N.H. Surprises Have Pundits Scratching Their Heads

Well, Judy, to follow up on that, was it your feeling that the polls drove the reporting or the reporting drove the polls? What's the relationship between -- because it wasn't just the polls. There was an awful lot of reporting going on that was kind of upbeat, and wavy, and all that sort of stuff.

JUDY WOODRUFF: Well, Jim, I have a terrible confession to make, and that is that reporters look at the polls.

JIM LEHRER: Oh, Judy, say it isn't so!

JUDY WOODRUFF: In fact, I would go so far as to say that polling, for better or worse, shapes much, if not most, of the news coverage in this country.

Reporters are every day looking at the polls, judging, frankly, in an era of increasingly limited resources for newsgathering, deciding where to put those resources.

And when you have six, seven, eight candidates in each party, you can't spend as much -- you're not going to spend as much time, energy and manpower, or womanpower, on the candidate who's polling at 2 percent as you are at the candidate that's polling at 30 percent or 40 percent.

So that is clearly part of what's going on here.

JIM LEHRER: But there were an awful lot of -- you all -- awful lot of reporters on the ground in New Hampshire. And the question, of course, is that the public asks, "Well, wait a minute. How could you all have been over there talking to all the voters and talking to everybody, and gotten it wrong, as well, anecdotally, in terms of reporting?"

Is there an answer to that question, Judy?

JUDY WOODRUFF: Well, there's not a clear answer, Jim. We were hearing on the ground -- I was actually focused more on the Republicans, so I have an out here, you could say -- but anecdotally...

JIM LEHRER: Oh, no, you don't.

JUDY WOODRUFF: Based on what I was hearing from people, I was hearing a lot of Obama. I would go around at McCain rallies and Romney events, but especially at McCain and Huckabee rallies, and I would hear people say, "Well, I'm looking at either Huckabee or Obama," or, "I'm looking at McCain, Edwards and Obama."

So I was picking that up. And I certainly didn't sense that it was going to happen at the last minute.

I will just quickly add this, Jim, that I do think there's something to be said for those who point out the polling ended basically on Sunday. The voting happened on Tuesday, and there was movement clearly in the last 24 to 48 hours.

JIM LEHRER: Yes.

Gwen, what is your contribution to that, the connection between the reporting and the polls?

GWEN IFILL: I was actually going to contribute to what Judy just said. But the connection -- one of the interesting things that happens -- you're right. There were hundreds, maybe thousands of reporters in New Hampshire on the ground from, you know, Norway and Japan, not just forming conventional wisdom among ourselves.

And what could happen in that kind of situation is that three people would get in the corner of a room and form conventional wisdom.

But the difference, however, I would say, in defense of those poor reporters on the ground, is that there were no fewer than nine polls that showed him with these kinds of leads.

It wasn't like someone decided that Barack Obama, absent any other evidence, was going to lead, that the Obama people had him up by 8 points in their internal polling. The Clinton people had him up in double digits.

It wasn't like someone said, "Oh, wait a second, you might be wrong. You should be careful. Things are turning," which sometimes happens. That did not happen this time.

So I think people were ill-equipped to see what would happen in the last-minute vote.

Amy, you go first.

AMY WALTER, The National Journal: OK, well, let's talk about a couple of things. The first is that, you know, for all the talk about how things were wrong on the Democratic side, the Republican polling was right on target.

I mean, there was clearly -- and it goes to Judy's point -- which is something happened on Monday that clearly changed the contours of this race. That a poll can't pick up on, because nobody was out there doing that. So I think that's a very important factor.

The other issue, too, is -- and this is why people love to follow politics, because it does change, and it can change overnight.

But the reality was there was something happening and moving that people were not -- and Gwen made this point -- people weren't just making this up or pretending that it was happening.

What happened in the polls -- and you could see this -- we can always guess as to why it happened. But here's what Hillary Clinton was able to do that the polls did not pick up on, whether it was the number of women who turned out, certainly bigger than anybody would expect, I mean, almost 60 percent of the electorate, women. That's not a typical number.

The other is that, for as much as we talked about the role of independents, I don't think that independents decided that they're going to go for McCain because suddenly Obama didn't need them anymore. What you found was that independents actually turned out at a higher percentage than the polls had predicted, but she did better among those independents.

That was, I think, the untold story.

JIM LEHRER: But, Stuart, don't you find it extraordinary that nine polls would all be wrong?

STUART ROTHENBERG: Yes, I do find that extraordinary. In fact, I find it impossible to believe.

Jim, look, in any given year, I look at hundreds or thousands of polls. And I believe that too many journalists, too many cable TV talking heads and anchors don't discriminate between the good polls and the not-so-good polls.

I never go on the basis of one survey. I think an individual survey comes out -- we've seen it over the past few months -- they've been overanalyzed. I think a lot of surveys done in the first half of the year, people weren't really paying attention. These were not strong judgments.

But the surveys that we got over the past week -- you know, one poll can be wrong, two polls, not eight or nine when they're showing the same thing, the same direction, and the same general magnitude.

I think it's just -- the evidence is clear that something happened between the end of the polling. Something happened Sunday, Monday that changed some people's view of this race. I believe it's women.

I believe they started to empathize and identify more with Senator Clinton, and she was humanized, and that changed some of the votes.

JIM LEHRER: Well, Stu, you say you've read -- you have looked at hundreds of polls, thousands of polls. Are you going to look at them any differently because of what happened in New Hampshire? Is this a big deal in the polling world, in other words?

STUART ROTHENBERG: Not for me. There were some polls that I think have a great track record and others don't. There are some pollsters who have done polling in particular states, who ask good questions, and other polls, I think, frankly, are not as good.

I never overreact to a single survey. I think it would be wonderful if reporters from now on would stop doing that.

But on the basis of what happened at the end of the game here, no, this was simply a question of there weren't polls when people were changing their minds. And that's because you can't do polls on election day and then report them the next day. They don't matter, because you actually have the election results.

All right, finally, back to you, David, and you, Mark. I asked both of you in our 6 o'clock NewsHour last night what your expectations were. And each of you answered. And, you know, and what did you base that on?

When you look back on it, now, David, do you find there was some flawed information that you had or some -- what? Tell us.

DAVID BROOKS: Yes, as I recall, I predicted Clinton by three. No, I could be wrong. I'm just rewriting history.

No, I think, a, it was the polls; but, b, we walked around. We saw the lines. We talked to the campaigns. It was unanimous. It wasn't a close call. So why were we wrong?

I mean, we've heard a lot of the theories that are out there, but it seems to me there were two basic organizations of the theory. The one is, was it a late swing, which is very plausible? But the problem with that is...

JIM LEHRER: It wasn't that it wasn't picked up either by the reporting or the polls?

DAVID BROOKS: Well, because they weren't asking. But the problem with that is that when you ask late-deciders who did you go for, Hillary had no advantage among those people. So that at least complicates the theory.

JIM LEHRER: What do you mean they had no advantage?

DAVID BROOKS: Well, they asked, "Did you decide in the last day? Did you decide in the last three days?" And the people who said they decided in the last few days, it was pretty split between Obama and Clinton.