Predicting Election Results: Who Knew?

For the most part, polls predicting the outcome of the 2012 presidential election were fairly accurate.

by / November 7, 2012

In a political race as close as the 2012 presidential contest, surveys were nearly continuous; news outlets and nonprofit organizations, among others, sampled voter responses to candidates, issues and even individual statements in an attempt to predict winners and losers. For campaigns, poll results offered an opportunity to shore up flagging efforts, redouble successful strategies, and ensure their candidates and issues won the day. For the candidate, polls offer the chance to, as a French politician once said, "Find out where the people are going, so I can lead them."

For the most part, polls were fairly accurate in their predictions. Of 16 major organizations that conducted polls between Oct. 29 and Nov. 5, eight had President Obama in the lead, two had Romney in front, and six had the two candidates tied. In most cases, the differences were within only a few percentage points. According to reports, however, Nate Silver, a statistics whiz, predicted 49 of 50 states in the 2008 presidential election. And this year? 50 out of 50 predicted correctly. One tweet opined that he was from the future.

Some less scientific measures also seemed to prove themselves. For example, no Republican candidate has ever won the presidency without winning Ohio, and Mitt Romney did neither on Nov. 6. And in an experiment, 30 million Twitter mentions of Romney or Obama predicted an Obama win, according to Slashdot. The experiment succeeded in predicting the winner based solely on sheer numbers of mentions, even though it did not differentiate tweets that were critical or laudatory of the mentioned candidate.

Attempts to predict the outcome of a presidential election can employ complex algorithms and a scientific approach, or some quirky but consistent predictor. But any predictions must come up against the Electoral College, in which electors decide who will become president. Four times in U.S. history, a candidate who lost the popular vote has been elected president. In 2000, for example, George W. Bush lost the popular vote to Al Gore but became president anyway. Even more bizarre, in nearly half the states, the appointed electors are not required to vote for the candidate who won their state's popular vote, effectively disconnecting the actual selection process from the will of the voters. Such a wild variable could one day blow all predictions out the window.

The 2012 election was no sooner decided than predictions began to appear about 2016 and who the candidates might be. However, as Niels Bohr once said, "Prediction is very difficult, especially if it's about the future."

Wayne Hanson

Wayne E. Hanson served as a writer and editor with e.Republic from 1989 to 2013, having worked for several business units including Government Technology magazine, the Center for Digital Government, Governing, and Digital Communities. Hanson was a juror from 1999 to 2004 with the Stockholm Challenge and Global Junior Challenge competitions in information technology and education.