MU's +11 Dem Skewed Poll
I have no doubt that the senate race has tightened -- that Tammy Baldwin has gained while Tommy has snoozed. As I said Wednesday, Thompson has to realize that this is isn't 1998 anymore. But the normally reliable MU Law School poll is nigh-on absurd. The 18-point swing in just a month should raise more than a few flags. How does Baldwin go from 9 points down to 9 points ahead? Could it be the skewed sample? Consider these internals:
August poll: Thompson + 9 (50/41), with a Republican/Democrat sample of +1 Democrat (43/44)
September poll: Baldwin +9 (50/41), with a Republican/Democrat sample of +11 Democrat (40/51).
Unless Wisconsin voters have suddenly swung hard left, this makes no sense. We are, at best, a closely divided state, and recent elections have demonstrated the durability and motivation of Republican voters.
What changed in one month? MU's sample. In August the MU pollsters found a closely divided state with Tommy out-performing the normal partisan loyalties. Just a month later, they issue a poll on a dramatically different electoral universe in which partisan voters line up with the nominees of their own party. This may reflect some erosion for Thompson, but the topline results: absurd.