Party Identification

Party identification is a term that is used to describe a psychological attachment to a political party. The concept was initially developed by Angus Campbell and his colleagues at the University of Michigan in their landmark studies of voting behavior in U.S. presidential elections in the 1950s. It is at the heart of the social psychological model of electoral choice and became a central concept in the study of public opinion and voting behavior. The Michigan scholars conceptualized party identification as a durable affective loyalty that tends to reinforce itself as it serves as a perceptual filter that screens out information that might weaken it. It is seen as shaping citizens' perceptions of politics, their evaluations of political objects like candidates and issues, and their behavioral responses to political stimuli, most important their voting behavior. By providing coherence and stability to political perception and political behavior, party identification conditions the effectiveness of communication strategies adopted by political elites. At the aggregate level, party attachments provide stability to public opinion and electoral outcomes.

The concept of party identification encompasses two dimensions. The directional component indicates which party a voter identifies with, while the intensity component denotes how strongly she identifies with this party. The Michigan scholars devised a simple instrument to measure party identification in the United States. First, respondents are asked, “Generally speaking, do you think of yourself as a Democrat, a Republican, an Independent, or what?” Persons who identify with a party then are asked how strongly they do, while the remaining respondents are then asked whether they lean toward a party. Subsequently, the answers to these questions are combined to form a seven-point scale ranging from strong and weak Democratic identifiers to weak and strong Republican identifiers. At the midpoint of the scale, pure Independents are located between Democratic and Republican leaners. Thus, to capture the concept that encompasses two components a one-dimensional measure is used.

This procedure has several conceptual implications that have not gone without criticism. First, some authors proposed to regard the directional component and the intensity component as being independent of one another, so that it would become easier to explain why some strong identifiers with one party turn to strong identifiers of another party. Second, the traditional measure assumes the identification with one party to be incompatible with an attachment toward another party, thereby ruling out the possibility of multiple party identifications. Third, partisan independence and party identification are also assumed to be mutually exclusive. Building on evidence on the attitudinal and behavioral correlates of independent leaning, critics objected that this assumption is not valid and conceives of independent leaners as “closet partisans.” However, more recent research revealed crucial differences between leaners and identifiers in attitudes toward political independence, thereby backing the traditional measure. Finally, critics pointed to the possibility that citizens may dislike a party while not identifying with any party. Building on this argument, they pleaded for considering positive and negative attachments to political parties independently. Despite these and related objections, the traditional measure of party identification is still in use.

...