Chester Thai ’14
Though an established part of any political race, election polls actually fail to help the public interest by misguiding and deceiving voters as a result of their constant fluctuation.
The purpose of election polls is to outline in clear data which candidates have the most support from voters at a certain time. However, it remains unclear whether the polls are reliable as indicators of the eventual nominee and if they should affect voter decisions. According to the American Research Group, the predictions of seven major polls showed Romney’s victory within a 5% margin of error. Moreover, The New York Times has shown that the average margin of error during the senate race two years ago among eight different firms was 4%. These numbers show that polls, to a degree, do represent how people would vote. However, these poll numbers reflect a candidate’s chances going into an election. The further back from the actual election, the more polls become misleading and liable to change. In the months prior to the Iowa primary, six different candidates spent time in first place. A week before the South Carolina primary, Romney carried a double-digit lead over fellow candidate Newt Gingrich, but lost to Gingrich by thirteen percent when, as noted by Charles Bryant, Chair of the History Department, “twenty percent of voters in South Carolina changed their minds in the last week.” “Polls are only a snapshot of the mood at the time,” he explained. “Emotions can change with the heat of the moment.”
However, there is another important debate: are the polls accurate? The answer lies in the way polls are conducted. Usually a small sample of the voter population is polled, through random phone calls or internet technology. For example, the Quinnipiac University Poll typically calls a thousand people over the course of a week through randomly generated telephone numbers. However, can the views of a thousand individuals truly represent the population of a state, or even the entire country? As Michael Barone of the Wall Street Journal wrote, “political polling is inherently imperfect.” Many factors contribute to this.
Bryant noted, “Telephone polls skew the results demographically because only those with landlines are polled … Also, internet polls lean towards the affluent because of more available technology.” In addition, Bryant said, “How you ask the question is as important as the question itself. If the question asks for a choice between two candidates, then you would be more likely to pick the candidate you hear last.” There is also the possibility of bias. The FiveThirtyEight blog of the New York Times found that a couple of the major polling firms were considerably biased towards one party during the senate races in 2010. Bryant also mentioned “push polls,” where a voter is asked a neutral question about two candidates, followed by additional details that favor a particular candidate.
How then, between the volatile swings and inaccuracies, can the ordinary voter hope to make sense of pre-election polls? Voters should not rely heavily on polls when deciding on a candidate. Even if polls, without their shortcomings, allow us to assess how the candidates are viewed in the eyes of the public, they are still not necessary. In fact, they may even hurt voters. The concept of polls is to inform the public of which candidate is leading his or her competitors. By doing so, people receive the impression of whoever is the “best.” This narrows the minds of voters because there should be no concept of one absolute best candidate except in our personal beliefs. We should be able to decide on our own whom we would vote for by our own assessment of the candidates, because each has his or her own views on different issues. Is this not the entire point of voting, to see who has the ability to meet the needs of each individual? Combined with the many faults of polls, our ability to freely elect our own representatives without outside influence may be compromised.