Ten Miles Square

Blog

October 08, 2012 8:23 AM The Debate Is Not a Campaign Reset

By Mark Kleiman

Gallup has a seven-day horserace tracking poll. For the past week, Obama has been at either 49 or 50, and Rommey at 44 or 45. This morning’s numbers are the first ones that reflect a day’s post-debate polling: Obama 50, Romney 44.

RAND also has a seven-day tracker, based on a panel: they accumulated a sample of 3500 voters in March and got them agree to respond by emails once a week (500 each day) as to their percentage probability of voting and their percentage probability of voting for their current favorite. The numbers reported yesterday (which I assume were mostly pre-debate) were good for Romney: he was up 0.6 and Obama down about the same amount. But today’s numbers (one-seventh post-debate) went the other way, Obama up 0.7, Romney down 0.8, for a seven-day average of Obama 49.9, Romney 44.

The online IPSOS poll has “pre-debate” and “post-debate” samples. Post-debate shows Obama holding steady at 48% among “likelies” while Romney picks up some from “undecided/don’t know” and “wouldn’t vote” to move from 38% to 43%. The debate seems to have pulled Romney above water (at least temporarily) on favorable/unfavorable.

Good night for Romney, bad night for Obama? Sure. Missed opportunity to put the android out of its misery? Probably. Game changer? I don’t think so.

Footnote Yes, the RAND approach builds in a bias because people willing to play the game will include fewer people who completely tune out of politics, and playing will tend to make the respondents more attentive to the campaign than they otherwise would have been. Since Romney’s crapola makes sense only if you don’t listen carefully (or simply filter everthing through Fox News bias) I’d expect the RAND electorate to be marginally more pro-Obama than it would have been without the intervention. But that doesn’t mean a really earth-shattering event would fail to show up in the trend.

Also note that the RAND results are equivalent to a “likely-voter” screen, with the enthusiasm gap (Romney supporters about five points more likely to actually vote than Obama supporters) already factored in.

[Cross-posted at The Reality-based Community]

Mark Kleiman is a professor of public policy at the University of California Los Angeles.