Three months out from this year's election that will decide the direction of this country for the next four years, this writer decided to crunch some numbers. As has been explained in past articles of this nature, I use a rather complicated system of analysis that tracks not only trends in the polls this year, but over a certain time period (20 years, or 5 Presidential election cycles) and then weight polls based upon their historical accuracy in predicting not only overall outcomes, but the more important state-by-state outcomes. Some states have a large pool of polls and these are generally the battleground states.
Since the last time I wrote an article of this nature as concerns the Obama-Romney match up, the result was an electoral vote of 290 to 248 in favor of Obama. Additionally, this system allows one to look at the target states that must be won by Romney for him to prevail and what the trend is in that particular state. Based on that trend, assuming the same rate of improvement or demise continues, one can project out to November where the race will stand then with respect to particular states. In this system, although Obama has a 290-248 lead in the electoral vote, the projected figures indicates that he would win Ohio thus dropping Obama's lead to 272-266 making the win razor thin. This way, the race would come down to a Romney victory in either New Hampshire or, if failing there, a western swing state like Colorado or Nevada (although I, unlike some pundits, put Iowa in this category also). Therefore, even though the raw numbers indicate an Obama victory, it is tenuous and the electoral scene is hopeful for Romney.
Speaking of swing states, I define these as states where the difference of the average of all the polls since the beginning of the year for each candidate are plus/minus 2.5 points. Using this criteria, we have eight states: Nevada, Colorado, Missouri, Ohio, New Hampshire, Florida, Virginia and North Carolina. In 2008, Obama won seven of those states (Missouri being the exception). This year thus far, Obama holds tenuous leads in four of the eight- Nevada, Colorado, Ohio and New Hampshire. Please note that this based on the average of polls, NOT the most recent poll. There is also a set of secondary "swing states-" those on the margins of the plus/minus 2.5% or with the real potential to reach that margin by the time November rolls around. These could be considered targets of opportunity for Romney and Obama won all four of these states in 2008- Iowa, Wisconsin, Michigan, and Pennsylvania. However, I personally am not so hopeful as many regarding Pennsylvania. Assuming Romney eventually takes Ohio, then the most likely western win (assuming Obama takes New Hampshire) would be Colorado or, if failing there, Nevada. Incidentally, this model predicts a Romney sweep of the south with the vote not being so close in North Carolina as Obama would like, a fairly good Romney victory in Florida and a razor thin Romney victory in Virginia.
Now, there is a somewhat competing model as many have noted in comments to previous articles of this nature. Specifically, many point out the Dick Morris observation that the undecided voters generally break towards the challenger over the incumbent in the election booth. So, using the average of these polls and assuming voter turnout roughly the same as 2008, this system was put to the test under three scenarios- when 75% of the undecideds vote for Romney, when 67% vote for Romney and a final analysis (explained below) percentage votes for Romney.
When 75% of undecided voters swing towards Romney come election day, Romney wins the election with 311 electoral votes to Obama's 227. This model would deliver Nevada, Colorado, Iowa, Ohio, Pennsylvania, and New Hampshire to Romney. I did not include the southern states or Missouri since my "trend" model also predicts Romney victories there.
But, a totally different electoral scenario develops when the number is dropped to 67% of undecided voters pulling the lever for Romney. In that case, the outcome is Obama with 322 electoral votes to Romney's 216. In this scenario, Obama would not only keep Nevada, Colorado, Ohio and Pennsylvania, but he would also win Florida and Virginia. Romney would keep both Iowa and New Hampshire. Neither of those states are particularly rich in electoral votes, but Obama's states in this scenario are the jewels of the battleground states and an exclamation point on his reelection.
The final scenario I looked at was what if the undecided voters broke for the challenger against an incumbent based on the most recent example in 2004 when John Kerry took on George W. Bush. According to CNN exit poll data, undecided voters who voted for Kerry, the challenger, did so at a rate of about 54%. That is considerably below the 75% postulated by Morris. In that case, the electoral outcome would come out as my original "trend analysis" predicts- Obama winning 290-248.
Obviously, this is based on polling data three months out from the general election and yes, that is the only "poll" that counts. It may very well turn out, although not likely, that all undecided voters will vote for Romney. In that case, then it would spell an unmitigated electoral disaster and repudiation of Barack Obama. Well, we can dream. Of course, we have not had any conventions yet nor have we had any debates. We do not even know who Mitt Romney's running mate is yet. So, a lot can change between now and any point in the future. And speaking of Romney's running mate, it cannot be overemphasized that his selection will not necessarily be the deciding factor with respect to any battleground state. Numerous studies have proven that the selection of a Vice Presidential running mate has no effect on swinging that selection's home state into their camp. Hence, choosing a Rob Portman will likely have no effect on delivering Ohio to Romney, nor will McDonnell assure a victory in Virginia or Rubio a victory in Florida. That is why I am steadfast in my belief that Romney will look beyond that consideration and select Senator John Thune of South Dakota.