American politicians and the trouble with public-opinion polls

Ronald Reagan, Walter Mondale, and ''Larry the Lobster'' all have one thing in common: Their survival (political or otherwise) can hinge on the public-opinion polls.

il4l,0,16l,9p The polls, which play a larger role in America than anywhere else in the world, are about to take center stage again in Election '84. This could be their biggest year ever. And for good reason. Savvy politicians have learned that good polls can mean the difference between victory and defeat.

There are, of course, all kinds of polls. There are public polls (like Gallup and Harris), and there are private polls used by the politicians (like Hart and Teeter). There are telephone polls and door-to-door polls. There are call-in polls and call-out polls. There are clip-out-the-coupon polls and by-mail polls and exit polls and - well, the list could go on and on.

Most important, however, there are some good polls and some very, very bad polls.

And that brings us to Larry the Lobster.

Larry can teach us something about polls - which ones we should take seriously, and which polls are only good for a laugh.

Larry, his claws tied shut and his feelers wiggling, was the subject of an NBC-TV call-in poll.It was obviously just for fun - part of the network's ''Saturday Night Live'' show. Actor Eddie Murphy hoisted Larry up and announced that listeners everywhere could vote on whether Larry should be cooked - or allowed to live.

''This is not a joke,'' Murphy solemnly declared. ''Vote to save Larry by calling 1-900-720-1808. Or vote to kill him by calling 1-900-720-1909. We'll be back a little later to see if Larry lives.''

The votes quickly poured in. After only a few minutes, an announcer gave the early, incomplete returns:

Save Larry, 56,893 votes.

Cook Larry, 65,743 votes.

One person who happened to tune in the show, Associated Press national political correspondent Evans Witt, said he became so interested that he couldn't turn off his set, even though it was past midnight. He had to know whether Larry would be dropped into a pot.

Such ''polls'' are all in good sport, veteran political analyst Burns Roper notes. But what if call-in polls were used for more serious political purposes, such as nuclear arms policy or Grenada? That could spell trouble.

Yet that is just what is happening, Mr. Roper says.

There are many examples. But the ones that have polling experts in a huff these days are those conducted by ABC-TV. On its late-night news programs, ABC has been promoting 900-number call-in polls (an ABC spokesman calls them ''straw polls'') on issues of great public import, even though such polls have no scientific validity.

Roper and other experts like Norman M. Bradburn, director of the National Opinion Research Center, charge that these ABC polls distort public opinion. They have ''more of a show-biz quality to them than an informing-the-public quality,'' Roper contends, and ''they should be sharply curtailed, if not eliminated.''

Are 900-number polls really that bad? The critics say ABC proved that when it recently ran a 900-number poll on whether the United Nations should get out of the United States. The call-in sample ran 2 to 1 in favor of ousting the UN.

But ABC also ran a parallel, scientific call-out poll that produced, in Roper's words, ''more than diametrically opposite results.'' When the results were tabulated from the scientific poll, it indicated that about 72 percent of the American people favored keeping the UN in the United States.

How can that happen? Because call-ins are next to worthless from a scientific standpoint. In pollster parlance, such surveys are called ''Self-selected Listener Opinion Polls,'' or SLOP. These polls are proliferating, which means the public is getting more SLOP, and less that is worthwhile.

There are other developments in polling - both positive and negative.

On the favorable side, competition has forced pollsters to speed up their processing of results. Polls today come out two weeks faster than they did four years ago, and that makes the data fresher and more pertinent.

The growing number of polls also gives the public a larger voice in local, state, and national affairs.

But there are trends less sanguine. Today more of the media - newspapers, TV , magazines - do their own polling. That raises the issue of conflict of interest. A newspaper, for instance, may push its own poll on Page 1, but ignore an equally valid - or perhaps, conflicting - poll by another newspaper. One wag says he has yet to see a TV network say: ''Yesterday we reported that our poll showed such-and-such.But a Gallup poll released today casts serious doubt on the validity of our poll.''

Another concern is the hurry-hurry syndrome. As noted above, data are being processed faster. That is fine. But too often pollsters are rushing to have the first poll after a major event. That may not be best, because while a quick poll may measure initial reaction after a major happening like Grenada, it may be more important to get the public's considered reaction several days, or even two weeks, afterward.

As the '84 campaign unfolds, how can voters and government policymakers protect themselves from influence by misleading polls? The experts offer these suggestions.

1. Disregard polls that are not scientifically drawn. There are many examples: White House tallies of telephone calls and telegrams for and against the President's latest speech; call-in polls on TV and radio; clip-the-coupon-and-mail-it-in polls in newspapers. All can yield distorted pictures of the public mood.

2. Check the source. Was the poll conducted by a professional group like Gallup? Even if the poll was scientifically conducted, does it come from a ''biased'' source? Campaign officials will sometimes release results of private polls that favor their candidates. But they may withhold disturbing data from the same survey.

3. Check the sample size. If only 100 people were interviewed, the poll is subject to large errors. Accuracy is expensive. A typical Gallup poll will sample about 1,500 adults.

4. Check for trends. A poll can be very valuable when one can look back over a period of months or years to compare the latest reading with others. For instance, do more Americans today favor capital punishment than in the past? Is Ronald Reagan's strength among Republicans and independents growing?

One more warning: Even with all these precautions, a poll can be very misleading. Some of the top pollsters in the country say that results can vary as much as 25 or 30 percent on specific issues because of small differences in the wording of the questions.

G. Cleveland Wilhoit and David H. Weaver, professors at Indiana University, studied this problem for the American Newspaper Publishers Association, and their booklet on polls offers several cases of widely disparate poll results on the same subject.

In May 1977, for example, a poll by Opinion Research Corporation asked: ''Do you favor the United States continuing its ownership and control of the Panama Canal, or do you favor turning ownership and control of the Panama Canal over to the Republic of Panama?''

Result: 78 percent for the US retaining control.

That same month, a poll by Cambridge Research Associates asked: ''Do you think the United States should negotiate a treaty with Panama where, over a period of time, Panama will eventually own and run the canal?''

Result: 51 percent opposed to such negotiations.

All of this is more than academic nitpicking. These are weighty matters in America, where the government is guided in large part by public opinion.

Polls, however, must be kept in perspective. Tom Bethell, Washington editor of The American Spectator, worries that some pollsters are politically motivated and shape their questions to get certain answers.He also frets that some politicians follow the polls too religiously. That is a poor way to lead, he suggests, and he asks:

''What do you think the polls would have shown inquiring whether some sort of military operation (in Grenada) would have been a good idea? We know what the answer would have been, don't we?'' Yet after the US landing there, the public supported it.

Oh, yes. One other thing. After staying up well past his usual bedtime, Evans Witt finally found out what happened to Larry the Lobster. The final vote: 239,096 to save Larry; 227,452 to cook Larry. Eddie Murphy exclaimed: ''Larry lives!''