Suddenly realizing his thirties are catching up to him, James decides to join a Facebook group about clean eating. As a busy family man who works full time and travels for work, he tends to grab takeout more often than he’d like. He figures some home-cooked meals for the road might help him shed a few pounds and get back into a fit lifestyle, setting a healthy example for his kids. 

At first, the group feels helpful. There are thousands of people going through similar struggles, swapping recipes, sharing label-reading tips, and celebrating small changes. But over time, the tone shifts. The practical advice he joined for—the kind that made getting healthy seem possible for normal folks like him—starts to fade. Posts about antibiotics and gut bacteria, inflammation caused by dairy, and hidden toxins in GMOs begin to dominate. Members who ask skeptical questions go quiet. The most popular voices speak with absolute certainty, and the tone becomes increasingly rigid and judgmental.

Six months later, the man who just wanted a few better habits has cut out meat and dairy entirely, avoids eating at friends’ houses, and argues with his spouse about what the kids should be allowed to eat. Sure, he lost weight and feels amazing—but at the expense of his closest relationships.

When we engage with others who share our outlook, we often come away with a stronger, more rigid version of our original view. Group polarization strengthens our oppositional stance, but it’s not always loud or dramatic. Often, it’s a quiet drift, fueled by validation, repetition, and the sense that those who hesitate to join others in their extremism might be missing something important.

Risky shift: A pattern first observed by James Stoner in the 1960s, where group decisions leaned more toward risk than individual ones. It became an early clue to help uncover the broader group polarization effect.

Cautious shift: The opposite of risky shift. When a group begins with a cautious mindset, discussion can push members toward even greater caution.

Social comparison: The evaluation of personal opinions and beliefs based on the perceived attitudes and positions of others. This can result in alignment with group norms, often leading individuals to adopt slightly stronger positions to maintain social standing.

Informational influence: A process where exposure to supporting arguments from like-minded individuals reinforces existing beliefs, making them feel more rational and justified.

Echo chamber: A social or digital environment where individuals are mostly exposed to views they already agree with, intensifying polarization.

Groupthink: A related but distinct concept, groupthink occurs when a group values harmony and consensus over critical thinking, often suppressing dissent to avoid conflict. While group polarization involves intensifying agreement, groupthink is about avoiding disagreement altogether.

Staying in sync with a group of like-minded people was a survival strategy for early humans. In tight-knit communities, expressing views that aligned with the group signaled loyalty and was therefore rewarded with safety. Showing conviction could build trust, and reinforcing shared beliefs became a way to maintain cohesion and reduce internal threats; the stronger the cohesion, the safer the group. This intense agreement among a group set the stage for what psychologists would later call group polarization.

The first systematic evidence of group polarization came in the early 1960s. Social psychologist James Stoner, a master’s student at the time, asked MIT students to make decisions about hypothetical dilemmas, like whether a fictional person with a life-threatening condition should undergo a risky operation or make healthier lifestyle choices.² When participants decided alone, they were moderately cautious. However, when they discussed the same dilemma in small groups, their final decisions leaned more strongly toward risk. This pattern became known as the risky shift. 

Follow-up studies by Marvin Zuckerman Wallach, Norman Kogan, and Daryl Bem showed that the shift wasn’t always toward risk.³ Sometimes it moved in the opposite direction, toward caution. The direction of the shift depended on the group’s initial lean, such as business students leaning towards risk and liberal arts students showing more restraint. When individuals began with a cautious stance, group discussion motivated a cautious shift, pushing them to become even more conservative. The core effect was less about risk—more about intensification. Group discussion seemed to move individuals further in whatever direction they were already inclined to go.

By the 1970s, psychologists such as Serge Moscovici and Marisa Zavalloni expanded on these findings and helped define the concept of group polarization.4 Their work showed that discussion among like-minded individuals tends to strengthen the shared position, making it more extreme and more confidently held. Researchers identified two reinforcing mechanisms that help explain the effect: social comparison and informational influence.⁵

We often assess our own views based on what others around us seem to believe. That’s social comparison, and it can lead us to take a stronger stance just to stay in line with the group. Then there’s informational influence: hearing arguments that support our beliefs makes them feel more solid, as if we have proof on our side. Together, these forces can make extreme views seem reasonable and necessary.

In later decades, group polarization moved out of the lab and into public discourse. Legal scholar Cass Sunstein applied the concept to political decision-making, showing how homogenous groups, whether in juries, government, or online spaces, often drift toward more extreme positions simply through internal discussion.⁶ He warned that when dissent is absent or discouraged, polarization becomes more likely—and more damaging.

Today, the concept of group polarization appears in studies of political ideology, organizational culture, online communities, and public health. Its dynamics shape how movements form, how conflicts escalate, and how shared beliefs harden into fixed identities. What began as a pattern in small-group decision-making now helps explain some of the most persistent and puzzling forms of collective behavior in modern society.

James Stoner

An American social psychologist based at MIT in the 1960s, Stoner conducted the first empirical studies that led to the identification of the "risky shift." His research showed that individuals tend to make more extreme decisions after group discussion, laying the foundation for what would become group polarization theory.

Marvin Zuckerman Wallach, Norman Kogan, and Daryl Bem

American psychologists whose 1962 study extended early research on group decision-making. Their work demonstrated that group discussion tends to amplify individuals’ leanings, reinforcing their initial direction. This helped lay the groundwork for the shift from the “risky shift” to the more general concept of group polarization.

Serge Moscovici and Marisa Zavalloni

Social psychologists—Romanian and Italian, respectively—who formally introduced the term group polarization in 1969. Working together in France, they demonstrated that group discussion tends to push shared attitudes toward more extreme positions, especially on emotionally or politically charged topics. Their research provided the theoretical structure that helped define group polarization as a distinct phenomenon.

Cass Sunstein

An American legal scholar and behavioral economist. Formerly at the University of Chicago and later a professor at Harvard Law School, Sunstein brought group polarization into public discourse with his research on law, politics, and online behavior. His work connected the theory to real-world issues like political echo chambers and policy decision-making.

When we spend time in groups that share our views, our confidence tends to grow. We become more certain, less questioning, and more likely to dismiss information that doesn’t fit with the prevailing narrative. These shifts can influence how we make decisions, how we listen (or don’t listen) to others, and how we respond when we feel uncertain. Over time, what starts as agreement can turn into a conviction harder to step back from—especially in spaces where dissent is rare or unwelcome.

Close relationships

Group polarization often plays out in our closest circles—in conversations with family, friends, or trusted peers. As we spend more time with people who see the world like we do, our views tend to sharpen. We may become more certain, less open to questioning, and more likely to feel distant from those who think differently—even when there was once a strong connection. Disagreements that once felt like part of a healthy relationship can become harder to move through.

This kind of disconnect can show up even in quick conversations. In a study involving high school students, people clustered into groups based on similar racial opinions and spent just two minutes discussing each prompt on race. After the conversation, members of each group showed even stronger versions of their original views.⁷ It doesn’t take long, drawn-out conflict to create distance. Brief, quiet agreement might be all it takes to push us further apart.

Workplace and team environments

In team settings, ideas can gain momentum quickly. When a group starts leaning in a certain direction, conversation often reinforces that path. People speak up in support of the dominant view, and those who see things differently may stay quiet. As the discussion continues, the original idea gathers weight, and the group moves further along a single track.

A recent study of board decision-making in Chinese publicly listed companies observed this pattern in action. When directors shared similar expectations—either for aggressive strategic change or caution—their group decisions became even more extreme after discussion. The study demonstrated both the risky and cautious shift, echoing early findings on how group discussions amplify initial leanings; the more aligned the board was at the outset, the more polarized the strategic choice became.⁸ Patterns like this don’t always lead to bad decisions, but they do make it harder for teams to recognize early missteps or stay open to revisiting assumptions and competing ideas. 

Politics and public life

Politics shapes how we see the world and how we decide who belongs in our communities and who shares our sense of values. In politically aligned groups, views often become more certain and less open to challenge. Over time, shared language and assumptions take hold, and small differences start to feel like deep divides. What begins as a position on an issue can harden into a line between “us” and “them.”

This pattern shows up in both research and everyday conversation. In a US-based study on public apathy and climate risk, researchers asked liberal and conservative participants to discuss the topic of climate change in small, ideologically aligned groups. Each group spent 15 minutes talking with peers who shared their general viewpoint. After the conversation, both liberals and conservatives showed more extreme versions of the original beliefs they held before the discussion began.⁹ As those polarized shifts settle in, political identity becomes a filter that shapes how individuals respond to ideas, evidence, and each other.

The idea of group polarization has held up across decades of research, but that doesn’t mean it explains everything, nor does it mean it’s always a bad thing. Scholars continue to explore where it applies (or doesn’t) and how it interacts with the messier realities of group life, such as competing social identities, structural forces, and deliberate attempts to shape public opinion.

Is group polarization sometimes strategic?

Group polarization is often treated as a byproduct of human psychology that emerges naturally from group dynamics. But some scholars argue that polarization can be encouraged deliberately, especially in political or media environments designed to amplify division. Cass Sunstein has written extensively about how group polarization can be intensified by institutional structures, such as segmented media or curated online spaces.⁶ He points to how cable news networks and algorithm-driven platforms like Facebook or YouTube often expose users to content that reinforces their existing beliefs. This design can create echo chambers where disagreement fades and extreme views gain traction simply through repetition and validation. 

Other researchers, like Yochai Benkler and colleagues, highlight the role of attention economies—platforms that reward sensational content with more visibility—and partisan messaging, where news outlets and influencers frame information to align with political agendas.¹⁰ Together, these forces reinforce echo chambers and drive individuals toward more extreme views by amplifying outrage and filtering out dissenting perspectives. These critiques shift the focus of polarization from a symptom of group effects to a strategy within the systems that shape and sustain them.

Does polarization reflect real belief change, or just social performance?

Group polarization theory suggests that individuals’ views grow more extreme through discussion with like-minded peers. But some researchers argue that what changes may not be core beliefs, but how confidently or publicly individuals express them. In a large-scale study, political scientists John Bullock and colleagues found that partisan respondents gave more exaggerated or incorrect answers on politically charged topics until they were offered a small cash reward for accuracy or guaranteed full anonymity.¹¹ Under those conditions, the partisan divide shrank sharply.

These findings suggest that polarization, in some cases, may be more about signaling group loyalty than deeply held conviction. If public pressure and reputation drive how strongly people express their views, then some polarization effects could reflect strategic performance more than persuasion—shifting how we understand and address them.

Does polarization reflect real belief change, or just social performance?

Group polarization theory assumes that individuals’ attitudes become more extreme through discussion. But some researchers argue that what actually changes might not be beliefs themselves, but how openly or strongly people express them. Political scientist Brian Schaffner found that Republicans were more likely to give hardline answers on immigration when they thought their responses would be seen by party leaders.  This observation implies that their public opinions were shaped by perceived social expectations more than internal conviction.¹¹

Similarly, political researcher Emily Ekins has shown that strong partisan views often soften when individuals are asked about issues in more neutral or private settings.¹² These findings raise questions about whether the effects we label as polarization reflect deeper belief shifts, or strategic signaling designed to show loyalty and belonging. If polarization is sometimes more about performance than persuasion, that could change how we approach solutions.

What happens when group identity gets complicated?

Most group polarization research focuses on individuals in clearly defined, like-minded groups. But real life is rarely that simple. We all carry multiple identities—cultural, political, professional, generational—and those identities don’t always point in the same direction. Some researchers have found that these cross-pressures can make extreme shifts less likely, especially when individuals regularly interact with others who challenge their views, for example, someone who holds strong political beliefs but still has close relationships with family members, coworkers, or community members who think differently.¹³ Others note that in any given setting, one identity often becomes more important than the others, and group alignment still takes hold.¹⁴ Group polarization may not disappear in complex identity settings—it often adapts to whichever group tie feels most relevant in the moment.

Online radicalization and the Christchurch attack

In March 2019, a 28-year-old man carried out a mass shooting at two mosques in Christchurch, New Zealand, killing 51 individuals and livestreaming part of the attack on social media. Investigations revealed that he’d spent years immersed in online communities centered around anti-immigrant and white nationalist ideologies, on platforms like 4chan and 8chan that reward extreme views with visibility and validation.¹⁵

Scholars analyzing the attack have pointed to group polarization in online communities as a key mechanism in the shooter’s radicalization.¹⁶ When individuals spend time in online spaces where everyone shares the same views, their beliefs can start to shift. On anonymous message boards, users often post extreme content—memes, rants, even manifestos—with no one pushing back. Over time, that kind of steady reinforcement can make those views feel more obvious, more justified, and harder to question. What starts as frustration or doubt can grow into a sense of urgent moral purpose.

In the Christchurch case, the attacker explicitly referenced these communities in a manifesto posted online before the shooting, using language, memes, and ideological cues commonly found in extremist digital spaces. He framed his violence as part of a shared ideological mission—further evidence that the radicalization wasn’t isolated, but socially intensified.

COVID-19, masks, and polarized beliefs

During the early months of the COVID-19 pandemic, public health officials urged widespread mask use to slow transmission. But what began as a safety measure quickly became a political signal. In the United States and other countries, opinions about masks split sharply along ideological lines. As lockdowns and stay-at-home orders pushed people to spend more time online, many found connection in digital communities where political and personal beliefs were shared and reinforced. For some, that meant entering spaces where skepticism about COVID-19 precautions became a dominant—and increasingly extreme—view. For others, it deepened commitment to public health measures and distrust of those who questioned them.

Researchers have documented this group polarization in both digital and in-person settings, but the effect became especially visible during the early stages of the COVID-19 pandemic. One study analyzing social media behavior found that exposure to like-minded content—whether in favor of or against mask mandates—amplified individuals’ views over time.¹⁷ These online spaces created self-reinforcing feedback loops, where supportive arguments were repeated, opposing views were dismissed or mocked, and users became increasingly certain in their positions. The result was a kind of hardening: more confidence, less openness, and shrinking room for curiosity or nuance. In the absence of mixed voices and the safety to question, even personal health decisions became litmus tests for group belonging.

Why do we support opinions as they become more popular?

Curious how group consensus starts in the first place? Read more about the bandwagon effect and why we tend to follow the crowd.

The more the merrier? The irrationality behind group chats

Want to see how group polarization sneaks into everyday life? Read this technology insight for a lighter—but no less revealing—look at how group dynamics shape what we say and believe.

1. Haidt, J. (2012). The righteous mind: Why good people are divided by politics and religion. Pantheon Books.
2. Stoner, J. A. F. (1961). A comparison of individual and group decisions involving risk (Unpublished master’s thesis). Massachusetts Institute of Technology, Cambridge, MA.
3. Wallach, M. A., Kogan, N., & Bem, D. J. (1962). Group influence on individual risk taking. Journal of Abnormal and Social Psychology, 65(2), 75–86.
4. Moscovici, S., & Zavalloni, M. (1969). The group as a polarizer of attitudes. Journal of Personality and Social Psychology, 12(2), 125–135.
5. Myers, D. G., & Lamm, H. (1976). The group polarization phenomenon. Psychological Bulletin, 83(4), 602–627.
6. Sunstein, C. R. (2009). Going to extremes: How like minds unite and divide. Oxford University Press.
7. Myers, D. G., & Bishop, G. D. (1970). Discussion effects on racial attitudes. Science, 169(3947), 778–779.
8. Chen, W., Ma, X., & Lan, H. (2023). Group polarization in board decisions about strategic change: Evidence from Chinese publicly listed companies. Management and Organization Review, 19(2), 201–232.
9. Kahan, D. M., Peters, E., Wittlin, M., Slovic, P., Ouellette, L. L., Braman, D., & Mandel, G. (2012). The polarizing impact of science literacy and numeracy on perceived climate change risks. Nature Climate Change, 2(10), 732–735.
10. Benkler, Y., Faris, R., & Roberts, H. (2018). Network propaganda: Manipulation, disinformation, and radicalization in American politics. Oxford University Press.
11. Schaffner, B. F. (2021). Priming or signaling? How politics affects immigration attitudes. Political Behavior, 43, 535–556. 
12. Ekins, D. (2020). What Americans really think about politics: The hidden tribes of opinion. Cato Institute. https://www.cato.org/blog/americans-are-not-political-extremists
13. Mutz, D. C. (2002). Cross-cutting social networks: Testing democratic theory in practice. American Political Science Review, 96(1), 111–126.
14. Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S. (1987). Rediscovering the social group: A self-categorization theory. Blackwell.
15. Institute for Strategic Dialogue. (2024, March 14). Five years on from Christchurch: Assessing the evolution of the threat landscape and policy response. https://www.isdglobal.org/digital_dispatches/five-years-on-from-christchurch-assessing-the-evolution-of-the-threat-landscape-and-policy-response/
16. Youngblood, M. (2020). Extremist ideology as a complex contagion: The spread of far-right radicalization in the United States between 2005 and 2017. Humanities and Social Sciences Communications, 7(49).
17. Jiang, J., Chen, E., & Yan, Y. (2021). Political polarization drives online conversations about COVID-19 mask-wearing. Journal of Medical Internet Research, 23(2), e24839.