... If history is any predictor, we are all likely to in for a few surprises on Election Day.
Elections rarely play out exactly as we think they might. As such, I think there are two interesting possibilities to consider. These are by no means the only two possibilities, but they are two which I believe are, at least, plausible ...
1. An Obama Landslide: An amazing number of new voters are registered to vote this year. This is happening, for example, in states like North Carolina. Are we entering a new paradigm which would finally prove existing "likely voter" models to be entirely obsolete? If so, it is possible Obama could win by a dramatic landslide that exceeds anything we would imagine...
2. "Dewey Beats Truman": Remember the New Hampshire Primary? Obama was winning in the polls by double digits. He had all the energy and all the visible ground troops (I was in the state and can vouch for this). Yet -- defying all the polls, pundits, and predictions -- Hillary Clinton won. What if this scenario plays out in a couple of states such as, say, Pennsylvania and Ohio? It doesn't have to happen nationally -- depending on how things go, flipping one big state could make the difference. It's not hard to imagine that if the so-called "Bradley effect" actually exists, that it would manifest itself in the Rust Belt.