Anthony Weiner and the cell advantage
Weiner leads the Democratic pack with 30 percent among cell-phone users, according to the poll. Quinn was the next closest candidate among cell phone respondents, with 16 percent. City Comptroller John Liu was third, with 11, Public Advocate Bill de Blasio and former comptroller Bill Thompson followed with 9 and 8, respectively.
Twenty percent of cell phone respondents said they were undecided.
Among those who responded by land line, the results more closely mirrored the poll's overall findings with Weiner in the lead, followed by Quinn, Thompson, de Blasio and Liu.
Weiner led among landline respondents, but barely, with 23 percent, followed by Quinn with 22 percent. Thompson had 15 percent, followed by de Blasio with 11 and Liu with 8.
What's the significance of the fact that Weiner's margin comes mostly from from cell-phone users?
For one thing, they tend to be younger voters, who, historically, turn out in lower numbers than their older counterparts, who disproportionately continue to rely on land lines.
(As Harry Enten at the Guardian pointed out to me, a survey from the Centers for Disease Control and Prevention found that young people and also minority voters "are the most likely to eschew landline telephones.")
So the question, for the purposes of measuring them in a poll, becomes whether to account differently for the likelihood of cell-phone respondents to turn up on primary day (or for a runoff) than for the likelihood that a land-line user will.
Mickey Blum, the director of Baruch College Survey Research, said, "People who have cell phones only are more likely to be younger, poorer and minorities and probably less likely to vote in a mayoral primary. Mayoral primaries get low turnouts. Only diehard voters show up, unless there is a reason for a particular group to vote. That was true of Latino voters when Ferrer ran."
Basil Smikle, a Democratic political consultant who worked for Weiner in prior elections, told me in an email that cell phone respondents "are younger, more mobile, more engaged in social media and vote in general elections but [it's] unsure if they have significant primary voting history."
Bill O'Reilly, a Republican political consultant, said in an email that cell phone respondents "would be less likely to vote -- younger audience."
All of which suggests that the cell-phone responses might be less accurate predictors of voter behavior than land-line responses.
But as the veteran Democratic pollster Jef Pollock pointed out, the screening of the respondents ought to have accounted for the voting likelihood of the participants already.
"Cell phone and land line respondents are different no doubt," he wrote. "They tend to be younger, so sure less likely to vote, but theoretically, the people in this poll are all likely voters. So not sure that should make a difference here."