A conversation begins over dinner at a friend’s house. People lean in, laughing, swapping headlines and half-remembered facts. Someone brings up a recent policy change, say, a local housing law. Heads nod. There’s general agreement: housing is a mess. The room feels warm, collaborative. But then, one guest makes a sharper comment: “The city’s just pandering to developers.” Another adds, “Typical—corporate greed always wins.” Tones start to shift. Voices grow firmer. Someone challenges, then doubles down. Another retorts, slightly louder. Lines begin to form. Sides solidify. What started as shared frustration turns into ideological hardening. By dessert, you’re no longer talking about housing. You’re defending worldviews. This is polarization in real time, an ordinary conversation morphing into division.

In psychology, polarization describes how attitudes and beliefs can grow more extreme over time, especially in groups where similar views are shared. What begins as moderate disagreement often hardens into rigid opposition. At the core is a simple idea: agreement creates confidence, and confidence pushes people further in the direction they already lean. Think of it like a group of hikers moving uphill. At first, the trail is steep and uncertain. They encourage each other to keep going. With shared effort, the pace quickens and confidence grows. But in their eagerness, they stop checking the map. They follow the group instead of the signs. Soon, they’re far off the original path—deeper into the woods, harder to retrace their steps. The terrain grows rougher, and disagreement about direction leads to even more splintering. They’re not just lost—they’re entrenched.

Inside our minds, this development is shaped by how we process information. Human beings like patterns. We like knowing who we are and what we stand for, so we look for signs that we’re right. When we find confirmation, it feels good. When we encounter disagreement, it feels uncomfortable, even threatening. Over time, we unconsciously start to curate our mental environment. We read, watch, and follow content that aligns with what we already believe. We filter out what doesn’t.¹

This isn’t a flaw. It’s a function. Our minds simplify the world to make it manageable. But those simplifications come at a cost: we stop seeing nuance. And once nuance disappears, so does the possibility of healthy disagreement. Group environments intensify this pattern. When people with shared beliefs come together, they amplify one another.² Small shifts become big leaps. Each member adjusts slightly to match the group, and the group adjusts to stay ahead. It’s like tuning a radio higher and higher: eventually, all you hear is the same note, louder than ever.

This dynamic doesn’t just affect what people think. It changes how they think. When surrounded by agreement, individuals feel more certain. That certainty reduces their willingness to question assumptions. It also increases emotional investment in the issue at hand. A disagreement no longer feels like a difference in opinion; it feels like a threat to identity. Technology accelerates this process. Online platforms are built to maximize engagement. They feed us content that keeps us scrolling, and what keeps us scrolling is emotion. The more provocative the post, the more likely we are to react. This creates echo chambers, where the only voices we hear are the ones that already agree with us. In those spaces, empathy withers.³

At a certain point, disagreement no longer sounds like an alternative view—it sounds like betrayal. People on the “other side” become caricatures, not individuals. Their motives are questioned. Their intelligence is doubted. Dialogue gives way to defensiveness, and curiosity is replaced by certainty. This is the danger of polarization: not that people hold strong beliefs, but that they lose the capacity to understand why others might hold different ones.

Still, polarization isn’t irreversible. Minds are plastic. Beliefs are shaped and reshaped by experience. Exposure to different perspectives, structured dialogue, and a willingness to be challenged can reopen the space for change. But that process takes effort. It means sitting at tables where not everyone agrees. It means holding space for discomfort without retreating into certainty. At its heart, polarization is about belonging. It offers people a sense of identity and place. To address it, we need more than facts; we need new ways of connecting, new narratives that hold tension without tearing. Even when we’ve wandered far off course, standing at what feels like a cliff edge, there’s always a way back. Sometimes, it just takes looking up, reorienting, and finding the path again, together.

Group Polarization: The tendency for a group’s dominant opinion to intensify after discussion. When like-minded individuals talk, they reinforce each other's views, making everyone more confident and more extreme. This process helps explain how social groups radicalize or why committees can make riskier decisions than individuals acting alone.

Confirmation Bias: The cognitive tendency to seek out, notice, and remember information that supports our existing beliefs and discount evidence that contradicts them. In polarized environments, confirmation bias helps people maintain their convictions and avoid exposure to opposing perspectives.

Motivated Reasoning: Occurs when people interpret information in ways that protect their existing identity or emotional investments. Instead of evaluating facts neutrally, we judge them based on whether they support our group or self-image. This helps polarization persist, even in the face of clear evidence to the contrary.

Echo Chamber: A social or digital environment where individuals only encounter ideas and opinions that reflect and reinforce their own. In echo chambers, alternative views are filtered out, ridiculed, or mistrusted. This environment accelerates polarization by limiting empathy and reinforcing in-group certainty.

Partisan Cues: Symbols, phrases, or endorsements that signal where a political group stands on an issue. These cues help people decide what to believe without evaluating the details. For example, if a well-known conservative leader supports a policy, conservative voters may adopt the same view, even without knowing much about it. These cues can be as overt as party logos or as subtle as a politician’s talking points. Instead of weighing every argument, people often lean on these signals to align with their group.

The first systematic articulation of group-driven extremity emerged in the late 19th century. In 1895, Gustave Le Bon, a French sociologist and early theorist in crowd psychology, published The Crowd: A Study of the Popular Mind. He argued that individuals, once part of a group, lose their rational autonomy and become swept up in a collective emotional current that can heighten aggression, impulsivity, and suggestibility.⁵ While his work was speculative and lacked empirical rigor, Le Bon’s core idea, that groups intensify emotion, profoundly influenced later psychological theorists.

The scientific investigation of polarization began taking shape in the 20th century. In 1961, MIT researcher James Stoner conducted a now-famous experiment that asked participants to assess risk-taking scenarios individually and then in groups. For example, one prompt asked whether a fictional character should leave a stable job to pursue a riskier but potentially more rewarding career. Surprisingly, group decisions were consistently riskier than individual ones. This effect became known as the “risky shift” and marked a key inflection point in the study of group polarization.⁶

David Myers and George Bishop extended this research in the early 1970s. In a landmark study, they grouped participants based on their existing attitudes toward racial issues and then observed how group discussion affected their beliefs. Those with prejudiced views became more entrenched, while those with egalitarian views became more progressive.² Their work broadened the scope of polarization research from risk-based decisions to include moral, political, and social values, demonstrating that group discussion amplifies preexisting attitudes.

Simultaneously, scholars began digging into the individual cognitive mechanisms behind this phenomenon. A major breakthrough came in the 1970s with research on confirmation bias. Ross, Lepper, and Hubbard showed that people not only favor information that aligns with their beliefs but also maintain those beliefs even after the original supporting evidence is debunked.⁷ This persistence, known as belief perseverance, revealed how humans are motivated to preserve cognitive consistency rather than challenge their own viewpoints.

The 1980s and 1990s introduced new complexity with the concept of motivated reasoning. Ziva Kunda theorized that individuals actively interpret evidence to support desired conclusions, especially when identity is at stake.¹ Instead of neutrally evaluating facts, people unconsciously recruit mental justifications to maintain alignment with their values and in-groups. This explains why two people can look at the same data and arrive at completely different interpretations, not due to ignorance but due to internal bias filters shaped by emotion and identity.

Technological advances dramatically escalated these tendencies. Legal scholar Cass Sunstein wrote in his book, Republic: Divided Democracy in the Age of Social Media, that the internet’s personalization features would create echo chambers, where users encountered only information that aligned with their views.⁸ This architecture wouldn’t just confirm existing biases; it would isolate people from opposing perspectives and discourage open dialogue. Sunstein’s concerns gained empirical support in the age of social media. A study by Brady and colleagues analyzed millions of Twitter posts and found that content containing moral-emotional language, phrases invoking strong feelings of anger, disgust, or moral violation, was significantly more likely to be shared.3 These emotionally charged messages didn’t promote discussion; they deepened tribal lines and reinforced in-group allegiance.

The real-world effects of these dynamics became increasingly visible in politics. During the 2004 U.S. election cycle, researchers documented rising geographic polarization, as voters began clustering into ideologically homogeneous communities.⁹ People were not just voting differently, they were living, working, and socializing within political echo chambers. This phenomenon, known as “The Big Sort,” reflected a society in which social identity and political ideology had become deeply intertwined.

Today, polarization is considered a systemic issue, affecting democracy, health, and education. It shows up in vaccine hesitancy, climate change denial, and growing mistrust of institutions. Psychologists, political scientists, and behavioral economists now collaborate to understand how identity, emotion, information ecosystems, and cognitive biases interact.¹⁰ While technology accelerated the pace, the underlying dynamics of group polarization, confirmation bias, and echo chambers are rooted in long-standing human tendencies.

Despite the challenges, this history also offers a roadmap for intervention. Programs that encourage cross-group dialogue, media literacy, and self-awareness of bias have shown promise. Some researchers suggest “inoculation” techniques: exposing people to weakened versions of misinformation so they can build resistance, like a vaccine for the mind.¹¹ Others are exploring algorithmic transparency and content curation reforms to reshape digital spaces that currently reinforce polarization. As we navigate this polarized world, understanding where the problem comes from is the first step toward solving it. 

Gustave Le Bon

A pioneering thinker in mass psychology. In The Crowd: A Study of the Popular Mind, he argued that individuals in crowds lose their sense of self and become emotionally suggestible.⁵ He believed this loss of individuality made groups more extreme, impulsive, and volatile. Although his work lacked empirical evidence, Le Bon's ideas foreshadowed modern concepts of group polarization by describing how collective dynamics intensify behavior. His writing influenced early 20th-century political leaders and continues to be referenced in discussions on mass influence.

James A.F. Stoner

A professor of Management Systems at Fordham University, Stoner launched the first empirical study of group decision-making in the context of risk in the 1960s.⁶ His discovery of the “risky shift” revealed that groups tend to make riskier decisions than individuals, overturning the assumption that collaboration leads to moderation. This phenomenon highlighted how consensus can distort risk perception and escalate behavior. Stoner's research introduced the idea that social settings magnify opinion intensity, providing a launching point for future investigations into polarization. His methodology also set a precedent for rigorous experimental design in decision psychology.

David G. Myers

In the 1970s, social psychologist David Myers expanded Stoner’s findings to show that group discussion amplifies attitudes, not just risk preferences. He conducted studies on racial attitudes, showing that after group dialogue, people’s views became more extreme in the direction they initially leaned.² This effect held across various domains, from politics to morality. Myers coined the term group polarization, framing it as a central mechanism of modern social division. His work is still foundational in both academic research and public conversations about ideological echo chambers.

Ziva Kunda

In the 1990s, Israeli social psychologist Ziva Kunda transformed the field with her theory of motivated reasoning, which revealed how people shape their thinking to protect desired beliefs.¹ She argued that cognition isn’t passive or neutral; it's shaped by emotional and identity-driven goals. Her work showed that people unconsciously gather, interpret, and remember information in ways that defend what they want to believe. This helped explain why facts often fail to sway minds in polarized environments. Kunda's insights remain a cornerstone of modern cognitive and political psychology.

Cass Sunstein 

In the early 2000s, legal scholar and behavioral economist Cass Sunstein warned that digital technologies would fragment public discourse.⁸ In Republic.com, he predicted that the internet would create echo chambers, spaces where people only encounter views they already agree with. Without dissenting perspectives, Sunstein argued, democratic dialogue weakens and polarization deepens. Once theoretical, his concerns have since been backed by numerous studies on social media dynamics.

Polarization isn’t just a buzzword on the nightly news; it affects how we learn, heal, vote, scroll, and speak. From the classrooms where students tiptoe around divisive topics to the hospital rooms where trust in public health is wearing thin, and even the algorithms that shape what we see online, polarization quietly reshapes our daily reality. If you’ve ever felt like people live in “two different worlds” despite sharing the same country or even the same home, polarization may be the unseen architect.

Classrooms under strain

You might expect schools to be places where differing ideas flourish, but in a polarized world, even facts can become political. A 2022 Education Week survey found that a majority of U.S. teachers feel pressured to avoid controversial subjects like politics in the classroom.¹³ Many fear being reprimanded by school boards or publicly criticized by parents. For some, teaching the wrong word or even displaying a certain book title could mean disciplinary action. As a result, lesson plans are increasingly sanitized, creating sanitized echo chambers that offer comfort but little challenge.

This environment contributes to the growth of what researchers call epistemic polarization: the belief that truth is not objective, but defined by group identity. A 2024 study on political development showed that those in polarized environments exhibited lower intellectual humility, decreased tolerance for ambiguity, and greater resistance to evidence.¹⁴ These traits aren't just abstract psychological constructs; they manifest as classroom behaviors such as avoidance of dissent, rote memorization over inquiry, and personal attacks during discussion.

Even digital learning environments aren’t immune. Some school platforms allow parents to pre-approve materials or opt their children out of certain topics entirely, reinforcing the idea that education is something to be filtered through ideological preference rather than broadened by diversity. Meanwhile, online political discourse often spills into student debates, leading to digital pile-ons or anxiety over participation.

Even in a climate thick with division, change is quietly taking root. At Miami University in 2024, more than 100 student leaders, from College Democrats and Republicans to Students for Justice in Palestine, gathered not to win arguments, but to sharpen a different skill: listening. Through a partnership with the Constructive Dialogue Institute, they engaged in Perspectives, an immersive program teaching students how to recognize bias, empathize across divides, and engage in respectful disagreement.¹⁵ No one was told to compromise their values. Instead, they learned how to sit with discomfort and resist the urge to villainize. What emerged wasn’t consensus, but something deeper: a renewed capacity for civil engagement. Students left feeling more confident navigating tough conversations, and the university has now committed to expanding the program. In a world where disagreement often feels dangerous, Miami’s initiative shows that polarization isn’t a permanent condition; it’s a skill gap we can close.

Health & misinformation

During the COVID-19 pandemic, science met skepticism in ways few public health officials could have anticipated. Recommendations on masks, vaccines, and social distancing became proxy battles for deeper ideological conflicts. Political identity, more than medical vulnerability, shaped people’s behavior. A 2024 master’s thesis from the University of Manitoba found that belief in COVID-related vaccine myths was strongly correlated with political and religious affiliation in both the U.S. and Canada.¹⁶ This pattern persisted even when individuals were presented with evidence-based corrections.

The impact was profound. Research from the Yale School of Public Health revealed that politically polarized U.S. counties had statistically higher COVID-19 mortality rates, even when controlling for income, access to hospitals, and age.¹⁷ In short, polarization itself became a risk factor. Individuals didn't just reject facts; they rejected the authority behind those facts. Motivated reasoning—the tendency to evaluate information in a way that confirms pre-existing beliefs—further hardened positions, leading to what researchers describe as identity-protective cognition.

Efforts to reverse these effects have taken a page from psychology. Inoculation theory, which likens misinformation defense to building a vaccine against falsehood, has gained traction in public health circles. Programs like Canada’s “Science Up First” and the WHO’s infodemic task force aim to “prebunk” false claims before they spread, using storytelling, graphics, and trusted community voices.¹⁸ These messages work not by overpowering misinformation, but by introducing small “doses” of fallacy that trigger cognitive resistance.

And yet, the challenge remains daunting. Anti-vaccine influencers generate millions of views with emotional anecdotes, while fact-checking sites struggle to keep pace. Public trust in health institutions has declined markedly among highly polarized populations. According to a 2023 study, only a limited number of respondents who strongly identified with partisan political identities expressed “a great deal of trust” in medical experts.¹⁹ As polarization continues to shape who we trust, even objective risk becomes subject to ideological interpretation.

The algorithmic divide

You scroll. You click. You pause. And the machine watches. Social media algorithms don’t just reflect our preferences; they shape them. Over time, they learn what keeps us scrolling: content that confirms, excites, and outrages. This is the backbone of selective exposure: our tendency to seek out information that aligns with our beliefs.²⁰ It’s different from confirmation bias, which kicks in after we’ve seen something, making us interpret it in a way that supports what we already believe. Selective exposure is about what we choose to look at in the first place. And online, the stakes are higher. Instead of quiet reinforcement, we get turbocharged echo chambers.

A 2019 study in Science found that false political claims spread faster and farther on Twitter when they resonated with the user’s existing ideology.²¹ These posts weren’t driven by bots; they were shared by real people acting on gut feeling, especially when messages included moral-emotional language like betrayal, injustice, or virtue signaling. The more provocative the post, the more likely it was to be amplified.

Some platforms have attempted to disrupt this cycle. Facebook introduced “context labels” for disputed posts. Twitter experimented with “read before retweet” prompts. YouTube has quietly downranked conspiracy-adjacent channels in its algorithm. And emerging platforms like Tangle and Ground News offer “bias comparison” tools that let users see how headlines differ across ideological lines. But such reforms remain voluntary, and user opt-in is low.

Meanwhile, the real-world implications are multiplying. In 2023, the Edelman Trust Barometer reported that users with strong partisan leanings were 40% more likely to distrust journalism, science, and even the judicial system.²² This distrust corrodes the connective tissue of society, shared facts. Without them, deliberation becomes difficult, compromise impossible, and extremism normalized. In the digital age, what you see increasingly determines what you believe, and who you’re willing to believe it with.

We often talk about polarization as if we all agree on what it is and how to fix it. But beneath the surface lies a battleground of ideas. Researchers, psychologists, technologists, and educators disagree not just on what drives polarization, but on whether our current solutions are working—or in some cases, making the problem worse. These are not academic squabbles. They influence the design of social media, the shape of public discourse, and the way we raise the next generation of citizens. Let’s dive into three of the most pressing debates in polarization research.

Is social media fuel or a mirror?

We’ve all seen the echo chambers: heated comment threads, personalized feeds, and memes that make nuance feel like an endangered species. But is social media creating polarization, or simply reflecting deeper social divides?

Many researchers argue that social media platforms don’t just reflect polarization, they stoke it. Their algorithms are built to maximize engagement, and nothing fuels clicks like outrage. One study put this to the test. Researchers asked Twitter users to follow bots that regularly shared posts from the opposite political side.²³ The idea? That exposure to diverse views might soften extreme positions. But that didn’t happen. Instead, people dug in. The more they saw opposing arguments, the more entrenched they became.

In parallel, other researchers have used simulations to model how algorithms sort content.²⁴ Over time, these virtual ecosystems showed a consistent pattern: people ended up in tighter ideological bubbles, surrounded by content that reinforced their beliefs and shielded them from dissent. These findings suggest that polarization isn’t just a social trend; it’s baked into the way our platforms work.

But here’s where the picture gets more complicated. If social media causes polarization, then we’d expect the most online generation, young adults, to be the most polarized. But they aren’t. In fact, researchers have found the sharpest rise in polarization among older adults, the same group that uses social media the least.²⁵ This suggests that something else is fueling division. Traditional media like cable news and partisan talk radio still carry enormous weight, especially among older audiences. Political leaders and party elites also play a powerful role in shaping how people see the world—and who they blame. Digital platforms may add fuel, but they’re not the fire itself. As sociologist Shamika Alam puts it, online spaces amplify what we already believe, but they rarely invent those beliefs from scratch.²⁶

So, which is it? A digital accelerant or a cultural mirror? The answer likely lies somewhere in between. Social media may not invent tribalism, but it certainly gives it steroids, pushing us toward more extreme versions of ourselves. Understanding that distinction matters. If tech is the driver, regulation and platform redesign may be the cure. But if it’s a reflection of deeper civic fractures, then media literacy, institutional reform, and face-to-face dialogue must be part of the solution too.

Does talking across divides help or hurt?

It sounds like common sense: when people talk to each other, they understand each other. But when it comes to deeply held values, that’s not always how it plays out. Can dialogue reduce polarization, or can it make things worse?

The hope behind initiatives like Braver Angels and the Constructive Dialogue Institute stems from a powerful psychological idea: Allport’s Contact Hypothesis. It suggests that structured, respectful contact between opposing groups can reduce prejudice. But does it actually work? In a landmark study, political scientists David Broockman and Joshua Kalla decided to find out.²⁷ They sent canvassers door to door in South Florida, asking 501 voters to engage in ten-minute, face-to-face conversations about transgender rights. The approach was gentle, nonjudgmental, deeply personal, and rooted in shared experiences. Remarkably, it worked. Even months later, participants showed lasting reductions in transphobic attitudes.

These findings sparked a movement toward deeper dialogue. In classrooms, community centers, and workplaces, groups are now experimenting with structured conversation programs. One such effort, the Perspectives course by the Constructive Dialogue Institute, used guided exercises to bring politically diverse college students into meaningful exchange. In a 2022 trial with 755 students, nearly three-quarters came away with significantly reduced signs of polarization.²⁸

If the conditions aren’t right, dialogue can backfire. Druckman et al. (2021) found that exposure to opposing views can increase polarization if people feel morally judged or socially threatened.²⁹ Known as the “backfire effect,” this phenomenon shows that people double down on beliefs when their identity feels attacked. Think about your last heated debate: did it change your mind, or just make you dig in harder?

There’s also growing concern about how online discourse mimics but doesn’t replicate real-world dialogue. Digital spaces often remove tone, context, and accountability—the very cues that help dialogue succeed. Even when intentions are good, comments can be misinterpreted, sarcasm mistaken for scorn, and debate can spiral into hostility.³⁰ In short, contact isn’t a cure-all. It’s a tool, and like any tool, it needs structure, intent, and careful use. When done well, dialogue fosters humility and connection. Done poorly, it builds walls instead of bridges.

Is polarization always a bad thing?

We often hear that polarization is ripping society apart. But some scholars argue that conflict, when rooted in values, can be necessary, even healthy. So, is polarization always a threat to democracy, or could it be one of its lifelines?

Supporters of this idea highlight that polarization isn’t just about anger or tribalism. Sometimes, it’s about moral clarity. Lilliana Mason (2018) argues that group identity can serve pro-social goals, creating solidarity, motivating action, and giving voice to the marginalized.³¹ After all, many of history’s most important social movements, from civil rights to women’s suffrage, were fiercely polarizing in their time. They forced uncomfortable conversations that society needed to have.

Critics point to a deeper kind of division: affective polarization. In the U.S., this shift has become particularly stark. In a widely cited 2015 study, political scientists Shanto Iyengar and Sean Westwood asked Americans to rate people based on party affiliation.³² The results were sobering: participants consistently viewed members of the opposing party as less intelligent, less trustworthy, and even less patriotic than their own. This tribal mindset bleeds into everyday life. It shapes hiring decisions, dating preferences, and even neighborhood choices. You’re not just disagreeing with someone, you’re avoiding them.

And the distrust runs deep. The Edelman Trust Barometer, an annual global survey that measures institutional trust across countries, shows that Americans’ trust in those with opposing political views has dropped to historic lows.²² When people stop believing the other side is acting in good faith, it’s not just relationships that suffer. The very foundations of democratic life, dialogue, compromise, and cooperation start to erode.

So is polarization the enemy, or the energy? Maybe the real issue isn’t disagreement, but how we disagree. When polarization reflects real moral stakes and is accompanied by respectful dialogue, it can drive change. But when it’s driven by fear, misinformation, and identity-based animosity, it tears at democracy’s seams.

Logging off to cool down

If you’ve ever wondered what would happen if Americans collectively logged off Facebook and Instagram during a major election, this study gives us a rare glimpse into that alternate universe, and what it reveals about polarization is more nuanced than you might expect.

In 2024, a team of researchers—Allcott, Gentzkow, Tucker, and colleagues—ran one of the largest social media experiments in political science history.³³ Over 19,000 Americans were paid to log off Facebook and Instagram during the height of the 2020 U.S. presidential race. From late September through Election Day, participants unplugged from two of their most-used digital spaces, and researchers tracked what happened next.

To measure the effects, the team used a mix of surveys and behavioral data, monitoring voter turnout records, news consumption patterns, and changes in political attitudes. What they found was nuanced. First, those who deactivated learned and engaged less. They read less news, knew fewer political facts, and were slightly less likely to vote. For many, social media wasn’t just a distraction; it was their main connection to the civic world.

But there was a twist. The same group also reported softer attitudes toward the opposing political party. In psychological terms, their affective polarization declined. They weren’t switching sides, but they were cooling down. Logging off didn’t make people apolitical, but it did seem to take some of the heat out of partisan hostility. The result? A paradox. Stepping away from social media made people a little less informed, but also a little less inflamed. Crucially, the researchers emphasize that the reductions in polarization were small and accompanied by a decrease in overall political engagement. Users reported talking less about politics, caring less about the issues, and feeling less motivated to vote. This raises a tough question: Can we reduce polarization without also dampening civic engagement?

The authors offer a careful interpretation: social media is a double-edged sword. It connects us to important civic conversations but also amplifies outrage, misinformation, and partisan narratives. Their study doesn’t support simple calls to “log off.” Instead, it points to the need for platform design reform. Encouraging diverse viewpoints, improving content quality, and reducing algorithmic amplification of divisive content may be more productive than digital abstinence.

This real-world case study underscores a critical insight: polarization is not just about people, it’s about the systems that shape what they see, feel, and believe.

Vaccine messaging, media bubbles, and public health in a pandemic

When COVID-19 vaccines rolled out in early 2021, public health officials hoped the end of the pandemic was in sight. But they hadn’t accounted for an invisible contagion spreading just as fast: polarization. As it turned out, beliefs about a needle in the arm had less to do with science and more to do with what news you watched, what party you voted for, and which influencers filled your feed.

A 2022 study in Nature Human Behaviour by Pennycook and colleagues examined over 18,000 U.S. participants to explore how media habits influenced vaccine attitudes.³⁴ They found that those who relied on right-leaning news sources were significantly more likely to express vaccine hesitancy, not due to misunderstanding the science, but because of ideological cues baked into the coverage. These included loaded language ("government overreach"), repeated emphasis on personal freedom, skepticism of pharmaceutical companies, and the strategic framing of mandates as partisan threats rather than public health tools.

To test how people might be nudged away from misinformation, researchers designed a simple online experiment. Before participants read or shared COVID-19 news headlines, some were shown a quick prompt: “Before you share, think about how accurate this is.” That was it, no fact-checks, no corrections, just a gentle reminder to pause and reflect. And it worked. People who saw that short message were over 25% more likely to choose reliable news articles to share, compared to those who didn’t see the prompt. Instead of measuring what people actually posted online, the researchers gave them a list of headlines and asked which ones they’d be willing to share. Those primed to think about accuracy picked more trustworthy ones. A small shift in attention made a measurable difference.

But the backlash was swift. Critics said interventions like this felt manipulative, like platforms were trying to steer people’s thinking. Some accused Facebook and YouTube of censorship when they added fact-check labels or downranked misinformation. In polarized environments, even truth itself became political. For some, being corrected felt like being controlled.

In highly polarized environments, being corrected can feel like being silenced, especially when the correction comes from a source the user already distrusts. A label that says “false information” doesn’t just inform, it signals a power dynamic. It can trigger defensiveness, reinforce group identity, and backfire. Psychologists call this the reactance effect, when people push back against efforts to influence them, not because they disagree, but because they feel their freedom is being threatened. And in a climate where every claim is suspect and every label feels loaded, even truth becomes tribal. Fighting the social narratives around it? That’s a whole different kind of contagion.

False Consensus Effect

Ever assume most people agree with you, just because the people around you do? The False Consensus Effect explains why we tend to overestimate how common our beliefs are, especially in tightly-knit social or ideological groups. In a polarized world, this bias can create dangerous illusions of consensus and reinforce “us vs. them” mentalities. This piece explores the psychological roots of the effect, its political and cultural consequences, and how recognizing it can help foster humility and perspective-taking.

Confirmation Bias

We tend to see what we expect and ignore what challenges us. This BSRG entry breaks down the psychological mechanics of confirmation bias: the brain’s tendency to prioritize information that affirms our preconceptions. In a polarized world, it’s a major fuel source. You’ll learn how this bias drives everything from selective media consumption to political entrenchment, and what behavioral interventions might help challenge it.

1. Kunda, Z. (1990). The case for motivated reasoning. Psychological Bulletin, 108(3), 480–498. https://doi.org/10.1037/0033-2909.108.3.480
2. Myers, D. G., & Bishop, G. D. (1970). Discussion effects on racial attitudes. Science, 169(3947), 778–779. https://doi.org/10.1126/science.169.3947.778
3. Brady, W. J., Wills, J. A., Jost, J. T., Tucker, J. A., & Van Bavel, J. J. (2017). Emotion shapes the diffusion of moralized content in social networks. Proceedings of the National Academy of Sciences, 114(28), 7313–7318. https://doi.org/10.1073/pnas.1618923114
4. Myers, D. G., & Lamm, H. (1976). The group polarization phenomenon. Psychological Bulletin, 83(4), 602–627. https://doi.org/10.1037/0033-2909.83.4.602
5. Le Bon, G. (1895). The crowd: A study of the popular mind. Fischer.
6. Stoner, J. A. F. (1961). A comparison of individual and group decisions involving risk. MIT Sloan School of Management.
7. Ross, L., Lepper, M. R., & Hubbard, M. (1975). Perseverance in self-perception and social perception: Biased attributional processes in the debriefing paradigm. Journal of Personality and Social Psychology, 32(5), 880–892. https://doi.org/10.1037/0022-3514.32.5.880
8. Sunstein, C. R. (2001). Republic.com. Princeton University Press.
9. Bishop, B. (2008). The big sort: Why the clustering of like-minded America is tearing us apart. Houghton Mifflin Harcourt.
10. Westen, D., Blagov, P. S., Harenski, K., Kilts, C., & Hamann, S. (2006). The neural basis of motivated reasoning: An fMRI study of emotional constraints on partisan political judgment in the 2004 U.S. Presidential election. Journal of Cognitive Neuroscience, 18(11), 1947–1958. https://doi.org/10.1162/jocn.2006.18.11.1947
11. Van der Linden, S., Roozenbeek, J., & Compton, J. (2020). Inoculating against fake news about COVID-19. Frontiers in Psychology, 11, 566790. https://doi.org/10.3389/fpsyg.2020.566790
12. Monin, B., & Jordan, A. H. (2009). The dynamic moral self: A social psychological perspective. In D. Narvaez & D. K. Lapsley (Eds.), Personality, identity, and character: Explorations in moral psychology (pp. 341–354). Cambridge University Press. https://doi.org/10.1017/CBO9780511627125.016
13. Schwartz, S.. (2023, June 27). Most teachers want politicians to stay out of their classroom decisionmaking. Education Week. https://www.edweek.org/teaching-learning/most-teachers-want-politicians-to-stay-out-of-their-classroom-decisionmaking/2023/06
14. Dawson, C., Julku, H., Pihlajamäki, M., & Svedholm-Häkkinen, A. M. (2024). Evidence-based scientific thinking and decision-making in everyday life. Cognitive Research: Principles and Implications, 9(1), Article 5. https://doi.org/10.1186/s41235-024-00578-2
15. Kissell, M. R. (2024, April 3). More than 100 student leaders participate in Constructive Dialogue Perspectives. Miami University News. https://miamioh.edu/news/2024/04/more-than-100-student-leaders-participate-in-constructive-dialogue-perspectives.html
16. Hallberg, A. (2024). How religious and political affiliation influence belief in COVID-19 vaccine myths in Canada & the United States (Master's thesis, University of Manitoba). University of Manitoba MSpace. https://mspace.lib.umanitoba.ca/items/08961df9-1e76-46a5-9e78-beed34953d33
17. Wallace, J., Goldsmith-Pinkham, P., & Schwartz, J. L. (2023). Excess Death Rates for Republican and Democratic Registered Voters in Florida and Ohio During the COVID-19 Pandemic. JAMA internal medicine, 183(9), 916–923. https://doi.org/10.1001/jamainternmed.2023.1154
18. Sonnac, N., & Smadja, D. M. (2025). Rebuilding trust in public health: Countering medical disinformation as a democratic value in the digital age. Journal of Pharmaceutical Health Services Research, 16(1), rmaf005. https://doi.org/10.1093/jphsr/rmaf005
19. Millar, B. (2023). Partisan epistemology and misplaced trust. Episteme. Cambridge University Press. https://doi.org/10.1017/epi.2023.46
20. Stroud, N. J. (2010). Polarization and partisan selective exposure. Journal of Communication, 60(3), 556–576. https://doi.org/10.1111/j.1460-2466.2010.01497.x
21. Grinberg, N., Joseph, K., Friedland, L., Swire-Thompson, B., & Lazer, D. (2019). Fake news on Twitter during the 2016 U.S. presidential election. Science (New York, N.Y.), 363(6425), 374–378. https://doi.org/10.1126/science.aau2706
22. Edelman. (2023). 2023 Edelman Trust Barometer. Edelman. https://www.edelman.com/trust/2023/trust-barometer
23. Bail, C. A., Argyle, L. P., Brown, T. W., Bumpus, J. P., Chen, H., Hunzaker, M. F., ... & Volfovsky, A. (2018). Exposure to opposing views on social media can increase political polarization. Proceedings of the National Academy of Sciences, 115(37), 9216–9221. https://doi.org/10.1073/pnas.1804840115
24. Chueca Del Cerro, C. (2024). The power of social networks and social media's filter bubble in shaping polarisation: An agent-based model. Applied Network Science, 9, Article 33. https://doi.org/10.1007/s41109-024-00679-3
25. Boxell, L., Gentzkow, M., & Shapiro, J. M. (2017). Greater internet use is not associated with faster growth in political polarization among US demographic groups. Proceedings of the National Academy of Sciences, 114(40), 10612–10617. https://doi.org/10.1073/pnas.1706588114
26. Alam, M. A. (2024). Social media and political polarization: Investigating the role of online platforms in shaping public opinion. Embar Publishers. https://www.embarpublishers.com/assets/articles/1724686809.pdf
27. Broockman, D., & Kalla, J. (2016). Durably reducing transphobia: A field experiment on door-to-door canvassing. Science, 352(6282), 220–224. https://doi.org/10.1126/science.aad9713
28. Constructive Dialogue Institute. (2022). Turning down the heat on campus: How an online educational program can reduce polarization and improve dialogue in college classrooms. https://constructivedialogue.org/articles/turning-down-the-heat-on-campus/
29. Druckman, J. N., Klar, S., Krupnikov, Y., Levendusky, M., & Ryan, J. B. (2021). Affective polarization, local contexts and public opinion in America. Nature human behaviour, 5(1), 28–38. https://doi.org/10.1038/s41562-020-01012-5
30. Coe, K., Kenski, K., & Rains, S. A. (2014). Online and uncivil? Patterns and determinants of incivility in newspaper website comments. Journal of Communication, 64(4), 658–679. https://doi.org/10.1111/jcom.12104
31. Mason, L. (2018). Uncivil agreement: How politics became our identity. University of Chicago Press.
32. Iyengar, S., & Westwood, S. J. (2015). Fear and loathing across party lines: New evidence on group polarization. American Journal of Political Science, 59(3), 690–707. https://doi.org/10.1111/ajps.12152
33. Allcott, H., Gentzkow, M., Mason, W., Wilkins, A., Barberá, P., Brown, T., Cisneros, J. C., Crespo-Tenorio, A., Dimmery, D., Freelon, D., González-Bailón, S., Guess, A. M., Kim, Y. M., Lazer, D., Malhotra, N., Moehler, D., Nair-Desai, S., Nait El Barj, H., Nyhan, B., Paixao de Queiroz, A. C., … Tucker, J. A. (2024). The effects of Facebook and Instagram on the 2020 election: A deactivation experiment. Proceedings of the National Academy of Sciences of the United States of America, 121(21), e2321584121. https://doi.org/10.1073/pnas.2321584121
34. Pennycook, G., Epstein, Z., Mosleh, M., Arechar, A. A., Eckles, D., & Rand, D. G. (2021). Shifting attention to accuracy can reduce misinformation online. Nature, 592(7855), 590–595. https://doi.org/10.1038/s41586-021-03344-2