The Game Theory of the Stanford Prison Experiment
https://www.simplypsychology.org/zimbardo.html
The Stanford Prison Experiment was a landmark study conducted at Standford University in 1971. The researchers wanted to investigate whether the violent nature of guards toward prisoners in the U.S. prison system was due to the inherent nature of the guards, or whether it was the rigid social hierarchy and environment of prisons which makes it more favorable for the guards to be violent; in other words, whether the social environment and roles of prisons creates a dominant strategy for guards to be violent towards prisoners.
The test subjects comprised of 24 students split into 9 guards and 9 prisoners (with the other 6 students as fill-ins in case they were needed). The prisoners had to live in the mock prison 24/7, while the guards were divided into groups of 3 and split into three 8-hour shifts (morning/evening/over-night). By the 6th day of the Stanford Prison Experiment (at which point the entire experiment had to be shut down because of the craziness which ensued), the guards, who were selected due to being seemingly good-natured and kind students, were physically and mentally abusing, humiliating, and dehumanizing the prisoners. On the other hand, the prisoners, who were originally working together and forming a unity of sorts, were now only caring and fending for themselves, doing seemingly anything so long as it was in their best interest.
I have extrapolated information about what occurred during these 6 days and created a 2×2 game theory table outlining the strategies of the prisoners and the guards:
P1: Disobey G’s, work with other P’s
P2: Be obedient towards G, don’t work with other P’s
G1: Follow standard procedure towards P’s
G2: Cause P’s undue suffering
At the beginning of the experiment the guards were following strategy G1: they were not overtly nice to the prisoners, however, they were not at all harsh and were just going about standard duties like keeping watch and conducting meals. Meanwhile, the prisoners were following strategy P1, and by the 2nd day of the experiment, they were barricading themselves in their rooms and making fun of the guards (in fact, it was even noted that some prisoners were glad to have the role of ‘prisoner’ during this experiment because it was fun and there was essentially no work or responsibility on their part). The guards, shocked and humiliated, began to change their strategy at this point. They realized that strategy G1 was failing them since they received very little payoff and had to deal with more work in managing the prisoners; they switched to strategy G2, and saw almost immediate results. The prisoners were at first defiant, so the guards started to spread rumors of false rewards and punishments, and even about things that the prisoners did. They also began to treat the prisoners harsher and more strictly (only addressing them by their ID numbers, not giving certain prisoners bedding, not allowing them to smoke cigarettes like before, forcing physical punishments like doing push-ups, solitary confinement, etc.). This was far easier for the guards because of the tendency of the prisoners to listen more out of fear and to be far more obedient than before, showing that G2 is a dominant strategy. Thus, the prisoners had to choose their best response knowing that the guards would choose G2. For them, they received the least ‘worst outcome’ when they chose to be obedient and fend for themselves, since being disobedient resulted in far worse punishments. This is evident by the fact that by the 6th day of the experiment, nearly all prisoners switched to strategy P2 and were even snitching on other prisoners to get better treatment themselves. Although P2 may not seem like a good outcome for the prisoners, since they were living in constant fear, it was the better of two bad strategies (they are prisoners in a model of the U.S. prison system, after all, so it is expected that they would not have a positive payoff – perhaps a testament to the debate that our country’s prison system is inherently flawed). I am fascinated by this and wonder how the game theory matrix looks like in foreign prisons (if one even exists).