Pew Research shows Bush leading Kerry 49-44% among likely voters, and 48-41% among registered voters. Only half of Democratic voters think that Kerry will win in November.
The CBS poll finds Kerry has made up five points among registered voters, and they are now even at 47-47%. As with all the polls, Bush’s favorables still look good, although Kerry has gained a few points in some categories: 52 percent of registered voters have confidence that Mr. Bush could protect the country from terrorism. Before the debate, only a quarter of registered voters had confidence that Kerry could; now, 39 percent do. And so on.
And these numbers are after the debate last week. Kerry must be getting desperate. Nothing he does gives him any traction. He picks a running mate with huge approval ratings. No bounce. The media pretends for two weeks that he had a successful convention. No bounce. He gives just about the best performance he could have at the debate, while Bush did poorly, and he can only pull within 6-7 points. Not good, unless you’re a Republican.
NYT/CBS, like the poll that came out over the weekend, has flipped the percentages of Democrats and Republicans in October from what they were using last month (when they showed Bush ahead). Also, it looks like NYT/CBS sampled self-described "liberals" pretty heavily, and did this poll over a weekend, which also has historically tended to produce a Democrat-leaning respondent pool.
The guys at Power Line have the goods:
https://powerlineblog.com/archives/008078.php
My take, FWIW: While Bush should by no means be complacent (I think his decision to schedule a major national-security and economic-policy address for tomorrow is a good move, BTW), Kerry should be really worried that even with his friends in the dinosaur media sandbagging the polls for him to set up his big "comeback," he still can’t do better than a tie in the horserace number, and can’t reverse Bush’s positives on the "internals" in these surveys.
historical trends for incumbent presidents show that they never acquire more of the vote than pre-election polls show and they often acquire less. So for Bush, polling between 47-49% is a zone in which he will lose. Challengers gain 2-8% with the average being 4%. Bush must crack 50% to win. In battleground polls, he is polling closer to 47%. That may not match one’s intuition, it’s just what history tells us. more information here:
https://www.prospect.org/web/page.ww?section=root&name=ViewWeb&articleId=8694
The story that you linked to was quite weak. It showed that Carter got 1 point less, Reagan actually gained 1 point, Bush Sr. got exactly what the polls predicted, and Clinton lost 2 points. With the possible exception of Clinton, who is an outlier because Ross Perot got nearly 10% of the vote in 1996 (far more than Nader has any chance of getting), these numbers actually show that polls are remarkably accurate in predicting performance. Even if it is generally true that undecideds break in favor of challengers, it does not follow, per the American Prospect’s suggestion, that Bush needs "The Big Five-Oh." Nor is the race closer in the alleged "battlegrounds" than it is generally. Bush’s margin is in fact larger in Missouri and Ohio than it is nationwide.