I'd like to call your attention to two things that happened over at Gallup.com on October 10, 2012, just 8 days ago and exactly one week after the first presidential debate. As usual, they released their tracking poll result covering the previous 7 days, Oct. 3-9, a period including what other polls have called Romney's two best polling days all year, Oct. 4 and 5. Here was the result:
Oct. 10: Romney 48, Obama 48
On the same day, Gallup released a statement outlining a bunch of changes they were making to their tracking poll methodology. I'll detail some of those in a minute. Here is what happened in the tracker after these changes were announced:
Oct. 11: Romney 48, Obama 47
Oct. 12: Romney 49, Obama 47 (Romney's best day falls out of tracker)
Oct. 13: Romney 49, Obama 47 (VP debate on the 11th should be felt now)
Oct. 14: Romney 49, Obama 47
Oct. 15: Romney 49, Obama 47
Oct. 16: Romney 50, Obama 46
Oct. 17: Romney 51, Obama 45
Oct. 18: Romney 52, Obama 45 (2nd presidential debate should be reflected)
To recap, a full week after the first debate they had it tied, then R+1, R+2 for a few days, R+4, R+6 and now R+7. All of this in the 8 days after announcing new procedures for the tracking poll they've been running all year.
What were those changes for the final month:
- they separated the presidential tracking poll from a larger survey it had been embedded within.
- they went from 40% cell phones to 50% which leads to...
- "This means that our weights to various phone targets in the sample can be smaller, given that the actual percentage of cell phones and cell-phone-only respondents in the sample is higher."
- And: "We have instituted some slight changes in our weighting procedures, including a weight for the density of the population area in which the respondent lives. Although all Gallup surveys are weighted consistently to census targets on demographic parameters, we believe that these improvements provide a more consistent match with weight targets."
- made some "slight" changes to the likely voter screening questions
- moved these 7 LV questions to the beginning of the survey
- the shorter presidential race only survey has increased response rate
Gallup concedes the following with regard to all the changes:
All in all, it is possible that these changes in methods, which we believe increase the representation of our overall samples, may have some impact on political or other measures included in the surveys.
And the changes announced on Oct. 10 may not represent the only changes Gallup made:
We will most likely make other changes to our survey procedures as our survey methods and procedures evolve.
Now, Nate Silver has a pretty
good piece up explaining the history of Gallup polling on presidential races. they have shown swings of even greater magnitude (Gore +26!) in several past races when other pollsters did not.
But my question to readers is this: could Gallup's changes announced Oct. 10 have something to do with the big swing toward Romney? A swing that occurred a week later than it should have given news events?