Those Fluttering, Stuttering Polls

To George Gallup, it is the mostunpredictable presidential election in his four decades as a pollster. Says Daniel Yankelovich: "Voters are in complete conflict. They will tell you one thing today and something else tomorrow." Muses Louis Harris: 'The voters out there are trying to tell us something." From all indications, their message may not be at all clear until they take part in the poll that matters most--the one on Nov. 2.

The ebb-and-flow character of opinion in 1976 has undeniably injected suspense into the campaign. But it also has intensified the hazards of voter sampling, producing wide disparities in the major polls and seriously unnerving the candidates and their chief strategists.

Perhaps most dramatic has been the plunge in Jimmy Carter's support. The Gallup poll showed him dropping from 62% (and a 33-point lead) in late July to 48% (a six-point margin) as of last week. Harris had him at 66% (a lead of 39 points) in July, which fell to 47% and a five-point edge last week. Yankelovich, who conducts opinion soundings for TIME, never gave Carter more than 48% and a ten-point lead, and right after the first debate had him running dead even with Gerald Ford.

Inevitable Descent. Carter's decline, however, is rather readily explained. Both Gallup and Harris gave Carter his biggest lead immediately after the Democratic National Convention in July, when his visibility was highest and when Ford was trying to fend off the challenge of Ronald Reagan. Yankelovich gave Carter 48% to Ford's 38% in April, and a 47% to 38% edge in June. The next Yankelovich poll in late August gave Carter 46%, Ford 40%. Gallup and Harris surveys taken at about the same time reflected Carter's inevitable descent from the heights, although both still gave him a substantial lead. All three polls had detected the same, expectable national trend away from Carter and toward Ford.

The variations in the most recent samplings can be accounted for by the standard margin for error in opinion surveys (3 points), by differences in polling techniques and by the fact that Gallup and Harris polled after the second debate, in which Carter did well. Also, Yankelovich does not try to push those who seem genuinely undecided into saying how they are leaning, while Gallup and Harris sometimes do. Thus the percentage of undecided in Yankelovich polls is generally larger. On the other hand, both Gallup and Harris try to weed out those who indicate that they are unlikely to vote.

But the greatest hazard for pollsters has been the volatility of the electorate in a year when neither major party candidate commands an unswervingly loyal national constituency. Thus relative trivialities (Carter's remarks to Playboy about lust, Ford's golfing trips from his congressional days) may prompt voters with a soft allegiance to one candidate to shift to an equally transient preference for the other. The debates have contributed heavily to the volatility --Ford gained after the first, Carter after the second--which underscores the importance of this week's third debate.

George Gallup, now 74, is still somewhat startled by the virtual evaporation of Carter's once-commanding lead, the largest such loss his organization has ever reported for a presidential candidate. Now Gallup sees indications that "Carter may be recouping his losses." The polls have "fluttered and stuttered," he says, because neither candidate has much stature in the minds of the voters --a fact that Gallup believes may result in an extremely low voter turnout next month. He argues that while Carter was seen as a conservative in the primaries, he appears more liberal when pitted against Ford. Says Gallup: "We are finding a strong trend to the conservative position not just in the U.S. but in the entire Western world." Gallup also thinks television, which shows both candidates "doing uninteresting things day after day," has turned off many voters and left others undecided.

Inner Conflict. Yankelovich stresses that a few days' difference in the polls can account for sizable variations. He also contends that there are two types of electorates: one that makes its mind up and stays put, as in 1972, when 60% of the voters had decided to support Richard Nixon before Labor Day; and the 1976 voters, who "are very unsure," torn by "inner conflicts" and who thus respond to a Ford gaffe one day, a Carter gaffe the next. "People are uneasy about Carter and find Ford an acceptable alternative," says Yankelovich. He emphasizes, as do Gallup and Harris, that polls are not supposed to predict future results. "The figures can only tell you how the voters feel at a given moment in time," he says, and the voters may not feel that way on Election Day.

Harris, who agrees with Gallup that the winner may well be determined by the size of the voter turnout, notes that polls have been marked less by zigzagging than by a persistent Carter decline. But the situation is so fluid that he plans to continue polling through Oct. 31 or even Nov. 1, the day before the election. "I feel that this election will be very close right down to the wire," says Harris. "But I don't think we are helpless in finding out what's going on."

Harris is undoubtedly correct. But a skeptical electorate can be depended upon to baffle the pollsters--and the politicians--until the ballots that will allow for no further shifting of views are counted the night of Nov. 2.

This file is automatically generated by a robot program, so viewer discretion is required.