Stanford Prison Experiment — What Really Happened

In August 1971, in the basement of Stanford University's psychology building, Philip Zimbardo turned 24 college men into guards and prisoners and watched what happened. Within six days he had to shut it down. For half a century, the Stanford Prison Experiment was taught as proof that ordinary people will become brutal when handed power and a uniform. In the last decade, that famous lesson has unraveled.

The Setup

Zimbardo recruited 24 male volunteers via newspaper ad ("$15/day for psychological study of prison life") and screened them for stability. He randomly assigned half to be guards and half to be prisoners. A wing of Stanford's Jordan Hall was rebuilt into a mock prison with cells, a "solitary confinement" closet, and surveillance.

Prisoners were "arrested" by real Palo Alto police, fingerprinted, and processed. They wore smocks, stocking caps, and ID numbers. Guards wore khaki uniforms and mirrored sunglasses, carried wooden batons, and worked 8-hour shifts. Zimbardo himself played "Prison Superintendent."

The study was planned for two weeks. It was terminated after six days, reportedly because the situation had spiraled out of control.

The Story Zimbardo Told

According to Zimbardo's published accounts and his subsequent books and TED talks:

• Some guards became spontaneously sadistic, humiliating prisoners, forcing pushups, and inflicting petty cruelties.
• One prisoner had a psychological breakdown within 36 hours.
• Prisoners became apathetic and compliant; a few rebelled and were punished.
• Zimbardo's then-girlfriend (later wife) Christina Maslach visited on day 5, was horrified, and convinced him to end the study.
• The lesson: situations override personalities. Ordinary people, given roles of unchecked power, behave monstrously. Roles of powerlessness produce learned helplessness.

The study became one of the most cited in psychology. It made Zimbardo a celebrity. It informed military training, prison reform debates, and his expert testimony at the Abu Ghraib courts-martial in 2004. It inspired films (Das Experiment 2001, The Experiment 2010, The Stanford Prison Experiment 2015).

The Story That Fell Apart

Starting in the early 2000s, and exploding into mainstream attention with Ben Blum's 2018 Medium piece "The Lifespan of a Lie" and Thibault Le Texier's 2019 book Histoire d'un mensonge ("History of a Lie"), archival research has dismantled most of the textbook narrative.

Le Texier obtained Stanford's archived recordings, memos, and pre-study briefings. What he found:

• Zimbardo and his team coached the guards to be tough. In a recorded briefing on day zero, Zimbardo's research assistant David Jaffe told guards to create "an atmosphere where the prisoner feels powerless... we have to be the ones who do this, and the way we'll do that is by giving them no privacy." This is not spontaneous role-induced cruelty; it is performance directed by the principal investigator.
• Guards who refused to be harsh were criticized. One guard, John Mark, later said he tried to be a "good guard" and was made to feel he was failing the study.
• The famous "breakdown" prisoner (Doug Korpi) admitted decades later that he was faking — he wanted out so he could study for his graduate exams, and the meltdown was theater.
• The script was tight. Guards followed instructions, not their nature. The infamous "count" rituals, the pushups, the cell stripping, even the sleep deprivation — many were Zimbardo's idea, not the guards'.
• The sample was self-selected and primed. The ad specified "prison life," which research by Carnahan and McFarland (2007) showed attracts volunteers higher in aggression, authoritarianism, and Machiavellianism than ordinary volunteers for "psychological research."
• There was no control group. No blind condition, no statistical analysis, no peer review of the methodology before publication. Zimbardo published the most famous version of the results not in a refereed journal but in Naval Research Reviews.

By any modern methodological standard, the SPE is not a controlled experiment. It is a piece of demonstration theater that Zimbardo himself increasingly came to inhabit as "the Superintendent."

The BBC Prison Study (2002)

When Stephen Reicher and Alex Haslam tried to replicate the SPE for a BBC documentary in 2002, with real ethical oversight and no coaching of guards, they got the opposite result.

• Guards were reluctant to assert authority and uncomfortable in the role.
• Prisoners formed solidarity, organized collective resistance, and eventually negotiated power-sharing with the guards.
• The study evolved into a fascinating exploration of group identity formation, not a confirmation of automatic role-corruption.

Reicher and Haslam concluded that brutal behavior in prison-like settings is not automatic — it emerges only when guards identify with a project (an authority's mission) that legitimizes it. Sound familiar? It is the same revised mechanism that came out of the Milgram experiment reanalysis: not blind obedience, but identification.

What Was Real

Strip away the inflated claims and a few things remain true:

1. The setting was theatrical and immersive. Even with coaching, the experience was distressing for real people, and that distress is documented.
2. Zimbardo himself lost perspective, becoming so absorbed in his Superintendent role that he authorized practices a researcher should have flatly refused. This meta-lesson — that the experimenter can become a participant — is arguably the study's most honest finding.
3. Authority structures can enable abuse even when no one in the structure is personally cruel. The Abu Ghraib parallels Zimbardo drew at the 2004 trials were not crazy — but they did not require the SPE to make.

Comparison: SPE vs BBC Prison Study vs Milgram

Feature	Stanford Prison (1971)	BBC Prison Study (2002)	Milgram (1961)
Question	Do roles corrupt?	Same	Will people obey authority?
Setup	Guards/prisoners, 6 days	Guards/prisoners, 8 days	Shock teacher/learner
Coaching	Explicit, recorded	None	Standardized prompts
Brutality emerged	Yes (per Zimbardo)	No — solidarity emerged	Most went to 450V
Control group	None	Behavioral monitoring	Multiple variation conditions
Methodology	Demonstrational	Peer-reviewed protocol	Controlled experiment
Modern verdict	Methodologically invalid	Robust, replicated framing	Real effect, overstated original numbers
Lesson today	How NOT to do science	Identity drives group behavior	Authority enables harm via identification

Why Textbooks Still Teach It

Mahzarin Banaji and others have argued that the SPE belongs in textbooks — but for the opposite reason than it was first included. Students should learn how a single charismatic researcher, lacking controls, can produce a celebrated finding that does not hold up. The SPE is now a case study in the replication crisis alongside ego depletion, social priming, power posing, and the marshmallow test reanalysis.

There is a meta-irony: the most famous demonstration of "situations corrupt people" is itself a demonstration of how situations (publishing pressure, media attention, charismatic authority) can corrupt science.

Real-World Parallels

The SPE's enduring cultural power comes from the disturbing pattern matches it offered:

• Abu Ghraib (2003) — U.S. soldiers torturing Iraqi detainees. Zimbardo testified that situational factors, not individual evil, drove the abuse. Critics counter that command policy explicitly encouraged "softening up" detainees — a coached condition more like SPE than a spontaneous one.
• Police brutality patterns — studies of department cultures suggest leadership and training matter far more than uniform-issued power.
• Workplace abuse hierarchies — the situational lens helps, but identity and management practice matter more than role labels alone.

In every case, the original SPE framing ("anyone in a guard uniform becomes a sadist") oversimplifies. The Reicher/Haslam framing ("guards become harsh when they identify with a leadership that legitimizes harshness") fits the evidence better.

Related Reading

• Read the Milgram experiment explained for the companion classic, with similar methodological caveats.
• Our piece on how the human brain works covers the neuroscience of social conformity.
• What is consciousness explores how role immersion changes subjective experience.

FAQ

Q: Did the guards really become sadistic on their own? A: The original story said yes. Archival recordings now show that Zimbardo's team coached guards to be harsh and criticized those who weren't. The "spontaneous" cruelty was substantially scripted.

Q: Did the prisoner who broke down really break down? A: Doug Korpi, the prisoner whose recorded screaming became iconic, admitted in interviews decades later that he was performing. He wanted to leave so he could study for graduate exams; faking distress seemed the fastest exit.

Q: Why was the experiment stopped early? A: Zimbardo's then-girlfriend Christina Maslach visited on day 5, was alarmed by what she saw, and told him to end it. Zimbardo credits her intervention with snapping him out of his "Superintendent" mindset.

Q: Is the experiment formally retracted? A: No, but the field has largely abandoned its conclusions. Most modern social psychology textbooks now teach the SPE alongside the critical reanalysis rather than as established fact.

Q: What did the BBC replication show? A: The opposite. With proper ethical oversight and no coaching, prisoners formed solidarity and challenged the guards, who were reluctant to assert authority. The study evolved into an examination of group identity, not role-induced brutality.

Q: What's the real takeaway about evil and power? A: Roles alone don't make people brutal. Brutality emerges when people identify with an authority structure that legitimizes harm, when explicit instructions push them toward it, and when norms support escalation. Selection effects, leadership, and culture matter more than the simple costume of "guard."

Sources

• Haney, C., Banks, C., Zimbardo, P. (1973). "A study of prisoners and guards in a simulated prison." Naval Research Reviews 9.
• Le Texier, T. (2019). Histoire d'un mensonge: Enquête sur l'expérience de Stanford.
• Blum, B. (2018). "The Lifespan of a Lie." Medium.
• Reicher, S. & Haslam, S. A. (2006). "Rethinking the psychology of tyranny: The BBC Prison Study." British Journal of Social Psychology 45.
• Carnahan, T. & McFarland, S. (2007). "Revisiting the Stanford Prison Experiment: Could participant self-selection have led to the cruelty?" Personality and Social Psychology Bulletin 33.
• Griggs, R. A. (2014). "Coverage of the Stanford Prison Experiment in introductory psychology textbooks." Teaching of Psychology 41.
• Zimbardo, P. (2007). The Lucifer Effect: Understanding How Good People Turn Evil.