Polling during presidential election campaigns has become a daily affair with unfortunate consequences for voters and politicians alike. Not only has the number of voter polls broadcast by network television news more than tripled since 1968, our research shows that polling reports have grown at the expense of issue coverage.
In 1968, for each minute of poll coverage in their news reports, the TV networks broadcast four minutes of campaign issue coverage. By 1992 this ratio had fallen to only 90 seconds of issue coverage for each minute of poll coverage.
These changes are not an accident. Television news producers believe that to hold viewer interest they must de-emphasize issue coverage and turn campaigns into horse races. Indeed, we've found that poll results in the 1990s are much more likely to be reported at the beginning of broadcasts as "news exclusives" than they were in the 1960s and '70s. This helps create the impression that the network has landed a scoop and that the story being reported is especially important and newsworthy.
Network news executives typically blame the public, rather than themselves, for this lack of issue orientation in their broadcasts. "People hate politics," they complain. "Viewers will change channels if we report on issues." But polling and constant horse-race coverage contribute to the public's lack of involvement and experience with campaign issues. Polls are used to characterize how each candidate and campaign is doing, but with little awareness that they often drive the whole tone and orientation of stories.
Of course, poll-generated stories also have enormous impact on the candidates' campaigns. Positive polls fuel enthusiasm and confidence for a front-runner, while negative poll reports cause campaign contributions to decline, endorsements to fall off, and staff energy and morale to wither.
Constant poll reporting also makes it much more difficult for the underdog to get his message out. Instead of being allowed to focus on issues and positions, the trailing candidate is hounded by reporters asking him to explain his poor showing in the polls and whether he will change his strategy.
Walter Mondale in 1984, Michael Dukakis in 1988, and George Bush in 1992 each faced one disparaging poll-driven question after another, as does Bob Dole in 1996. The questioning and the stories become self-perpetuating. Negative stories and sound bites saturate the news, contributing to the decided view of the underdog as ineffectual, a certain loser, even though these stories would never have been broadcast or written had the same candidate been leading in the polls.
Polls can also reduce the appetite of reporters to do tough stories on the front-runner. Because correspondents are often convinced of the outcome weeks before election day, some will back away from highly critical stories about the front- runner in the final days of a campaign knowing that they need to maintain good working relations with the new or reelected administration. Conversely, there is little or no such restraint shown toward the underdog because it is assumed that he and his staff will soon be virtually irrelevant as news sources.
Pre-election poll projections also rob voters of the sense that they are part of the process. In those instances where the presidential election may be the only significant contest on a ballot, one can understand why some citizens may choose not to vote if the news media have spent weeks chanting the same mantra that the outcome is already certain.
Just as the television networks claim that their election-day predictions based in exit-polling are completely benign - even though millions of voters in the Far West are routinely told by the networks that the presidential election has already been decided before the polls in their states have even closed - news organizations remain oblivious to their role in shaping the tenor of our electoral politics through incessant preelection polling.
Recognizing these problems, other countries have outlawed polling altogether, or for the final few weeks of an election. Such a solution seems extreme to some, and in the United States it would fly in the face of the First Amendment. Ideally, news organizations would stop conducting so many polls, and editors and reporters would exercise much greater restraint in using polls to both shape coverage and spike audience ratings and reader interest.
Although the competitive and dramatic elements of a campaign may often be more exciting, they are less substantive. We must ask ourselves whether incessant election polling furthers the goals and ideals of democracy, or merely the financial and competitive interests of media companies.
*Robert Kubey is associate professor of communications at Rutgers University in New Brunswick, N.J., and director of the Media Education Lab at Rutgers, Newark. Vincent M. Fitzgerald is assistant professor of communications at the College of Mount Saint Vincent in Riverdale, N.Y.