An Election Prediction

Since tomorrow is election day, it is time for my prediction of election results. I strongly believe that Romney will win the presidency despite the never ending hype about the various state polls. There are a few factors which make this inevitable in my opinion.

1) Romney is leading in every nation poll with independents. Indeed, when one averages the latest poll numbers together, the Romney lead is quite large with this group. For years, the conventional wisdom has been that if a candidate wins independents by more than 5%, he or she will win the election. Romney's lead seems to be in excess of 10% among this group.

2) The electorate has shifted in composition from 2008. During the 2008 election, the exit polls found that the Democrats led by 8% when voters were asked to identify their party affiliation. That lead was the result of a huge surge of minority voters and young voters who were voting for the first time due to the Obama mystique and his status as the first black president. The lead was also the result of a great many Republican voters deciding not to turn out. Whether it was fatigue with the Bush administration or the perilous economic times, GOP turnout was way down.

There have been two huge studies done in 2012 to determine the party identification of likely voters tomorrow. Gallup interviewed tens of thousands of likely voters and Rasmussen did the same. All together about 90,000 people were questioned about their party identification. Since the normal poll questions less than 1000 folks, the accuracy of these two studies is much greater. Both Rasmussen and Gallup found that among likely voters, party identification favored the GOP (Gallup by about 1.5% and Rasmussen by close to 6%). That is a swing of just about 10% from 2008. Since all polls show that about 95% or more of self identified Republicans and Democrats vote for their own party's candidate, the shift in the composition of the electorate has enormous consequences. A great many polls adjust their results to meet the 2008 makeup of the electorate. If there is a 10% shift towards Romney (or even half that much) it makes a supposed tie into a big win for the GOP.

3) None of the national polls show Obama with more than 50%. On average, Obama reaches only to the mid 47% range. It is a well documented phenomenon that the few undecided who get to the polls tend to break strongly for the challenger. If 5% undecided break even 3 to 2 for Romney, it gives the GOP another 1% to add to its margin. Indeed, if the break is 4 to 1 which is more likely, it provides another 3% for the Republican margin.

There are a myriad of other factors, but they pale in comparison to the ones above.
Right now, I expect that the following battleground states will go to Romney: North Carolina, Virginia, Florida, Iowa, New Hampshire, Ohio, Colorado, Wisconsin and Pennsylvania. I expect Obama to carry Nevada and Michigan. Among non-battleground states, there is a chance for Romney also in Minnesota although that may be wishful thinking.

Assuming these results, Romney wins the election with 315 electoral votes to 223 for Obama. If Minnesota flips to Romney, that changes to 325 to 213.