Policy Memos

Anger and Prejudice after the Beslan School Hostage Taking

Policy Memo:


Publication Date:




On September 1, 2004, Russia experienced its most appalling act of terrorism in recent history: the taking of approximately 1,200 hostages at School No. 1 in Beslan, North Ossetia, where they remained for 53 hours, thirsty, hungry, frightened, and in sweltering heat before meeting a gruesome ending that included explosions, a burning and collapsed gym roof, gunfire, and widespread chaos and trauma. After these horrific events, many observers expected retaliatory violence. The conventional wisdom was that the 330-plus deaths and countless injuries sustained after the unprecedented hostage taking of schoolchildren, parents, and teachers would now revive longstanding bloody rivalries that had existed between Ossetians and neighboring Ingush and Chechens. The assumption was that the horror of the hostage taking would trigger anger, and anger would lead to retaliatory violence.

As it turns out, however, anger was actually a productive force in fueling peaceful political participation in response to the hostage taking and interestingly had little to do with support for retaliatory violence. The single most important factor explaining support for retaliatory violence was not anger but ethnic prejudice, regardless of whether or not that prejudice induced an emotional response. These findings are based on interviews held in 2007 with 1,098 victims of the hostage taking, or 82 percent of victims identified, along with focus groups of politically active and inactive victims (See Appendix). The findings of this research have implications for anticipating the aftermath of violence in other interethnic contexts.

Anger after Violence

Anger is a negative and often disparaged emotion, one that we are taught from childhood to try to minimize or at least control. It is a response to a perceived misdeed, an unpleasant emotional state resulting from the perception of unwanted, unfair, or undeserved consequences, especially when the inflicted harm is perceived as intentional.

By this definition, anger was a justifiable reaction to the Beslan hostage-taking. Children were the primary targets and victims, their victimization undeserving and therefore maddening. Moreover, the intentionality of the perpetrators to inflict harm was clear. The deaths, physical injuries, and emotional devastation of the incident were all the result of deliberate acts of terrorism, and thus justified moral indignation. The intentions of government actors before, during, and after the hostage taking were arguably less clear, but the still unresolved discrepancies in information and the perceived callous indifference of the authorities to the victims mean that even government-inflicted harm was often perceived as intentional and warranting anger. As has been established in places like South Africa, truth-telling can “quench anger,” but the victims of Beslan and other North Ossetians to this day believe they do not know the truth about most aspects of the hostage taking and its gruesome ending. Without a true accounting, victims may ruminate and excessively attribute intentionality and feel anger.

However, not everyone experiences anger equally. People vary in their readiness for or susceptibility to anger. When asked how many days in the past week they have felt anger, victims reported modal responses of zero and two days, but responses varied across the entire range from zero to seven days in the past week.

Prejudice after Violence

Ethnic prejudice is defined as negative preconceived opinions of others based on their ethnicity. Prejudice between Ossetians and Chechens and Ingush existed well before the events of 2004. Ossetians and Ingush have had a decades-old conflict over ownership of the Prigorodny district of North Ossetia that was formerly part of Ingushetia. The conflict became violent in 1992, leading to hundreds of deaths, hostage takings, missing persons, property destruction, and the forcible displacement of Ingush to refugee camps in Ingushetia and Chechnya. Atrocities were committed by both sides, and the conflict was never quite resolved.

Historical animosity also exists between Ossetians and Chechens. Federal air raids were launched at Chechnya from Beslan during the first Chechen war, and while Ossetians are not ethnic Russians, their perceived participation in the war casts them in the minds of some as proxies for the Russian government.

The perpetrators of the violence in Beslan in 2004 were mainly of Chechen and Ingush descent. Victims and other North Ossetians often think of the perpetrators not simply as individual terrorists but as representatives of their presumably violent and barbaric ethnic groups. Survey data confirm that most Beslan victims felt negatively about dining with an Ingush, having a relative marry an Ingush, or sending their children to school with Ingush children. Over three-quarters of the victims blamed the Ingush and Chechen people for causing the tragedy in the school.

Importantly, however, not all victims held these sentiments, and among those who were prejudiced, sentiments were expressed with varying levels of intensity.

Anger Fuels Participation

According to established theories of emotion, anger leads to aggression. The urgent need of angry people to right a perceived wrong supposedly encourages confrontation and violence. The tendency of angry people to overestimate their chances for successful retaliation and to use selective memory, which heightens their prejudice and stereotyping, can also encourage further confrontation and violence. Thus, the so-called “action tendency” of anger is thought to be retaliatory harm.

Conversely, other evidence shows that anger is not always channeled aggressively. Action tendencies are socially regulated, meaning they are inhibited or promoted by social norms. Victims of violence who are angry may refrain from retaliatory violence due to the social unacceptability of violence in many, if not most, circumstances. Cases of retaliatory violence have occurred mainly in contexts where social norms promote violence or where political entrepreneurs exploit anger and increase the acceptability and necessity of a violent response. At the same time, social norms often do not promote violence, and elites often do not provoke it.

If anger does not provoke victims to engage in retaliatory violence, what does it do? Anger is an empowering emotion in the sense that anger demands expression and action. An angry victim often has stronger motivation than a less angry victim, more optimistically appraises risk, and thus may be more likely to do something in response to violence. What exactly such an angry victim may do is less clear. How will that anger be expressed or acted upon?

Constructive action such as political mobilization is a likely contender. This is precisely what we have seen in Beslan. The angrier the Beslan victim, the more likely he or she was to participate in politics. Victims were asked about their participation in thirty-one different and often very prominent participatory acts, including signing petitions, writing to newspapers, attending rallies, meeting with political officials, blockading a highway, and staging a courtroom sit-in. The angriest victims were roughly six to nine times more active than the less-angry victims. Only about a quarter participated in none of these activities, whereas non-participation was the norm for the majority of those who were never angry or angry only one day a week.

However, the angriest victims and least angry victims showed no difference in their support for retaliatory violence. Victims were asked an extremely blunt question about whether they approved of killing Chechens as retaliation for the hostage taking. Only a minority expressed such support, which is consistent with the fortunately minimal occurrence of actual retaliatory violence after the hostage taking. Well over three-quarters of victims somewhat or completely disapproved of violent retaliation, regardless of their level of anger.

Prejudice Fuels Violence

The literature on violence often discounts the importance of prejudice when emotions are not aroused. Feeling averse to people because of their ethnicity is supposedly insufficient to generate attack, but when an emotion such as hatred is involved, support for violence is then encouraged. However, there is a more likely, direct, and non-emotional causal mechanism linking ethnic prejudice to ethnic violence, which is the grievance itself, when that grievance is prior violence and individuals are contemplating retaliation. Individual support for ethnic violence might be ignited not by emotion but by actual grievances experienced by already prejudiced individuals. It is not that ethnic identities alone lead to violence. Such an argument ignores or even trivializes the grievance or dispute between two rival ethnic groups. In the North Caucasus, for example, had interethnic killing not occurred, especially on the part of the Russian government, it is unlikely that the negative opinions of Ingush and Chechens toward Ossetians and vice-versa would in themselves have led to violence. Rather, the argument is that, in the context of a grievance such as victimization by violence, retaliatory violence may be more likely among those who have negative feelings toward the ethnic group of the perpetrators, regardless of their individual level of emotional arousal.

In the North Caucasus, retaliatory violence for the Beslan school hostage taking was a much rarer occurrence than journalists and other students of the region had anticipated. Among victims, attitudes matched this aggregate behavioral outcome; support for retaliatory violence was also low. However, to the extent that support for violence existed, prejudice was a very strong correlate. The more prejudiced the victim, the more likely he or she was to approve of killing Chechens. This does not mean that prejudice predicts support for violence; prejudice toward Ingush and Chechens is widespread among victims, but approval of retaliatory killing was not widespread. Prejudice simply makes support for violence much more likely, and lack of prejudice is very strongly correlated with disapproval of violence.

Ramifications for Other Interethnic Contexts

Violent individuals are sometimes excused for their violent acts with the rationale that some legitimate circumstance angered or frightened them. In the context of retaliatory ethnic violence, that excuse may be weak. The evidence presented here does not suggest that every individual brought to some high emotional threshold would support retaliatory violence against coethnics of the perpetrators. Instead, it suggests that aversion to other individuals based on their ethnicity is dangerous in itself. The legitimate circumstance, while very real and likely anger-provoking, is nonetheless a pretext for violence based on preexisting prejudice.

When individuals support retaliatory violence in a context of interethnic conflict, they are not simply taking revenge on the initial perpetrator of violence. Instead, they are generalizing the revenge and endorsing harm brought to people of the perpetrator’s ethnicity. It is not clear that anger or other emotions can account for this leap from logical revenge (kill the killer) to generalized, prejudiced revenge (kill the killer’s coethnics) without numerous other intervening steps. Prejudice, however, can indeed account for the leap, since the prejudiced individual thinks in ethnic terms and is likely to see a killer as a representative of his or her ethnic group.



About the Study

We defined a victim as a surviving adult (18+) hostage, parent, or guardian of an underage hostage, or next of kin of a deceased hostage. Using lists of hostages from the Procuracy, the City Social Provision division that administers social aid to victims, Mothers of Beslan, and various journals, we compiled a rough inflated list of 1,479 hostages’ names. After correcting for misspellings, duplicates, non-hostages, and non-existent individuals, we whittled the list to 1,226, which is close to the average number of published estimates. We then targeted this entire population, trying to get one victim respondent for each former hostage or two in the case of parents of underage hostages. According to these rules (one targeted respondent per former hostage or two in the case of parents), we identified and contacted 1,340 victims. Of these, 38 were out of town or otherwise unavailable for the entire duration of the survey, 7 moved and had no forwarding address, and five did not participate for other reasons. Only 192 victims refused to participate (44 former hostages and 148 parents or other relatives of former hostages), a remarkably low number given the sensitivity of the topic. Reasons for refusal for the most part involved victims not wanting to recall a painful event and instead wanting to forget and move on rather than live in the past. A smaller number of victims thought that surveys were useless and would not help them personally. The resulting sample of 1,098 victims of the 1,340 initially contacted represents an 82 percent response rate. Face-to-face interviews were conducted in respondent homes in May, June, July, and August of 2007 (213, 649, 163, and 73 respondents, respectively). Six focus groups of 6 to 9 participants each were conducted in Beslan in December 2008. Focus group participants were selected at random from the database of respondents to the victim survey, using sex, age, and level of activism as selection criteria.

About the author

Associate Professor of Political Science
University of Notre Dame