Right now, there is all sorts of gossip and speculation about who John McCain and Barack Obama are going to pick as their vice presidential nominees. Most of that speculation focuses on the possible impact of the vice presidential nominee on the 2008 campaign. However, that is probably the wrong issue to focus on. Very rarely does a vice presidential nominee have a decisive impact on a presidential race. It hasn't happened since 1960, in fact (when John F. Kennedy's selection of Lyndon Johnson might have tilted Texas and some other southern states in a very close election).
No, the real impact of a vice presidential nominee is that whoever it is has a strong chance of becoming a future president himself. Four of the last eight presidents (Lyndon Johnson, Nixon, Ford, and George H.W. Bush) were former veeps. Two other veeps (Walter Mondale and Al Gore) won their party's presidential nominations during that period; and one of them (Gore) came as close as you can get to winning the presidency without actually doing it. In this March 11 post, I explained why the role of the vice presidency as a stepping stone to the top job creates some harmful incentives: presidential candidates have an incentive to select VPs who might help them win the current election, even if those individuals are unlikely to make good presidents themselves (although the VP nominee is unlikely to have a decisive electoral impact, candidates are usually reluctant to give up even a slight political advantage).
Why is the vice presidency such a valuable stepping stone to the top job? There are several possible reasons. One especially important one is the name recognition that comes with being VP. Most voters are rationally ignorant and know little about politics and politicians. With the exception of the president himself, the vice president is usually the only political office-holder whose name is known to a majority of the public. As I note in this article, the majority of Americans don't know the name of their representatives or senators (much less anyone else's). By contrast, surveys show that some 60 to 70% can name the vice president. This relatively high name recognition is a tremendous advantage (see pg. 1213 of this article for one such survey result). Such relatively uncharismatic VPs as Nixon, George H.W. Bush, Mondale, and Gore would probably have been unable to win their party's presidential nominations without it. Besides the president, the only politicians who can compete with the VP in name recognition are those who were famous prior to winning elected office (e.g. - Hillary Clinton and Arnold Schwarzenegger).
In addition to the way name recognition helps VPs win their party's nomination, there's also another obvious path by which the veep can reach the presidency: if the president dies, is incapacitated, or is forced to resign. Of the forty-three presidents in American history, nine were veeps who reached the top job by one of these routes, with Gerald Ford being the most recent.
The bottom line: the really important stake in this month's veepstakes is not the 2008 presidential election but the selection of a person who may well become president in the future.
UPDATE: The flawed link to my related March 11 post has been fixed. Thanks to those who pointed it out.
UPDATE #2: As some of the commenters point out, a total of 14 vice presidents have become president in American history. This makes up about 30% of the total number of presidents (43). It also means that about 20% of all veeps have eventually ascended to the top job. A VPs odds of becoming president are vastly higher than that of any other type of officeholder, including governor, representative, senator, Supreme Court justice, etc. There have been about as many former governor presidents as former Veep presidents, but of course the total number of governors is vastly higher than the total number of vice presidents.