So, you say you've noticed a poll! Maybe it was a poll in an early primary state. Maybe it was a national poll. Maybe it was a series of head-to-heads in which Hillary Clinton was paired against Marco Rubio in Florida, or Donald Trump was paired against Bernie Sanders in Wisconsin, or Martin O'Malley was paired against a candelabra in the parlor. Did you see the results of that poll? That poll had such results, man!
Now you're worried, or confused, or giddy, or your body is so stuffed with endorphins that you might never come down.
But you've been burned before! You can remember that time when the fears you felt after reading about a poll did not end up being realized. You have a distinct memory of another time a poll gave you a big old sugar high when you saw the candidate you liked at the top, and the crushing emotions to which you were reduced weeks later, after some nice Midwestern folks said, "You know what? Nah," to the earlier results that made you feel so nummy-yummy warm inside.
Why do polls do this to us? Should you believe them? Should all pollsters be pushed out to sea on an ice floe? How many multitudes are contained within the 288 people in New Hampshire called between Dec. 7-10? When angels dance on the head of a pin, what is the margin of error: the number of angels or the size of the pinhead?
Ugh. Can you calm down? Let's try all of this again.
Is election polling accurate?
Sure! It depends, though. How far away was the election date when the poll was conducted? We have some very good political science suggesting that the most accurate presidential polls are those conducted right before an election, and that the further we are from that election, the less accurate the polls get.
So, let's see -- we are what, 300 some-odd days out from the election? Okay. Election polls are basically garbage right now, sorry. Polls of the Iowa caucus and New Hampshire primary are a little better, but still not great.
Agh! You're kidding!
No, I'm not. It's always been this way and it always will be this way until that blessed day the sun goes supernova and we perish from this earth.
If polls aren't accurate, why do pollsters take them?
Well, to practice their craft, for one thing. Polling is a fairly meticulous science and pollsters spend a considerable amount of time just calibrating their instruments so they're in peak form when respondents are likely to give salient and predictive responses.
Pollsters would have no hope at accurately predicting an election if they just jumped into it late. Imagine if the cast of "Hamilton" was asked to perform the musical on the day they all first met. It would be a terrible performance, and you would feel terrible watching it, and you'd tell your friends, "'Hamilton' was terrible!" and then "Hamilton" would be a huge flop and Lin-Manuel Miranda would have to go work as a government contractor or something and all his wondrous talents would be wasted filing government procurement forms. In the real world, though, that cast got to practice their work for a period of time and it helped make it a success. The only difference is that they had the luxury of practicing privately.
Well, why does the media constantly report on the parts of polls that are just inaccurate snapshots?
Take your pick.
a) The media favors short-attention-span garbage over depth.
b) The high demand for short-attention-span garbage is how revenues are raised to fund things that aren't short-attention-span garbage.
What if the media just decided to stop publicizing polls?
What would happen in that situation is the media would be accused of suppressing information.
What's the solution?
Calming down and becoming more accepting would be a good start. Short of that, let's recognize that there's more to these early presidential polls than the top line that tells you who's "winning" at the moment. Down in the crosstabs is interesting information that can guide both pollsters and campaigns toward greater success.
I'm not convinced that polling shouldn't just be phased out, or to some extent discredited. After all, they keep getting things wrong, like the U.K. elections in 2015.
Are you British? Because if you're not, I don't see why you need to be so hung up on their elections, dude.
But, yes, it's true that was a high-profile failure. And it's also true that as technology moves people onto new platforms and away from the ones pollsters rely on to contact them, pollsters could have to make some adjustments to keep up. Sometimes, the old way of doing things breaks before the new way of doing things is perfected. It's possible that we're at some inflection point right now that's going to force the science of polling to evolve. Someone should do a poll of pollsters, I guess?
But people should calm down. There's a world of polling work beyond election polling. And pollsters are helping us explain the world every day. At The Huffington Post, we have a series of polls that help explain the fundamentals of the political environment. Public opinion on how the economy is faring, whether the country is on the right track, whether approval of Congress is trending up or down -- these are the measures of the parameters and forces of the political world that underpin all the metaphysical stuff about how Bernie Sanders might do against Rand Paul in Ohio if the election were held today. (The election is not being held today.)
Beyond that, pollsters offer us a wealth of information about the world we live in. Take for example, Pew Research Center's in-depth look at how Americans view their governing institutions or Brookings' American Values Survey, which both explore the state of the American mind on many key issues of the day.
Are you worried that the middle class is losing ground? They are! Those were some good instincts. (Did you think the middle class was doing okay? I am so, so sorry!)
You should thank a pollster for doing the work that allows you to feel confirmed in your view of the world.
And, more importantly, I'll thank you for picking up that phone and sharing your perspective with a pollster. It has helped me learn more about modern life. In the end, your view of the world is invaluable, and pollsters are among the few people in the media who actually, literally care deeply about your opinion.
So what should I do between now and the election?
That's easy: Quit smoking and wear your seat belt whenever you're in a car.
~~~~~
Jason Linkins edits "Eat The Press" for The Huffington Post and co-hosts the HuffPost politics podcast, "So, That Happened." Subscribe here. Listen to the latest episode below.