Question: I’m calling from a national research firm. We’re conducting a survey and I’d like to ask you a few questions about issues that affect people here in Wisconsin. This will only take a few minutes of your time and no one will try to sell you anything as a result of this survey. May I speak with a male/female 18 years or older who lives at this address and is now at home?
Answer: What address? This address? The laundromat? I’m folding a load of tidy-whities, dude. Just a minute… anyone here want to talk to some survey weasel? Sorry duder. Everyone’s busy, and nobody’s home. Did I tell you it’s the laundromat?
The above exchange or one like it would take place 460.3 times out of every thousand polling contacts if polling companies used lists that included all cell phone numbers and landlines. Household polling is no longer meaningful. Nobody’s home except my dad and he had a stroke.
In 2008, a study found that polling results including cell phone users leaned about 5% more in a progressive direction than polls relying on landline contact alone. Public polling is a black art, the goal of which is to trend toward the midline of all other poll results while ultimately predicting more “wins” than the competition. Public polling of political races is a profit center for polling companies. They want to get it right, and to get it right they want to do it like the competition does it.
The results this year should shake up the industry, since the polls themselves seemed based as much on media noise as on genuine public opinion and are skewed far from the most likely outcomes. When Russ Feingold wins his Wisconsin senate seat, the polling companies will have to congratulate him on a big come-from-behind victory. In fact, the results have been only in doubt due to the noise in the system from the propagandists, noise that doesn’t much influence voters, but that creates the static impression paid for by the propagandists.