Want to play a little pre-election game to pass the time and assuage the anxiety? For curiosity's sake, I went back and read a blog post I filed almost exactly a year ago to see what I was thinking. Lo and behold, I discovered this gem: Jon Huntsman will win this November.
More interesting is that this post was my first serious introduction to Nate Silver, who seems sure to be remembered as one of the more notable figures from the 2012 elections.
Last November, Silver let readers of his New York Times blog play a game that is still fun and relevant today. (In order to make sense of this post, it helps to actually play the game.)
The game used three basic inputs to predict the outcome of the 2012 election: economic growth (measured in gross domestic product), President Obama’s favorability rating and an ideological “rating” for the Republican challenger, a number based on the candidate’s voting record, the ideological composition of his fundraisers and where voters themselves place the candidates on ideological scales in polling.
A year later, we can still play Mr. Silver's game with two of the variables known: GDP grew 2.0 percent in the third quarter of 2012, and Obama's favorability rating is about 50 percent.
We can't tinker with Romney’s ideological rating; Silver gave him a score of 49 on a 100-point scale, where zero is a flaming liberal and 100 is Rush Limbaugh. The more moderate you are perceived, according to election models, the better, so Romney's number is good. (But not as good as Jon Huntsman's score, which is why Silver predicted — correctly in my opinion — that he would have been the harder candidate for Obama to beat.) We might contest that number now, given what happened to Romney in the primaries and in the early part of the general election, when he moved to the right and was held to it by Obama’s attacks. But we have to leave it static.
With a GDP growth rate of 2.0% and an Obama approval rating of 50 percent, Silver’s model predicts that Obama has a 57 percent chance of winning. Obviously, this is below where Silver has it today, with Obama at 86.3 percent. But, as Silver indicated a year ago, the closer we got to the election, the more that polls, and not the fundamentals, would drive the models. Or perhaps said more precisely, the fundamentals would drive the polls, which in turn would inform the models
Anyway, here, finally, is what I find interesting: Obama was able to raise his approval rating by at least 7 percentage points over the course of a year. That fact, more than anything, may give him the election. At a 2 percent GDP growth rate, Silver predicted that any Obama approval rating below 47 percent tipped the odds of victory toward Romney. So close, but perhaps still too far for Mr. Romney.