October 9, 2012
Last weekend I posted my presidential election prediction. Here are the factors that went into my conclusion.
- Polling data
- What people have seen before the debates
- What people saw during the debate
- Voter turnout
- A tipping point.
First off here is my presidential election prediction. I don’t expect it to be close and am going to predict a 320 to 218 Romney victory. I believe there is about a 25% chance of a 350 to 188 margin. I’ve believed since early spring this would be the case. Of course this is based on no major game changers. If we drop bombs on Iran then I suppose it will be time to reassess.
Whoa, you say! How can that be. Well I believe a part of what has kept this “close” is the dishonest polling that happened all the way through September along with months of unchallenged commercials smearing Mitt Romney’s character. Pollsters have been over-sampling Democrats based on the 2008 Election results. I wrote about how polls were being skewed here and here. The 2008 election turnout margins were a special case not likely to be replicated. In fact there is no chance it could happen this time. In addition I believe polls are becoming more and more unreliable because fewer and fewer people respond to them. Pew research recently released data that showed only 9% of those reached participate in the poll. This compared to 36% as recently as 1997. And if you don’t have a land-line you are not going to be included either.
Pollsters and the talking heads make this overly complicated in an effort to look smart and in some cases to to influence the public.
Here is the simple way to look at this. If the Republic and Democrat turnout is equal, as it was in 2004 and 2010, then we would have a 50/50 election. But 28%, give or take a percent or two, of the votes will be cast by independents. These folks are now consistently moving towards Mitt Romney. If Romney wins these independents by 10% that works out to about a 3% swing. A 53% to 47% total overall. There really is a correlation between the overall final vote and which swing states mover to the winner.
A 53% to 47% margin for Romney would mean that Ohio, Florida, Virginia, Iowa, Colorado would go to Romney. But between Wisconsin, Pennsylvania and Michigan, it is very possible that two of those could go to Romney as well.
So what are the indicators that this could happen? There are two factors I looked at. One is the numbers and the other has to do with human nature.
The Numbers
- Democrats have a reduced advantage in voter registration compared to 2008. This is true in swing states.
- Democrat enthusiasm is lower than found among republicans. This means a larger percent of Republicans are likely to vote and a lower number of Democrats are likely to vote. This is a reversal from 2008. Pollsters have generally ignored the significant of this change in their polling and have used 2008 assumptions.
- In the future I believe studies based on GIS data (Geographical Information Systems) will provide much more reliability then the old polling methods. A GIS study predicts that Ohio will go to Romney by a much larger margin then what skewed polls are predicting.
Human Nature
Do you know anyone who voted for McCain that plans to now vote for Obama? Me either. However I am aware of several people who voted for Obama that are of the “never again” persuasion.
The Obama campaign spent millions of dollars in the Spring and Summer running ads in swing states trying to paint a picture of Romney as a bumbling greedy rich guy who causes peoples spouses to die and, gasp, is probably a felon.
For many people in these states the debate last week was the first time they saw Romney as himself, unfiltered by the media. He came across as presidential. One of the keys to closing a sale is for the customer to see themselves as comfortable with their purchase. I believe at the end of the debate millions of people who weren’t sure before, left feeling comfortable with the idea of Romney as president.
Finally, once undecideds start come out in favor of a change it becomes easier and easier for others to join in. This is what is referred to as a tipping-point. While I believe that Romney was already leading going into the debate I suspect the debate created the tipping point that will result in landslide. Much as the 1980 race between Carter and Reagan was considered to close to call only one week before the election. That 1980 election resulted in a 489 to 49 Reagan Landslide.
So there is my presidential election prediction. Have I missed something? What are your thoughts?
