Paul J. Richards/AFP/Getty Images
Allan Lichtman, a professor at American University, discusses his 13 keys to a successful election campaign on April 13 in his office in Washington, D.C.
In January 2010, more than a year before Mitt Romney had formally announced he was running for president, political historian Allan Lichtman predicted President Obama would be re-elected in 2012.
On Tuesday, Lichtman extended his record of correctly forecasting the winner of the popular vote to eight straight elections.
What makes Lichtman interesting is that he makes predictions early, long before eve-of-election polls, long before October surprises, and sometimes even before the nominees have been chosen.
Lichtman says he sees elections the way geophysicists see earthquakes — as events fundamentally driven by structural factors deep beneath the surface, rather than by superficial events at the surface.
He said he came to this idea after happening to meet a Russian geophysicist. They got to talking about earthquakes and asked themselves whether elections might follow the same principles as earthquakes.
"Everything we know about elections, we've already stolen from geophysics," Lichtman said in an interview shortly before Tuesday's election. "Tremors of political change, seismic movements of the voters, volcanic elections, political earthquakes. It's all geophysics anyway."
Rather than think of elections as battles between liberals and conservatives, or even between two candidates, Lichtman said he decided to test the idea that elections follow earthquake principles: You either have stability, or you have upheaval.
Translated to elections, if the incumbent party in the White House kept the White House after the election, that meant you had stability. If the incumbent party lost, that meant there was upheaval — an earthquake.
Lichtman analyzed presidential elections between 1860 and 1980. Over that 120-year period, he looked for markers of stability and markers of upheaval.
Much of what he found is intuitively obvious: When the country was in recession or there was a foreign policy disaster during the tenure of the last administration, the incumbent party was likely to lose. When there was a major domestic or foreign policy success, the economy was doing well, or an incumbent president was running for re-election, the party in power tended to hold on to power.
What Lichtman did was take his data seriously: He found that in every election between 1860 and 1980, when the answers to six or more of the 13 questions he devised went against the party in power, there was an upheaval — the challenger won.
He applied the model to subsequent elections. Starting in 1984, the model has correctly predicted the winner of the popular vote in every election — sometimes months or even years before the election takes place.
Lichtman's model raises questions about the way the media cover campaigns and the way candidates run for office because it suggests that the ups and downs of horse-race coverage, gaffes and ad campaigns may not be as important to the outcome as most people believe.
"Primarily, elections are responsive to these much deeper forces," Lichtman said. "Focusing on the campaign is like focusing on the froth of the wave, instead of the wave itself."
Lichtman said the model actually shows the American political system in a positive light. Basically, it suggests good governance tends to get rewarded. If a party is in power, it doesn't need to have a perfect track record to keep the White House. It can afford as many as five strikes against it — any more, and there is an earthquake and the challenger wins.
Before the 2012 election, Lichtman said his model showed the answers to only three of the 13 questions — he calls them "keys" — turning against Obama: One was the long-term state of the economy. A second was the fact that the incumbent party in the White House had taken a shellacking during the previous midterm elections. The third was that Obama's sizable disapproval ratings meant he could not be considered a once-in-a-generation charismatic leader.