Polls are both art and science

Did you see the latest poll?

No doubt you’ve heard that sentence several times over the course of this election season. It seems that every day there is a new poll showing this candidate ahead or that one behind.

But what do these polls really tell us?

There is a famous picture that shows a grinning Harry Truman holding up a newspaper with the headline “Dewey defeats Truman.” According to exit polls, the prognosticators of that election called the race for Thomas Dewey. Truman went to bed assuming he’d lost the race only to wake up to find out that he had, indeed, won.

We see similar polling contradiction in Utah. According to three different polls released this week, the Congressional District 4 race between Mia Love and Doug Owens is either 28 points apart, 9 points apart, or simply too close to call. How can a poll that is, at least in theory, about the same subject have such differing results?

Polling is far from an exact science. As Columbia University Sociology Professor, Herb Gans notes “polls are answers to questions rather than opinions.”

Polls measure the answers to a very specific and highly designed set of questions. When a poll is created a great deal of time, and often money, is spent on the wording and context of the questions that will be asked.

A smart campaign pollster can create questions that are expressly designed to steer the person answering the poll toward a very specific outcome.

Even when the pollster is seeking an unbaised result, the questions themselves are inherently biased, even if only subconsciously, by the preconceived ideas of those writing the questions.

Now, this isn’t to say that polls are bad.

Here in Utah we have an abysmal voter turnout rate, some of the lowest numbers in the whole country. There is a lot of apathy regarding elections. Lack of competitive races is a significant factor. Polls draw attention to a race and typically, the more attention any particular race gets, the higher the voter turnout.

But just as polls can increase our interest we need to be cautious that the results of a poll don’t drive us away from the voting booth, by assuming the results are a forgone conclusion and that our vote won’t matter.

Polls are just one facet of elections and while they may give some insight, they should not be seen as a definitive prediction of an election outcome.

And as we’re seeing right now, if you don’t like the results of a poll, wait a few days and you’re likely to see a different result.

The Better UTAH Beat airs Tuesday afternoons on KVNU’s For the People. Podcasts of previous episodes are available here.

Leave a Comment