Tuesday, December 20, 2011

The Compliant Juror: Part III

In Part I of this series, I described an experiment in which a subject was willing to apply lethal electric shocks to a fellow human simply because the subject was instructed to do so by the person conducting the experiment.

In Part II of this series, I explained that the subject's response to authority was not unique, that two-thirds of the subjects in that particular experiment refused to disobey the authority figure even when instructed to apply what the subjects believed to be potentially deadly shocks.

I challenged to you, the readers, to decide whether you would have disobeyed the authority figure. I suggested that our natural reluctance to defy authority might explain the behavior of compliant juries. I promised to buttress that suggestion in a separate post. I'll attempt to do so now.

Stanley Milgram conducted 18 variations of the experiment I described. Many of the variants might be instructive to anyone attempting to understand jury behavior. I'll limit myself at the moment to Experiment 18. In that experiment, an additional confederate was introduced. I call him the Executioner, since he was the person who executed the sentence. He was the one who would apply the shock should the Learner fail to properly answer the question posed by the actual subject, the Teacher. The Teacher now had only to read the questions, announce whether the Learner responded correctly or incorrectly, and announce the shock level that was to be applied if the answer was incorrect.

The Teacher was the only subject in the experiment. The Experimenter, the Learner, and now the Executioner all understood that the Learner was not being shocked, that the experiment had to do with response to authority rather than the effect of negative feedback on learning.

Clearly, a juror's role is more similar to the subject in Experiment 18 than to the subject in Experiment 5, discussed previously.
The juror will not escort the convicted defendant from the courtroom. 
The juror will not impose the sentence. Except in capital murder cases, jurors usually are not allowed to know beforehand what the possible or likely sentence would be. Sentencing normally takes place after the jurors are dismissed. The jurors frequently never know the consequence of their decision. 
In capital murder cases, the jurors will not administer the lethal injection.
Experiment 18 was designed to study the impact of separating the subject from the final punishment. Here's how Milgram described that variation.
To examine this phenomenon within the laboratory, a variation was carried out in which the act of shocking the victim was removed from the naive subject and placed in the hand of another participant (a confederate). The naive subject performs the subsidiary acts which, through contributing to the over-all progress of the experiment, remove him from the actual act of depressing the lever on the shock generator. ... 
Any competent manager of a destructive bureaucratic system can arrange his personnel so that only the most callous and obtuse are directly involved in the violence. The greater part of the personnel can consist of men and women who, by virtue of their distance from the actual acts of brutality, will feel doubly absolved from responsibility. First, legitimate authority has given full warrant for their actions. Second, they have not themselves committed brutal physical acts.
For comparison, recall that 26 of 40 subjects in Experiment 5 refused to defy authority; 26 of 40 subjects actually applied (or thought they applied) 450 volts to a person who had long since gone quiet after screaming in pain and after complaining of heart problems. The subjects did so only because they were told by a person in a lab coat that such shocks were necessary to complete a word comparison test.

In Experiment 18, 37 of 40 subjects continued to participate in the test to the very end. They continued to read the questions, to announce whether the answers were right or wrong, and continued to identify the level of shock that was to be applied. They did everything except throw the switch.

The only difference between Experiment 5 and Experiment 18 was who threw the switch. If the real test subject, the Teacher, had to throw the switch, the obedience rate was 65%. If the real test subject did not have to throw the switch, the obedience rate was 92.5%.

Assuming you believed you would have defied the Experimenter if participating in Experiment 5, do you believe you would have defied the Experimenter if participating in Experimenter 18? Would you have been among the 7.5% who did not, or are you simply suffering from Lake Wobegon syndrome?

If I discomfort you near Christmas (or Hanukkah or Kwanzaa or Festivus for the rest of us), consider it a cautionary gift. Alternatively, consider me the Gingrich who stole Christmas.

3 comments:

Anonymous said...

As I said in the other one too, I think it does a good job of explaining things we can't imagine happening at the individual level but do happen, like the holocaust. But group dynamics is a little different. I also think it's hard to study what happens because jury deliberations aren't recorded and only in big cases are they asked about their reasonings.

Here is a book I just saw the basics of, but said that we should charge juries if they choose guilty, but later exonerated.

http://www.amazon.com/More-Sex-Safer-Unconventional-Economics/dp/1416532226/ref=pd_sim_b_5/marginalrevol-20

But then that raises questions of any efficiency.

Mike

Anonymous said...

One thing I did want to add, even though we say that the burden of proof is on the state to prove that the defendant is guilty, just by getting arrested and sent to a trial the burden is more on the defense side instead of the state side. That would go along with the an argument from authority. There are many times we can say something but it isn't the case, like what exactly is the difference between beyond a reasonable doubt for criminal cases to the burden in a civil case.

Mike

tsj said...

Mike,
I'll be discussing the burden of proof issue soon, once again.

Post a Comment