Over at NRO's The Corner, Yuval Levin makes the (hilarious) case that on the economy, the past week has been better for McCain than Obama:
Both candidates started the week pretty poorly on the economic crisis, but McCain seemed to gain his footing by week's end while Obama didn't. Perhaps voters noticed, though we will need to see a few more days of data to know if there's a real trend here.
To make his case, Levin cherry picks two data points from the Diageo/Hotline tracking poll showing that on Tuesday, Obama held an 11 point advantage on the economy but by Friday Obama's advantage was just 3 points. Levin's numbers are accurate, but all he's doing is accurately measuring statistical noise.
The poll only asked this question of a subsample of respondents, so the margin of error is relatively high -- 4.6%. And if he'd picked Monday as the reference point, the change would have been a far less dramatic 4 point swing. Or if he picked September 10th, Obama would actually have gained ground, because back then, McCain actually had the advantage. The point is that these questions asked of subsamples tend to bounce around a bit.
Most importantly, Levin ignores the fact that virtually every poll shows Obama maintaining or expanding his lead over McCain. On Tuesday, Obama enjoyed a 1.5% average lead over McCain in the tracking polls. Today, it's 4%. I suppose there's a theoretical chance the economic debate has been good for McCain, but I wouldn't bet a dime on it.
Now, that all being said, the polls of today don't mean anything because the only poll that counts is the one taken on election day. But if Levin seriously thinks this debate advantages McCain, then that must mean he'd be willing to focus on the economy every single day for the rest of the election.
And something tells me he wouldn't accept such a proposition.
© Jed Lewison