This “Taylor Swift effect” in Tennessee might be real after all
This morning, in response to questions from several Palmer Report readers, I addressed new polling data in the Tennessee Senate race from Reuters, which had Democratic candidate Phil Bredesen suddenly within three points. The question, of course, was whether this could be attributed to the “Taylor Swift effect.” The answer, as of this morning, was that we’d have to wait for more data. As it turns out, thanks to another new poll, we haven’t had to wait long.
The confusion began when the New York Times released a poll showing Republican candidate Marsha Blackburn up by a whopping fourteen points, right around the time Taylor Swift was endorsing her opponent Phil Bredesen. Cable news played up this poll as if it were the only poll that mattered in the race, while Palmer Report pointed to the fact that the average of the major polls in that race showed Blackburn being up by something closer to 5.5 points.
The trouble with yesterday’s new Reuters poll is that it was the first time Reuters had weighed in on this race. Sure, it has Bredesen down by just three points, but we don’t know where the Reuters model would have had Bredesen pegged before the Taylor Swift endorsement. So while it certainly helped Bredesen in the polling averages, it wasn’t clear if this was a trend. But now a new poll from Vanderbilt University says that Bredesen is ahead by one point. So now what?
Again, there’s nothing more dangerous in polling analysis than comparing apples to oranges. But in a race where none of the polling outlets are moving quickly enough to give us apples to apples comparisons, we’re left looking at the fact that Phil Bredesen’s two best poll numbers this month – by far – have come in the aftermath of Taylor Swift’s endorsement and the huge number of people who subsequently registered to vote. In any case, the current poll numbers matter less than how much time and effort the Resistance puts into the Tennessee race over the next three weeks.
