The Milgram Obedience Experiment: Why Ordinary People Followed Harmful Orders

Why Milgram ran the study

The Milgram experiment investigated obedience when an authority instructed people to carry out an apparently harmful task. Participants acted as teachers in a supposed learning study and were directed to increase apparent shocks after wrong answers. The research measured how far they continued, alongside their protests and visible distress.

Milgram framed the work around the problem of destructive obedience. A laboratory task could expose pressures that influence behavior, although it could not reproduce the history, ideology, institutions, or consequences involved in real atrocities. Comparisons with those events require care.

The central question remains relevant when people face conflicting responsibilities. A person might trust an institution's expertise while also feeling concerned about someone affected by its instructions. The experiment created a highly controlled conflict of this kind, using deception.

How the experiment worked

Participants were recruited for a supposed study of memory and learning. A rigged role assignment made the participant the teacher and an accomplice the learner. The teacher was instructed to increase the apparent shock level after each incorrect response.

The learner received no shocks. Participants did receive a real sample shock to make the apparatus convincing. Milgram's original report describes a generator with labeled settings rising to 450 volts. These settings represented apparent punishment in the task, rather than actual shocks delivered to the learner.

The procedure varied across Milgram's studies. In the condition reported in the original paper, the learner pounded on the wall at a high setting and later stopped responding. Accounts involving spoken protests or heart complaints concern other variations and should not be merged into a single script.

When a participant hesitated, the experimenter used prompts intended to keep the task going. The instructions and the apparent authority of the setting created pressure to continue, even when the participant expressed concern.

What Milgram found

In Behavioral Study of Obedience, published in 1963, Milgram reported that 26 of 40 male participants reached the maximum 450-volt setting. That is 65% of this particular sample. The remaining participants stopped before the maximum.

The study also recorded sweating, trembling, nervous laughter, and verbal objections. Continuing with the task could coexist with discomfort and protest. Observed compliance alone does not establish enjoyment of suffering or a stable personality trait.

The sample and procedure matter. A percentage from a particular laboratory condition cannot estimate how humanity would behave in every conflict with authority. Milgram's broader program varied the circumstances precisely because the setting could affect participants' decisions.

The variations that revealed what mattered

Milgram changed features such as the authority's proximity, the learner's proximity, the institution associated with the study, and the presence of other people who resisted. These variations are discussed in his book Obedience to Authority. They make the setting part of the explanation.

A person who sees a peer object has a different social example from someone facing the experimenter alone. Similarly, distance from the learner changes the immediate cues available. A comparison across conditions must specify which features changed before interpreting differences in obedience.

Reporting only the best-known result hides this variation. When evaluating a summary, check whether it identifies the condition, recruitment, endpoint, and procedure. Those details determine what the observed behavior can reasonably tell us.

How Milgram explained it

Milgram's agentic-state explanation proposed that people may see themselves as carrying out an authority's wishes and shift responsibility toward that authority. This was an interpretation of the behavior, rather than a directly observed switch inside a participant's mind.

Reicher, Haslam, and Smith proposed an account emphasizing identification with the experimenter and the scientific project. In this view, believing that the work served a valued purpose could help explain cooperation. Their article Working Toward the Experimenter develops this alternative.

The debate matters because explanations imply different questions. Was a participant deferring responsibility, trusting the institution, identifying with science, responding to pressure, or combining those processes? Similar outward behavior can have different motivations.

Ethics and lasting controversy

The study used deception and exposed participants to substantial distress. Prompts to continue also complicated their practical freedom to withdraw. These features raise ethical questions about consent, risk, and whether debriefing can adequately address what a participant experienced.

Modern evaluation would consider safeguards before recruitment, an effective withdrawal process, monitoring of distress, and follow-up. A study's scientific value must be considered alongside the burden imposed on participants. The absence of real shocks to the learner does not erase distress experienced by the teacher.

Jerry Burger's partial replication, published in 2009, stopped at the 150-volt decision point and used additional safeguards. Its design provides evidence about that earlier point in the procedure. It cannot directly establish how participants would have behaved at Milgram's maximum setting.

Criticisms of the Milgram experiment

Milgram's results have also drawn criticism aimed at the method itself, separate from the ethical concerns already described. The psychologist and historian Gina Perry examined Milgram's original archives and argued that some experimenters departed from the standardized script described in his published reports, raising questions about how uniform the procedure really was across sessions.

A separate line of criticism concerns whether participants believed the shocks were genuine. Perry and colleagues reanalyzed an unpublished measure from Milgram's own research and found that a meaningful share of participants expressed doubt about the setup, which complicates any simple reading of the obedience rate as evidence of how people behave when they fully believe they are harming someone.

These criticisms do not overturn the basic finding that many participants continued a distressing task under an experimenter's direction. They do mean that a single obedience percentage should be read alongside its procedural details and the uncertainty about what each participant believed was actually happening.

Applying the questions without labeling people

The psychology concepts hub includes fundamental attribution error, which concerns how we explain other people's actions, and social identity theory, which addresses group membership. Both offer useful questions when considering obedience, alongside their own limits.

If you struggle to express disagreement, the people-pleasing self-check can prompt reflection. It does not measure how you would behave in Milgram's experiment. The assertive communication worksheet can help you practice asking for clarification or naming a concern.

For example, you might ask a supervisor to explain the reason for a request, state what worries you, or seek independent advice. The safest response depends on the setting and possible consequences. A licensed therapist can help when fear of disagreement or pressure in relationships is affecting your well-being.