Monday, November 5, 2012
A Recount Election?
Raz and Gallup: 49-48 Romney by 1 point. All but a couple mainstream media polls: Tied or Obama by 1 point by Clifford F. Thies Iowa Electronic Market - Vote Share Market: Obama by about 1/2 point Winner-Take-All Market: Obama a heavy favorite (implying a decided tilt in the Electoral College) There it is. About as close as you could imagine. About the only thing that is certain, at this point, in the minds of Democrats is that they have a decided edge in the Electoral College. They will win in the Electoral College even if they lose the popular vote, because of their "firewall states" of Ohio (most importantly), along with Iowa, Nevada, New Hampshire and Wisconsin. Frankly, I do not see a Democratic tilt in the Electoral College. I see no reason to change my view that the winner of the popular vote will almost certainly be elected President, if the difference is by 1 point or more. In particular, if Romney were to win the national popular vote by 1 point or more, and somehow lose Ohio, he will nevertheless be elected President by reason of one or more of his alternate paths to the White House. Ohio - a toss up according to Raz, a slight Obama advantage according to the mainstream media polls. If Romney wins Ohio (along with Colorado, Florida, Indiana, North Carolina and Virginia, and the McCain states) he wins. Obama just absolutely has to carry Ohio and this is a toss up state or very close to it. Wisconsin + one other state - another "out" for Romney. Let's say the Democratic GOTV in Ohio makes the difference. It is possible Romney could win by taking Wisconsin and, say, New Hampshire. Sorry Wisconsin + Maine's 2nd C.D. won't do. This road to the White House is not very promising if Ohio, because of the tendency of the close ones to fall together for one or the other candidate. But, if Ohio goes Obama only by reason of the Democratic GOTV in that state, if Romney loses Ohio even while winning the national popular vote, it is not improbable that he will carry Wisconsin and one other state. Iowa + Nevada + Maine's 2nd C.D. - a third "out" for Romney. According to the mainstream media polls, we're a couple points behind in both Iowa and Nevada, and who really knows what will happen in Maine's 2nd C.D. But, the early votes and absentee ballots give us reason to be hopeful in these two states. As for Maine's 2nd C.D., if we pull Iowa and Nevada but not Maine's 2nd C.D., it's 269 to 269 and the election of the President will be thrown in the U.S. House of Representatives, where we have a decided advantage. Minnesota + one other state or Pennsylvania - a fourth and a fifth "out" for Romney. Only one of the recent polls in each of these states puts us ahead or in a tie. Some of the polls are badly skewed and can be disregarded. But others, when unskewed, still have Obama ahead. So, while I'm not ecstatic about these paths to the White House, if each of the first three "outs" fails, and we win by reason of Minnesota + one other state or by reason of Pennsylvania, we'll take it. For me, the real reason for Romney to add Minnesota and Pennsylvania to the battleground states is not a desperate move because we know we're going to lose in the firewall states. It's to develop a large enough majority in the Electoral College to obviate any post-election legal shenanigans, and also to build the argument for a mandate for governing. Finally, I need to point out the difference between margin of error, and model uncertainty (or, unknown bias). The polls, especially the mainstream media polls, weight the interviews in their samples to reflect what they believe to be the demographic and even the partisan mix of those who will vote. This is unknowable until the election is actually held. The application of "likely voter" screens helps reduce the problem of model uncertainty and, so, the difference between my average of the unskewed polls and the simple average of the polls has collapsed in the most recent two weeks. But, there is the possibility that the polls are still biased.