By Andrew Kohut
n its 1992 post-election survey, the Times Mirror Center for the People and the Press found that, unlike four years earlier, voters felt that they had learned enough from the presidential campaign to make an informed choice. Most elements of the campaign were given markedly better performance grades than in 1988.
There were two notable exceptions: Evaluations of the press and the pollsters changed little. On average, the press was graded C and the polls C+ for their roles in the 1992 campaign.
A public grown increasingly cynical about the political process has come to view both the press and the polls as part of the problem, not part of the solution. Both are seen as having too much influence on the outcome of elections. It is ironic that modern polling, which was defined by its founding father, George Gallup, as an “instrument of democracy,” would come to be seen as disenfranchising people.
But these days politics, polling and the media are more closely linked in the public’s mind than ever before. Political campaigns and the policy process in Washington have become integrally tied to the practice of reporting public attitudes. Pollsters and the media are seen as partners in a process from which the average American feels detached. Voters’ complaints about campaign coverage focus on negativity, on lack of coverage of issues, and on the extent to which the horse race dominates. The critical public sees the hand of polling in each of these shortcomings. To many people, election polls are about “who’s up, who’s down” and little else.
But polls can be, and often are, much more than this, as the surveys made available to the Citizens Election Project attempted to demonstrate. Polls can play an integral role in reconnecting news organizations to their audiences. Survey research is an important method for reaching out to citizens. It can bring their voices and ideas into the foreground of news coverage in a systematic, representative manner. The public mood, the public agenda, and the public’s political and policy preferences, as measured by polls, are crucial source material for journalism and they attempt to encourage public engagement.
What does this mean in practical terms? How then can polling better serve the citizenry? The answer is not to banish horse race polls. Observing the development of candidate preferences is an important element of election coverage. It is, however, only one element. In our view, there are other components. For example, citizen-oriented polls should include:
Questions that find out how issues, problems and policy alternatives relate to people’s lives and personal experiences. For example, asking about the most important problems facing the nation is one step. But, further questions that ask respondents about problems that they face in their communities and lives give a more people-oriented understanding of that agenda.
Open-ended questions that ask respondents to put in their own words views about issues, candidates and policies. Such questions are too often underused because of cost factors, complexity of reporting and researchers’ biases against seemingly unwieldy question results.
Information questions that test public knowledge. They can clarify opinions not yet fully formulated. Learning more about where people get their information and how they form their opinions is essential to a more accurate and more nuanced reading of public opinion.
An examination of attitudes on a multi- dimensional basis. Exploring both the conflicts and consistencies inherent in opinion offers a more realistic portrayal of the public than the all-too-frequent “one-question” judgments. Polls that measure the values and the fundamental attitudes that underlie policy and candidate preferences provide a richer and fuller understanding with which the public itself can identify.
The Pew Research Center’s surveys made available to the Citizens Election Project were guided by these principles. Four nationwide polls, as well as a survey of New Hampshire voters, were conducted during the primary season. The surveys were in addition to focus group research by The Harwood Group that was commissioned for the project.
The Pew Research Center also augmented its surveys with an experimental Poll Watch, which attempted to enrich the press’s understanding of public opinion through an ongoing secondary analysis of polling data throughout the primary campaign season. The Pew Research Center synthesized the results of the wide array of national and key primary state polls. Its Poll Watch reports looked critically at the successes and failures of surveys as they charted candidate preferences and voter attitudes and were used as a resource by citizens as well as by journalists. 