The Iowa caucus, the first nominating contest of the year, has finally come and gone (at least by print time, anyways). And, gratefully, the caucus and upcoming primaries will have taken with them the scourge of pre-election polls. Finally, for the first time since this election cycle began last April, Americans will have a true sense of which politician has the lead in this year’s battle for the nomination.
The issue here is that polls, especially the kind of pre-election polls that have dominated the news cycle for nearly a year now, are usually terrible. Whether they lack accuracy or predictive ability, pre-election polls are very good at giving a false sense of what’s actually going on.
First and foremost, national primary polls are just pure entertainment. If there was a national primary, then the fact that Donald Trump leads the rest of the GOP contenders, often by a margin of 2-1 or greater, would actually be important because it would spell certain victory for Trump.
The only problem is that there is no national primary.
What matters instead are the state primaries, all 50 of them. The lead that Donald Trump held over the once No. 2 contender, Texas Sen. Ted Cruz, was in the single digits in Iowa. Yet after the dust of the caucus was settled, Cruz came out on top. However, things don’t look so good for Mr. Cruz in New Hampshire, where Trump still leads him and the rest of the field by 22 points, according to the RealClearPolitics polling average.
Add on to that the fact that not all polls are created equal, and suddenly the horse race begins to lose its meaning. Take, for instance, the Internet straw poll. After every single debate so far, whichever outlet ran the debate took to Facebook or their own websites to conduct a straw poll to find out who won. The problem here is that there is no scientific basis for this poll and the results are often skewed beyond usability.
While Bernie Sanders may or may not have won whichever debate, we will never truly know how the public felt about it because ardent Sanders supporters, who are more likely than not to perceive a Sanders victory regardless of his performance, will quickly flood Facebook or Reddit with their assertions of his victory. Once that happens, the results are skewed and the opinion of actual undecided voters, who often actually care who wins a debate, is buried.
That’s not to say there aren’t good polls or that polls can’t be predictive. Established pollsters like Gallup and the Pew Research Center and even new internet pollsters like YouGov pay close attention to methodology and try their hardest to minimize sampling errors and other kinds of bias. Being even more specific, pollsters like the Des Moines Register’s Ann Selzer are held up as “the gold standard” of polling, according to both the Register and industry professionals, for a remarkable string of historically accurate polls. It is a cruel irony, however, that Ann Selzer was proven wrong on caucus day as Cruz pulled ahead of Trump — a reversal of her own prediction.
We might also take solace in the fact that as the various caucuses and primaries draw closer, the polls become more accurate and more predictive. To look at a poll from June of last year as a real measure of who will win the election is really a silly idea. But to look at Saturday’s poll out of the Des Moines Register is a relatively sound notion.
Even so, polls just aren’t always right and they shouldn’t be used, under any circumstances, for any kind of voting decisions. For a historical example, turn to the famous 1948 Chicago Daily Tribune headline, “Dewey defeats Truman,” which incorrectly called the victor of that year’s presidential race. For a more recent example, take Mitt Romney’s 2012 loss to the incumbent Barack Obama — a loss no one save a few outliers predicted.
With that in mind, it is evident that polls can be useful for tapping the pulse of the American public, if and only if, those polls are solid in their method and taken close to the actual contest. Look for how the poll was conducted, who did the surveying and how many people remain undecided and
Despite the fun in watching the political horse race, be sure to look past polls when you make your final decision at the caucus. Use policy, review platforms and really get to know what a candidate wants to do in the Oval Office before you cast a vote their direction. Polls will never tell who’s actually qualified for the job.
The editorial staff can be reached at firstname.lastname@example.org and on Twitter @TheSagebrush.