Back to the directoryFind in the timeline
Milgram's obedience experiment
At Yale, 26 of 40 men gave what they believed were shocks up to 450 volts when told to continue. The result is documented. Its ethics, its archive and its explanation as simple obedience are contested.
- Where it sits
- Academic psychologyResearch published for other researchers and students.
- Evidence level
- ContestedThe evidence conflicts or is disputed, for example after failed replications.
- Sources inspected
- 5 of 53 with the relevant text, company document or archive item checked, 2 as an abstract or reference article, 0 not verified.
What happened
At Yale, Stanley Milgram recruited 40 men aged 20 to 50 by newspaper advertisement and direct mail for what they believed was a study of memory and learning. Each was told to give a learner a shock for every wrong answer, one level higher each time. The shocks were not real, and the learner was a trained confederate. The first report appeared in 1963.
The psychological idea
The shock generator had 30 switches labeled from 15 to 450 volts. When a subject hesitated, the experimenter used verbal prods to continue. The learner made no protest until 300 volts, then pounded on the wall and stopped answering. Milgram asked how far subjects would go when an authority told them to carry on.
In practice
No commercial application is documented. In the baseline study 26 of 40 subjects (65 per cent) went to 450 volts, and the other 14 stopped between 300 and 375 volts. Fourteen Yale seniors had predicted that about 1.2 per cent would go all the way. Fourteen subjects laughed or smiled nervously and three had uncontrollable seizures, one so severe that the session was halted. Subjects were interviewed afterward and the deception was explained.
Why it mattered
The baseline result showed most subjects continuing to the top of the scale despite the learner's protests. Haslam, Loughnan and Perry pooled 21 of Milgram's conditions, with 740 participants, and found that 43.6 per cent reached 450 volts, with lower obedience under group pressure to disobey and with closer teacher-learner proximity. Any link to consumer persuasion is interpretation.
Limits and qualifications
Baumrind published an ethics response in 1964, and Milgram's own report describes subjects who sweated, trembled and stuttered. Perry's archival work, as cited by Haslam and colleagues, points to discrepancies between how the study was presented and how it ran, and to skepticism among many participants. Haslam and Reicher argue the data fit engaged followership: compliance reflected identification with the experimenter's scientific aims, and a direct order tended to produce refusal. Burger's 2009 partial replication stopped at 150 volts with 70 adults and found obedience only slightly lower than Milgram's.
Evidence behind the claims
Milgram's subjects were 40 men aged 20 to 50 recruited by newspaper advertisement and direct mail, who believed they were joining a Yale study of memory and learning; the generator had 30 switches from 15 to 450 volts, and the learner, a trained confederate, did not protest until 300 volts, when he pounded on the wall and stopped answering.
26 of 40 subjects (65 per cent) obeyed fully to 450 volts, and the 14 who defied the experimenter stopped between 300 and 375 volts.
Fourteen Yale seniors predicted a mean of 1.2 per cent full obedience; 14 of 40 subjects showed nervous laughing or smiling; three had full-blown uncontrollable seizures, one so violently convulsive that the session was halted; subjects were observed to sweat, tremble and stutter; and subjects were interviewed and the hoax explained afterward.
Across 21 of Milgram's conditions with 740 participants, 43.6 per cent (323) reached 450 volts, and obedience was lower with group pressure to disobey and with closer teacher-learner proximity.
Haslam and Reicher argue that participants were engaged followers whose compliance reflected identification with the experimenter, report that the prod 'The experiment requires that you continue' tended to work while the direct order 'You have no other choice' typically led to refusal, and note that Milgram ran 30 or so variants.
Haslam, Loughnan and Perry cite Perry's 2012 Behind the Shock Machine for troubling discrepancies between how the study was represented and how it was run, and note apparent skepticism among many participants; Baumrind published an ethics response in American Psychologist in 1964.
Burger's 2009 partial replication of Milgram's Experiment 5 with 70 adults stopped at the 150-volt point and found obedience rates only slightly lower than those Milgram found.
People involved
- Stanley Milgram. Experimenter, Yale University
- Diana Baumrind. Author of a 1964 ethics critique
- Gina Perry. Author of an archival study, Behind the Shock Machine (2012), and co-author of the Meta-Milgram synthesis
- S. Alexander Haslam. Psychologist, co-author of the engaged followership account
- Stephen Reicher. Psychologist, co-author of the engaged followership account
- Jerry M. Burger. Author of the 2009 partial replication
In the Psy101 course
- OpenStax Psychology 2e, section 12.4: Conformity, Compliance, and ObedienceChapter 12, Social Psychology. Concept. The section discusses this topic.The reviewer opened the page: it describes Milgram's obedience experiment at Yale and says 65 per cent continued to the maximum voltage.
- OpenStax Psychology 2e, section 2.4: EthicsChapter 2, Psychological Research. Ethics. The section discusses this topic.The reviewer opened the page: it covers informed consent, deception, debriefing and review boards. It does not name Milgram.
This is course relevance only. It is not evidence for the historical or corporate claims on this page.See the full course map.
Sources and inspection
5 of 5 sources were opened and read at the level shown. A source marked as not verified supports no claim on this page by itself.
- [1]Behavioral Study of ObedienceMilgram, S.. Journal of Abnormal and Social Psychology, 67(4), 371-378 (Internet Archive copy), 1963. Journal article. Primary source.Relevant text checkedRead the OCR text of the 1963 article from an Internet Archive scan as returned by the fetch tool: subjects, apparatus, procedure, results, tension and interview passages. The scan is a community upload and the OCR is partly garbled at the debriefing passage.
- [2]Contesting the Nature of Conformity: What Milgram and Zimbardo's Studies Really ShowHaslam, S. A.; Reicher, S. D.. PLOS Biology, 10(11), e1001426, 2012. Journal article. Critique or replication.Relevant text checkedRead the open-access essay as returned by the fetch tool, for the Milgram and Stanford passages. It argues for a social identity account and relies partly on the authors' own data.
- [3]Meta-Milgram: An Empirical Synthesis of the Obedience ExperimentsHaslam, N.; Loughnan, S.; Perry, G.. PLoS ONE, 9, e93927, 2014. Journal article. Critique or replication.Relevant text checkedRead the open-access article as returned by the fetch tool: abstract, pooled results, discussion and limitations.
- [4]Some thoughts on ethics of research: After reading Milgram's Behavioral Study of ObedienceBaumrind, D.. American Psychologist, 19(6), 421-423, 1964. Journal article. Critique or replication.Abstract or publisher record onlyCrossref bibliographic record only (title, journal, volume, pages, year). The article was not opened, so its arguments are not described.
- [5]Replicating Milgram: Would people still obey today?Burger, J. M.. American Psychologist, 64(1), 1-11, 2009. Journal article. Critique or replication.Abstract or publisher record onlyRead the abstract through the Europe PMC record. The article text was not opened.
Access note: Perry's book and Baumrind's article were not opened, and Burger's paper was read as an abstract only. Statements about Perry and Baumrind rest on Haslam and colleagues or on a bibliographic record.
Independent source review: Claude review agent, independent of the research agent, 2026-10-21. Verdict: pass-with-edits.