July 09, 2005
No More Polls?
As Kevin notes, there are now more cell phone subscribers than landline subscribers in the US. My girlfriend and I are a good example -- two cells, no landline. The question, then, is how long before this starts violently skewing poll results. Pollsters are legally barred from calling cell phones. Cell phone users, to some degree or another, make up a different demographic profile than the rest of the country (skewed young and economically mobile), and may have different political opinions than the land users. This got a lot of attention in 2004 but, in the end, the polls turned out almost exactly right (indeed, those who harp on the exit polling forget that nearly every poll in the country got the results within the margin of error). As the country switches to cell phones, though, that won't last forever. So when's the tipping point?
TrackBack URL for this entry:
Listed below are links to weblogs that reference No More Polls?:
» Cellphone Numbers Overtake Land Lines from Outside The Beltway
Cellphone Numbers Overtake Land Lines (LAT) Hold the phone. The number of mobile-phone users in the U.S. surpassed the number of conventional land-based phone lines in the second half of 2004, the government said Friday. By the end of the year, ... [Read More]
Tracked on Jul 9, 2005 4:05:04 PM
» Some Call It A Bonfire/Carnival Of Classiness... from WILLisms.com
We call it "Classiness, All Around Us." Click to explore more WILLisms.com. In no particular order, WILLisms.com presents classiness from the blogosphere (now with 50% more classy!): 1. Nancy Pelosi and Eminent Domain- Vodka Pundit (via Daly Thoughts) ... [Read More]
Tracked on Jul 12, 2005 7:07:44 PM
At some point, pollsters will no longer be legally barred from calling your cell phone.
Step one is for your cell phone to be listed in the phone book by default, which will probably happen pretty soon. After that, allowing direct marketing calls to your cell phone will probably come pretty quickly, perhaps in exchange for requiring that the first minute of all incoming calls not count towards your minutes. Or something.
Posted by: Nick Beaudrot | Jul 9, 2005 7:56:24 PM
I remember the left thought that the polls were wrong in the last election cuz they weren't calling cell phones. The left surmized, for some reason, that most cell phone owners were liberal. Just cuz someone is young and doesn't have a landline phone doesn't mean they are liberal.
Another mis-interpretation of the left was the number of registered Democrats. I have seen accusations of fraud just cuz a precinct had more registered Democrats than Republicans and Bush won that precinct.
My dad and granpa are both registered Democrats in Ohio, both voted for Bush. My dad even got a thank you call from Bill Clinton and a couple of thank you mailings from the DNC. Alot of union guys, alot of retirees, and alot of old (Zel Miller) Democrats don't really vote for the Democratic candidate anymore. But they also have not bothered to change the voting registration.
Alot of union guys couldn't stomach voting for Kerry. A couple of friends of mine work at the Akron Firestone plants. I used to tell them if Kerry won that they would be able to marry each other before long. Yeah, they used to hate it when I would remind them that their union wanted them to vote for gays and affirmative action. I also reminded them that the Dems hadn't done anything for organized labor in a long time.
They just didn't bother to vote, and who could blame them.
Posted by: Captain Toke | Jul 9, 2005 11:57:45 PM
Polls are generally weighted for different demographic characteristics, so the pollsters figure that as long as they get a few surveys from a specific demographic group, they can extrapolate for the whole class in that demo.
BUT, one could make the argument that the early adopter no landline/cell user is psychographically distrinct from the landline holder. At one time I think one could make a better argument for this. But, we're far far past the early adopter stage now.
Posted by: Roxanne | Jul 10, 2005 1:43:51 PM
I used to work for a polling company in a technical aspect and this issue has been looked at for a long time now.
However, in some ways it is less of a problem than you might think. First off, as Roxanne noted, sampling is a fairly detailed science where you make sure that your sample is as representative of your demographic as possible. Secondly, the characteristics that Ezra ascribes to cell phone only people, young and economically mobile, are virtually impossible to survey anyway as they are seldom home and usually not willing to take surveys. The fact that they don't have a home phone just means survey companies arn't wasting their time calling a home phone that isn't answered for the most part.
Anyway, the point is, that this isn't a new problem and there remain for now methods to compensate for it. over time, I expect that other methods will evolve to gauge public opinion, perhaps even including calling cell phones, although I expect that a sample controlled internet/email survey method will be the replacement.
Posted by: Dave Justus | Jul 11, 2005 3:30:49 PM
Posted by: peter.w | Sep 17, 2007 3:22:05 AM
The comments to this entry are closed.