"Surprising" Polls: Republicans plan to vote Bush

I'm usually not keen on discussing the "horse race" side of the election. Who's ahead toda? Who will be ahead tomorrow? It's exhausting and potentially depressing. But the media continues to do this and it appears that the candidates (and maybe the voters) pay attention to this coverage. I've read recently, for example, that John Kerry's campaign has pulled ads from Arkansas (and other states) because polls show the alleged swing state is becoming a lost cause.

Hmmmm. The Zogby interactive poll on the Wall Street Journal website has the state blue for Kerry as of September 20, by 0.1%.

The Milwaukee Journal Sentinel had this telling headline on September 22: "On average, Bush leads state, but poll numbers vary widely" . Since Wisconsin is a swing state that went very narrowly for Al Gore in 2000 (roughly 5000 votes), the Kerry campaign is taking it very seriously. Indeed, Kerry has spent parts of 4 days this week in Wisconsin prepping for the first debate.

So, what do the Wisconsin polls show?

According to a new statewide Badger Poll, Bush enjoys a clear advantage in Wisconsin, a state Democrats have carried since the 1980s and can ill afford to lose.

Bush...leads Democrat John Kerry by double digits in the poll....

A new ABC News poll of likely voters in Wisconsin offers similar findings. Bush holds a 10-point lead in that survey...A poll released last week by Gallup also showed Bush leading Kerry, in that case by eight points.

But that's not the whole story.

In non-partisan Wisconsin surveys released this week by the American Research Group, Mason-Dixon and Zogby Interactive, Bush and Kerry are within two points of each other.

Anyone following the national and state polls knows that these widely varying poll numbers are not unique to Wisconsin. These fluctuating poll results are being found everywhere.

Look closely at the Wisconsin polls, however, and one readily gets a feel for the problem:

Kerry spokesman George Twigg termed the Badger Poll flawed.

"The majority of polling continues to show this race is a dead heat. If there's any advantage (for Bush), it's very slight," said Twigg...

Twigg also contended that the pool of voters surveyed in the Badger Poll was skewed toward Republicans. Republicans accounted for 36% of those surveyed in the Badger Poll, Democrats 29%. In the ABC poll, Republicans made up 35% of the likely voters surveyed, Democrats 29%.

Twigg pointed out that Democrats were a bigger share of the state's 2000 voters than Republicans, according to exit polls.

[Political Scientist Charles] Franklin, of the University of Wisconsin, said that's one reason to view these polls as a snapshot rather than a forecast, since the makeup of the state's electorate in November is unlikely to be as Republican as it is in the latest Badger Poll.

The pollster John Zogby had a piece on this precise concern on September 7. He emphasized the known data about party identification as compared to the party ID used in recent prominent polls like Newsweek and Gallup:

If we look at the three last Presidential elections, the spread was 34% Democrats, 34% Republicans and 33% Independents (in 1992 with Ross Perot in the race); 39% Democrats, 34% Republicans, and 27% Independents in 1996; and 39% Democrats, 35% Republicans and 26% Independents in 2000. While party identification can indeed change within the electorate, there is no evidence anywhere to suggest that Democrats will only represent 31% of the total vote this year....

This is no small consideration. Given the fact that each candidate receives anywhere between eight in ten and nine in ten support from voters in his own party, any change in party identification trades point for point in the candidate's total support.

Point-for-point. That's enormous. Look at the 2 Wisconsin polls using 35 amd 36% Republican likely voters but only 29% Democratic. Make the new totals something like 34% Democratic and 33% Republican and the results would be dramatically different. The apparent and widely publicized 8 to 10 point Bush lead suddenly becomes something like 2 or 3%. Statistically, that's almost a tie...or nearly a coin flip.

It's the same in the national polls. Because Gallup and other organizations have been using the same methods all year, it also means that Bush was probably very far behind for much of the year, when the race was portrayed as close.

Bush has clearly gained, and probably has a small lead, but it is not at all time to panic. Movement by just 1% of the electorate from Bush to Kerry would make this race a virtual tie again. Movement by 2% and Kerry would have a small lead.

Keep working. Don't despair. In 2000, late October polls showed Bush with a 5% lead (or more), Bush took a day off from campaigning, and then Karl Rove sent him to California to demonstrate his confidence in the cakewalk.

We know what happened. The same political team pulled a similarly deceptive stunt on Iraq.

Keep the faith.