New analysis suggests most Milgram participants realised the “obedience experiments” were not really dangerous

By Christian Jarrett

Stanley Milgram’s experiments in the 1960s – in which ordinary volunteers followed a scientist’s instruction to give what they apparently thought was a deadly electric shock to another participant – have been taken by many to show our alarming propensity for blind obedience. Milgram’s own interpretation, his “agentic state theory”, was that we readily give up our own sense of responsibility when following instructions from an authority figure. However, his “obedience studies” have come in for recent criticism and re-interpretation (not that you’d know this from the textbooks). The most prominent contemporary theory is that the studies don’t demonstrate blind obedience at all, but rather “engaged followership” – people’s willingness to do bad things when they see them as morally good because they serve a grander cause, in this case science.

Now Matthew Hollander at the University of Wisconsin, and Jason Turowetz at the University of Siegen, have conducted the first in-depth analysis of the interviews that many of the participants gave immediately after taking part in the now infamous research. The new findings, published in the British Journal of Social Psychology, provide little evidence for engaged followership. Instead most of Milgram’s participants showed scepticism that anyone had been seriously harmed at all.

Little studied before, the secretly recorded interviews were conducted by the actor who played the role of experimenter. Hollander and Turowetz listened to 91 of these interviews featuring 46 “obedient” participants who’d applied the most extreme level of shock, and 45 who defied the experimenter and at some stage refused to continue. Focusing for now on the obedient participants, Hollander and Turowetz say that they offered four main reasons for why they continued to the end of the experiment.

Just under 60 per cent of these participants said at least once that they had been following instructions, which provides some support for Milgram’s agentic theory. Around 10 per cent said at least once that they had been fulfilling a contract: “I come here, and yer paying me the money for my time“. The most common explanation was that they believed the person they’d given the electric shocks to (the “learner”) hadn’t really been harmed. Seventy-two per cent of obedient participants made this kind of claim at least once, such as “If it was that serious you woulda stopped me” and “I just figured that somebody had let him out“.

If there is merit to the engaged followership theory, then the most common, or at least most salient, explanation from the “obedient” participants ought to be about the virtues of the science. Yet fewer than 25 per cent of them referred to the importance of the experiment (e.g. “I thought the experiment, uh, depended on my going ahead“) even though Hollander and Turowetz interpreted this category as broadly as they could. Only six participants (out of all 91 interviews) even mentioned the words “science” or “scientific” at all. While some participants identified with the experiment, “the majority do not appear to have continued out of commitment to science,” Hollander and Turowetz said.

The new results contrast sharply with evidence adduced by proponents of the engaged followership theory. For instance, for a 2015 paper, Alex Haslam, Steve Reicher, and others, analysed responses to a survey that Milgram mailed out to his participants in the weeks and months after they took part. The most salient feature of the participants’ feedback was reportedly their happiness and sense of privilege at having taking part in important science. However, there is a key difference from the new findings. These surveys were completed a good time after the participants had been debriefed about the purpose of the research. In other words, they were delayed, post-hoc justifications, coloured by the debrief. In contrast, the newly analysed interviews provide the participants’ immediate, personal take on why they behaved the way they did.

“Certainly, all the evidence suggests that Milgram’s participants were ‘happy to have been of service’,” wrote Haslam et al in 2015. “To the contrary,” conclude Hollander and Turowetz, “an essential and hitherto neglected body of evidence – the interviews – suggests that the puzzle of Milgramesque behaviour is more complex than engaged followership at present allows.”

Normalizing trust: Participants’ immediately post-hoc explanations of behaviour in Milgram’s ‘obedience’ experiments

Christian Jarrett (@Psych_Writer) is Editor of BPS Research Digest

20 thoughts on “New analysis suggests most Milgram participants realised the “obedience experiments” were not really dangerous”

  1. While I haven’t read the original paper, the description provided above suggests that this was a very facile analysis; the same demand characteristics at play that led the subjects to be obedient were at play in post-hoc questioning. The authors’ explanation that this in some way repudiates Milgram’s analysis is not compelling, at least not on the basis of this article.

    1. How could there be demand characteristics? The researchers didn’t collect the data but analysed Milgram data

    2. Agree, though would need to read whole paper. First, I wonder if the average non-scientist participant would even think to mention the ‘science’ aspect of participation, so the fact that most didn’t seems relatively unimportant. Second, if they really did think they had harmed someone, one way of excusing, or justifying, their behaviour when faced with the experimenter in debrief would be to say they hadn’t really believed the learner was being harmed. Whatever, this will make a useful resource for getting my students to think about issues of qualitative analysis in research.

      1. My concern with this study is that it treats the comments from participants in the interviews as objective reasons, whereas they are being interviewed about why they apparently repeatedly applied electric shocks to someone. Trapped between the motivation to comply (due to blind obedience or engaged followership) and the apparent situation they were hurting someone, they trusted that the ‘experimenter’ had things under control and suspended their own responsibility.

      2. What about cognitive dissonance as an explanation for those who said that they knew all along that nobody was harmed?

Comments are closed.