City Journal has posted my latest op-ed, which argues that the Electoral College gives Mitt Romney a slight advantage. That is, my analysis predicts that, if one candidate wins the popular vote while the other wins the Electoral College, then Mitt Romney is more likely to be the one who wins the Electoral College.
The main argument involves an analysis of "Political Quotients." Specifically, I find that Iowa, with an average political quotient of 50.7, is the state that is most representative of the U.S., which has an average political quotient of 50.4.
In 2008, Iowa was also the median state in the Electoral College. That is, suppose you ordered all the representatives of the Electoral College according to the PQ of the representative's state, starting with the lowest. In such an ordering, the first six members would the representatives from Utah (the nation's most conservative state), the next three members would be the representatives from Wyoming (the nation's second-most conservative state), and so on. Such a list would contain 538 members, and the 270th member (the number necessary to win a majority of the Electoral College) would be a representative from Iowa.
In 2012, however, the median PQ state shifted to Colorado, a slightly more conservative state than Iowa, with a PQ of 48.2. The reason: conservative states (those with PQs lower than Iowa's) gained six electoral votes in the last census. Thus, if the nation votes exactly 50-50, then Iowa would also be likely to vote almost exactly 50-50 (since its PQ is nearly the same as the nation's PQ). But Colorado would tilt toward Romney, and with increased conservative electoral representation, so would the Electoral College.
Other evidence of Romney's Electoral College advantage involves some recent polls. Using a nationwide sample, the RealClearPolitics average gives Obama a 2.6% lead over Romney.
Meanwhile, an NBC News/Wall Street Journal poll, using a sample of voters from twelve swing states, gives Romney an eight-point lead. The average PQ of the twelve swing states happens to be almost exactly the PQ of the average Colorado voter. Thus, I argue, the twelve swing states are good predictor of how the Electoral College will vote.
Expect soon for liberals to complain about the unfairness of the Electoral College.