Posted July 9, 2009 | Leave a comment
Of politics, polling and pollsters
It's easy to tell when campaign season gets rolling here in Virginia. Signs clutter the highways, political e-mails back up like leaves in the storm drain and more polls than you can shake a stick at.
I've heard it said that newspapers shouldn't report polls, since they're only a snapshot of public opinion and there's always a chance they could be biased or just plain wrong. There's some merit in that position.
The Associated Press is very choosy about the polls it reports on, at least according to this story. There's merit to that position. If you only report on the polls that your organization has complete control over, the chance of some outside bias making it into the polls are much less.
That being said, polling is expensive. An editor and I talked to a polling company once about commissioning a survey for one of our congressional districts. I wasn't privy to the phone call to the pollster, but I was told that it included the phrase "How much?!"
Here in the cheap seats, we've always been inclined to watch as many polls as possible Statistically speaking, the more data points are involved, the more accurate the results.
So to a large extent, we watch publicly available polls to help get a read on which way the wind is blowing in the Shenandoah Valley and the rest of Virginia. Generally speaking, there are four pollsters that I watch with some regularity:
• Rasmussen Reports: These guys are polling machines, churning out data on a daily basis. They're also good at building long-term trends for races in Virginia. Scott Rasmussen is very transparent about methodology and questions.
• Public Policy Polling: They might be employed by Democrats more often than not, but everything I've seen about this outfit gives me confidence. They're transparent with methodology, and throw the cross-tabs out for the world to see. Tom Jensen also blogs the results and cuts through the clutter well.
• SurveyUSA: I've been watching this polling company since I started this gig in 2005. I started paying close attention to them when they made me look like an idiot by predicting a sizable win by Democratic Lt. Gov. Tim Kaine over Republican Attorney General Jerry Kilgore. And we all know that can't happen. Right?
• Mason Dixon: The only reason you don't see Brad Coker's numbers in our paper more often is because of Mason Dixon's unfortunate habit of releasing results for publication on Sunday, the only day we don't publish. Mason-Dixon is transparent with all their information, and they have the best geographic handle on Virginia of any pollster I follow.
Bottom line, they're all transparent in their methods and results, they have a pretty good track record for accuracy and none of their head honchos are afraid to get on the phone with a small-town paper and answer questions about their work.
Occasionally, though, we will grab a poll from someone who polls Virginia less often: like Suffolk University in Massachusetts, or Christopher Newport University here in Virginia.
Taking a lower-frequency poll has bitten us before, like this Majority Watch survey in 2006 that showed the 10th Congressional District to be a 5 point race. In reality, Rep. Frank Wolf won re-election by 16 points.
Is the poll from a campaign, pitching only one or two data points, not their entire results? No thanks. If we can't see the entire poll, then it's probably just spin from a campaign looking to improve their image.
Bottom line, if the poll isn't transparent, it's probably not going to be mentioned in our pages. At least until the funding for the Cheap Seats Institute for Public Opinion and Awesome Stuff Research comes through...
Copyright © The Northern Virginia Daily | nvdaily.com | 152 N. Holliday St., Strasburg, Va. 22657 | (800) 296-5137