Friday, August 06, 2004
By Brad Bannon
Watching the polls is not nearly as interesting as watching Catherine Zeta Jones but it is an election year and a presidential election year at that and we are political professionals so we should be keeping close tabs on the national political surveys. And if we use the national polls to understand the presidential race, we should use them for good not for evil. The biggest problem with the national political surveys conducted by news organizations is the reliance and the prominence of the trial heats. The head to heads are the first and sometimes only thing that the sponsors report. But if you really want to understand the public opinion environment that governs the presidential race, you have to go to the poll websites to look under the hood and kick the tires.
Read more after the jump…
You can look at trial heats until the cows come home, but the only set of numbers that really mean anything at this point in the presidential campaign are the right direction and wrong track figures. Bush may be a few points ahead of Kerry or a few points behind but recent polls indicate that a majority of Americans feel the country is going in the wrong direction. A June Gallup national poll showed that two out of three Americans think the country is headed in the wrong direction. It will be very difficult for President Bush to win a second term if Americans still feel this pessimistic on Election Day.
As Democrats learned the hard way, the popular vote that the national polls measure is not the bottom line in a presidential election. In fact the networks could eliminate their polling programs and save millions of dollars and keep an eye on the people who actually elect the president; the justices of the United States Supreme Court. But if we want to keep tabs on voters through the national polls, we should look at the surveys intelligently or not use them at all.
The focus of the activity in 2004 presidential race is in 18 battleground states which I will not take the trouble to list since we all know which states they are. Because these battleground states are so important to the outcome of the presidential election, the national polling totals are not particularly useful. In other words, it doesn’t help John Kerry much if he is ahead in a national Pew Center survey by five points, if he is losing the battleground states by five points. When CNN commissions a Gallup survey, it usually reports the results of the presidential head to head nationally and cross tabulations for the voters who live in the swing states.
In life, the devil is in the details and in survey research the angel is in the crosstabs. It is nice to know that John Kerry is ahead by a few points but it is a lot better to know how the presidential candidates are faring with key voter groups such as the people who are living in the battleground states, soccer moms, NASCAR dads or political independents.
It is not uncommon to see differences in the heads to heads among national surveys conducted at about the same time. There are several reasons for the discrepancies in the trial heats.
The order of the questions affects the head to heads. When Ronald Reagan was President, he always did better in surveys when the trial heat was at the beginning of the survey than he performed in surveys where the head to head was towards the end. Why? Apparently voters who supported President Reagan were less likely to do so after they had the chance to hear and respond to questions about the issues of the day.
The methodology that pollsters use to select respondents also influences the outcomes of the trial heats. Some surveys interview Americans they consider likely to vote; some question all registered voters and others interview all Americans 18 or over whether they are registered voters or not. Why interview unregistered Americans?
Because both the Democrats and Republicans are working hard to register millions of new voters between now and Election Day so pollsters should be trying to figure out what politically inactive Americans might do if they get involved. In a June Gallup survey, Kerry had a 2 point lead among all Americans; a 3 point lead with registered voters and a 6 percentage point lead among likely voters.
Trial heats have the shelve lives of J-Lo’s marriages. Even accurate head to heads give the misleading impression that attitudes are set in stone. In March, the Pew Center conducted a national survey which showed Senator Kerry leading President Bush by a margin of 48% to 46% with only 6% of the voters undecided.
But when the Pew Center factored in voters who had only a tentative commitment to supporting their choice, the result was that almost a third of all voters were up for grabs. That’s way there will be more twists and turns in this year’s presidential race than there are in a rollercoaster.
Brad Bannon is president of Bannon Communications Research which for 20 years has designed poll driven messages for Democrats, labor unions and issue groups. You can reach him at Brad@BannonCR.com
