How easily we split into hostile "us versus them" groups
It takes shockingly little to trigger tribalism. In the 1970s, psychologist Henri Tajfel showed that when participants were divided into arbitrary groups based on trivial criteria—like whether they preferred a painting by Klee or Kandinsky—they immediately began favoring their "in-group." Even when they had never met, they consistently awarded more points to members of their own group and penalized the "out-group."
The Search for a Neutral Baseline
Before the 1970s, many social psychologists assumed that hostility between groups required tangible conflicts of interest. The dominant view held that prejudice, discrimination, and intergroup conflict arose primarily when groups competed for scarce resources, such as territory, jobs, or social status. To better understand how deep-seated prejudices developed, the Polish-born social psychologist Henri Tajfel set out to design an experimental baseline: a completely neutral setting stripped of history, competition, and personal animosity, upon which researchers could gradually introduce conflict variables to see when hostility first emerged.
To create this baseline, Tajfel and his colleagues devised what became known as the minimal group paradigm. The methodology aimed to isolate the bare minimum conditions necessary for an individual to favor their own group over another. Participants were categorized into groups based on utterly meaningless or arbitrary distinctions, with no prior history between the groups, no face-to-face interaction, and no direct material gain at stake for the participants themselves. To the researchers' surprise, this neutral baseline never materialized, because the mere act of categorization proved sufficient to spark immediate group favoritism.
Designing the Minimal Conditions
To ensure that the groups were truly minimal, Tajfel established strict criteria for the experimental design. First, there could be no face-to-face contact among participants during or before the experiment, preventing the formation of interpersonal bonds or social hierarchies. Second, the criteria used to divide participants had to be completely trivial. In early experiments, participants estimated the number of dots flashed on a screen and were randomly labeled as either overestimators or underestimators. In other well-known trials, participants expressed preferences for abstract paintings by either Paul Klee or Wassily Kandinsky and were categorized accordingly.
Crucially, the experimental procedure ensured complete anonymity. Participants only knew that they belonged to one group and that other individuals belonged to either the same group or the other group, identified only by code numbers. Furthermore, there was no overlap between personal self-interest and group allocation. When participants were asked to distribute monetary rewards or points using specially designed matrices, they could only allocate resources to other anonymous participants, never to themselves. Despite having nothing personal to gain and knowing nothing about the other individuals, participants consistently awarded more resources to members of their assigned in-group.
The Choice to Maximize the Difference
The distribution matrices used in these experiments revealed that participants were not merely showing passive generosity toward their own group; they were actively trying to outperform the other group. The matrices presented participants with trade-offs between several distinct strategies: Maximum Joint Profit, which maximized the total amount of points awarded to both individuals regardless of group; Maximum In-group Profit, which gave the highest possible absolute score to the in-group member; and Maximum Difference, which maximized the numerical gap between the in-group and the out-group in favor of the in-group.
The findings demonstrated a consistent preference for Maximum Difference over Maximum In-group Profit. When forced to choose, participants routinely opted to give fewer absolute points to their own group member if doing so ensured that the out-group member received even less. For example, rather than choosing an option that awarded 19 points to an in-group member and 17 to an out-group member, a participant would frequently choose an option that awarded 13 points to the in-group member and only 7 to the out-group member. Establishing a relative advantage over the out-group was prioritized over maximizing the absolute welfare of the in-group.
Social Identity and Positive Distinctiveness
To explain why arbitrary categories produced such pronounced bias, Tajfel and his collaborator John Turner developed Social Identity Theory. The theory proposes that an individual's self-concept consists of two components: personal identity, derived from personal traits and interpersonal relationships, and social identity, derived from the social categories to which the individual belongs. Because people possess a fundamental motivation to maintain a positive self-image, they strive to perceive their social identities in a positive light.
Achieving a positive social identity requires social comparison. Since the value of a group label is relative, individuals evaluate their in-group by comparing it with relevant out-groups. In the minimal group paradigm, where no objective information about the groups existed, the only way participants could create positive distinctiveness for their group was by allocating more resources to it. By establishing that their group was superior in the point-distribution task, participants conferred status onto their group and, by extension, bolstered their own social identity.
Alternative Explanations and Reciprocity
Although Social Identity Theory became the dominant explanation for the minimal group effect, other researchers proposed alternative mechanisms. One prominent critique suggested that the findings might be driven by experimental demand characteristics, meaning that participants simply deduced that the experimenters expected competitive behavior and complied with subtle cues in the experimental setup. However, subsequent variations controlling for demand characteristics continued to observe in-group favoritism.
Another influential perspective, advanced by Toshio Yamagishi and other researchers, focused on bounded generalized reciprocity. This evolutionary and rational-choice framework argues that group favoritism is not necessarily driven by self-esteem or psychological distinctiveness, but rather by an implicit expectation of mutual cooperation. Under this view, humans have evolved heuristic decision-making rules that treat in-group members as potential exchange partners. Even in anonymous, one-shot laboratory settings, participants may allocate more resources to in-group members under the intuitive assumption that in-group members are more likely to return the favor in the future.
Asymmetries and Experimental Limits
Later research revealed important boundaries to the minimal group phenomenon. One major qualification is known as the positive-negative asymmetry effect, identified by Amélie Mummendey and colleagues. While participants readily discriminate in favor of their in-group when allocating positive outcomes like points, money, or pleasant stimuli, this bias significantly weakens or disappears when they are tasked with distributing negative outcomes, such as unpleasant noise or monetary penalties. In negative contexts, participants tend to follow a strict fairness norm rather than penalizing the out-group.
Additionally, the minimal group paradigm demonstrates in-group favoritism rather than active out-group hostility. The baseline behavior observed in these experiments reflects a desire to lift the in-group above the out-group, not necessarily an active desire to harm or destroy the out-group. While minimal categorization creates the foundational bias that allows hostility to develop, the emergence of severe real-world conflict typically requires additional factors, such as ideological narratives, perceived threats, historical grievances, or competition over genuine resources.
Key takeaways
•The minimal group paradigm demonstrated that simply assigning people to arbitrary, meaningless categories is sufficient to generate immediate in-group favoritism.
•When distributing resources, participants frequently prioritize maximizing the gap between groups over maximizing the absolute gain of their own group.
•Social Identity Theory explains this behavior as a drive for positive distinctiveness, where individuals seek a positive self-concept through favorable social comparisons.
•The phenomenon is primarily characterized by in-group favoritism rather than direct out-group hostility, and it diminishes significantly when distributing burdens rather than rewards.