On Nov. 4, the American populace will vote, the electoral college will decide and, on Jan. 20, 2009, we will have a new president. But who will it be?
Poll results indicate a Senator Barack Obama lead by an average of 5.9 percent, according to the aggregation at http://www.RealClearPolitics.com. State-by-state polling indicates that the Electoral College heavily favors Obama and puts him with a 375-163 lead.
These snapshots indicate a probable Obama victory. But election night reality might be very different. Why might pollsters predicting an Obama victory be wrong?
There is the infamous Bradley Effect, named after Tom Bradley, an African American who ran for governor of California in 1982. Bradley had a double-digit lead in the polls prior to Election Day. His campaign had already opened the champagne and started the party well before the day was over. As the voting tallies poured in, however, the Bradley campaign experienced a distinct sense of horror. Out of 7.5 million votes, Bradley lost by just 1.2 percent.
Many black politicians fare much worse in elections themselves than in polling presumably because voters lie to pollsters. In a study of 133 senate and gubernatorial elections between 1989 and 1996, black candidates did 2.7 percent worse than they polled. However, that same study also determined that between 1996 and 2006, polling bias was negligible. One can only hope the Bradley Effect no longer exists.
State-by-state polling can be highly misleading. Because there are 50 states, it is difficult for pollsters to conduct frequent polls in all the states. Therefore, swings in more up-to-date national polling can indicate significant swings in individual states not seen in state polling. If the race tightens just prior to Election Day, the electoral map may still look like a 375-163 Obama landslide when it is really much closer.
In aggregating poll results, one should expect to find a normal distribution of results a bell curve. Polling for 2008 does not follow this logic. Obama’s support ranges from 44 to 54 percent, while Senator John McCain’s lies between 39 and 46 percent (excluding undecided voters for both). CBS and the New York Times have found a 13-point margin for Obama, while polling firm IBD/TIPP (considered the most accurate in 2004) has a statistically insignificant three-point Obama lead.
These widely varying results indicate that, barring errors in sampling, pollsters have vastly different weighing criteria of who they think will be likely to vote. Polling firms must ask themselves many questions about what they think will happen on election night. Will there be a significant youth vote this year? What will the black voter turnout be? Will currently undecided voters suddenly decide to favor one candidate? It all depends on how each firm weighs its data.
One polling firm, Gallup, has even declared that it is so unsure of how to weigh its polls that it has plotted three entirely different scenarios. Gallup’s results vary between a two- and seven-point Obama victory. This is a five-point swing, all from just one polling firm!
There have been times during this year’s primary season where not just one, but every single pollster got the results wrong. In the New Hampshire Democratic primary, polls indicated that Obama would win by 8.3 percent. Instead, Clinton won by 2.6 percent. For pollsters, it can sometimes be a guessing game, and they might all be wrong.
Yes, these are reasons for uncertainty. But the most conspicuous reason not to write this election off lies within the remaining few days before Nov. 4 the American electorate is fickle. By Sept. 2, after the Democratic Convention, Obama was leading by 6.4 points. Just six days later, McCain was leading by three points. The current economic crisis then led to McCain’s recent demise his fate is tied to the Bush administration. This prior volatility indicates how quickly voters can change their minds in accordance to news.
Should the economy magically stabilize and news of impending economic doom leave the headlines, it is possible that we will have a much closer race.
Regardless of all these factors, I have a vague sense of je ne sais quoi that this election will be much closer than it seems right now.
Otis is a member of the class of 2011.