By John Tabin on 10.4.04 @ 1:24AM
The numbers offensive is just getting started.
Some readers -- and not just the usual liberal gadflies -- think
my Friday "Bush Won" column
was nuts. To review: As I wrote
before the debate, if the match-up either shifted the dynamics of
the race in the President's favor or left them unchanged, it would
be good for Bush, whereas Kerry needed a significant move in his
favor. After the debate, which Kerry certainly "won" on points, I
saw little reason to think the electorate would be moved very much.
I predicted that the Senator "might get a small bounce in the
polls, but probably not enough to fundamentally change the
trajectory of the race," and by default declared Bush the
winner.
Guess what? There's ample reason to believe that I was
right:
• When ABC took an instant viewer poll last week, their
sample supported George W. Bush vs. John Kerry, 50% to 46% before
last Thursday's debate. After the debate, the same interviewees
supported the President 51% to 47% -- no net change.
• In Rasmussen's daily tracking poll, 6 percent of
respondents said Friday that the debate changed their minds: 3
percent who moved into Kerry's corner, 2 percent who moved into
Bush's, and 1 percent who moved into the Undecided column -- a
paltry movement in Kerry's favor.
• The L.A. Times poll showed that among survey
participants who watched the debate, who already favored Kerry by 1
point last week (when the total sample of registered voters in
general favored Bush by 4 points), his lead increased to 3
points.
• In a survey done in conjunction with Knowledge Networks
by Democracy Corps, a Democratic firm, in a
demographically-balanced sample polled over the Internet, Kerry
garnered a 2-percent gain in support during the debate.
(Incidentally, partisan polling firms -- this one is run by James
Carville, Stanley Greenberg, and, when he's not on leave playing
guru to a campaign, Bob Shrum -- have a statistically-proven bias
toward the home team in the data they release publicly.)
Against all this stands the Newsweek poll, showing
Kerry leading 49-46, a net 7-point swing from September 9/10, when
the magazine showed Bush leading 50-46. But just as
Newsweek oversampled Republicans after the New York
convention and overstated Bush's bounce, here they seem to have
oversampled Democrats and overstated Kerry's. For the first night
of the poll, conducted after the debate, most of the Midwest and
all of the South were excluded from the sample -- it was too late
to call people in the Central and Eastern time zones. Mark
Blumenthal, Mickey Kaus's longtime "Mystery Pollster" whose young
blog has quickly become an invaluable psephological resource,
writes that the Friday and Saturday portion of
the sample might also be suspect because "weekend-only interviewing
yields respondents that are more aware of current events and
political figures even after demographic weighting."
Newsweek's trendline toward the president swung from a 39%
Republican/34% Democratic sample on September 9/10 to a 34%
Republican/36% Democratic sample after the debate. Voters' party ID
does change (John Zogby is almost alone among pollsters in using a
method of weighting that assumes that it doesn't), but probably not
this much.
The USAToday/CNN/Gallup poll shows, among likely
voters, a 49-49 dead heat, compared to an 8-point lead for Bush
before the debate. Blumenthal's point about the tendencies of
weekend polls may also apply to Gallup (the party breakdown of the
sample is not yet public as I write). But since this is also the
first poll to include a sample taken on Sunday, it might also be
registering Kerry's tremendous win in the weekend spin war,
something that was hard to predict; both sides came out of the
match with plenty of ammunition. But Saturday Night Live,
where the debate skits were dismayingly influential in 2000, was
harder on Bush than Kerry, and our esteemed press corps has reached
a clear consensus that Kerry scored huge and has the wind at his
back. Will this storyline persist in their coverage? It's
impossible to tell; with a vice-presidential debate tomorrow, a
town-hall presidential debate on Friday, and a potential
mini-scandal brewing over whether Kerry is a big
fat cheater, we're in for a week full of surprises.
topics:
NATO
