Network News

X My Profile
View More Activity

Virginia: Pollster Pushback

Eager to tamp down a bubbling storyline that Terry McAuliffe has a double-digit lead in the three-way contest for the Democratic nomination for Virginia Governor, former delegate Brian Moran's campaign today released its own poll showing the two locked in a very close race.

Campaigns poll a lot, but release their internal data infrequently, usually in an effort to nudge coverage their way or attract support. The move comes on the heels of a second publicly-released poll putting McAuliffe, former chair of the Democratic National Committee, atop the field.

Those two polls - one by SurveyUSA, and the new one by PublicPolicyPolling (PPP) - use automated telephone calls to gather data, a newer methodology that remains controversial. (The Washington Post does not typically report these types of polls.)

The poll commissioned by the Moran campaign, was conducted Al Quinlan of Greenberg Quinlan Rosner Research (GQRR), using live telephone interviewers. While SurveyUSA and PPP disclose much of their collected data online, GQRR's release was substantially more sparse, limited to a few results favorable to their client.

The polling hubbub masks a central point: polling primaries is difficult, particularly this one. This is the first contested Democratic gubernatorial primary in years, Virginia has no party registration so any registered voter can opt to vote in the June race and Barack Obama's reshaping of the electorate creates new unknowns.

Just what will turnout be? What types of voters will come out on Election Day?
Will turnout be similar to the 2006 Democratic primary for Senate, where about 150,000 participated, or somewhere closer to the nearly 1,000,000 who voted in the Democratic presidential primary last year? No one expects the higher number, but estimates vary widely, and matter, perhaps as much polling methodology.

The Post will explore each of these questions and its polling team will describe how the paper decides which polls to report -- and which to ignore -- as the June 9 primary approaches.

And one commonality in all the data now in the public domain is that there are high numbers of undecided voters, meaning that the contest is fluid, no matter what the current horse-race standings may be

By Jon Cohen  |  May 5, 2009; 6:19 PM ET
Save & Share:  Send E-mail   Facebook   Twitter   Digg   Yahoo Buzz   StumbleUpon   Technorati   Google Buzz   Previous: On Gay Marriage, A Near Even Split
Next: Obama at Notre Dame: Catholics' Take


I don't know enough about polling to know what methodology is most accurate, but I do know enough about politics to know that with the level of undecided voters and with all three candidates loaded with cash, there will be a barage of TV ads which will move these numbers considerably as we get closer to the election.

It would be helpful in evaluating these polls to know when the candidates went on TV in the various markets. For example, Creigh Deeds just went on TV in the last week. All three of these polls were taken before Deeds launched his ads. Meanwhile, McAuliffe has been on TV for weeks. I don't know when or if Moran has gone on TV.

That kind of context would be helpful in evaluating these snapshots.

Posted by: fakevirginia | May 5, 2009 11:49 PM | Report abuse

A new Rasmussen Reports national telephone survey found that 40% would vote for their district’s Republican candidate while 39% would choose the Democrat. Thirty-one percent (31%) of conservative Democrats said they would vote for their district’s Republican candidate. This after being down 14 points in a June 2008 poll. That's a 15 point swing.

More impressive, Republicans have not won this poll in five years. The tide is turning, dramatically. Rassmussen is the only pollster to get the last two election percentages exactly right. Fordham University recently released a report on 23 pollsters naming Rassmussen the most accurate in America in 2008's election.

Fordham's report showed that in 2008 17 of 23 (74%) pollsters reported that Obama was more popular than the votes showed. Can you have a clearer picture of pollsters trying to sway an election than that? And no, they are not incompetent. They are the richest media in America and have no excuse for inaccuracy other than it is intentional. The bottom five for accuracy: CBS; Gallup; Reuters/C-SPAN/Zogby; CBS/New York Times; and finally dead last, Newsweek.

Posted by: BothSides | May 6, 2009 11:07 AM | Report abuse

The comments to this entry are closed.

RSS Feed
Subscribe to The Post

© 2010 The Washington Post Company