Booknotes

DAVID MOORE, AUTHOR, "THE SUPER POLLSTERS" Probably the first one was a straw poll in the 1820s, when a newspaper on the East Coast wanted to find out in the local election who was going to win. The reporters went out and talked to a bunch of people and reported it in the newspaper. That was probably -- that's the first one we know of.

MOORE: Well, the first scientific poll, or at least the most important scientific poll that was conducted, was in 1936, when George Gallup was conducting a series of surveys, actually, for subscribers. He started his column called America Speaks, and he predicted then that Roosevelt would win the election, while the Literary Digest, using straw-polling techniques, predicted that Alf Landon would win the election. And of course, George Gallup was right, using these new scientific methods. Now it wasn't the first scientific poll, but it's the most important threshold of when we consider scientific polling to begin. And so, for the most part, George Gallup is considered by most pollsters as the father of the American polling.

MOORE: Well, George Gallup was originally a journalist. He worked in market surveys, he started doing some market surveys. And in 1934, actually 1933, a year before the 1934 election, he said, "You know, this scientific stuff is pretty good. I think I'll ..." -- he sent out a bunch of ballots or letters to what he considered to be a scientific sample of people across the country, so that he could predict the congressional elections a year later. And oddly enough, even though it was a full year difference in the polling, his results were very, very close. So he figured, "Well, I've got a neat technique here; now what can I do with it?" And that's when he started doing his political polling for the newspapers. And of course, today, everybody -- certainly the Gallup organization, a major international polling organization, and George Gallup was clearly the person who thrust polling into the American scene.

MOORE: No, he is dead, and his wife died a couple of years ago.

MOORE: Originally from Iowa. He went to school in Iowa and then eventually moved to Princeton, and Princeton, New Jersey, is where his name is most closely associated with. And, by the way, the reason he moved to Princeton, starting his American Institute for Public Opinion in Princeton, New Jersey, he thought that the name Princeton would be associated with Princeton University. So he thought that would, to the extent that it might help him, get people to respond. And since then, his organization's been there.

MOORE: No, I didn't have the opportunity to. But he has two sons now, George Gallup Jr. and Alec Gallup, who are still major players in the Gallup organization, still at Princeton University.

MOORE: I'm not quite sure what kind of a firm it is, but a larger firm called SRI that owns the Gallup organization. But the Gallup organization still does conduct these political polls, although SRI, for the most part, is market-research oriented, not oriented toward the political policy polls that the Gallup organization has done.

MOORE: Well, a straw poll is a poll where people participate in the sample because they want to; call-in polls, 900-polls, for example, or if you mail out the questions and ask people to respond -- some do, some don't. Unless you have a very systematic follow-up to get everybody into it, then you've got a distorted sample in a way. But a scientific poll is one where the pollsters go out, reach the individuals and essentially try to persuade them to participate so they get a good cross-section of everybody across the larger population, so...

MOORE: Right.

MOORE: Well, there are two kinds of manipulation. One is when you take the information itself and try to use it to manipulate public opinion. And for the most part, that's essentially what campaign pollsters want to do. They want to get information about what the public is thinking, what they're feeling, and then try to tailor-make their candidates' responses to fit those feelings. The other kind of manipulation is when a pollster will ask questions that are biased or perhaps not reveal all the information. And probably the biggest example I give -- or the most, I would say, flagrant example I give of that kind of manipulation was from one of the major pollsters in American history, and that's Lou Harris. He's a very highly respected, very creative, very innovative person. But in 1980, he was working for ABC -- he was doing polls for ABC during the election -- and during the period from January through June, when Ted Kennedy was challenging Jimmy Carter for the nomination, Lou Harris, who favored Ted Kennedy, would report the polls and the results that were favorable toward Kennedy. But when there was a surge of support for Jimmy Carter or, in some cases, when there was negative information about Ted Kennedy, Lou Harris either would not present that right away -- wait for several weeks -- or not present it at all. I give several examples of where he just didn't present his negative poll results toward Kennedy at all. And that is a very unusual situation to find, particularly for a pollster who would be working for the media.

MOORE: Well, I just came across it. It's not anywhere in a book or anything. But all of the information -- all of the polls that Lou Harris has conducted are now archived at the Lou Harris Data Center at the University of North Carolina. And you can actually -- through your computer, you can dial in and get the polls, the question wordings and the actual results. Also what's available, which I got from the Lou Harris Data Center, were all of his reports during that same period of time. So I could compare what his polling results were, what his questions were, what he got for polling results and what his reports were. And I will say he never lied; he never gave results that weren't there. But what he did do was sometimes not present results, and sometimes, in the case when there was a surge of support for Jimmy Carter, he waited two or three weeks before he would finally reveal it, and usually then only once Kennedy maybe had regained some of the ground. It was a very unusual set pattern, but it was so clear that anyone who was taking a look at it would come to the same conclusion I did, that this was a clear -- it must have been a deliberate effort to try to manipulate public opinion.

MOORE: As far as I know. No one else has ever mentioned it. Lou Harris has always had -- a lot of pollsters get reputations, and Lou Harris, in particular, for having a bias toward being more liberal. And you could tell, in the questions, sometimes, if the questions seemed to favor one side or another. But that's straight up. Somebody presents the questions. They say, "This is the way I ask them; if you've got a better way of asking, you can do it." But this is very different, because this is a case where nobody had really looked behind the scenes and looked at the actual poll results and the reports to see whether or not the poll results were adequately reflected in the reports. And I was, I must say, flabbergasted when I saw the kind of bias that was coming out during this period of time.

MOORE: When? Last fall. In about October or November is when I was -- it was part of looking at ... I have a chapter there on Lou Harris, and he's done a lot of creative things. He was the first pollster for a presidential candidate and so on. But I happened to be looking at a time -- I had heard some people say, "Well, you know, he was really biased because he for Ted Kennedy." I thought, "Well, you know, people can say that, but what evidence is there?" And then, when I read that the University of North Carolina had the polls there, and I called up and they said, "Yeah." I thought, Well, I'll take a look. And I looked at a couple and it looked strange. And I looked more and more and then I saw that there was a systematic pattern. It's not been very long ago.

MOORE: No. I didn't talk to him about it. I had interviewed him earlier, but I didn't go back to him and discuss it with him.

MOORE: No. Not as far as I know. It's in the book, and maybe this is the first that it's been expressly mentioned. I've mentioned it a couple of times when people have interviewed me, but no one has really done anything with it. Lou Harris retired from Lou Harris & Associates in January of this year. He decided to form his own organization. He had been, of course -- it had been his own organization, Lou Harris & Associates, but it had been bought out many years ago by one company and then another company. And so, while he was still CEO of his own organization, he really didn't have as much control over it as he said he wanted.

So, after turning 70 in January of this year, he decided he was going to go out and start his own research organization. So there's a new organization, called LH Research, and that's what he is starting, ginning up right now. I don't know -- I haven't heard of polls done by LH Research, per se, but my impression was that he was going to try to focus on international polling. But I don't know. He was a little bit vague when I did get a chance to talk to him, very briefly, by phone, about his new organization.

MOORE: Well, Lou Harris had been in the Army. He went to the University of North Carolina. He, after getting out of the Army, started working with Elmo Roper. When he left Elmo Roper in 1956 to start his own firm, it was a kind of a controversial move, which he himself acknowledges. He said he took three of Elmo Roper's clients -- he said because they wanted him to. Elmo Roper, I don't think, felt that way, because, as I point out in the book, Roper was very upset with Harris for having taken his clients and starting his own firm. And then, of course, he was in New York City in Rockefeller Center for all the years since he started his own firm.

MOORE: In 1960, he was the first presidential pollster, and he worked for John Kennedy. There is some report that a Claude Robinson worked for Richard Nixon at the same time in 1960, but there's nothing on the record that shows exactly what Claude Robinson did. But Theodore White, who covered -- in his book, "The Making of the President," in 1960 -- covered a lot about Lou Harris and what he did. And Lou Harris himself, of course, went on for, you know, a much longer career afterwards. Claude Robinson died in 1961, so we don't precisely know what he did. But that was one of the candidates that Lou Harris worked for. And he said he worked for 240 candidates in a three-year period, which included candidates from Canada and the United States. So after three years of doing this, he said he felt like he'd aged 240, so he decided to stop doing political polling -- that is, partisan candidate-oriented polling -- and instead do client polling for businesses, but also public policy polling. You can't really do both. You can't release your polls as a public policy poll at the same time that you're working for a candidate, So he decided that he wouldn't do any more candidate-oriented polls.

MOORE: Kennedy used the polls in two ways. He used them during the primaries to try to figure out which primaries he would enter. At that time, party bosses still made the decisions as to who the candidates were going to be, and so a candidate like John Kennedy, who didn't have support of the party leaders, would go out -- he would have to do the same thing in 1960 -- would go to certain primaries to try to demonstrate their electoral strength, that they have some ability to draw votes. And Lou Harris was helpful in identifying the states that might be most beneficial for Kennedy. During the general election, among the many things that Harris did was to identify those states that were very strongly anti-Catholic. And if they were anti-Catholic, Kennedy just simply wasn't going to campaign in any of them according to Lou Harris. He also identified some themes that Kennedy was saying, themes about education and so on, that turned out to be very important. And as we know, the election was, you know, whisker thin, to the extent that Lou Harris did help Kennedy in any area, just like any other one of those factors that could have been crucial in helping Kennedy to victory.

MOORE: Right.

MOORE: Well, that's probably the way his polls showed; 51 to 49 -- it was really 50:50 essentially. No one would argue with a margin of accuracy of that level. In any kind of a national poll of that level, it's too close to call. There are some criticisms, or at least reservations, that people've expressed. Theodore Sorensen expressed reservations about some Harris polls, saying that he felt that the results Harris got seemed to reflect more what Harris thought Kennedy wanted rather than what was truth out there.

MOORE: Theodore Sorensen was the campaign manager for John Kennedy at the time and later worked in the Kennedy administration. And then later wrote a book called "Kennedy," a running account of the Kennedy administration. But on the whole, I think the assessment was that there had been some mistakes. This was the first time around, and then, on balance, Lou Harris' efforts did prove beneficial to Kennedy.

MOORE: Well, I went to high school in California. Before that my father was in the Army and moved around quite a bit.

MOORE: In New Hampshire. The University of New Hampshire is where I work, and Durham, New Hampshire, is where I live.

MOORE: Probably in graduate school. I went to Ohio State and worked with a guy there named Dick Hofstetter, who was doing, at that time, face-to-face polls in the Columbus area. That was in '67 to '69 framework. And I was still in the Army; I was going back for a master's degree and ultimately a Ph.D, so I could go back and teach at the military academy. And after I got out of the Army and went to the University of New Hampshire -- New Hampshire politics is interesting politics, and at that point I got much more involved in doing polling in the state and then, ultimately, across the country.

MOORE: Well, the polling I do now -- I'm director of the service center at the University of New Hampshire, and we do client polling for mostly state organizations, but also for other private, even profit groups. But the most interesting polling we do is the statewide polling for public policy issues. We work with WMUR-TV, which is a ABC affiliate, and we do polling for them, and then for a consortium of newspapers in the state to keep track of what's going on in New Hampshire.

MOORE: Oh, we did pretty well. We did pretty well. We were the first polling organization to show the surge of support for Clinton. Initially, he had been at about 3 percent level and one of our polls -- this is no big thing -- I'm not trying to take any credit; it just happened that ours was timed, and we detected this real surge of support the first time Clinton overtook Tsongas. And then we also happened to be the first polling organization that detected his real dip in support where Tsongas took over him again. But there are other polling organizations that were doing polls at the same time, so I don't think we were particularly prescient or anything, we just -- in terms of timing. But it was very interesting. The results were interesting. We were able to show the volatility of the vote in a way that I had never even suspected.

MOORE: Well, we do student -- since 1976, the UNH poll was really a student-oriented poll. The very first poll I did was with students. It was a class exercise. We thought we'd see what would happen. We didn't have the amount of controls over student interviews that we do today; we've got, you know, computer-assisted telephone interviewing, people monitoring on the telephone. It was a group. We got together in '76 and did a poll. We did a poll first in the primary, and we did a poll in the general election. And we turned out to be, you know, very accurate with respect to the final results. And so, little by little, that polling operation grew. 1980 was when that person was talking about. It was still mostly a class project. By 1982, we were doing work for the television station, WMUR. And by 1984 till now, we've had, you know, the evolution of polling in our group has been very great, so we have much sophisticated ways of doing it now.

MOORE: Right.

MOORE: That's right.

MOORE: Well, we have a very interesting gubernatorial race because we have a female candidate on the Republican side and a female candidate on the Democratic side.

MOORE: For governor. And prior to that, we've only had three candidates ever run for governor in the state of New Hampshire. So here are two this year alone.

MOORE: Three women candidates, I'm sorry. Yeah, three women candidates. So now we have two women candidates, one on either side -- in both parties. And New Hampshire has gone through, as the Northeast has, probably more of a recession -- has felt the effects of recession more than others, and both of these women have been very important in the legislature. The Republican woman is chair of the House Appropriations Committee.

MOORE: And that's Liz Hager. And Deborah Orne-Orneson is a Democrat, so therefore, since this is a Republican legislature, she doesn't have a leadership role in terms of committee chair, but she has been very actively involved in the budget battles in the state. And so both of them, as a natural outgrowth of their participation, are now participating in the primary. And I should say that there are a lot of men that are running as well. Our incumbent governor, Judd Gregg, who's been in for four years -- two terms, because we have two two-year terms -- is now running for the Senate in an attempt to replace Senator Warren Rudman.

MOORE: He was going to run, and he kept saying -- everybody thought he was, but he's got a business with international trade, or something like that, going on, and he said he couldn't afford to.

MOORE: In the Senate race, Judd Gregg, who's our incumbent governor, is being challenged by a couple of Republicans who haven't won public office before. On the Democratic side, there is a Democrat who ran two years ago and lost. But in the process, at least in New Hampshire, winning and losing is almost a requirement before you can win, because you have to get your name out there in some way -- particularly for Democrats. Republicans win most of the time. The only way a Democrat could ever win -- I mean, could ever -- name-known, is to run and lose. That's the best way they can do it. I think it will be an interesting race there because the amount of dissatisfaction, essentially with the economy, and then a number of people blaming the governor for the economy, in a way -- taking the heat for it -- may make the Senate race a much closer race than one might normally expect.

MOORE: Yes, we are. We are doing polls for the media consortium. We do quarterly polls and then certainly we'll be involved when the general election comes to see how people are doing.

MOORE: I decided to write that book a couple of years ago. I myself have found polling very interesting, and what really prompted me to start writing it was the fact that much of the research these days shows that pollsters who want to be honest and objective and get truth still are frustrated by the fact that sometimes slight changes in question wording and sometimes the order in which questions are asked can, even though we don't anticipate it, influence the responses that we get. So it's really interesting to see how much this area of polling has developed. And when I started writing about that, and then started getting into it more and more, I realized that, as much as the ideas about what's going on, it would be interesting to, I think for some people at least, to find out what has been the history of polling. Who are these people that have had such an impact upon our political and social life through the method of polling? And that's how I got into it. And so, essentially, that's a critique as well as applause for polling, but mostly, I would say, a kind of recount of the history of polling in America and its role today.

MOORE: Well, I'm not going to say who I think is doing the best. I think there are several organizations that are doing well, and it's certainly difficult to tell, unless you can go back and take a look at everything that they're doing. But I would put a great deal of confidence in what I call the super pollsters, the media polls. When I see an ABC/Washington Post poll or CBS/New York Times or NBC/Wall Street Journal or Gallup/CNN poll, I have confidence that those are pretty accurate polls, as far as the limits of polling are concerned.

And then there are other organizations that have come out -- Los Angeles Times poll's another one -- and those polls really do provide pretty consistent, reliable results. And they're going to get -- we do find that there are tremendous differences sometimes in the results they give -- not tremendous, but significant differences -- as many as 7, 8, 9 percentage points. But usually those can be explained either by margin of error or by the fact that some polling groups try to screen out non-voters and some do not at this stage of the game.

MOORE: It must be the exit polls, the VRs, the voter research and survey exit polls. Both in New Hampshire and two subsequent states, the exit polls are conducted by VRS, Voter Research and Surveys, and this is a consortium of CNN, ABC, CBS and NBC, the four networks, that will conduct the exit polls.

And we don't know what the mistake was, but I talked to Warren Mitovski, who's a head of VRS, and what they showed, at least in New Hampshire, was that Pat Buchanan came very close to George Bush. He came within 3 or 4 percentage points, according to the exit polls. But according to the actual polling results, there was about a 15-point difference. That's a very surprising difference. It caused great consternation among the newsrooms around the country that were getting early exit-poll results suggesting that this could be an upset in the making, and that here was Buchanan clearly getting above the 40-percent level -- and it turns out he actually got 37 percent. And I don't think Mitovski knows yet. And it may be difficult to know precisely why Buchanan voters got disproportionately represented in the exit poll. But it happened again. So there's certain phenomena going on there at the time. Of course, that's moot now, but at the time it was a problem.

MOORE: It was really formed after the 1988 election and before the 1990 election. By 1988, there were three networks -- CNN wasn't doing any exit polls -- but there were three networks going into the major states spending literally millions of dollars to conduct separate exit polls. It was just too expensive. They were all trying to beat each other, getting the exit polls, projecting the results and coming on first. But there was a matter, I guess where it decided: Is it worthwhile to be first by minutes over your competitor for all the millions of dollars that you spend, or should we try to consolidate the effort? And that's what they did. A lot of the academic pollsters are not too happy because, with three different polls, you've got to check on each other. But the media pollsters, the networks, say they just simply could not afford to continue to do exit polls the way they had been doing them.

MOORE: That's right.

MOORE: From the same source. And the interesting thing about it is what I call the "many lies of television." They'll all present it as their own. They'll say, "According to" -- and the local affiliates do it, too. They'll project results because they can subscribe to the VRS results. They pay a healthy sum of money for those results, and newspapers can, too. And it's perfectly legitimate. But they present the results as though they're their own, kind of to give their viewers the impression that they've done all this work, which, in a sense, they have. They've paid for it. But it really is a consortium, rather than just any one network getting the results by themselves.

MOORE: Before VRS went into effect -- VRS not only does the exit poll, but VRS does the projection of the elections, so that projection is based not just upon the exit polls, but it's also based upon early returns from a sample of precincts around the state. So a number of precincts will be chosen, and then the network would send out interviewers, or actually data collectors, to stand at the precincts and, as soon as those votes are tallied for that precinct, immediately they'd be called in, put into the computer. Those vote totals would be compared with vote totals from previous years or other kinds of characteristics in a model that would then ultimately predict whether or not X is going to win or Y is going to win. So it's a combination both of exit-poll results and of early precincts. That's the way they're predicted. Now each of the networks had their own different model, so when they got together under VRS, there was a long period of negotiation as to who was going to be head of it, what models they would use, what kind of input they would have and so on. But that's essentially the way it's done even today.

MOORE: I think most pollsters would say that no poll has influenced the election; they provided information. But in 1988, there was a 50-state poll; very creative, innovative poll, done by ABC and Washington Post. I say innovative because most polls these days are not done on all 50 states. We do a nationwide poll, but the truth is that, even if a candidate is going to win by 4 percentage points in a nationwide poll, that person could lose because of the electoral votes. So what's important is to know how each one of the states are going to do.

But to do 50 state polls is a very humongous undertaking and a very expensive one. But ABC and the Washington Post did do it. And they used certain procedures, and they did it over a several-week period. ABC announced the results right before the second debate between Dukakis and Bush. And they announced it in such definitive terms that it appeared as though ABC was saying, "You don't have to worry about the election; it's already been decided. Our polls have shown that Bush has an electoral lock"

Now an electoral lock implies that there's no real variation in opinion within the states, which clearly wasn't the case. And so ABC came under a great deal of fire for presenting the results of that poll in the definitive way they did, and in the timing of the poll, which was just the night before the second debate. The argument is that, as a consequence of that, all the reporters said, "Well, in order for Dukakis to win, he's got to really hit a home run in this big debate." Of course, that was the debate where he was asked what he would do if Kitty Dukakis was raped. And nobody ever felt he hit a home run, that he really fell flat on that, and therefore the reporters and the press all essentially reported it as though the election had been decided, because he hadn't hit the home run; ABC poll had shown that he was going to win. And even the Washington Post criticizes ABC for the way it presented it.

I personally think it was a very creative poll, and I would hate to see ABC and the Washington Post not do it again. But I think there could be some modifications in the timing and even in the language so it isn't presented as though it's God is speaking and telling us the results here two weeks before the election.

MOORE: Yes, I did. And they were off on some states, but generally speaking, they were pretty correct. Jack Germond still was critical of it. Jack Germond of the Baltimore Sun was still critical. He said that even if they were right, it didn't make it right for them to have presented it that way. But for the most part, the poll was a pretty good picture of what the 50 states were doing, although they were wrong on several states. But ultimately they were right with respect to the overall picture.

MOORE: Well, a very important piece of information that should be presented about the poll, of course, is when it was taken. That may seem very trite, but a lot of times polls are announced -- they may have been taken several weeks ahead of time and yet the announcement will be, "According to a recent poll." In the flux of the campaign, things can change dramatically. Also very important is who commissioned the poll. If it's one of the media polls, you can assume that there is no ax to grind, that it's objective.

But sometimes there are polls done about the environment or about health that are specifically commissioned by a health organization or by an environmental organization or by a business organization. But they may commission a Gallup or a Harris poll. They won't commission a media poll to do it because the media polls won't do it. And then it will be reported, say, as a "Gallup poll." And so you think, "Well, it's objective." But the fact that it was commissioned by an organization could very well mean that the tenor of the report and the whole subject of the report is almost inevitably biased in favor of a client. So that's very important.

The number of people that are interviewed and the question -- the exact question wording -- is very important, and what the results were for each question. Now the unfortunate thing is that the average person who is watching the poll is not going to know all these things. My biggest advice is: Be skeptical of all polls, and only accept poll results to the extent that they seem to be reinforced by other ones. I mean, if you get a poll result that's really weird with the respect to other polls at the same time, particularly during election time, it's interesting, look at it. But you know, don't think it -- I don't even trust my own polls completely because I know that it's a snapshot in time that, in a very highly volatile situation, that what you get right now is not necessarily reflective of what's going to happen in two or three days. So I'm skeptical of all polls, and I think, particularly if you don't know who the organization is that's doing it, you ought to be skeptical.

MOORE: Yeah.

MOORE: CBS and New York Times do their own in-house polling. ABC/Washington Post do their own in-house polling. NBC and the Wall Street Journal commission polls with two campaign pollsters -- Peter Hart for the Democrats and it was Bob Teeter for the Republicans. Although now that Bob Teeter is working for President Bush, they have Vincent Brigglio who is the Republican pollster. And those two guys get together. So they figure that's a check on each other because if you got a Republican and a Democrat and they agree upon the interpretation of the questions, you can't have any kind of particular bias. And the Los Angeles Times does its own in-house polling in it, so those major media polls are really, in a sense, insulated from any kind of real effort to try to manipulate them. And so I don't think that we have the same kind of problem. And I also think that what Lou Harris did then was a real aberration. It's not the kind of behavior that one would have ever expected from him, and it's certainly not the kind of polling that is reflective in most of the polls today.

MOORE: OK.

MOORE: Bob Teeter is the campaign manager now for President Bush, and he did the polling for President Bush in 1988. He also did polling for President Ford in 1976, and he worked for the committee to re-elect the president, President Nixon, in 1972. In 1980 and '84, he helped the Reagan-Bush campaign. So Bob Teeter is probably the most experienced of all of these pollsters in terms of his length of time in working both as a pollster and a strategist.

MOORE: He's given very high marks. He worked with Peter Hart on the NBC/Wall Street Journal poll. And Peter Hart considers him to be a person of very sound judgment. I read in the papers -- I don't have any separate way of analyzing it -- that some people are criticizing him for not being a strong campaign manager. But as a pollster and as a strategist and a person who had judgment about what was going on, he seemed to be very highly respected.

MOORE: Peter Hart is probably the pre-eminent Democratic pollster. He's worked for more senators and members of the House and governors on the Democratic side. He worked with Morris Udall a little bit in 1976; in 1984, for Walter Mondale. He was Walter Mondale's pollster. And he is very definitely considered on a par, certainly, with Teeter, in terms of a person with very sound judgment. He also has a -- kind of a Garren/Hart -- Jeff Garren is his assistant and works with the Garren/Hart Strategic Research. It's kind of separate from a lot of the other work that Peter Hart does. So Jeff Garren seems to be taking over much of the political work that Peter Hart did.

MOORE: Richard Worthland. He's a guy who comes from California, comes from Salt Lake City, Utah, and then worked with Reagan in California; moved his firm out from California to Vienna...

MOORE: ...in Virginia. He did the polling for Ronald Reagan when Ronald Reagan first ran for governor -- excuse me, when he ran for re-election for governor and when Ronald Reagan ran for the president. He is the most mathematical of all the pollsters, using very sophisticated techniques. And he's generally -- the pollsters in there very highly regarded. He's a little bit different in the sense that he tends to work for more conservative candidates. Very strong, very friendly person to speak with.

MOORE: The wunderkind of 1972. He started polling when he was an undergraduate at Harvard University in the 1970s. He said he was animated by an idealism from Robert Kennedy. And in 1972, he worked for George McGovern and, surprise of all surprise, helped McGovern get the nomination. At that time he was widely perceived by the press as a, you know, very brilliant kind of greasy-haired, rotund, hippie kind of guy with brilliant insights into the American psyche.

In 1976, he worked for Jimmy Carter; in 1980, became more sophisticated, more presentable, and looks very dapper these days. In 1986 -- he said he got out of politics because in 1986 he came up with what must be one of the worst strategies that a pollster would come up with. Here was a guy who was animated to get into politics because of an idealism about democracy, but in 1986, he's working for Alan Cranston who was trying to get re-elected as senator from California. According to the polls, the younger people in California were more likely to vote for Ed Sherry, who was the Republican opponent. But they were also more likely to be turned off by a negative campaign, and therefore not vote.

So Pat Cadell said that his strategy that he came up with was to make the campaign so disgustingly negative that it would depress the turnout, particularly among younger voters, and allow Cranston to win. And he says -- and it was -- the strategy was successful. Cranston won by 2 percentage points. And Cadell said he felt very bad about that and felt that he had come to an end that he had never expected. He was poisoning politics rather than helping it, so, he said, he got out. Now he's back in, but not for money. He's an informal adviser to Jerry Brown, and has been on the airwaves recently talking about H. Ross Perot. But personally he said he's not going to get back into politics, at least as far as being a paid adviser.

MOORE: There appears to have been -- it's hard to know, because neither the pollsters nor the presidents nor their advisers really want to admit it. But there is evidence that Pat Cadell did a great deal of polling for Jimmy Carter when Carter was in office, that Richard Worthland did about two to three times as much polling, proportionately, for Ronald Reagan when he was in office.

It's not clear right now -- I know that there has been continuous polling going on and I don't know how much, but I've also heard that, during the Persian Gulf War that there was continuous polling going on to keep the pulse of the democracy, or what the people were thinking about the war and the way it was going, and that perhaps even George Bush terminated the war too quickly in the fear that, if he went on and actually got rid of Saddam Hussein and allowed the war, maybe at that point, to cause more American casualties, that he might lose support of the American public. So the public's view, in that sense, was very closely monitored to make sure things weren't getting too far out of hand. I would say that these days the president -- not this president, but the presidents are, in fact, using polls to constantly keep in touch with what -- what's happening.

MOORE: Right

MOORE: Well, the polls kept showing that Jimmy Carter was doing worse and worse. He was getting high disapproval ratings -- and there didn't seem to be anyplace, really, for Jimmy Carter to go. He was going to give a speech, in July of the year before the election, in July of 1979. He was scheduled to give a speech on the energy crisis, the energy problem. But Pat Cadell convinced him that there was a real crisis of confidence that was going on, that people were very alienated from the government and that Jimmy Carter should address that particular question. Jimmy Carter cancelled the speech -- or postponed the speech. He retreated to Camp David, where he had meetings for 10 days with all sorts of leaders. And then afterwards, he kind of came down from the mountain and gave a speech. And it was a speech which not only talked about the energy crisis, but it also talked about this crisis of confidence. It could be that there's a crisis of confidence, but it certainly, in the long sense, backfired for the president to say, "There's a crisis of confidence, it's against government, and we've got to do something about it."

Initially, according to Cadell, and the polls did show him at the time -- people responded very favorably. Here was somebody trying to deal with this problem of the broader question of how much people trusted government. But a few days later, Jimmy Carter fired all of his Cabinet and then accepted a few of them back. And it seemed like such a political ploy that he really got savaged in the press for the speech and for the firing of his Cabinet, so that that malaise speech really has gone down, at least in the press' perception, as a devastating speech for Jimmy Carter's potential re-election.

MOORE: Yes, right.

MOORE: Sounds like somebody could be saying the same thing today. That kind of alienation of the American public is probably -- there's probably always a strain of some alienation among some of the people who are not benefiting from the way things are happening or going in the country. But it's particularly pronounced in times of economic distress, such as we have now, such as there was in the latter part of the 1970s. What he's talking about -- the distress and the paralysis -- is exactly what people are talking about in Congress today. And it is interesting because Pat Cadell, in 1972 and in 1976, ran his candidates as outside candidates against that alienation. Jimmy Carter ran as being an outsider in Washington. In 1980, that malaise speech was an attempt to put Jimmy Carter outside of Washington again, so that he could run against Washington. It was difficult, because he's an incumbent president.

Of course, we see George Bush today doing the same thing that Jimmy Carter did, trying to make himself appear as the agent of change, as somebody who's not a part of Washington. But in 198 -- -excuse me, in 1990, when I interviewed Pat Cadell, he said, "If we think there was alienation in the 1970s," he said, "there's really alienation now. If we could talk about it being on a Richter scale, and it maybe was five on the Richter scale then, it must be eight on the Richter scale now." That's what he was talking about about a year and a half before -- about a year and a half ago. And, in that sense, I think he's probably accurately identified the kind of turmoil that's going on in the country today.

MOORE: He's an informal adviser to Jerry Brown. Jerry Brown met with him. Apparently Jerry Brown, in his opening speech, lifted a passage that was a part of a novel that Pat Cadell had started writing, "Mr. Smith" -- talking about alienation and corruption and so on. And I don't know that he's an informal adviser to Ross Perot. All I do know is that he has been interviewed and has made very positive comments about the way in which Ross Perot's going about it. But I have not had an opportunity to talk with either Cadell or Perot about whether they've met or discussed -- I'm sure journalists may have better access than I to those people.

MOORE: I interviewed all of the campaign pollsters. I interviewed the five that we've just gone through, including Tubby Harrison. We didn't mention him, but Tubby Harrison worked for Michael Dukakis. And I interviewed the major media pollsters. I've interviewed most everybody that I cover in the book.

MOORE: I think that is a very fanciful idea that really ignores human nature. Our polling shows, you know, people don't pay a lot of attention to public policy. What he's asking -- of course, I don't know precisely how he's going to implement that -- but what he's asking is for the public to pay a greater amount of attention to public policy issues than they've shown any inclination to do so now.

He was talking, for example, about health care. He says, "Tell you what I'm going to do. I'm going to go out and find all the experts and we're going to find out the three most important ways in which to deal with health care. And then we're going to put it to the American people and they can vote." Well, how are you going to put it up to the American people? That's the big question. Are you going to take 20 minutes, 20 minutes and 20 minutes out of an hour TV time? And who's going to make the presentation? And how are you going to ensure that it's not a biased presentation?

We know that, in polling, we can't ask a question precisely in an objective way; it's almost impossible. And now we're talking about a whole presentation of a whole issue, and then the people are going to vote on it. Electronically, I'm not sure we necessarily have the technology to do it. But even if we did, it just seems to me a very unrealistic approach. We have elected representatives; that's the job. That's what they're supposed to do -- find out the nuances and what goes on, and if we don't like it, we vote them out of office. If you really want to reform government, it's not going be through electronic voting; it's going to be getting rid of the -- it's going to be, I think, through public financing or getting rid of the impact that lobbying groups can have by buying a representative. That's the way to go, not by suddenly inserting the public in a way in the policy process. I just don't think it's going to work.

MOORE: I've been doing polling and working with polls and survey research for over 20 years.

MOORE: Well, there's not a humongous difference. I think there is probably a greater degree of alienation -- more people these days than certainly at any time in the past 20 years -- more people are expressing discontent with what's going on. But it seems to me more a matter of degree. For example, people still feel very strongly and favorably about the Persian Gulf War. They still feel favorably about US image overseas. And so there is ... I would suspect that the degree of alienation that we find now is in large part a reflection of the state of the economy, and that if things get better, then that alienation will subside. But I think there will always be some people who are alienated from the government because they don't feel as though there's true democracy, or as that they're getting a part of the benefits. And I think to the extent that there are inequalities that exist -- strong inequalities, not just in actual income, but in opportunity -- that that alienation will continue.

MOORE: Yeah, for the most part. You wouldn't know it from the polls. And this is where the polls kind of manipulate the results and they don't necessarily mean to. Polls will ask a question -- they might say: "Do you think the United States should or should not give aid to the Contras?" Well, if I ask you that question and you don't know -- I haven't said, "or are you unsure?" or "Haven't you thought about it?" -- you kind of feel obligated to give an answer, and most people will. And so you may come up with maybe 5, 6, 7, 8 percent of the people who volunteer, "I don't know." And a lot of public opinion polls like that overestimate the extent to which people have an opinion about matters.

But one thing the polls don't overestimate is the amount of information they have. Numerous polls -- many, many polls have indicated that people are not aware of a lot of foreign policy issues, a lot of domestic policy issues. And I think, in a sense, that's the nature of democracy. We actually vote for people to do that kind of work for us, and we pay attention to the things that maybe get the most attention in the media and that are very important to us -- and, of course, the economy is one. But all of the other public policy issues -- there are so many of them, so complex, that I think it's unrealistic to expect the public to have a really thorough knowledge of those issues.

MOORE: Yeah. Well, there are a lot of different reasons that might account for it. I'm not sure that there's only one. You have to remember that, if you took a look over the past several elections, although it's gone down by a fraction of a percentage point, essentially it's remained stable at about the 50-percent level. It was about 60-percent level in 1960 when Kennedy ran. So there's been about a 10-percent drop. Part of the drop could be that there are more people who are eligible to vote. For example, we had the amendment that allowed young people to vote. So there's a bigger base. Young people don't vote as much as older people. And there was a big drop -- once they became eligible to vote, there was a big drop in the percentage since, disproportionately, they don't vote as much. But we also have other restrictions on voting.

If you want to vote, you've got to work it into your workday. It's not a holiday, even for a presidential election. You've got to go out and you've got to try to get registered. There are ways to have, you know, regular registration or immediate registration and to avoid fraud, but for the most part, we as a country have not decided we want to do that. So there are ways we could increase the amount of turnout. But there is this opposition, I think, to letting everybody vote if they want to. We want to put certain obstacles in their way to make sure that they've paid enough attention or something like that.

MOORE: Yes. Typically in New Hampshire it is. And the reason is that it gets so much attention it's like a general election. I think we got on the order of 65-percent turnout of registered voters. That's a very high percentage. But remember, you were up there and thousands of media people were up there and there was a tremendous amount of attention. The newspapers constantly, and the headlines, were focusing upon that particular election. So the fact that people get energized to go out and vote, because of so much coverage, is to be expected.

MOORE: Yes.

MOORE: Oh, I don't know. I hadn't thought about what would make me happy. It was a lot of fun writing the book because I learned a lot of things that I hadn't known before. And I find very interesting, when I'm interviewed, as we are today, to be able to talk about it. I think I would be very pleased if those who are interested enough to buy it, read it and said, "This is a good read. It's fun to read. And there's a lot of useful information in it." Some people have told me that but, of course, we know there's a bias whenever you talk with people themselves. So that's what I would hope for it, I think.

MOORE: Well, I don't have any step-level change that I want to see. What I'm really interested in is the extent to which current research is revealing to us the ambiguity of polls. There is an uncertainty about any poll, and I think the more educated we become about these uncertainties, the better off we are, so that we don't take to heart any single given poll. And we recognize that the complexities of public opinion, in order to be reflected in a poll, have to be examined using numerous questions, not just one single question. So I think the research that's going on now, revealing this uncertainty of the results is, very useful.

MOORE: Thank you very much, Brian. I enjoyed it.