Let’s be honest. Anyone who thinks that only 27% of registered voters in Virginia consider themselves to be Republicans is either a.) stupid; b.) high; c.) incredibly biased or d.) all of the above. Well that is exactly what the number the Washington Post used in their latest poll. To put that in context, the last time Republicans were less than 30% of the general election turnout in Virginia was when LBJ was elected in 1964. Let’s look at the last 3 statewide elections held in Virginia:
2006 – 39% Republican, 36% Democrats, and 26% Independents
2008 – 33% Republicans, 39% Democrats, and 27% Independents
2009 – 37% Republicans, 33% Democrats, and 30% Independents
That gives us an average of:
36.3% Republicans, 36.0% Democrats, and 27.6% Independents
Even during the Democratic wave of 2008, GOP turnout in Illinois was 28% and in California it was 30%. Again, anyone who thinks only 27% of voters in Virginia consider themselves to be Republicans is either a.) stupid; b.) high; c.) incredibly biased or d.) all of the above. Furthermore, even within their Independent sample their appears to be some skew as well – 46% “moderate” among whom Obama leads Romney by 23%, 58% to 35%? I don’t think so…
So, what impact did the WashPo have by undersampling Republican voters by 9% and oversampling Independent liberals and labeling them as “moderates”?
The best they could do was move Tim Kaine into a tie with George Allen at 46% each.
They did manage to get Obama to 51% with Romney at 44%, which was probably their overall goal for this poll.
In order to get Timmy! over 50%, the GOP sample would have probably plunged into the teens and even the WashPo figured they couldn’t get away with that. (Which probably partially accounts for their flawed Independent moderate sample as well.) The fact that he is trailing Obama by 5% in his own state in a poll this skewed to favor them is a sign of incredible weakness. No wonder the DSCC has already reserved $7.4 million worth of air time for the month before Election Day in Virginia to run ads on behalf of Kaine.
Let’s adjust the Post’s findings using the above averages for party turnout. (These will still skew low because of the Post’s flawed sample within Independents who consider themselves “moderates” who responded to questions as liberals would.)
McDonnell’s approval rating – 59%.
George Allen is at 46.3% while Kaine is at 45.7% (still close, but again look to the so-called “moderates” in their Independent sample.)
Mitt Romney clocks in with 45.6% while Obama drops to 50.7%.
Unfortunately, due to the Post’s incredibly sloppy methodology in this poll, it is impossible to further adjust these numbers to more accurately reflect the Independent vote so as to remove liberals masquerading as moderates.
The take away from this poll? The WashPo is still biased and sloppy. You’re better off utilizing one of the more accurate polls such as Survey USA, VCU, Roanoke College or PPP (when they’re being honest and not trying to pump up numbers for their Dem clients.)