Skip to content
Link copied to clipboard

Commentary: Dear pollsters, shaddup

By Eddie Chambers It's time to put pollsters under the spotlight. Prior to the election, the prevailing view of the news media was that Hillary Clinton would win. As we now know, Donald Trump was able to rack up his Electoral College votes with significantly greater ease than many had imagined. What we witnessed on the night of the election was a lesson in the unreliability of polls.

By Eddie Chambers

It's time to put pollsters under the spotlight.

Prior to the election, the prevailing view of the news media was that Hillary Clinton would win. As we now know, Donald Trump was able to rack up his Electoral College votes with significantly greater ease than many had imagined. What we witnessed on the night of the election was a lesson in the unreliability of polls.

Instead of waiting for the electorate to cast their votes and for whatever result might come, the media seemed more intent on peddling the quack science of polling in an attempt to predict elections or predict the ways in which the electoral mood might be swaying.

But time after time, polling has been exposed as a sham and deeply flawed means by which the electoral mood can be judged. There is simply no credible or justifiable method of second-guessing an election. Nor should there be. Rather, it should be for the candidates or for the political parties to advance their arguments and for the ballot booth chips to fall where they may.

A sizable industry has grown up around polling, reflecting perhaps the lucrative nature of the business. The science of the industry is further compounded by pollsters' attempts at legitimacy in the form of plus/minus margins of error.

Coming from Britain, I know the wave of pollsters getting things spectacularly wrong began as far back as 1992, when John Major "unexpectedly" led the Conservatives to victory. In fact, the most recent general elections in Britain have been called incorrectly by pollsters - this after weeks and months of febrile reporting of this or that shift in the polls.

But the ultimate failing is the very existence of polling itself. The most recent failing of U.K. pollsters was their inability to successfully call or predict the Brexit referendum vote. Here in the United States, it was the latest election.

It's time for the media to dispense with its reliance on polling and simply accept, as all of us must, the results of elections, as and when they take place.

This may be anecdotal, but ask yourself the question: When have you ever been polled? Or, who are these people who supposedly make up pollsters' samples? It seems that polling has the appearance of something we suspect relates to or refers to other people and not ourselves.

Politicians and opposing camps are, of course, free to commission their own private polls, though this seems to be somewhat akin to employing a carnival or circus fortune teller. What we should not have to put up with is the reporting of polls by the news media that commission the very polls they then regard as news. The world has plenty of real news to report without the need for the mainstream media to manufacture their own news in the form of the results of polls they have commissioned or consulted.

A number of countries maintain election silence in the day or days leading up to an election, so for 24 or 48 hours at least, the electorate are spared the sorts of phony news and pollster reporting that elections often generate.

Rather than a reliance on polls for evidence of dramatic or perceptible shifts in the electoral mood, the media would better serve us by reporting on the substance of whatever policies candidates or political parties might be advancing. It is within this context that the media should focus more of their attention on critically examining whatever policies are proposed. This, rather than reporting of polls, would help the electorate enormously. We need to see an end to the media's irrational and self-serving reliance on polling and the results of polls as a stand-in for real news.

It might be something of a cliché, but the only poll that matters is the one carried out by the nation as a whole on Election Day. All other polls should be treated with skepticism. With few people, if any, openly declaring themselves as susceptible to the influence of polls, we all should question their usefulness or validity. At the very least, the news media can do a better job and stop relying on polling.

Eddie Chambers is a professor of art and art history at the University of Texas at Austin. eddiechambers@austin.utexas.edu