Partisan loyalty and the six component model

Abstract

In a recent article William Shaffer contends that the classic SRC six-component model is degenerate in that all six dimensions of partisan attitude reduce to one: party identification. Shaffer reaches this radical conclusion through a factor analysis of the six-component scores for the 1952 to 1968 Presidential election surveys. A close look, however, demonstrates that Shaffer's position is both methodologically and theoretically indefensible. His factor analytic results simply do not justify the conclusion he draws. Moreover, even if the six components did reduce completely to one underlying dimension, the implication does not follow that party loyalty is the operative mechanism. A consistent reality can be just as productive of attitudinal consistency as a hypothetical underlying psychological factor.