Stacking the Deck?

ABOUT 60 PERCENT SAY NO, 30 percent say yes, and the other 10 percent aren't sure. The question is whether polls affect the outcome of presidential elections, and the experts are divided.

Those who say that polls influence the electorate rather than simply mirror its attitudes cite three main trends; the bandwagon effect, the underdog effect and the media effect.

Those who say political polling generates a bandwagon effect claim the surveys cause voters and financial backers to follow the crowd and abandon unpopular candidates.

"Polls affect things rather dramatically," says Dan Callegari, Sen. Gary W. Hart's (D-Col.) Manchester, N.H. campaign coordinator. "If you're not showing strength in the polls, then you have to sell your candidate twice as hard. Mondale's campaign theme is that he is doing best in the polls."

Evidence suggests a correlation between success in the polls and success in fundraising. Those with money to donate are less likely to invest in the unpromising future of a candidate struggling in the polls. Contributions translate into political favors only if the candidate gains office.

"The strongest effect of these polls is to financial contributions," says Adam J. Clymer '58, assistant to the executive editor in charge of polling at The New York Times. "The people who give money are not the average voters, and they want to be winners."

During the 1976 presidential campaign, Jimmy Carter did not receive ample contributions to promote a viable campaign until he made a huge jump in the national polls, Clymer adds.

It is more difficult to show the effect of poll standings on a candidate's success on election day. One branch of the bandwagon theory predicts that supporters of those performing poorly in the polls often fail to vote because they believe their cause is lost.

Another major theory predicts a diametrically opposite effect. According to the underdog theory, a large portion of the electorate, believing that victory is assured, simply don't bother to vote. Supporters of this theory also contend that because people naturally sympathize with the underdog, voters may cast their ballots out of pity for, rather than in support of, a candidate.

Many people subscribe to neither theory. "I don't think that people vote for the loser because of polls," says Michael Barone '68, formerly vice-president of a private polling agency and now a Washington Post editorial writer. But "the ordinary voter does not get a huge charge from being on the winning team," Barone adds.

To the extent that there is a bandwagon effect, Barone says, it is unfair to blame it on polls. "People don't live in a vacuum where polls are their only access to reality," he contends. "The poll is only one of many items that gives people information to who is ahead. The media, word of mouth, neighbors, advertising, leaflets, door-to-door campaigning--all relay certain impressions to the voter," he adds.

Experts agree that both underdog and bandwagon effects tend to be stronger in primaries than in general elections. "Most of the evidence suggests that the polls in general elections have a very limited bandwagon or underdog effect," says Gary R. Orren, a Kennedy School professor. In a general election, partisanship provides a broad base of support that "outweighs" any potential effect the polls might have, Orren adds.

This partisanship does not exist in primaries, and the differences among candidates may not be as well defined and tangible as they are in general elections. "In these primaries," Orren says, "people tend to vote strategically."

In a primary there is a tendency not to "waste votes," Clymer says. Some may gain comfort in casting "symbolic votes"--ones for candidates whose ideas the voter agrees with but who have no chance of winning. "But others could say, 'I should vote for the guy I like second best because my guy has no shot, and I don't want my least favorite candidate to win,'" Clymer says.

Some analysis maintain that even in primaries, the amount of strategic voting is relatively small. "The public polls which appear throughout the media are, by and large, a product of the spectator sport element of politics," says Geoffrey D. Garin '75, president of Peter D. Hart Research Associates, a private polling company, whose clients include Democratic frontrunner Walter F. Mondale.

"These polls do not have a very direct impact. For instance, you may end up rooting for the Yankees because they're in first place. But most people do not give their loyalties to a team--or a candidate--based on its standings. Even when there are many candidates," Garin adds.

Whether or not voters are directly effected by polls, polls may govern the amount of media attention each candidate receives. Many people believe that, if nothing else, there is a strong media bandwagon effect.

"The only thing the press covers is polls," Callegari asserts. "If you're not a frontrunner then the polls don't mean anything. When they show some upward movement that's when things start ticking and you receive some cover age. The media is not that ambitious to cover...those who aren't high in the polls," he adds.

Members of the media and many other observers claim that polls don't have the virtually limitless influence Callegari suggests. Barone points out that The Washington Post has run editorials and articles about Reubin O. Askew, who has fared poorly in the polls.

"Askew's gotten coverage from us," Barone says. "If he doesn't do well, it's not really the fault of The Post. He's had his chance to make his case to the public."

Garin says that though Jesse L. Jackson has tallied less than five percent in national preference polls, he has received abundant media coverage.

"True, you can get extensive coverage because of the polls," Barone says, "but also by being interesting. Jackson's not popular, but he's covered. The polls are only a small part of what makes media coverage."

The many democratic forums, debates and road shows have resulted in a moderate amount of exposure for each candidate, Orren says. "I think that the media coverage has been rather evenhanded," he adds. "There's no evidence that Mondale is getting excessive coverage. The voters know who the candidates are."

Whether or not polls affect the outcome of elections, candidates faring poorly in polls have traditionally sought to discredit them. During the 1968 presidential campaign, when most polls showed Hubert H. Humphrey trailing Richard M. Nixon by a wide margin, Humphrey once called a press conference to change that the Gallup Poll the views of Blacks. More recently, Boston mayoral candidate Melvin H. King complained that he underperformed in polls that failed to include newly registered voters among those surveyed.

The French have settled the great poll debate by law; it is illegal to publish poll results during the week preceding French presidential elections