IN POLLING VOTERS, DUBIOUS NUMBERS ABOUND
Election prediction business offers stats typically most trusted by those who agree with the result
Now that the somewhat-silly season of primaries and runoffs — which somehow lasts longer than the NBA playoffs — has ended, there’s one sure thing leading up to the November general election: A surge in the number of polls proffering a look inside voters’ minds.
Like any good politician or Volkswagen executive, though, most of us only believe studies that support our already held positions and beliefs.
Therefore, it wasn’t exactly a sigh of relief for Dems when an end-of-May Quinnipiac poll showed Republican U.S. Sen. Ted Cruz with an 11-point lead over upstart Democratic fatted calf Beto O’Rourke, deflating an April poll that called the race “too close to call.” More of an “I told you so.”
The same poll showed Gov. Greg Abbott opening up a 53-34 percent lead over his Democratic opponent, Lupe Valdez, the only surprise in the result being how Valdez attracted 34 percent of responders.
Polling in the United States (and apparently in Britain, where Brexit caught people by surprise) seems these days to be less of a science and more akin to phrenology. Researchers into the decline in reliability of such surveys say the migration to cell phones, with their caller ID and ease of blocking unwanted callers, has been a large factor.
Well, sure. The only people who answer unknown callers now are those with really good credit or who don’t mind robocalls about student loan forgiveness. That surely skews results.
According to a just-released study by the National Center for Health Statistics, 53.9 percent of U.S. households rely entirely on cellphones, abandoning landlines at a quick pace. The government, smartly, does its polling in face-toface interviews.
Looking anecdotically, in lieu of conducting a survey about it, there definitely seems to be a growing distrust in the polls that inundate our lives each election season. The Pew Research Center, a self-described “fact tank” that has conducted surveys since the mid-1990s, felt compelled in May to release a six-and-a-half-minute video defending its raison d’etre.
Talking heads in the video insist that polling is more accurate these days, comparing modern presidential election polling (2 percent margin of error) with those conducted in 1936 (12 percent margin of error, perhaps attributable to those giving up their home telegraphs).
“It’s also important to remember that election polls are just one kind of poll and that they’re not the best barometer for the accuracy of polling in general,” wrote Courtney Kennedy, director of survey research for Pew in the text accompanying the video. “Because an election poll has an extra hurdle to jump: It not only has to measure public opinion, it also has to predict which of the people interviewed are going to vote and how they will vote — a notoriously difficult task.”
“So, yes, we can still trust polls. But it’s important to be realistic about the precision they can provide.”
A lack of precision doesn’t stop sometimes contradictory headlines, such as, “The polls are swinging to the Democrats,” or “The weakening blue wave.” Nor does it keep bloviating TV commentators from predicting the political direction of the country. It just reduces credibility of such to the level of a sports radio personality predicting who will win this year’s Super Bowl (faultless prediction: Not the Texans).
Over the past few years, a number of organizations and websites have tried to take the pulse of voters through aggregation, not exactly the most scientific of methods. Sites like RealClearPolitics and FiveThirtyEight simply round up all these problematic polls and average them together.
Nate Silver, the founder of FiveThirtyEight, recently launched his own impassioned defense of polls.
“The media narrative that polling accuracy has taken a nosedive is mostly (expletive) …,” he wrote on his website.
“Polls were never as good as the media assumed they were before 2016 — and they aren’t nearly as bad as the media seems to assume they are now. In reality, not that much has changed.”
So, a quick poll of those who rely on polling for their operations suggest that polls are just as accurate — or inaccurate — as they’ve ever been. No reason to doubt their conclusions.
Polls are a lousy way to look at important things like elections or the national ardor. But they’re really the only tool available.
As long as they’re kept in proper perspective, they can shed some light, but that’s always going to be in a pace filled with dark corners and shadows cast by a dim bulb.
Accuracy may not be the main issue. Most of the time, polls will be either dismissed as unbelievable or serve as an echo chamber of our own thoughts. Sometimes, we just believe what we want to believe, especially if the subject ignites our passion.
Such as in Texas, where RealClearPolitics has averaged its available polls to determine the standings in our current Senate race and finds Cruz to have a nine-point lead over O’Rourke. Sounds about right. Must be accurate.