Running to win: The utility of the ticket-splitter theory in presidential elections
Theory suggests that ticket-splitters often are the balance of power in elections, and constitute at least one-fourth of the American electorate. Ticket-splitting is not, however, motivated by party identification. Instead the voter's evaluation of the candidate, pertinent issues, and party attitudes contribute to their choice. Candidates, therefore, can attract electoral majorities by communicating certain messages through the identification, location, and targeting of ticket-splitters. This thesis will examine election data from 1952-1990. If the theory proves reliable and valid, the usefulness of the traditional two-party analysis of the modern electoral environment will be questioned.