After several weeks of panic over the Kerry campaign's supposed inability to take advantage of the recent bad news for the Bush administration, a new meme has risen to the fore. With a more sophisticated look at the polls, the new thinking goes, we can see that John Kerry's going to win. Big. The logic, implicit in a recent Andrew Kohut op-ed, is made more explicit by Mark Mellman in The Hill and stated even more clearly by Hotline Editor Chuck Todd in the current issue of The Washington Monthly.
The key move in all three pieces is to take a look at history: Incumbent presidents almost never win in close elections. Sometimes, as in 1984 and 1996, the public is basically satisfied with the direction in which the country is headed, and the incumbent wins easily. Other times, as in 1980 and 1992, the public is not. In the latter cases, misgivings about the challenger tended to keep the race looking fairly tight for a while, but eventually those disenchanted with the incumbent found themselves voting for the only alternative. Because polls make it clear that the public is not at all happy with George W. Bush's conduct in office, the logic goes, Kerry will emerge to win by a healthy margin -- current horse-race polls are a lagging indicator, and Bush's low approval ratings are the key thing.
I have my doubts. Looking to the past in hopes of discerning immutable laws of history that can guide our predictions of the future is a difficult -- some would say dangerous -- enterprise. The sample of American presidential elections taking place under anything resembling modern conditions is perilously small. Much, moreover, turns on questions of definition. Expand the notion of "incumbent" to include races where a sitting vice president runs as the successor to the incumbent president and the extremely close elections of 1960 and 2000 enter the picture. So, too, does the 1988 election, the results of which were more consistent with the "incumbent landslide" theory.
The year 1968 represents an even more confusing case. As in 1960, 1988, and 2000, one party had a sitting vice president as its nominee. Hubert Humphrey, however, ran not as the successor of a president barred from running again by term limits but, rather, to succeed an incumbent president whose policies had been so thoroughly discredited that he was forced to withdraw from consideration. The result: a very tight election, with Richard Nixon besting Humphrey by just 500,000 votes out of more than 70 million cast.
Or was it? It was a blowout in the Electoral College, with Nixon receiving 56 percent to Humphrey's 35. More important, the Nixon-Humphrey popular-vote margin was given a fairly spurious closeness by segregationist George Wallace's strong 13 percent showing, the exclusion of which from the analysis masks the extent of the public's repudiation of the Democratic Party. In 1972, Nixon picked up all the former Wallace voters and trounced George McGovern.
Consideration of the 1968 election brings us back to the 1980 and 1992 races: All three were characterized first by strong primary challenges to the incumbent and then by third-party candidacies aimed at members of the incumbent's political coalition. Jimmy Carter would have lost the 1980 election even if he'd obtained all of John Anderson's votes, and the precise role Ross Perot's candidacy played in producing Bill Clinton's 1992 victory remains controversial. Still, both independent races -- presaged by primary insurgencies led by Ted Kennedy and Pat Buchanan, respectively -- symbolized the depth of disaffection with the incumbents who later went down to defeat. In both elections, voters who were not willing to question the basic ideology of the president were quite eager to question his personal leadership. Such voters by and large returned to the fold, but relatively nonideological swing voters who choose mostly on the basis of such personal factors ultimately broke for the opposition.
Thus, to the historical observation that incumbent presidents with low approval ratings typically lose big in November, we can add a second observation: No incumbent president since Herbert Hoover has been defeated without first facing a primary challenge. Moreover, with the exception of the post-Watergate election in 1976, no incumbent has been defeated without being weakened by a third-party challenger. Despite Tim Noah's best efforts, Bush faces no such opponent. Instead, it's Kerry who does, as he needs to deal with Ralph Nader siphoning off votes to his left.
So does Bush have it in the bag? Again, no. There is that pesky trend about unpopular incumbents. And Hoover did, after all, lose the election. What's more, Bush is on pace to become the first president since, well, Hoover to preside over a net decrease in employment. But then again, despite the job market's poor performance over the Bush years as a whole, things are getting better right now and will likely continue to do so through November.
Historical arguments, in other words, can be marshaled to prove just about whatever point you like. My favorite, though, comes from the not-so-distant year of 2003. Back then, those of us who followed politics closely were treated to one exciting show in the campaign for the Democratic presidential nomination -- the rise of the insurgent Howard Dean, the emergence of Wesley Clark as the leading “stop Dean” contender, Dick Gephardt's surprising failure to slip into irrelevance. Who would win? Well, John Kerry, of course, who, having been written off months ago, became the favorite within a week of the Iowa caucuses and then fought out a desultory battle with the eerily optimistic John Edwards.
Apparently the swing-voter types who decide elections start paying attention pretty late in the game. And as much as has happened over the past two months, much more could happen between now and November. The administration seems to be at least contemplating the idea of withdrawing forces from Iraq at the request of the soon-to-be-unveiled interim government, a step that would surely transform the dynamics of the race. Most notably, the United States might, like Spain, witness a pre-election terrorist attack. Would the public react by rallying around its commander in chief or by holding him accountable for his failure to better defend the country? I, for one, don't care to hazard a guess, and surely history can provide us little guidance in answering the question.
Matthew Yglesias is a Prospect writing fellow. His column on politics and the media appears every Tuesday.