Agenda masquerading as analysis

Froma Harrop's column "Palin Drove Stake Into Centrist Hearts" (here) is a fine example of a superficial analysis cooked up to justify a pre-formed conclusion. In this case Ms. Harrop does not care for Gov. Palin's social conservatism and so sets out to prove that picking Palin destroyed McCain's chances.

Just look at the RealClearPolitics poll averages dating back to early September. The McCain-Palin numbers started cratering about a week after the convention, which was two weeks before the stock market did.

Except they didn't.

McCain did not "crater" until Sept. 24th through 26th, and the economic crisis really began to dominate the headlines beginning Sept. 15th with Lehman Brother's bankruptcy filing.

This is clearly evident if you look at the post-convention detail chart below.

I have plotted the mean of the Rasmussen and Gallup daily tracking polls over the period in question (using a couple of reliable tracking polls instead of the RCP average gives a better indication of responses to specific events because the methodology is consistent throughout the time period). Note that the McCain line jumps dramatically during the GOP convention (Sept. 2 through Sept. 4) and stays higher than the pre-convention level thereafter, taking a small hit around the Lehman Brothers bankruptcy and AIG rescue but stabilizing again before falling off sharply on Sept. 24-26. This corresponds exactly to the announcement of the "bailout" proposal, McCain's suspension of his campaign to go back to Washington, and the off-again on-again first debate.

Clearly, voters were not impressed with McCain's reaction. Governor Palin was not the cause, however much Froma Harrop wants to blame McCain's problems on picking a pro-life running mate.