Not to worry. The polls will guide us to light and truth. After all the polls are accurate reflections of the will of the people. If the election were held today, which presidential candidate would you vote for? A simple (and certainly fair) premise. Absolutely no skewing. No adjustments for past elections, ethnic background, age, or past voting pattern are necessary.
If this were true, the unemployment numbers might spike again, this time due to the flood of suddenly un- (or at least under-) employed prognosticators, talking heads, bloggers, reporters and their like hitting the streets.
Plus Americans would be deprived of one of their favorite reality shows carried on all networks, major, cable, and otherwise: The 2012 U.S. presidential election.
One compelling story line for the 2012 elections is the focus on an ever expanding number of competing polls. The resulting cacophony makes it nearly impossible for the average person to understand what is actually happening.
Gallup, 538, Rasmussen, Quinnipiac, Reuters/IPSOS, and so on now focus on ever smaller demographic clusters, so much so that there may be no actual living person that fits into them. And so we now have the situation where people are bombarded with probabilistic data about segments of the population so specific that they can seem irrelevant and certainly removed from the real world.
People then become so inured to the campaign that they begin to perceive it as background noise and go about their private lives and business.
The result is a voting public who listens to this information and is more confused than informed.
So where do you go from here? Social Media.
Social media is now being recognized as the voice of the people. While polls are inherently unreliable because they are tuned and skewed and ‘balanced’, social media tracks voice of ‘the people speaking’ rather than taking pseudo-random measurements with any number of data points that are difficult to put into any type of pattern – without the attendant tuning or tweaking.
With this in mind we thought to turn the Internet and social media analysis tool, NarrativeTracker to the 2012 Presidential election.
Though we had used it in 2010 to measure the overall outcome of the Mid-term Election which it did, we had never applied to a presidential race.
We find the results compelling: according to NarrativeTracker, Mitt Romney will beat Barack Obama next Tuesday by 52.1% to 47.9%, a margin of 4.2% in the popular vote.
This is a prediction, neither a poll (like Gallup) nor a statistical analysis (like that of Nate Silver). There is no 'margin of error'; our analysis includes data from the preceding ninety days incorporating long-, medium-, and short-term movements that polls cannot capture (except in aggregate). The predictive element only adds to NarrativeTracker’s power.
Interestingly enough, and in contradiction to the traditional polls, Romney was leading Obama at the beginning of the measurement period with the President building positive momentum in the final days (increasing by about 8%). The data suggests however that Obama will not close the gap by November 6th.
The end-result of the analysis is a NarrativeTracker Index number, in this case translated into a percentage. The analysis used August 31, 2012 as the initial time stamp and concluded earlier today. No consideration of the Electoral College result was taken in the analysis; it is strictly related to popular sentiment.
The NarrativeTracker Index is based on the national discourse, providing a real-time, accurate picture of what the public is saying about any topic, at any point in time. NarrativeTracker analyzes the Internet, some 2,000,000 of the most prominent blogs, 275,000 of the top print and electronic media sites, and new social media as they emerge.
As opposed to polls, which only test choices under a pre-determined set of scenarios, the narrative of the people is the true vox populi and it continually manifests itself through social media.
Payack is president of The Global Language Monitor and Peters is a high tech executive and author.