How—and how well—does the modern process of vice presidential selection work? What can be learned by considering the nominations of Texas Sen. Lloyd Bentsen by the Democrats and Sen. Dan Quayle of Indiana by the Republicans in 1988?

Most students and practitioners of American politics agree as a matter of principle that vice presidents should be chosen with their constitutional role as presidential successor uppermost in mind. Nine vice presidents, more than a fifth of those who have held the office, have succeeded to the presidency. (They served as president for forty-two years, twenty-nine of them in this century). During one recent period from 1945 to 1977, vice presidents who became president by succession occupied the Oval Office fully half the time. Illnesses, impeachment proceedings, and assassination attempts have made succession an active possibility during twenty of the nation's forty presidencies (Goldstein 1982, 207-08). Every postwar vice president except Walter F. Mondale has become the subject of unusual public concern because of some event or condition that raised the possibility that he would succeed to the presidency.

To the extent that the concern for presidential succession is taken seriously, it implies two governance criteria for evaluating the vice presidential selection process. The most important governance criterion is the competence of nominees for vice president to be president. Historically, six of the nine successions have occurred during the vice president's first year in office, suggesting that even the best on-the-job training is no substitute for a wise initial choice.