Polls apart

In one sense, it's just another pollster book. Barry Sussman's What Americans Really Think, and Why Our Politicians Pay No Attention (Pantheon, New York, $18.95) draws on his experience at the Washington Post, where he was special pollster for more than a decade. Typical of its kind, the book ranges across years of surveys. It discusses various issue areas, including Watergate, the Iran-contra affair, social security, abortion, Israel, and the news media. It's full of hard numbers and softer generalizations. Such pollster books, while not uninteresting, are increasingly common. They do remind us what Americans think - that people distrust government, want no more Vietnams, fear crime and drugs, and worry about economic malaise. But we knew that already.

Sussman does something more. He broods. Not about America, or the political process, or even President Reagan, whom he trashes unmercifully at every opportunity. No, what Sussman broods about is Sussman - and, in that way, about the whole business of pollstering. He doesn't do it overtly: This is no autobiography. But what nags at the back of his mind is a question as pressing as a subpoena and as murky as the Potomac: What's the value of my chosen occupation, and why have I become a pollster?

One wishes he had asked it plainly, so relevant is the question. It touches one of the most central and least understood phenomena of modern political life: the opinion survey. Our reliance on polls is astonishing. The public loves them and believes in them. Elected officials worship them and pay for them. Sussman tells us so. Then why, in this book, does a defensive tone creep in around the edges? Why is it that, when he insists that as a nation ``we are far better off with [polls] than without them,'' he sounds like a man convincing himself?

The answer, I think, lies in our widely shared but rarely voiced misgivings about polls. To his credit, Sussman names the obvious pitfalls in several useful but too-brief chapters: the problems of question-order bias (through which you can lead respondents into the answers you want), of interviewer bias (through which you can intimidate them out of giving answers you don't want), and of a poorly selected sample. But there are two less obvious ones that Sussman mentions only in passing.

One concerns complexity. Many issues, he writes, are ``complicated matters that are generally unsuitable for ordinary polling.'' That's true, he notes, in many tangled areas of foreign affairs. But it's also true of some domestic issues. ``When it comes to busing,'' he notes, ``as in so many matters dealing with the sensitive issue of race relations, what Americans really think is fairly complex.''

The uncomfortable conclusion? Sussman doesn't draw it. If he did, he would have to say that much of life is complicated, and that survey questions regularly elicit grossly oversimplified responses. And that, he would have to admit, is pretty bad press for his chosen occupation, which survives by reducing the three-dimensional complexity of human response to flat caricature.

Cause enough for brooding, perhaps. But the second point is even more telling. He notes that the great majority of people ``may give an opinion when pressed, but it is not worth very much.'' Translation: In many areas, there's no three-dimensionality to be found. Randomly sampled respondents have little to say. Yet we continue to press them with questions. So they give answers - perhaps because, when an organization so full of opinions as the Washington Post seeks your views, it's hard to admit you don't have any.

And that gets to the heart of the matter. Do we or don't we have a citizenry whose views are worth soliciting? Does polling give real answers to trivial questions and trivial answers to real questions? In other words, does grass-roots democracy work, or are people barely knowledgeable enough to elect better-informed citizens to represent them?

No wonder Sussman broods. So should we. Polls are here to stay. But the principles behind them are far from clear. The sooner we learn how and why to doubt them, the sooner we'll be able to keep them in perspective.

A Monday column

We want to hear, did we miss an angle we should have covered? Should we come back to this topic? Or just give us a rating for this story. We want to hear from you.