ad info

Editions | myCNN | Video | Audio | Headline News Brief | Feedback  





Bush signs order opening 'faith-based' charity office for business

Rescues continue 4 days after devastating India earthquake

DaimlerChrysler employees join rapidly swelling ranks of laid-off U.S. workers

Disney's is a goner


4:30pm ET, 4/16









CNN Websites
Networks image


Do the Polls Really Reflect What Voters Feel About Election 2000?

Aired October 27, 2000 - 7:30 p.m. ET


ROBERT NOVAK, CO-HOST: Tonight, one day Gore is up. The next day it's Bush. What's going on with the polls? Do they really reflect how voters feel about election 2000?

ANNOUNCER: Live from Washington, CROSSFIRE. On the left, Bill Press; on the right, Robert Novak. In the CROSSFIRE, Democratic pollster Geoff Garin, and in New York, Republican pollster John McLaughlin.

NOVAK: Good evening, welcome to CROSSFIRE.

Today's latest tracking poll from Gallup for CNN and "USA Today" shows a 13-point lead by George W. Bush over Al Gore -- Wow! That's a landslide. But wait a minute, the CNN/"TIME" poll gives Governor Bush only a six-point lead; and ABC News four points; "The Washington Post" only three points; and John Zogby's polling for Reuters and MSNBC shows -- guess what? -- Vice President Gore in front by two points.

What's more, the polls have been bouncing around like scores in a basketball game. At least the candidates have a variety of polls to choose from as they campaign in battleground states. Today Al Gore was in Pittsburgh basking in the charisma of Bill Cosby and George W. Bush was in Kalamazoo, Michigan talking tax cuts.

But what about the poll-takers? Do they know what they're doing? Is somebody cooking the books or is it all just fuzzy math? Bill Press.

BILL PRESS, CO-HOST: Fuzzy math, to borrow a phrase. Thank you, Robert.

John McLaughlin, good evening, thanks for joining us on CROSSFIRE. I just want you and Geoff to know I'm not taking sides tonight. I don't trust either one of you. And here's why, if I may quote one of your colleagues and a guy that Bob and I both admire, John Zogby -- I think the only one to nail the 1996 race right on its head -- John Zogby's quoted in "The Boston Globe" the other day as saying: "Response rates are not what they used to be. When I started in this business in 1984 they were averaging about 65 percent, now we're down around 33 or 35 percent," which means roughly six out of 10 people you can't reach or they refuse to talk to you.

How can we trust a poll that's based on only 40 percent of this country?

JOHN MCLAUGHLIN, REPUBLICAN POLLSTER: Well, I think the one thing you have to keep in mind is that a lot of us, like myself and Geoff -- we do the political polling and it's a tougher thing to do, but you have to take more time. Unfortunately for our candidates, it gets more expensive, but you have to be much more careful and keep a lot of quality control on it.

These media polls, you're going to see a lag where -- they move the electronic news, where in the news cycle somebody could have a good day, it could affect the polls by a few points, and then it could disappear within a couple days. So there's a volatility, there's a closeness of it, and the one thing we'll all agree on, it's going to be a close election.

The one who has the best week next week could win the election.

PRESS: Yes, but I don't think you talked about the 60 percent. Again, if 60 percent won't talk, to give any validity to a poll based on the 40 percent, you have to assume that the 40 percent are going reflect the opinions of the 60 percent you can't reach or won't talk to you.

I mean, that's bogus, isn't it?

MCLAUGHLIN: It is -- well, some of these media polls are going to be way off. Most of us that do it for a living, and the only thing we can do, is we have to be much more careful. We have several quality-control checks where we set up models, we pick lists of registered voters who are likely to vote, with past history, you add new registrants, you build the model.

You also screen properly, and then when the results come off, if you're a professional pollster you have past results to measure to that are fairly immediate, as well as long term that you can model trends on.

So what I'm saying to you is I think some of these political polls that are done by the parties and the candidates and the candidate pollsters are probably better than the media polls.

PRESS: Well, let me just ask you this frankly about these people who do talk to you, because, frankly, I think the phone is a pest. I think anybody who's got any brains today has caller ID or they have an answering machine or they just don't answer the phone. Right?

MCLAUGHLIN: It's a problem.

PRESS: What can you say about the kind of person who has spent 30 minutes on a Friday night talking to you instead of having a drink? I mean, come on. They're idiots aren't they?

MCLAUGHLIN: I do less polling on Friday and Saturdays than, probably, other people, probably -- certainly, than the media. But the other part about it is you have to be very careful. We get a good response because most Americans are not called. The hard part of what you said is getting through the answering machines.

We usually have one contact for every 10 attempts that are successful. And 1/3 of the calls that we hit are answering machines. Now, you have to be very careful doing it. And there's another element: A lot of Americans have cell phones now. They don't use the hard phone like they used to and you can't get those.

So I'm -- what I'm saying to you is, you are right in your assessments. There are a lot of pollsters that do this professionally that are working through that. What's bad about it is you're going to find some media polls that could be way off because they don't take that into account. They're done very cheaply and they're done very quickly.

Geoff Garin, I wanted to quote another of your colleagues to you. You know, by the way, for those of you who don't read Shakespeare, he was the one who said kill all the lawyers. I'm sure he'd say kill all the pollsters now.

But Frank Luntz, who is a pollster said this, he said: "Shoot us all. Basically, line up every pollster and just get rid of them. Quite frankly, everyone in my profession looks pretty bad right now."

Agree or disagree?

GEOFF GARIN, DEMOCRATIC POLLSTER: Well I think a lot of pollsters do look bad. It's important to understand what's happening, that there are these challenges that Bill and John have talked about, that people -- fewer people answering. And different pollsters deal with those challenges in different ways and that's what produces the different results.

Most of the polls -- most of the media polls now try to weight their sample so that it looks like what we know about the voting universe. We know how many men and women participate in an election. We have a pretty good sense from previous exit poll data what the age distribution is; and so you can weight to those things.

The real issue is, do you weight to some model of what the partisan split is in the electorate? Most of the polls do. The Gallup Poll, which is who "USA Today" and CNN does not weight to parties so that -- that's why there's a lot more bouncing around in that poll. But different pollsters have different approaches to dealing with these challenges. They are real challenges.

The question is, if you're -- all the work that we do and that John does is to help advise candidates and you need to have some information to make good political decisions is -- how do you get that information as accurately as possible. I think you can still come pretty close if you recognize what you're dealing with.

NOVAK: Geoff, let me ask you a dirty little question if I could. There's been something true about all the public polls, what John is calling "media polls;" at least since 1980, 1984, '88, '92, '96 -- you know what the thing in common with all of them is? The Republican's candidate for president has always done considerably better -- considerably better -- than the polls indicated he would do. Why is that?

All those elections, whether it's a landslide by Reagan or a debacle by Dole, they did better in the polls.

GARIN: Well that can't be true any longer because you have so many different polls and so many results to choose from. You've got such a wide range, I don't know. I don't know that that's particularly true.

NOVAK: It is true. I can show you the information. But let me go here: Is there a possibility that the Republican vote is consistently underestimated in the polls and it's being underestimated now?

GARIN: I don't think so; and look, I mean, it's just hard to make any generalization about the polls now because the polls now are so different from one another. And, you know, there are other, you know, just in terms of, again, trying to use the polls to predict the turnout and the ultimate result -- most of these polls, for example, showed that labor union households make up about 16 or 17 percent of the sample.

Well, in the last couple of elections unions have gotten their membership up to representing close to 22, 23 percent of the turnout. So all of those things can make a difference.

NOVAK: Let me just ask one other thing. I want to take up something that Bill said about what kind of person agrees to talk to a guy like you.

GARIN: I agreed to talk to a guy like you.

NOVAK: Ed Goeas, a pollster, he said this -- he said: "Quite frankly, a mother who will take 20 minutes on a weekend to talk to a pollster is not a normal person. Is that true?

GARIN: There are plenty of people who are willing to share their opinions.

NOVAK: Normal people?

GARIN: Yes, definitely. People welcome the chance in a well- crafted questionnaire to express their opinions about important issues of the day; there's nothing abnormal about that at all.

PRESS: Show me one of them.

John McLaughlin, at the risk of never -- putting my own job on the line, I guess is what I'm saying, I have to be honest with you and tell you that the poll I have the least confidence in is our own. And let me tell you, maybe, why.

MCLAUGHLIN: Because you don't like the guy who's ahead.


NOVAK: Very good, John.

PRESS: Here was CNN's tracking poll on Tuesday of this week, OK: Gore, 46, Bush 45. All right? Here's CNN's tracking poll today: Bush 52, Gore 39.

Now, there's been no major event this week. There's been no major -- no debate, none of the other polls show a swing like that. How could that possibly be true?

MCLAUGHLIN: Maybe you're biased towards Republicans. But actually what I think in that kind of a poll, you are seeing it's a fallible science. And as Geoff said, there's a lot of us who have statistical backgrounds, advanced degrees, but we know what the margins of error are in these polls.

Now, there are also elements for some common sense. I'll give you an example: I do a lot of polls in New York. One night when the Yankees were playing, the game went long, it was a tight eighth inning, the Mets played after it; I got a survey back from the state legislative district I threw out because it was imbalanced. The men who answered the phone that night would not be the type of men I'd want to see in a poll. I redid it the next night, and I got what I thought was a good poll. But there's an element of common sense; there's an element of science.

Now the other factor here that I think you may see and it's just emerging now, remember, no poll ever showed Ronald Reagan winning the presidency. There was none. None had him certainly ahead of Carter. None had him over 50 percent. There were fewer polls back then. The difference is you have an incumbent administration right now that is in effect Gore represents the Clinton-Gore years. He's polling under 50 percent and Bush clearly has a higher favorable rating, a more likable personality than Gore. And Gore's trying to even up the issues.

What I'm saying is the late-breakers. If Bush has a good week on the issues and a good aggressive week you could see a very volatile election where the late-breakers do break to him. And that -- what you may see is a see-saw effect where it's swinging up on a good day and swinging down but there's a margin of error you have to calculate.

PRESS: All right, pollsters, we're going to take a break here, just a second. And for you folks at home, here's just what you wanted -- another pollster for you. Yes, after tonight's show you can get the latest on our own polling when CNN's polling director Keating Holland takes your questions in our chat room at

We'll take a break. We'll be back with more CROSSFIRE. So if anybody calls you during the break to take a poll hang up on them. We want you back.


PRESS: Welcome back to CROSSFIRE. Most Americans complain about polls, yet everybody reads them. But this year there are so many polls and they are so all over the place they're driving us crazy. So tonight as a public service and with only 10 days left, we ask two people who make a living in this unsavory profession to defend their product. They are John McLaughlin, a Republican pollster in New York who does not work for the Bush campaign and Geoff Garin in the studio with us, a Democratic pollster who does not work for the Gore campaign -- Bob.

NOVAK: Geoff, since you're not a part of the Gore campaign, Vice President Gore today and last night was in West Virginia. Now if anybody had told me three months ago that 12 days, 11 days before the election the Democratic nominee would be in West Virginia, I'd say they were crazy. Do you know why he's in West Virginia?

GARIN: Well, because the polls quite accurately show that West Virginia is close, just as quite accurately show that Florida is close. Who would have thought six months ago that George Bush would have to be to campaigning in Florida?

NOVAK: Florida's always been a swing state. But West Virginia -- "The Charleston Mail" newspaper yesterday had Bush 10 points up.

GARIN: Well, I don't think it's a 10 point state but it's clear that -- look, I think that in this election the impulse in all kinds of states has been to closeness. And let me say one other thing about the polls is that I think the polls are all over the place and there are methodological problems. I think the voters are also all over the place in this election.

I think there are about 25 percent of the electorate who are really conflicted about this. Not that they hate the candidates, but they are painfully aware of the flaws of each candidate, And one day they're more concerned about one guy's flaws, the other day they're concerned about the other guy's flaws, and whatever the last thing they heard is focusing their attention on one flaw or another. So there really is a lot of movement within the voters from day to day. They're having a hard time making their minds up.

NOVAK: I want to ask you about Ralph Nader in the polls.


NOVAK: I have been hearing from everybody, particularly from Democrats, he's going to fade away. He's going to fade into nothing. I read the polls, he doesn't fade. He's at 4 percent, and as far as usually you think at the end of a campaign, a third party candidate with less than 5 percent of the vote disappears from the media. There's more media on Ralph Nader now than there was a month ago. Do you think he can hold that 4 percent right to the election?

GARIN: My guess is that the 4 percent will not disappear to nothing but it will drop in half. I think a lot of those voters either will turn out not voting or they're going to say, look, there is too much at stake, that a President Bush could end up making abortion legal, wrecking our environment...

NOVAK: That's a Democratic line.

GARIN: ... doing all those kinds of things and people will recognize that this is not a free vote.

NOVAK: It hasn't worked so far.

GARIN: Well, look, I think that there's two weeks to go and that this argument is being made in a much more forceful way than it's been made at any other point of the election.

NOVAK: You will grant me, Geoff, that John Anderson, who got 7 percent of the vote in 1980 -- he didn't fade at all at the end. All his fading came in September and early October. He stood firm by the time he got to October 30th.

GARIN: Right, but I don't think that there was a campaign -- first of all, neither side was clear who John Anderson was hurting, so none of them invested great resources in trying to pull that vote away. It is clear...

NOVAK: You think that that's going to decide the election?


GARIN: I think there's a couple states where Ralph Nader has a significant impact


PRESS: Let's get back to John McLaughlin. John...


PRESS: I don't know whether you would describe polling as a science, but let's assume you would. And yet, when you look at it...

MCLAUGHLIN: Fallible science.

PRESS: A fallible science, yes. When you look at it, there are so many variables in a poll, right, whether you're talking about registered voters or likely voters? What time of the day? What day of the week? How you phrase the question, the order of the question, on and on.


PRESS: So again, how can -- what is the exactitude we can expect when there are so many variables?

MCLAUGHLIN: Well, I think you have to look at track history and you have to also make a judgment, as Bob said, you've got to look at past election results. And certainly in '94 and '96 the polls underreported the Republican vote. In '98 in some states they underreported African-American, the minority voter, Democratic turn- out and they were wrong.

You also have to take a look at what's happening in last week. If these polls stay close you're going to have a larger election turnout because the vote means more to the average person. So you also have to model for turn-out.

I think you have to make a very careful decision about the polls, which ones you're using, which ones you're talking about. And what's really interesting is the media seems obsessed on reporting these polls more so than the issues these days. So, you know, whether good or bad, if you're going to trash them and you keep giving them press, they're going to keep coming out with them.

PRESS: I take your point. You talked about turnout, which gets to the point of likely voters. Obviously, you're talking to registered voters. It's almost meaningless because who knows who's going to vote. But how do you know when people tell you they're a likely voter, how do you know they are a likely voter? How do you know they're not just lying to you?

MCLAUGHLIN: In different states, you can actually get data history of who's voting and who's new registrants and who voted in '96 and '98.

PRESS: By name.

MCLAUGHLIN: Yes, by name. You can absolutely get -- you can get, it's all computerized in certain places. In other states you can't. Some places you're better off doing random digit dial. For example, with motor-voter now you get a lot of voters on the list who don't really exist and aren't really going to vote. They registered to get their license or some government service, but then they have no history of voting. So you have to use your brains.

PRESS: You mentioned history, I just want to ask a quick question, maybe Geoff wants to comment, too, about the future. They started doing, I guess, mail polls and then door-to-door. Now with cell phones used today and so many answering machines, is phone polling really dead? I mean, is this like the last year the phone pollings going to be around? Are we going to move to Internet polling or what?

MCLAUGHLIN: It's going to change, but you're definitely going to move to more Internet polling and you're going to move to something maybe even where you have to give people incentives to be part of a panel where you're pay them. But Geoff, as he knows, everybody answers. We live in a democracy and there's a lot of people who -- they haven't spoken to Bob Novak or Bill Press, but when they get a call, they think they're talking to somebody important like yourself and their opinions count and they're willing to tell us.

PRESS: So you're not out of business, Geoff?


GARIN: Not yet, but this is a business -- when I started with Peter Hart in 1978, 75 percent of the work we did was door-to-door. And we'd spend 45 minutes with people and it was a wonderful experience and zero percent of the work we do now is door-to-door. And 10 years from now, I would think the majority of what we do will be done by the Internet or some other means by what we do. I don't think the phone will even be a phone in 10 years from now. So we're going to have to reinvent what it is we do and we're in the process of doing that. But for right now, I feel confident that we can get a reasonably accurate poll over the telephone.

Bob: OK, we have to go. We have to have a reasonably accurate departure. Thank you very much, Geoff. Thank you John McLaughlin in New York. And Bill Press and I will poll each other with closing comments.


NOVAK: CNN's polling director Keating Holland answers all your questions about polling right after the show at And tune into Monday's CROSSFIRE when former White House Press Secretary Joe Lockhart sits in for Bill Press with co-host Mary Matalin. Their guests: "The New Yorker's" Joe Klein and "The Washington Post" Bob Woodward. Their topic: the latest from the campaign trail.

Bill, 40 years ago when I covered this campaign, we were coming down to the end and nobody knew how John F. Kennedy or Richard M. Nixon -- who was going to win. We didn't have any polls. Nobody took the media polls. They took a rare poll. And we were flying by.

Now we've got so many polls we are swimming in it and we still don't know who's going win. So we haven't progressed anything in 40 years except that we get a lot more data.

PRESS: But actually, Bob, because I was reading today, Gallup was doing some door-to-door work at the time and Gallup, our pollster, predicted that Nixon was going win, which gives me even less confidence.

NOVAK: I'll tell you a little bit about...

PRESS: Thirty years, Bob, I've never, never received one call from a pollster. Neither has Carol. I don't know about your household. What are we, not on the map?

NOVAK: But let me tell you this, Nixon really did win the popular vote in 1960.

PRESS: Are you accusing John Kennedy of stealing the election?

NOVAK: No, I'm just telling you he won the electoral vote and lost the popular vote.

PRESS: Oh, Bob, get off it. From the left, I'm Bill Press. Have a good weekend, everybody. Good night from CROSSFIRE.

NOVAK: From the right, I'm Robert Novak. Join us again next time for another edition of CROSSFIRE.



Back to the top  © 2001 Cable News Network. All Rights Reserved.
Terms under which this service is provided to you.
Read our privacy guidelines.