Friday, July 25, 2008

Election Pollsters Making the Wrong Phone Call

A long time ago, I had a teacher who sometimes asked us to compose newspaper headlines for logic problems. His point was that one way to solve problems is by looking at them through the lens of a different paradigm.

The professor and his headlines exercise came back to me this morning as I pondered the 2008 election polls showing a close race between McCain and Obama -- so contrary to everything I see around me. (Admittedly, I live in California. But despite the notion that California exists in an alternate universe, it's also the state that elected Ronald Reagan to his first political office.)

These days I don't hear anybody -- even people I know will probably vote for him -- expressing any positive interest in McCain. But everybody is talking about Obama, including people who've probably never pulled the lever for a Democrat but are Obama supporters. 

Yet poll numbers show voters almost evenly divided between the McCain and Obama. But let's turn the problem on its head and, instead of the end point -- the answers to polling questions -- look at the process for getting to that end point. Then another explanation for the result percolates up.

Election polls reflect the pre-Internet -- or, rather  pre-Internets -- communication perspective of middle-aged political pundits and pollsters. One that assumes that everybody has a wired "home phone" on the PSTN. That's the Aha! here: Pollsters are only calling people who answer listed phone numbers supplied by the "phone company."

Do any of these pollsters and pundits have children? Or friends under 35? 

A lot of people aren't reachable on "home phones" anymore. Like my 17 year-old, they use cell phones almost exclusively. Many don't have wired phones at all. For a growing number, the "home phone" is an VoIP phone service. None of these services are listed in the "white pages."

I'm betting on lots of egg-on-pundit-faces in November. 

1 comment:

Anonymous said...

You are so right- I found a study done by the National Inst. of Health for July-Dec 2007, 15% of households are cell only, and of those 90% are under 44. They are so missing a large voting block. It's here
and the demographics are at table 1.