The Stanford Prison Experiment

In brief

In August 1971, the psychologist Philip Zimbardo (1933 to 2024) converted the basement of Stanford's psychology building into a mock prison and randomly assigned 24 male college students to play prisoners or guards for a planned two weeks. He shut it down after six days. Several guards had grown cruel, several prisoners had broken down, and Zimbardo drew a sweeping conclusion: the prison itself, not the character of the people inside it, had produced the abuse. For nearly fifty years the study stood as the textbook demonstration of situational power over personal disposition. In 2018 the French researcher Thibault Le Texier, working from Zimbardo's own archived recordings and papers, published an investigation showing the guards had been coached toward harshness, the study's most cited breakdown was staged, and the exercise looked less like a controlled experiment than a demonstration built to prove a point Zimbardo already believed.

The full treatment

The problem it answers

By the early 1970s, Zimbardo wanted to understand why real prisons produced such reliable brutality. Was it the guards, drawn from a rough pool, or the institution itself, the uniforms, the keys, the walls? The U.S. Office of Naval Research funded the study because the Navy and Marine Corps wanted to understand conflict inside military prisons. Zimbardo's method was to remove the usual explanation, bad character, by handpicking the opposite: two dozen young men chosen from roughly seventy newspaper-ad respondents, selected because psychological screening found them stable, non-criminal, and unremarkable. If ordinary men turned cruel once placed inside a prison structure, the structure, not the men, would be doing the work.

How the simulation was built

Real Palo Alto police helped Zimbardo arrest the nine men assigned to be prisoners at their homes without warning, then booked and delivered them to the mock prison. There they were stripped, deloused, dressed in smocks with no underwear, chained at one ankle, and addressed only by a number stitched to the smock. Nine other men were assigned to be guards, given uniforms, batons, and mirrored sunglasses, and worked eight-hour shifts. Zimbardo did not stand outside the study as a neutral observer: he took the role of prison superintendent himself, a dual position he later admitted compromised his own judgment.

What happened over six days

By the second day, prisoners staged a rebellion, barricading their cell doors. Guards broke it with fire extinguishers and began escalating punishments: forced exercise, sleep deprivation, solitary confinement in a small closet participants called "the hole." One prisoner, Douglas Korpi, given the number 8612, had an apparent breakdown that became the study's most famous moment, screaming that he was "burning up inside." Zimbardo released him, and four more prisoners followed for similar distress before the study's planned end. It stopped entirely on day six because Christina Maslach, a recent Stanford psychology PhD (and Zimbardo's girlfriend, later his wife), visited the basement, watched guards march hooded prisoners to the toilet in chains, and told Zimbardo plainly that what he was doing was wrong. Of roughly fifty outside visitors who saw the study in progress, she was the only one who objected.

Deindividuation: the mechanism Zimbardo proposed

Zimbardo explained the guards' cruelty with a concept called deindividuation, first named by Leon Festinger, Albert Pepitone, and Theodore Newcomb in a 1952 paper describing how anonymity inside a group loosens a person's normal inner restraints. Zimbardo extended the idea in a 1969 chapter for the Nebraska Symposium on Motivation, arguing that anonymity, diffused responsibility, and immersion in a role reduce self-awareness, freeing behavior to follow the situation's norms rather than personal conscience. In the prison, he argued, the guards' sunglasses, the prisoners' numbers, and the novelty of near-total power inside a closed institution combined to strip away individual identity and substitute the role's own logic, one that trended toward domination once nothing checked it.

Related distinctions

The study is often confused with Stanley Milgram's obedience experiments of 1961 to 1963 at Yale. Milgram tested obedience to a direct order from a visible authority figure; Zimbardo's study, on his own account, tested immersion in an unscripted role with no explicit orders. If the guards were in fact given explicit instructions, the study collapses into a variant of Milgram's paradigm rather than the separate phenomenon it claimed to reveal.

Lineage

The concept of deindividuation traces to Gustave Le Bon's 1895 account of crowd psychology, in which individuals supposedly dissolve into an irrational collective mind, and was given empirical form by Festinger, Pepitone, and Newcomb in 1952. Zimbardo's 1969 elaboration sat inside a broader current of postwar social psychology, running through Solomon Asch's conformity studies and Milgram's obedience work, that treated the atrocities of the Second World War as evidence that situations, not monstrous individuals, could produce monstrous acts. The Stanford study was conceived as this tradition's most vivid proof.

The strongest case for it

Even granting every flaw in the design, the broader claim the study is used to illustrate, that institutional roles and permissive authority can push ordinary people toward cruelty, has independent support elsewhere. Milgram's obedience experiments, run under tighter control, found high compliance with instructions to deliver apparently painful shocks. Zimbardo later testified as an expert witness in the defense of a military police guard implicated in the 2004 Abu Ghraib prison abuse case, arguing the parallels were exact: minimal training, unclear rules, night shifts, no accountability. Whatever the flaws in its execution, the underlying claim, that circumstances matter more than intuition suggests when predicting who commits cruelty, remains a defensible corrective to the habit of explaining bad behavior entirely by bad character.

The strongest case against it

The strongest challenge arrived in 2018, when the French researcher Thibault Le Texier, given access to Zimbardo's archived audio recordings and unpublished materials at Stanford, published his findings first as the book Histoire d'un mensonge and then, in English, as "Debunking the Stanford Prison Experiment" in American Psychologist (2019). Working from the tapes, Le Texier documented that the guards were not left to invent their own behavior. David Jaffe, an undergraduate researcher who played the "warden," coached guards directly, telling at least one hesitant guard that the study aimed to demonstrate how bad prison conditions could get and pushing him to act tougher. Le Texier also found that not all guards behaved cruelly; several were, by participants' own accounts, fair or friendly, a fact that sits awkwardly with the claim that the situation reliably converted anyone placed inside it.

The same year, the journalist Ben Blum published "The Lifespan of a Lie" (2018), reporting that Douglas Korpi, whose breakdown had been the study's most famous moment, admitted decades later that he had faked it, wanting only to leave and study for an exam, and that "anybody who is a clinician would know I was faking." Blum and Le Texier together made the case that participants were less spontaneously transformed by their roles than responding to demand characteristics: cues within an experiment about what the experimenter wants, which participants then supply.

A further empirical challenge came earlier, in 2002, when the psychologists Stephen Reicher and S. Alexander Haslam ran a partial replication for the BBC, published as "Rethinking the Psychology of Tyranny: The BBC Prison Study" (2006). Unlike Zimbardo's guards, theirs failed to identify with the assigned role, hesitated to impose authority, and were overpowered by the prisoners, who built an egalitarian system of their own. Tyranny arrived only once participants identified strongly with a group whose norms sanctioned it, a process the authors call engaged followership rather than automatic deindividuation, contradicting the claim that any group placed in guard uniforms drifts toward brutality on its own. Critics have also noted that the original study was never published as a peer-reviewed journal article in the ordinary sense, ran with no control condition, used a self-selected sample of 24, and was led by a researcher who had abandoned neutrality by taking the superintendent's role himself.

Where it stands now

The events of August 1971 are documented and not seriously disputed: prisoners were stripped and humiliated, guards did impose degrading punishments, and the study did end early after Christina Maslach's objection. What has collapsed is the interpretation Zimbardo built on those events. Most academic psychologists now treat the study as a compromised demonstration rather than a scientific finding, and a growing number of introductory textbooks either drop it or reframe it around its own debunking. It survives in popular culture through The Lucifer Effect (2007), several feature films, and constant citation whenever institutional abuse invites a situational explanation. The live debate is no longer whether situations shape behavior, few psychologists doubt that, but how much weight one poorly controlled demonstration should carry.

Test yourself

Think of a role you have stepped into, a job title, a uniform, a position of minor power, that came with unstated expectations about how to act. Ask honestly whether you read those expectations off the bare situation, or whether someone, a boss, a culture, a script, was quietly coaching you toward them the whole time.