Abstract
In previous studies about multi-member districts under the SNTV, the “M+1 rule” was proved using the Lower House election data. (Reed (1990, 1997), Kawato (2004)) In this paper, we show the difficulties that occurred by using the Lower House data and propose the use of the prefectural election data of ordinance-designated cities. There are two advantages in using the prefectural election data of ordinance-designated cities. The first is that the numbers of the candidates who have few possibilities to win the election are few. The second is that there are many cases that the numbers of the seats of the districts are changed. Using these two advantages, we confirm the “M+1 rule” for the number of the whole candidates and the stability of “M+1 rule.”
