Recent polling consistently suggests that McCain is only around 7 points down in Michigan, but is trailing by some 15 points in Pennsylvania. For those who went through a modern western education system, that makes Michigan 8 points more pro-McCain than Pennsylvania is. Surprising, then, that the McCain camp is pouring resources into Pennsylvania, soon after announcing the abandonment of Michigan as a lost cause. What's the reasoning, then?
"We pay attention to our own polls," senior McCain aide Mark Salter is quoted as saying. "Pennsylvania looked much better than Michigan."
In other words, the McCain camp is referring to polling data that reads Michigan as being quite a bit more McCain-friendly than mainstream polling does.
Now, this is kind of interesting. There are really three explanations I can see, and it's hard to pick between them. Firstly, it's possible that Salter is just referring to imaginary polling data (some people call this lying), because the decision to pull out of Michigan and stay in Pennsylvania was not made on any sensible basis at all.
Secondly, the McCain camp could be using polling methodology that is significantly different to the mainstream polling. They could be asking different questions, or they could be putting their questions to a different, but still somehow representative, population sample. But for this to explain their decision, then not only would their methodology have to skew the results quite drastically, it would have to do so in Michigan but not in Pennsylvania.
The third possibility is that, in referring to "our polls", Salter really meant that they apply a layer of interpretation to polling data based on other considerations - so by "our polls" he kinda means "our take on the polls", which I guess is a reasonable use of language. This would mean that, when they look at the 7-point margin in Michigan, and the 15-point margin in Pennsylvania, they say to themselves, "ah, but there is an extraneous factor that makes Pennsylvania a much better bet for us; a factor that just doesn't show up in the polls."
Now I have to be honest - I think the first explanation is the most plausible. I think that maybe they recognised pulling out of MI (or at least their announcement of it) as a mistake, and decided not to repeat that mistake in PA. Maybe they are making rash decisions, because they are addled, temperamental, confused, or senile.
But let's say their "own polls" are not imaginary. That means they believe they've identified something starkly different about Pennsylvania compared to Michigan, either through actual private polling, or through some interpretive lens they are applying to the polls the rest of us are seeing. Something that makes the polling results misleading, so that voters will go for McCain on election day but not say so to pollsters, and that applies to Pennsylvania but not to Michigan. The only such thing I can think of is the so-called Bradley Effect.
To be clear, I am not saying that I think Pennsylvania is a more racist place than Michigan. On that question, I have absolutely no idea. But I wonder if maybe McCain's strategists believe exactly this.
Is there some other explanation that has eluded me?


Salon.com
Comments
Maybe they, like Bush, make their own reality. It has to be better than the one they are now finding themselves in.
great post, thanks for raising the question.
Pennsylvania has a larger native GOP machine, including the very popular Tom Ridge, that can be mobilized to help McCain. Because PA isn't dominated by a single blue collar industry, like Michigan, it is considerably less hard hit than Michigan. In other words, it's not as much of a stretch to make PA competitive at this stage than Michigan, budgetwise and logistically.
They don't want to look like they are pulling out of PA. And they want Obama to continue to spend money and time in PA, diverting his attention from other places where McCain can definitely compete.
They could also be banking on a 5 point Bradley effect, effectively driving PA'a numbers closer. They may also be discounting first time voters, meaning that their "likely voter" criteria will screen out Obama voters, and make the race tighter. So it is quite possible that their internals are closer than those done by the media pollsters.
Q: What do you mean, "your own polls"?
A: Well, we firmly believe that Pennsylvanians can't stomach voting for a black man, but are afraid to admit that in polls.
My feeling is that people don't like it when assumptions like that are made about them. And remember if they are counting on that 5-point Bradley Effect, then for some reason they believe it will hit PA but not MI, which is a pretty strong judgment to hit a whole state with.
I do still think their "own polls" probably don't say what they're claiming. But I would like to hear them try to explain it anyway.
Or perhaps they were there in order to keep their numbers up in the face of increased McCain activity/resources. In which case, the McCain campaign has figured out just what it's worth in time and money in order to keep the Clintons and Biden from spending their time in states where the race is closer.
Or maybe the Obama campaign invested the Clinton/Biden time in PA just so that McCain would have to put resources there that his campaign would rather spend in a state with a tighter race. In this scenario, McCain's campaign is forced to spend resources in a state that they expect to lose in order to avoid sending another message to their base that they're giving up on yet another key state.
I lean toward the second or third hypothesis. But all I know for sure is that neither candidate is doing much in Oregon where I live.
I think #1 is closest.
I believe that at the time they made the MI decision the two were about even, and presumably they decided they only had enough money for one. The auto industry was looking so bleak that MI seemed harder to salvage, so they put their money on PA.
Since then, Obama shot up in the polls much faster in PA than MI, making it look, for the moment, at least, like a bad bet. I think they're trying to paper over that.
I bet they wish they had kept their mouths shut. If they had quietly pulled out of MI, they could now change their minds, and say they were just taking a break or something. Now, they would look particularly foolish, and erratic if they dropped out of PA and plunged back into MI. (And would the deserted MI voters forgive them?) They made their bet, and announced it; now they're stuck with it.
I believe the polls and I believe McCain is all but dead in PA right now. But I can see where he believes the abortion issue is a "come home" issue for huge numbers of people in PA. And if this meltdown continues to correct, as it did today, he might not be entirely wrong.
In MI, he's now ahead by 16 points: 54 percent to 38.
That's just done.
It's bleak for McCain in PA, but over in MI. (I normally never say never, but since he's down 16 and has given up, that's pretty over.)