In February, when the Signal unveiled its model for predicting presidential elections, we wrote that, "while campaigns and candidates matter, they don't matter all that much. Despite the varying quality and positions of the campaigns and candidates over the last 10 presidential elections, variables beyond their immediate control describe the outcome very well."
At the time, our model predicted that Barack Obama will win reelection with 303 electoral votes. In the past five months, as the campaigns have spent tens of millions of dollars on advertisements and fought tooth and nail for the smallest advantages, the model has remained steady. Only Virginia has switched columns at any point, wavering between camps as our predictions show and incredibly tight match there. At the moment, we have Obama eking out a victory in Virginia, keeping him at 303 electoral votes. Should he lose the state, he would retain the presidency with 290 votes to Mitt Romney's 248.
We could be dead wrong, of course, and Romney could walk off with a landslide victory. The main reason we have confidence in our prediction is that it now combines immediate factors like polls and prediction markets with big-picture factors like economic growth and a state's traditional ideology. In the past, the data has shown that these grand crosscurrents have more sway over elections than flavor-of-the-week scandals. Then again, it's easier to predict the past than the future.
You can explore the prediction in the following map. If a state is at least 75 percent certain to go to a candidate, it is colored dark blue or dark red. States more in the balance have a lighter color.
One does not need a model to see that this is an extremely close election. While Obama currently has the advantage, he is defending seven states where Romney has a very decent chance of winning: Virginia, Ohio, Colorado, Iowa, New Hampshire, Nevada, and Wisconsin. Virginia, Ohio, and any one of those other five states is enough to secure Romney victory.