No clear picture from the polls before Election Day


Obama won his re-election bid before the end of election day, despite a race that seemed too close to call, especially in several swing states.

During the election, polls showed states like Florida and Ohio flipped-flopping between Romney and Obama, which is a reflection of how close the race was in these states.

For example, Mitt Romney took his first lead in an Ohio state poll on Monday, Oct. 29. The poll, conducted as a telephone survey by Rasmussen Reports, showed Romney with a 2 percent lead over Obama, giving him 50 percent of the vote as opposed to Obama’s 48 percent.

This lead could have proven paramount in the race for the White House as Ohio itself had a 49.5 percent chance of providing the decisive electoral vote, according to FiveThirtyEight, a political forecast. The forecast, which is maintained by Nate Silver, a political reporter with the New York Times, identified eight other states as tipping states, or swing states: Colorado, Florida, Iowa, Pennsylvania, Nevada, New Hampshire, Wisconsin and Virginia.

Obama won all these states is his easier-than-predicted re-election.

Behind Ohio, Virginia and Wisconsin respectively had the second and third highest chances of providing the deciding electoral vote, although Florida was a close fourth.

Despite Romney’s first lead in Ohio, most polls showed Obama with a slight advantage. For example, in a Quinnipiac University poll, Obama maintained a 5 percent lead over Romney. This same poll, conducted in conjunction with the New York Times and CBS News, showed Obama leading 49 percent to 47 percent in Virginia. In Florida, Obama led Romney 48 percent to 47 percent.

Given the margin of error, however, the presidential race in these key swing states seemed closer than one or two percent.

“There is error in every poll,” said Dan Jones, founder of the Dan Jones & Associates polling firm and a political science professor at the University of Utah. “What I’m finding this year, more so than ever, is a lot of disparity. There is more disparity in these polls than I have ever seen, and I’ve been polling since 1959.”

Jones explained that in polling, there are many key factors. Gender, age and sample size all have an effect on the polls, as well as over polling. “There are too many polls,” Jones said, “and I poll everyday.”

He also emphasized the importance of making sure a poll’s sample is representative of the demographic. For example, a Colorado poll conducted primarily in Denver, which favors Obama by 12, will undoubtedly reveal a preference for Obama, although many other parts of the state favor Romney, according to Jones.

“Not all polling firms are credible,” he added, as some use biased questions and techniques called push polling, where the person polled is given a disparaging yet hypothetical situation involving a candidate and asked to answer accordingly. Some pollsters also use automated polling services that sometimes call in the middle of the night and seem more intrusive.

When it comes down to predicting the winner, though, Jones admitted before the election, “I really can’t tell who will win this election.”

While disparities did exist in many state polls, especially those in swing states, Nate Silver proposed that averages from state polls usually predict an election.

After compiling poll data from previous elections, starting with the 1980 election, Silver identified 44 instances in which state polls revealed a single-digit difference in percentages between two candidates. In all but nine instances, the favored candidate, as determined by the average of at least three credible polls, won the state and its electoral votes.

Further, the instances in which a candidate won the state despite trailing in pre-election polls occurred with less than a two percent difference between the candidates. Meaning, historically, a candidate who led a state by two or more percentage points always won the state.

This still implied Romney had a chance at winning key swing states despite Obama’s lead in a majority of the polls. In other words, though history was on Obama’s side, there are cases where the trailing candidate won the state, especially when he trailed by less than two percentage points.

Ironically, national polls painted a different picture. A number of prominent and credible polls revealed that, on  average, Romney led the popular vote, even by as much as five percent. These polls included Gallup, NPR, ABC News, the Washington Post and Rasmussen Reports.

Both candidates seemed fully aware of these polls. After a brief pause in campaigning due to Hurricane Sandy, both returned to the campaign trail in full fervor. Romney returned on Oct. 31 with a rally in Tampa Bay, Fla., the first of three rallies in Florida.

The following day, Nov. 1, Obama hit the trail running with a quick succession of rallies in Green Bay, Wisc.; Las Vegas, Nev.; and Boulder, Colo.

Romney and Obama were not on the trail alone. Even BYU students involved themselves in the election, offering their support to the candidates at campaign events.

Ty Nielson, a junior from Alpharetta, Ga., majoring in mechanical engineering, recently spent time in Las Vegas, Nev., working with the Romney and Ryan campaign.

As a volunteer, he was responsible for going door-to-door to the houses of registered Republican voters and encouraging them to vote.

“It was a super sweet experience to take part in the Romney-Ryan campaign,” he said. “There were volunteers from Idaho, California, Utah, and, of course, Nevada. Surprisingly, a substantial amount were non-Mormon.”

According to Nielson, this was his way of making a difference in the campaign.

“I am from Georgia, so my vote doesn’t really matter. Whether I vote or not, the state will go red. Going to Nevada was my way of making a difference in this election.”

Print Friendly, PDF & Email