Did Deez Nuts Break Political Polling?
By now, you've almost certainly heard: An Iowa teenager registered to run for president as "Deez Nuts," and suddenly found himself polling beside Hillary Clinton and Donald Trump in North Carolina, as well as the subject of at least one kinda hilarious/kinda cringeworthy super-cut.
Fine. It's August. Everyone's off on vacation, silly and sand-addled.
Here's the thing: In the last year, amid plummeting response rates and unresolved questions about how technology has disrupted political forecasting, pollsters have conspicuously mis-called the U.K. and Israeli elections, the U.S. midterm elections, and the Scottish independence referendum, prompting star-statistician Nate Silver to wonder if polling is "in crisis." On top of all that, seeing Deez Nuts—whose real name, according to media reports, is Brady Olson and who is 20 years shy of meeting the Constitution's age minimum on running for president—break into a Public Policy Polling survey does little to reassure politics watchers on the polling industry and the electorate it purports to survey.
"PPP is to some degree almost like the crack cocaine of polling," said Zac McCrary, a partner in Anzalone Liszt Grove Research political polling firm. By that, he said he meant that its polls are cheap, quick, and fun. "They’re in the volume business. They need to find a way to get media attention to get clicks."
Somehow, McCrary seems not actually to mean this as a particularly negative comment about polling or PPP. Sure, he said, it's not how "a traditional strategic research firm" might spend a client's money, but the survey might actually say something about voter attitudes.
"It’s a more salacious, a more interesting way, to say, 'none of the above,'" McCrary said. "Fourteen, 15 months out from the election a lot of people are certainly in the 'none of the above' head space."
Historically, a handful of people have also embraced joke and protest candidates even at the moment that they went into the ballot box, said Bryon Allen, a partner in WPA Opinion Research, a political data company.
"I do think maybe it tells us about the degree to which the Trump thing and everything else that’s going on in the early period of the presidential campaign has made people more susceptible to treating the whole thing as a bit silly right now," Allen said, referring to the surge in the polls for real estate mogul-turned-reality TV star-turned presidential contender Donald Trump. Other things included controversy surrounding Hillary Clinton's e-mail usage, he said.
Allen and others said Nuts' success could also be a methodological artifact: PPP tends to use robo-calls and Internet questionnaires, rather than live surveys, and people seem more likely to joke or to protest on those forums.
In fact, more than half of Internet respondents (who, of course, skew younger) were willing to rate Nuts' favorabability, with one in five rated him favorably, Ken Goldstein, a political science professor and pollster, pointed out. Only two percent said they liked him on the phone.
"If we were to do this survey by traditional phone methods, I don’t think Deez Nuts would be doing so well," Goldstein, who has written about polling for Bloomberg, said.
What does PPP have to say for itself? The firm's director explained that his survey asked about Nuts' favorability after social media followers, who usually get to suggest a question, urged him to put Nuts in.
"People tweeted at us. They said, 'You should poll this. It’d be funny,'" the director, Tom Jensen, said. "We just don’t think that politics has to be serious all the time."
(Reports have also said things got started when Nuts e-mailed Jensen, which Jensen said is true but wasn't the deciding factor.)
Jensen didn't respond specifically to comparison of PPP to a drug, but he did say being humorous doesn't mean he can't poll seriously. He said the survey of about a 1,000 voters didn't ask about Nuts until the end, making sure it didn't contaminate other questions, and Jensen talked about seeking out representative samples and screening respondents to make sure they met pre-selected criteria.
"They’re still real voters," he said.