Landlines, cell phones, and polls
Why my opinion has never mattered…
Here’s something nobody ever talks about, but nevertheless seems like it should be very important: Public opinion polls are almost all done by calling landlines. Do you own a landline? I don’t. In fact, I’d be hard pressed to name a friend in my generation who has one.
The demographic implications are pretty obvious – landline polls are likely to skew heavily toward older populations. Comparative studies show that landlines also favor females and whites. (So just replace the word “Americans” in poll results with “Elderly white women” and you’ll get a better picture)
This is a tough issue to get around, and understandably, polling companies have been quick to downplay its significance. Pew says that mobile-only and landline-only polls produce “virtually identical” results. But the same study they used to draw that conclusion provided the numbers that demonstrated the demographic gap described above. Mobile- and landline-based populations will likely converge on some issues, but certainly not all. Politically, it would seem to follow that polls err to the right.
In my view, the increasing scarcity of landlines poses a major obstacle to an already unreliable service; pollsters already have to worry about people without any phones, people who don’t answer phones, and people who aren’t sincere or who don’t understand the questions. Now, with some estimates suggesting that half of adults 30 years old and younger use only cell phones, we can be fairly certain that our voices just aren’t being represented.
It’s unlikely that polls will go away, however. I can’t imagine what the media and anyone with an agenda would do without them (and their infinite, and rather convenient, layers of interpretation).
On the bright side, polls are no more misleading than other mainstay methods of generalizing the “general will” (hat tip to Rousseau), such as, say, voting. The voting demographic is as small and unrepresentative as any, yet we don’t exactly stop political commentators mid-sentence and protest, “Well, technically, New Yorkers didn’t elect Senator Schumer, a relatively small number of voters in New York elected him.”
In other words, whether we’re talking about polls, elections, workers, readers, or viewers, we’re likely to be generalizing the views of a smaller subset onto a much larger population. We’re often reminded after references to the beloved “town hall” government of our Forefathers that civic participation was actually quite limited. Things may not have changed as much as we might like.
UPDATE – I’ve received some rebukes and note here that Gallup, for its part, claims that it’s known about this for years and is continually fixing the issue and including cell phones in their polls. This may be true, but if it wasn’t still a problem, they wouldn’t be calling for so much research on the topic. Someone call me and poll me so I can get over my skepticism.
-Colin
Related posts:
ShareComments
2 Responses to “Landlines, cell phones, and polls”
Leave a Reply
Colin,
You and I played a few jobs together about 8 or 9 years ago when you lived in St. Louis. I hope you are doing well and still playing drums.
Also sharing a passion for politics, I recently discovered your opinion pieces. Even though my political paradigm is the antithesis of yours, and I disagree with just about everything you write, I love reading your blog.
It seems like anytime the polls are negative for one political party, members of that persuasion diminish the surveys. My side does it too. But the fact is if elderly women did not vote in this country, Democrates would have a tough time getting elected. Especially when they are picked up from the assisted living facility by some left wing activist, bussed to the nearest polling facility, and told to pull the lever for the Democrat or some Republican is going to take away their medicine.
For the most part, I think the polls have been very accurate over the past few election cycles. We knew at least a week ahead of time that Scott Brown was going to feed the Democrates a big shit sandwich. I think most polls were within a point or two.
By the way, I received a polling call on my cell phone, but I could not unhook it from the gun rack in my pick up truck fast enough. Hey, wait a minute, maybe you do have a point!
Just trying to be light, I really do wish you future success.
Hey John! It’s been a while since our old-timer swing band days… Glad to hear you enjoy the blog, despite our apparent differences.
I’ve received a pretty solid amount of disagreement on this post, mostly on whether pollsters do a good job of including cell phones or achieving a representative sample. I wasn’t making a partisan point here (at least not intentionally), other than to point out that if cell phones are indeed underrepresented, that would likely tilt polls to the right. I don’t think there is much controversy there.
Ultimately, I am still pretty skeptical of polls in general – perhaps not so much because of their (in)ability to reflect what people actually think, but because often what people actually think is blatantly wrong or a result of ignorance on an issue.
Winston Churchill once said that the best argument against democracy is a 5-minute conversation with a voter.
(I know, I’m such a liberal elitist!)