Philip G. Zimbardo, Craig Haney, W. Curtis Banks · 1971–1973

The Stanford prison experiment

If you put ordinary young men into the roles of guards and prisoners, does the situation alone make some of them cruel?

  • A mock prison in the basement of Jordan Hall, Stanford University's psychology building
  • Contested

The question

One common way to explain brutal prisons is to blame the people in them: violent inmates, sadistic guards. Philip Zimbardo and his students at Stanford wanted to test a different idea. What if you took young men with no history of violence, made some of them guards and some prisoners, and let the setting do the rest? If cruelty appeared anyway, the cause would be the situation, not the people.

How it was done

An ad in local newspapers asked for male students for a “psychological study of prison life”, paid $15 a day. About 75 men replied. After questionnaires and interviews, the researchers chose 24 judged the most stable and mature, and split them into guards and prisoners by the toss of a coin.

The “prison” was a short stretch of basement corridor in Jordan Hall, the psychology building, fitted with three small cells and a dark closet used for solitary confinement. With the help of the Palo Alto police, the prisoners were picked up at their homes in August 1971 without warning, charged with burglary or armed robbery, handcuffed, searched, fingerprinted and driven to the mock prison.

Guards worked eight-hour shifts and went home in between; prisoners stayed day and night. Guards were told not to use physical punishment. Zimbardo himself acted as prison “superintendent”, and an undergraduate assistant, David Jaffe, played the warden. The study was meant to last up to two weeks.

What came out

According to the published account by Craig Haney, Curtis Banks and Zimbardo (a 1972 report, published as an article in 1973), events moved fast:

On the sixth day Zimbardo stopped the study. The night before, Christina Maslach, a young psychologist who had just finished her doctorate at Stanford and was dating Zimbardo at the time, came to interview participants. She saw prisoners marched to the toilet with bags over their heads and confronted him about the suffering she had witnessed. Zimbardo has said that this was his wake-up call.

The authors’ conclusion was situational: the prison itself, not the personalities of the people in it, produced the cruelty and the breakdowns. They noted, though, that their sample was too uniform to test personality properly. Zimbardo returned to the study for decades, most fully in The Lucifer Effect (2007), where he linked it to the abuse of prisoners at Abu Ghraib.

What really happened (replications and critiques)

Doubts from the start. In 1975 the psychologists Banuazizi and Movahedi gave college students a description of the study, and asked what they thought would happen. Most guessed the researchers’ hypothesis, and about 90% predicted that guards would be oppressive and hostile. Their argument: participants may simply have played the roles they knew were expected, a problem psychologists call demand characteristics. Other researchers answered at the time that demand characteristics exist in all studies and do not by themselves explain the results.

What the archives show. After Zimbardo gave the study’s archive to Stanford’s library in 2011, the French researcher Thibault Le Texier went through the recordings and documents and interviewed 15 participants. His book appeared in French in 2018, and his main article in American Psychologist in 2019. His central points:

The breakdown. In a 2018 Medium article, the journalist Ben Blum reported that Douglas Korpi, the first prisoner released, now said he had faked his breakdown to get out and study for a graduate-school exam. Korpi had earlier described the experience as deeply upsetting.

Zimbardo’s reply. In a 2018 response, Zimbardo argued that a researcher who believes a breakdown is real must treat it as real, and that Korpi’s accounts had changed over time. He said the warden’s push for toughness was mild, that physical force was forbidden, and that the guards chose how far to go. If the guards were simply doing what they thought was wanted, he asked, why did some stay kind? He accepted that the study had flaws, and called the lack of an independent scientific observer a major flaw.

Who volunteers. In 2007 the psychologists Carnahan and McFarland placed two nearly identical ads, one mentioning “prison life” and one not. People who answered the prison ad scored higher on aggressiveness, authoritarianism, Machiavellianism, narcissism and social dominance, and lower on empathy and altruism. The authors said this favours an interaction between person and situation, though its implications for the 1971 study are still uncertain.

The BBC Prison Study. In December 2001 Stephen Reicher and Alex Haslam ran a prison study with the BBC (broadcast in 2002). It was planned for up to ten days and ran for eight, under an independent ethics committee. Guards were randomly assigned and given power, but no instructions on how to use it. The guards did not come to identify with their role and were uneasy using their authority. The prisoners built a shared identity and eventually overpowered them. A self-governing commune followed, then fell apart, and some participants began to favour a harsh new regime. Reicher and Haslam concluded that people turn to tyranny more readily when their groups fail and they feel powerless, not because roles turn people cruel automatically. Zimbardo responded that a televised study was not a valid replication.

Ethics

The study was approved by Stanford’s human subjects committee and funded by the US Office of Naval Research, and in 1973 a review requested from the American Psychological Association concluded that the guidelines of the time had been followed. Participants signed consent forms, though these did not warn of the intense distress that followed.

The main problems are clear in hindsight, and Zimbardo acknowledged several himself. As superintendent, he was both the person running the prison and the scientist meant to watch it, and he later wrote that this role absorbed him and pulled his attention away from the research. In The Lucifer Effect he described his failure to intervene as a “sin of omission” and apologised.

Today the study would not be approved in this form: no independent person with the power to stop it, a researcher in charge of the situation he was studying, and participants who did not feel free to leave.

What it means today

The Stanford Prison Experiment is better understood as a dramatic demonstration than as a controlled experiment. It does not prove that anyone handed a uniform will become cruel; that strong claim is contested, and the evidence for it is weaker than its fame suggests.

What survives the critiques is more interesting:

If they are right, that is uncomfortable in its own way. It suggests that harm is often done by people who think they are doing the right thing, which also means it can be questioned and refused.

Would you have passed?

Coming soon. There is no honest way to simulate this one, so we won’t try. Instead, a few short questions: what you would have done as a guard if a colleague pushed you to be tougher, as a prisoner if you could not tell whether you were allowed to leave, and as Maslach, walking in from outside. There are no scores. At the end you’ll see what the real participants did, and how each answer connects to the debate above.

Questions people ask

How long did the Stanford Prison Experiment last?

It was planned for up to two weeks and was stopped after six days, in August 1971. Zimbardo ended it the morning after Christina Maslach, a young psychologist who had just finished her doctorate at Stanford, objected to what she saw.

Was the Stanford Prison Experiment faked?

Not faked, but seriously compromised. Archives studied by Thibault Le Texier show that guards were briefed on the study's aims and encouraged to be tough, which the original report did not make clear. Zimbardo rejects the charge that this explains the guards' behaviour.

Has the Stanford Prison Experiment been replicated?

Not in full. The BBC Prison Study (2001–2002), run by Stephen Reicher and Alex Haslam, gave guards no instructions on how to behave. The guards did not bond around their role, and the prisoners organised and challenged them.

Was the Stanford Prison Experiment ethical?

It was approved at the time, but it would not pass today. Zimbardo later named his double role as researcher and prison superintendent, and the absence of an independent observer, as serious flaws.

Sources

  1. Haney, C., Banks, C., & Zimbardo, P. (1972). Interpersonal dynamics in a simulated prison (Technical Report Z-09). Stanford University, prepared for the Office of Naval Research. apps.dtic.mil/sti/pdfs/AD0751041.pdf
  2. Haney, C., Banks, W. C., & Zimbardo, P. G. (1973). Interpersonal dynamics in a simulated prison. International Journal of Criminology and Penology, 1, 69–97. www.ncjrs.gov/pdffiles1/Photocopy/10301NCJRS.pdf
  3. Stanford Prison Experiment official website (prisonexp.org). Frequently asked questions. www.prisonexp.org/faq
  4. Zimbardo, P. G. (2018). Philip Zimbardo's response to recent criticisms of the Stanford Prison Experiment. prisonexp.org. academicweb.nd.edu/~rwilliam/ndonly/readings/Methods/01-Experimentation/ZimbardoDefense2018.pdf
  5. Zimbardo, P. (2007). The Lucifer effect: Understanding how good people turn evil. Random House.
  6. Le Texier, T. (2019). Debunking the Stanford Prison Experiment. American Psychologist, 74(7), 823–839. doi:10.1037/amp0000401
  7. Le Texier, T. (2018). Histoire d'un mensonge: Enquête sur l'expérience de Stanford. Paris: La Découverte (Zones).
  8. Blum, B. (2018, June 7). The lifespan of a lie. Medium (Trust Issues).
  9. Live Science (2018). One of psychology's most famous experiments was deeply flawed. www.livescience.com/62832-stanford-prison-experiment-flawed.html
  10. Banuazizi, A., & Movahedi, S. (1975). Interpersonal dynamics in a simulated prison: A methodological analysis. American Psychologist, 30, 152–160.
  11. Reicher, S., & Haslam, S. A. (2006). Rethinking the psychology of tyranny: The BBC Prison Study. British Journal of Social Psychology, 45(1), 1–40. doi:10.1348/014466605X48998
  12. Reicher, S., & Haslam, S. A. (2006). Tyranny revisited: Groups, psychological well-being and the health of societies. The Psychologist (British Psychological Society). bps.org.uk/psychologist/tyranny-revisited-groups-psychological-well-being-and-health-societies
  13. 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(5), 603–614. doi:10.1177/0146167206292689
  14. Wikipedia. Stanford prison experiment (accessed November 2026). en.wikipedia.org/wiki/Stanford_experiment
  15. Ratnesar, R. (2011). The menace within. Stanford Magazine, July/August 2011. stanfordmag.org/contents/the-menace-within