Thursday, September 18, 2008

Math Lesson 2

Following up on the lesson from Jed regarding cable TV math, there's this from today's NY Times:

And the New York Times/CBS News poll [taken this week] found no evidence, at least to date, that Ms. Palin has allowed Mr. McCain to expand his appeal to women voters or independent voters. Polls taken immediately after the convention had found evidence of a sharp increase in support for Mr. McCain among white women, but this poll suggests that that effect was, so far at least, limited. White women were evenly divided between Mr. McCain and Mr. Obama; before the conventions, Mr. McCain led Mr. Obama among white women, 44 percent to 37 percent.

Got that? The latest poll shows that McCain is now doing significantly worse among white women than he was before the convention, despite a temporary convention bounce/novelty blip. Realize that replacing "were" with "are now" in that last sentence would clarify things immensely.

I guarantee you that if the candidates were switched, the NYT would phrase it something like this:

And the New York Times/CBS News poll found no evidence, at least to date, that Ms. Whomever has allowed Mr. Obama to expand his appeal to women voters or independent voters. Indeed, his choice of her for a running mate seems to have hurt. Polls taken immediately after the convention had found evidence of a sharp increase in support for Mr. Obama among white women, but this poll suggests that that effect was short-lived. White women are now evenly divided between Mr. McCain and Mr. Obama; before the conventions, Mr. Obama led Mr. McCain among white women, 44 percent to 37 percent.

This seven-point plunge has caused serious worries, say some observers, who spoke on condition of anonymity.

3 comments:

Anonymous said...

I would like to know, for the sake of having more complete information what the polls are for all women and not only white women.

It is possible that non-white women had favored Obama both before and after and the "Palin's effect" was negligible.

bjkeefe said...

That's a good point, and you should send Adam Nagourney an email and tell him so. (Email link on this page.)

I can imagine, though, that the polls are not that finely sliced, for several reasons.

First, the number of non-white women in any random sample of "all voters" or "likely voters" is likely to be small enough that the margin of error becomes too large to be able to say much about that sub-group.

Second, "white women" are one of the perennial narrative themes for the MSM -- soccer moms, security moms, and now hockey moms. Not something to be happy about, but it is a fact of conventional-wisdom-reality.

Third, I'd bet tall dollars that Obama has at least a 70-30 lead among non-white women, if not a lot higher, and this has probably been so since South Carolina. A reporter would say that this is not news, since it hasn't changed.

But, as I said, you're right to be annoyed by this omission, so I urge you to let Nagourney know. I've written to him several times to complain, and he's pretty good about responding.

Anonymous said...

I wasn't really annoyed. And I figured some of the reasons why only white women are included. I was just curious...

ShareThis