality in the value of votes is a constitutional requirement and the requirement for equality in the value of votes should not be taken more lightly just because the election in question is for members of the House of Councillors in light of the facts that both the House of Representatives and the House of Councillors are assemblies provided for in the Constitution and that members of both houses have a common constitutional position as the representatives of the people.

(4) Incidentally, the cumulative judicial precedents of this Court have held as follows: equality in the value of votes is not the sole and absolute criterion, but it must be realized in harmony with other policy purposes or grounds that the Diet is authorized to consider in determining an election system at its own discretion; consequently, as long as a specific decision made by the Diet can be found as a reasonable exercise of its discretion, such a decision cannot be judged to be unconstitutional even if the decision compromises equality in the value of votes to a certain extent. Therefore, it is necessary to consider, from this perspective, whether the discretion of the Diet is reasonable in determining the election system consisting of two merged constituencies from four prefectures and 43 prefecture-based constituencies. I think that the reasonableness of the Diet's discretion should be strictly determined as long as it is the issue of whether the Diet is permitted to take the constitutional requirement for equality in the value of votes lightly to the extent of accepting extreme inequality at its discretion.

On the premise of the aforementioned idea, I consider the reasonableness of the reason for the Diet's using a prefecture as a unit of constituency when determining the system consisting of two merged constituencies from four prefectures and 43 prefecture-based constituencies at its discretion after the 2015 Amendment. First of all, as mentioned in (1) above, the reason is that consolidation of the will of residents by using a prefecture as a unit has significance in the same manner as the reason for having used a prefecture as a unit of constituency before the same Amendment. However, as also pointed out in the 2012 Grand Bench Judgment, the Diet determined that an acceptable election system can be established even without using an administrative unit, prefecture, as a unit of consolidation of the will of the people for the four prefectures subject to the merger, while recognizing that use of a prefecture as a unit of constituency is not a constitutional requirement. Taking this into account, as long as the other 43 prefectures are also administrative units, it is rather reasonable to think that the Diet itself has not denied that the significance or necessity of consolidation of the will of the people on a prefecture-by-prefecture basis as a whole is not so strong and that such consolidation can be replaced by consolidation of the will of the people based on other units (merged constituencies fall under such units in the 2015 Amendment; however, not limited thereto). The more important point is the following fact: as the system consisting of two merged constituencies from four prefectures and 43 prefecture-based constituencies has the same basic structure as the election system determined in the 2012 Grand Bench Judgment in that a prefecture is inflexibly used as a unit of constituency for over 95% of constituencies, it potentially includes a mechanism that continuously causes extreme inequality in the value of votes arising from the disparity between prefectures in terms of population in the sense as pointed out in the same judgment over a long period of time; actually, extreme inequality in the value of votes, the maximum disparity of over 3 to 1, consecutively arose in two elections held under the same system. In light of the 2012 Grand Bench Judgment and as long as these facts are seen from the premise of my stance of considering the maximum disparity of 3 to 1 as indicating extreme inequality in the value of votes, it can be said that the aforementioned reason that consolidation of the will of residents on a prefecture-by-prefecture basis has significance is obviously not reasonable to the extent that extreme inequality in the value of votes can be justified.

Needless to say, there will be difficulty in reshaping procedures, organizations, personal relationships, etc. for election that have been created on the premise of the conventional election system that has been maintained over the decades in a practical and political sense in association with the fundamental change of the system. However, although such difficulty is a burden that is necessarily associated with the fundamental change of the system, it cannot be a reason in determining whether the Diet's discretion is reasonable, that is, whether or not the Provisions on the Apportionment of Seats is unconstitutional. In the end, I think that the Diet's discretion in having decided to use a mechanism consisting of two merged constituencies from four prefectures and 43 prefecture-based constituencies cannot be considered reasonable enough to justify the Diet's taking the constitutional requirement for equality in the value of votes lightly to the extent of accepting extreme inequality, just because consolidation of the will of the people on a prefecture-by-prefecture basis is not conducted in all prefectures and it only has some significance in relevant prefectures.

In addition, regarding the House of Councillors, it is said that the maximum disparity becomes large as it is necessary to apportion an even number of seats to each constituency due to a constraint based on the constitutional requirement for election for half the members every three years. However, extreme inequality in the value of votes, the maximum disparity of 3 to 1, does not arise only because of the existence of that constitutional requirement. The cause of extreme disparity in the value of votes over a long period of time existed rather in the inflexible use of a prefecture as a unit of constituency, as pointed out in the 2012 Grand Bench Judgment. If that point is fundamentally altered, it becomes possible to design an election system in a manner that the aforementioned constraint based on the constitutional requirement does not cause the requirement for equality in the value of votes to be taken lightly, as pointed out in multiple separate opinions attached to the cumulative judicial precedents of this Court and the separate opinions of other justices attached to this judgment. Therefore, the constitutional requirement for election for half the member every three years also cannot be considered as the basis for the reasonableness of the Diet's discretion in relation to the fact of causing the disparity of 3 to 1.

For the reasons described above, I think that the disparity between constituencies in terms of the value of votes indicated the existence of extreme inequality to such an extent that it could raise a question of unconstitutionality at the time of the Election.

3. Concerning evaluation of the 2018 Amendment

On the other hand, the majority opinion states that the 2018 Amendment gave consideration so as to maintain the direction of the 2015 Amendment based on the fact that the maximum disparity was corrected, albeit only slightly, as a result of increasing the number of seats for members to be elected from constituencies by two and apportioning those seats to Saitama Prefecture, for which the number of voters per member was the largest in 2016 Election, and aggressively evaluates the 2018 Amendment as a circumstance that leads to the conclusion that the Election was constitutional. However, I cannot agree with that evaluation for the following reasons.

The election system under the 2018 Amendment is the system consisting of two merged constituencies from four prefectures and 43 prefecture-based constituencies created through the 2015 Amendment. It is certainly undeniable that adoption of the merger of constituencies through the 2015 Amendment can be considered as the first step to the direction of creating constituencies by using a broader area (merged constituency) than a prefecture as unit. On the other hand, the election system after the 2015 Amendment has the same basic structure as the election system before the same amendment in that it also adopts the prefecture-based constituency system. Paying attention to the fact that its basic structure is maintained with respect to over 95% of constituencies (97.8% of voters), it is also possible to think as follows: under the constraint that the maximum disparity cannot be reduced to within 4 to 1 only by reapportioning seats among constituencies in the conventional election system, the system of creating two constituencies by merging two prefectures each, only targeting prefectures with the smallest populations, was adopted as a measure for reducing the maximum disparity to 3 to 1 by increasing the size of constituencies with the smallest populations in response to the requirement for the reduction of the maximum disparity with the basic structure maintained. In that case, this adoption of the merger of constituencies can also be evaluated as being hardly oriented toward the fundamental reform of the prefecture-based constituency system. From such standpoint, only the fact that the direction of the 2015 Amendment was maintained does not mean that the fundamental reform required in the 2012 Grand Bench Judgment was made.

Next looking at the content and background, etc. to the 2018 Amendment, for members to be elected from constituencies, [i] the system consisting of two merged constituencies from four prefectures and 43 prefecture-based constituencies created through the 2015 Amendment was maintained, and [ii] the number of seats for members to be elected from constituencies was increased by two and those seats were apportioned to Saitama Prefecture. However, the fact mentioned in [i] means that the election system that has the same basic structure as the election system before the 2015 Amendment was maintained without making any change to the basic structure, the prefecture-based constituency system. The amendment mentioned in [ii] cannot avoid giving the impression of a stopgap measure for reducing the maximum disparity that is designed to quicky reduce the maximum disparity to get it close to less than 3 to 1 by increasing the number of seats only in Saitama Prefecture, where the maximum disparity of 3.08 to 1 was recorded in the 2016 Election, because it is impossible to reduce the maximum disparity only by reapportioning seats among constituencies. In addition to the fact mentioned in [i] and the amendment mentioned in [ii], amendment was made to increase the number of seats for members to be elected by proportional representation by four and thereby establish a specific frame. However, this cannot be considered as leading to the reform and change of the basic structure, the prefecture-based constituency system. Furthermore, no sign of aiming at consensus to expand the merger of constituencies as a method for reducing the disparity in the future was seen in discussions over the 2018 Amendment at the Diet, taking into account the facts that there had been a torrent of negative discussion on the merger of constituencies and that the option of increasing merged constituencies were not discussed at all.

In such manner, I cannot find out any factor indicating that the fundamental reform of the system required in the 2012 Grand Bench Judgment had been made or was being made in the 2018 Amendment. If any amendment indicating the potential that the maximum disparity can be made infinitely closer to 1 to 1 from around 3 to 1, albeit gradually, was made in the 2018 Amendment, I could have made an evaluation that the 2015 Amendment established a foothold for the fundamental reform of the system and that the 2018 Amendment took a further step toward resolution of the problem pointed out in the 2012 Grand Bench Judgment, while seeing hope in said potential. However, in actuality, I must say that the 2018 Amendment fits in with the evaluation that it was an amendment that only intended to keep the maximum disparity less than 3 to 1 (according to estimate at the time of the amendment, the maximum disparity was 2.99 to 1; however, the actual value was not less than 3 to 1) without changing the basic structure, the prefecture-based constituency system. Looking at the period before the Election after the 2018 Amendment, I cannot find any circumstance based on which this evaluation should be changed. The maximum disparity of 3.00 to 1 certainly indicates a reduction of the maximum disparity by 0.08 compared to the 2016 Election. However, from my standpoint of considering the maximum disparity of 3 to 1 as indicating extreme inequality in the value of votes, based on this level of reduction of the disparity, I cannot valuate the 2018 Amendment as having maintained the direction of the fundamental reform of the system that was left behind in the 2015 Amendment.

Turning to a different point and verifying whether the Diet has performed its obligation as provided in Article 7 of the Supplementary Provisions of the 2015 Amendment Act that it imposed on itself by providing that "review" is to be made continuously, aiming to definitely reach a "conclusion" on the fundamental reform of the election system in preparation for the next ordinary election (referring to the Election), the Diet is certainly recognized as having made "review." In light of the timing of the 2018 Amendment, it is natural to consider the same amendment as a "conclusion" reached as a result of the "review" as referred to in the aforementioned Supplementary Provisions. Therefore, the aforementioned obligation set forth in the Supplementary Provisions was performed in that sense. However, the "conclusion," the 2018 Amendment, cannot be considered as the outcome of the fundamental reform of the prefecture-based constituency system as required in the 2012 Grand Bench Judgment, and it basically maintains and keeps the basic structure of the system and is thus hardly considered to have the content that meets the requirement mentioned in the same judgment, as mentioned above. Although the Diet can be considered to have made "review" and reached a "conclusion," the "conclusion" reached by the Diet can be evaluated only as one that avoid facing the problem without fully tackling the fundamental reform of the structure of the system as pointed out in the 2012 Grand Bench Judgment.

As described above, in terms of evaluation of the 2018 Amendment, I have a different opinion from the majority opinion to the effect that the amendment had not lost the attitude oriented to the correction of the disparity, and the conclusion mentioned in 2. above is not affected by the 2018 Amendment.

4. Concerning the continuity of the issue of the structure of the system as pointed out in the 2012 Grand Bench Judgment and the limits of the Diet's discretion

(1) The election system after the 2015 Amendment has the same basic structure as the election system before the same amendment, and extreme inequality indicated by the maximum disparity of 3 or more to 1 consecutively arose in two elections that were already held. In addition, inequality in the value of votes does not arise only in constituencies with the largest number of voters per member, and in constituencies with relatively large populations (for example, constituencies pertaining to nine prefectures with populations of 5,000,000 or more), a tendency of the disparity being close to the maximum disparity is also prominently recognized, and extreme inequality in the value of votes continuously exists there over a long period of time. This phenomenon has not changed throughout the period before and after the 2015 Amendment. Comprehensively observing this state, the Diet should be evaluated as having maintained a mechanism that had caused extreme inequality over several decades without reasonable grounds and thereby having produced the result of continuously causing extreme inequality in the value of votes to such an extent that it could raise a question of unconstitutionality (unconstitutional state).

This just means that the Diet avoided facing the problem by making legal amendments without showing the direction of eliminating inequality in the value of votes: which had continued since several decades ago, on the occasion of the 2018 Amendment, and thereby made that inequality further continue to exist. Based on such idea, it is difficult to draw a conclusion that it is safe to prolong decades-long extreme inequality in the value of votes. In addition, from the perspective of balance with the fact that measures to meet the requirement for equality in the value of votes for the House of Representatives have already been enshrined into law, I must say that it is an urgent necessity to make it possible to hold election for members of the House of Councillors in a proper form as early as possible because of the following reason: on the premise of the current state and trend of distribution of population in prefectures, inflexible use of a prefecture as a unit of constituency has adverse effects on equality in the value of votes, not once and in the short term, but with certainty and for a long period of time; in particular, voters in constituencies with large populations are continuously deprived of their rights, but it is impossible to compensate for (have them exercise) their deprived rights later.

(2) The 2012 Grand Bench Judgment clearly pointed out that the Diet has the obligation to overcome extreme inequality in the value of votes (unconstitutional state). In addition, in the 2009 Grand Bench Judgment rendered prior to the same judgment, this Court also made a holding that points out the same issue as pointed out in the 2012 Grand Bench Judgment and encourages the Diet to make review thereon. Looking further back, it was also clearly pointed out in the aforementioned report of the Expert Committee in 2005 that there are limitations to the reduction of the maximum disparity as long as the prefecture-based constituency system is maintained. Therefore, it can be said that the Diet and its members could fully predict the following points based on the content of the same report: there are limitations to the significant reduction of the maximum disparity as long as an election system that has the same basic structure as conventional system is maintained; a stopgap measure at the level of slightly increasing the size of constituencies with the smallest populations through introduction of two merged constituencies from four prefectures cannot fundamentally resolve the problem. Based on these, it is safe to say that the Diet had been able to sufficiently recognize what it should do before the Election during that period, based on the 2012 Grand Bench Judgment at the latest. It is considered that the Diet imposed the obligation to make review on itself in Article 7 of the Supplementary Provisions of the 2015 Amendment Act for that reason.

Based on the above, I think that the fact that the Diet maintained an unconstitutional state that had continued to exist over several decades without making legal amendments to correct the unconstitutional state during the period from the rendering of the 2012 Grand Bench Judgment to the Election, which was held about seven years later, should inevitably be evaluated as going beyond the limits of the Diet's discretion pertaining to the election system for the House of Councillors.

(3) Comparatively speaking, the following idea may also be possible: the maximum disparity ceased to exceed 4 to 1 (or extreme inequality indicated by the maximum disparity of around 5 to 1 as pointed out in the 2012 Grand Bench Judgment ceased to arise) owing to adoption of the merger of constituencies through the 2015 Amendment; therefore, extreme inequality in the value of votes as pointed out in the 2012 Grand Bench Judgment was overcome once there; and with regard to the election system adopted after the 2015 Amendment, it is impossible to determine that it went beyond the limits of the Diet's discretion without seeing whether the maximum disparity continues to be around 3 to 1 under the system consisting of two merged constituencies from four prefectures and 43 prefecture-based constituencies over a certain long period of time, if not over several decades. However, I cannot agree with that idea.

The reason therefor is as follows: on the premise of my evaluation of the 2018 Amendment as mentioned in 3. above and based on my opinion as mentioned in 2. above that the maximum disparity of 3 to 1 indicates extreme inequality in the value of votes and that no reasonable ground for justifying the Diet's taking the constitutional requirement for equality in the value of votes more lightly to the extent of accepting the maximum disparity of 3 to 1 with regard to the fact that the Diet established an election system pertaining to the Provisions on the Apportionment of Seats; even if the maximum disparity of around 5 to 1, which had arisen before the 2015 Amendment, was reduced to around 3 to 1 after the same amendment, the reason for consolidation of the will of the people on a prefecture-by-prefecture basis is not reasonable at either of those times (before the same amendment, for the ground held in the 2012 Grand Bench Judgment; after the amendment, for the ground mentioned in 2.(4) above); therefore, I must determine that the fact remains that extreme inequality in the value of votes continues to exist without any reasonable grounds.

(4) Incidentally, there may be an issue of whether the fact that the 2016 Election was determined to be constitutional in the 2017 Grand Bench Judgment can be a legitimate ground for suspending "review" made by the Diet on the issue as pointed out in the 2012 Grand Bench Judgment. However, it should be said that the following fact can be easily read from the text of the judgment: the 2017 Grand Bench Judgment neither made a determination that the 2016 Election was constitutional on the grounds that the issue as pointed out in the 2012 Grand Bench Judgment was resolved through the 2015 Amendment nor made a determination that the question of unconstitutionality does not exist with regard to the disparity in the value of votes if the maximum disparity is only around 3 to 1; and it was to the effect that this Court expects the Diet to continue the fundamental reform of the prefecture-based constituency system until the Election through further discussions. Therefore, I think that existence of the 2017 Grand Bench Judgment is not a legitimate ground for suspending review.

5. Conclusion

For the reasons described above, I think that the following fact should inevitably be evaluated as going beyond the limits of the permissible exercise of discretion by the Diet and that the Provisions on the Apportionment of Seats as a whole had violated Article 14, paragraph (1) of the Constitution: despite the fact that the disparity between constituencies in terms of the value of votes under the Provisions on the Apportionment of Seats still cannot be considered to have escaped from the decades-long extreme inequality at the level recognized as absolutely impossible to ignore, in light of the importance of the constitutional requirement for equality in the value of votes, the Diet maintained the prefecture-based constituency system with respect to 43 constituencies, accounting for over 95% of the total number of constituencies (the number of voters in those constituencies account for 97.8% of the total number of voters) in the 2018 Amendment without any reasonable grounds and thereby avoided facing the problem and had left extreme inequality in the value of votes in a state where it would continue to exist.

However, at present, I consider it reasonable not to invalidate the Election but to dismiss the claim for a judgment to the effect that the Election is to be invalidated under the doctrine of judgment in consideration of circumstances for the public interest and to only declare the illegality of the aforementioned election in the main text of the judgment in respect for the mechanisms of the Constitution, which state that legal amendment by the Diet's resolution is required for the reform of the election system in the end.

Finally adding a remark, although I have no objection to the point that a prefecture is a useful unit of administration, various changes in society and diversification of the values of people are not occurring on a prefecture-by-prefecture basis but are expanding without regard to the boundaries between administrative districts. This phenomenon is being increasingly accelerated by the dissemination of the Internet. Nobody can deny that it has become increasingly more necessary than before for the State to consolidate the will of the people in depth from diversified standpoints. I strongly hope that the Diet will promptly carry out the fundamental reform of the election system not only from the members' standpoint but also from the standpoint of the people: who exercise the right to vote, so as to sufficiently meet the constitutional requirement for equality in the value of votes from a higher perspective.

The dissenting opinion stated by Justice UGA Katsuya is as follows.

Unlike the majority opinion, I think that I unfortunately have to say that the Provisions on the Apportionment of Seats are unconstitutional for the following reasons.

1. Right to vote as the basis of the sovereignty of the people

It goes without saying that the Constitution requires equality in the content of the right to vote, that is, equality in the influence of voters' votes on the election of a member. In addition, this requirement for equality is extremely strong, and unlike the case of a qualification system in which different treatment according to ability is justified, one person one vote must be equally guaranteed for persons who are above the prescribed age, irrespective of their knowledge of politics and social experience, etc. This is because equality in the right to vote forms the basis of the sovereignty of the people and democracy. Therefore, I think that the issue of equality in the value of votes must be liable to a strict judicial review and that deviation from the principle of equality in the right to vote is not permitted unless it is truly unavoidable. If the value of one vote differs depending on the place of residence, it substantially means admitting multiple elections based on the place of residence. This violates the principle of equality referred to in Article 14, paragraph (1) of the Constitution and also raises the issue of violation of Article 15, paragraph (1) of the Constitution, that is, infringement of the right to vote that includes equality.

2. Diet's legislative discretion

For the House of Councillors, the Diet certainly has legislative discretion to determine the number of seats, the types of election to be used (a national constituency-based election or election using constituencies, or combination of these types of election), the way of apportioning seats in the case of using a combination of election types, whether to hold a national constituency-based election by proportional representation, and the unit to be used as a constituency. The Diet can establish an election system that differs from that of the House of Representatives for the House of Councillors by exercising its legislative discretion so as to ensure that the significance of the bicameral legislative system can be exerted. In that sense, it may be said that the Diet has a wide range of discretion. However, as this legislative discretion of the Diet is given within the framework of the Constitution, it is on the premise of the best effort to make the value of one vote equal as far as possible. Therefore, it is considered that equality in the value of one vote is not just one of the factors to be considered when it is comprehensively considered with various other factors but constrains the legislative discretion as a top consideration.

3. Diet's accountability

In light of the fact that the right to vote is an extremely important right that forms the basis for the sovereignty of the people, the Diet must design an election system while considering the state of no disparity in the value of one vote as a default. If there are unavoidable circumstances where the disparity arises in the value of one vote due to technical or time constraints, the Diet takes accountability therefor, and if the Diet gives no reasonable explanation, the relevant situation must be considered as being in an unconstitutional state.

Then, in terms of the question of whether the Diet has fulfilled accountability for the existence of unavoidable circumstances pertaining to the existence of the remaining considerably great disparity in the value of one vote in the election held based on the 2018 Amendment Act, I must probably say that the Diet has not indicated any ground for justifying inequality in the value of the right to vote, which causes a situation where one person is substantially regarded as having three votes, because although I can understand that it is politically difficult to merge constituencies, existence of political difficulty does not become a justification for sacrificing equality in the right to vote provided in the Constitution and as mentioned later, there are also methods of achieving improvement toward equality in the right to vote other than the merger of constituencies and it seems to be possible even to achieve improvement toward equality in the right to vote while maintaining the existing constituencies.

4. Absence of grounds for more loosely accepting the disparity in the value of one vote in the election system for members of the House of Councillors than that in the election system for members of the House of Representatives

The Constitution certainly admits the superiority of the House of Representatives with respect to certain matters under the bicameral legislative system. However, on the other hand, the Constitution sets a longer term of office for members of the House of Councillors than those of the House of Representatives, and the House of Councillors is never dissolved. Therefore, at the House of Councillors, deliberations can be held on a stabler and longer-term basis than the House of Representatives. However, this fact is not directly related to the issue of equality in the value of one vote. Even considering the fact in real terms, it causes occurrence of the "divided Diet," and if a bill passed by the House of Representatives is rejected by the House of Councillors, it cannot be enacted unless it is passed again by a majority of two-thirds or more of the members present in the House of Representatives. Therefore, the House of Councillors has large power. Consequently, the aforementioned difference between the systems of the House of Representatives and the House of Councillors seems not to become a ground for justifying the greater disparity in the value of one vote in election for members of the House of Councillors than that in election for members of the House of Representatives.

In addition, the Constitution also provides that election takes place for half the members of the House of Councillors every three years. However, these provisions also cannot be considered as obliging the apportionment of an even number of seats to each constituency without fail in the case of establishing constituencies. For example, the constitutional requirement seems to be fulfilled if it is ensured that election takes place for half the total number of seats for members of the House of Councillors every three years by apportioning half the total number of seats for members of the House of Councillors to election by proportional representation, which is now held based on a national constituency, and the remaining half to election using constituencies, and by holding election by proportional representation alone for a certain year and then holding election using constituencies alone at the time of the next election three years later. Therefore, it is considered unnecessary to apportion an even number of seats to each constituency.

Given that a system is designed so that single-seat constituencies are created by making it possible to apportion an odd number of seats to constituencies and election using constituencies is held once every six years for single-seat constituencies and once every three years for multiple-seat constituencies, the opportunities for elections using constituencies becomes fewer in single-seat constituencies than multiple-seat constituencies. However, this is the result of election in proportion to population and probably cannot be considered as unconstitutional.

As mentioned above, the system of election for half the member of the House of Councillors also seems not to become a ground for loosely accepting the disparity in the value of one vote under the election system for members of the House of Councillors compared to that under the election system for members of the House of Representatives. In addition, the aforementioned introduction of constituencies with an odd number of seats is also possible under the current demarcation of constituencies.

As mentioned above, although the Diet has the legislative discretion to establish an election system that differs from the House of Representatives for members of the House of Councillors in light of the significance of the bicameral legislative system, it seems to be difficult to find a constitutional ground for considering that the requirement for equality in the value of votes may be taken more lightly for the House of Councillors than the House of Representatives.

5. Difficulty of justifying the disparity in the value of votes on the grounds of the necessity of regional representatives

The existing constituencies for the House of Councillors are basically on a prefecture-by-prefecture basis though some constituencies have been merged. In Japan, prefectures are certainly historically important regions and there is no doubt that residents have a sense of belonging to a prefecture from political and social perspectives. However, an election system using a prefecture as a unit of constituency was introduced in the House of Councillors in 1947 not for the purpose of electing prefectural representatives but because it was judged appropriate to adopt prefecture-based constituencies as a means for securing members who are acquainted with actual regional conditions under the situation of the time, where both means of transportation and means of information and communication were far behind what they are today. Today, when both means of transportation and means of information and communication have dramatically developed, it has become far easier for the Diet to investigate the actual regional conditions, and it can thus be said that it has practically become less necessary to maintain prefecture-based constituencies. In addition, it seems to be difficult to find a ground for putting emphasis on prefectural representatives in the Constitution.

Originally, the terms "prefecture" and "municipality" are not used in the Constitution though the term "local public entity" is used therein. It is only at the level of the Local Autonomy Act that ordinary local public entities are defined as prefectures and municipalities. For this reason, it is now generally considered possible to abolish prefectures and introduce the Doshu system (local administrative system in which Do and Shu are established as administrative districts) under the current Constitution, and the Local Government System Research Council of the Cabinet Office has also submitted a report toward promotion of the Doshu system.

In this manner, it is not only impossible to find any ground for putting emphasis on prefectural representatives at the House of Councillors in the Constitution but I also think that it is difficult under the Constitution to justify the disparity in the value of one vote in order to have the House of Councillors more generally have regional representative-like character. That is, as Article 43, paragraph (1) of the Constitution provides that "Both Houses shall consist of elected members, representative of all the people," not only members of the House of Representatives but also those of the House of Councillors must not be regional representatives but be representative of all the people. These provisions not only indicate a code of conduct that the members of the Diet should act not as regional representatives but as the representatives of all the people as long as they were once chosen in an election but is also considered to mean that it is not permitted to sacrifice balance in the value of one vote in designing an election system in order to give regional representative-like character to all or some members, as held in 2010 (Gyo-Tsu) No. 207, the judgment of the Grand Bench of the Supreme Court of March 23, 2011, Minshu Vol. 65, No. 2, at 755 (hereinafter referred to as the "2011 Grand Bench Judgment"). The same judgment is related to the rule of reserving one seat per prefecture for the House of Representatives. However, the following holding is also considered to be applicable to election using constituencies for the House of Councillors: "It can hardly be justified as being reasonable to cause, only for the purpose of coping with problems arising from regional circumstances, inequality in the value of votes between voters in particular regions (prefectures) and those in other regions (prefectures)." Regarding the defendant's argument that it is necessary to make it easier for the opinions of the representatives of underpopulated areas to be delivered to national politics by using a prefecture as a unit of constituency, it is true that measures for underpopulated areas are certainly important issues of national politics, and I can agree with that argument in the point that it is not safe to discount underpopulated areas only because the areas are thinly populated. However, importance of measures for underpopulated areas seems not to become a reason for increasing the value of votes in underpopulated areas, as stated in the following holding in the 2011 Grand Bench Judgment: "members …are, irrespective of in which regions their constituencies are located, required to take part in national politics as representatives of all the people. Consideration to relatively less populated regions is a matter that these members should take into account when making laws and performing other duties from a nationwide perspective in the course of carrying out such political activities." In addition, although there are various minorities in the world and it is important to make it easier for the opinions of such minorities to be delivered to national politics, as long as it is not permitted to increase the value of one vote of such persons for that purpose, it seems to be difficult to justify adding extra value to the value of one vote of residents in underpopulated areas for the purpose of making it easier for the opinions of residents in underpopulated areas to be delivered to national politics though that purpose is the important issue of national politics.

Incidentally, I also do not deny taking into account the significance and substance, etc. of each prefecture serving as a political unit as one factor, as held in the 2017 Grand Bench Judgment. However, this is just under the condition that it is compatible with the requirement for equality in the value of votes, and I consider that it is impossible to put emphasis on the requirement for prefectural representatives even by impairing the requirement for equality in the value of votes.

6. Concerning an unconstitutional state

For the reasons described above, I regret to say that the Provisions on the Apportionment of Seats were in an unconstitutional state as long as there was still inequality that could not be ignored in the value of votes and the Diet had not given a reasonable explanation that the inequality is unavoidable. Incidentally, for confirmation, I would like to add a remark that making a determination that the Provisions on the Apportionment of Seats were in an unconstitutional state neither denies the Diet's legislative discretion nor forces the Diet to adopt a specific option. As adoption of election using constituencies is not required under the Constitution, it is also possible to adopt a system in which all the members are elected in national constituency-based election by proportional representation. Even if election using constituencies is maintained, it is also possible to adopt block constituencies, increase merged constituencies, increase the total number of seats for members to be elected from constituencies and increase the number of seats apportioned to constituencies with small values of one vote (in addition to a method designed to increase the total number of seats for members of the House of Councillors, there is also a method designed to decrease the total number of seats for members to be elected by proportional representation and transfer the reduced seats to the total number of seats for members to be elected from constituencies), and establish single-seat constituencies, as well as combining these methods as appropriate. Therefore, I think that the scope in which the Diet can exercise its legislative discretion toward realization of equality in the value of votes as required under the Constitution is by no means narrow.

7. Fact that the reasonable period for the correction of an unconstitutional state has passed

I wonder if, in a suit to seek invalidation of an election, the election may be determined to be unconstitutional, irrespective of passage of the reasonable period, if it is in an unconstitutional state while passage of the reasonable period must become an issue in determining the existence or absence of negligence in a suit to seek state redress. However, aside from this point, I also state below the reasons for considering the election in question as unconstitutional based on the theory of the reasonable period, which is a framework for determination established by this Court.

In 2008 (Gyo-Tsu) No. 209, the judgment of the Grand Bench of the Supreme Court of September 30, 2009, Minshu Vol. 63, No. 7, at 1520, this Court pointed out that it is difficult to significantly reduce the maximum disparity as far as the existing mechanism of the election system is maintained and that it cannot be denied that it is necessary to reform the mechanism itself. This Court then held as follows: in light of the facts that an election system that can properly reflect the will of the people is the foundation of democracy and that equality in the value of votes is a constitutional requirement, it is desired that the Diet should promptly set out to carry out an appropriate study on this issue, while fully taking into consideration the importance of equality in the value of votes. In addition, in the 2012 Grand Bench Judgment, this Court admitted that inequality in the value of votes is in an unconstitutional state and held that legislative measures designed to reform the prefecture-based election system should be taken in order to correct inequality as soon as possible because such level of inequality raises the question of unconstitutionality. On the other hand, in the 2017 Grand Bench Judgment, this Court held that the provisions on the apportionment of seats on which the 2016 Election held under the 2015 Amendment Act relied are not in an unconstitutional state. The issue is how to understand the relationship between these judgments.

I think that the Diet could recognize that an unconstitutional state was not overcome only by the reduction of the maximum disparity between constituencies in terms of the value of votes to about 3 to 1 through the 2015 Amendment, irrespective of the conclusion of the 2017 Grand Bench Judgment. This is because the 2017 Grand Bench Judgment first pointed out not only the fact that the maximum disparity between constituencies in terms of the value of votes was reduced to around 3 to 1 but also the fact that the Supplementary Provisions of the 2015 Amendment Act provided that review is to be made continuously, aiming to definitely reach a conclusion on the fundamental reform of the election system in preparation for the next ordinary election and thereby indicated the direction and the strong will of the legislative branch toward further correction of the disparity in the value of votes in the future, and then held that the provisions of the same Supplementary Provisions were not in an unconstitutional state (incidentally, as it is also clear in 1. or 6. above, I myself am not aligned with the position of taking into account the strong will of legislative branch to make efforts for the future in determining the constitutionality of the provisions on the apportionment of seats). These Supplementary Provisions of the 2015 Amendment Act indicates that the Diet itself had recognized at that time that the 2015 Amendment alone was not sufficient to fully meet the constitutional requirement for equality in the value of votes, and it is probably necessary to consider that based on the same idea, the Supreme Court also held that the provisions on the apportionment of seats were not in an unconstitutional state by taking into account the expression of aforementioned Diet's strong will as an important factor to be considered.

Looking at the 2018 Amendment that was actually made, introduction of a specific frame system for members to be elected by proportional representation has significance as a response to complaints about the maintenance of the existing merged constituencies and is not intended to reduce the disparity in the value of one vote. The Diet only increased the number of seats in the Saitama Prefecture by two toward realization of equality in the value of votes. As far as the disparity was only slightly reduced by this measure, the Diet should have also been able to recognize that the "fundamental reform of the election system" as referred to in the Supplementary Provisions of the 2015 Amendment Act had not been realized through the 2018 Amendment.

Based on the aforementioned circumstances, the Diet should have been able to recognize that an unconstitutional state cannot be overcome without significantly reducing the maximum disparity in the value of votes in election for members of the House of Councillors since the time of the 2012 Grand Bench Judgment at the latest. In light of the aforementioned circumstances, I think that the 2017 Grand Bench Judgment also does not become a ground for justifying the alteration of such recognition.

It is recognized that it is politically considerably difficult to fundamentally reform the election system that has been established since the inauguration of the House of Councillors and that the Diet has not stood by fecklessly but has continued enthusiastic discussions by establishing the Expert Committee on the Election System under the Council on the Reform of the House of Councillors even after the 2016 Election. However, taking into account the fact that the goal that the Diet set on itself in the Supplementary Provisions of the 2015 Amendment Act: the fundamental reform of the election system toward the next ordinary election, was not realized in the end, it must be said that the reasonable period has already passed. Therefore, regretfully, I must say that the Provisions on the Apportionment of Seats are unconstitutional.

8. Effect of the Election

As long as the Provisions on the Apportionment of Seats are determined to be unconstitutional, the Election held under the Provisions of the Apportionment of Seats is, in principle, to be invalidated in light of Article 98, paragraph (1) of the Constitution. However, I think that a suit against the disparity in the value of one vote under the provisions of Article 204 of the Public Offices Election Act substantially created a special suit on constitutionality as case law by formally using the provisions of the same Article because it is unreasonable if there is no path to judicial remedy for the right to vote, which is a right that forms the basis of the sovereignty of the people, though such suit is not planned to be filed under the same Article. Therefore, with regard to the method of a judgment, a flexible determination seems to be exceptionally permitted in light of the division of roles between the judicial branch and the legislative branch, unlike in general cases.

There may be an interpretation that makes it possible to give a response that does not cause the confusion of national politics even with the rendering of a judgment invalidating the election, and some proposals have already been presented in the separate opinions of the justices of this Court. At present, various discussions are being held on such issues as what scope of members are to lose their positions, but it cannot be necessarily said that discussions on this point have been sufficiently accumulated even in the academic world. It must still be too early to go so far as hold that the election is invalid beyond declaration of unconstitutionality of the election and render a judgment to leave the issue to the Diet's response under such circumstances. I consider it appropriate for this Court to render a judgment only declaring unconstitutionality at this time with expectations regarding the Diet's response, and then render a judgment invalidating the election while indicating the method of responding to the judgment when it becomes clear that such judgment declaring unconstitutionality is no longer effective.

Justice OTANI Naoto

Justice IKEGAMI Masayuki

Justice KOIKE Hiroshi

Justice KIZAWA Katsuyuki

Justice KANNO Hiroyuki

Justice YAMAGUCHI Atsushi

Justice TOKURA Saburo

Justice HAYASHI Keiichi

Justice MIYAZAKI Yuko

Justice MIYAMA Takuya

Justice MIURA Mamoru

Justice KUSANO Koichi

Justice UGA Katsuya

Justice HAYASHI Michiharu

Justice OKAMURA Kazumi