Obedience and Power

One of the fundamental aspects of social interaction is that some individuals have more influence than others. Social power can be defined as the ability of a person to create conformity even when the people being influenced may attempt to resist those changes (Fiske, 1993; Keltner, Gruenfeld, & Anderson, 2003). Bosses have power over their workers, parents have power over their children, and, more generally, we can say that those in authority have power over their subordinates. In short, power refers to the process of social influence itself—those who have power are those who are most able to influence others.

Milgram’s Studies on Obedience to Authority

The powerful ability of those in authority to control others was demonstrated in a remarkable set of studies performed by Stanley Milgram (1963). Milgram was interested in understanding the factors that lead people to obey the orders given by people in authority. He designed a study in which he could observe the extent to which a person who presented himself as an authority would be able to produce obedience, even to the extent of leading people to cause harm to others.

Like his professor Solomon Asch, Milgram’s interest in social influence stemmed in part from his desire to understand how the presence of a powerful person—particularly the German dictator Adolf Hitler who ordered the killing of millions of people during World War II—could produce obedience. Under Hitler’s direction, the German SS troops oversaw the execution of 6 million Jews as well as other “undesirables,” including political and religious dissidents, homosexuals, mentally and physically disabled people, and prisoners of war. Milgram used newspaper ads to recruit men (and in one study, women) from a wide variety of backgrounds to participate in his research. When the research participant arrived at the lab, they were introduced to a man who the participant believed was another research participant but who was actually an experimental confederate. The experimenter explained that the goal of the research was to study the effects of punishment on learning. After the participant and the confederate both consented to participate in the study, the researcher explained that one of them would be randomly assigned to be the teacher and the other the learner. They were each given a slip of paper and asked to open it and to indicate what it said. In fact both papers read teacher, which allowed the confederate to pretend that he had been assigned to be the learner and thus to assure that the actual participant was always the teacher. While the research participant (now the teacher) looked on, the learner was taken into the adjoining shock room and strapped to an electrode that was to deliver the punishment. The experimenter explained that the teacher’s job would be to sit in the control room and to read a list of word pairs to the learner. After the teacher read the list once, it would be the learner’s job to remember which words went together. For instance, if the word pair was blue-sofa, the teacher would say the word blue on the testing trials and the learner would have to indicate which of four possible words (house, sofa, cat, or carpet) was the correct answer by pressing one of four buttons in front of him. After the experimenter gave the “teacher” a sample shock (which was said to be at 45 volts) to demonstrate that the shocks really were painful, the experiment began. The research participant first read the list of words to the learner and then began testing him on his learning.

The shock panel, as shown in Figure 6.9, “The Shock Apparatus Used in Milgram’s Obedience Study,” was presented in front of the teacher, and the learner was not visible in the shock room. The experimenter sat behind the teacher and explained to him that each time the learner made a mistake the teacher was to press one of the shock switches to administer the shock. They were to begin with the smallest possible shock (15 volts) but with each mistake the shock was to increased by one level (an additional 15 volts).

Once the learner (who was, of course, actually an experimental confederate) was alone in the shock room, he unstrapped himself from the shock machine and brought out a tape recorder that he used to play a prerecorded series of responses that the teacher could hear through the wall of the room. As you can see in Table 6.2,”The Confederate’s Schedule of Protest in the Milgram Experiments,” the teacher heard the learner say “ugh!” after the first few shocks. After the next few mistakes, when the shock level reached 150 volts, the learner was heard to exclaim “Get me out of here, please. My heart’s starting to bother me. I refuse to go on. Let me out!” As the shock reached about 270 volts, the learner’s protests became more vehement, and after 300 volts the learner proclaimed that he was not going to answer any more questions. From 330 volts and up the learner was silent. The experimenter responded to participants’ questions at this point, if they asked any, with a scripted response indicating that they should continue reading the questions and applying increasing shock when the learner did not respond.

Table 6.2 The Confederate’s Schedule of Protest in the Milgram Experiments

Before Milgram conducted his study, he described the procedure to three groups—college students, middle-class adults, and psychiatrists—asking each of them if they thought they would shock a participant who made sufficient errors at the highest end of the scale (450 volts). One hundred percent of all three groups thought they would not do so. He then asked them what percentage of “other people” would be likely to use the highest end of the shock scale, at which point the three groups demonstrated remarkable consistency by all producing (rather optimistic) estimates of around 1% to 2%.

The results of the actual experiments were themselves quite shocking. Although all of the participants gave the initial mild levels of shock, responses varied after that. Some refused to continue after about 150 volts, despite the insistence of the experimenter to continue to increase the shock level. Still others, however, continued to present the questions, and to administer the shocks, under the pressure of the experimenter, who demanded that they continue. In the end, 65% of the participants continued giving the shock to the learner all the way up to the 450 volts maximum, even though that shock was marked as “danger: severe shock,” and there had been no response heard from the participant for several trials. In sum, almost two-thirds of the men who participated had, as far as they knew, shocked another person to death, all as part of a supposed experiment on learning.

Studies similar to Milgram’s findings have since been conducted all over the world (Blass, 1999), with obedience rates ranging from a high of 90% in Spain and the Netherlands (Meeus & Raaijmakers, 1986) to a low of 16% among Australian women (Kilham & Mann, 1974). In case you are thinking that such high levels of obedience would not be observed in today’s modern culture, there is evidence that they would be. Recently, Milgram’s results were almost exactly replicated, using men and women from a wide variety of ethnic groups, in a study conducted by Jerry Burger at Santa Clara University. In this replication of the Milgram experiment, 65% of the men and 73% of the women agreed to administer increasingly painful electric shocks when they were ordered to by an authority figure (Burger, 2009). In the replication, however, the participants were not allowed to go beyond the 150 volt shock switch.

Although it might be tempting to conclude that Milgram’s experiments demonstrate that people are innately evil creatures who are ready to shock others to death, Milgram did not believe that this was the case. Rather, he felt that it was the social situation, and not the people themselves, that was responsible for the behavior. To demonstrate this, Milgram conducted research that explored a number of variations on his original procedure, each of which demonstrated that changes in the situation could dramatically influence the amount of obedience. These variations are summarized in Figure 6.10.

This figure presents the percentage of participants in Stanley Milgram’s (1974) studies on obedience who were maximally obedient (that is, who gave all 450 volts of shock) in some of the variations that he conducted. In the initial study, the authority’s status and power was maximized—the experimenter had been introduced as a respected scientist at a respected university. However, in replications of the study in which the experimenter’s authority was decreased, obedience also declined. In one replication the status of the experimenter was reduced by having the experiment take place in a building located in Bridgeport, Connecticut, rather than at the labs on the Yale University campus, and the research was ostensibly sponsored by a private commercial research firm instead of by the university. In this study, less obedience was observed (only 48% of the participants delivered the maximum shock). Full obedience was also reduced (to 20%) when the experimenter’s ability to express his authority was limited by having him sit in an adjoining room and communicate to the teacher by telephone. And when the experimenter left the room and had another student (actually a confederate) give the instructions for him, obedience was also reduced to 20%.

In addition to the role of authority, Milgram’s studies also confirmed the role of unanimity in producing obedience. When another research participant (again an experimental confederate) began by giving the shocks but then later refused to continue and the participant was asked to take over, only 10% were obedient. And if two experimenters were present but only one proposed shocking while the other argued for stopping the shocks, all the research participants took the more benevolent advice and did not shock. But perhaps most telling were the studies in which Milgram allowed the participants to choose their own shock levels or in which one of the experimenters suggested that they should not actually use the shock machine. In these situations, there was virtually no shocking. These conditions show that people do not like to harm others, and when given a choice they will not. On the other hand, the social situation can create powerful, and potentially deadly, social influence.

Before moving on to the next section, it is worth noting that although we have discussed both conformity and obedience in this chapter, they are not the same thing. While both are forms of social influence, we most often tend to conform to our peers, whereas we obey those in positions of authority. Furthermore, the pressure to conform tends to be implicit, whereas the order to obey is typically rather explicit. And finally, whereas people don’t like admitting to having conformed (especially via normative social influence), they will more readily point to the authority figure as the source of their actions (especially when they have done something they are embarrassed or ashamed of).

Social Psychology in the Public Interest

The Stanford Prison Study and Abu Ghraib

In Milgram’s research we can see a provocative demonstration of how people who have power can control the behavior of others. Can our understanding of the social psychological factors that produce obedience help us explain the events that occurred in 2004 at Abu Ghraib, the Iraqi prison in which U.S. soldiers physically and psychologically tortured their Iraqi prisoners? The social psychologist Philip Zimbardo thinks so. He notes the parallels between the events that occurred at Abu Ghraib and the events that occurred in the “prison study” that he conducted in 1971 (Stanford Prison Study. Retrieved from http://www.prisonexp.org/links.htm).

In that study, Zimbardo and his colleagues set up a mock prison. They selected 23 student volunteers and divided them into two groups. One group was chosen to be the “prisoners.” They were picked up at their homes by actual police officers, “arrested,” and brought to the prison to be guarded by the other group of students—the “guards.” The two groups were placed in a setting that was designed to look like a real prison, and the role-play began.

The study was expected to run for two weeks. However, on the second day, the prisoners tried to rebel against the guards. The guards quickly moved to stop the rebellion by using both psychological punishment and physical abuse. In the ensuing days, the guards denied the prisoners food, water, and sleep; shot them with fire-extinguisher spray; threw their blankets into the dirt; forced them to clean toilet bowls with their bare hands; and stripped them naked. On the fifth night the experimenters witnessed the guards putting bags over the prisoners’ heads, chaining their legs, and marching them around. At this point, a former student who was not involved with the study spoke up, declaring the treatment of the prisoners to be immoral. As a result, the researchers stopped the experiment early.

The conclusions of Zimbardo’s research were seemingly clear: people may be so profoundly influenced by their social situation that they become coldhearted jail masters who torture their victims. Arguably, this conclusion may be applied to the research team itself, which seemingly neglected ethical principles in the pursuit of their research goals. Zimbardo’s research may help us understand the events that occurred at Abu Ghraib, a military prison used by the U.S. military following the successful toppling of the dictator Saddam Hussein. Zimbardo acted as an expert witness in the trial of Sergeant Chip Frederick, who was sentenced to eight years in prison for his role in the abuse at Abu Ghraib. Frederick was the army reservist who was put in charge of the night shift at Tier 1A, where the detainees were abused. During this trial, Frederick said, “What I did was wrong, and I don’t understand why I did it.” Zimbardo believes that Frederick acted exactly like the students in the prison study. He worked in a prison that was overcrowded, filthy, and dangerous, and where he was expected to maintain control over the Iraqi prisoners—in short, the situation he found himself in was very similar to that of Zimbardo’s prison study.

In a recent interview, Zimbardo argued (you can tell that he is a social psychologist) that “human behavior is more influenced by things outside of us than inside.” He believes that, despite our moral and religious beliefs and despite the inherent goodness of people, there are times when external circumstances can overwhelm us and we will do things we never thought we were capable of doing. He argued that “if you’re not aware that this can happen, you can be seduced by evil. We need inoculations against our own potential for evil. We have to acknowledge it. Then we can change it” (Driefus, 2007).

You may wonder whether the extreme behavior of the guards and prisoners in Zimbardo’s prison study was unique to the particular social context that he created. Recent research by Stephen Reicher and Alex Haslam (2006) suggests that this is indeed the case. In their research, they recreated Zimbardo’s prison study while making some small, but important, changes. For one, the prisoners were not “arrested” before the study began, and the setup of the jail was less realistic. Furthermore, the researchers in this experiment told the “guards” and the “prisoners” that the groups were arbitrary and could change over time (that is, that some prisoners might be able to be promoted to guards). The results of this study were entirely different than those found by Zimbardo. This study was also stopped early, but more because the guards felt uncomfortable in their superior position than because the prisoners were being abused. This “prison” simply did not feel like a real prison to the participants, and, as a result, they did not take on the roles they were assigned. Again, the conclusions are clear—the specifics of the social situation, more than the people themselves, are often the most important determinants of behavior.