I went and voted for Sarah Palin for Vice President, and had to vote for her running mate, too. So, if McCain wins Utah by only one vote and wins in the Electoral College because he won Utah, then my vote made the difference.
Of course, the likeliest scenario is that, just as the polls say, Obama will win the election.
Here’s my prediction:
However, I call your attention to this page, which I archived on Election Day 2004. (I updated it with actual margins and Zogby’s error on some swing states.) So the polls margins can have errors of up to 7% compared to the final result — and that was without the Bradley Effect.
My prediction above is based off my instincts regarding which way the tossup states will fall. Going off the Real Clear Politics averages, there are a couple more scenarios I want to point out.
What if the combination of poll bias and Bradley Effect is 6%? If we give all states where Obama’s margin on RCP is less than that to McCain, here is the result:
Unless there’s a stray Electoral College vote in Maine or Nebraska (where they allocate then by Congressional district), we have a tie. Then the Presidential election goes to the House of Representatives. Since the Democrats will control a majority of the state delegations and each state gets one vote, Obama will win.
What if the combination of poll bias and Bradley Effect is 7%? If we give all states where Obama’s margin on RCP is less than that to McCain, here is the result:
That flips Nevada, and that’s enough to flip the election to McCain.
Obviously, that’s the scenario I’m hoping for, along with less likely ones like McCain wins Pennsylvania.
Please note that I believe the Bradley Effect may be large this year not because of a large amount of racism, but because so many Obama supporters have been saying that opposition to Obama is racist. That makes people who favor McCain for non-racist reasons less likely to disclose their preference.