Electoral Projections: Still No Changes, but Looking Up
Generally speaking, every time I write one of these entries, I get comments to the effect that polls at this time mean little, or that this state or that state is in play, or complaints about the accuracy of polls. Let me state right up front, as I do every time, that with a little over 4 months to the election, anything can change. We have not even had the conventions and the usual bump in the polls each candidate receives coming out of them. Romney has not chosen a running mate, although I do NOT subscribe to the theory that VP selections decide elections. There have been no debates and, although some believe so, the campaign has not gottten down and dirty yet. So YES- this writer knows that this far out, polls sometimes need to be taken with a grain of salt. But, it does not make them any less important in determining where the state of the race stands at this point in time.
In a previous entry, I explained the methodology looking at long term and short term presidential electoral trends. I also look at polls as concerns this election and not just the most recent poll. A lot of comments have called me “crazy” because “the most recent poll out of state X shows Romney ahead by X points.” If we went that route, incidentally- the “most recent poll method”- Obama would win in a landslide. Instead, the trends in the polls in 2012 along with the relative accuracy of the poll- based on past performance- is taken into account. With the weakest polls, the margin of error and the fact that “undecideds” break for the challenger are taken into account. Given the historical accuracy of the poll involved, the margin of error, the “undecided break,” both or neither are taken into account to modify the poll results. By “undecided break,” I mean the fact that undecided voters do tend to vote for the challenger on Election Day.
One thing this method does not really care about is the internal poll numbers. Where there is detailed information with respect to particular polls, the demographic and ideological data somewhat approximates actual voter data from exit polls in the last two Presidential elections. This analysis does not get into how either candidate is doing with Hispanics or women or independents or any other group in any state.
One thing this method can discern is the effect of certain actions before and after the fact. For example, since a baseline can be established, one can look at how Obama’s great gay marriage epiphany played in key states, or his immigration non-enforcement announcement, or the Supreme Court Obamacare decision, or anything else that comes along in the future. It also discerns trends in states that modify, along with the longer term trends, the point differences between the candidates on a month-to-month basis. For example, Obama may be showing a 5 point average lead over Romney in Ohio. However, in one the previous month, Obama may have averaged a 6 point lead in polls in that state. In the current month, that average may be three points. Although Obama may lead overall, it is apparent that that support is waning in that state in favor of Romney. That trend modifies the overall score and allows one to project out a Romney victory in Ohio on Election Day.
So without further ado, going east to west, the results by region are as follows:
In New England, Obama will sweep all the electoral votes here. Many have speculated on the possibility of Romney taking New Hampshire. Although I would still categorize it in the “swing” category, Obama looks to, at this time, take the state. Moving a little south to the Middle Atlantic region, again its an Obama sweep EXCEPT for West Virginia, although some would argue with that state’s inclusion in this region. And, incidentally, many have speculated on Pennsylvania going for Romney. Every year I hear the same thing and every year it is a blue state at the national or statewide level. Perhaps, if the Philadelphia and Pittsburgh areas were shipped to a neighboring state, then Republicans would have a better chance.
Moving over to the Rust Belt, I believe it a foregone conclusion that Romney will flip Indiana. And although polls out of Ohio show Obama ahead, Romney is gaining momentum there. Assuming he plays the campaign correctly, I think he will take this state also. However, not so much Michigan. A real interesting state is Wisconsin which shows Obama ahead, but Romney has made some gains there of late. Although I would not elevate it to swing status just yet, if I was Romney, I would concentrate my resources here rather than Michigan.
Now things get a little more interesting as we head into the electoral vote-rich South. There are also three swing states in this region- Virginia, North Carolina, and Florida. Romney needs all three to win the Presidency assuming there are no major upsets elsewhere. North Carolina appears to be a lock for Romney as Obama’s gay marriage announcement has affected his standing there and it is understandable. Doing so the day after North Carolina voters decided against gay marriage by a 3:2 margin is a slap in their collective faces. Virginia will be tough and Florida tougher still. However, the good news at this point is that I have them in the Romney camp by very slim margins. Hence, it looks like a Romney sweep in the South. One thing I would look for here, in terms of Democratic/Liberal dirty tricks, would be a whisper campaign against Romney’s Mormon faith.
In the Midwest, again Romney seems to have the upperhand except most in Minnesota. The state to watch here is Iowa which I now consider a swing state (it wasn’t two months ago). If Romney could take Iowa, then it could be toast for Obama. If not, then one must go west to look for victory. In the Mountain states, Romney should prevail in all states except Colorado where Obama has maintained a fairly steady lead. In fact, if anything, Obama has gained a little in Colorado in the past two months.
Finally, there is the West. Its a mixed bag with Obama taking the majority of them except Arizona and Alaska. The good news is that Obama believed he had a chance in Arizona (can you say, “Pander to Hispanics?”), but the state is actually trending away from his grasp. The key state here is Nevada which is in that swing category, but I am giving it to Obama at this point.
That leaves the electoral count 272-266 for Obama. If Romney loses any of North Carolina, Virginia, Florida and Ohio early in the night, assuming everything else holds true, then the election is over fairly early in the night. If, however, Romney prevails as predicted in these four states, we have to wait for either Iowa or Wisconsin, more likely Iowa. If the early states go for Romney along with Iowa, its over and Romney is the President. If Iowa goes for Obama, then it is either Colorado or Nevada that would give it to Romney with the best bet being Nevada. To extend Election Day later into the night, it is imperative that Romney take Florida, North Carolina, Virginia, and Ohio. To reiterate, the task is difficult, but not impossible.