This paper reviews the study on Japan's electoral institution called Chû Senkyoku Sei (the Medium-sized District System, MDS). I focus on the working of MDS during the one-party rule by the LDP, and show two distinct approaches that I call MDS research and SNTV research are based on different understandings of the same electoral system and political actors under it. MDS research is better equipped with detailed knowledge of various aspects of electoral systems. A lesson one can learn from MDS research may be that how politicians understand the electoral system definitely affects the working of the system. SNTV research introduced new perspectives into the electoral research in Japan. It presented theoretically important insights into Japan's electoral system and guided recent innovative research. Perhaps the greatest contribution is that it has made the study of Japan accessible to a broad audience of political science community. Both approaches have contributed to enriching the Japanese electoral research.

View full abstract

I define the state of equilibrium as a state that is relatively easy to enter but relatively difficult to exit. So defined, equilibrium becomes a dynamic concept which must be analyzed as change over time. Using the graphical technique of a “return map” borrowed from chaos theory in physics, I show that the M+1 equilibrium (that there should be one more serious candidate per district than the number of seats in the district) holds for Japanese House of Representative elections from 1947 through 1993. I also show that the M+1 equilibrium is stronger in smaller district magnitudes.
I further examine another definition of equilibrium found in the literature: that the effective number of parliamentary parties should equal the effective number of electoral parties. Though the logic of this definition is impeccable, it does not work well empirically.

View full abstract

This paper advances an institutional hypothesis of voter behavior which postulates that the electoral system influences parties' strategies with respect to determining their issue positions and placing relative weights on issues and traits. And, in turn, their strategies influence the relative importance of issues and traits in voting and the empirical utility of a theoretical model of voter behavior. In a weak system such as proportional representation, parties, regardless of whether they are vote-maximizing or policy-seeking, may be motivated to take distinct issue positions. By contrast, in a strong system such as plurality rule, office-seeking candidates may take centrist, homogeneous issue positions and highlight traits in their electoral campaigns. Such electoral strategies may in turn affect voters' evaluation functions. In my empirical analysis, I use an integrative model based on nested logit in order to examine the effects of issues and traits on voter behavior in the plurality and the proportional representation component of the 1996 general election for the Japanese House of Representatives.

View full abstract

Since the early 1990s, Steven Reed and Gary Cox have changed our understanding of Japan's multimember SNTV electoral system by high-lighting its institutional effects similar to what is known as Duverger's law in the Anglo-American context. Both Reed and Cox presented evidence to show that Japan's M-member districts tended to produce competition among M+1 candidates, thus claiming that a generalization of Duverger's law, called “M+1 rule” exists and applies to the Japanese case. This paper reevaluates Reed's and Cox's claims. We offer some additional evidence to demonstrate that the findings of these two studies are reconcilable, both pointing to the basic pattern of Japanese voters' strategic behavior under the above electoral system. We, however, also address an issue left unexplored in these previous studies, namely the role of political parties. Under Japan's multiparty system, party labels matter in elections: while Japanese voters are generally willing to abandon the candidates without affiliation with the established parties, the partisan effects produce constraints for voters' strategic coordination for both informational and ideological reasons.

View full abstract

In this paper I review from historical perspective electoral studies in japanese constitutional law and discuss some issues that deserve further consideration. Electoral rules can be thought of as the most important components of one country's constitution. The Japanese Constitution of 1946 has several extensive provisions concerning elections. Japanese constitutional law scholars have analyzed the normative meaning of these provisions in the context of one's right to vote, the right to stand for election, election canvassing and malapportionment. This paper summarizes these normative analyses of some eminent constitutional law scholars.

View full abstract

Students who write about elections in Japan has been largely concerned of party identification or partisanship, or the party-in-the-electorate. Very little attention has been paid to the role of party organizations in elections, mainly on the assumption that they had little role simply because Japanese parties lacked organized mass membership. But this assumption is unexamined, like in the U. S. Recent studies of American political parties revealed the greatly enlarged activities of national party committees, especially in congressional elections during 1980s, suggesting that parties could have adjusted to candidate-centered politics as useful intermediaries rather than as dominant actors. Also in Japan, more inquiry of parties' roles in elections is essential to a realistic understanding of Japanese elections.

View full abstract

A shift of paradigm in electoral studies is not as fundamental shift in epistemology of universe as that in natural science or in physics. However, those characteristics of paradigm shift described by Thomas Kuhn seem to be applicable to miniparadigm shifts in electoral studies. In electoral studies, the paradigm has shifted from the traditional political science to behavioralism, and further to rational-choice theory. Are these shifts of paradigm in electoral studies progressing toward the truth? It is possible that, as Kuhn suggests, from a different perspective the truth can be seen as a different configuration. If so, a paradigm shift means a shift of a perspective about the truth. Then, interactions between differnst paradigms or perspectives will lead us to a new direction in our research (or our better understanding of the electoral phenomena). Especially, interactions between the empirical or inductive behavioral approach and the deductive rational-choice approach should be fruitful. They should not be mutually exclusive but mutually supplementary.

View full abstract

This article is an introduction and an analysis of “voters' satisfaction with party's policy”. It shows (1) that “voters' satisfaction with party's policy” is based not only on their “policy distance from party' position” but on their perception on party's policy competence, (2) that it can explain their voting decision better than “policy distance”, and (3) that it represents a part of party's job performance.

View full abstract

By using two decision models based on cognitive social psychology, voting behavior in the 1998 Japanese House of Councilors election was analyzed. JEDS national panel survey (1996-1998 “Japanese Election and Democracy Study” accessible in WWW (http://www.iss.u-tokyo.ac.jp/pages/ssjda-e/)) was the source of analysis.
Political party schemas in 1996 and their change indicators between 1996 and 1998 were both highly significant in predicting the 1998 vote, which shows the effectiveness of cognitive accessibility model of voting behavior.
Retrospective voting, which reflects the use of cognitive shortcut in voting decision, was also significant in determining the vote (by multinomial logit analysis). Much unclear was the role of sociotropic judgment on the vote, although the egocentric vote was clearly existent. That is, sociotropic judgement has positive relationship with the retrospective evaluation of the 1998 Hashimoto Cabinet, but its relationship with the vote was slightly negative when controlling the egocentric evaluation of economy as well as other variables. (151 words)

View full abstract

After W. W. II, Japan had been re-developed in economic sphere and democratized drastically in social and political sphere. The more people had been affluent, the less leftist parties, SDPJ, JSP, SDL, or JCP, had got their votes in post war Japan. So far as five SSM data-sets from 1955 to 1995, following four aspects are found within the socio-economic change, at first, extention of regular schooling, secondly, increase of employees in industry, thirdly, diffusion of consumer goods, and at last, persistence of property differentials. In this paper a socio-economic model concerning the diachronic micro-macro link between aforementioned decline of leftism and societal change totally, is presented. Hypothesis on intelectuals' leftist dissolution, hypothesis on employees' leftist dissolution, hypothesis on consumer goods diffusion, and hypothesis on property differentials, in the socio-economic model, are confirmed by logistic regression analyses through five SSM datasets, except the hypothesis on employees' leftist dissolution. Althogh there are a few methodological unsatisfactions, which derived from cross-sectional data-sets, about diachronic micromacro link, the socio-economic model is durable enough, roughly speaking, and a structural background on dissolution of leftism is found systematically.

View full abstract

‘Public goods problems’ arise when the effort of several self-interested individuals is needed to accomplish a task. In this paper, we examine whether or not free riders exist in political activities of a civil group and a related local party. Through an experimental method supposing an election campaign, we test how varying levels of benefits from public goods, selective incentives, and costs influence cooperative behavior. The results show that a few free riders do emerge under particular conditions. We also suggest that levels of political interest slightly distinguish free riders from simple non-cooperators, and that cognition of costs distinguishes free riders from cooperators. The study indicates that the groups are supported by their devoted members, but at the same time know they constantly need new members' participation to maintain themselves. Finally, we find the need for further study of the ‘members’ turnover cycle' and the resolution of methodological problems.

View full abstract

This paper reexamines how the “Clean-Up” (Shukusei) election campaign functioned in Japanese political history. The campaign in the late 1930s was not merely an official enlightenment movement to prevent electoral corruption but an all-out campaign launched nationwide by the Home Ministry and the military intending to get rid of political parties. The authorities aimed to establish a new political system which would replace “a party cabinet” to deal with a national emergency. However, the scheme was not a success. Consequently, this failure resulted in the preparation of the “Yokusan” election to establish a one-party government.
The research reveals the following points:
1, The campaign did not necessarily destroy the constituency of the political parties. However, the solid relations between voters and parties did not exist in those days.
2, The campaign neglected parliamentary government and party politics, and on the other hand it advocated the “sacredness” of elections and prevention of abstention. This contradiction caused an increasing rate of abstention.
3, To emphasize the sacredness of elections, visiting shrines and flying the national flag were forced on the people. And gradually the campaign turned into a movement affiliated with state control for unifying the Japanese.

View full abstract