Short answer: Groups can underperform individuals for a couple of distinct, well-studied reasons. Group polarization, where discussion pushes a group’s attitude toward a more extreme version of what members already leaned toward, has solid experimental support. Groupthink, the more famous idea that cohesive groups suppress dissent and rush toward flawed consensus, is the more well-known explanation, but later research has actually struggled to fully substantiate Irving Janis’s original model, suggesting the real story is a bit more complicated than the popular version.
Group Polarization: The Better-Supported Effect
Group polarization describes a well-documented pattern: when a group already leans toward a particular view and then discusses it together, the group’s collective position after discussion tends to be more extreme than the average of what individual members believed going in. If a group starts out mildly favoring a risky decision, group discussion tends to push that group toward more strongly favoring the risky choice, not toward moderation. This effect isn’t about suppressed disagreement or a domineering leader, the elements central to groupthink; it can happen even in groups with open, healthy debate, because of two overlapping processes researchers have identified: people get exposed to a wider pool of arguments supporting the majority view during discussion than they’d generate alone, and people also tend to shift their stated position to align with what they perceive as the group’s consensus, a form of social comparison.
This is a genuinely important distinction, since group polarization and groupthink get conflated constantly in casual conversation but describe different failure modes. Group polarization is fundamentally about attitudes shifting through the exchange of information and social comparison, not about people privately doubting the group’s direction and staying quiet to preserve harmony.
Groupthink: Famous, But More Contested Than Its Reputation Suggests
Groupthink is the more culturally famous concept, developed by psychologist Irving Janis after studying historical foreign policy disasters. Janis described it as a deterioration of mental efficiency, reality testing, and moral judgment that results from in-group pressures, and proposed that it’s especially likely in cohesive groups with a powerful, directive leader, under conditions of stress or crisis, where members suppress private doubts to preserve group harmony and a sense of invulnerability.
Here’s the honest complication that’s often left out of popular summaries: subsequent research attempting to test Janis’s model directly has had real trouble confirming it. One paper proposing a significant revision to the theory noted plainly that it was developed in response to a growing body of research that has failed to substantiate the groupthink model as Janis originally proposed it, and offered an alternative explanation centered on groups having inflated confidence in their own collective ability, sometimes called collective efficacy, that exceeds their actual capability. This doesn’t mean bad group decisions driven by conformity pressure and suppressed dissent don’t happen, historical examples that inspired the theory are real, well-documented cases. It means the specific causal mechanism Janis proposed, and the specific conditions he said would reliably produce it, haven’t held up as cleanly under later, more rigorous testing as the theory’s popularity might suggest.
What Seems to Actually Drive Group Decision Failures
>
Beyond these two named phenomena, researchers studying group decision-making have identified more basic, mechanical problems that plague groups even without dramatic conformity pressure or a domineering leader. One well-documented issue involves information sharing: group members often spend more discussion time on information everyone already shares in common, rather than surfacing the unique information any individual member might hold, meaning groups can fail to draw on their full collective knowledge even when no one is being actively silenced. This is a more mundane, structural explanation than either groupthink or polarization, but it’s a genuinely well-supported one, and it suggests that simply having diverse information distributed across a group’s members doesn’t guarantee that information actually gets surfaced and used.
How to Hold These Ideas Together Honestly
The fair summary keeps a few distinct, partially overlapping phenomena in view rather than reaching for a single catch-all explanation. Group polarization has solid experimental support and can push already-leaning groups toward more extreme positions purely through discussion and social comparison. Groupthink, while historically influential and a useful description of certain real, dramatic decision fiascoes, is a less well-substantiated causal model than its fame suggests, and more recent theorizing points toward inflated collective confidence rather than simple conformity pressure as a likely driver. And separately from either named phenomenon, groups frequently fail simply because they don’t effectively pool the unique information their individual members hold.
What This Means in Practice
- Watch for extreme positions emerging from group discussion, especially in already like-minded groups. Given how well-supported group polarization is, a homogenous group discussing a topic they mostly already agree on is a genuine risk factor for drifting toward an extreme, poorly examined conclusion.
- Don’t rely on “prevent groupthink” checklists as a complete solution. Since later research has struggled to confirm groupthink’s specific proposed mechanism, techniques designed narrowly around it may not address the full range of real ways groups go wrong.
- Deliberately surface information that isn’t already common knowledge within the group. Given how well-documented the shared-information bias is, actively asking each member what unique information or perspective they’re holding back can address a genuinely different, often-overlooked failure mode.
- Be more cautious of high-confidence, cohesive groups than you might instinctively expect. Since inflated collective confidence, not suppressed dissent alone, appears to be a strong predictor of decision fiascoes in updated models, a group’s strong sense of shared capability is itself worth treating as a risk signal, not just a reassurance.
Group decisions can genuinely turn out worse than an individual’s would have, but the popular explanation, groupthink, is a less settled, more contested piece of psychology than its fame suggests. Group polarization and simple failures to pool unique information appear to be the better-supported culprits behind many real group decision failures.
