1998 年 13 巻 p. 5-16,268
The concept of partisanship has been one of the central topics of the voting research in Japan as in other democratic countries. While whether the original concept of party identification developed in the U. S. is transportable to other countries is still under debate, the party support variable (measured by the question, “which party do you usually support?”) is often treated as functional equivalent to the party identification measure in Japan.
This review article reminds researchers of Japanese voting study of a need for a careful look at the party support variable. It does so by going “back to the basics.” It evaluates the party support variable against the four basic assumptions of the original party identification concept: the sense of identification, its stability, its unidimensionality, and its transitiveness.
Citing the existing works and drawing some new data, the article concludes that 1) the party support variable is not exactly measuring the sense of self-idenification with a party, 2) it is not as stable as its counterpart is assumed to be, 3) it is increasingly difficult to map the Japanese current political parties on the left-right uni-dimensional scale, and 4) whether its operational definition meets the transitiveness assumption is questionable.