Why I'm Ignoring the Polls This Year (Hint: Cell Phones)
I took an informal poll of my friends last week: "How many of you have landlines?" Out of a group of ten late-twentysomethings, four raised their hands. If they had been early-twentysomethings, I'm sure that number would've been lower still. But it wasn't always this way: back during the initial internet boom of the mid-nineties, the online explosion actually proved a huge boon to wired telephony, with the number of lines rising nearly 24 percent from 142.4 million in 1992 to 186.6 million in 1999. But this was because so many people were using dial-up internet, and hooked up dedicated phone lines just so they could "chat with buddies" and "surf the web" on AOL, Compuserv, et al.
But since 2000, landline usage has dropped off the edge of the planet, back down to pre-1991 levels. Broadband killed dial-up, which in turn killed (well, hurt) landlines -- nowadays a cell phone is considered a must-have by many, and conversely, a landline is considered an extra. (The only reason I have a landline in my house is because we have a security system, which requires a wired phone line to call the cops if someone tries to break in. I don't know what's more 20th century: the fact that my security system uses a landline, or the idea that criminals would rather steal stuff from me actually-physically rather than digitally-virtually.)
So far, I'm betting none of this is a surprise. Neither is the demographic breakdown of landline-vs.-cell userhood:
"¢ 25% of people 18-25 don't have a landline.
"¢ Landline-havers tend to own their own homes, and tend to be older and less mobile (no pun intended).
"¢ Cellphone-only folks are more likely to live in cities and rent their homes.
But here's the thing that took me by surprise: even though as many as 16% of American households are cellphone-only (and possibly 25% by the end of this year), most pollsters don't call cellphones. And that means every one of the Obama-vs.-McCain horserace statistics you'll see over the next 6-8 weeks will be flawed.
There are a couple of reasons why pollsters usually don't call cells: firstly, because they're not allowed to use autodialers to do so; they've got to punch in your cell number by hand -- and that's more expensive. Also, cell users are harder to reach; they tend to ignore calls from unfamiliar numbers, might be driving or in the middle of something when they're called, and because they're often being charged by the minute, many aren't willing to participate in lengthy telephone surveys.
So what kind of impact could this have on poll results? Here's what Salon had to say:
This year, the increasingly inexcusable failure to count a growing pool of voters could prove mathematically embarrassing. Let's say that with the campaigns' increased focus on the Web, Facebook, phone-texting and other targeted ways to communicate to younger Americans, voter turnout rises and this cellphone-only universe climbs from under 10 percent of the electorate to something closer to 20 percent. If these voters' preference is 60-40 for Obama, they alone will increase his national total by 2 percentage points. And those could easily be conservative projections. In fact, Gallup Poll results from earlier this year (prior to Obama's designation as the presumptive Democratic nominee) had a 4-point swing in favor of Obama once cellphone-only respondents were folded into the overall sample.
After 2000 can any public or private polling organization willingly use a sampling methodology that understates a candidate's support by 4 points, or more than 3 million voters?
And here's the funny thing (in case you're not already laughing): our history-minded readers will remember Harry Truman's famous upset victory over Thomas Dewey in the 1948 presidential election -- the newspapers had called the election for Dewey before all the votes had been counted because they were relying on poll data. Poll data collected via landlines. Back in 1948, it seems, Democrats were also less likely than Republicans to have landlines -- or any phone lines at all for that matter -- so they were underrepresented in the polls.