Friday, October 1, 2010

So, Which Polls Should We Trust? (Hint: None of Them)

Morgan Griffith at his best (right)>

Polls are both a blessing and a bane during election season, which, like professional sports, seems to have no beginning or end.

A great example of this Jekyl-Hyde duality is the SurveyUSA poll yesterday that had the despicable Morgan Griffith (boooooo!) trailing the intelligent and effective Rick Boucher (yaaaaaaaaaaay!) in Virginia's 9th District by 15 points, 53-38. Meanwhile, Tom Perriollo (hooooray!), has been trailing his Republican opponent Robert Hurt (hisssssssss!) over in the 5th District Congressional race by as much as 46 points recently, an assertion that has brought the Perriollo camp to the brink of apoplexy, especially when a conservative group's poll has the race at about four points' difference.

Griffith has been using the "He's Nancy Polosi and Barack Obama rolled into one" tactic, which is simply preposterous, and he's also "blaming" Boucher for support of cap and trade, which was the right vote. The problem with Griffith thus far, it appears, is, as one of the old boys from the coal fields might put it, is, "You ain't from around here, are you?" Griffith doesn't live in the 9th, but as a pal o' mine says--using the Palin reference every time she gets the chance--"Griffith can see the 9th District from his house." I said months ago that I'd be surprised if Griffith got 40 percent of the vote (the same 40 percent that votes "no" to everything).

Perriello is on familiar turf looking at an opponent's backside. He trailed Virgil Goode two years ago up until just about the last week, but this time his opponent is not taking incumbency for granted because he doesn't have any. Goode had been routinely elected by wide margins for years, but Goode faced the recognition of his incompetency in a year when people finally figured out that George Bush had screwed up just about as completely as a human being can and that his policies were spelled "ruination."

According to, pollster Pete Brodnitz, "SurveyUSA’s polls have been a consistent outlier in the race for Virginia’s 5th Congressional District." Perriello's people have called the numbers "mathematically impossible." quoted a Democratic Congressional Committee spokesman as saying,“High-quality, reliable polling done by Democrats and Republicans have shown this race to be neck-and-neck ..."

The Virginia Board of Elections notes that far fewer than one percent of voters have cast ballots early, but Survey USA reportedly says that number is seven percent, a significant difference.

In any case, wherever you go, it looks like Perriello is trailing an empty political suit, a situation that is unacceptable. Perriello is the very definition of what a good representative should be.

1 comment:

  1. Dan, I conduct marketing research for a living, and I can absolutely aver that what you say is true - you can trust *none* of the polls that are conducted for political purposes. The "pollsters" are typically volunteers, and their sample is no more random than David Bowers' Rolodex. Their purposes are not to inform, but to sway.

    Enjoyed, as always! Best,