In his TNR column last week, my esteemed colleague and mentor William Galston expressed one of the more regularly repeated convictions about presidential politics: Reelection campaigns are a referendum on the incumbent. As he wrote:
One of the best established findings of contemporary political science is that in presidential contests involving an incumbent, the incumbent’s record is central to the public’s judgment. A race for an open Oval Office is about promises and personalities; a campaign for reelection is about the record and performance of the person currently occupying the White House.
In other words, the results of November’s election will hinge on what voters think of Barack Obama, not what they think of Mitt Romney. But is this really true?
Judgments of the incumbent’s record are certainly central to any campaign for reelection—but so are judgments of the challenger’s record, character, ideology, agenda or party. After all, life today looks a little better if you know that tomorrow could be worse. This argument isn’t just academic: It affects how Obama runs his campaign. Should he focus only on his own achievements? Or should he emphasize Romney’s failings and the agenda of the G.O.P.?
It turns out that political science is not an infallible guide to this particular subject. The sample size of recent presidential reelections is limited, and the most recent, in 2004, cut against the “referendum” hypothesis, and the closely associated belief that undecided voters will break sharply against incumbents late in the election cycle. The 2004 election also showed that drawing attention to doubts about the challenger is not always a waste of time—it certainly wasn’t for George W. Bush. Arguably, 1996 was similar: The Clinton campaign spent the early stages of the cycle feeding on negative perceptions of the opposition party.
A more nuanced version of the referendum hypothesis holds that challengers to even the weakest incumbents must cross some threshold of credibility before achieving victory. Take 1980, one of the prior elections that is often cited by both supporters and doubters of the referendum hypothesis. The contest between Jimmy Carter and Ronald Reagan was very close until Reagan, in that contest’s sole televised debate, famously asked, “Are you better off than you were four years ago?” Reagan was urging voters to make the election a referendum, just as Carter, in the same debate, was urging them to choose between two different agendas for the future. (If, of course, presidential re-elections are automatically referenda, then Reagan’s peroration would have been unnecessary and irrelevant.) Reagan’s entreaty ultimately won out, of course—but it did not necessarily have to be so. Even 1980 was, in an important respect, a “comparative” election. Jimmy Carter was a famously wounded incumbent representing a deeply divided party, and (towards the end of the cycle) bleeding support to a third-party candidate as well. That set the threshold for Reagan quite low, but he still had to achieve a certain level of credibility and approval to win. His campaign message and debate performance satisfied that criterion, overcoming Carter’s efforts to describe him as an extremist.
If economic conditions—or, for that matter, the perceived security and status of the United States—deteriorate to the levels bedeviling Carter in 1980, then the threshold Mitt Romney must cross to win may well be very low. But it still exists, and with Romney’s favorability ratings—and just as importantly, those of his party—at perilously low rates, Democrats would be foolish not to keep them as low as possible. And if, as is presently more likely, objective conditions in the country improve at a modest rate, then all sorts of factors could be decisive: including the populist “base mobilization” efforts that Galston cautions against, the arguable responsibility of past Republican administration and current Republicans members of Congress for poor economic conditions, and yes, comparisons of the two futures offered by the candidates.
Interestingly enough, just as Galston is warning Democrats not to ignore the “referendum” hypothesis, a smart Republican commentator, National Review’s Ramesh Ponnuru, has warned Republicans not to rely on it. Ponnuru cites 2004 as a precedent, suggesting that partisan polarization has changed the usual dynamics:
The strategic insight of the Bush re-election campaign in 2004 was that times had changed. The nation was divided 50-50 between the parties. The number of committed partisans had increased, and the number of true swing voters — as opposed to voters who say they are independent but reliably vote for one party — had shrunk. In a polarized country, no president could hope to achieve high approval ratings for very long.
It followed, though, that a president could win re-election even with approval ratings that would once have spelled doom. In a 50-50 America, every presidential election was a choice between the incumbent and the challenger and not just a referendum on the incumbent. If voters who didn't approve of the incumbent could be persuaded to prefer him to the challenger, the incumbent would win.
That was Bush's game plan. Republicans portrayed Kerry as an effete liberal who would raise taxes and wouldn't assert the national interest. They didn't try to persuade Americans that Bush had been a terrific president or even that Kerry was unpresidential. They just made the case that Bush was better than the alternative.
When they had to contemplate a choice between Bush and Kerry, some voters found themselves starting to approve of the president by comparison: At the end of February 2004, when Kerry locked up the nomination, Bush's approval rating was at 44 percent in the RealClearPolitics average of polls. By Election Day, it had climbed five points.
Bush did not win in a landslide. In fact, no president has ever won re-election with a lower percentage of the major-party vote. But win he did. (Then, with Kerry out of the picture, Bush's numbers resumed falling.)
If Ponnuru’s analysis is correct (and I see no reason why it isn’t), then the hyper-polarized atmosphere of 2012 may be even less conducive to a referendum.
In any event, as Galston himself notes cogently, Obama is not entirely the “master of his own fate.” There is only so much he can do to make life happier for Americans between now and November, and other actors, including the Iranian regime and the U.S. Supreme Court, could have an important impact on the dynamics of the election. So what is Team Obama to do? Certainly it should make every effort to defend the president’s record. But campaigns must work with the material at hand. Given the behavior of the GOP during the Obama administration, the ideological agenda it has foisted on its nominee, and the characteristics that have made that nominee such a hard sell to his own party, a strong “comparative”—or if you insist, negative—campaign against Romney is both wise and inevitable.
Ed Kilgore is a special correspondent for The New Republic, a blogger for The Washington Monthly, and managing editor of The Democratic Strategist.