The Bystander Effect and Humanitarian Intervention in Mass Atrocity Crimes:
An Interdisciplinary Analysis of Factors Influencing Bystander Intervention
Milan Ciganovic
There is an implied understanding in society that individuals should come to the aid of persons requiring assistance during an emergency. A similar understanding exists in international politics- in that the international community often feels compelled to intervene in order to stop large-scale human suffering. Although there is a shared sense of obligation to intervene, this obligation does not always lead to action. Notable explanations as to why individuals do not always intervene and help others during emergencies are found within social psychology. This discipline has conducted significant research into what is known as the bystander effect: an observed social phenomenon demonstrating that when there are more witnesses to an emergency, it is less likely that individual bystanders will react and intervene to assist those in danger. Literature on the bystander effect offers a number of variables which may impact the willingness of a bystander to intervene during an emergency: the cohesiveness of the group witnessing the incident, ambiguity concerning the severity of the emergency, social processes of the diffusion of responsibility, social influence and audience inhibition, the severity of the emergency, etc..
This paper is similarly designed to explain why the international community and international organizations such as the United Nations are not always willing to intervene in cases of mass atrocities such as genocide. The inconsistent nature of the international community’s response to mass atrocities (particularly during the 1990s) through humanitarian intervention measures is the main topic of interest in this paper. This paper seeks to discover whether concepts from the study of the bystander effect and social psychology provide any explanatory power in examining the factors which influence the international community to intervene for humanitarian purposes during mass atrocities. Given the existence of both societal and international humanitarian/helping norms (shared understandings) relating to the duty to assist those in need of help, it is possible to examine the factors which both negatively and positively influence individuals and global actors to act on these norms.
This paper argues that although humanitarian norms exist both at the societal and international levels, concepts from the study of the bystander effect offer limited insights into the willingness of the international community and its members to respond to mass atrocities through military intervention. Instead, there are structural and institutional complexities present in the international context which play a larger role in influencing the international community and its members to intervene during such crises. However, there is evidence that social processes such as social influence and audience inhibition, and group cohesiveness may affect the behaviour of global actors. The ambiguity and nature of international law and the structure of the UN and its institutions are more important variables affecting the behaviour of states.
Part one of this research paper examines literature on the bystander effect and some of the key concepts from the associated research. Part two examines shared social and international norms concerning the duty to intervene, and the legal dimensions of the duty to intervene in facets of domestic and international law. Part three develops an IR framework examining the factors influencing global actors to undertake humanitarian intervention during mass atrocities.
PART 1: Social Psychology and the Bystander Effect
Latane and Darley, Kitty Genovese and the 38 Bystanders
The bystander effect is a well-established observable phenomenon in social psychology which became an important topic of research in the mid-1960s. Pioneering studies and experiments performed by psychologists Bibb Latane and John Darley led to findings suggesting that the more witnesses there were to an emergency situation, the less likely that one of those individuals was to intervene. These findings would start a wave of research into this topic which would further cement the idea that group size is negatively related to the willingness of witnesses to intervene. This willingness to intervene came to be known as ‘helping behaviour’.[1]
Based on findings from their 1968 experiment(s), Latane and Darley were the ones who coined the term “bystander effect”. The two psychologists became interested in the phenomenon following the well-publicized murder of a young New York woman named Kitty (Catherine) Genovese in 1964.[2] Such a murder in New York City would not have gathered much public attention had newspapers not disclosed that at least 38 witnesses had observed the attack, but not a single witness did anything to assist the victim. The murder of Kitty Genovese is believed to have played out for over half an hour, with not one of the 38 witnesses even calling the police to report the attack and murder.[3]
After these additional details surrounding the murder of Kitty Genovese came to light, Latane and Darley became interested in discerning why given “obvious humanitarian norms” about helping, individuals remain bystanders during emergencies.[4] In 1968, the two psychologists set out to research the willingness of witnesses to intervene in a number of ground-breaking and informative social experiments. The first of these experiments revolved around a number of volunteers isolated in booths being led to believe that they were overhearing someone suffering an epileptic seizure. The results of the experiment showed that 85 percent of volunteers who believed that they were alone and knew what was happening ended up reporting the incident. On the other hand, only a third of volunteers who overheard the seizure and were led to believe that others knew about the seizure did anything to intervene or report the incident. This experiment highlighted the phenomenon that the larger the number of witnesses to an emergency, the fewer will do anything to help.[5]
In one of their other experiments, Latane and Darley placed one, two, or three volunteers in a room under the guise that they were filling out a questionnaire. While the individuals worked on the questionnaires, the room slowly began to fill with smoke. The purpose of the experiment was to see under which conditions the volunteers would react and respond to the situation: while alone, or while in the company of others?[6] The results of the experiment were rather astounding. Within two minutes of the smoke entering the room, a volunteer left alone in the room would intervene by looking for the source of the smoke or getting help. In rooms where two people were left in a room, only one in ten of the volunteers did something and reported the smoke.[7] Similarly, in rooms where three volunteers were present as smoke filled the room, only one in 24 individuals reported the smoke within the first four minutes. The others volunteers did nothing as smoke filled the room and as they began to cough and rub their eyes.[8]
These experiments influenced many other experiments aimed at replicating the same results given different variables, including variance in the composition of the witnesses, and variance in the type of emergency (robbery, falls, screams, crashes, etc.). The results of each of these mirrored and confirmed the results of the original: the more people, the less likely someone will respond.[9] The lack of action and intervention by witnesses during these experiments does not mean that these individuals were bad or immoral people. All individuals are thought to be capable of behaving in such a matter in group situations. Even with the wide circulation of cellphones and other technology, the reporting of incidents is becoming easier, but non-intervening behaviours have remained.[10]
By the same standard, the lack of intervention in international crises by global actors does not mean that these actors are bad or immoral, but that there are circumstances at play which lead to the ultimate decision to not intervene. The reality is that for a number of different reasons, both individuals and global actors end up not helping others who are in danger even though they feel obliged to intervene. In describing the process by which individuals end up remaining bystanders, Margaret Heffernan notes that:
The nonintervening subjects of Darley and Latane’s experiments had not, they said, decided not to intervene. Rather they were frozen in a state of indecision and conflict about whether and how to intervene. Looking for a way out of that discomfort, they chose the easier path, a kind of moral shortcut.[11]
Similarly, Latane and Darley state in one of their 1968 papers that “In certain circumstances, the norms favouring intervention may be weakened, leading bystanders to resolve the conflict in the direction of non-intervention.”[12] Individuals may understand that they should intervene, but there are a number of explanations from social psychology as to why individuals do not intervene. In turn, this paper now turns to an examination of the said explanations.
Social Processes: The Diffusion of Responsibility, Social Influence, and Audience Inhibition
One of the social processes (modes of interaction in society) used to account for the occurrence of bystander effect is what is known as the diffusion of responsibility. The diffusion of responsibility refers to the hypothesis that witnesses to emergencies may indeed recognize that help is needed but ultimately fail to intervene themselves, believing that others present at the scene also have the responsibility to help.[13] Latane and Darley theorized that the diffusion of responsibility explained the findings of their 1968 experiments.[14][15] The concept is simple: if a single individual is a witness to an emergency, they alone are responsible for dealing with the emergency; if there are multiple witnesses to an emergency, individuals may feel as if their own responsibility to help is lessened, thus making them less likely to intervene.[16] A group of witnesses cannot know exactly what each person is thinking or how they will react; therefore the responsibility to help is diffused among the group. The lack of intervention similarly diffuses any potential blame for the non-intervention among individuals in the group.[17]
Unlike in situations where there is only one witness present, the presence of two or more witnesses in an emergency leads to a situation in which the pressure (be it social or legal) to intervene does not focus on any single witness. In such cases the responsibility to intervene is shared among the group of witnesses as a whole, and therefore no one intervenes and helps.[18] Individual responsibility and accountability for intervention is diminished in group settings. In the case of the experiment where smoke flooded a room with volunteers, the diffusion of responsibility explanation does not fully account for why volunteers in groups did nothing. As Latane and Darley explain, “It is hard to see how an individual’s responsibility for saving himself is diffused by the presence of other people.”[19]
As for why few individuals in groups during the smoke experiment did not react when their own life may in fact have been in danger, Latane and Darley describe how ‘social influence’ may be an explanation. During the process of deciding whether or not to intervene/act during such an emergency, it is likely that the individual in question will in some considerable way be influenced by the decisions they perceive others to be taking. If others do not perceive the situation as serious and choose to do nothing, they may influence the perceptions of another and inhibit their likelihood to intervene as well.[20] What one individual is perceived to be thinking or doing, and the inaction of one individual may influence others to do the same. Alongside the diffusion of responsibility, social influence is the second of three major social processes which are believed to inhibit helping behaviour which leads individuals to decide that inaction is the expected form of behaviour during emergencies.[21]
The third social process believed to inhibit helping behaviour and intervention during emergencies is ‘audience inhibition’. In this case, a bystander who may want to intervene may fear embarrassing themselves if in fact they wrongly interpret an event as an emergency. The presence of other bystanders can inhibit another person’s desire to intervene and help by making said person fearful that their behaviour (the intervention) can be seen and judged negatively by others.[22] The three social processes of the diffusion of responsibility, social influence, and audience inhibition describe how the presence of multiple people can influence an individual’s behaviour in different, yet related ways. Beyond just the presence of a group of bystanders, the actual interpersonal relationships of the group can also influence a group member’s willingness to intervene and help in an emergency.
Group Cohesiveness
The general make-up of a group of bystanders can have an impact on whether individuals in that group are willing to intervene. Latane and Rodin were among the first to test whether the relationship between bystanders can influence helping/intervening behaviour. Latane and Rodin found that although pairs of friends playing the roles of bystanders showed signs of social inhibition when overhearing someone crying in pain, these bystanders intervened much faster than pairs of strangers.[23] The researchers found that friends were more likely to discuss the incident and decide on a mutual plan of action. This is because friends are less likely than strangers to misinterpret each other’s lack of action, and because friends are also less likely to feel embarrassed about acting/intervening in front of each other.[24] Other explanations as to why friends are more likely to intervene than strangers includes the suggestion that friends are concerned about maintaining high opinions of each other, and that friends do not diffuse responsibility between themselves as much as strangers. The latter explanation proposes that there is a mutual sense of obligation of responsibility between friends so that friends may in fact believe that they share the responsibility to assist as opposed to strangers between whom the responsibility is diffused.[25]
Beyond simply looking at the bystander friendship as a factor influencing individual intervention, overall group cohesiveness is considered to be an important variable for understanding the bystander effect. The meaning that the group has to an individual within said group is argued to affect an individual’s willingness to intervene.[26] Cohesiveness can be defined as “the degree of attraction members of a group have for one another”.[27] Cohesiveness can range from a total lack of cohesiveness among complete strangers, to lower levels of cohesiveness among acquaintances, and ever greater levels of cohesiveness.[28] Cohesiveness is argued to increase individual responsiveness to, and acceptance of, social norms such as the social-responsibility norm- the common belief that people should help those who need assistance and who are dependent on them for assistance.[29] Norms, which will be discussed later on in the paper, can be understood as shared understandings of what are considered acceptable behaviours.
This means that if cohesiveness among a group of strangers is considered to be low, the individuals in that group have a decreased responsiveness to helping norms. The inverse of this is also true: the greater the group cohesiveness among bystanders, the more significant and acceptable the social-responsibility norm. Therefore, more cohesive groups are argued to be more likely to help in an emergency.[30] As a result of the greater acceptance of social norms such as the social-responsibility norm, group cohesiveness is argued to play a central role in facilitating helping/altruistic behaviour.[31]
The Victim
As much as the dynamics within a group of bystanders can affect their willingness to intervene in an emergency situation, so does the identity of the person in danger. Research has shown that manipulating victim characteristics during experiments can impact how bystanders intervene. Experiments testing whether racial prejudice plays a role in a bystander’s willingness to intervene show that both the diffusion of responsibility and the social influence to not intervene may play more of an inhibitive influence in scenarios when victims are black and the bystanders are white. When bystanders are alone and hence not influenced by the presence of others, black and white victims have been found to be helped equally quickly by bystanders.[32]
These findings were found to be true in both scenarios involving low- and high-prejudiced bystanders. However, white bystanders who were rated as being more prejudiced, and who were in the presence of other passive bystanders were found to respond to black victims more slowly than to white victims.[33] Granted, these tests were performed over 30 years ago when racial relations may have been different than as in present day, these findings are still relevant because the experiments were performed using individuals with different levels of racial prejudice. These findings also demonstrate that perceptions and prejudices regarding victims can influence a bystander’s willingness to intervene.