I was mostly concerned here to challenge several of Cost’s assumptions and question whether ideology, as he was calling it, was really so irrelevant to the GOP primary electorate in recent years. I grant that the inclinations of the GOP primary electorate in Pennsylvania are not very significant in the larger political picture at the moment, but analysis that purports to explain those inclinations still needs to make sense.

Your analysis focuses on western PA (and the “T”) and its effects on the Republican primary, although I think this represents rearranging the Titanic’s deck chairs. The western and central parts of that state simply lacks the population, and therefore votes, to meaningfully impact state-wide elections.

You simply have to look at a population-by-county map of PA to see this. Philadelphia County and the ring counties (Montgomery, Bucks, Chester, Delaware) alone, which are packed in the state’s extreme southeast, possess roughly 4 million residents, a third of the state’s population. Throw in the next ring of counties (Lancaster, Berks, Lehigh), all possessing mid-sized urban centers (Lancaster, Reading Allentown respectively), and they together contain about one and a quarter million residents.

This second ring of counties are more rural and conservative, but trust me as a former longtime resident of the region they’re not socially conservative, Appalachia flavor or otherwise, like, say, denizens of Mifflin County PA. For example Berks and Lehigh went for Obama overall, as did Philadelphia and its ring counties. All together the Philadelphia market and the outer edge of its metro area together account for over 5 million residents, about 40% of the state’s population.

The Allegheny County residents are another one and a quarter million residents, so this county alone and the ones mentioned in the east, nine counties out of over 60, account for half the states’ available votes.

Allegheny County overall tends to vote Democrat in state wide races and has for a long time. The “T” tends to vote Republican and has for a long time. Philadelphia tends to vote Democrat and has for a long time.

This leaves the Philadelphia ring counties, alone possessing almost 25% of the state’s electorate, as *the* battleground in the state.The residents of these counties used to vote Republican a lot, thus empowering someone like Specter. But the GOP has become more southern/religiously dominated, and this suburban population, moderate, northeastern, economically-oriented conservatives, began to turn away. Combine this with the fact that the southeastern part of the state is growing even as the “T” and the west shrink in population, and its electoral clout only grows.

This is the significant change in Pennsylvania politics of the last 20 years. Everything else, population changes non-withstanding, has stayed the same, but Philly ring counties went blue. That is why Specter has to change parties to stay alive, and why parsing this conservative’s chances or that in Butler county (pop. ~182,000) is meaningless. There simply aren’t enough votes from those populations to make a difference for one, and two, they’re going to vote Republican in all likelihood anyway, to zero effect. (of course if the vote Democrat it’s even worse for the GOP and we go even beyond spankings like Santorum’s loss to Casey – a.k.a. Potted Plant – by 17 points to something almost cathartic in its devastation).

It’s how Santorum and McCain got clocked, how Lynn Swann got clocked (sacrificial lamb status or no). How Mike Fisher got clocked. All in statewide races. Everyone does the same thing – Allegheny for Dems, Philly for Dems, perhaps this county or that (like Centre or Dauphin) for Dems, *all the rest* for Republicans. Except for when the Philadelphia rings counties vote Democrat, and then the Republican candidate has the floor mopped with himself.

It will happen again and again, provided the Democrats can hold onto majority votes in Philadelphia ring counties. At this point, discussing appeal anywhere else is somewhat pointless.