More than fifty years ago, then Yale psychologist Stanley Milgram conducted the famous—or infamous—experiments on destructive obedience that have come to be known as “Milgram’s shocking experiments” (pun usually intended). Milgram began his experiments in July 1961, the same month that the trial of Adolf Eichmann—the German bureaucrat responsible for transporting Jews to the extermination camps during the Holocaust—concluded in Jerusalem. The trial was made famous by the philosopher Hannah Arendt’s reports, later published in book form as Eichmann in Jerusalem. Arendt claimed that Eichmann was a colorless bureaucrat who blandly followed orders without much thought for the consequences, and whose obedient behavior demonstrated the “banality of evil.”
Milgram himself was Jewish, and his original question was whether nations other than Germany would differ in their degrees of conformity to authority. He assumed that the citizens of America, home of rugged individualism and apple pie, would display much lower levels of conformity when commanded by authorities to engage in behavior that might harm others. His Yale experiments were designed to set a national baseline.
Half of a century after Milgram probed the nature of destructive obedience to authority, we are faced with the unsettling question: What would citizens do today?
The Original Experiment
In Milgram’s original experiment, participants took part in what they thought was a “learning task.” This task was designed to investigate how punishment—in this case in the form of electric shocks—affected learning. Volunteers thought they were participating in pairs, but their partner was in fact a confederate of the experimenter. A draw to determine who would be the “teacher” and who would be the “learner” was rigged; the true volunteer always ended up as the teacher and the confederate as learner.
The pairs were moved into separate rooms, connected by a microphone. The teacher read aloud a series of word pairs, such as “red–hammer,” which the learner was instructed to memorize. The teacher then read the target word (red), and the learner was to select the original paired word from four alternatives (ocean, fan, hammer, glue).
Did Milgram’s experiment demonstrate that humans have a universal propensity to destructive obedience or that they are merely products of their cultural moment?
If the learner erred, the teacher was instructed to deliver an electric shock as punishment, increasing the shock by 15-volt increments with each successive error. Although the teacher could not see the learner in the adjacent room, he could hear his responses to the shocks as well as to the questions. According to a prearranged script, at 75 volts, the learner started to scream; from 150 volts to 330 volts, he protested with increasing intensity, complaining that his heart was bothering him; at 330 volts, he absolutely refused to go on. After that the teacher’s questions were met by silence. Whenever the teacher hesitated, the experimenter pressed the teacher to continue, insisting that the “the experiment requires that you continue” and reminding him that “although the shocks may be painful, they cause no permanent tissue damage.”
Milgram was horrified by the results of the experiment. In the “remote condition” version of the experiment described above, 65 percent of the subjects (26 out of 40) continued to inflict shocks right up to the 450-volt level, despite the learner’s screams, protests, and, at the 330-volt level, disturbing silence. Moreover, once participants had reached 450 volts, they obeyed the experimenter’s instruction to deliver 450-volt shocks when the subject continued to fail to respond.
Putting Milgram’s Work in Context
Milgram’s experiment became the subject of a host of moral and methodological critiques in the 1960s. These became somewhat moot with the publication of the American Psychological Association’s ethical principles of research with human subjects in 1973 and the restrictions on the use of human subjects included in the National Research Act of 1974, which effectively precluded psychologists from conducting experiments that, like Milgram’s, were likely to cause serious distress to subjects. This was undoubtedly a good thing for experimental subjects, but it also blocked attempts to answer a question that many were asking: Did Milgram’s experiment demonstrate that humans have a universal propensity to destructive obedience or that they are merely products of their cultural moment?
Cross-cultural studies at the time provided a partial answer. The U.S. mean obedience rate of 60.94 percent was not significantly different from the foreign mean obedience rate of 65.94 percent, although there was wide variation in the results (rates ranged from 31 to 91 percent in the U.S. and from 28 to 87.5 percent in foreign studies) and design of the studies. However, a historical question remained: Would subjects today still display the same levels of destructive obedience as the subjects of fifty years ago?
There are grounds to believe they would not. In the 1950s, psychologists and the general public were shocked by the results of Solomon Asch’s experiments on conformity. In a series of line-judgement studies, subjects were asked to decide which of three comparison lines matched a target line. Crucially, these judgements were made in a social context, among other participants. Only one of the subjects in the experiment was a naive subject, while the other six were confederates of the experimenter instructed to give incorrect answers. When they gave answers that were incorrect, and seemed plainly incorrect to the naive subjects, about one-third of the naive subjects gave answers that conformed with the majority. In other words, people were happy to ignore the evidence before their eyes in order to conform to the group consensus.
For some observers, these results threatened America’s image as the land of individualism and autonomy. Yet in the 1980s, replications of Asch’s experiment failed to detect even minimal levels of conformity, suggesting that Asch’s results were a child of 1950s, the age of “other-directed” people made famous by David Reisman in his 1950 work The Lonely Crowd. Might the same failure to replicate be true today if people faced Milgram’s experiment anew?
Understanding Milgram’s Work Today
Although full replications of Milgram’s experiment are precluded in the United States because of ethical and legal constraints on experimenters, there have been replications attempted in other countries, and attempts by U.S. experimenters to sidestep these constraints.
A replication conducted by Dariusz Dolinski and colleagues in 2015 generated levels of obedience higher than the original Milgram experiment, although the study may be criticized because it employed lower levels of shock.
More intriguing was the 2009 replication by Jerry Burger, who found an ingenious way of navigating the ethical concerns about Milgram’s original experiment. Burger noted that in the original experiment 79 percent of subjects who continued after the 150 volts—after the learner’s first screams—continued all the way to the end of the scale, at 450 volts. Assuming that the same would be true of subjects today, Burger determined how many were willing to deliver shocks beyond the 150-volt level, at which point the experiment was discontinued.
Given social support, most subjects refused to continue to administer shocks, suggesting that social solidarity serves as a kind of a defense against destructive obedience to authority.
About 70 percent were willing to continue the experiment at this point, suggesting that subjects remain just as compliant in the 21st century. Nonetheless, Burger’s study was based upon a questionable assumption, namely that 150-volt compliance has remained a reliable predictor of 450-volt compliance. Subjects today might be willing to go a bit beyond 150 volts, but perhaps not to the far end of the scale (after learners demand that the experiment be discontinued etc.). In fact, this assumption begs the critical question at issue.
However, French television came to the rescue. One game show replicated Milgram’s experiment, with the game show host as the authority and the “questioner” as the subject. In a study reported in the 2012 European Review of Applied Psychology, J.-L. Beauvois and colleagues replicated Milgram’s voice feedback condition, with identical props, instructions, and scripted learner responses and host prods. One might question whether a game show host has as much authority as a scientific experimenter, but whatever authority they had managed to elicit levels of obedience equivalent to Milgram’s original experiment (in fact somewhat higher, 81 percent as opposed to the original 65 percent). It would seem that at least French nationals are as compliant today as Milgram’s original subjects in the 1960s.
In fact, the replication suggests a darker picture. One of the optimistic findings of the original Milgram experiment was his condition 7, in which there were three teachers, two of whom (both confederates of the experimenter) defied the experimenter. Given this social support, most subjects refused to continue to administer shocks, suggesting that social solidarity serves as a kind of a defense against destructive obedience to authority. Unfortunately, this did not occur in the French replication, in which the production assistant protested about the immorality of the procedure with virtually no effect on levels of obedience. And unfortunately, not in the Burger study either: Burger found that the intervention of an accomplice who refused to continue had no effect on the levels of obedience. So it may be that we are in fact more compliant today than Milgram’s original subjects, unmoved by social support. A dark thought for our dark times.