When you think about it, you see that Gore bungled even worse in 2000 than Kerry is blowing it this year, because Gore _should_ have won; the two traditional benchmarks of peace and prosperity were solidly in his favor. Kerry has a much tougher row to hoe this year, in truth.

The 2000 poll cited also came just before The New York Times unleashed its barrage of claims about the Bush's campaign's subliminal advertising -- the "DemocRATS" controversy (that was actually reported first 10 days earlier by Fox News, but treated as a joke by Brit Hume & Co.). It helped add several points onto Gore's lead later in the month, but that turned around when the first debate arrived and Al gave Howard Dean his template on how to run his 2004 campaign.

I fully expect the Times and other outlets to come up with some other insubstantial claim that they'll try and blow up into a major controversy in the upcoming days in another effort to drive down Bush's pre-debate numbers. The problem is that unlike 2000, Bush is a known quantity to the American public; the issues surrounding this election are far more life-and-death than what the public could perceive in 2000, and even if they do stir up a kerfuffle about nothing, John Kerry's still going to have to defend his senate records at some time during the debates, which even the Times' edtiorial writers haven't made the effort to defend in the wake of 9/11.

Is there any support for Nagourney's assertion that Labor Day weekend polls are 'quite unreliable'? I understood that weekend polls tended to overstate Democratic input, but why would this make the TIME/Newsweek poll results showing a Bush lead unreliable?

Labor Day polling is thought to have inherent unreliability because voters, especially more affluent ones, are away. Traditionally, they have polled more Democratic.

Maybe, Labor Day is big on Fire Island this year so Nagourney thinks that enough of his rump-rustler cohorts aren't being polled(over the phone anyway)?