Over the past few decades, political scientists have come up with increasingly elaborate statistical models for predicting presidential election outcomes. The basic idea is that voters evaluate presidents largely on the basis of how the economy has fared under them. That extends to potential successors in the same party. As such, McCain basically inherits Bushs record and Obama inherits Clintons record. These statistical models have also gotten quite accurate in their presidential forecasting. In 2004, during a very close race, eleven out of twelve studies correctly predicted a Bush win.
So why is Obama certain to prevail? Much of it boils down to what youve already heard in the press: what political scientists call the fundamentalsprimarily presidential approval, the state of the economy, and incumbencyare indisputably in the Democrats favor this year. And dont be deceived by fluctuating poll numbers, either. At this time in 1988, George H.W. Bush was solidly trailing Michael Dukakis in Gallup polls. Likewise, at this time in 1992, George H.W. Bush was solidly leading Bill Clinton. Enough said.
While political science models differ from one another in the variables they take into account and in how they weigh them, nearly all of the models are quantitative. That means they not only pick a winnerthey also forecast what his or her margin of victory will be. So what are the models saying for 2008? Douglas Hibbs, who has been studying the relationship between voting and the economy since the 1970s, has predicted that the Republican candidate will win only 46.7 percent of the vote. (That means the Democrat would win 53.3 percent of the vote, since these models assume only a two-party race.) In two different models, Michael Lewis-Beck and Charles Tien of Iowa, while using quite generous assumptions for Bushs presidential popularity (35 percent) and GNP growth (3 percent annual), still predict either a narrow Republican defeat (with McCain winning 48.5 percent of the vote) or else a far more resounding one, with McCain winning only 41.4 percent. Alan Abramowitz of Emory, assuming 2 percent economic growth for the first half of the year and a 30 percent difference between the presidents approval and disapproval numbers, estimates that McCains vote total will be just under 44 percent. Ray Fair, a Yale economist, puts McCains total at 47.8 percent. And the outlier in this crowd, Helmut Norpoth of SUNY at Stonybrook, has said that the Democratic candidate would beat McCain by only a single percentage point or lessstill far from encouraging as a best-case scenario for the GOP.
So, then, if its already been proven scientifically to be a Democratic year no matter what, why even bother to have a campaign? Or an election? Why not just hand the White House over to Obama right away? Well, heres the catch. First of all, political science isnt saying that if the election were held today the Democratic candidate would win. The models assume a November election. Nor are political scientists saying that Obama could win without campaigning. Indeed, their models assume that both sides will campaign. Rather, what theyre saying is that the function of the campaigns will merely be to awaken the latent preferences already existing within voters. Sure, bulging campaign coffers and sky-high IQs are important assets, but theyre unlikely to be significantly disparate between the two campaigns. Political scientists therefore tend to view them as a wash. (And, yes, this does tend to make us boring Beltway dinner guests.) Think of campaigning as irrigating a field in which the seeds are already planted and the weather isnt too crazy. You cant change the crop, but you do need to add the water.
And what about the worry du jour, the supposed split in the Democratic coalition? Again, take it from a political scientist. Dont sweat it. Democrats will come home. John Sides of George Washington has shown that loyalty has been on the rise since the early 1970s, when only about 60 percent of Democrats voted for the Democratic nominee. Todayand since the 1990sits been closer to 90 percent. Sides has also dismissed the partys great education dividecollege versus no college. Its at times non-existent and, when it exists, has not widened over time. More than 80 percent of Democrats with or without college degrees have supported the Democratic presidential candidate in the general election since 1988, and that number has been closer to 90 percent over the past decade. Bitter as Hillarys supporters have been, theyre not going to make a meaningful dent in this trend.
Now, you may still think Im being overly confident in my soothsaying. And Ill concede its a strange year. For the first time since 1952, the incumbent partys candidate is neither a president nor vice president. In theory, just as Bill Clintons strong economic record did not rub off on Al Gore, George Bushs foreign policy errors will not rub off on John McCain. But I seriously doubt it. My friends, as John McCain might say, the die is cast. As long as Democratic donors write checks, the Obama ground team performs in November as it did in February, and (let us pray) the country avoids a major terrorist attack, the story of 2008 will end happily for Democrats and for the reputation of political science. And at this point, with all the bets I plan to make, its the only outcome I can afford.
