In the aftermath of the 2012 Presidential Election, I am continuing my look at verifying the predictions (also known in this blog’s vernacular as veridiction) made by people in the weeks and months before the votes were actually tallied.
My previous post on this subject regarding Newt Gingrich saw a first with one quote containing three predictions, all of which were wrong.
This post also sees a first as I believe I have never had two individual people in the header of one of my posts about veridiction.
Today’s subjects are Kenneth Bickers and Michael Berry, two political science professors from the state of Colorado who have a model to help them predict the winners of presidential elections.
Here’s a link to a story written about them so you can read about this academic duo and their model in greater detail.
The article was published on October 4, a month before the election. The gist of the pair’s model is that it uses economic data, not polling numbers, to predict who will be the occupant of the White House next January.
Based on their model, Bickers and Berry said that President Obama would win only 208 votes in the Electoral College and Governor Romney would earn 330.
Now that the final final results are in and with the news that a winner has been declared in the state of Florida (Barack Obama), it can be shown that Bickers and Berry’s model came extremely close to predicting the actual number of Electoral College votes allocated, which turned out to be 332 to 206.
Unfortunately for the model from the pair from Colorado, it was President Obama who had the higher number.
The article I linked to had this to say about the model…
The state-by-state economic data used in their model have been available since 1980. When these data were applied retroactively to each election year, the model correctly classifies all presidential election winners, including the two years when independent candidates ran strongly: 1980 and 1992.
To this, I can only add the thought that a model is only as good as its last prediction.