Party Identification

Party identification (often shortened to "party ID" or PID) refers to a citizen's self-described affinity with a political party, typically measured on a survey scale running from "strong Democrat" through "independent" to "strong Republican," or analogous scales in other party systems. The concept was developed by Angus Campbell, Philip Converse, Warren Miller, and Donald Stokes in The American Voter (1960), based on University of Michigan election studies, and remains one of the most-used variables in voting behavior research.

The Michigan model treats party ID as a long-term, affective identity learned early in life—largely through parental socialization—and relatively stable across the life course. In this view, it functions as a perceptual screen that shapes how voters evaluate candidates, issues, and economic conditions.

A competing revisionist account, most associated with Morris Fiorina's Retrospective Voting in American National Elections (1981), argues that party ID is better understood as a "running tally" of accumulated judgments about how parties have performed. Under this view, identification can update in response to events, though more slowly than vote choice.

Key empirical patterns associated with party ID include:

• Predictive power: in U.S. presidential elections, partisans vote for their party's nominee at rates that have exceeded 90% in recent cycles, according to American National Election Studies (ANES) data.
• Turnout: strong identifiers vote at higher rates than weak identifiers or independents.
• Motivated reasoning: partisans tend to accept information congenial to their party and discount contrary information.

Outside the United States, party ID is measured but tends to be weaker and less stable in multiparty systems, and scholars such as Russell Dalton have documented partisan dealignment in many advanced democracies since the 1970s, with rising shares of voters declining to identify with any party. Independents who "lean" toward a party, however, often behave like weak partisans rather than true neutrals.