UPDATE: How did the model perform? By my calculation (based on reports at noon today), McCain captured 47% of the popular vote; the Fair model predicted he would win 48%.
Ray Fair, a Yale economist, provides a simple statistical model to predict the vote share of today's presidential and House elections. Click here to see the last pre-election predictions. My sophisticated experiments (i.e., plugging in different values for 2008 growth rates) indicate that we needed at least a 3.15% growth rate for a McCain victory and there is no feasible rate of growth that would deliver a Republican House majority. All the more reason to stay home.
HT to Mankiw.
|Compute Vote Predictions: Input Values|
|The equation to predict the 2008 presidential election is|
|VOTEP = 46.61 + .680*GROWTH - .657*INFLATION + 1.075*GOODNEWS|
|The equation to predict the 2008 House election is|
|VOTEC = 43.37 + .397*GROWTH - .384*INFLATION + .628*GOODNEWS|
|Values to be computed:|
|?||Republican share of the two-party presidential vote in 2008 (V0TEP)|
|?||Republican share of the two-party House vote in 2008 (V0TEC)|