Polls

Polls are a useful method for collecting information about the public's opinions, attitudes, and behavior in a variety of political contexts. Good-quality polling results in information that describes these attributes of a relevant population to which inferences can be drawn from the sample of the respondents who were interviewed. Polls and surveys are employed by political campaigns, news organizations, and interest groups who want to understand the public's reactions to actual or potential events in order to develop strategy, report on current events, or contribute to the development of public policy.

A poll is a form of data collection that typically involves a sample of respondents, drawn to represent an underlying relevant population, who are asked a standardized series of questions in a fixed form. The results are analyzed for the entire sample, with inferences drawn back to the population from which it was drawn, as well as for specific subsamples that represent subgroups in the population. In some cases, the relevant population is well known and easily operationalized, as in the case of “adults age 18 and over residing in telephone households.” But in other cases, the relevant population has to be constructed during the course of the interview, as in the case of “likely voters.”

Polls and the news media have been closely linked ever since the start of commercial polling in the 1930s. In this initial period, many pollsters built the commercial, private side of their businesses by having a relationship with news organizations that would distribute the results of their polls of public opinion and build the visibility of their firms. This was true of George Gallup, who produced a syndicated newspaper column, and Elmo Roper, who started out doing quarterly polls for Fortune magazine. They anchored their public work around their performances during presidential election campaigns; their general success in estimating those outcomes, leaving aside the 1948 election, validated the method and increased its public acceptance. In the 1960s, major news organizations [Page 622]formed partnerships to conduct their own polls, linking large circulation metropolitan dailies and national television networks.

In the 1970s, broadcast networks were engaged in a contest with newspapers about the nature and content of election night coverage. The networks began to develop the exit poll methodology as a way to capture the views of voters leaving the polls to support immediate coverage of who won and why. Up until this time, newspapers had been the primary source of information about elections, although the full tabulation of the votes would often not be available until Wednesday's or Thursday's editions. This exit polling technique transformed the coverage of elections by providing more immediate information and fuller analyses of what happened and why. Exit polls did not come without controversy, however.