Historically, there has been no case of a candidate winning the presidential election while losing all three top bellwethers: Missouri, Nevada, and Ohio. Nev. was the last of the three to join the union, and get the vote, in 1864.
While Republicans and Democrats have squred off since 1856, the late-1800s saw some winners prevail in only one of the three states (GOP William McKinley won Ohio in 1896 and 1900; no Republican president has ever won election without the Buckeye State). Since 1908, the last 25 elections saw Mo. back the loser in 1956, Nev. going for the 1908 and 1976 runners-up, and Ohio did not picking the winners of 1944 and 1960. Agreeing as a trio in 20 of the last 25 elections—for 80 percent—Mo. has been right 96 percent, and Nev. and Ohio both 92 percent.
Mo., Nev., and Ohio are each included in the states that I predict in this year's election—one that I believe will show Sen. Barack Obama (D-Illinois) prevailing in the Electoral College.
Since Sam credits Obama with Nevada, I just wanted to present this for anyone else finding it worth consideration.
While Nevada's demographics may be changing once again, they are not, nor will they ever be, in my mind a "bellweather." Their population will remain fairly low for one thing, even with the rapid growth of Las Vegas.
Ohio's a better choice, but it was on the losing Republican side in '48. Missouri is more of a true bellweather since it's always had an effective mix of demographics that can predict the election. There's this one book I read that said Missouri was almost like two different states when it camt to it's regional makeup. But here, it appears the blue-collar moderate conservative vote has more influence there than in more liberarian Colorado, methinks. Exactly why it's that way when it went for Obama in the primary is beyond me, but that's how it goes. sometimes.