Predicting The Presidential Election

PREV 3 of 4 NEXT

This calculation doesn't account for the possible economic slowdown in 2008. If the economy turns down slightly during the next nine months—as many economists expect—the predictions don't change dramatically. For example, should the annual income growth rate fall to 1 percent, the Republican share of the two-party vote would slip to about 52 percent; if income doesn't grow at all between now and the election, the Republican voting share would be approximately 50.9 percent. Bad news for the economy is good news for the Democrats.

Of course, noneconomic factors will also influence the election, none more so than the war in Iraq. The Bread and Peace model acknowledges that American voters dislike seeing their soldiers die in foreign wars. In the 1952 and 1968 elections, anger at the rising number of fatalities in Korea and Vietnam, respectively, may have helped unseat the Democrats, who took America into both conflicts. This year, Hibbs believes, the Republicans will pay a price for starting wars in Iraq and Afghanistan.

How high a price? According to Hibbs' model, every 1,000 American military fatalities costs the incumbent party 0.3 percent of the two-party vote. Given that the U.S. military death toll in Iraq and Afghanistan is now close to 4,500—and may well reach 5,000 before Election Day—the model predicts that the war will decrease the Republican vote by as much as 1.5 percentage points. In a tight race, that could prove decisive.

At least one eminent economist believes that the Bread and Peace model understates the electoral importance of Iraq. In a paper published in 2005, William Nordhaus, a colleague of Fair's at Yale, argues that the Iraq war almost cost Bush the 2004 election. Given the healthy state of the economy that year and the fact that he was an incumbent, Bush should have defeated John Kerry by about 15 percentage points. But his actual margin of victory was just 3 percent. If this logic holds up in 2008, the Democrats should win comfortably.

All of this leaves me persuaded to put some money on the Democrats. However, when I logged on to WillHill.com, a British betting site, I found that the odds of a Democratic victory had already tightened to 1 to 2. That means if I wagered $100, I stood to win just $50. (The Republicans were on offer at 6 to 4, meaning I could have put down $40 to win $60.) With the candidacies of Michael Dukakis and Al Gore firmly imprinted on my memory, I couldn't bring myself to back the Democrats at such unfavorable odds.

These odds indicate that bookmakers believe the likelihood of a Democrat triumphing in November is 66.6 percent, whereas they see the chances of a Republican winning to be 40 percent. In short, the betting market concurs with the voting models: The Democrats hold a significant advantage, but the race is far from over.

PREV 3 of 4 NEXT