REESTABLISHING "THE SOCIAL" IN RESEARCH ON DEMOCRATIC PROCESSES: MID-CENTURY VOTER STUDIES AND PAUL F. LAZARSFELD'S ALTERNATIVE VISION

Voter studies conducted in the United States during the first decades after World War II transformed social scientific research on democracy. Especially important were the rapid innovations in survey research methods developed by two prominent research centers at Columbia University and the University of Michigan. This article argues that the Columbia and Michigan voter studies presented two visions for research on democracy. Where the Michigan research produced quantitative measures expressing the 'political behavior' of the electorate, the Columbia studies, and especially Paul F. Lazarsfeld, presented an alternative vision for qualitative research on political choice. Largely ignored by later voter studies, this vision prefigured much contemporary research on democracy that embraces a qualitative or interpretive approach. This article reconstructs Lazarsfeld's alternative vision, describes the institutional context in which scholars disregarded it in favor of formal quantitative models, and argues for its recognition as a forerunner to qualitative research on democratic processes.