Steve Says

Tonight I covered a story about a growing consumer trend that may start to effect the validity of political polls- cell-phone-only-callers.

Right now between 14 and 22 percent of Americans have only, or use only a cell phone as their means of communicating.  When political polls are conducted via telephone, pollsters traditionally only call landline numbers. They do so because federal law prohibits them from auto dialing cell phones (auto dialing is the cheaper faster way to poll).  If pollsters want to call cell phones they have to hand dial the numbers at a cost that is up to three times more expensive than calling a landline.  Why?

- Cell phone customers pay for their minutes and therefore have to be reimbursed for them.

- Cell phone users have a higher rate of refusal (how busy are you when you answer your cell)

- Cell phone users have a higher rate of error because they are usually doing two or more things at once.

- Cell phone users are harder to track geographically (remember you can port your NY number to Dallas if you move here)

Ok.. so they're harder to reach- so? 

This all potentially becomes a problem when you consider that the majority of cell-phone-only users are male, are minorities and support Barack Obama.  HOWEVER these same consumers are also LESS likely to be registered to vote, and if they ARE registered they are less likely to CAST A vote in the upcoming election.

In an attempt to make up for this "missing" demographic statisticians "weight" polls (counting some answers more than once to make up for under represented groups) and say it all comes out in the wash. 

Considering that "weighting" increases error in data, what do YOU think? 

Do we need to start looking at ways to include more cell-phone-only users in political polls or are the results good and accurate as they are?

Have at it: