Beginning to Measure The Bounce In State Polls

National polling is not where the race is won and lost, as we learned eight years ago.  The state polling is what matters, and we're starting to see a trickle of poll data where we can measure the bounce there.

SurveyUSA is out with its first post-convention polling in Virginia, and it shows little-to-no movement toward McCain.  This is out of line with other national polling, which as I've noted, shows a big move.  The Rasmussen reports poll shows a slight bounce toward McCain, but not as big as his nationwide bounce. 

The internals of the SurveyUSA poll are particularly interesting -- there is movement toward Obama in the DC suburbs, while there is movement toward McCain in the Hampton Roads area.  This is completely consistent with their candidacies and with the national polling -- Obama moves up with upper middle class, wealthy, white suburbanites, while McCain moves up with a more traditionally middle class contingent (this area of the state also has a huge African American population in the cities and also rural counties, hence Obama's relatively strong performance here).

The other interesting thing about the SUSA polling is the type of person who is undecided.  Looking it over, the undecideds are older (0% undecided from 18-34!!), politically independent, moderate (0% liberal undecided!!), and not located in NoVa.  Now there are error margins to the subsamples that render a lot of this irrelevant, but the bottom line is that Obama appears to have come pretty close to maxing out his strongest groups, and is left with needing to close the gap with demographics that he doesn't perform especially well with.

In Colorado, Obama seems to have received a nice bounce, and leads in the latest Rasmussen poll.  Given the results in Virginia and the location of the Democratic convention, that does not surprise me.

Similarly, in Florida, Obama has actually made some gains, and is now tied with McCain.

The big movements for McCain have come in Ohio, where he went from leading 45-41 to now leading 51-44.  Similarly, in Pennsylvania, McCain has gone from trailing 45-40 to trailing by only 47-45.

I think what we're going to see in the next couple of days is that in the states with large industrial, blue collar populations -- states where Obama did not fare well in the primary -- and in states that are traditionally conservative but Obama had overperformed, like the Dakotas and Montana, that we will see a fairly substantial movement toward McCain.  The McCain (and the Obama speech) speech will bring the former, while the Palin speech (and the Obama speech) will bring the latter.  This may also be the case in liberal states with a large blue collar base like Massachusetts.

But in states with large, near-dominant suburban and exurban populations like Virginia, Colorado, and Florida, McCain's bounce may be muted, or even non-existant.  The question is whether McCain can go over the hump in some of the blue collar states, and/or whether he can sufficiently hold onto the traditionally red exurban states (and vice versa).