Maybe, maybe not

America Wants to Know is going to go out on a limb here and predict that in tomorrow's election, the Republicans will hold their majorities in both the House and the Senate.

The polls certainly don't show that.

But take out your phone bills and count the services that did not exist just a few elections ago. Caller ID. Cell phones. Blackberries. Voice mail.

Pollsters take lists of phone numbers and call them and ask the people who answer how they're going to vote.

They use scientific sampling techniques to select the numbers they will call.

But what happens to a poll's accuracy when a significant number of people don't answer their phone or tell the pollsters they don't have time to take a survey?

Are the polls skewed to include a disproportionately high number of bored retirees and a disproportionately low number of frantic working parents?

Do the pollsters talk to non-citizens who don't admit to being non-citizens when some stranger calls on the phone?

Do the pollsters take into account the high number of people (every workplace has them) who constantly voice their dissatisfaction but don't ever really do anything about it? ("Oh, was the election this Tuesday?")

If the Democrats fail to recapture the House, the candidates and the leadership may not be to blame for failing to turn out all those voters who said they were planning to vote for them. It could be that in the final analysis, their majority was just a math error.


Copyright 2006

.