The Man in the Gray Coat

In a plain room at Yale University in the summer of 1961, a man sat before a long gray box studded with thirty switches. Each switch was labeled with a voltage, rising in careful increments from fifteen at the far left to four hundred and fifty at the far right. The final switches carried an additional marking, blunt and unambiguous: a row of red Xs and the single word DANGER. In the next room, out of sight but not out of earshot, another man was strapped to a chair with an electrode taped to his arm. The man at the switches had been told his job was simple. Every time the learner answered a memory question incorrectly, he was to administer a shock, and with each mistake he was to climb one switch higher.

The learner made mistakes. The voltage climbed. And somewhere past the middle of that gray box, the man in the chair began to scream.

Most people, asked in advance what they would do, are certain of their answer. They would refuse. They would stand up, walk out, refuse to touch another switch the moment a stranger cried out in pain. This certainty is nearly universal, and it is one of the most reliable predictions a person can make about their own character. It is also, according to the psychologist who built that gray box, almost always wrong.

A question born from Jerusalem

The experiment did not begin as a study of memory, despite what its subjects were told. It began as an attempt to understand a war. Stanley Milgram was the son of Jewish immigrants, and he had come of age in the shadow of the Holocaust: an atrocity of such scale that it seemed to demand explanation not only in political terms but in psychological ones. How does a society produce enough willing hands to murder millions? The comfortable answer, popular in the years after 1945, was that the perpetrators were monsters, or that Germans in particular carried some cultural defect that made them uniquely susceptible to tyranny.

In 1961, that comfortable answer came under public pressure. In Jerusalem, Adolf Eichmann, one of the chief logistical architects of the deportations to the death camps, stood trial in a glass booth. His defense, repeated in various forms throughout the proceedings, was that he had merely followed orders. He was a functionary, a man who processed timetables and quotas, not a fanatic. The philosopher Hannah Arendt, covering the trial for The New Yorker, would later distill this into a phrase that has never stopped provoking argument: the banality of evil ¹. Eichmann, in her reading, was not a snarling ideologue but something more unsettling, a thoughtless bureaucrat who did his duty and never quite grasped what his duty meant.

Milgram, then a young assistant professor at Yale, watched this unfold and asked a question that turned the accusation inward. If obedience to authority could produce mass murder, was that obedience a peculiarly German trait, or was it a feature of human beings as such? He suspected the latter, and he designed an experiment to test it. His original plan even included a comparative arm: run the study in New Haven, then run it again in Germany, and measure the difference. He never needed to book the flight. What he found in Connecticut was disturbing enough to make the comparison beside the point ².

The four-dollar advertisement

The recruitment could not have been more mundane. Milgram placed an advertisement in the local New Haven paper offering four dollars, plus fifty cents carfare, to men willing to spend an hour helping with a study of memory and learning. The pay was modest even by 1961 standards. The volunteers who answered were a deliberate cross-section of ordinary working life: postal clerks, high school teachers, salesmen, engineers, laborers. They were not students, not psychology majors primed to guess the game. They were the general public, arriving one at a time at an elegant laboratory bearing the name of one of the most respected universities in the country.

Each arrival was met by two men. One wore a gray technician’s coat and carried himself with the flat, professional calm of a person doing routine work. This was the experimenter, an authority figure assembled from small signals: the coat, the clipboard, the institutional setting, the unhurried voice. The other man was warm and a little soft, middle-aged, introduced as a fellow volunteer named Mr. Wallace. He mentioned, in passing, that he had a mild heart condition. He was, in reality, a mild-mannered accountant hired to act, and he would never receive a single shock. But the true subject of the experiment, the man who had answered the advertisement, did not know any of this.

The two volunteers drew slips of paper to decide who would play teacher and who would play learner. The draw was rigged. Both slips read teacher, and the actor always claimed the learner’s role. From that moment the machinery was set. The real subject watched as Mr. Wallace was led into an adjoining room, seated, and strapped down, with an electrode paste applied to his arm to prevent, the experimenter explained, blisters and burns. Then the teacher was taken to his own station, seated before the gray box, and given a sample shock of forty-five volts so that he would understand the apparatus was real. It stung. That sting was the last true shock anyone in the building would feel that day.

The climb

The task itself was tedious, which was part of its terrible power. The teacher read word pairs, then tested the learner’s recall. Each wrong answer required the next switch up, fifteen volts higher than the one before, announced aloud before it was pressed. For a while there was nothing dramatic. Then, at seventy-five volts, the first grunt of discomfort came through the wall. At one hundred and twenty volts, the learner shouted that the shocks were becoming painful. At one hundred and fifty, he demanded to be let out, invoking his heart. At two hundred and seventy, the protests became agonized screams. At three hundred, he pounded on the wall and then refused to answer any further questions at all.

And here the experiment revealed its cruelest hinge. The teacher was instructed to treat silence as a wrong answer. A refusal to respond meant another shock, and the voltage kept climbing into the region marked with red and the word DANGER, into the two switches at the end labeled only with an ominous XXX. Past a certain point the learner had gone completely silent, which any reasonable person would read as unconsciousness or worse. The teacher was told to continue regardless.

What did the teachers do? They suffered. Milgram’s records and film footage document men sweating through their shirts, trembling, digging their fingernails into their palms, breaking into fits of nervous, disbelieving laughter. This was not sadism. Nobody wanted the man in the next room to be hurt. Again and again, teachers turned to the experimenter in visible distress, asking whether they could stop, asking who would take responsibility, asking to check on the learner. The experimenter never raised his voice and never threatened. He simply delivered, in sequence, a short series of scripted prods. Please continue. The experiment requires that you to continue. It is absolutely essential that you continue. You have no other choice, you must go on ³. That was the entire arsenal of coercion: four calm sentences, spoken by a man in a gray coat with no power to enforce anything.

Two out of three

Before he ran the study, Milgram described the setup to a panel of senior psychiatrists and asked them to predict the results. Their estimate was reassuring and, as it turned out, spectacularly wrong. They believed that only a pathological fringe, perhaps one in a thousand, would proceed all the way to the final four-hundred-and-fifty-volt switch. Ordinary people, they reasoned, would break off long before then, most of them somewhere around the point where the learner first demanded release ⁴.

The actual figure from Milgram’s baseline condition was sixty-five percent. Nearly two out of every three ordinary men pressed every switch on the board, delivering what they believed to be a potentially lethal shock to a silent, possibly unconscious stranger, because a soft-spoken stranger in a lab coat told them the experiment required it. Not one subject, notably, stopped before the learner’s first serious protest at one hundred and fifty volts. The disobedience, when it came, came late and cost the disobedient enormous distress.

When Milgram published the findings in 1963 in the Journal of Abnormal and Social Psychology, they detonated across the discipline and beyond it ³. The result was so counterintuitive that people reached for reassuring explanations. The subjects must have known it was fake. They must have been a strange or aggressive sample. But the interviews and the physiological signs of stress suggested that most genuinely believed they were harming a man, and the point of the study was precisely that these were not strange people. They were the population, drawn from the newspaper for four dollars.

Not monsters, but situations

The most persistent misreading of Milgram is that it revealed a hidden reservoir of cruelty inside ordinary people, as though everyone is secretly waiting for permission to hurt. Milgram’s own conclusion was almost the opposite and more unsettling for it. His subjects were not cruel. They did not want to inflict pain, and inflicting it made them miserable. What overrode their conscience was not any inner sadism but an external arrangement: the presence of a legitimate authority who assumed responsibility, in a context the subjects had voluntarily entered and did not know how to gracefully leave.

Milgram called the psychological shift the agentic state, the moment a person stops seeing themselves as the author of their actions and begins to see themselves as an instrument carrying out another’s will ². In that state, the moral question quietly changes. It is no longer am I willing to hurt this man but am I willing to disrupt this situation, disobey this official, ruin this experiment. Those are much smaller questions, and most people, most of the time, will not disrupt.

What makes the research genuinely scientific rather than a single sensational demonstration is that Milgram did not stop at one version. He ran more than twenty variations, changing a single variable at a time to see what moved the needle. The results form a map of the conditions under which conscience reasserts itself. When the experimenter gave his orders by telephone rather than in person, full obedience dropped sharply, and some subjects secretly cheated by administering lower shocks than instructed. When the learner was moved into the same room, or when the teacher was required to physically press the learner’s hand onto a shock plate, obedience fell further still. And in perhaps the most hopeful variation, when the subject was placed alongside two confederate peers who defiantly refused to continue, the overwhelming majority took courage from their example and quit. Rebellion, like obedience, turned out to be contagious ².

The lesson embedded in these numbers is not that human beings are helpless before authority. It is that behavior is shaped by situation far more than we like to admit. The same man who presses every switch when an authority stands at his shoulder will stop the moment that authority steps out of the room, or the moment a single peer models refusal. We prefer to explain conduct in terms of character, good people and bad people, and Milgram’s variations suggest that character is doing less of the work than the arrangement of the room.

The long argument

No experiment this provocative survives sixty years without challenge, and Milgram’s has faced two serious lines of attack. The first concerns ethics. His subjects were deceived, subjected to acute distress, and in some cases left believing, for a time, that they had nearly killed a man. The experiment could not be run today in anything like its original form, and the modern framework of informed consent and ethics review boards exists in part because of studies like this one. The debate over whether the knowledge gained justified the distress caused has never fully closed.

The second challenge concerns accuracy. When the psychologist Gina Perry examined Milgram’s archived tapes and notes for her book Behind the Shock Machine, she documented that the experimenter often departed from the four-prod script, improvising additional pressure, and that a significant number of subjects appear to have doubted the shocks were real ⁵. Other scholars have argued that the headline figure of sixty-five percent applied cleanly only to one of many conditions, and that Milgram, a gifted communicator, sometimes presented his findings with more tidiness than the raw data allowed.

These criticisms matter, and they have complicated the simple story. What they have not done is overturn it. A carefully controlled partial replication conducted by Jerry Burger in 2009, designed to satisfy modern ethical standards by stopping at the one-hundred-and-fifty-volt point that historically predicted final outcome, found obedience rates strikingly close to Milgram’s own ⁶. Studies using immersive virtual reality, in which subjects know the victim is not real yet still show measurable stress, have echoed the pattern. Across cultures and across decades, when a legitimate authority issues a harmful instruction inside a structure the subject has agreed to enter, a large fraction of people comply. The exact number wobbles. The phenomenon does not vanish.

The rare one

It is tempting to file all of this under history, a grainy black-and-white lesson about the generation that built the bomb and watched the Eichmann trial. But the machinery Milgram exposed was never about the specific horrors of the twentieth century. It was about the ordinary human response to a uniform, a title, a calm confident voice, and the small daily reluctance to make a scene. That response is not confined to laboratories. It sits inside every institution: hospitals where junior staff hesitate to question a senior physician’s error, corporations where employees carry out policies they privately find indefensible, chains of command in which each link can honestly say the decision came from above.

Milgram’s most valuable gift may be the one his subjects lacked: a warning. He showed that the feeling of certainty about our own goodness is a poor guide to how we will act when the pressure is applied and the authority is standing close. The subjects who eventually walked out did not do so because they were braver by nature. They did so, the variations suggest, because something in the situation gave them permission: a phone call instead of a face, a colleague who refused first, a moment to remember they still had a choice. The person who steps back from an order that feels wrong is genuinely rare. But rarity is not the same as impossibility. And the first step toward being that person is to stop assuming you already are.

Sources

1. Arendt, Hannah, Eichmann in Jerusalem: A Report on the Banality of Evil, Viking Press, 1963. — https://en.wikipedia.org/wiki/Eichmann_in_Jerusalem
2. Milgram, Stanley, Obedience to Authority: An Experimental View, Harper & Row, 1974. — https://en.wikipedia.org/wiki/Obedience_to_Authority:_An_Experimental_View
3. Milgram, Stanley, ‘Behavioral Study of Obedience,’ Journal of Abnormal and Social Psychology, 1963. — https://doi.org/10.1037/h0040525
4. Blass, Thomas, The Man Who Shocked the World: The Life and Legacy of Stanley Milgram, Basic Books, 2004. — https://en.wikipedia.org/wiki/Stanley_Milgram
5. Perry, Gina, Behind the Shock Machine: The Untold Story of the Notorious Milgram Psychology Experiments, The New Press, 2013. — https://en.wikipedia.org/wiki/Behind_the_Shock_Machine
6. Burger, Jerry M., ‘Replicating Milgram: Would People Still Obey Today?’ American Psychologist, 2009. — https://doi.org/10.1037/a0010932

Related reading

• The Basement That Never Was
• The Safe Sadness of a Song

More from the Mind edition →