New poll numbers coming out Tuesday from Public Policy Polling (PPP) are like “pretty much all of them,” meaning, their website says, “within a few points in one direction.” Their preview blog post goes on to say conventional wisdom would dictate a tie means a loss for the incumbent, but “in the case of Nevada there are a couple big reasons why the tie might go to Harry Reid.”
As evidence, PPP cites the fact that 2008 polling in Nevada turned out to be the worst of any swing state. Obama won Nevada by 12 points but PPP, Mason-Dixon, and Rasmussen had him up by only 4, and CNN had him up by only 7. (To be fair, PPP points out that Suffolk showed Obama up by a 10 point lead, Zogby showed him up by 11, and AP had him up by 12.)
PPP concludes that there is at least some precedent, then, for pollsters to underestimate Democratic performance in Nevada, which can skew polls.
Will this be the case again? Maybe; maybe not. If the Obama-wave voters don’t come out in droves for the midterms, the pollsters showing Sharron Angle trending ahead may be right. On the other hand it is possible, as PPP speculates, that “the polls are systematically underestimating Harry Reid.”
Most public polling is assuming a large gap between Democratic and Republican turnout. PPP says their poll tomorrow consists of a sample showing they voted for Barack Obama in 2008 by only 2 points, compared to his actual 12 point victory. And even with that dropoff in turnout, PPP says their poll shows the race still very close.
When asked what he thinks, veteran journalist Jon Ralston–who says he has seen more “crappy public polls” than he cares to this campaign season–says he thinks Reid may be up slightly.
“I do not trust any of the public polls. They all have sampling problems,” said Ralston in an interview with Battle ‘10. “However, any single poll is only as good as its model. If Republicans swamp Democrats in turnout, the public polls could be borne out. If not–if the Dems can come close to holding their registration advantage, or only lose a couple of points–it will be very close.”
Ralston said bad public polls are a problem because they “drive the narrative” and many in the press have no idea how to properly analyze and report them. He also said a pollster surveying for anything less than a full ballot test at this point–including “None of the candidates” and ‘Tea Party of Nevada’ candidate Scott Ashjian–diminishes accuracy.
As for whether the notorious Nevada Democratic get-out-the-vote machine can compensate for the enthusiasm gap, Ralston said that is “the” question of the political season.
“There has never been an operation like it, but there may never have been a headwind like they have to deal with,” said Ralston.
“There is a five-point gap in registrations between the two parties and the Dems believe, worst case, their edge goes down to one point at the polls,” said Ralston. “If you look at raw numbers, that would take an incredible win by the GOP in turnout numbers.”
September voter registration numbers showed Democrats with 466,512 active voters statewide compared to 406,986 for Republicans and 171,842 non-partisan voters.
Early voting in Nevada starts Saturday.