In an early test of Paul Ryan's impact on the presidential campaign, pollster John Zogby finds the race tied, matching recent findings from both Gallup and Rasmussen. Most notably, Zogby claims Romney has crested 40 percent of the 18- to 29-year-old age bracket, for the first time.
This is a critical threshold that the Obama camp can be none too pleased about. President Obama scored 66 percent of the youth vote in 2008. In an interview with the Washington Examiner, Zogby thinks Ryan may be opening up the youth vote to Romney.
"'It could be his youthfulness,' said Zogby of Ryan," the Examiner reported. "Plus, he said, more younger voters are becoming libertarian, distrustful of current elected officials and worried that they are going to get stuck with the nation's looming fiscal bill."
Zogby's poll comes at a moment when the polling industry is engaging in sustained naval-gazing, trying to figure out what to do with the increasing prevalence of cellphones and the ever higher refusal rate when they reach someone.
This uncertainty amidst changes results in vastly different philosophies producing markedly different results. The latest polls, aggregated by Real Clear Politics, could show Romney up by 4 points (Rasmussen) or Obama up by 9 (Fox News).
Rasmussen uses online interactive polls to try to compensate for the fact that its polling machinery cannot dial cellphones. Other pollsters hand-dial cellphones and lock in 25 or 30 percent of the slots for those who have only cellphones and no land line.
"The cellphone is not new; hence it is not a new problem," wrote John Zogby at Forbes last week. "There have been colleagues who initially eschewed calling cellphones, then began arbitrarily setting 'standards' for the proper numbers of cellphones in each sample, then determining who was a genuine cellphone respondent and who was not. In short, some of my best-known competitors were exercising the art of polling, just as I had suggested they would have to do. Those who deny new technologies and methodologies, criticize others for using them and then arbitrarily embrace them are only causing confusion. But ultimately, a good pollster is one who embraces new means of communicating with consumers and voters can reasonably explain why his methodology works, and can prove that it does work over time."
A key point of contention is the split among Democrats, Republicans and Independents in a given poll. Republicans often contest polls that show markedly more Democrats than Republicans, arguing that exit polls in 2008 and 2010 suggest something much closer to parity.
An ABC analysis of exit polls after the 2010 elections showed that Republicans had pulled even, at least in turnout. "Democrats and Republicans were at parity in self-identification nationally, 36-36 percent, a return to the close division seen in years before 2008, when it broke dramatically in the Democrats' favor, 40-33 percent."
In short, a wipeout 2008 stemmed from a "dramatic" 7-point Democratic advantage, while a Republican tide in 2010 showed an even split between the parties, with Independents tilting to the GOP.
One might thus conclude that the measure of a poll is how closely it predicts the likely turnout of Democrats, Republicans and Independents, and how accurate it is in predicting the tilt of the latter. Any poll — or any news report of a poll — that does not front these ratios would thus be suspect.
"In 2004, re-weighting polls to reflect the 2000 exit polls was all the rage among Democratic bloggers," wrote Sean Trende at Real Clear Politics. "The argument went that Republicans hadn’t had parity with Democrats in polling in a very long time, so we should ignore polls showing Republicans even with Democrats, or perhaps even ahead of Democrats in terms of ID. Of course, the final exits showed a tie between the parties, as Republicans managed to turn out their base at 'supercharged' levels."
Keeping with this tradition, the pundits at the conservative Hot Air blog routinely shoot fish in a barrel, critiquing the latest polls by their DRI distributions.
"The CBS/NYT model has Democrats a +9 in Florida," wrote Ed Morrisey on Aug. 1, when in 2008 they were only a +3 and an even split in the 2010 midterms. Ohio’s sample has exactly the split in 2008 (D+8), which is 9 points better than Democrats did in the midterms. Pennsylvania’s numbers (D+6) come closest to a rational predictive model, somewhere between 2008′s D+7 and 2010′s D+3, but still looking mighty optimistic for Democratic turnout.
The consensus among many pundits is that 2012 could split the results of 2008 and 2010, which were high water marks for the Democrats and Republicans, respectively.16 comments on this story
Two weeks ago, John Harwood assessed the confused state of the polling industry for the New York Times, noting that an NBC/Wall Street Journal poll showed Obama up by 6 points, after enforcing a 30 percent cellphone-only requirement, while the online-centric Rasmussen, polling within hours the same day, showed Romney up by 3 points.
"Nobody has the answers," Scott Rasmussen told Harwood. “We’re all experimenting with the same thing. How do you reach people in a way they communicate?”
Eric Schulzke writes on national politics for the Deseret News. He can be contacted at firstname.lastname@example.org.