Making the Japanese House of Councillors’ elections 4 times more fair….

There has been one issue where the lack of political killer instinct of the DPJ (and to a lesser degree other parties such as Komeito and Your Party that won seats in the larger multi-member constituencies in the recent Upper House election) has been quite puzzling. On a number of occasions various Japanese courts have ruled that the 一票の格差 (ippyou no kakusa) – the relative difference between one constituency and another in terms of the number of votes it takes to elect one candidate to the House of Councillors and/or Representatives – in Japanese elections has been unconstitutional. While no change should be expected under the LDP government which had generally benefited from rural votes essentially being worth more than urban votes, it was curious that the DPJ at least did not set the ball rolling on such reform sometime earlier than now on this issue, perhaps by setting up a parliamentary commission to look at the issue. After all, it is unconstitutional for one and any changes likely benefit the DPJ for another.  It is worth considering that electoral changes in the other interlude between LDP regimes (1993-94) are often cited as being crucial for the DPJ becoming the major party after the 2009 elections, even if 15 years delayed.

Nevertheless, the government has finally got around to looking at the issue at least for the Upper House – perhaps a recent statement from the Tokyo High Court, in ruling on the ippyou no kakusa’s unconstitutionality, that tinkering with the system was not going to cut it anymore, was useful here. The Asahi reports (ja) that the DPJ is going to propose a plan to be considered next spring  that attempts to equalize the vote discrepancy throughout the country by ensuring no one area has an ippyou no kakusa of more than 1.2. They will do this by setting up 11 multi-member constituency blocks spanning the country(image here Japanese but pretty straightforward), with the intention that such as system be put in place by the 2013 election. There will be no requirement for each prefecture to be represented by at least one member of parliament at every election and proportional representation will be abolished. This will also be accompanied by a reduction of seats from 242 to 200.1

This is a similar plan to one that I discussed in an earlier post as put forward by Takenaka Harukata. Takenaka, in considering how to make the Upper House more democratic, also very sensibly focused on the “functions”of the respective houses in terms of how they fit within the policy making process (rather than being obsessed with the “numbers” of MPs as some Japanese politicians have been). Suggesting that the Upper House had become no more than a disruptive clone of the Lower House, he suggested that the Upper House should  provide for a more pluralistic politics, rather than being a site for the continuation of partisan conflicts. In terms of its relationship with the Lower House, the Upper House should be a more independent/deliberative, rather than political party-centred check on the Lower House. This increasing partisanship is, according to Takenaka, a reasonably recent development and has somewhat undermined the function and value of a separate Upper House.

The new proposal is not completely dissimilar to the thinking behind the US HoR and Senate, although the roles somewhat reversed in terms of how the Houses are elected. Idealistic as the above expectation may be, I think it serves as a pretty useful principle – institutional design is not the be all and end all of fostering a desired political/democratic culture, but it certainly is a major component. Such a change has great potential to provide different (and importantly, more democratic) incentives for political actors to embrace certain policies and engage in a different style of politics.

Needless to say, despite the greater constitutional appropriateness of such a system, it will not come about so easily and there may need to be some pragmatic compromises. Parties such as the Communist Party and the Socialist Party will may find it more difficult to even maintain their current low numbers, given that they have been more or less PR only parties for sometime. (edited: Actually on rethinking this these PR parties may do well, although it depends on how many seats are up for election in each bloc – the larger the bloc the better it seems for parties like the communists and socialists who sometimes only garner a few percent, which would be enough in the larger blocs. Parties with more support on the PR such as Komeito and YP might do quite well)

Komeito also generally benefits from the PR system, although is and also picks up a lot of urban support. Your Party also benefited from PR at the last Upper House election, although and Your Party in particular might want to take a bet on an expanded multi-member district system which gives more strength to urban voters. While the Communist and Socialist parties are unlikely to agree to the proposal no matter what, That said, the DPJ will have a hard time convincing some of its own members to get behind the plan – a fair number will effectively see their seats abolished and the specific local concerns of their constituents being given less consideration in Upper House election campaigns at least (Although if the latest Upper House election is anything to go by, from a purely self-serving point of view, these DPJ members may lose their seats anyhow).

Such a system may open up opportunities for new parties to form where older parties (including the bigger ones) might stumble in this new system- as suggested Your Party for example could be a beneficiary from such a proposal. The proposed system is not at all inherently contrary to the interests of small parties and could be quite profitable for up-and-coming small-medium size parties. Up until now Upper House elections, given the PR/local constituent split in seats, tended to be run on very broad but non-specific national issues, or on very specific local issues (although also often served as referenda on the ruling government). This might sound good in theory, but in reality it appears to me at least that it allows/incentivizes politicians to avoid the bigger pressing issues and focus on local issues. Local issues are of course important, but one could reasonably ask why have two parliamentary houses emphasizing such concerns. The proposal would seem to transform Upper House elections to being more about regional (as opposed to “local”) and national issues, which could allow for more compromise and a more deliberative, less partisan approach being taken in the Upper House.

One of the “negative” outcomes of such a system would be more minority/divided governments as a major party would struggle without quite substantial support to achieve a Upper House majority. Then again, given the state of Japanese politics right now (where a tired and out of ideas LDP cannot pick up support from voters deserting the DPJ), and the likelihood of there being no majority in either house if an election was held right now, this may not be such a downside. It would certainly compel parties to consider their long-term strategies to managing the legislative process – mainly because parties themselves would have less control over their members in such a system, similar to how Senators in the US have traditionally been seen to be less partisan and more independent of their parties (although, not necessarily more “reasonable,” to be sure).

Nevertheless, “because it is unconstitutional” should be a worthwhile response to the above concerns in a country seemingly committed to democracy and constitutional government. Certainly if the current incarnation of the LDP were to somehow climb back into power without any alterations to the way politicians are elected then it does not bode well for reform and evolution in the Japanese political system – however badly one perceives the DPJ to be managing the government, it is less than clear that any other party interested in reform of any kind would be all that successful in the current political environment. Certainly parties like Your Party, and any party focused on regional issues that might pop in the next year or two, who represent Japanese constituencies with some reasonable things to say, could well find themselves slowly withering away without any long-term institutional insurance through the electoral system.

1 I think there surely must be more detail to be worked out. According to the proposal, the Kinki bloc will have 32 members being elected, with 5 other blocs have 20 or more members. Upper House elections are three years apart so in these blocs 10 to 16 members will be elected each time. If they are treated as one large electoral bloc then assuming the most popular 2 or 3 candidates get a significant proportion of the vote, then other candidates might only have to get a few percent to get elected. This seems very low to me (even if a few percent of a lot of people is, well, a lot) – the 11 bloc system would make more sense I feel (in terms of the logic I outlined in this post) if the number of parliamentarians was reduced to 100 (politically impossible I am sure), or, the blocs were smaller. Or some combination of reduced membership and slightly smaller blocs.

6 thoughts on “Making the Japanese House of Councillors’ elections 4 times more fair….

  1. Hi sigma, been following your interesting blog for some time now as a way to learn more about the going-ons in Japan.

    Anyway, upon seeing your mention of this DPJ proposal, the geek in me decided that as I have a bit too much free time, I decided to do some number crunching. 🙂

    Assuming these 11 blocks each use a simple proportional representation system. (As the national proportional seats would be abolished under this new proposal, I’ve disregarded the data from there and only focused on the district seats. )

    Under the new system, the DPJ would have won 41 seats in this year’s HoC election, i.e. 41% of the 100 proposed seats up for grabs, which would correspond neatly with the 41% of votes they’ve garnered. (Currently it is 38% of the district seats. )

    Similarly, the LDP/Komeito coalition would win a combined 39 seats, also corresponding to the 38-39% of votes they have. (In contrast to 57% of the district seats they have won. )

    (LDP/Komeito results are skewered due to most Komeito voters voting for LDP candidates in the district vote, so I’ve lumped them both together. )

    The two parties to benefit most from this new system would be Your Party and the JCP.

    Your Party would win 12% of seats, up from 4% they would have won without relying on the national proportional vote (NPV). The JCP would win 7% up from 0%, as all their current seats come from the NPV.

    So it’s quite likely these two parties would support this proposal, pending a few details to be ironed out and some concessions given to them by the DPJ.

    (Including the NPV seats, YP would have gone up from 8% to 12%, JCP from 2% to 7%)

    Currently, the DPJ has 106 seats in the HoC, well short of the 121 halfway votes needed to pass this proposal. However, as Your Party and the JCP has 11 votes and 6 votes respectively, even without the Komeito, there’s a chance this proposal might be passed with these 17 extra votes alone.

    And to address your concern about candidates getting in with a very low percentage of votes, a minimum percentage criteria may be imposed, as being practiced in Germany and New Zealand, like 3% to 5%, to prevent extremist microparties from squeezing in.

    • Kinny, first of all thanks for reading.

      Secondly, I fully support the diversions an inner geek sometimes lead to! Yes my initial mistake was surely in lazy reading – thankfully I realized it didn’t make logical sense. A change in the ippyou no kakusa would most likely lead to the kinds of numbers you have suggested.

      A couple of things however – I am not sure if the proposal is to make each of these blocs use a simple PR system. From my reading they would became very large “multi-member” districts. For example a district with a 20 seat allotment would have 10 seats available each time for the staggered elections. From here top 10 vote getting candidates, not parties, would get elected from throughout this whole district. Parties would support certain candidates but the centrality of the party in the actual process itself is less important. It is a peculiar system but one that has worked ok in Japan before – the difference I guess is that on first appearances the “districts” are considerably bigger than they have been. Having a threshold I guess might not work in some of the electorates as you might not fill the required number of seats.

      Ie In our hypothetical district Candidate A is popular – gets 22 percent of the votes from the whole district, B gets 18, C gets 15 percent , D gets 10% – this would leave 35% for the remaining 6 candidates and basically they would have to equally get exactly 5-6 percent each to get over a threshold of 5%. Which is unlikely considering how many candidates would enter into the competition in such a big bloc. (The biggest is 32 from Kinki if I remember, so 16 seats to fill!). I wonder if they could just simply halve the districts, ie have a West and East Kinki district which would only put 8 up at each time and the others less. Alternatively the could increase the amount of times they “stagger” elections ie every 3 years to every 2 years – so most districts except Kinki would have less than 10 seats to fill each time.

      But interesting numbers nevertheless – back of the hand math I guess you could say that if the communists got uniformly 7 percent for a single candidate across all “super-districts,” then they would be all but guaranteed of winning one seat each in each of the districts with 8 or more seats up for grabs (ie 16 overall). They would thus in a 3 yearly staggered system acquire 7 candidates each election – 14 overall out of 200. Your Party with 12 percent should win one seat in each of the districts with 6 or more up for grabs each election, (9 of the 11 districts) and probably acquire 2 more if they were willing to split the vote between candidates in 2 of the larger districts. This would increase their strength to 22 plausibly overall out of 200 (so a small relative gain), although again it is assuming their vote is uniform across each district and that party discipline is good. The JCP would have much better party discipline (ie hard core supporters who will vote as directed) than Your Party which attracts swing voters. So perhaps it is a gamble for Your Party……DPJ plus PNP, plus SDP and JCP only comes to 120 out of 242 – to attract the Komeito (19) and Your Party (11) they might have to half the district sizes or increase the staggering so to guarantee a better relative strength compared to now to attract these two parties.

      (Note I think the Komeito vote was lower in the in the district vote because in the single member districts Komeito supporters voted for the LDP candidate – so the LDP number could be lower while Komeito might be higher and more in line with the vote from the PR side of things.)

      But yes, convoluted – be interesting to see if they get anywhere with it!

  2. Thanks for replying.

    Anyway, I am aware that the HoC elections have been using Single Non-Transferable Votes (SNTV) for its multiple member districts (MMDs) since time immemorial (1947), which is more or less a de facto proportional representation method with only one choice out of many, only that voters pick individual candidates rather than a party list (opened or closed).

    Though as you have noted yourself, these existing MMDs usually only have up to 5 slots up for grabs for every staggered election, as anything larger than 6 slots would be a nightmare to speculate using the SNTV format and would simply make my head explode, as I have absolutely no idea who’s who in which district have better chances of winning, I’m not as well networked as your average Ozawa-like election strategist, after all. 😉

    Which was why I used the assumption of a simple list-based PR instead, like how the current national PR bloc is elected (with 48 seats up for grabs).

    From how I see it, these 11 super-districts are basically derived from the existing giant nationwide PR bloc divided into 11 smaller PR blocs. As for what kind of PR is to be adopted, be it open/closed-list PR (used in most European countries) or Single Transferable Vote (STV, used in the Australian Senate), that’ll be for the MPs to decide amongst themselves.

    As long as it’s not SNTV, then IMHO I don’t see any problem with a district being allocated 10 seats or more (up to 16 for Kinki). Though that’s just my two cents.

    From my calculations: the JCP would win one seat each in North Kanto, South Kanto, Tokyo, Tokai and Kyuushuu and two in Kinki, while YP has one seat each in Hokkaido, Tohoku, North Kanto, Tokyo and Kyuushuu, two in Tokai and Kinki and three in South Kanto.

    (Note I think the Komeito vote was lower in the in the district vote because in the single member districts Komeito supporters voted for the LDP candidate – so the LDP number could be lower while Komeito might be higher and more in line with the vote from the PR side of things.)

    For the district seats, the LDP had won 35% of the votes to the Komeito’s 4%, totalling 39%.
    For the nationwide PR bloc, LDP has 24% while Komeito has 13%, which would total roughly 37% – more or less the same.

    That was why I decided to lump their seats together as most LDP seats are won with Komeito votes.

    PS Most of the data for my calculations came from this incredibly useful website, which has election result archives all the way back to the first Meiji elections in 1890.

    • Thanks for the clarification. Yes I agree the SNTV would be a nightmare which is certainly where I was going with that. My thoughts were exactly that STV might be an interesting system having seen it work in Australian elections. A major concern for the Japanese has always been with “wasted votes” which was one of the motivations for bringing in the PR system in the first place given a large number of SMDs, so it would certainly meet that criteria. I would be surprised if it was not considered as an option in 1993 but a superficial search does not bring anything up. You are right, certainly in principle if it was STV then this does not lead to electoral distortions even in big districts. But I guess my concern is with what kind of elections in terms of coverage and discussion of issues there would be, with maybe 50 or more candidates competing over a wide area.

      Oh and thanks for the website link! I meant to ask in the first comment actually, but it has been bookmarked. Cheers 🙂

  3. Pingback: A bolder approach to electoral reform? « σ1

  4. Pingback: Constitutional change in post-3/11 Japan. « σ1

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s