News operates in cycles, and the current news cycle has clearly been focused on the fact that George W. Bush is leading John Kerry in the polls. Just what is the current election polling situation?
Every poll with which I am familiar that has been conducted since the conclusion of the Republican National Convention shows Bush with at least a small lead over Kerry. The latest CNN/USA Today/Gallup poll has Bush at 52% and Kerry at 45% among likely voters. That seven-point gap roughly reflects the average of what other national polls have shown in the last several days.
Gallup data show that Bush gained slightly between the last national poll we conducted before the Republican convention and the first national poll we conducted after the convention. Such a convention bounce is typically a reflection of intense focus on a candidate during the four days of his convention, and isn't surprising.
What's more interesting this year is the fact that Kerry did not enjoy a bounce after his convention.
Why don't all polls show the exact same figures?
There are a number of reasons.
Polls are based on samples taken from a large population. Those samples are used to estimate what the results would be if everyone in the population were interviewed. The good news is that these estimates -- if based on scientific random probability sampling procedures -- do an excellent job of approximating the opinions of everyone in the population. The other news (I hesitate to call it bad news) is that the data from a sample are estimates, not exact figures that would be found if the entire population was indeed interviewed.
So from sample to sample, there is variation in the estimates. Even three Gallup Polls conducted simultaneously would probably not show exactly the same results, but the results would be within a few points of one another. That's why every poll has a margin of error.
Also, in an election year, different polling firms ask the "whom are you going to vote for" question in slightly different ways. Polling firms also have different procedures for determining likely voters, which is an important part of pre-election polling. All in all, it is a wonder that poll estimates are as close to one another as they are.
Back to the election. Does the fact that Kerry is behind by seven points mean that he is in really big trouble?
Not necessarily. Candidates' standings in the polls can change significantly during the last two months of a campaign.
Also, it is useful to remember that a seven-point lead may look more significant than it really is. If Bush loses just three points, and Kerry gains those three points at the same time, the race would suddenly be 49% to 48%. Of course, it can work the other way around. If Bush gains three points and Kerry loses three points, Bush would then have a more dramatic 13-point lead. In short, nothing is set in stone at this point.
The Commission on Presidential Debates has scheduled three debates, although there are news reports that they might not all occur. Can debates really make a difference in the election outcome?
Yes. The debates are the last remaining major planned news event of the campaign. No matter how many debates are finally agreed upon by the two campaigns, they have the potential to affect the race significantly.
Most analysts believe that in 1980, Ronald Reagan improved his chances significantly in his debate with Jimmy Carter. Al Gore may have lost the 2000 election with his less-than-stellar performances against George W. Bush in that year's three debates. And of course Richard Nixon's sweaty performance in the 1960 debate is credited with helping John F. Kennedy establish his credibility as a legitimate presidential contender.
Does Gallup play a role in finding undecided voters to ask questions in the second planned debate?
Yes. For the last three elections -- 1992, 1996, and 2000 -- the Commission on Presidential Debates has called on Gallup to use random scientific sampling procedures to select undecided voters who come to the debate venue and ask the candidates questions. We did this for the Clinton-Bush-Perot debate at the University of Richmond in 1992, for the Clinton-Dole debate at the University of San Diego in 1996, and for the Bush-Gore debate at Washington University in St. Louis in 2000.
This year, the town hall debate is scheduled for Oct. 8, again at Washington University in St. Louis.
As a pollster, it's fair to say that you like the town hall debate format?
Absolutely. I believe it is a great idea to have real people ask the candidates real questions. Journalists can ask smart, informed questions, but not always the ones that the American public would like to see answered. I hope this St. Louis debate goes forward as scheduled.
Some press accounts indicate that the Bush campaign is concerned about the possibility of partisan questions being asked at the debate -- even though the sample is "undecided voters." In that regard, it's important to remember that Gallup will recruit the respondents as part of a random probability sample of the St. Louis metropolitan area, and the people we recruit will have no idea why we are asking them questions until they have met the screening qualifications. Plus, all of the screens used to identify undecided voters are reviewed and approved by representatives from both parties before the recruitment goes forward.
Does the attention given to Kerry's Swift Boat situation and Bush's National Guard service indicate that character is becoming a more important dimension of this year's campaign?
Gallup's Sept. 3-5 poll suggests that in general, there may be more of a focus on the candidates, rather than the issues, this year than in 2000.
We periodically ask voters: "Which of the following do you think will be most important to you when you decide who to vote for -- where the candidates stand on issues that matter to you, or the leadership skills and vision that you think the candidates would have as president?" Both times when we have asked the question this year -- on Jan. 2-5 and Sept. 3-5, likely voters have chosen the "leadership skills and vision" side by a 10- to 12-point margin.
In 2000, the "stance on the issues" response won out every time Gallup asked the question in the final months leading to the election.
It is also important to note that there has definitely been deterioration in Kerry's overall image positioning in Gallup's most recent poll, regardless of whether it was a result of the Swift Boat ads or the Republican convention.
Only 55% of Americans now say that Kerry can "handle the responsibilities of commander in chief of the military," down from 63% just after the Democratic convention. During the same time period, the percentage of Americans saying Bush can handle these responsibilities rose from 61% to 68%. Plus, Kerry's unfavorable ratings have risen from 37% after the Democratic convention to 43% in the Sept. 3-5 poll.
What are the key issues driving this campaign?
As is true with any presidential campaign, there are a wide variety of issues that come in and out of focus as the days of the campaign wind down inexorably toward Election Day itself.
This year, in particular, has included extensive coverage of international issues, particularly terrorism and the war in Iraq. And of course there is the major focus on the Swift Boat/National Guard controversies, suggesting that people believe that what Kerry and Bush were doing 30 to 40 years ago is an indicator of what type of president they will be over the next four years.
While all of this is going on, Gallup data continue to show that the average American remains vitally concerned about more current, and perhaps more prosaic, issues: the economy and the cost of healthcare.
A recent New York Times article by James Dao used Gallup data to compare the public's reaction to the Vietnam War in 1965 with the public's reaction to the Iraq war now. Has dissatisfaction with Iraq built much quicker than dissatisfaction with Vietnam?
Yes. It took about three or four years of involvement in Vietnam before a majority of Americans told Gallup that the war there had been a mistake in August 1968. A majority of Americans were saying that the Iraq war was a mistake in June of this year, less than a year and a half after the initiation of military action in Iraq.
However, opposition to the war in Iraq (based on the results of Gallup's classic "mistake" question) has actually dropped in recent weeks after reaching 54% in two polls in June and July. According to the Sept. 3-5 poll, only 38% of Americans say that America's involvement in Iraq was a mistake while a clear majority, 57%, say that it was not. It's not clear why this "mistake" number has declined, particularly when the number of U.S. casualties is going up.
Does Kerry have any advantage over Bush in terms of Iraq?
No. It's the other way around. Fifty-four percent of Americans say that Bush can do a better job than Kerry on Iraq, while 41% choose Kerry. After the Democratic convention, the two men were tied on this measure.
There is continued speculation that Osama bin Laden will be found or killed before the November election. Would that make a difference in the outcome?
Apparently it would. The Sept. 3-5 poll asked voters how important it is for the United States to capture or kill Osama bin Laden. Two-thirds (68%) feel it is extremely or very important, while only 9% think it is not too important or not at all important.
By the way, 66% of Americans say that it is either very or somewhat likely that Osama bin Laden will eventually be captured or killed. That's down from the 78% who thought so in November 2001.