Author(s): Ikuo Kabashima and Ryōsuke Imai
This paper examines the impact that the Japanese electorate's negative
evaluation of Prime Minister Mori Yoshiro had on its voting behaviour in the
2000 General Election.This study uses an eight-wave panel survey covering the
period from 1993 to 2000.The last survey which was used most, was conducted
through the mail immediately after the 2000 General Election. It was found that
party leader evaluation did in fact affect voting behaviour, with voting in
proportional representation districts affected more significantly than voting
in single-member districts. We also examined how party leader evaluation
affected split voting between proportional and single-member districts, and
found that negative party leader evaluation was a major cause of split voting.
Finally, it was found that the evaluation of Prime Minister Mori himself was
affected most significantly by a series of problematic speeches Mori made in
the run-up to the election.
Introduction
Research into voting
behaviour in Japan has not focused very much so far on the relationship between
voting behaviour and evaluation of party leaders. Despite the heavy media
coverage of party leaders, and especially of prime ministers, it has been
assumed that, compared with the support rating of a party or evaluation of
candidates, this factor has had little impact on election outcomes, an assumption
reflecting the relatively low profile of Japanese premiers over the years
(Miyake 1989). Certainly with the exception of Nakasone Yasuhiro and one or two
others, few party leaders have had distinct personalities or exerted strong
leadership in post-war Japan. In the case of the 42nd general election (lower
house election), held on 25 June 2000, however, considerable atten- tion was
devoted to party leader evaluations for negative reasons: a series of very
insensitive remarks by prime minister Mori Yoshiro' had led many to question
his fitness as premier. On the other hand, his successor, Koizumi Jun'ichiro,
enjoys very strong support, and this was widely seen as the decisive factor in
the strong showing by the Liberal Democratic Party (LDP) in the Upper House
election of July 2001.
The aim of this
paper is to clarify the relationship between party leader evaluation and voting
behaviour, focusing on the 2000 general election. By analysing election results
from this hitherto neglected perspective, we hope to provide new theoretical
input into voting behaviour in Japan. The data used in this paper are findings
from a survey conducted immediately after the June general election by a group
including the authors. Previous surveys led by Kabashima had carried out a
series of seven linked nation-wide panel surveys between 1993 and 1996, and
this was the eighth in the series (cf. Kabashima 1998). As nearly four years
had elapsed since the seventh survey, many respondents had died or moved away
in the interval. However, valid responses were received from 803 persons, a
response rate of 29.9%. As this was a mail survey with a low retrieval rate,
certain reservations apply to the findings. However, since this is the only
survey of party leader evaluation in the 2000 general election, we believe its
findings are nevertheless of interest.
PRIOR
RESEARCH
Research in the US
(Pierson 1975; Abramowitz 1985) and in Britain (Stewart and Clarke 1992) has
tended to indicate that leader image is a substantial factor influencing voter
behaviour in those two countries. There has been less research in this field in
Japan, but two studies deserve mention, one by a group under Araki Toshio at
Hokkaido University (Araki et al. 1983) and the other by Kawato Sadafumi
(Kawato 1988). The Araki group researched the connection between party leader
image and voting behaviour in the 1979 lower house election and the dual upper
and lower house elections in 1980 and 1986. For the 1979 and 1980 elections, Araki
et al. conducted a panel survey which asked respondents to evaluate prime
minister Ohira Masayoshi, president of the LDP, and Asukata Ichio, Japan
Socialist Party (JSP) chairman, on a five-point scale for 11 items relating to
leadership qualities. The follow- ing findings were obtained:......
continue.........