The gradual supplanting of traditional pollsters by poll aggregators as our national tea leaf readers is just the most obvious ripple from a sea change in the public opinion industry. Experts who speak from the gut about Americans’ mood and sentiment, informed by decades of immersion in questionnaires and focus groups, are in the process of being replaced by experts who speak Big Data. As analytics start crowding results-based intuition, the risk is that intuition—and with it, the empathy, early trend-spotting and other benefits of decades in the field—may be crowded out altogether.
I recently witnessed a panel discussion about the future of public opinion polling in the face of not just mounting methodological challenges but also broader ones related to public distrust, perceived bias and inconvenience. (Full disclosure: the panel was hosted by my company, Kantar, and expertly moderated by this publication’s national editor, Amy Walter.)
Watch the panel on C-SPAN: http://cs.pn/1dMgYdy
Republican pollster Bill McInturff wasn’t the chattiest member of the panel, which also included President Obama’s pollster Joel Benenson, Mark Blumenthal of The Huffington Post, Kantar’s Ken Goldstein; and Pew Research Center chief Alan Murray. All brought different but ultra-informed perspectives to the exchange. Benenson spoke from his winning work and his warfare with the news media over their poll coverage in 2012, Blumenthal is a methodology savant, etc.
But the variety of perspectives set off McInturff’s long view. As co-founder of Public Opinion Strategies and co-pollster for the NBC News/Wall Street Journal national survey for a decade, McInturff doesn’t so much talk as exhale insights based on his direct engagement with Americans through phone polls and focus groups since 1987. He has worked in the industry long enough to have developed his party’s widely used question gauging interest in voting back in 1992—and to recognize, 20 years later, that the question no longer accurately assesses voters’ likeliness to turn out.
In the audience were dozens of pollsters and poll consumers of varying levels of experience. The youngest hailed from the Big Data set: staffers from new Democratic firms started by Obama campaign alumni and Greenberg Quinlan Rosner Research, which has launched a digital practice. Also present was Brent McGoldrick, head of paid media for the Romney 2012 presidential campaign, whose background bridges traditional polling and analytics.
These folks represent where the public opinion industry is going—an industry in which the seekers stop waiting by the phone and place themselves at all the proliferating junctures of devices and platforms where public opinions are expressed. The new industry is looser knit and more fragmented. The guy who gathers the datasets may not be the guy who deciphers them. A smaller percentage of people involved in the process will ever actually interview a human being.
This diverges from the public’s own view of what public opinion polling is. At least in the abstract, Americans prefer human contact when it comes to offering their opinions on politics and policy. In Kantar’s “poll of polls” released at the event, the option of being interviewed in person had the most positive impact on people’s willingness to take part in a poll (24 percent more likely to participate), followed by interviewing by landline phone (23 percent). The option of participating in a poll by IVR or social media had the most negative impact on people’s likeliness to take part (58 percent and 60 percent less likely to participate). (Of course, it stands to reason that people who are willing to take part in a phone poll are somewhat more likely to favor person-to-person contact.)
But as Blumenthal declared at the start of the panel, the future is upon us. For decades, polling has been a partnership between pollsters and the public who have answered the door, then the phone. Now, to overcome the challenges it faces in getting people on the phone, the industry must get ahead of the public. People’s opinions won’t just be solicited—they’ll be scraped and bought in bulk from vendors.
Also driving the shift from the pollster’s gut toward the analyst’s stats are changes inside the political consulting community. People entering the field today are being trained differently from McInturff and former Democratic campaign pollster Blumenthal, both of whom said flat-out that their backgrounds are insufficient to interpret Big Data.
On top of being more tech-savvy, younger political consultants are more pluralistic. They harbor an “anti-guru” sentiment that isn’t directed against any particular practitioners, for the most part, but against the idea that instinct—particularly the instincts of one consultant—should ever outweigh the math of many.
Watch the panel on C-SPAN: http://cs.pn/1dMgYdy
One practitioner pointed out to me that the wiseman pollster is an idea reinforced by a mainstream media that needs crowned prognosticators to provide informed speculation about public opinion. The explosion of cable especially has led to the crowning of more and more (and less and less experienced) prognosticators.
For the record, anchors Tom Brokaw and the late Tim Russert haven’t kept McInturff and longtime NBC News co-pollster Peter Hart close at hand on election nights to spout soundbites, but to merge their knowledge of public opinion with the exit poll findings. Still, it makes sense that the wisemen and women who can eloquently sum up the people of the United States are as endangered as the traditional media world that created them.
The national media also continues to foster the idea that when it comes to national politics and policymaking, national polls matter. The uncomfortable but gelling reality is that all opinions are not created equal. A president can get elected without commissioning a single national survey. A majority of the public can support new gun laws but a handful of lawmakers can block them. This occasional static between the nation’s opinions and actual outcomes already casts doubt on the commitment of lawmakers to any higher purpose than getting re-elected. But it may be starting to cloud the perceived value of national polls, as well.
For the near future, as we bridge the gap between traditional polling and whatever’s next, those in need of tea-leaf reading will keep relying on intuitive expertise. Alan Murray of Pew predicted that within 20 years, the substance of the panel discussion will seem quaint. What might seem quaint a lot sooner than that, for better but also for worse, is the idea that one public opinion practitioner is able to authoritatively explain us to us.