SeriesHidden Power: Dynasties, Secret Societies and the Invisible Mechanisms of Power Part 47 / 55

Why Do People Obey Authority?

Someone tells us to continue. The instruction comes from a person who speaks with confidence, wears a recognizable sign of professional standing, or claims to represent a respected institution.

Editorial illustration, not an archival image of Milgram’s or Asch’s experiments.

Someone tells us to continue. The instruction comes from a person who speaks with confidence, wears a recognizable sign of professional standing, or claims to represent a respected institution. At the same time, something within us warns that the action should stop. What prevails at that moment: our own judgment, trust in authority, the desire to belong, or the fear of being the only person to say no?

Research in social psychology offers no single explanation. But it does reveal how decisions can change when a person is no longer acting alone, but amid others' demands, the appearance of legitimacy, and responsibility for the consequences. That is precisely why these studies deserve careful reading—even where popular accounts have turned them into an overly simple lesson about “human nature.”

An Order in the Name of Science

In the early 1960s, psychologist Stanley Milgram studied at Yale University how far people would go in following an experimenter's instructions. Participants were told that the research concerned the effects of punishment on learning. In the role of “teacher,” they were required to press switches on a device labeled with progressively higher voltages, up to 450 volts, whenever the “learner” supposedly gave an incorrect answer. The “learner” was part of the experimental staging and received no such electric shocks; the real participant did not know that. To make the device more convincing, the participant was given an actual, weaker sample shock before the experiment began. [1]

In 1963, Milgram reported that, in the condition described in his published paper, 26 of 40 participants—65 percent—reached the highest marked level. Fourteen refused to continue before that point. That last fact is not a minor footnote: the experiment demonstrated not only compliance but also resistance, hesitation, and serious signs of distress. Milgram described tension, sweating, and nervous laughter; continuing did not necessarily mean that participants believed what they were doing presented no moral problem. [1]

An often-overlooked detail matters here. In the precise condition described in the 1963 paper, the “learner” did not loudly protest at 150 volts: the account describes pounding on the wall after the 300 V mark and again at 315 V. In later, better-known voice-feedback variations, the complaint came at 150 V. Combining these two scenes may make for a compelling story, but it sacrifices historical accuracy. [1] [4]

In the same paper, Milgram also recorded a prescribed sequence of four prompts to continue. An initial request escalated through references to the needs of the experiment to an explicit statement that the participant had no choice. So the question is not merely whether an “authority” was present, but how its demand was expressed and how the participant understood their own role. [1]

Why the Famous Percentage Is Not Enough

The figure of 65 percent became emblematic of Milgram's research. Yet a symbol quickly becomes misleading when presented as the proportion of all people who would carry out any harmful order in any system. Milgram did not conduct one single, unchanging experiment. The variations altered the circumstances, the physical proximity of the “learner,” the experimenter's position, the instructions, and the behavior of other people present. [1] [3]

In 2014, Nick Haslam, Steve Loughnan, and Gina Perry analyzed data from 21 conditions in which reaching the final level provided a comparable measure. Across those conditions, 740 people took part, and 323—43.6 percent—reached 450 V. The authors began with a broader set of 23 conditions but excluded two from this particular calculation. The combined figure is not a new universal measure of obedience; above all, it shows how much the story changes when we examine several experimental arrangements instead of just one. [3]

The synthesis identified associations with the way instructions were delivered, the consistency of the experimenter's demands, the relationship between “teacher” and “learner,” and pressure from others present to refuse. The authors themselves caution that many conditions were assembled as the research progressed and that statistical modeling cannot resolve every difficulty of comparison. A relationship between an experimental circumstance and an outcome does not fully explain the inner reasons of every individual participant. [3]

Across this range of results, a different research proposition becomes more persuasive: what matters is not only what kind of person someone is, but also the relationship in which they find themselves with the person giving orders and the person bearing the consequences. A laboratory result can help us ask questions about an institution, but it cannot, by itself, establish how any particular contemporary government, company, religious community, or secret society operates.

Sometimes No One Gives an Order—The Majority Is Enough

Where Milgram examined the relationship to an experimenter, Solomon Asch explored a different experience: a person faces a simple task of comparing the lengths of lines, yet the others present give one clearly wrong answer after another. The individual is the only person unaware that the others' responses have been coordinated in advance. When required to state their judgment in public, a correct perception can suddenly become socially uncomfortable. [2]

Schematic illustration of Asch’s task: one reference vertical line on the left and three comparison lines A, B and C on the right.
Illustration of the line-length comparison task in Asch’s conformity research; not an archival photograph of the experiment. Nyenyec, CC BY-SA 3.0, Wikimedia Commons. Image: Nyenyec / Wikimedia Commons CC BY-SA 3.0 Unported

In 1955, Asch reported on 123 participants in the series under discussion. On critical trials, when the majority deliberately offered an incorrect judgment, participants followed the mistaken majority in 36.8 percent of responses. This is not a claim that 36.8 percent of people were always conformist: what was measured was their responses in a particular situation. Roughly one-quarter of those studied never went along with an incorrect majority answer. [2]

Asch also systematically examined the effect of unanimity. The presence of one person who publicly offered a different judgment could greatly reduce the pressure of a unanimous majority. Significantly, the “ally” did not always have to give the completely correct answer: even a crack in the group's apparent unity could open space for independent judgment. But this was a finding from a particular visual task and experimental arrangement, not a guarantee that a lone dissenter can overcome every form of social coercion. [2]

Conformity and obedience can intersect in institutional life, but they should not be equated. In the first case, people adapt to peers or a group norm; in the second, they act in relation to instructions from an authority. Between the two lie voluntary cooperation, negotiation, belief in the rightness of a goal, and apparent consent. If we call all of this “blind obedience,” we fail to understand what actually happened. [2] [10]

Do We Follow Authority Because We Believe in It?

An established interpretation of Milgram's work emphasizes that a person within a hierarchy may hand responsibility over to a superior. But in 2012 Stephen Reicher, Alexander Haslam, and Joanne Smith proposed a different, complementary explanation: some participants cooperated actively because they identified with the experimenter and with what seemed to be an important scientific purpose. The authors call this engaged followership—actively following leadership that the participant experiences as legitimate. [5]

This explanation shifts attention from the bare order to the way leadership presents a shared undertaking. A participant may regard themselves as helping worthwhile research, while overlooking or pushing aside the question of what the “learner” is supposedly enduring. For understanding harmful actions, this is an uncomfortable but necessary insight: people may take part even while believing they are doing something good. A subjectively good intention does not, by itself, remove responsibility for one's actions. [5]

Yet this does not mean that engaged followership has been proved to be the only explanation. The interpretation draws on statistical associations across experimental variations, analyses of spoken exchanges, and archival material; it cannot reliably reconstruct every participant's motive. Fear of disagreement, confusion, trust, differing perceptions of the reality of the staging, and situational pressures remain relevant to the debate. We must choose between the evidence, not between two complete stories. [5] [7] [9]

Stephen Gibson's analysis of spoken exchanges draws particular attention to the fact that participants were not helpless recipients of instructions: they also negotiated with the experimenter. Gibson used actual audio recordings and transcripts from two conditions (“voice-feedback” and “women as subjects”); his examples raise questions about how standardized the procedure really was. The famous final prompt telling the participant that they had “no choice” did not emerge in these analyses as a universally effective means of securing compliance; it was often associated with refusal. Yet that is no reason to conclude that direct orders never work in real-life settings. [6] [8]

When Do We Say No?

Milgram's story is often told as though we are watching a person who simply cannot stop. But in the original report, fourteen people refused to continue despite facing the same basic experimental pressure. Their actions show that refusal was possible, though not easy or automatic. Individuals could respond differently to the same formal demand. [1]

The difference between isolated and supported dissent also matters. In some of Milgram's variations, resistance by other participants reduced continuation; in Asch's work, the presence of an ally greatly reduced conformity to a mistaken majority. But intellectual honesty also requires a less convenient finding: in Burger's 2009 partial replication, a single staged refusal did not produce a statistically significant reduction in continuation beyond the 150 V point. Different examples and different situations do not have identical effects. [2] [4]

With additional safeguards, Burger replicated part of the well-known procedure: 70 adults participated across two conditions, and the experiment was stopped at the crucial 150 V threshold. In the basic condition, 28 of 40 people—70 percent—indicated willingness to continue, compared with 19 of 30—63.3 percent—in the condition with a model who refused. In Milgram's comparable original condition, 33 of 40 people—82.5 percent—reached the corresponding point. These figures compare behavior at the 150 V threshold, not actual observed continuation by the modern participants all the way to 450 V. [4]

If we look to these studies for the possibility of personal freedom, we should not pretend they provide a simple manual for resistance. It is more useful to notice that doubt is not always a sign of weakness, that support from another person can change the balance of power, and that people do not respond identically even in the same test. The question of whether someone can object without suffering disproportionate punishment, however, already takes us beyond the laboratory and into the organization of actual institutions.

Researchers Were Not Outside Relations of Power

The obedience experiments did not merely study authority. They also created a sharply unequal relationship between researcher and participant: the researcher knew the true purpose, whereas the participant made decisions on the basis of a deceptive presentation. That very situation made it possible to study reactions, but it also generated serious ethical objections concerning distress, consent, and the consequences of the research. [1] [7]

In 1963, Milgram described procedures after the experiment intended to explain the staging to participants and reduce their tension. Later historical and archival examinations have questioned whether such explanations were always provided to the extent publicly described. Drawing on archival material, Ian Nicholson argued in 2011 that there may have been discrepancies between the public account of the research and the participants' experiences. His argument should be treated as a documented scholarly critique, not as a conclusion that all the data from Milgram's studies are therefore worthless. [1] [7]

More recent replications also have limits. Burger's study was not identical to the original experiment precisely because the degree of deception and exposure to stress had to be restricted. Its ethical improvement was also a methodological difference, preventing an unqualified comparison with 1963. A study that reveals human vulnerability does not thereby earn unlimited permission to expose people to that vulnerability. [4]

This also raises the question of how we interpret results. If we begin by believing that people are obedient by nature, we will see compliance above all in the records; if we are looking for courage, we will emphasize refusals. The scientific answer is more demanding: we must also examine original recordings, different conditions, participants' experiences, alternative explanations, and evidence that challenges our preferred thesis. [3] [7]

From the Laboratory to Real Hierarchies

Institutions depend on some degree of coordination. In a hospital, in traffic, or during emergency rescue, the complete absence of common rules could cause considerable harm. Obedience is therefore not inherently immoral. The morally and politically decisive question is when an instruction is legitimate, whom it serves, who checks its consequences, and whether there is room for responsible objection.

Laboratory experiments cannot tell us who secretly makes decisions within a particular network of power, who funds a particular operation, or why an individual chose to cooperate in a specific historical event. Such conclusions require documents about actual organizations, procedures, interests, and actions. Milgram's and Asch's figures cannot replace the investigation of real power relations. It would be equally mistaken to use those experiments to assign collective guilt to any nation, profession, or religion.

They do, however, show why it makes sense to look beyond a formal organizational chart when examining an institution. Do employees know who bears responsibility? Is an appeal to a “good cause” used instead of assessing concrete harm? Can someone request an explanation, report wrongdoing, and consult others without losing their dignity? These are verifiable questions about decision-making, not direct findings from a laboratory.

Individual courage matters, but it would be unfair to place the entire burden on the weakest member of a hierarchy. Institutions shape the rules, the hierarchy, and the consequences of resistance; those with power often have more opportunity to change that arrangement. The philosophical demand for inner freedom therefore does not rule out the need for external safeguards, transparency, and accountability in leadership.

The Boundary Between Trust and Surrendering Judgment

Why, then, do people obey authority? Sometimes because they rightly trust expertise and shared rules. Sometimes because they want to belong, believe in an important goal, fear standing apart, or do not know how to shoulder responsibility for refusal. And because a seemingly legitimate position can weaken attention to the person who bears the consequences. None of these possibilities alone explains every person or every event. [2] [3] [5]

Perhaps the most striking lesson of this research is not how quickly people relinquish judgment, but how often they must decide anew whether to exercise it. The studies show both compliance and resistance, and warn that people may justify their own part in harmful conduct even with good intentions. Inner freedom is therefore more than simply opposing every instruction: it is the willingness to assess the purpose, the means, and the consequences, and to accept responsibility for one's own actions.

Authority needs a measure of trust if it is to serve a community. But when it demands that trust replace conscience and reasonable questioning, it has crossed a boundary that its position alone cannot justify. The question for both individuals and institutions is the same: do we allow a person to say “no” in time, before someone else suffers the consequences?

Sources and further reading

  1. Stanley Milgram (1963), “Behavioral Study of Obedience,” Journal of Abnormal and Social Psychology, 67(4), 371–378. DOI: source. University-hosted copy of the article. Source 1 Source 2
  2. Solomon E. Asch (1955), “Opinions and Social Pressure,” Scientific American, 193(5), 31–35. DOI: source. University-hosted copy. Source 1 Source 2
  3. Nick Haslam, Steve Loughnan, and Gina Perry (2014), “Meta-Milgram: An Empirical Synthesis of the Obedience Experiments,” PLOS ONE, 9(4), e93927. source. Source
  4. Jerry M. Burger (2009), “Replicating Milgram: Would People Still Obey Today?”, American Psychologist, 64(1), 1–11. source. Published text at APA. Source 1 Source 2
  5. Stephen D. Reicher, S. Alexander Haslam, and Joanne R. Smith (2012), “Working Toward the Experimenter: Reconceptualizing Obedience Within the Milgram Paradigm as Identification-Based Followership,” Perspectives on Psychological Science, 7(4), 315–324. source. Source
  6. Stephen Gibson (2013), “Milgram's Obedience Experiments: A Rhetorical Analysis,” British Journal of Social Psychology, 52(2), 290–309. source. Author's version in a university repository; it may differ from the final publisher's version. Source 1 Source 2
  7. Ian Nicholson (2011), “‘Torture at Yale’: Experimental Subjects, Laboratory Torment and the ‘Rehabilitation’ of Milgram's ‘Obedience to Authority’,” Theory & Psychology, 21(6), 737–761. source. The publicly available abstract and bibliographic record were reviewed, not the full text. Source
  8. S. Alexander Haslam, Stephen D. Reicher, and Kathryn Millard (2015), “Shock Treatment: Using Immersive Digital Realism to Restage and Re-examine Milgram's ‘Obedience to Authority’ Research,” PLOS ONE, 10(3), e109015. source. Source
  9. S. Alexander Haslam, Stephen D. Reicher, Kathryn Millard, and Rachel McDonald (2015; first published online in 2014), “Happy to Have Been of Service: The Yale Archive as a Window into the Engaged Followership of Participants in Milgram’s Obedience Experiments,” British Journal of Social Psychology, 54(1), 55–83. source. The bibliographic description and abstract were reviewed; the full text was not examined. Source
  10. OpenStax (2020), Psychology 2e, § 12.4, “Conformity, Compliance, and Obedience.” source. Secondary textbook source. Source