yet another election day
John Micek passes on the final Q.... (I can't spell common nouns, I sure am not going to try to spell the name of that school up in Connecticut) poll and it has Sestak up by a meaningless 1 % point and Dan way ahead of the pack. So the Sestak/Specter race is about as even as it gets, in line the last Muhlenberg poll which showed a literal tie which is pretty amazing in itself. Why they have the election I suppose.
For Gov on the D side the persistent curiosity is the amazingly high undecided going into the final day. According to Q 30% are still undecided despite a bunch of choices... maybe its so many choices that is keeping folks from commiting to one name or the other? I'd love to see the cross-tab between the undecided voters for governor and senate polling with its much lower undecided count. Who are the D's undecided for Gov, but decided for senate? There must be a fair number of folks like that. It's curious.
On the D senate race... if the polls really are showing a tie I suspect it bodes well for Sestak. I get a strong sense that the clear support of the greater D establishment could be impacting the pre-election polling. Sort of like bias that is debated over the role of race in polling. Some races see polls showing greater support for minority candidates than actually show up in the final election returns. Similarly I suspect there are some number of folks who feel pressured to support, or at least show support, for Sen. Specter, but who will not vote for him behind the curtain. I'm not saying it's going to be a big effect, but if the race is showing a tie and if there was a potential stealth Sestak support of just a few percent then this race swings. and I just can't think of a counterveiling reason why folks might be saying they are voting for Sestak, but really vote for Specter in the end. I suspect there could be more Specter support than is showing up in the polls just from sheer name recognition built up over decades. How it all balances out I have no idea. But with the last polling near even, it makes for a good natural experiment to study polling bias after the fact like the uber-statisticians are wont to do. So at least there will be something to blog about in perpetuity.
* I honestly have no idea who Justin Bieber is. I mean... he was on SNL earlier in the year so I can place a face with the name, but I have no idea if he sings, dances, tells jokes or solves differential equations for a living. I am deliberately not asking Mr. Wikipedia just to see how long I can go without knowing.