Polls and Politics: The Dilemmas of Democracy

5. Do Polls Give the Public a Voice in a Democracy?

5 Do Polls Give the Public a Voice in a Democracy? Michael W. Traugott At the advent of the modern polling period, there was an extended debate about the role that public opinion polls could play in a democracy . It pitted the pollsters against some academics in a conversation about the contributions that polling could make to an informed debate about policy and the appropriate voice that the public could have in such deliberations. With the passage of time—now more than half a century—there is almost no discussion any longer about whether polls should be used to measure public opinion or that the results should be widely disseminated. The increased number of polling organizations and the advent of polling operations within news organizations have rendered most parts of that debate moot. Nevertheless, serious questions remain about how the public’s voice is being measured, how it is interpreted by those who analyze and report on poll results for public consumption, and whether or not new technology will endanger the collection of representative data by presenting a set of distorted views about what the public thinks or prefers. In a parallel development to the rise in telephone usage (which reduced the cost of conducting polls compared to face-to-face surveys in households in both time and money), the rise of Internet and Webbased polls is luring public pollsters to collect more data faster, often to feed the insatiable appetite of Web sites for new content. But now the problem of representativeness takes on new meaning and raises a different set of concerns. In this chapter, I trace the origins of modern polling as it relates to recording the public’s voice, focusing on the early justifications and rationalizations for legitimizing the technique and the industry. Then 77 I trace some patterns of organizational development and adaptation that have spread the use of polls and the dissemination of results. And finally I look at problems of contemporary reporting of polls and the strategic use of data that may be distorting the public’s voice as reflected in poll results. Polling as a Representation of the Public’s Voice Since the development of commercial polling by George Gallup and others in the 1930s, proponents have argued that the technique is an ideal forum for expression of the public’s sentiment outside the context of elections.1 Archibald Crossley, writing just after Gallup successfully challenged the Literary Digest in projecting the outcome of the 1936 presidential election, viewed the public opinion poll as a way to correct for “false presentations of public opinion” and as “the longsought key to ‘Government by the people.’”2 Gallup and Rae described public opinion as: “the reserve force in democratic politics [that] can play its part only if the common run of people are continually encouraged to take an interest in the broad lines of public policy, in their own opinions, and in those of their fellows, and if clear channels exist through which these opinions can become known.”3 These pioneers claimed that polls had an important role to play because elections in the United States occurred systematically but infrequently. There could be issues that politicians avoided or never discussed during campaigns, or certain issues that arose during a presidency or legislative session that could not have been previously discussed because no one knew they would arise. Measurements of public opinion about the U.S. government’s responses to the attacks on 11 September 2001 or the disclosures about Enron’s business practices and its political contributions are more recent examples of such issues. Polls present an opportunity for citizens to express their opinions and preferences on these matters when the ballot box is not available to them. And the broad dissemination of results keeps the population informed about other citizens’ views on important issues of the day.4 There have always been critics of the development of public polling. Blumer was one of the earliest; he railed in his 1948 presidential address to the American Sociological Association against the simple aggregation of attitudes as a measurement artifact without any theoretical basis.5 Ginsberg is a more contemporary critic who is concerned about the ways that the act of polling has transformed the meaning of “public opinion.” He describes the measurement of opinions as 78 Michael W. Traugott “externally subsidized” by polling agencies that formulate questions and field surveys on their own...