With the mass elation today from Republicans cheering the Gallup poll that shows Romney inching to his biggest lead yet, 7 percentage points among likely voters, it is important to note a few caveats about how to read Gallup’s daily polls:
1. Gallup uses seven-day tracking for the head-to-head contest. This means they (should) have less noise in their poll, because a small bounce will be averaged out over seven days. It also means that it can take longer for movement to register, which is what happened after the first debate. Today’s polling includes only one day since the second debate. Whether or not that debate ultimately changes anything either way remains to be seen.
3. One of the most frustrating aspects of Gallup’s polling is that their job-approval numbers are based on all adults. This gives a much rosier picture of Obama’s real approval than what Election Day voters will feel, and makes comparisons with the head-to-head numbers impossible. On average, net job approval drops between 4 and 6 percent when moving from registered voters to likely voters. Going from all adults to likely voters would likely be an even larger change. Today’s 50–44 would at best be even among likely voters at, you guessed it, 47–47. The one advantage of the three-day rolling average for approval ratings is that if you see big swings (and with Gallup, approval swings all the time), you can make an educated guess whether it was a better or worse day for Obama. But even with that said, it’s still just a guess, because “all adults” is such a wider population than either registered or likely voters.
We won’t know for at least a few more days if the second debate had any real impact on the trajectory of this race, but these early signs from Gallup today, along with tracking from Rasmussen (Romney gained a point from yesterday) and IBD/Tipp (Romney gained two points from yesterday), suggest that if Obama gained anything from the debate, it will be nowhere near what he lost in the first round.

View Comments