It bears repeating: the electorate is awfully polarized this year. Gallup’s latest poll showed among likely voters only a 13% band of swing voters: 5% voting for Kerry who say they may change their minds, 5% voting for Bush who say they may change their minds, and 3% with no opinion. The balance have picked a candidate and say they’re certain: 42% for Kerry, 45% for Bush. All signs indicate that this years contest will be nearly as close a squeaker as 2000.
That said, the polls following last week’s Democratic Convention should give supporters of John Kerry pause. It could be predicted — and it was, by several pollsters — that Kerry’s bounce out of the convention would be modest by historical standards. And it was modest indeed:
* In the days following the convention, Newsweek showed a net gain of 4% for Kerry from three weeks ago for a 7-point lead among adults 18 and over.
* CBSNews showed a statistically insignificant net-1% gain in Kerry’s favor for a 7-point lead among registered voter.
* American Research Group showed a statistically insignificant net-1% gain in Kerry’s favor since a month ago for a 4-point lead among registered voters.
* And ABCNews/Washington Post showed a net 6% gain for Kerry since before the convention, for a statistically insignificant 2-point lead among likely voters.
I’ve listed those in ascending order of significance: generally, counts of likely voters are more accurate than counts of registered voters, which are more accurate than counts of adults.
What’s surprising is that, by some measures, Kerry actually lost ground. Much attention has been paid toward the USAToday/CNN/Gallup poll that showed Gallup’s first post-convention negative bounce since 1972: among likely voters, it showed Bush leading by 4%, a net loss for Kerry of 3% from the week before the convention. Many have dismissed this as a fluke, but unless the usual models for likely voters are wrong — that is, voters who’ve voted consistently in the past and say they plan on voting this year are not as much more likely to vote as pollsters assume — then Gallup, along with ABC poll showing Kerry achieving a tie, is probably among the more accurate post-convention polls.
Of special interest is the Rasmussen daily tracking poll of likely voters. It’s been stubbornly tied for months; with the exception of a brief period after Kerry selected John Edwards as his running mate, neither Bush nor Kerry had opened up a sustained significant lead up until last week. Rasmussen did show a bounce for Kerry during the week of the convention, with 2-to-4 point leads throughout the week. (Each day’s results represent an average of the previous three nights.)
But in yesterday’s results, the first taken completely after Kerry’s acceptance speech, the bounce was all but gone: Kerry had an insignificant 1-point lead. That suggests that likely voters in the survey were either turned off by Kerry’s speech, or had, over the weekend, lost any affection they may have accrued for the Democratic ticket.
Perhaps worse for Kerry is that, for all the militaristic imagery of the convention, the convention didn’t help Kerry grab the center. Before the convention, 41% saw Kerry as “moderate,” 43% as “liberal.” After the convention, 38% call Kerry “moderate,” and 46% call Kerry “liberal.”
So Kerry’s attempt to ingratiate himself to middle America has resulted in at best a small bounce in the polls and at worst a negative bounce and an increased perception of his liberalism.
The electoral dynamics point to a race so close that either candidate might win. But John Kerry is nonetheless an exceptionally weak candidate.