I've been hearing people constantly say that this election will be very close, which is certainly possible. But here's my issue with that: Since 1916, only two patterns have existed with incumbent presidents running for reelection. They either win a second term by a larger popular and electoral vote margin, or they lose big against their challenger. It seems to be a pretty consistent pattern over the past century, which leads me to believe that one or the other will happen this year. Obama wins big or he loses big.
1924, 1948, 1980, 1992, and 1996 are distorted by the presence of a significant third party campaign. 1936 and 1956 were but slight improvements for the incumbent on their blowout of four years previous. 1964, 1972, and 1984 each had message candidates nominated as opponents who were not near the political center. 1932 had the shock of The Great Depression to cause a massive change in the electorate. The only incumbent election since 1916 that was roughly similar was 2004, in which the incumbent improved slightly over his results of four years earlier. Romney is not a message candidate, so absent some major new shock, there is no reason to expect a major change from 2008.
Let's not forget that 1980 only turned into a blowout in the last week, after the debate. Had the debate gone differently (or never happened), Carter may well have eked out a narrow victory, Anderson or no Anderson.
And I wouldn't quite put Mondale in the same category as McGovern or Goldwater.
