McManus: Presidential crystal balls
Unemployment is mired at 9%, and President Obama’s poll ratings are mired too. Democrats are dispirited. Republicans are fired up and ready to go. Activists on both the right and the left are demanding change.
So is there any way Obama can win reelection next year? Sure he can. Or so say America’s presidential election soothsayers.
Economists, political scientists and historians have created a minor academic industry out of forecasting election outcomes. They all have different formulas for drawing their conclusions, but many of the predictions I’ve looked at in recent days give Obama an edge next year no matter who his Republican opponent turns out to be.
The forecasters hedge, of course. Everything could change if the economy slides back into recession or unforeseen calamity hits. But many of them see Obama as the likely winner.
Under a model developed by Larry Bartels, a political scientist at Vanderbilt University, “Obama is very likely to win even if economic conditions do not improve between now and election day.” That’s what Bartels posted last week on a political science blog called the Monkey Cage.
What prompts such a conclusion a full year before election day?
Easy: 5.17 + (3.49 x income growth in 2012).
That’s a simplified version of a formula that forecasts the likely margin of victory for a president who’s running for a second term. Bartels has back-tested it on every election since 1948, and he says it works pretty well.
Under the formula, even if growth stays low, Obama should still get a boost from two factors. One is that American voters rarely turn a president out of office if his party has held the White House for only one term. (That happened only once in the last 123 years, when Jimmy Carter lost in 1980.) The other is that voters may look at the economy next year and give Obama credit for ending the recession that came before.
“Obama’s fate will hinge not only on how the economy fares in 2012, but also on whether and how voters think about the economic mess he inherited in 2009,” Bartels explained in an email. “If they grade him on a curve, he is likely to be reelected; if not, he is likely to lose.
“This is not a forecast,” Bartels insisted; it’s mostly a hypothesis aimed at figuring out what influences voters’ decisions.
But others aren’t so cautious. A historian at American University, Allan J. Lichtman, has compiled a list of what he calls 13 “keys” to winning the presidency, and he says Obama has basically won reelection already. “It’s conceivable that Obama could lose if the economy goes into another recession or if he gets hit by a major scandal, but I don’t see it,” Lichtman told me. His keys include incumbency, the economy, foreign policy success and charisma, and he says Obama is hitting on 9 of 13. Most political scientists and economists dismiss Lichtman’s model as too subjective, but he points out that he’s called the winner correctly seven elections in a row.
If you’d like a one-stop digest of almost every conceivable model, there’s a website called pollyvote.com that puts them together in one combined forecast — figuring, basically, that they can’t all be wrong. Pollyvote even includes “facial competence,” an experiment in which people rate candidates on how impressive they look. (Obama beats Mitt Romney, but Hillary Rodham Clinton beats Obama.) Pollyvote’s combined forecast predicts that Obama should win 50.4% of the popular vote — presumably enough to squeak through to a second term.
Facial competence aside, what are the forecasters saying?
Essentially, three things.
First, at this stage, polls don’t matter. At this point in the 1984 presidential campaign, President Reagan had the support of only 53% of the voters, but he went on to win by a landslide; President George H.W. Bush had the support of 59% of the voters, but he went on to lose. No modern president has won reelection with a job approval rating as low as Obama’s current 43%, but this year’s polls can’t forecast next year’s votes.
Second, the economy does matter — a lot. There’s vigorous debate among academics over which economic index is the best predictor of voter behavior (overall growth or growth in disposable income), and when it matters (during the election year only or during the president’s entire term). And there’s debate over what noneconomic factors count, and how to count them. But almost all of the forecasting models confirm James Carville’s expert analysis: It’s the economy, stupid.
Finally, incumbency matters. Obama may look battered and beleaguered, but history suggests that incumbents can often improve their standing through executive actions, and that voters often hesitate before voting against someone they supported four years earlier.
Does this mean that our long and noisy presidential campaigns are just an expensive charade, and that the only thing that matters is whether the economy goes up or down?
Not really. “None of these models is perfect,” said Ray Fair, an economist at Yale University whose own forecast method is based mostly on growth in gross domestic product. “There’s plenty of room for other factors to count.”
Especially, perhaps, this time. Last month, Fair plugged the latest growth forecasts into his formula and issued a new estimate of Obama’s likely share of the popular vote: exactly 50%.
If you don’t like any of those formulas, there’s always this old, unscientific one: the height index. More often than not, the taller candidate wins. Obama is 6 feet 1; Romney, 6 feet 2; Herman Cain, a reported 6 feet even.
Any way you measure it, it’s likely to be a close election.