Begue et al (2014)

Historically, it was believed by the likes of William Shirer that a personality flaw made the Germans overly obedient to malevolent authority resulting in the heinous crimes committed during the Holocaust. This was investigated by Adorno et al (1950) who created the F scale as a measure of receptiveness to the ideals of fascism and he created a psychodynamic theory explaining that harsh and punitive child rearing was to blame for the problems in the German psyche. However, the philosopher Hannah Arendt argued that Adolf Eichmann was far from a sadistic, monster and was in fact very ordinary. She uses the phrase the banality of evil to describe what she saw as she observed him at his trial.

Milgram (1974) however, and later Burger (2009), demonstrated the power of the social situation in overriding personality variables. This said, Milgram himself revealed that those who administered the highest shocks did indeed score highest on the F scale. In recent years the pendulum has swung back to the investigation of individual differences that account for differences in obedience and dissent.

Aim: To investigate which personality traits best predict levels of obedience in a ‘Milgramesque’ situation. Milgram would no doubt have been pleased by this turn of events given that he himself said “I am certain that there is a complex personality basis to obedience and disobedience, but I know we have not found it” (Milgram, 1974).

Begue et al (2014) felt that socially desirable traits such as Conscientiousness and Agreeableness may in fact contribute to destructive obedience as they predict normative behaviour and should facilitate submissive behaviour toward authority.

The second aim of the study was to look at how political opinions relate to obedience to authority. It was predicted that people who reported more left wing political attitudes would be more disobedient and people inclined towards past rebellious and unruly behaviour, or a readiness to perform such behaviours, would be more disobedient to authority in a Milgram paradigm.

Procedure: Participants were 35 males and 31 females aged 26–54 from the general population of Paris who were contacted by phone 8 months after their participation in a study transposing Milgram’s obedience paradigm to the context of a television game show, (Beauvois et al 2012). The original group of 80 were selected for the programme by an independent company from a consumer database. They received 40 Euros for their participation. Four participants were excluded due to familiarity with Milgram’s research. 76 were left (40 males and 36 females, aged 25 and 55 years old (M = 39.7 years, SD = 8.51). Participants who had already participated in a game show were not eligible, nor were those with health conditions or those taking any kind of medication.

The experiment took place in a TV studio, and the authority of the television world was represented by a host. An original game show was set up onstage with the help of technical devices (i.e., cameras, lighting, giant screen, control room) and human resources. When participants arrived at the television studio, an alleged producer greeted each participant, along with another person who was a confederate. The producer told them that they would be filmed as they participated together as players on a TV game show. Because the filming was said to be for a pilot show aimed at testing the game “under real conditions” and improving it if need be, they were informed that they would not win any money, unlike the future game contestants, who would try together to win one million Euros. For one of the players (“the questioner”), the task consisted of asking questions; for the other (“the contestant”), the task was to answer correctly. They were told that the penalty for each incorrect answer would be an electric shock delivered by the “questioner” to the “contestant.”

The alleged producer then had the players draw straws to determine which person would play which role. The drawing was rigged so that the participant was always the questioner and the accomplice was always the contestant. The two players were then led onstage, where they were awaited by the game host (a female weather forecaster for a French national TV station). Then, in front of the camera, the host explained the game as follows. The contestant would be given a limited amount of time to memorize a list of 27 pairs of related words (e.g., cloudy-sky, tame-animal). Then the questioner would say the first word in each pair and the contestant would have to select the related word among the four words. If the contestant made a mistake, the questioner was to deliver an electric shock, increasing the shock’s intensity each time. The team of players would win if they were able to get through all 27 questions (whether the answers were correct or incorrect). Once these rules were stated, the contestant was taken into a chamber where he would not be seen by the questioner or the audience. However, in front of everyone, before the door of the chamber was closed, the contestant was strapped to the chair in which he would receive the shocks and was shown the buttons to press to choose the right answer. The questioner was seated at a desk in the centre of the stage, under the projectors and cameras, with his or her back to the audience and facing a giant screen where the questions would be displayed in succession. The questioner was shown the handles to be used to deliver the (fake) electric shocks to the contestant, with an increase of 20 volts for each new wrong answer.

The shocks ranged from 20 volts to 460 volts and were grouped together in sets labelled as follows: “slight shocks” (20, 40, 60); “moderate shocks” (80, 100, 120); “strong shocks” (140, 160, 180); “very strong shocks” (200, 220, 240); “intense shocks” (260, 280, 300); “extremely intense shocks” (320, 340, 360); “danger: severe shocks” (380, 400, 420); and “XXX” (440, 460).

According to a predetermined schedule, the contestant responded incorrectly 24 times out of 27. If the questioner began to hesitate or refused to go on, the host intervened by giving at most five prods to urge the players to continue. The first four prods were worded in a Milgram-like fashion (“go on with the questions”; “go on, don’t let yourself get upset”; “the rules say you must go on”; “go on, we are taking all responsibility for this”). Only the fifth prod, “you can’t make him lose; what will the audience think?” was specific to the television situation. As in Milgram’s paradigm, starting from when the shock reached 80 volts, the contestant’s pre-recorded reactions were heard: grunts at first, then loud cries of pain accompanied by refusal to continue, then screams and pleas to stop the game. After 380 volts, the contestant refused to answer. After 420 volts, nothing more could be heard. The game ended either when the questioner had asked all 27 questions (which involved two shocks of 460 volts) or when, in spite of the host’s five prods, the questioner decided to stop.

Participants were assigned to one of four conditions. In the standard condition (n = 32), which was similar to Milgram’s “voice-feedback” condition, a “questioner” (the participant) had to ask 27 questions to a “contestant” (an accomplice of the experimenter) who could be heard but not seen. Every time the contestant gave an incorrect answer, the questioner was to punish him by delivering an (alleged) electric shock. The shocks ranged between 20 and 460 volts and were to be increased by 20 volts with each new mistake.

The social-support condition (n = 19) differed in one aspect from the standard condition: When the voltage reached 120, the production assistant (an accomplice) rushed out onstage and asked that the game be stopped because it was too immoral. The assistant was brushed aside by the host, who went on with the game.

The TV-broadcast condition (n = 18) was similar to the standard condition, except that upon arrival, the questioner and alleged contestant were informed that the TV station would broadcast the pilot show. The players would be on TV but would still not win any money.

Finally, in the host-withdrawal condition (n = 7), the situation was similar to Milgram’s condition in which the researcher leaves the experiment (like Experiment 7). Upon reaching 80 volts, the host explained that from now on, the players would continue on their own. Then the host went offstage and did not come back until the game was over. There were then no more prods after the host left the stage.

Findings of original study (Beauvois et al.2012): Results showed that in the standard condition, there was 81% obedience, whereas in the host-withdrawal condition, there was 28% obedience. Those conditions were not significantly different from Milgram’s conditions. There was 74% obedience in the social-support condition and 72% in the TV-broadcast condition. Only the standard condition and the host-withdrawal condition differed significantly.

8 months later, the original PPs were called and asked to take part in a survey for 20 Euros. They did not know it was related to the game show. The response rate was 89%, leaving a total sample of 35 males and 31 females aged 26–54 years (M = 39.66, SD = 8.51). Participants did not differ on obedience when compared to the whole sample.

The Big Five Mini-Markers questionnaire (Saucier, 1994) is a 40-item adjective checklist and as such provides an abbreviated, yet valid and reliable version of 100 trait-descriptive adjectives of the Big Five personality domains.

Political ideology was measured with standard items from the World Value Survey Questionnaire (Inglehart & Welzel, 2005): NB: Please click here to see this questionnaire: wv6_official_questionnaire_v4_june2012-1

Political orientation was based on the following single item measure: “In political matters, people talk of ‘the left’ and ‘the right.’ How would you place your views on this scale, generally speaking? (1 = extreme left, 10= extreme right).”

The Political Activism Scale (PAS) was composed of four behaviours, and for each behaviour, participants had to select between “have done” (coded 3), “might do” (coded 2), or “would never do” (coded 1). The PAS included the following behaviours: signing a petition, attending lawful demonstration, joining unofficial strikes, and occupying buildings or factories.

The measure of obedience was the intensity of shocks delivered by individuals. Intensity of shocks ranged from 100 to 460 volts and exhibited a significant negative skew similar to Milgram’s original studies. Therefore, we utilized nonparametric (i.e., Spearman’s rank correlation coefficients) statistics.


  • The intensity of shocks administered did not significantly differ between male and female Pps.
  • There also was no relationship between obedience and age, nor between obedience and the familiarity of individuals with TV game shows.
  • Conscientiousness and Agreeableness were significantly (p = .039) associated with willingness to administer higher-intensity electric shocks to a victim.
  • Political orientation was also significantly related to obedience; the more the participants defined themselves as on the “left” of the political spectrum, the lower the intensity of shocks they agreed to give to the contestant.
  • Rebellious political activism and obedience were also significantly correlated, although this was not only marginally significant, (p = .10). Analysis by gender showed that the relationship was significant for females, (p = .03), but not for males, (p = .95)

Conclusions: The results provide empirical evidence suggesting that individual differences in personality and political variables matter in the explanation of obedience to authority. As expected, Conscientiousness and Agreeableness predicted the intensity of electric shocks administered to the victim. Second, disobedience was influenced by political orientation, with left-wing political ideology being associated with decreased obedience. Third, we showed that women who were willing to participate in rebellious political activities such as going on strike or occupying a factory administered lower shocks. The results suggest that situational context, even though a powerful determinate of behaviour, does not necessarily overwhelm individual-level behavioural determinants. It is interesting to note that personality traits such as Agreeableness and Conscientiousness, which are widely related to positive outcomes such as better mental health, longevity, academic performance, parenting, reduced aggression, and prosocial behaviour, may also have darker sides in that they can lead to destructive and immoral obedience.  “Evil” behaviour such as destructive obedience may indeed be “banal” in the sense of not relying on extraordinary cruelty or ideological hate, but it also may also be facilitated by dispositions that are consensually desirable elsewhere with family and friends.  Although our results suggest that adaptive traits in the interpersonal domain may be maladaptive in a context involving destructive authority, they also suggest that some behaviours that may disrupt social functioning, such as political activism, may express and even strengthen individual dispositions that are both useful and essential to the whole society, at least in some critical moments.