Quinnipiac (2/22-28, registered voters, 12/8-14 in parentheses):
Arlen Specter (D-inc): 49 (44)
Pat Toomey (R): 42 (44)
Undecided: 8 (11)
Joe Sestak (D): 36 (35)
Pat Toomey (R): 39 (40)
Undecided: 24 (22)
(MoE: ±2.6%)
Arlen Specter (D-inc): 53 (53)
Joe Sestak (D): 29 (30)
Undecided: 14 (15)
(MoE: ±3.9%)
This ought to shake up some of that growing pessimism about the Pennsylvania Senate race -- although most of that pessimism seems to come from the fact that only Rasmussen and Franklin & Marshall have been polling this race reliably for the last few months. (And recall that Franklin & Marshall offers both LV and RV models, where Toomey wins the LV matchup and Specter the RV matchup, meaning that, well, Rasmussen and Quinnipiac -- which is polling registered voters right now -- may both be right.)
The improvement in the matchup between Arlen Specter and Pat Toomey is all the more interesting because Specter and Joe Sestak have been going hammer and tongs at each other while Toomey is out of the fray. In fact, Toomey's favorables indicate how unbesmirched he really is: he's at 26/8, with a whopping 65% "haven't heard enough" (note to Specter or DSCC or somebody: start defining Toomey as Club for Growth wingnut ASAP). Sestak is in a similar boat, with favorables of 18/6, while pretty much everyone has already decided about Specter, who has an approval of 48/45, which has changed very little since December's 47/45. I hate to start prematurely acting like the Democrats have weathered the worst of the storm and the momentum going into November has changed direction, but with nothing else having significantly changed about this race in the last few months, it seems like changes in the national environment have to be at least partially responsible here for the changed toplines. (Discussion underway in conspiracy's diary too.)
http://www.swingstateproject.com/diary/6491/pasen-quinnipiac-gives-specter-7point-lead