Thursday, October 31, 2013

Six Polls In Two Days: Cuccinelli Is Definitely Losing, But Pollsters Disagree On Whether He's Closing Or Fading

By far his highest profile campaign surrogate, Hillary Clinton's widespread popularity may indeed help McAuliffe to put it away next week. And more than a few DC journalists have reported that the McAuliffe '13 campaign serves the dual purpose of being a test run Hillary's old inner circle. A.P. photo.

For anyone trying to get a handle on the state of the Virginia Governor's race with just a week to go, today was not the best day to tune in. No less than SIX polling firms have released new numbers since Tuesday, all without a lot in common. Though one overiding theme is present in all of them: Ken Cuccinelli is losing.

In fact, McAuliffe has led his opponent in the last 32 consecutive polls. But agreement on who is winning is where the similarities stop. Consider the small table below, which groups all five recent surveys into one chart:

The Democrat leads by anywhere from a 4-pt to 15-pt spread, with McAuliffe ranging between 42-51% of the vote, Cuccinelli 33-41%, and Sarvis 8-12%.

So take your pick, right?

The problem for Republicans, however, is no matter how you mix and match those polling numbers, there just isn't a winning scenario for Ken Cuccinelli. McAuliffe's worst performance (42% in the Hampton poll) is still better than Cuccinelli's best (41% in the Quinnipiac poll).

Digressing back to the recent poll variation; the trajectory of the race is hard to decipher not simply because of the topline result, but because of divergent trendlines as well. The Washington Post and Roanoke College polls seem to think the ground is crumbling beneath the Republican nominee. In mid-September, McAuliffe led Cuccinelli 47-39%. That lead grew to 51-39% five weeks later. Roanoke found a similar trend, though to a greater extent. Just 3 weeks ago, the local Virginia pollster found the Democrat up against the Republican and Libertarian, 41-36-9%. He's now at 48-33-10%.

On the other hand, Quinnipiac University and Rasmussen Reports both find Cuccinelli making late gains. He cut McAuliffe's lead from 7 to 4 points in the former, and from 17 to 7 points in the latter.

As if  that weren't enough, both Hampton College and Zogby find no real change from their previous surveys one month ago.

What's the cause of all this polling discrepancy? All six pollsters were in the field in the same geographic region at approximately the same time. How can one find a near-tied race, and the other a double-digit landslide? For an explanation, take a glance at some of the particular demographic and political characteristics of the poll respondents:

Zogby and ABC/WaPo didn't provide the demographic make-up of their surveys.

There's pretty remarkable consistency between the four pollsters' demographic findings and the make-up of the 2009 Virginia Governor race. Quinnipiac and Hampton University find a more racially diverse electorate than the one that showed up four years ago, though not enough to have altered their topline results in a significant manner.

Where the pollsters do diverge decidedly, however, is with their respective partisan identification findings. As you can see in the chart above, party I.D. ranges from a two point Republican advantage to a 9 point Democratic advantage (the WaPo poll was D+7). Though none of them have found an electorate as Republican-identifying as the one that turned out in 2009.

As the Washington Posts's Scott Clements and Ben Pershing noted yesterday, it appears to be this disparity in party ID between the pollsters that's the most responsible for the disparate toplines and trendlines. And reexamining the results, this theory makes sense. It's probably not a coincidence that the pollster finding the largest Republican advantage in party ID also found the closest overall margin between Cuccinelli and McAuliffe (Quinnipiac), or that the D+9 pollster also happens to be the McAuliffe +15 pollster (Roanoke College).

But no matter how many African Americans, or men, or Independents turnout to vote next Tuesday, the writing has been on the wall for some time. Cuccinelli is headed for defeat. The only question appears to be whether it'll be close enough to give the race a bit of excitement. Though if the current survey average is any indication, the most exciting thing in Virginia on Tuesday night may be the pollster accuracy competition.

Sidenote: As an illustrate of how poor a shape Cuccinelli is in for election day, consider the chart below, which highlights poll results from the four pollsters had their partisan identification findings been reweighted to match the electorate that turned out in 2009. Despite the strong Republican turnout advantage, Cuccinelli can't get out in front of McAuliffe in any of them:

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.