People's Choice, the

To this day, the study known as the People's Choice has influenced academic discussion far beyond the field of political communication. In sociology, the study is [Page 541]regarded as methodological spadework, in political science as the first systematic explanation for voting behavior, and in communicational science as the beginning of the empirical research on media effects. Thus, the study is a milestone for any of these disciplines.

Methodically, Paul Felix Lazarsfeld (1901–1976) continued to develop for People's Choice the instrument of the descriptive survey in order to explain differences between and changes of attitudes. From the constituency of the Erie County community in Ohio, he took a representative sampling that was divided into four evenly composed groups of 600 persons each. One of these groups was interviewed once a month over a period of 7 months during the presidential election campaign of Franklin D. Roosevelt and Wendell Wilkie in 1940 in order to trace individual developments. The other three groups were used for monitoring purposes by interviewing them once, but not at the same time—a groundbreaking design (“panel technique”).

The substantial output of this study can be summarized in three theses:

• 1
  Political predispositions. Social structures explain voting decisions. The assumption is that the decision for a vote can be explained by considering three lines of conflict: city/country, socioeconomic status, and religion. “Social characteristics determine political preference” (p. 27). Through an “index of political predisposition,” the interviewees were grouped into different classes that permitted for prognostication about their voting decisions. The election campaign had the following effects:

  • The primary political orientation of more than half of the interviewees is reinforced during and by the election campaign (“reinforcement effect”).
  • The political orientation of 14% is not activated until the election campaign, but this orientation can be predicted from their positions in the social structure (“activation effect”).
  • Only a small number (8%) undergoes conversion of its previously stated party preference during the election campaign; these are usually unconcerned people with little interest in politics (“conversion effect”).
• 2
  Opinion leadership. Political preferences are mediated by communication in homogeneous groups. The social structure forms separate social networks. The communication in these social homogeneous groups is of determining significance, especially the interaction between the “opinion leaders” who are intensively involved in the election campaign and the “opinion followers” in the according network. Both types are distinguished by psychological and communicative variables, that is, willingness of articulation. So in all social groups there are “opinion leaders.” They provide orientation for the followers, that is, they mediate “interpretations” and exert social pressure, for example, in their families or at work. For this reason the group specific opinions are intensified—including those of who will win the election (“bandwagon effect”).
• 3
  Two-step flow. Media messages have limited effects—mediated by the opinion leaders. Because of the group communication, latent political predispositions become apparent in the citizens. They are the selection pattern with which the media messages are filtered (“protective screen”). The opinion leaders follow the media in a comparatively intensive manner; they arrange the election propaganda according to the pattern of analysis and pass them on afterward. The authors are able to show that those who follow politics in the media very intensively change their opinions very rarely; their positions were also confirmed by messages that were intended to achieve the absolute opposite. Conversely, the more someone is susceptible to change, the less she or he is likely to be reached. The authors find interpersonal communication by far more effective than media communication; they see the reason for this in advantages such as flexibility, trust, and commitment: “more than anything else people can move other people” (p. 158).

All three results have to a great extent formed the understanding of public opinion. The explanations for voting decisions out of the social structure became the groundwork for election prediction models. Compared to competing approaches, the sociostructural explanation has lost ground as traditional social relationships are dissolving and thus the regular constituencies decrease and the volatile voting potential increases. The hypothesis of the relevance of group communication has been reviewed many times und developed to the network approach. The hypothesis of the twofold relativized media influence (through social structures and interpersonal communication) has dominated the discussion, until in the 1970s long-term and specific—for example, cognitive—media effects became the center of empirical research. In this process, a central [Page 542]role is played by television, which is imputed with a comparatively high credibility and suggestive power and therefore with a higher influence on voting decisions than the press and the radio.

...