May 12th, 2009: Cell Phone only Households outnumber Landline only Households

The CDC has been in the news of late in regards to the swine flu (sorry the H1N1). The polling world though is very grateful to the CDC though for a survey they do twice a year. They do this survey using in-person interviews, as opposed to telephone interviews, which is not really done today elsewhere. The CDC does this because they cannot take the chance of the cell phone problem that all the typically polling organizations do. The cell phone problem is that pollsters can’t call cell phones like they do with landlines.
 

The results of the CDC poll make sense on the face of it. The number of households who only use a cell phone went up, and the number of households that only use a landline went down. The amazing thing is that for the first time, the percentage of cell phone only households (20.2%) outnumbers the number of households with only a landline (17.4%). Look at the graph below

The trend is clear. So is the problem for pollsters. They have been contacting people via landlines for basically 50 years without any huge mistakes. Still, the fault lines are breaking. In the 2008, there were a few mistakes. One can not forget the horrible call made by pollsters after the New Hampshire Primary. While I don’t believe the cell phone problem was the whole reason why pollsters got it wrong, it certainly was a factor.  I think of this problem like a Jenga game. Pollsters are able to fix the problems in the sampling by using weighting. As that Cell Phone only number goes up, they are pulling out more bricks in the Jenga game. I hope that we are able to fix the problem without making a huge mistake like the 1948 election, but the history of polling says we won’t change until a bill egg is thrown on our faces.