Never Lie to a Pollster

The article I chose to critique, “Eighty-Four Percent Say they’d Never Lie to a Pollster,” October 15, 2008, can be found on Ann Coulter’s blog website. Ms. Coulter is the author of six New York Times bestsellers, is a legal correspondent for Human Events, and writes a well-known syndicated column for Universal Press. She has gained popularity with many due to her controversial and confrontational style, and was on the cover of TIME Magazine, April 25, 2005. On the other hand, she has been fired as a correspondent (more than once) for inflammatory remarks made during interviews and is considered by many to be an extreme right-winger. Coulter has described herself as a polemicist who likes to "stir up the pot" and, unlike broadcasters, does not "pretend to be impartial or balanced."

Coulter’s intended audience is directed toward readers who are interested in the correlation between election results and voter polling. Ms. Coulter presents an argument against the current buzz making its way through the media--the so-called “Bradley Effect.” This is “the notion that some material number of voters will lie about their intentions to pollsters, claiming that they will vote for a black candidate when in fact they will vote for the white guy.”

It is interesting to consider whether this phenomenon does exist and if it really does influence the outcome of an election. Does the American idea of political correctness make us lie to pollsters to avoid the impression of racism? She says the idea of institutional racism is just a way for democrats to “whine” about American voters. Coulter claims that not only is there social pressure to continuously convince others that you are not a racist but now it seems people are also feeling pressured into defending why they are republicans.

Ms. Coulter argues that polling results over the past 32 years (since 1976) have typically been skewed toward democratic wins. In fact, the election results indicate that not only have the polling numbers and projections been wrong, but they have been very wrong. It is not until the recent election of 2004 that the polling results actually provided a correct prediction of an election outcome.

Based on Ms. Coulter’s analysis of the previous 32 years of polling and election results, I believe that there is not a correlation that can be used to accurately predict outcomes. The theory that we can randomly sample a few voters in a few selected locations and determine an outcome is questionable. Human nature is to not share personal information with strangers, and let’s face it, voting is very personal. I believe that not only what kind of question is asked, but who is asking the question, and under what circumstances will influence how a voter might answer a pollster—truthfully or not. Perhaps if polling could be done anonymously, for example, a polling station at a precinct where a voter is chosen at random to stop and answer polling questions without anyone around. The voter could be selected at random through the computerized voting machine and given a ticket with a unique password so that polling fraud could be reduced. They would then take the ticket and login to a polling station. This would also reduce polling bias by the pollster, and results would be immediate. In other words, the system we are currently using is antiquated and not scientifically robust. With our new technological world, it is time to investigate methods that are more substantial.