Suppose that this November’s presidential election is so close that one candidate wins the popular vote and the other wins the Electoral College (and thus the election). Which candidate, Mitt Romney or Barack Obama, or neither, holds an edge in the Electoral College system for choosing U.S. presidents? Though his advantage is small, I believe it belongs to Romney.

My analysis begins with what I call a “political quotient.” I’ve constructed this device to measure a person’s political views quantitatively. Higher PQs correspond to more liberal views, with “100” indicating an outlook approximately as liberal as Nancy Pelosi’s or Barney Frank’s, while “0” indicates positions approximately as conservative as Jim DeMint’s or Michele Bachmann’s. According to my estimates, the PQ of the average American voter is 50.4.

In my book, Left Turn, and on my website, I estimate the PQ of the average voter in each of the 50 states and find that Iowa is the most moderate state in the nation: its average PQ, 50.7, is closest of all states to the national PQ 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 2008 Electoral College according to the PQ of the representative’s state, starting with the lowest. In such an ordering, the first six members would be 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.

One PQ point translates into about a half percentage point in terms of votes in a national election. For instance, Colorado is about 2.2 PQ points more conservative than the national average (50.4 minus 48.2). Accordingly, in a national election, we can expect Colorado to vote for the more conservative candidate by about 1.1 percentage points (one-half of 2.2) higher than the nation. If in the upcoming election the nation splits its vote 50–50 between Obama and Romney, then Colorado would vote 1.1 percent higher—or 51.1 to 48.9—for Romney.

Thus, according to my analysis, if the election produces a split decision, with a majority of the electorate picking one candidate and the Electoral College picking the other, Romney would win the Electoral College and become president.

Some recent polls support this forecast. Currently, the RealClearPolitics average poll, which surveys voters across the nation, gives Obama a 2.6 percent lead over Romney. Meanwhile, an NBC News/Wall Street Journal poll, which surveys voters in twelve swing states, gives Romney an 8-point lead over Obama. The average PQ of the 12 swing states is 48.7—almost exactly the PQ of Colorado. The 12 states should therefore predict very well the outcome of the Electoral College. These polls, like my PQ analysis, suggest that the Electoral College tilts slightly toward Romney, compared with a pure popular-vote system. Expect soon for liberals to renew their complaints about the unfairness of the Electoral College.

Tim Groseclose is the Marvin Hoffenberg Professor of American Politics at UCLA and the author of Left Turn: How Liberal Media Bias Distorts the American Mind.