Often Distorted Early Polls Supply Fuel to Political Fires

Science always has been something of a cold-blooded intruder into the instinctive world of politics. This is evident once again as the science of public-opinion polling handicaps California election races still a year or more away.

Polls that supposedly tell us who is ahead and who is behind amount to the most tangible political currency in circulation this early in the election cycle. Everyone with even a passing curiosity reads them. And people who live politics are absorbed by them.

Polls, in short, do not just set the agenda this early before an election. In many ways they are the agenda.

And that is cause for worry, not only among politicians and those who watch them and work for them, but surprisingly, among pollsters themselves. They see that opinion samplings this many months before an election frequently are given too much importance, often are misunderstood and occasionally are misused.

‘I Have My Doubts’

“Basically I feel uncomfortable with these polls. . . . I have my doubts about them,” conceded Mervin D. Field of San Francisco, whose 40 years of public opinion sampling make him the dean of California pollsters and one of the most recognizable names in the business.

What bothers Field is that early polls such as his tend to glorify the status quo and “freeze the dialogue.”

He continued: “These polls very early on, unfortunately, have a disproportionate influence on the (political) insiders. What happens is that they are misread. Early polls generally can enhance the position of someone with better ID at the expense of someone who doesn’t have such good ID. . . .”

“Frequently,” he added, “they don’t have a lot of meaning.”

For just such reasons, pollster Lance Tarrance of Houston advised one of his clients, Assemblyman and U.S. Senate hopeful Robert W. Naylor of Menlo Park, not to bother polling so far in advance of the June, 1986, primary. “Polls shouldn’t be thrown around like confetti,” Tarrance said.

I. A. Lewis, director of the Los Angeles Times Poll, is emphatic.

“Any poll taken more than six weeks before an election is a waste of time as far as learning anything,” Lewis said.

Still, who can deny the power of little rows of numbers, encased in objective science, that hold promise of insight into ourselves? Who can deny their allure? Who can deny their impact?

We are, after all, what pollster Field described as “a nation of scorekeepers.”

Demands for Poll ‘Data’

Contributors demand poll “data” before opening their checkbooks. Reporters peruse polls before opening their notebooks. And for the growing legions of political staff professionals and consultants, polls are The Book.

That is the paradox.

The public’s views on far distant elections may not, in themselves, be well thought out, complete or of vast consequence. But take these opinions, cast them in science and publish them for all to see, and the results take on great significance.

Before the last election for governor, for instance, The Times’ Lewis surveyed voters on the subject. He found that voters answered differently depending on whether they were aware of current poll standings. People who knew of polls were more likely to favor the reported front-runners than people unaware of survey standings.

In at least two cases in recent years, campaign managers believe they were able to influence the outcome of Field’s California Poll to the advantage of their candidates.

The first was in 1977-78, when state Sen. Ken Maddy of Fresno ran for the Republican nomination for governor. Campaign aide Doug Watts, of a consulting company called Russo-Watts, said he went to libraries and determined when Field might be expected to conduct his interviews with voters. “Then we bought time and conducted media events for the days just before we guessed he would be in the field . . . and it turns out we weren’t far off,” Watts recalled.

“All we were doing is trying to raise money. And the way to do that is to show well in the Field Poll,” Watts continued.

Something worked. A relatively unknown candidate at the start, Maddy gained quick credibility in the race. And by April of 1978, Field himself wrote: “Maddy has made a sizable gain in the standing--his support tripling in the last few months.”

Maddy went on to raise enough money to make other candidates jealous and finished a strong third in the crowded race, collecting 20% of the Republican primary election vote, quite an achievement for a Central Valley lawmaker.

The fourth-place finisher, then-San Diego Mayor Pete Wilson, was one of those who was impressed. Aide Otto Bos recalled that Wilson, who got only 9% of the vote in the 1978 gubernatorial primary, tried to copy Maddy’s poll-conscious strategy at the next election, in 1982, when he ran for and was elected to the U.S. Senate.

Anticipating a Poll

Again by researching Field’s patterns over the decades, Bos said that Wilson, as the campaign progressed, decided to spend “the entire bankroll of $200,000” in television advertising just before he anticipated a poll would be conducted. Bos said the campaign beat the start of Field’s survey by a week.

“We shot up 6%, so something worked,” Bos recalled.

Field hotly objects to any suggestion that his poll can be second-guessed and manipulated.

“If this was a small state, you could make a nobody a somebody,” he said. “But you can’t do that here.”

Field added that there may be “some truth” to candidates moving up in his poll via early advertising. “But for every case like that, I can show you many others where it didn’t work.”

Often the use of results of early polls is less sophisticated. For instance, Rep. Daniel E. Lungren of Long Beach sought to increase his standing among Republican voters at last winter’s state GOP convention in Sacramento by papering the convention hotel with scores of copies of Field’s first poll in the 1986 elections. It showed him as one of the least-disliked contenders for the 1986 primary election for the U.S. Senate.

Only 2% of the Republicans queried had anything unfavorable to say about the congressman, while 15% said they had a favorable reaction. That ratio of favorable-to-unfavorable was better than for any other potential GOP candidate--enough to drive Lungren to the nearest photocopy machine, never mind that fewer than one in five voters even recognized the congressman’s name.

“If the polls have some good news, you cannot afford to ignore it,” Lungren said.

Over the years, pollsters themselves have pondered how to stay in business, remain interesting to the public but reduce the potential for polls to distort, not just measure, voter sentiment.

One of the boldest such efforts involved Lewis in the 1976 presidential elections, when he was director of polling for CBS News, which joined with the New York Times to produce election surveys.

“That year we never said who was ahead. And that was because there is a lot to be said for de-emphasizing the horse race aspect of polls,” Lewis said. Instead, the poll stuck with how candidates were doing with various interest groups and in different regions, but not the sum of all the parts. “It got to be absolutely ridiculous,” Lewis recalled.

More recently, Field said it was a “good question” to ask why he continues producing his horse race polls for newspapers and broadcasters even as his doubts about them linger.

‘Editors Want Them’

“I started out without the feelings I have now,” he said. “. . . Now, frankly, we need the financial support in order to continue our other activities. And editors and news directors want them. They are closer to their audiences than I am.”

And besides, the veteran pollster said, “there have been enough cases recently where candidates who didn’t show well in early polls stuck with it and ended up doing quite well to ameliorate some of my concerns.”

If polls have a distorting effect on campaigns, many people believe that it is not the fault of the data but of those who read and use it.

“I think the biggest impact of these early polls is on the press,” said Tom Quinn, who has managed campaigns for two Democrats, Los Angeles Mayor Tom Bradley and former Gov. Edmund G. Brown Jr. “Reporters tend to take seriously those candidates who do well in polls and dismiss those who don’t.”

So, how do pollsters read polls?

Pollster Gary Lawrence of the Santa Ana firm of Decision Making Inc., which has done major polling for President Reagan, shared his formula: “I’m not interested in which horse is ahead. I’m interested in how fast each horse is moving in relation to the others at that moment.”

Far outnumbering the published horse race polls that the public sees are private polls commissioned by candidates.

Field estimated that for every $1 worth of polling data seen by the public, another $9 to $10 worth is undertaken and kept private by candidates.

In fact, the most compelling argument for public polls such as Field’s and the others conducted by individual news organizations is to act as a check on the selective and always self-serving disclosure of private polling information by candidates.

“A public poll puts all the data on the table for everyone to see,” Field said. He observed that the selective and sometimes widely misleading leaks of private polling results occur “constantly” in campaigns. “That’s worse than the early public polls.”

Private polls this early in the election cycle, according to pollster Lawrence, can serve several needs for candidates, depending on the situation. In the case of an incumbent running for reelection, research--which is what most pollsters now call their work--can locate “vulnerabilities.” Or, in the case of a crowded field and no incumbent, Pollster John Fairbank, of Fairbank, Canapary & Maullin, said that early campaign polling before the 1982 election helped “validate” candidate Bill Honig’s belief that the public was ready for major educational reform in California, before it was obvious to others in politics. Honig went on to be elected superintendent of public instruction in California.

Some candidates use polls for much, much more.

For instance, one candidate who has become strongly identified with the use of early polls is Ed Davis of Valencia. The Republican state senator and former Los Angeles police chief has a widely known reputation for political independence and strikes many as the kind of square-jawed politician who has little doubts about where he is going. So why does he refer to polls-this, or polls-that in almost every political discussion?

“Polls are a very indispensable part of the planning process,” he replied. “I like to know if I’m ahead of so-and-so, or behind him.”

Davis recalled that in 1978, when he ran unsuccessfully for governor, his campaign handlers urged him to “try to hide the fact that I was a policeman.” Now, he said, thanks to polls, he can face voters utterly confident “that one of the things they like about me is that I was police chief. I can be proud of my past.”

Because of the widespread reliance on polls, perhaps it was inevitable that some iconoclasts in politics would find success by ignoring them and mocking their wisdom.

Against the Grain

In California, this honor goes to the campaign duo of Michael Berman and Carl D’Agostino, who manage the political arm of the Westside Los Angeles Democratic organization of Reps. Henry A. Waxman of Los Angeles and Howard L. Berman of Studio City.

Last year, the pair engineered the successful Democratic campaign that beat Republican Gov. George Deukmejian’s ballot measure to reapportion the state Legislature and congressional districts, a matter of political life and death for some incumbent Democratic officeholders.

Because of the multimillion-dollar cost of the television advertising campaign and the high political stakes, Berman and D’Agostino said that some Democrats who provided the financing also insisted upon commissioning a major poll of voter attitudes.

According to the two, the results of the poll suggested that Democrats should campaign on issues of “fairness” of the governor’s plan and how it was a partisan matter. As it turned out, this was the very thrust of the campaign Deukmejian himself mounted. Meanwhile, Berman and D’Agostino deliberately ignored the poll and shaped a campaign that focused tightly on how judges would be drawn into politics if they were given reapportionment powers, as Deukmejian’s plan had it.

These days, the two hold up that example of how a campaign can be more creative in shaping opinion than a poll in trying to measure it.

Still, Berman and D’Agostino are a decided minority. Interviews with nearly 20 pollsters and campaign experts found all but a handful believed that candidates should spend in the vicinity of from 10% to 20% of their early budgets on survey research.

“We have bred a race of people,” pollster Lewis said, “who cannot understand politics without a poll.”