Illusions of Ticket-Splitting

Abstract

Studies of ticket-splitting have been based on flaws of evidence and conception, whether using aggregate or survey data. A common basis for error has been the "two-office split- result" approach. Errors devolving from this approach, for example, assume that split congressional districts or split gubernatorial/ state legislative alignments are indeed based on ticket-splitting. Survey studies, with rare exception, suffer from the problem of voter recall, reliability of reported information, and the "missing GOP vote" explanation. Both aggregate and survey studies commonly either ignore the incomplete problem or understate it significantly. However, the high number of incomplete ballots suggests, along with other data, that ticket-splitting is not necessarily widespread or increasing. Approaches to remedy these problems are suggested.