Pennsylvania in T-346 Hours

By Fester

Two weeks until I vote in a presidential primary that matters.  Actually I'll vote early in the morning as my polling place is the senior citizen center down the street  where they sell really good cookies and coffee which are gone by 7:15. I am overly motivated by bakery products. But what is the state of the Pennsylvania race and what can we expect? Right now we are seeing two very different polling pictures.

The first is Pennsylvania is voting its historical demographics as a Democratic but conservative Democratic state and supporting Hillary Clinton by about eight to ten points. She is winning big in the central part of the state, and other areas that are Republican base areas in the general election while keeping it close in Greater Pittsburgh and being blown out in Philadelphia. Obama has been closing, but will come up a couple of points short.  This is the story that I have been expecting.  It is also the story that Kyle at Comments from Left Field is somewhat expecting:

I think without any external influences, Obama is likely to pick up a couple more points on Hillary, but will still likely lose the state by a margin of three to five points.  This is, of course, assuming nothing happens, the aggregated trends are fairly accurate, and both candidates maintain their current trajectory.

Yes, because those three things happen all the time in politics.

The other story is that the race is fundamentally a tie as of sometime in the past week and momentum is moving heavily towards Obama.  The most pronounced poll in this regard was the Public Policy Polling result of a twenty eight point pro-Obama swing from mid-March.  The only significant and readily transparent difference in these two groups is the tie +/- MOE group of polls were conducted slightly later, on average (by about a day or two) than the polls showing Clinton with a significant but shrinking lead.  I don't remember seeing anything on April 1st or 2nd that would have driven that type of movement to Obama. [UPDATE: The Post-Gazette has a new Quinnipiac Poll with a Clinton margin of +6 versus the previous +9 with all net movement coming from undecideds going to Obama and this poll was in the field within the past couple of days, so timing could be the factor here]   

I am guessing that the difference in these two groups of polls are the turnout models.  I think the tighter the likely voter screen, the more likely the poll will show a decent size Clinton lead.  The one exception is Quinnipiac which is screening only for registered voters which is a much larger universe than any likely voter screen.  Interesting. 

Pennsylvania is seeing a massive influx of new Democratic registrations from non-registered and other registered voters (switches from unaffiliated and the GOP dominate this category)  and the question is where do new Democrats in areas such as Montgomery and Bucks County lean?  Right now the evidence suggests that Obama's campaign did a better job of registering their likely first time voters than the Clinton campaign although both campaigns did an excellent job overall.  The looser the likely voter screen with a higher probabiltiy that a newly registered Democrat will vote most likely favors Obama.  And will they vote in numbers proportionate to their strength?

So we are two weeks out, things are closing hard by almost all metrics, the airwaves are getting saturated (I was on the eliptical at the gym a couple of nights ago and on the six screens tuned to four channels, I saw at least eight Obama ads and three Clinton ads in my five miles), and the candidates are touring hard.  I'll still give Clinton a slight lean to win the primary, but I will no longer be shocked if it is too close to call until 4:00AM Wednesday. 

 

   

Comments

[Jason]

I wish I had waited to vote in PA. The Dem Abroad primary was pretty one-sided, and now I feel like should have waited.