2005 年 56 巻 1 号 p. 11-35,349
Since the electoral reform a decade ago, three elections to the House of Representatives have been held. At the time of the reform, the problems of political funding and widespread corruption were commonly attributed to the mediumsized constituency system. It was then argued that the introduction of the dual system, consisting of the small constituency and proportional representation systems, would counter these problems by stimulating policy debates. This study examines the validity of these claims by comparing the patterns of voting behavior at elections held under the new and old systems. The analysis empirically tests five hypotheses built around the following points: 1. The range of candidates' policy positions, 2. The distance between candidates' issue positions and voters' ideal points, 3. Whether issue voting has increased/decreased? 4. Correspondence between subsidies to a constituency and the number of votes gained, 5. Correspondence between subsidies and the number of votes gained at House of Councilors elections. The results of this study show that the dual system has not brought about the expected changes in voting behavior. It also suggests that the problems regarding political funding did not originate in the medium-sized constituency system but still persist under the new system.