Consider this hypothesis about modern presidential elections: Whenever American voters elect a new president, they choose someone who is, along a critical dimension, the antithesis of the incumbent. The Incumbent Antithesis hypothesis, as I'll call it, fits recent history, and it may be correct. If so, it suggests a real challenge for the next Democratic nominee, even if it is Hillary Clinton -- perhaps especially if it is.
In the modern era, almost all presidential elections seem to support the hypothesis. In 1960, John F. Kennedy campaigned as the antithesis of Dwight Eisenhower -- young and vigorous, the voice of a new generation. Richard Nixon, emphasizing the theme of "law and order," succeeded Lyndon Johnson, who was widely thought to have unleashed social forces responsible for rebellion and violence.
Jimmy Carter, the peanut farmer from Georgia, was the model of integrity, the anti-Nixon. Ronald Reagan defeated Carter with precisely the qualities Carter lacked: He was decisive, optimistic and fiercely protective of U.S. interests.
Pledging to "put people first," Bill Clinton portrayed George H.W. Bush as distant, old, indifferent and out of touch. In a not-so-subtle slap at the Clinton years, George W. Bush pledged to restore honor and dignity to the presidency. For his part, Barack Obama campaigned as a sharp critic of the Iraq war who would replace Bush's famous gut reactions with a careful, deliberative and inclusive approach to government.
In recent decades, the only exception to the pattern seems to have been the election of George H.W. Bush, Reagan's vice president. But Bush succeeded in part by distancing himself from his boss, promising a "kinder and gentler America," emphasizing personal characteristics that would appeal to swing voters and to some of Reagan's harshest critics.
Insofar as the Incumbent Antithesis hypothesis applies to two-term presidents, it can be linked directly with another pattern, which is that presidential approval ratings plummet in the second term. This happened with Harry Truman, Eisenhower, Johnson, Nixon and Bush. (Reagan and Clinton are the exceptions; Obama's trend lines fit with the majority.) This seems to explain why American voters would support a candidate who seems, in crucial respects, to be the opposite of the incumbent.
So the Incumbent Antithesis hypothesis is plausible, but is it true?
That's not as easy to say. It's based on a small sample size -- just eight elections since 1960. And many characteristics separate the eight incumbents from their replacements; it's hazardous to claim that any of them is causal. The winner is usually younger than the incumbent, too, and has darker hair, but we should hesitate to conclude that greater youth or darker hair makes for success in a presidential election.
We may as well say that what matters most is to belong to the opposing political party. Since 1960, candidates who succeed two-term presidents almost always have. (George H.W. Bush is the exception.)
In this light, the Incumbent Antithesis hypothesis cannot be taken as hard fact. But it also can't be ruled out. Whenever a two-term president's popularity ratings are low, swing voters may be especially interested in candidates who are, in important ways, the opposite. That's a warning for potential Democratic candidates, including Hillary Clinton -- and an important lesson for Republican candidates as well.
Sunstein is a nationally syndicated columnist.