Obedience to Authority: Milgram, Agency, Responsibility, and Social Influence
Updated: 4 days ago
Author: Ukrainian Psychological Hub · Published: September 24, 2026 · Editorial Policy
Obedience to authority is a form of social influence in which a person follows a directive from someone understood to have the standing to issue it. The phenomenon is psychologically important because an order can reorganize attention, choice, responsibility, and action without physically eliminating the person's capacity to refuse. Obedience therefore sits at the intersection of authority, legitimacy, social pressure, agency, moral judgment, and institutional structure.
The best-known evidence comes from Stanley Milgram's laboratory studies at Yale University. In the baseline condition reported in 1963, 26 of 40 participants continued to the maximum 450-volt switch—65 percent—even though many showed marked tension and protested along the way (Milgram, 1963). That result became one of psychology's most famous numbers. It also became one of its most frequently oversimplified numbers.
Milgram did not discover that “65 percent of people obey authority.” His program contained many conditions, and obedience changed substantially when the social configuration changed. A later empirical synthesis of 21 Milgram conditions involving 740 participants found that 43.6 percent reached the maximum shock level overall, while features such as experimenter legitimacy and directiveness, peer behavior, physical and psychological distance, and the relationship between experimenter, teacher, and learner were associated with different levels of obedience (Haslam, Loughnan, & Perry, 2014). The central lesson is contextual: people can comply with disturbing instructions under authority, but obedience is neither automatic nor fixed at one universal rate.
Modern research has also changed how psychologists interpret what happened inside the Milgram paradigm. Milgram's historical “agentic state” account proposed that people can experience themselves as instruments executing another person's wishes and shift responsibility upward to the authority. Contemporary work on coercion does support a narrower claim: being ordered can reduce implicit measures of the sense of agency and alter how people process the relation between their actions and outcomes (Caspar et al., 2016; Akyüz et al., 2024). That evidence does not establish Milgram's full theory, and a reduced sense of agency does not erase causal, moral, professional, or legal responsibility.
What Is Obedience to Authority?
In psychological research, obedience usually refers to behavior performed in response to an explicit command, instruction, or demand from a person or institution perceived as having authority. The authority relation matters. A stranger asking you to move a chair, a physician issuing a medical order within a clinical role, and a supervisor directing an employee may use similar words, yet the psychological meaning of the request differs because the perceived entitlement to direct behavior differs.
Authority is therefore not synonymous with force. Tom Tyler's influential review defines legitimacy as a psychological property that makes an authority or social arrangement seem appropriate, proper, and just; legitimacy can produce voluntary deference because people feel they ought to accept rules or decisions, rather than merely fearing punishment or seeking reward (Tyler, 2006). This helps explain why authority can guide behavior even when immediate coercion is weak.
At the same time, legitimate authority and coercive pressure can coexist. A person may initially accept an institution as legitimate and later encounter an instruction they regard as excessive, unethical, or outside the authority's proper scope. Obedience is therefore best understood as a process rather than a binary personality property: people interpret who is giving the order, what role that person occupies, whether the order belongs within that role, what consequences are expected, how others are behaving, and whether refusal feels possible.
Obedience, compliance, conformity, influence, and control
Obedience overlaps with other forms of social influence but should not be used as a synonym for all of them. Compliance is the broader act of changing behavior in response to a request or demand. Conformity usually concerns adjustment toward group norms, expectations, or the behavior of other people rather than a direct order from a recognized authority. Influence is broader still: it can occur through persuasion, information, modeling, norms, status, incentives, relationships, or control of outcomes. Personal control concerns influence over one's own outcomes, while social power concerns capacity to affect other people's outcomes. The distinction between power and recognized authority is developed in Power vs Authority: Weber, Legitimacy, Influence, and Obedience. For a direct comparison of group conformity and authority-based obedience, see Conformity vs Obedience: Asch, Milgram, Social Pressure, and Authority.
This distinction matters for evidence. A study showing conformity to a group cannot automatically be treated as evidence about obedience to authority. A study of coercion may illuminate part of obedience without capturing legitimacy. A study of leadership may involve voluntary followership rather than direct commands. Keeping the constructs separate makes it possible to ask what the authority relation itself contributes.
Why Do People Obey Authority?
There is no single psychological mechanism that explains obedience across settings. The evidence points to a combination of perceived legitimacy, role expectations, uncertainty, gradual commitment, social identification, interpersonal pressure, proximity, peer behavior, and the immediate structure of the situation. Different mechanisms can operate together, and the same instruction can be experienced differently by different people.
Legitimacy and role expectations
People enter institutions with expectations about who is entitled to decide what. Hospitals, laboratories, schools, courts, workplaces, emergency organizations, and families assign roles that coordinate action. When a directive fits a recognized role, compliance can require less active justification because the authority relationship already provides a default rule for whose judgment should guide behavior. Tyler's legitimacy research explains why accepted authority can generate deference without constant surveillance or punishment (Tyler, 2006).
Legitimacy is also conditional. Perceived expertise may support authority in one domain but not another. Formal office can authorize some decisions while leaving other decisions outside the role. A person's willingness to obey can therefore depend on whether the command seems procedurally appropriate, substantively credible, and connected to the authority's recognized function. This is one reason Authority in Psychology: Legitimacy, Expertise, Trust, and Obedience is a neighboring but distinct canonical topic: that article owns the psychology of authority itself, while the present article owns the behavioral phenomenon of obedience.
Uncertainty and informational dependence
Authority becomes especially influential when people are uncertain about what is happening, what the correct procedure is, or what consequences a decision will have. In ambiguous situations, the authority may appear to possess superior information, responsibility, expertise, or institutional backing. Obedience can then be partly informational: the individual treats the authority's instruction as evidence about what is appropriate.
The presence of uncertainty also complicates moral interpretation. A participant who continues because they genuinely believe the experimenter knows the learner is safe is psychologically different from a participant who believes harm is occurring but continues because refusal feels socially difficult. Observable behavior can look similar while the beliefs and motives behind it differ. Research on obedience therefore benefits from examining interaction, interpretation, and subjective experience rather than classifying behavior only by the final switch reached.
Gradual escalation and commitment
Milgram's shock generator increased in small increments. This matters because each step was close to the previous one. Refusing at one point could require a participant to explain why the immediately preceding step had been acceptable but the next one was not. Gradual escalation can create behavioral momentum: past compliance becomes part of the situation that the person must reinterpret if they want to stop.
This does not mean that gradual escalation mechanically produces obedience. Participants repeatedly hesitated, questioned the experimenter, expressed concern, and sometimes stopped. The more precise claim is that sequential commitment can increase the psychological cost of drawing a new boundary after a series of smaller decisions. It operates alongside, rather than instead of, authority, legitimacy, uncertainty, and social pressure.
Proximity, social distance, and the presence of others
Milgram varied how close the teacher was to the learner and how close the experimenter was to the teacher. He also manipulated institutional setting and peer behavior. The later Meta-Milgram synthesis found that the likelihood of reaching the maximum shock was associated with multiple features of the authority–participant–learner configuration, including experimenter directiveness, legitimacy, consistency, peer pressure toward disobedience, and the indirectness or intimacy of the teacher–learner relationship (Haslam, Loughnan, & Perry, 2014).
That pattern is one of the strongest reasons to resist trait-only explanations. The same broad paradigm produced materially different behavior when relationships and social cues changed. Obedience is not simply something located “inside” an obedient person; it is produced in an interaction between a person and a structured social situation.
What Milgram's Experiments Actually Showed
Milgram's original 1963 report described 40 men recruited from the New Haven area, aged 20 to 50 and representing a range of occupations and educational levels. They were instructed by an experimenter to administer increasingly severe shocks to a learner after errors on a memory task. The learner was an actor and did not receive the shocks, but the participant was led to believe the apparatus was real. Twenty-six participants continued to the 450-volt endpoint; 14 refused before it (Milgram, 1963).
The study established a striking possibility: ordinary adults could continue a harmful-seeming procedure under an experimental authority despite conflict, protest, and visible distress. It did not establish that most people always obey authorities, that people become mindless under orders, or that responsibility disappears when an authority is present.
The 65 percent figure is a condition-specific result, not a human constant
The baseline 65 percent is often repeated as if it measured a stable population tendency. Milgram himself conducted variations that changed the relation among experimenter, participant, learner, peers, and institution. Later synthesis across 21 conditions found an overall maximum-voltage rate of 43.6 percent and systematic variation associated with features of the situation (Haslam, Loughnan, & Perry, 2014).
The scientifically responsible interpretation is therefore conditional: under some authority structures, substantial proportions of participants complied with instructions that they experienced as troubling; under other structures, resistance was much more common. The phenomenon is robust enough to demand explanation, but the famous percentage should not be universalized beyond the procedure that produced it.
Obedience was not emotionally effortless
Milgram reported intense nervous tension in many participants, including sweating, trembling, stuttering, nervous laughter, and in some cases convulsive behavior (Milgram, 1963). Those observations matter because they show that behavioral continuation and internal endorsement are different things. A participant can comply while objecting, experiencing distress, attempting to negotiate, or trying to reduce the perceived harm.
The distinction between action and endorsement remains central in modern interpretations. A person who carries out an order is behaviorally obedient at that moment, but the behavior alone does not tell us whether they approve of the order, identify with its purpose, feel trapped by the interaction, accept the authority's legitimacy, or believe that responsibility has shifted elsewhere.
Milgram's Agentic State Theory
Milgram's later theoretical account, developed most fully in his 1974 book Obedience to Authority, proposed a shift from an autonomous state to an “agentic state.” In the autonomous mode, people experience themselves as directing their own actions and responsible for their consequences. In the agentic mode, they experience themselves as executing the wishes of a legitimate authority and locate responsibility for the larger action in that authority.
This is a historical theory, not a settled empirical fact. It was influential because it captured a recognizable feature of institutional obedience: people often describe themselves as “just following orders,” “doing my job,” or carrying out someone else's decision. But those descriptions do not prove that agency literally disappears, and the original Milgram studies did not directly measure a discrete psychological state called the agentic state.
What modern sense-of-agency research supports
Modern experimental work allows a narrower part of the theory to be tested. Caspar and colleagues used the phenomenon of intentional binding—an implicit temporal measure often used as a proxy for sense of agency—and found that coerced actions produced weaker binding than freely chosen actions. Coercion also changed neural processing of action outcomes in their experiments (Caspar et al., 2016).
Subsequent work showed that the psychology is not exhausted by a simple transfer of agency from subordinate to commander. In an experimental study where participants both gave and received commands, command structures altered implicit sense-of-agency measures in both roles, suggesting that responsibility and agency under hierarchical interaction are distributed more complexly than “the subordinate loses agency and the authority gains it” (Caspar, Cleeremans, & Haggard, 2018).
A 2024 study by Akyüz and colleagues used a different action–outcome timing paradigm and found that forced choices were associated with a longer perceived interval between action and result than free choices, which the authors interpreted as an agentic shift under obedience (Akyüz et al., 2024). The study is useful because it reproduces a reduced-agency pattern with another design; it still does not show that the full Milgramian agentic-state theory is established.
Agency is not the same thing as responsibility
Sense of agency is a psychological experience: the feeling that “I caused this.” Causal agency concerns whether a person's behavior actually contributed to an outcome. Moral responsibility concerns whether the person can appropriately be praised or blamed. Professional responsibility concerns duties attached to a role. Legal responsibility depends on legal standards, jurisdiction, intent, coercion, capacity, and other elements. These layers can interact without being identical. For the wider psychology of power as obligation, stewardship, and consequential burden, see Power and Responsibility: Leadership, Moral Obligation, and the Burden of Control.
An order may reduce a person's subjective sense of authorship while leaving their action causally necessary for the outcome. A person may experience strong coercive pressure yet retain some capacity to refuse. Conversely, a person may feel fully agentic while acting within a legitimate chain of command. Psychological evidence about agency can therefore inform responsibility without deciding it by itself.
Responsibility Under Orders
The phrase “I was following orders” is psychologically meaningful because authority can reorganize how responsibility is experienced. It can narrow attention to role performance, make refusal feel deviant, frame the authority as the decision-maker, and make the subordinate feel like an instrument rather than an originator. Experimental work on coercion shows measurable changes in the subjective and neural processing of action (Caspar et al., 2016; Caspar et al., 2021b).
Yet psychological displacement of responsibility is not equivalent to actual transfer of all responsibility. Commanders can bear responsibility for issuing orders; subordinates can bear responsibility for executing them; institutions can bear responsibility for creating incentives, norms, and reporting structures; and bystanders or peers can affect what seems possible. Responsibility can be distributed across a system without becoming nonexistent.
This distinction is essential for interpreting obedience research. Psychology can investigate perceived choice, coercion, agency, conflict, identification, and behavior. It cannot convert a laboratory finding into a universal rule that absolves anyone who acted under authority. Moral and legal responsibility require additional normative and legal analysis.
Beyond the Myth of Blind Obedience
Popular retellings often turn Milgram's participants into symbols of “blind obedience,” as though they stopped thinking and mechanically pressed switches. The archival and interactional record is more complicated. Participants questioned the procedure, worried about the learner, bargained with the experimenter, sought reassurance, interpreted the experimenter's words, and sometimes resisted particular aspects of the procedure while continuing others.
A 2022 review of what has been called a second wave of critical engagement with Milgram concluded that renewed archival and interactional scholarship has complicated the traditional story by focusing on the experiment's social production, affective dynamics, ethical history, and the interpretation of what participants believed was happening (Kaposi, 2022). The result is not that the experiments “showed nothing.” It is that their meaning is richer than a single obedience percentage.
Engaged followership
Stephen Reicher, Alexander Haslam, and John Smith proposed an influential alternative to the idea that obedience results primarily from passive surrender of agency. Their “engaged followership” account argues that people may follow an authority when they identify with the authority's project and see themselves as contributing to a valued collective goal—in Milgram's case, scientific research (Reicher, Haslam, & Smith, 2012). Haslam and Reicher later reviewed the broader obedience literature through this identification-based lens (Haslam & Reicher, 2017).
Engaged followership should be treated as an important contemporary theoretical reinterpretation, not as the final replacement for every other explanation. It highlights identification and shared purpose, while other scholarship emphasizes legitimacy, uncertainty, interpersonal pressure, coercive communication, incremental commitment, and the local structure of the interaction. Different mechanisms may explain different moments or participants.
Interaction and implicit pressure
Recent archival work has examined what the experimenter actually did when participants hesitated. Kaposi, Kacsuk, and Hegarty analyzed 136 sessions across four conditions and found that the phrase “Please continue” was used much more systematically than many other nonprocedural interventions, supporting the view that the experimenter's interventions carried interactional force beyond a simple, neutral repetition of a standardized script (Kaposi, Kacsuk, & Hegarty, 2026b).
Another 2026 archival analysis coded violations of the experimenter's procedural instructions and found that even participants traditionally classified as fully obedient did not necessarily comply with every procedural demand. The authors argue that obedience in the sessions was heterogeneous and cannot be reduced perfectly to whether the maximum switch was eventually reached (Kaposi & Sumeghy, 2026a). This evidence strengthens a process view: obedience can contain negotiation, partial resistance, selective compliance, and conflict.
Legitimacy, Coercion, and Social Influence
Legitimacy and coercion describe different routes through which authority can shape behavior. Legitimate authority encourages deference because people accept the authority's right to make decisions. Coercion changes the cost of refusal through threat, pressure, dependence, or constrained alternatives. Real institutions can involve both at once.
Tyler's review emphasizes that legitimacy is especially valuable to institutions because it supports voluntary cooperation when constant surveillance and punishment are impractical (Tyler, 2006). Milgram's paradigm, by contrast, also shows how direct interpersonal pressure can operate at the moment of hesitation. The experimenter did not simply announce a rule at the start; he responded to participants' attempts to stop and urged them to continue.
The distinction also clarifies why power and authority should not be collapsed. Someone can have power through control of resources without being regarded as legitimately entitled to command. Someone can hold a legitimate office yet have limited practical capacity to enforce an order. Obedience depends partly on how the person receiving the instruction understands this combination.
Situational Factors That Increase or Reduce Obedience
The strongest recurring lesson from the Milgram research tradition is that obedience is sensitive to social configuration. The Meta-Milgram synthesis identified several properties associated with variation in maximum-voltage obedience, including experimenter directiveness, experimenter legitimacy, consistency among experimenters, peer pressure toward disobedience, the indirectness of the participant's relation to the learner, the experimenter's distance, and the intimacy of the teacher–learner relationship (Haslam, Loughnan, & Perry, 2014).
These factors should be interpreted as features of the studied conditions rather than a universal formula. Their effects can depend on how participants understand the task, the institution, the authority, and the person affected by the action. Still, they provide a better evidence base than the popular claim that obedience simply reflects weak character.
Authority proximity and directiveness
A physically or socially present authority can monitor hesitation, answer questions, restate expectations, and make disengagement interactionally difficult. Directiveness clarifies what the authority expects. Distance can weaken that pressure by creating room for independent interpretation or delay. The relevant mechanism is not physical distance alone but what distance changes in monitoring, communication, accountability, and perceived control of the situation.
Victim proximity and psychological distance
When the consequences of an action are more immediate and the affected person is harder to ignore, empathic and moral cues become more salient. Milgram's variations and later syntheses suggest that the structure of the teacher–learner relation matters. This does not mean proximity always produces resistance, but it can change the balance between the authority's demand and information coming from the person who bears the consequence.
Peers and models of dissent
Other people's behavior supplies information about what is possible and legitimate. A peer who refuses can demonstrate that disobedience is an available action, reduce the isolation of dissent, and challenge the impression that the authority's expectation is uncontested. In the Meta-Milgram synthesis, peer pressure toward disobedience was among the features associated with lower obedience (Haslam, Loughnan, & Perry, 2014).
Can Milgram Be Replicated Today?
A literal replication of the original procedure raises major ethical problems because participants were exposed to intense stress and were deceived about the apparent harm being inflicted. Modern studies therefore modify the procedure. This protects participants, but it also means that “replication” usually refers to reproducing a psychologically meaningful portion of the paradigm rather than recreating every original condition.
Burger's 150-volt partial replication
Jerry Burger designed a widely cited partial replication that stopped the procedure at 150 volts, the point at which the learner first strongly demanded release in Milgram's script. Participants who continued beyond that point in the original studies were highly likely to continue further, so Burger used it as an ethically defensible stopping point. The study found substantial continued compliance and concluded that obedience pressures comparable to those observed by Milgram could still be demonstrated under the modified procedure (Burger, 2009).
Because the study stopped at 150 volts, it cannot answer how many contemporary participants would have reached 450 volts. Its value is narrower: strong authority-driven continuation can still be observed in a modern, ethically constrained paradigm.
Modern transparent and real-consequence designs
Caspar developed a modern experimental approach in which participants could choose whether to administer real but mild and ethically controlled shocks, allowing researchers to study obedience and disobedience without reproducing Milgram's extreme deception and distress (Caspar, 2021a). Such designs move the field toward experimentally tractable mechanisms—choice, coercion, agency, responsibility, and refusal—rather than treating the original procedure as the only possible model.
A 2025 Scientific Reports study also tested whether institutional prestige and the stated scientific versus commercial purpose of a modified Milgram-style procedure changed obedience. Fifty-eight participants remained in the final analysis across three conditions. The study found no significant condition difference in reaching its key obedience threshold and no significant gender difference in that small sample (Grzyb et al., 2025). The finding challenges a simple explanation based on prestigious university setting or scientific justification alone, but its modest sample, single-country context, and ethical modifications limit generalization.
Individual Differences: Is There an Obedient Personality?
There is no well-supported single personality profile that can identify who will obey an authority across situations. The dramatic situational variation in the Milgram program itself is evidence against a purely dispositional explanation. Personality can still matter, but it interacts with roles, institutions, beliefs, relationships, and immediate social cues.
One study using a televised game-show adaptation of the Milgram paradigm found that Agreeableness and Conscientiousness were associated with higher shock levels among 66 participants (Bègue et al., 2015). The result is interesting but far too narrow to justify an “obedient personality” label. A single small study in one unusual setting cannot establish a stable trait profile for obedience.
Authoritarianism is also a neighboring construct rather than a synonym for obedience. A person can score high or low on authoritarian attitudes without behaving predictably in every authority situation. The historical “authoritarian personality” tradition and contemporary authoritarianism research ask different questions from the behavioral question of whether a person follows a particular order.
Culture, Gender, and the Limits of Universal Claims
Obedience research has been conducted in different countries and institutional settings. A cross-cultural review of Milgram-paradigm studies found substantial evidence that high obedience could appear outside the original U.S. context, but differences in procedures, decades, samples, institutions, and definitions make precise cross-national ranking difficult (Blass, 2012). The defensible conclusion is that obedience to authority is not unique to one culture, while its level and meaning remain context-sensitive.
Gender findings are similarly unsuitable for a simple universal claim. Some studies report little or no difference, while individual samples are often too small to estimate modest effects reliably. The 2025 modified-paradigm study found no significant gender difference, but with only 58 participants in the final sample it does not settle the question (Grzyb et al., 2025).
Culture and gender should therefore be treated as potential moderators embedded in social roles, institutions, norms, and opportunities—not as deterministic explanations. The research base does not support formulas such as “people from culture X are obedient” or “gender Y obeys more.”
Obedience in Everyday Institutions
Obedience is not inherently pathological or harmful. Coordinated action often requires people to accept legitimate instructions: an air-traffic controller follows safety protocols, a surgical team responds to role-specific commands, a firefighter follows operational direction, and an employee complies with lawful procedures. Social life would become costly and dangerous if every instruction had to be renegotiated from first principles.
The psychological problem begins when the authority relation suppresses information that should change the action. Expertise can be mistaken for infallibility. Role hierarchy can make a subordinate hesitate to report an error. Responsibility can become diffused across a chain of command. Gradual commitment can make an escalating procedure feel harder to interrupt. Loyalty to an institution can compete with duties to patients, clients, colleagues, the public, or one's own ethical standards.
A classic field study of nurse–physician relations illustrated how role authority could produce compliance with a questionable medication order: most nurses in the experimental situation began carrying out a telephone instruction that violated several hospital rules before an observer intervened (Hofling et al., 1966). The study is historically important, but it reflects a specific mid-20th-century hospital context and should not be treated as a direct estimate of contemporary clinical behavior.
When Does Obedience Become Psychologically Risky?
Obedience becomes more concerning when several conditions converge: the order produces or risks serious harm; the authority discourages questions; the role's legitimate scope is unclear; responsibility is pushed downward while decision-making is concentrated upward; independent information is blocked; dissent carries disproportionate social or material cost; peers appear uniformly compliant; or the action escalates gradually so that each step seems only slightly different from the last.
These conditions do not prove that a specific order is wrong. They are warning signs that ordinary social processes may be making independent evaluation harder. The relevant question is not “Am I an obedient person?” but “What features of this authority relationship are shaping what feels possible, normal, responsible, and legitimate right now?”
How to Evaluate an Authority Demand
A psychologically informed evaluation begins by separating the authority's role from the content of the instruction. Ask what gives this person or institution the right to direct this action, whether the instruction falls within that role, what evidence supports it, who bears the consequences, and which rules govern refusal, consultation, or escalation. Legitimate authority is strongest when its scope and procedures are clear rather than when questioning is treated as disloyalty.
It is also useful to separate social discomfort from actual impossibility. In many authority situations, refusing feels awkward, disrespectful, or disruptive long before it becomes materially impossible. Naming that distinction can restore deliberation. Seeking a second opinion, consulting a written protocol, asking the authority to state the instruction clearly, or involving another responsible person can change the social configuration without requiring immediate confrontation.
For high-stakes professional settings, the relevant standard is the field's own law, ethics, safety procedure, and escalation pathway. Psychological research can explain why authority pressure is powerful; it is not a substitute for the rules governing a medical, legal, military, aviation, workplace, or emergency decision.
What Psychology Can Say With Confidence
Established empirical evidence supports the conclusion that authority can exert strong effects on behavior and that obedience varies substantially with social context. Milgram's original work and later syntheses show that proximity, legitimacy, directiveness, peer behavior, and relationship structure can matter. Modern ethically modified studies continue to demonstrate authority-driven compliance, although they do not recreate the entire original procedure.
Milgram's agentic state is a historical theory. Contemporary coercion experiments support the more limited empirical conclusion that acting under orders can reduce implicit measures of sense of agency or alter action–outcome processing (Caspar et al., 2016; Akyüz et al., 2024). They do not demonstrate that people lose agency altogether or that responsibility automatically transfers to the authority.
Engaged followership is a major contemporary interpretation emphasizing identification with the authority's project (Reicher, Haslam, & Smith, 2012). Interactional and archival critiques emphasize the experimenter's interventions, participants' beliefs, negotiation, and partial resistance (Kaposi, 2022; Kaposi, Kacsuk, & Hegarty, 2026b). These accounts refine the field rather than yielding one universally accepted replacement mechanism.
Popular simplifications should therefore be rejected. Psychology has not established that “65 percent of people will obey any authority,” that obedience is equivalent to conformity, that following orders eliminates responsibility, or that an “obedient personality” explains the phenomenon. The strongest contemporary account is plural and situational: authority works through relationships, institutional legitimacy, pressure, identification, uncertainty, agency, and the structure of available alternatives.
Obedience to Authority in the Psychology of Power
Obedience reveals one of the central problems in the psychology of power: how another person's capacity to direct outcomes becomes behavior inside the person who receives the directive. The process can involve accepted legitimacy, dependence, information, threat, duty, identification, habit, and role expectations. Authority is the social claim that a directive ought to be followed; obedience is one possible behavioral response to that claim.
This is why obedience should remain conceptually separate from status, dominance, prestige, leadership, conformity, and personal control. Those constructs can shape an authority relationship, but none is identical to following an order. The psychological value of the obedience literature lies precisely in showing how a structured relationship can transform the meaning of an otherwise ordinary choice.
FAQ
What is obedience to authority in psychology?
Obedience to authority is behavior performed in response to an explicit command, instruction, or demand from a person or institution perceived as entitled to issue it. The defining feature is the authority relationship, not merely the fact that behavior changed.
What did Milgram's obedience experiment show?
Milgram showed that substantial numbers of ordinary adults could continue administering what they believed were increasingly severe shocks when urged to continue by an experimental authority. In the 1963 baseline condition, 26 of 40 participants—65 percent—reached the maximum switch (Milgram, 1963). Across different conditions, however, obedience varied greatly.
Does the Milgram study mean 65 percent of people always obey authority?
No. Sixty-five percent was the result of one baseline condition. A synthesis of 21 Milgram conditions found an overall maximum-voltage rate of 43.6 percent and substantial variation associated with the structure of the situation (Haslam, Loughnan, & Perry, 2014).
What is the agentic state?
The agentic state is Milgram's historical theory that a person under legitimate authority may experience themselves as an agent carrying out another person's wishes and shift responsibility upward. Modern research supports a narrower link between coercion and reduced sense-of-agency measures, but it has not established a literal all-or-none agentic state.
Does obeying an order remove responsibility?
Psychologically, an order can reduce perceived agency or shift where a person feels responsibility lies. That does not automatically remove causal, moral, professional, or legal responsibility. Those are distinct questions governed by additional ethical and legal standards.
What is the difference between obedience and conformity?
Obedience concerns a directive from authority. Conformity concerns adjustment toward group norms, expectations, or behavior. Both are forms of social influence, but the source and structure of pressure differ.
What is the difference between obedience and compliance?
Compliance is broader: it means responding to a request or demand. Obedience is typically treated as an authority-specific form of compliance in which the person issuing the directive is perceived as entitled to do so.
Why do people obey harmful orders?
Research points to interacting mechanisms rather than one cause: legitimacy, role expectations, uncertainty, direct pressure, gradual commitment, identification with a collective purpose, distance from consequences, peer behavior, and constrained alternatives can all matter. Their importance varies by situation.
Can people resist authority?
Yes. Resistance occurred throughout Milgram's research, and situational changes could make it more common. Peers who model dissent, reduced authority pressure, closer contact with consequences, contested legitimacy, and clearer alternatives can all alter the probability of refusal. Resistance is a separate psychological process deserving its own analysis rather than the mere absence of obedience.
Is obedience the same as authoritarian personality?
No. Obedience is behavior in a specific authority situation. Authoritarian personality is a historical personality and social-psychological tradition, while contemporary authoritarianism research studies broader attitudes and dispositions. Neither construct determines behavior in every authority encounter.
Have modern studies replicated Milgram?
Modern studies have replicated important parts of the phenomenon using ethically modified procedures. Burger's 2009 study stopped at 150 volts; newer designs use real but mild consequences or modified authority settings. These studies show that authority-driven compliance remains observable, while ethical changes prevent an exact recreation of Milgram's entire original procedure.
Related Articles
References
Akyüz, N., Marien, H., Stok, F. M., Driessen, J. M. A., de Wit, J., & Aarts, H. (2024). Obedience induces agentic shifts by increasing the perceived time between own action and results. Scientific Reports, 14, 16769. https://doi.org/10.1038/s41598-024-66499-8
Bègue, L., Beauvois, J.-L., Courbet, D., Oberlé, D., Lepage, J., & Duke, A. A. (2015). Personality predicts obedience in a Milgram paradigm. Journal of Personality, 83(3), 299–306. https://doi.org/10.1111/jopy.12104
Blass, T. (2012). A cross-cultural comparison of studies of obedience using the Milgram paradigm: A review. Social and Personality Psychology Compass, 6(2), 196–205. https://doi.org/10.1111/j.1751-9004.2011.00417.x
Burger, J. M. (2009). Replicating Milgram: Would people still obey today? American Psychologist, 64(1), 1–11. https://doi.org/10.1037/a0010932
Caspar, E. A., Christensen, J. F., Cleeremans, A., & Haggard, P. (2016). Coercion changes the sense of agency in the human brain. Current Biology, 26(5), 585–592. https://doi.org/10.1016/j.cub.2015.12.067
Caspar, E. A., Cleeremans, A., & Haggard, P. (2018). Only giving orders? An experimental study of the sense of agency when giving or receiving commands. PLOS ONE, 13(9), e0204027. https://doi.org/10.1371/journal.pone.0204027
Caspar, E. A. (2021a). A novel experimental approach to study disobedience to authority. Scientific Reports, 11, 22927. https://doi.org/10.1038/s41598-021-02334-8
Caspar, E. A., Beyer, F., Cleeremans, A., & Haggard, P. (2021b). The obedient mind and the volitional brain: A neural basis for preserved sense of agency and sense of responsibility under coercion. PLOS ONE, 16(10), e0258884. https://doi.org/10.1371/journal.pone.0258884
Dolinski, D., & Grzyb, T. (2025). Obedience. In Research Handbook on Social Influence. Edward Elgar Publishing. https://doi.org/10.4337/9781035309672.00021
Grzyb, T., Dolinski, D., Sudoł-Malisz, M., Kulik, G., & Mielczarek, Ł. (2025). Doubting the power of prestige: Obedience to authority beyond institutional and research justifications. Scientific Reports, 15, 24911. https://doi.org/10.1038/s41598-025-10331-4
Haslam, N., Loughnan, S., & Perry, G. (2014). Meta-Milgram: An empirical synthesis of the obedience experiments. PLOS ONE, 9(4), e93927. https://doi.org/10.1371/journal.pone.0093927
Haslam, S. A., & Reicher, S. D. (2017). 50 years of “obedience to authority”: From blind conformity to engaged followership. Annual Review of Law and Social Science, 13, 59–78. https://doi.org/10.1146/annurev-lawsocsci-110316-113710
Hofling, C. K., Brotzman, E., Dalrymple, S., Graves, N., & Pierce, C. M. (1966). An experimental study in nurse-physician relationships. Journal of Nervous and Mental Disease, 143(2), 171–180. https://doi.org/10.1097/00005053-196608000-00008
Kaposi, D. (2022). The second wave of critical engagement with Stanley Milgram's “obedience to authority” experiments: What did we learn? Social and Personality Psychology Compass, 16(6), e12667. https://doi.org/10.1111/spc3.12667
Kaposi, D., & Sumeghy, D. (2026a). From legitimate to illegitimate violence: Violations of the experimenter's instructions in Stanley Milgram's obedience experiments. Political Psychology, 47(2), e70112. https://doi.org/10.1111/pops.70112
Kaposi, D., Kacsuk, Z., & Hegarty, P. (2026b). “Please continue”: Implicit communication and the experimenter's interventions in Stanley Milgram's obedience to authority series. European Journal of Social Psychology, 56(1), 292–306. https://doi.org/10.1002/ejsp.70033
Milgram, S. (1963). Behavioral study of obedience. Journal of Abnormal and Social Psychology, 67(4), 371–378. https://doi.org/10.1037/h0040525
Milgram, S. (1974). Obedience to Authority: An Experimental View. Harper & Row. https://search.worldcat.org/title/Obedience-to-authority-%3A-an-experimental-view/oclc/804851564
Reicher, S. D., Haslam, S. A., & Smith, J. R. (2012). Working toward the experimenter: Reconceptualizing obedience within the Milgram paradigm as identification-based followership. Perspectives on Psychological Science, 7(4), 315–324. https://doi.org/10.1177/1745691612448482
Tyler, T. R. (2006). Psychological perspectives on legitimacy and legitimation. Annual Review of Psychology, 57, 375–400. https://doi.org/10.1146/annurev.psych.57.102904.190038