Good journalists strive to uncover and share truth with others. That’s what attracts many people to the field or to supporting it, including myself.

It’s also why in recent years I have become interested in the study of “perception gaps,” or deeply distorted understandings people have of each other. I saw polls, like one from the nonprofit More in Common showing that Americans can vastly overestimate the number of people who hold extreme views. In terms of politics, the most partisan and politically active Americans are often the most misguided on views of “the other side,” while those who are less engaged are as much as three times better in guessing the views of the opposing side. I had begun to wonder how journalism, or poor journalism, might make these gaps worse, even without malintent. 

It’s easy enough to imagine what in journalism could make these gaps worse, some of which is likely beyond the control of any one journalist or the public. Platform incentives. Deadline pressure. Sloppy journalism that nonetheless gets amplified on social media alongside the good. Biases. But I also wondered about how journalism might reduce these gaps.

Noelle Malvar, Senior Researcher, More in Common USA,
New York

That’s why I was grateful that Joy Mayer of Trusting News and I got to connect recently with Noelle Malvar, a researcher on many of More in Common’s studies. We talked initially about the pull of binary categories — “this or that,” “party A or party B,” “identity x or identity y” — and  their influence on perception gaps. And we discussed how polling and storytelling, if we’re not careful, can obscure a more complex reality about the many identities we hold, strongly or loosely.

After that, I emailed with Malvar about how she thinks journalism might intentionally avoid or break up binary narratives. Our discussion could be considered part of a growing conversation between journalists and researchers about journalism in a time of “high conflict,” as writer Amanda Ripley has put it, or “truth-telling” in a time of misinformation and polarization, as we at API put it in 2019. Importantly, we discussed how journalists might learn from studies about perceptions and identity to better cover the anti-democratic actions that are becoming far too common in the years since.

How might journalism avoid these binary constructions, reduce perception gaps and better reflect more people’s complicated, hard-to-categorize identities? Our conversation is presented below, lightly edited for clarity.


Loker: Noelle, I’m curious if we could start simple — and concrete. What are two or three examples of public sentiments you think people who read a lot of media might be surprised to learn? 

Malvar: People who engage with a lot of media tend to think that more people support political violence, especially violence toward other Americans, than actually do. I think the media sometimes paint a picture that Americans from across political groups perceive violence as acceptable or even positive. That’s not entirely true. Our research shows that most Democrats and Republicans do not condone physically attacking ‘the other side.’ Recent polls (influenced by recent exceptional events) imply increasing numbers, and with violence either against the government or toward each other, even having a few supporters is troubling. But it’s also important to contextualize these findings and note that they are driven by strong partisans and Americans who hold extreme views. People are angry and exhausted, with pandemic stress and grief and uncertainty it can feel like a tinder box –  but the majority remain non-violent nor do they condone violence.

Another is that most Americans, regardless of race, age or political party, are grateful and proud to be American. There are of course variations based on group identities and fluctuations over time, but gratitude and pride are consistent emotions attached to the American identity – our data show that most Americans will still choose to live in this country when given the choice to live anywhere. The media frequently engages in upward social comparisons with other Western, democratic societies, but the majority of Americans are relatively thankful to be here.

Lastly, Americans who are attuned to the news or the headlines might feel that the nation is irrevocably divided along partisan lines, and there are profound and fundamental schisms between strongly identifying Democrats and Republicans. But divisions among the majority of Americans are not as extreme as we think. Our research compares the extent to which Republicans and Democrats think they disagree with the amount they actually disagree: a “perception gap.” On some issues, Democrats and Republicans imagine that almost twice as many people on the other side hold extreme views than really do. Consumption of most forms of media, including talk radio, newspapers, social media, and local news, is associated with a wider perception gap. For example, people who consume news “most of the time” are almost three times as inaccurate as those who consume it “only now and then.”

Loker: I found that striking in More in Common’s work: The perception gap between two sides can be so large. And alongside things like educational attainment, news behavior with even quality sources seems correlated with some of the gaps. 

It’s also probably not something most journalists want to hear. Most journalists want their readers to have an accurate sense of reality. Are there ideas from your work or others for how journalists might make this better? Or at least ways to not widen gaps further?

Malvar: Reality is complicated, and people are complex, so stories that reflect these complexities, and stories that complicate existing narratives, are helpful. Journalists and media in general play such a significant role in shaping social norms, and the more we can read stories that involve characters thinking complex thoughts on complex issues, the better. For instance, we know that most Americans are not strongly identified partisans (e.g. their political identity is not their most salient identity), but what does it mean if all they read are views of strongly identified Republicans and Democrats (or using our Hidden Tribes framework: Progressive Activists and Devoted Conservatives)? This relates to the perception gap. If we know that X% of Republicans have said opinion on an issue, but Democrats think Y% of Republicans do, how does a story that focuses on the most extreme Republican opinions (held by a minority) influence people’s perceptions of the other group?

More in Common shows the political “wings” of the country as made up by the fewest Americans.

The perception gap increases at these “wings.”

There’s a great article by Hélène Biandudi Hofer for Solutions Journalism Network that talks about other steps, such as asking questions that get at people’s life experiences and underlying motivations for their opinions and beliefs. We (More in Common) try to do this through our research methods too. We don’t just rely on public polling, but also do focus groups, in-depth interviews and longitudinal qualitative research. That’s because we know that it’s impossible to compress context, experiences, and reasons into a single statistic that then becomes a fully accurate representation of them or their group. So it’s this work of trying to break up the binary, or encouraging people to realize there’s sometimes more than two ways to look at issues.

Loker: The binaries common in national political reporting do seem aligned with “the wings” of your Hidden Tribes study — Progressive Activists and Devoted Conservatives. But you saw a similar trend in the state of Texas, and on local matters, correct? 

If a local journalist wants to avoid inadvertently worsening perceptions in their state or community, what does your Texas project suggest they might consider?

Malvar: We did observe the same trends in Texas. On issues related to immigration, or confederate statues, the percentages when looking at Democrats vs. Republicans would make it seem like there is a clean and clear partisan split. But when viewed through our “threads of Texas” framework, perceptions are more complex. For example, segments that are not defined by their political ideology are more varied in their responses, and how one leans on a perceived controversial issue does not necessarily always correlate with how they will view another issue.

There is a sizable portion of the Texas population — the “Apolitical Providers” and “Diehard Texans” for instance — whose opinions and stories are often overlooked because they don’t fit a clean narrative of what a Democrat or Republican is, so having their stories amplified (as the norm, not the exception) is helpful. Polarization begets polarization. Though perceptions of polarization have the same effect – being accurate about what issues are polarized and to what extent is important.

Another finding — and consistent with other research — is simply put: Demography is not destiny, and certainly not electoral destiny. In Threads of Texas, Hispanic Texans were found in all of the different Texan Threads (from “Lone Star Progressives” to “Heritage Defenders”). Their opinions about Texas’s economic future, for instance, were predicted more by their perception of available opportunities and belonging to that future (or whether they will be left behind) than by their political orientation or ethnicity. We often think and talk about other groups as monolithic voting groups (e.g., how will X candidate secure the Hispanic vote). I think our threads do a good job challenging that narrative and highlighting the diversity of viewpoints. Journalists can remember this when writing about public opinion on salient issues.

Loker: What I’m hearing is that a lot of polling may skew our understanding of reality, by perhaps focusing too heavily on political identity, which isn’t always salient or maybe most relevant. The result can give power to people who benefit from the binary construction — probably “the wings” — and we lose a rich complexity that exists in many people. Am I getting that right, or not exactly?

Malvar: Yes, that’s about right. Note that it’s not polling per se that causes the skew of reality. Polling is helpful especially because most of the time, elections and policy decisions ultimately have binary outcomes. But people aren’t as easily classifiable in their opinions and the identities they hold and that’s the reality that the labels and numbers can obscure. 

Politics is almost always framed, unwittingly or not, as a team sport, but we know that there are significant portions of the country that are disengaged, ideologically inconsistent, or simply ignored.

Loker: What is the response from policymakers and politicians to some of your work? 

Malvar: Politicians and people who work on campaigns have to contend with the reality that elections and policy decisions usually have binary outcomes (win/lose), but from our experience they do value knowing the complexity behind the electorate’s (and in general, people’s) thinking and motivations. With folks from politics and the policy world, we’ve gotten the most positive response on our segmentation studies – where we make the case that understanding the population (whether nationally or in a state) isn’t simply a case of breaking them up into demographic categories but understanding their psychologies and drivers to their behaviors and decisions – which ultimately also help when thinking about messages and narratives.

By the way, the preference for simple responses and explanations is a human bias, playing into people’s tendency to default to what’s cognitively easier and confirms previous beliefs. Unfortunately (but really, fortunately), things are rarely so clear cut, especially when examining people and their perspectives at multiple levels.

Loker: I’m curious what conversations are like in your field about the challenge of showing complexity of identity and not appearing as if you’re minimizing the threat of people (policymakers or otherwise) trying to undermine democracy. News people face that, too. Do you have any insights that may be applicable to journalists?

Malvar: I think there is justified tension in the field of bridge-building/polarization space; and that this tension is necessary to push our work (I heard someone say once that this work is a full-contact sport, and that resonated). There is a distinction, though, between deliberate conflict entrepreneurs (and people in power who incentivize them) and the much bigger number of people who are prone to believing and supporting narratives that undermine democracy. Our work and research focus on the latter — understanding their motivations, their perceived lack of and need for belonging or respect, and how these can translate to support for undemocratic principles. 

Our approach to addressing anti-democratic narratives is to try to understand why these messages appeal to particular audiences. From there, we can find ways to de-escalate conflict and re-assert values of democracy in ways that resonate. 

Journalists are in a position to provide context for the public; they can focus on people’s personal narratives and complex identities, and they can probe the why behind people’s decision-making and the filters with which they arrive at their beliefs. In short, they can contribute to reducing perception gaps by not giving oxygen or legitimizing views that undermine democracy, but rather illuminating how people arrive at these views. 

People perceive themselves as so much more than their political affiliations.

Loker: Ok, so say you’re a journalist writing about people in power actively pursuing anti-democratic actions. If you want people to read that story, what would your work suggest journalists consider?

Malvar: A useful initial framing would be to explain or spell out what makes their actions anti-democratic, and what makes the issue urgent – referencing these with sources that are trusted across political parties. We know from our research that most Americans strongly support democracy as a system of government, and that democracy is a key thread of the perceived American fabric. So journalists can start by emphasizing this shared narrative, then frame anti-democratic actions as detrimental for everyone’s shared goal.

It might also be helpful to veer away from an exclusive Democrats vs. Republicans framework. Politics is almost always framed, unwittingly or not, as a team sport, but we know that there are significant portions of the country that are disengaged, ideologically inconsistent, or simply ignored. And people perceive themselves as so much more than their political affiliations. As psychologist Jay van Bavel said in his Twitter post, identities are bundles of sticks. A Dems vs. Reps framing activates a social identity that relates to a different aim (winning) than more constructive goals that bolster democracy, or in this case, calling out actions that work against it.

Finally, narratives that elevate the human story and describe the impact of anti-democratic actions on people’s lives can be powerful. We know from our work on election integrity that the workings of democracy are felt acutely at the local levels; a productive lens can be to describe the effects or implications of anti-democratic actions at the local community level, and how this ultimately affects individual American lives.

Loker: I think I have just two questions left. First, how might existing institutional distrust affect attempts to do any of this? 

Malvar: Unfortunately, institutional distrust is the reality we are currently operating under. Not just distrust toward journalists or the media, but towards experts, science, and government institutions – exacerbated by this seemingly endless pandemic. I think that last factor matters, because people are just in a psychological state right now where sense-making is ever challenging and the schema for what is normal life is shifting constantly.

One helpful way to think through this is to qualify the type of distrust at play. There is the us-vs-them distrust that drives much of the current polarization. I think journalists can mitigate this through framing differences beyond party affiliation when relevant, contextualizing datapoints with human stories, and framing the effects of anti-democratic actions — regardless of the source — as ultimately detrimental at the individual, in-group, and national level.

The other type of distrust, social distrust, is one that is rooted in experiences of belonging, dignity, and equality. People have to feel like their communities are represented fairly and that they are represented with dignity in the stories they read. There’s a great recent piece in CJR that in one section talks about how journalism is currently structured in such a way that the people writing the most well-read stories are in significantly different contexts and have completely different lived experiences from the people and communities they are writing about. And I’ve been thinking about that from a researcher standpoint as well. So I think to answer your question, this type of social distrust can be addressed by research and journalism that is participatory and community-driven and frames truth and accuracy as a public, not partisan good.

Loker: Lastly, a thought experiment: Say you were well-resourced and could start something new, and you wanted to position it to address perception gaps and related issues. If you were to start a More in Common-inspired newsroom, what would be its marks? What are its beats? How does it inform its story selection? How does it reach a wide range of people?

Malvar: We would report on stories that highlight instances of social cohesion operating at the interpersonal, group/communities, and macro level. For instance, stories that illustrate healthy conflicts and/or describe collective settings that allow not just for a deep description of the problem, but also highlight productive actions or potential pathways to solutions. By the way, also differentiating these from stories describing truly anti-democratic behaviors.

An MIC-inspired newsroom would be more deliberate in selecting journalists and in the relationship-building between journalists and the people whose stories they report on, as well as its contextualization of Americans with opposing views. As for the reach, I think there’s currently a need and hunger for local news in formats accessible for audiences that are less politically engaged, though there is also the danger of disinformation in grassroots initiatives when the ethics or mission isn’t clear. So the incentives of this newsroom would have to be around building a shared aspirational vision – not of winning in an us-vs-them context, but of advancing the idea that accurate information and truth bolsters democracy and benefits us all.


Trusting News, a project of the Reynolds Journalism Institute and API that helps journalists earn trust and demonstrate credibility, recently began a new initiative: the “Road to Pluralism.” If you’re a journalist interested in the themes in the above interview, consider getting involved.

You might also be interested in: