1990: The Impact of Public Opinion Polls

© Cartoon courtesy of Doug Marlette.

[This article originally appeared in the Autumn 1990 issue of Nieman Reports.]

At the AAPOR meeting 10 years ago I talked about a user’s view of the polls. I talked then about my experience as an editor of The New York Times and of the enormous benefit which The New York Times/CBS poll had brought our reporters and readers, especially during political campaign show our own polling capability had freed us from dependence on self-serving analysis by candidates; had given us an independent check upon the course and integrity of the campaign process. We had, it seemed, achieved at least a part of the dream of progressive reformers that a disciplined, scientific approach to public opinion surveying would free the voice of the people from control by subjective party bosses and the tyranny of the smoke-filled room. Democracy of permanent referenda; constant accountability.

I know some of you were at the meeting because when my article appeared on the Op-Ed page of The New York Times last year, raising questions about the use of public opinion surveys by the press, I received a number of letters asking how I squared that article and that speech.

It was a troubling question. Consistency is not a hallmark of daily journalism, but now that I’ve shifted to a more academic setting it is a characteristic which seems to receive more attention. So I dug out my old speech and must admit I was relieved to find the roots of my present concerns in that speech. And they are clearly concerns which have only grown with time. As I put it then:

"All in all I guess it’s safe to say that I have become a believer in the careful use of polling in my work and fully understand the value of it as a tool to construct a better and more informative story. However, there are some things that disturb me still and these troubling thoughts have grown with the proliferation of polls in daily journalism."

Briefly, the concerns I listed then were:

First, the use of political polls as horse race reporting devices to focus on who’s ahead at a given point in the campaign.

Second, the impact on the sequential primary process by which presidential candidates are chose nan impact I feared could frustrate the democratic process as poll results created unrealistic expectations of performance or whipsawed public emotion by creating an almost daily contest of popularity which campaigns attempted to control.

Finally, whether the increased use of polling in the process was creating a closed, self-feeding system which reduced rather than expanded the public dialogue by including in the debate only those questions which attracted pollsters and campaign managers.

And I must report tonight, I do not believe those fears were misplaced. If the 1988 presidential campaign did anything it fundamentally challenged the hope that public opinion surveys would strengthen the public’s informed participation in the process. In the high tech political Star Wars of 1988, what the political process produced for the people was either a paid advertising media visit with old friends, or the product of a reported media which was mesmerized by and incredulous of the extent to which campaign organizations were capable of dictating the context within which the electoral decision would be made. We may have been able to blow away the smoke which filled the rooms in which political decisions were made. But we have replaced it only with a carnival sideshow house of mirrors in which a potential voter is hopelessly trapped in a disorienting hall which is reflecting and re-reflecting the same images.

The real shift in the political campaign of 1988 was the degree to which the independent judgment of the editors and reporters covering the campaign was neutralized by campaign strategies and tactics. This domination may best be represented by the extent to which the dominating themes struck early during the campaign Willie Horton and the Pledge of Allegiance began as paid advertisements, but became the focus of the news reports. A paid media based on focus group analysis of emotional appeal in the end determined the news agenda.

The tools of persuasion and manipulation are awesome and they are cynically used. And just as we use many public opinion surveys to measure the most obvious questions, these tools are used to measure prejudice, but never to plumb understanding or the level of awareness behind the opinion. The political system adjusts its messages to appeal to these measured emotions in order to move the opinion in one direction or another suppress or increase its intensity.

But where is the system which attempts to counter this process of using public opinion to round up and herd voters like so many cattle, with the liberating force of information? To use these tools of measurement more effectively and creatively to balance appeals to emotion and prejudice with knowledge which offers understanding and balance?

Increasingly our use of technology the computer, television, demographic targeting permits individuals, isolated in their private places, to respond to direct appeals. Less and less is it necessary for citizens to attend public places and meetings in which his or her interests are put into a public context are required to relate to the needs and desires of a neighbor. The context within which personal opinion must contest with public responsibility. Such personal isolation, which encourages selfish concerns and threatens democracy with a mean-spirited Balkanizationa competition for power uninspired by a sense of community or common good.

Advertising by candidates fine-tuned on a daily basis, when necessary, by continuous tracking polls and combined with the ability to create targeted audiences about whom increasing amounts of information is known offers constantly expanding opportunities to manipulate public opinion on the basis of narrow, tightly focused, and highly emotional issues.

And into this volatile atmosphere the American press contributes what? Essentially a measure of the relative success of the manipulations. A study by Gary Orren at Harvard University found that almost 50 percent of the political stories which ran in three newspapers during a period of 22 days preceding the 1988 presidential election, cited poll results. During the four months from September 1987 through January 1988, 113 horse race polls were published on the Republican candidates and 123 on Democrats all before the primary season officially began.

Each of these stories arguably came at the expense in time, thought, energy, resources and space of a story which could provide basic information about the state of our society, about the issues confronting us, about the alternate solutions which might be considered, about the true state of our personal tax burden or the military budget or the quality of education.

The question now is whether the press has the time or the will to concern itself with its contribution to education and understanding.

In most news organizations the journalists who plan campaign coverage are, by and large, a sandlot pickup team which comes together annually at best (usually only quadrennially) to plan a strategy for covering "this year’s elections."

And when they do they are stepping into a world in which the opinion researchers, advertising strategists, public relations packagers, have been working day and night for the preceding year or two or three under the prod of competition to find new and better ways to sell an idea, create a demand, understand a market. Even the best, most dedicated journalist under such circumstances is a babe in the woods at the beginning of each campaign he or she covers.

And while journalists are trying to pull together yet another ad hoc system, the system of political manipulation is recruiting from the most successful marketing and advertising companies in the world. They even have now an advanced training school. The Graduate School of Political Management in New York school which features disciplined academic study in polling, political management of the media, campaign advertising and promotion, demographic targeting, and "using polling information and orchestration of the news." A school whose funding by Philip Morris and Ford Motor Company reflects the growing commitment to and investment in the process of political manipulation by corporate America.

In a correspondence from my friend Adam Clymer pointing out inconsistencies in the article I had written for the Times last year, Adam said:

"[T]he best reason for public opinion polling on issues lies in the nature of our society, i.e. a Democracy. I think people do have opinions and more thoughtful ones than a lot of their governors believe. I think the people should have something to say about how they are governed."

As with much of what Adam believes about journalism generally and polling specifically, I emphatically agree with that observation. But I find myself increasingly concerned with the role of the press in a self-governing society to provide the information upon which those opinions are based. And here is the nub of my criticism of the use to which the press has now put the instrument of public opinion sampling:

The focus is too narrowly fixed on the process and the course of the campaign the dynamics and thus the excitement and entertainment value inherent in campaign coverage.

Public opinion surveys could be of enormous benefit to a responsible news organization’s approach to all coverage, not just campaign coverage. But not the way we use public opinion surveys. Not simply to learn who is ahead and how this or that issue is cutting.

In 1988 the print press seemed bewitched by the made-for-television nature of the campaign and offered reams of copy about the staging of events, the manipulation of candidate behavior, and the crafting of personal images. How is a citizen to make an intelligent decision on the basis of such understanding?

Into the vacuum left on reporting issues of substance the public opinion survey is thrown. But to determine what? What issues seem to be important? How many agree with the Bush or the Dukakis position? What racial, ethnic, or regional appeal each candidate has, and so forth. My question now is: How much is a bushel basket full of such data to a voter? What clue does this give a potential voter on the ability of one or the other to manage the S&L crisis or to address the quality of American education?

Our democracy is not merely a matter of registering preconceived notions and opinions of individuals. That approach was rejected in the design of our government in favor of a representative system by which the matters of government would be debated and issues resolved by consensus achieved by compromise. Responding to opinion in terms of its public impact not its private attraction. Most public opinion surveys now conducted by news organizations, in effect, record private opinion, and the consumers confuse the results with public thought. Few surveys examine the depth of understanding behind an opinion or the context of an opinion. When, for example, was the last time you saw a report on political opinion which required a response to the same question in more than one context? Questions which require the respondent to consider the consequences of an opinion?

I will admit to being narrow-minded, even simplistic, in my concept of the journalist’s role in a democracy. I believe it is the journalist’s role to inform public opinion. That in a self-governing society, the daily press is the only widely available system of education we have. If the voters are to receive the information they need to make informed choices on issues which confront them, it must come from the press.

To fill this role which I believe is the only one which justifies the protection given the press in the Bill of Rights then the press must know the extent to which opinion is based upon prejudice, emotion or information. The press must know what information the public needs in order to make more informed judgments. For that opinion informed or not is likely to be transmuted into a political position from which laws and policy will be fashioned. To the extent that the opinion is fathered by prejudice and ignorance so too the law or policy will harbor the same public poison.

So I worry that our current use of public opinion surveys does much less than it can to fulfill the public purpose to which political coverage is committed. Rather, because it focuses on the surface movements of opinion rather than their informing depths, I am afraid the press is unwittingly a part of the process of manipulating opinion devised by the political campaigns; that by concentration on a constant measuring of the success or failure of a campaign we have become a sort of extended focus group another in the corridor of mirrors in the campaign fun house yet again reflecting the same light in a closed system not introducing new lights to the process.

It is part of the old question: Do opinion polls shape opinion or do they measure opinion? I think a compelling argument could be made that in the absence of strong and sustained reporting on the facts underlying an issue polls can and do shape and create opinion.

And by measuring campaigns in terms of the questions they ask of themselves, I am concerned that the independent polling strategy, which I felt 10 years ago had freed the press of dependence upon the subjective reading of polls by campaign organizers, may now be making the press an even more integral part of the strategy of the imaginative campaign manager.

Hannah Arendt has said: "Freedom of opinion is a farce unless factual information is guaranteed."

Judge Learned Hand has said: "We have staked everything on the rational dialogue of an informed electorate."

As the fragmentation of American society hurtles ahead, we are increasingly becoming a nation of individuals who share less and less common information. Advertisers purveyors of information designed to influence our taste and our economic behavior seek to fix us, like insects impaled in a collection tray, as part of a narrowly defined group with set prejudices, tastes and desires. As these groups become more clearly separated one from the other narrowly focused vehicles to reach them with tailor-made messages are designed. The result threatens a constantly shrinking pool of Americans who begin each day with some sort of shared knowledge and understanding of events and issues and experiences.

It is imperative to the continued health of self-government that the press compensate for this trend. That the daily reports keep filled the common pool of information and shared experience of the body politic. In this endeavor public opinion surveys could be of enormous benefit to a responsible news organization’s approach to all coverage, not just campaign coverage.

Consumers of news, I am convinced, look to the daily press for information which they can use. Information which helps make a confusing and complicated world a little easier to understand, to confront. A news report which very simply helps them make it through the day. During the election season the potential voter depends even more fundamentally upon the daily press. To help editors design such news reports, public opinion surveys can be key tools. But not the way we use them now. Not simply to learn who is ahead or the appeal of this or that issue.

What is missing from our use of the tool is the key ingredient of context upon what information or misinformation is the opinion based? What do the people know about the issues? And knowing this the press can then perform its most vital function providing the information for enlightened self-government.

This past May in Lancaster, Pennsylvania, Bill Kovach, Nieman Foundation Curator, gave this talk at a meeting of the American Association for Public Opinion Research.