08/15/2013 06:46 pm ET

HUFFPOLLSTER: Christine Quinn, Bill De Blasio Ahead In NYC Mayoral Primary


Another New York City mayoral poll finds momentum for de Blasio (though not first place). The New Republic's Nate Cohn spars with PPP over the chances that Georgia turns blue. And for dessert...pie charts. Made out of actual pie. This is HuffPollster for Thursday, August 15, 2013.

MARIST POLL SHOWS QUINN, DE BLASIO AHEAD - Jon Schuppe: “The topsy-turvy race for New York's Democratic mayoral nomination has shifted again, with Public Advocate Bill de Blasio surging to a statistical tie with City Council Speaker Christine Quinn and former comptroller Bill Thompson in striking distance, a new poll shows...The poll, conducted Monday through Wednesday, found Quinn's support at 24 percent, statistically unchanged from the last poll in July. De Blasio jumped to 21 percent from 14 percent. Thompson followed at 16 percent, just a couple points higher from July, and Weiner dropped to 12 percent from 16 percent...In the Democratic primary race for comptroller, former Gov. Eliot Spitzer widened his lead over Manhattan Borough President Scott Stringer, to 56 percent to 34 percent." [NBC New York]

Even closer in a runoff - If, as seems likely, no candidate reaches 40 percent of the vote on Sept. 10, the top two vote-getters will head into an Oct. 1 runoff. If de Blasio and Quinn faced off, de Blasio would take 44 percent of the vote, and Quinn 42 percent, according to the Marist poll, which was conducted Aug. 12-14 and surveyed 679 registered voters. In a runoff between de Blasio and Thompson, de Blasio would lead by 44 percent to 36 percent.

Poll confirms de Blasio surge - A Quinnipiac University poll of likely voters released on Tuesday had showed de Blasio running ahead of Quinn (30 percent to 24 percent), followed by Thompson (at 22 percent) and Weiner (at 10 percent). So the two new surveys conflict on the order of candidates, but both found big jumps for de Blasio. More from NBC New York: "As Weiner flounders, the race's momentum has turned in favor of de Blasio, a liberal, outspoken critic of Mayor Bloomberg and New York's policing tactics -- particularly stop-and-frisk, which a federal judge deemed unconstitutional this week. De Blasio's turnaround is due in part to a swell of support from black voters; 20 percent of black Democrats said they'd vote for him, up from 10 percent last month. His surge in the polls has come as he has sought to showcase his racially mixed family on the trail, and released an ad narrated by his teenage son, Dante. Thompson, the only black candidate in the race, has the support of 22 percent of black Democrats, the poll found."

Advertising pays off - De Blasio’s higher standing in the polls comes shortly after his campaign began airing commercials on August 8. Previously, only Quinn had been runnings ads. Eliot Spitzer has also run four spots since entering the NYC comptroller’s race, while Stringer, his lesser-known rival, is debuting his first ad this week. Over the final three weeks of the campaign, the television air wars are likely to be far more competitive.

NATE COHN: GEORGIA IS DEMS' BEST CHANCE FOR 2014 SENATE GAIN - "The conditions for an upset might be brewing in the Peach State, where Saxby Chambliss is retiring and Democrats have stealthily become close to competing on a more regular basis. It might even be a better bet for Democrats in 2014 than Kentucky...Georgia's black turnout wasn’t so bad in 2010, despite a bad national climate and without a competitive federal election. According to the Georgia Secretary of State, the black share of the electorate only declined slightly, from 30 percent in 2008 to 28.3 percent in 2010. The white share of the electorate increased from 64.1 to 66.3 percent...Demographic change helps counter some of the drop-off. The state’s non-white population is booming and, as a result, the white share of the electorate dropped from 64.1 percent in 2008 to 61.4 percent in 2012. A similar decline was observed among registered voters...To be clear: Georgia is not a toss-up. The state retains a clear Republican lean in neutral conditions, and the Democratic path to victory is quite narrow with so few persuadable voters. Nunn probably needs the Republicans to blow it. But across the board, Georgia seems like a more winnable state for Democrats than Kentucky. [The New Republic]

And Cohn raises a red flag about a PPP poll... - In his article, Cohn refers to a recent PPP poll that showed Michelle Nunn, Democratic candidate in Georgia's upcoming U.S. Senate race, leading or tied with all of the potential Republican candidates as "meaningless" and "capriciously weighted." He explains in a footnote: "PPP weights to their own targets for race, based on a secret recipe of Census data, exit polls, or whatever else they feel like. In fact, this secret recipe is supposedly the secret to their success. But PPP's recent poll of Georgia found that whites constituted 71 percent of the electorate, while blacks were at 24 percent. This is not a state where there should be so much guesswork. According to the Georgia Secretary of State, it was 66 percent white, 28 percent black in 2010. If for some reason PPP didn't notice that, the Census estimate is even less white: 64 percent white, 32 percent black. In 2014, the electorate should be more diverse, but certainly not significantly more monochromatic. So how did they get 71 percent white, 24 percent black? Who knows. But I'm sure that a good method wouldn't have yielded those estimates. And in a state like Georgia, where racial turnout is 90 percent of the game, it's a little ridiculous that they could be off by so much." [ibid., see also PPP poll]

...clarifies that his beef is unrelated to the result... - Via email, Cohn stresses that although he believes "there are clearly too many whites [in PPP's sample], and, if the poll was weighted properly, they would have found Nunn doing better...I'm not complaining about the result, just the demographic composition of the poll. In fact, I think the underlying sample is far too friendly to Nunn. The easiest way to illustrate that: the poll has Romney winning by 7, just as he did last November, even though the electorate is about 10 pts whiter and 6 pts less black than 2012. In a racially polarized state, that's just not possible. They show Nunn with a larger share of the white vote than any federal Democratic candidate in a long time, despite plenty of undecided voters."

...and starts a "Twitter war" with PPP - Via Twitter, Cohn called on PPP to "explain how it got that whites = 71% of GA voters." After an initial reply from PPP's Tom Jensen noting that Republicans had also "bitched about the sample being too Democratic," Cohn pressed: "but what metrics point toward 71% white? I don't need the magic formula, just which ingredient pushes you near that figure?"

Jensen responded: “In the 2008 runoff, the last GA election without a major black candidate on ballot, everyone overestimated black turnout... And as a result significantly undershot Saxby Chambliss' margin of victory. If we weighted blacks any higher [than their percentage in 2008] on this poll...We'd be looking at a nearly split Obama/Romney electorate and a Democratic ID advantage."

Cohn and Jensen then went back and forth about the import of a 28 percent black composition of the 2010 election as reported by Georgia's secretary of state. Which led to this exchange:

Jensen: "There was a black Senate candidate in 2010. 2006 was 24% black."

Cohn: "The idea that you can run back the '06 electorate in '14 is a little ridiculous."

Jensen: "I think our poll reflects how Nunn stacks up to the GOPers at this point. If you think she should be up by 5-6 that's fine by us."

Cohn: "[A]re you saying you weighted to get the result you wanted?...because I'm reading this as 'our sample showed Nunn doing way too well among whites, so we reduced the number of blacks.'"

Jensen: "As I said the last time there was a big GA-Sen race everyone overshot black turnout and undershot Saxby...Maybe we're being over cautious but we try not to make the same mistakes twice. That's all I have to say." [@Nate_Cohn & @ppppolls via Storify]

GOP POLLSTER'S CONTRARIAN TWITTER 'RANT' ON PREDICTING VIA TWITTER - Via Twitter, Republican pollster Alex Lundry adds his perspective on a controversial study promoted in a Washington Post op-ed this week that claims counting tweets can predict election outcomes as well as polling: "Everyone is overlooking the most important point in this 'twitter predicts elections' controversy. Forget about prediction - the research shows a relationship between tweets and votes. THIS is what’s important. Even after introducing completely reasonable controls into the model, tweet share was still significantly related to votes. & forget about whether [Twitter] is representative of population. Q is, can Twitter dialogue drive bigger conversation that affects larger pop? This research indicates it can. & it reaffirms what we saw over & over again in 2012. The way you move #’s is by driving a conversation. We linked up TV ad data, social media volume/sentiment, news media vol/sent and polling. Each time we saw the same pattern…An ad starts dialogue on social media that peaks 2-3 days after release, news media next 2-3 days after, then 3-4 days later ballot effects. So its not whether Twitter predicts or whether its representative. Its that the conversation that happens there matters. And pollsters/strategists/pundits that stick their fingers in their ears and want to ignore Twitter data are fools. But also, analysts that declare the death of polling because of Twitter are equally foolish. [@alexlundry via Spotify]

HUFFPOLLSTER VIA EMAIL! - You can receive this daily update every weekday via email! Just enter your email address in the box on the upper right corner of this page, and click "sign up." That's all there is to it (and you can unsubscribe anytime).

THURSDAY'S 'OUTLIERS' - Links to more news at the intersection of polling, politics and political data:

-President Obama’s economic approval rating drops to 35 percent in a Gallup poll. [Gallup]

-Jonathan Bernstein calls on pollsters to ask what exactly people mean when they talk about the deficit. [A Plain Blog About Politics]

-Most Americans don’t realize that the federal deficit shrunk last year. [YouGov]

-Sean Trende reviews the track record of "party makeovers" and finds little record of success. [RCP]

-Robert Blizzard sees parallel trends in the approval ratings of Barack Obama and George W. Bush. [POS]

-NYC baseball fans don’t want to see Alex Rodriguez banned for life. [Quinnipiac]

-"A video of pie charts, made out of pie, about pie. Booyah." [Flowing Data]