How polling can stop making America’s polarization problem worse

Friedman is a former president of Public Agenda and the founder of its Hidden Common Ground initiative.

Survey research can serve democracy by illuminating people’s views, values and concerns. But, as an exhausted nation with a shredding social fabric faces a fateful midterm, I worry about the ways polling can also exacerbate America’s polarization problem.

Pollsters often ask questions in ways that exaggerate America’s divisions. This feeds the narrative, trumpeted by news outlets and weaponized by demagogues, of two monolithic nations vehemently opposed on every conceivable question. Yes, our national parties are polarized to the point of dysfunction, as are significant segments of the American public. But in many instances the narrative mischaracterizes the American people overall, feeding the anxiety and depression sweeping the land like omicron and depleting our democratic imagination when we need it most.

For example, an NPR/PBS News Hour/Marist Poll in June 2021 asked which concerned respondents more, “making sure everyone who wants to vote can do so” or “making sure that no one votes who is not eligible.” The poll found strong partisan polarization, with 85 percent of Democrats concerned with access and 72 percent of Republicans with security.

But what if you ask the question in a less binary way, as a Public Agenda/USA Today poll did in July? Rather than an either/or question, respondents were asked if their biggest priority was “preventing voter fraud,” “making voting simpler, convenient and hassle free for everyone,” or “preventing fraud AND making voting simple, convenient and hassle free”? A consensus of Republicans (67 percent), Democrats (64 percent) and independents (73 percent) chose the third option. Rather than extreme polarization, we see considerable common ground.

Which is the more accurate representation? Binary questions can sometimes reveal important partisan differences, such as on America's racial reckoning. But election security vs. voter access? Just because politicians pretend it’s an either/or proposition is no reason for pollsters to mimic a false choice and create a false impression.

A Pew study from April 2021 shows that the common ground on voting extends to concrete proposals, including requiring paper backups for electronic voting machines and permitting in-person voting for at least two weeks. More, the Public Agenda/USA Today poll found that super majorities of Republicans, Democrats and independents agree that the federal government should ensure voting access for everyone.

We see a similar pattern on culture war flare-ups about teaching American history, as in a recent USA Today article based on an Ipsos poll. “[N]ot surprisingly,” it contends, “the issue is firmly politicized: More than 8 in 10 Democrat parents believed their children should learn about the lingering impact of slavery and racism in schools, compared with fewer than 4 in 10 Republican parents.”

While the issue can be politicized, is it so starkly polarized among most Americans? Another Public Agenda/USA Today poll casts doubt. Respondents were asked which school curriculum would do the most to bring the country together. One that emphasizes “America’s achievements and greatness and honors its traditions,” “America’s shortcomings, mistakes and how it needs to change,” or “both America’s shortcomings and achievements”? Majorities or pluralities across the political spectrum favored the both/and response, showing a great many Americans willing to engage the kind of question posed by Eddie Glaude in reflecting on James Baldwin: “What does the story of slavery … look like when told in a way that neither glosses over the cruelty and failures of the country nor demonizes every aspect of the society… ?”

There are plenty of reasons pollsters sometimes frame questions in ways that elicit polarized responses, including that news outlets find those stories easier to tell than more nuanced ones. Another was suggested to me by a leading academic researcher: Pollsters take their cues from the way political leaders frame issues. He asked if I thought they should take it upon themselves to do otherwise. My response is that researchers should frame questions in ways that enable people to express what they really think.

Polls should do more than ask people to react to the limited choices offered by powerful, dysfunctional elites. If our democracy is to renew itself in this time of existential threat, pollsters can help by neither exaggerating nor papering over our differences. They should pay as much attention to the common ground upon which solutions and coalitions can be built as to the authentic disagreements the nation must navigate. They can help by letting the people speak.