... which, of course, we already knew. But political junkies like to look at poll internals because otherwise, you know, we might have to concentrate on our work on a rainy Monday afternoon, so I did just that with the new Muhlenberg/Morning Call PA-Senate poll (full pdf here).
First of all, the numbers show Toomey 46, Sestak 39, which is pretty much precisely in line with what a slew of other polls have been showing over the past couple of weeks. That is to say, Sestak is within ten points, but is consistently down by between six and nine, such that Nate Silver and other prognosticators give him very little chance of climbing the hill.
Looking at the internals confirms what we already know all too well: it isn't that Club For Growth Toomey is more popular in swing state Pennsylvania, but that pollsters believe that a lot of Democrats won't be voting on November 2nd.
The first indication of this in the crosstabs (again, pdf) is the breakdown by Party ID:
Democrats: Sestak 68, Toomey 19
Republicans: Sestak 14, Toomey 71
Independents: Sestak 41, Toomey 41.
Now these numbers on their own would look quite decent for Joe Sestak. He is losing only a little more of his partisan vote than Toomey is, and is exactly even with independents in a very difficult year electorally. And if you are like me (a native Pennsylvanian) you know that Dems have a massive Party ID advantage, one that received a very significant bump in 2008 registrations.
And those numbers haven't changed. According to quite current official stats (excel file here), Democrats make up 51% of the total PA registered voter rolls, Republicans are at 37%, and independents/other party voters are at roughly 12%. But the Muhlenberg crosstabs predict a radically different likely voter makeup: only 44% of their LVs are Dem, 47% are Rep, and 9% independent!
Muhlenberg doesn't give a RV figure (this close to the election fewer pollers will), but a rough calculation using the registration rolls and Muhlenberg's own by party breakdown is that if, very hypothetically, voters came out in numbers equal to their percentages on the voter rolls, Sestak would be leading 44 to 41%.
A similar indicator is that Muhlenberg's sample consisted 46% of 2008 Obama voters, and 47% of 2008 McCain voters. But as we all know, Obama took PA rather handily: 54.7% to 44.3% (despite Grumpy Grampa's desperate last efforts and a thousand SCLM stories about racist Pennsylvanians and the Bradley effect). So again, Obama voters are by and large not turning against us, they are, Muhlenberg and other pollsters believe, not going to vote in nearly the numbers that Toomey's Teabagger friends will.
So -- let's GOTFV.