washingtonpost.com  > Live Discussions > Politics

Election 2004: Poll Numbers

John Zogby
Friday, October 29, 2004; 4:00 PM

With Election Day less than a week away, the presidential race is still too close to call. Both sides are feverishly courting undecided voters, rallying their base and aggressively pushing their candidate.

What are the latest battleground numbers? How reliable are tracking polls? What problems do modern pollsters face?

John Zogby, president and CEO of Zogby International, was online to discuss his newest polling numbers and the 2004 election.

The transcript follows.

Editor's Note: Washingtonpost.com moderators retain editorial control over Live Online discussions and choose the most relevant questions for guests and hosts; guests and hosts can decline to answer questions.


John Zogby: It is another close one. Brand new numbers will be out in the states. I am certainly happy to answer any questions right now.


New York, N.Y.: I saw you on the Daily Show last night. So I have one big question. Are you willing to make the same prediction today, in this forum, as you did last night?

John Zogby: Yes. To the degree that I can predict, yes. I don't have a crystal ball I just have a hunch based on my numbers. And I still think that an incumbent president polling 48 or under nationally and in key states in trouble. The burden is still on John Kerry to persuade the undecideds but if history is any judge undecideds break against the incumbent.


Normal, Ill.: Which of the states which now seem most "in play" would you say are actually the most likely to go to Kerry and which of them to Bush?

John Zogby: The think is that the daily numbers just confound me,. Just when I think there is a trend one way it seems to be stopped in its tracks and goes the other way. For now Pennsylvania looks good for Kerry so does Ohio. Minnesota and Wisconsin look like they are breaking for Kerry right now. My new Reuters numbers will show Kerry moving into the lead in Florida. But obviously this is a very volatile situation.


Doha, Qatar: There has been lots of press regarding large number of early voters. Is there any indication whether these early voters are largely Republican or Democratic?

John Zogby: Characteristically the early voters we are capturing in our polls are not leaning one way or another. They are a dead heat.


San Bruno, Calif.: Under what conditions would you throw out a poll result, and take a 2nd poll?

John Zogby: If I see that there is something seriously flawed in our sampling. For example, a minority group significantly under represented. Or a group whose results are shockingly different than one would expect. For example, if I saw African Americans voting for Bush 70 to 30 percent. But it doesn't really happen like that and it is very very rare that a poll is so way off.


Snoqualmie, Wash.: What's your point of view of the Gallup methodology regarding likely voters? Some criticize them for overstating likely Republican turnout or or understating Democratic turnout.

John Zogby: We have our methodological disagreements and they are only methodological. But I feel that their screen for likely voters is too tight and some of their math that they use in factoring likely voters is too fuzzy. But they are good people and good pollsters. We just disagree.


Tripoli, Libya: Can you please talk about the use of the term "likely voters" and how first-time voters are captured (or not) in polling?

John Zogby: A very good question. Likely voters are those who for reasons of history or likelihood tell us that they are going to vote on Election Day. It is very important to accurately capture likely voters so that we can develop what we call a turnout model as close to what the turnout will be on Election Day. Most of that is good science. But some of it is artistry.


Mililani, Hawaii: Will Hawaii be a real battle ground state this year? Could our 4 electoral votes really be the deciding factor or will the past history of solidly and safely Democratic continue?

John Zogby: It is very hard for me to see Hawaii as a battleground state. I have not polled it. However, if the polling that I have seen is true and the scenario plays out with this election being so close than indeed Hawaii can be pivotal in who gets elected. But I still see it as a blue state.


Phoenix, Ariz. Do you have a sense about some of the other 'Bush leaning' states out west -- New Mexico, Colorado, Nevada, and even my state -- Arizona?

Are they still pretty much Bush leaning?

John Zogby: They are all Bush leaning. My polling is showing the president opening up wider leads in Nevada and New Mexico than I had previously anticipated. I felt that Arizona was competitive but it is less so since Kerry pretty much threw in the towel last month.


Highland Park, N.J.: How are you factoring in the large number of young people who seem to be eager to vote this time?

John Zogby: We always get an under representation for young people and have to apply a weight to bump them up in our sample. We are bumping them up a little higher this year because there is a considerably greater amount of enthusiasm and likelihood of their voting. I feel confident that we are capturing young voters.


Toronto, Ont.: Mr. Zogby-

I see that you are teaming up to poll 12,000 mobile phone users. Would you care to speculate a guess on what direction you think these folks might be leaning, or at least how their issues are similar and/or different from your other samples?

John Zogby: Yes, we are. It is an exciting project with both Motorola and Rock the Vote. No-one of us partners claims that this will be scientific but it is path breaking and at least the first window into the world of cell phone users and political habits. I am very excited about it because it is pioneering.


Washington, D.C.: A question about exit polling:

With so many early voters this year, will exit polls be able to accurately reflect what is actually going on next Tuesday?

Will calling a state with confidence for either candidate be possible?

John Zogby: There is a lot of good questions there. First of all there shouldn't be any reason to believe that early voters are any different than election days voters but they won't be captured in exit polls. Number two I don't think that exit polls should ever be used for projections. There are too many states that are too close to call and exit polls are very good but they involve sampling. Instead I think that we should all just wait for the returns to come in.


Ithaca, N.Y.: Love you. Love the fact that you love Utica. Why do the campaigns keep internal polling so close to their vests? How does it mesh with your very good work?

John Zogby: Thank you Ithaca, one of the prettiest cities in the United States. It really is. Private polling is for internal campaign usage and it is used for strategic purposes and tactical purposes and many states have laws that require pollsters to release the entire poll if only a part of it is released, and so not surprisingly there is a lot of information in internal polls that campaigns don't want to see the light of day. But interestingly it is always intriguing to see the strategy and the messaging play out validating what my own polls are showing.

What I try to do is open up the window to citizens so they can see the inside world.


Washington, D.C.: What percentage of eligible voters will participate in the 2004 presidential election? In the last several election cycles the participation rate has been around 50 percent although this year it could increase to 55-60 percent do to the large in increase in voter registrations.

John Zogby: I am in the category that is predicting around 55 percent, which is high. Last time we had 55 percent was 1992 with Ross Perot. The modern day record in 1960 with about 60 percent.


Concord, N.H.: As I recall, the national polls in late October 2000 were about where we are today -- slight lead for Bush. Is there any reason to expect a different dynamic (late shift to the Democrat) this time?

John Zogby: It is eerily similar to 2000, in that we have been posting in our daily news releases the day-by-day 2000 numbers which track very similarly to know. The key thing here is that the race is close like 2000. Neither candidate gets more than a two day advantage like 2000. There are two big differences are 1) that there is an incumbent president 2) higher energy on the democratic side this year than in 2000.


Deming, N.M.: What issues seem to be motivating voters in who they are choosing for president: is it the economy, is it the war, is it terrorism, or is it all three, or two out of three, or what?

John Zogby: The top four issues are the economy, the war on terrorism, the war in Iraq and healthcare. The economy has stayed #1 all year, but the most emotional and intense issue is the war in Iraq. Of the four top issues Kerry wins double digits in three of the four, but the president leads overwhelmingly regarding the war on terrorism.


John Zogby: I always enjoy these. The questions are great. I wish I had time to answer all of them, but I thank The Washington Post and its readers for the opportunity to do this. Thank you everybody.


© 2004 Washingtonpost.Newsweek Interactive
Viewpoint: Paid Programming

Sponsored Discussion Archive
This forum offers sponsors a platform to discuss issues, new products, company information and other topics.

Read the Transcripts
Viewpoint: Paid Programming