At Bloomberg.com, Al Hunt puts forth an impossible-to-prove theory as pundit wisdom:
Over the past 50 years, 17 men and one woman have been chosen by the major parties to run for the vice presidency of the U.S. Only one -- Lyndon Johnson in 1960 -- demonstrably affected the outcome of the presidential race.
This is worth remembering as the nation enters the quadrennial feeding frenzy over completing the tickets.
Journalists who want to seem more worldly than the rest of the press pack must periodically state that something that everyone likes to talk about means nothing at all -- and minimizing the importance of a running mate is one of the most popular ways to say that you know better than everyone else. But there's no way to measure the impact of vice-presidential nominees on the outcome of a presidential election. Maybe George W. Bush would have lost in 2000 had he picked another Dan Quayle; maybe Al Gore would have won had he picked a labor favorite like Richard Gephardt rather than the relatively hawkish Joe Lieberman. Maybe if Ronald Reagan had picked someone other than his chief primary rival, George W. Bush, in 1980, he would have beaten Jimmy Carter and John Anderson with 49 percent instead of 51 percent, and his status as a minority-vote president would have made Congressional Democrats less willing to work with him. How can anyone know for sure?
Hunt gives himself a little room by saying that Lyndon Johnson was the only VP pick with a "demonstrable" effect on an election, implying that one is free to believe that other candidates mattered (as long as you don't mind Washington sages laughing at your ridiculous theory). The truth is, no one knows whether the running mates will be important this year. And almost no one believes it will be the deciding factor, so there's no need to refute that notion. So when writing about the possible running mates in 2008, just dispense with the caveats and go right into the baseless speculation.