“Dewey Defeats Truman” read the headline on Nov. 3, 1948. The Chicago Tribune had given the election to Gov. Thomas E. Dewey of New York. The paper had depended on polls to forecast the election results. Gallup had given the election to Dewey 49.5 percent to 44.5 percent. We know what really happened. America is a country obsessed with polling.
Politicians’ actions are dictated by polling results rather than their own convictions. Polls not only indicate what issues a constituency will support, but also who they will support. In the lead up to presidential elections, which nowadays begin before the last ballot is cast for the prior election, polls are especially prevalent, or annoying depending on your persuasion. Yet, with so many polls given, the data is often contradictory. Doesn’t that defeat the purpose of polls?
Merriam-Webster Dictionary defines a poll as “a questioning or canvassing of persons selected at random or by quota to obtain information or opinions to be analyzed.” Once polls are analyzed, they are supposed to be an indicator of what will happen in the future. In reality, this couldn’t be further from the truth. For example, a Washington Post/ABC News poll had Republican presidential candidate Mitt Romney beating President Obama in a head to head matchup 49 percent to 46 percent. The media was whipped into a frenzy. ABC News reported this was “a clear reflection of Obama’s vulnerability to a well-positioned challenger.” Among registered voters, Obama would lose to Mitt Romney, the same man who lost the GOP nomination in 2008 to John McCain. It has to be true; the poll said so.
Not So Fast
Interestingly, A Ipsos poll conducted for Reuters showed Obama shellacking Romney by 13 points. If you didn’t notice, the articles discussing the polls’ results were from the same day! Let’s review: two polls designed to predict the same event contradict each other. Not only do the polls disagree; they are 16 points apart, which in politics is the equivalent of being beat by about five touchdowns. Need more evidence of the uselessness of polls?
According to CNN, the leader of the GOP at this point is — wait for it — Rudy Giuliani. So, Mitt Romney would beat incumbent Obama in a presidential race, but lose his own party’s nomination to a man who may not even run.
Politics Isn’t Math
I know what the critique of this article will be: there are many variables that determine the outcome of a poll. How do the questions differ? Who is being asked: registered voters or likely voters? Were the respondents called on the phone or was the poll given in person?
But this is exactly my point, there are too many variables. Politics is a people sport that can’t be packaged into a one size fits all box. One thousand people responding to a question is not representative of a country of 300 million. Some mathematician may say it is, but it’s not. After Osama bin Laden was killed I could have written the script of upcoming events.
In the weeks following bin Laden’s death, the media would report that Obama’s approval ratings were rising; then, after about a month, they would report that the bump was gone. Bingo! That’s exactly what happened, and what was used to back it up: polls! A CNN poll found Obama’s approval rating had dropped from 54 percent in mid-May to 48 percent in early June.
Somehow, 6 percent of Americans approved of President Obama one day, and disapproved the next. Are we to presume the electorate changes their opinion of the president that quickly? You would think that people are solidified in their opinion of President Obama and every other politician on the national stage.
Polls serve a purpose. They can help us in seeing how the political winds are shifting. But let’s not use them as the end all be all of decision-making. If you do, you’re going to be one confused individual.