Police Brutality: An Experimental View

October 5th, 2013   Submitted by Davi Barker

PoliceEye200In “Renegade Psychological Experiments” I announced my winning of the Agorist Pitch at PorcFest X and outlined my plan to proceed with the experiment. In “Authoritarian Sociopathy” I explored previous relevant research on obedience and authority. Now, I am in the design phase of my Agorist Pitch. This is the first draft of the design for a renegade psychological experiment on obedience to authority, specifically on police brutality.

If you are reading this I’d appreciate your help. Unfortunately, I am more of a writer than a scientist, psychologist or ethicist, although I’m consulting with numerous like-minded experts. I’m publishing this is a rough draft because I’m hoping to solicit general feedback to help me perfect the design.

The Problem

We are living in an increasingly militarized society, and I would argue that this has a primarily psychological cause, not merely a political cause. If allowed to continue this could have disastrous consequences, as it has throughout history. Further, I would argue that this problem stems not only from the psychology of those in authority, but also the psychology of those in obedience, specifically their tendency, whatever the reason, not to intervene when authority steps beyond its bounds. This sentiment was perhaps most eloquently expressed by Thomas Jefferson in this seldom quoted passage of the Declaration of Independence:

“All experience hath shewn, that mankind are more disposed to suffer, while evils are sufferable, than to right themselves by abolishing the forms to which they are accustomed.”

The militarization of society cannot be fought only with votes, or with cameras, or even with rifles, if the underlying impulses for compliance are not first addressed in the mind of every subject who slavishly accepts their subjugation. That is why the psychology of obedience is not merely a tool, it is a map of the problem itself.

Ethical concerns raised by previous experiments lead to changes in the APA ethical guidelines, which have made it almost impossible to study the psychology of obedience and authority. Because of the new limitations the more recent experiments on authority are far less dramatic than the previous ones, and have failed to penetrate mainstream culture.


Purpose Of Further Research

What is clear from previous experiments is that human nature is neither good nor evil, but essentially adaptive. If you take an otherwise good person and invent for them a role that incentivizes evil they will adapt to the new circumstances. And if you deeply internalize “obedience to authority” as a core personality trait you will become capable of the worst forms of murder, and tolerant of the worst forms of abuse. But even those deeply familiar with the Stanford Prison Experiment and the Milgram Experiment have usually never heard of the less dramatic studies. Devoid of shock value this research doesn’t impact the culture, and so it fails to safeguard society from the dangers of obedience. Changes to the ethical guidelines have essentially neutered research on authority and obedience. It has been relegated to water cooler banter among academics.

APA ethical guidelines are not law. They are essentially a criterion for public funding. So, if the ethical guidelines of legacy institutions hamstring meaningful research on obedience to authority, for the sake of safeguarding against the disastrous consequences of militarizing society, it is time for us to cast off these restrictions, and devise new ethical guidelines. It is time to conduct our own renegade psychological experiments, to show the world beyond doubt that power corrupts absolutely, and corrupt power deserves no obedience.


· Given the opportunity, a significant portion of the general population will not intervene in a clear incident unprovoked police brutality.

· There will be a statistically significant difference between the percentage of people who will intervene in an incident of police brutality, and people who will intervene in an identical incident of brutality by someone in civilian clothes.

· Demographic information can be discovered which correlates with higher rates of intervention in an incident of police brutality.

Methods and Procedures


Renting a meeting room in a shopping mall commonly used for consumer surveys, we will present ourselves as soliciting feedback on a new movie trailer from the general population as a convenience sample. Volunteers in unremarkable clothes and holding clipboards will offer shoppers some small reward if they agree to participate in the survey.

The volunteers will lead participants down a hall to the waiting area. The hall will have a fully visible surveillance camera. Volunteers should make some gesture to the camera to bring attention to it. In the waiting area there will be a couch from which the participants can see the door back to the hallway, and the door to a back room, both labeled “exit.” They can also see a surveillance screen displaying footage of the hallway they just came from, but no sound.

The volunteer will knock on the door to the back room, and the surveyor will emerge, at which point the volunteer will exit back down the hall, and be visible on the surveillance screen. The surveyor will explain to the participant that they need to fill out a short questionnaire while they set up the trailer. The surveyor will retreat into the back room while the participant fills out the clip board.

The questionnaire will ask for all relevant demographic data: age, sex, ethnicity, income, education, political affiliation, etc. It will also contain misleading questions about movies. How often do they go to the movies? What genre of film to they enjoy? What movies have they seen recently? etc. Embedded in that list must be the question, “Are you comfortable watching violent footage?”

When the surveyor returns for the clip board he will also thank what appears to be a previous participant, but is actually a confederate of the experiment. The surveyor asks the participant to wait just a few more minutes as he sets up the video and returns to the back room. The confederate exits through the hallway, leaving the participant alone in the waiting area.

As soon as the confederate exits the room the surveillance screen will begin displaying a prerecorded incident of unprovoked assault. A sound recording of the incident will be played in the hall, to create the illusion that the confederate is being attacked for no reason right outside. The brutality will begin with shouting, escalate to shoving, then a beating which could reasonably result in serious injury, until finally the assailant drags the confederate off camera.

Half of the participants, selected at random, will see a video in which the assailant is wearing a police uniform, and the other half will see the same scene except the assailant is wearing civilian clothes. The clothing of the assailant should be changed digitally so there is no disparity in the performance of each scene.

If the participant opens the door to the hallway that will be counted as an intervention. If the participant either takes the exit toward the back room, or stays in the waiting area until the end of the footage that will be counted as not intervening. Regardless of the outcome, once the participant has made their choice the illusion will be revealed and the entire scope and purpose of the experiment will be explained.

The surveyor will conduct an exit interview. All participants will be asked to complete an emotional survey describing how they felt and what they were thinking during the experience. They will be asked what motivated them to make the choice they did. Participants who intervened will be asked what they intended to do once they entered the scene. Were they going to yell at the assailant? Would they physically intervene? Or record the incident? Participants who took the other exit will be asked where they were going. Were they searching for an exit, or seeking help from the surveyor?

Ethical Concerns

Preventing physical harm:
There was concern that involving subjects in a staged incident of police brutality could put the confederates at risk of injury if the subject decided to intervene physically. The illusion created by the prerecorded footage, and the use of the door as a measure of intervention was devised as a way to protect everyone from physical harm.

Preventing forced witness:
There was concern that it may be unethical to force a subject into a situations where their only choices was to witness potentially traumatic brutality, or to intervene. To mitigate this risk the level of brutality in the footage should not exceed what may reasonably be seen in mainstream news, and the exit sign above the door to the back room should be added to create a third option of neither witnessing, nor intervening.

Reconciling trauma:
There was concern that in spite of the violent footage waiver in the questionnaire and the other precautions a subject may still find the experience traumatic, or emotionally distressing, especially if they have a personal history of brutality. To remedy this unfortunate result,, if it occurs, someone will be on hand to offer private exit counselling after completion of the experiment. If subjects take this offer, these sessions are not considered part of the exit interview, and will not be part of any analysis. In addition, every subject will be given the contact information of local counselling services, in case they experience distress after leaving, or even days later.

There is concern that it may be unethical to deceive subjects, both by asking them if they’d like to review a new movie trailer, and by creating the illusion of the assault. Subjects will be fully informed of the scope and purpose of the experiment afterward, however, there is no way to avoid deceiving the participants without biasing the results. We acknowledge this concern, but regard it as an acceptable risk given the potential value of the study.


Sample Quality:
A convenience sample from a shopping mall is not a perfect random sample of the general population. It will be weighted by socioeconomic status, age, and lifestyle factors that make one likely to shop in a mall.

Victim association:
Any apparent demographic or lifestyle information that can be gleaned from the appearance of the victim could impact the decision to intervene. Future studies should include variations where the victim and assailant are of variable race, gender, creed, sexual orientation, age etc.

Risk assessment:
A police uniform does not only indicate authority, is also indicates weaponry. The civilian assailant should be similarly armed, but it may not be as obvious. Disparity in rates of intervention may be influenced by disparity in perceived risk. The exit interview should be crafted to account for this.

Subject Isolation:
Subjects may be reluctant to intervene if they are alone, which does not reflect the typical real world incident of police brutality. This may bias the data toward non intervention. Future studies should include variations accounting for group dynamics.

Contrived Scenario:
Knowledge that this is a contrived scenario, and that they may be being observed, even believing it to be a consumer survey, may bias the data toward intervention. Further, if the performances of the actors, or the oddities of layout cause them to suspect a set up, they may doubt the reality of the footage.

Data Analysis

The relevant statistical data will be the rate of intervention in the group which saw footage of a police assailant, the disparity with the rate of intervention in the group which saw footage of a civilian assailant, and any demographic information which shows a statistically significant difference within each group.

Further data on emotion and motivation gleaned from the exit survey will be used to identify potential causes, avenues for future research, and general discussion of the issue, but is not directly relevant to the hypotheses.


If the rate of people willing to intervene in an incident of unprovoked police brutality is startlingly low that will give us some indication of the severity of the problem of obedience in society, and lend weight to the argument that police militarization is made possible in part by the complacency of civilians. If the rate is surprisingly high it will invalidate this hypothesis and indicate that other causes of militarization should be explored.

If the rate of intervention in the police assailant group is significantly lower than the rate in the civilian assailant group that will confirm that aggression from authority figures is more tolerated than aggression in general, indicating that authority itself may increase or even incentivise aggression. If no significant disparity between the rates is discovered that will invalidate this hypothesis and also indicate that other causes of militarization should be explored.

If specific demographic information is found to be correlated strongly with intervention that will indicate that obedience to authority is learned behavior, and not innately human nature. That should guide avenues for future discussion and research aimed at discovering both the root of the learned behavior, and the factors which contribute to learning the opposite behavior. If no correlation is discovered that will invalidate this hypothesis and indicate either that obedience to authority is innate, or that weaknesses in the experimental design failed to screen for the causal factors.

The raw data from the experiment will be made public record online, and the video of subjects who allow us to publish their image will be made into a documentary which will be available online. We will also solicit interviews with experts in the field to provide commentary and analysis.


If discovering the psychological cause of obedience to authority is the key to preventing the militarization of society, than research such as this is the key to avoiding the disastrous consequences of militarization. Further, if APA ethical guidelines prevent meaningful and influential research in this area, than we must be willing to fund research such as this privately, and to trumpet the results publicly so as to influence the culture directly.

Tags: , ,

27 Responses to “Police Brutality: An Experimental View”

  1. Seth KingNo Gravatar says:

    Let me see if I understand this correctly.

    We know from the Milgram and Stanford experiments that people are willing to obey perceived authority to great lengths. What we don’t know is why? Correct?

    And the object of this experiment would be to find out why? I’m not sure I understand HOW this experiment will find out why people obey perceived authority.

    • Davi BarkerNo Gravatar says:

      To some extent this is a different area of the field than Milgram or Stanford. This isn’t a study of what people are willing to do, or willing to endure, but a study of what they are willing to witness, and when they are willing to intervene. I think this is potentially more important data, because in the typical tyrannical scenario the perpetrators are a minority of the population, and the victims are a minority of the populations, but the witnesses are the majority, and thereby the most capable of meaningful intervention.

      To definatively determine why is probably the task of a future study, but homefully the demographics survey in this study can point us in that direction. For example, if we learn that home schoolers intervene more than public schoolers, young intervene more than old, poor intervene more than rich… or vise versa, whatever. That would be useful data not only to direct outreach toward high intervention demographics, but also to guide the design of the next version of this experiment.

  2. Nathan T. FreemanNo Gravatar says:

    I love this idea. I think it’s brilliant. Is there a Kickstarter or something where I can help make it happen?

    • Davi BarkerNo Gravatar says:

      There will be. I’d rather run it as a Bitcoin Starter, but it’s a tough call. There’s nothing particularly bitcoin about it. That’s just my preference. But I wouldn’t want to refuse funding in any way. I’ll probably start the fund raising in February. We’re planning on recording a trailer where we go through the experiment with all actors, just so that donors can see exactly what they’re contributing too.

  3. David LeemanNo Gravatar says:

    Setting of new and restrictive guidelines for these types of experiments is, in effect, showing that they are willing to succumb to authority. And you didn’t have to raise funding for that data.

    • Davi BarkerNo Gravatar says:

      Why? We’re not claiming any authority to impose ethical guidelines on anyone. We’re merely being transparent about the guidelines we’re using for ourselves.

  4. Don DuncanNo Gravatar says:

    Exit interviews should include contact info and instructions so the subject can give more data after reflection. People may not be able to give all the data if in an emotional state. When they calm down their input may increase or change.

    Also, it would be good to know if their experience was a “teaching moment” after reflection, or if they would like to receive the results, or if they would like to volunteer or help in any way.

  5. Don DuncanNo Gravatar says:

    Davi: As I witnessed the unofficial martial law at Watertown I got a flashback to the films the Nazis took when they rounded up and pushed around Jews. IT WAS THE SAME! I was sickened. I always wondered what I would do, and why those Jews did not fight back or resist, even a little. It’s too late to ask, but not too late to ask the Watertown folks.

    I recommend you go to Watertown and interview on camera as many as possible. There must be hundreds. This study could be made into a very low cost documentary which could be used to fund a larger study.

    I would like to know: Besides fear, did they feel shame? Were they angry at the military? Did they resist? If not, did they consider it? What did they think/feel later when they thought about what happened or saw themselves on TV? Would they do anything different now? Have they made a formal complaint? Do you have a gun? Does it make you feel safer from criminals? From the military? Do you consider your home a refuge and your castle? Is it necessary to force people out of their home for hours, partly dressed? Is it moral? Is it legal?

    If they answer: “No” to the last 3, What are you going to do about it?

    Or, are they glad they were removed from their homes? Do they feel safer when the military is rounding them up? Do they support this and want to see more?

    A question to ask first, and at the end of interview: Do you believe you live in the land of brave and home of the free?

    • Davi BarkerNo Gravatar says:

      That’s a neat idea. Additional man on the street interviews would be a nice compliment to a finished documentary.

      It would be kind of neat to include the entire World’s Smallest Political Quiz, so we could actually create a plot graph of the participants political views.

  6. state haterNo Gravatar says:

    I have thought of an additional potential weakness: contamination. Since this experiment will presumably be done in serial rather than parallel (that is, subjects from the same mall will be taken sequentially to the room, rather than using one subject per mall in scores of malls, which would probably be prohibitively expensive), there is a strong risk of contamination, especially from lemmings who take offense to (1) being part of an experiment without knowing it, and (2) a group of individuals trying to undermine the precious status quo that they know and love. If done serially at one mall, I suspect that there is a high chance that at least one subject will inform other shoppers via small talk (perhaps in a food court, or while shopping), and one or more of these shoppers will be selected and feign ignorance of the experiment, and intentionally skew the results (most likely by intervening against the cop, whereas they would not do so if they were not aware of the experiment). Worse, the subject may complain to the mall owner and have the experimenters ejected before the experiment is completed.

    I really like this idea, and think that the design is very good, overall. However, in light of my constructive criticism, perhaps it would be worthwhile to to do some hardcore fundraising and rent space in several malls (ideally at the same time, if there are enough experimenters to allow this), and limit the subject to one or a few in each mall. This would eliminate, or at least greatly reduce, the risk of contamination.

    • Davi BarkerNo Gravatar says:

      That’s a good point. I figured asking if they were tipped off in the exit interview would have resolved the contamination problem, but if they’re dishonest I don’t know how you solve that.

      As for the mall owner, we would necessarily be transparent with them as they are the property owner. If we can’t find a mall owner who consents we’ll have to figure something else out.

      Renting space in several malls is prohibitively expensive. The goal sample size is 100. If we’re only doing a few in each mall we’re talking about hundreds of thousands of dollars to solve an experimental weakness that can be addressed with a paragraph in the analysis. We need to start with a first study, even if it’s imperfect, even if it’s small, so we can leverage the success into fundraising for a larger study. All the variables that people want us to study require additional funding, and will probably have to come later, and hopefully once the base study is complete others will attempt to repeat it independently, to verify the conclusion.

  7. Just reading alongNo Gravatar says:

    I’m a former investigator and current anarchist/outlaw. I have lived both sides of obedience to authority my whole life. Its obvious that you have put a lot of time and thought into this, but you may have overlooked a critical weakness of data collection.
    What mall manager is going to allow you to traumatize their customers? Obviously, you can’t tell them the true intention of the test, so “misrepresentation of purpose” when renting space and immediate, police involved eviction comes to mind. And should you get an ok, all the releases and disclaimers in the world are not going to stop mall mgmt. from suing as soon as one single customer comes out of the test crying. First call will be to the police, who will ask the testers about the test. When coppers find out their actions are the subject of the tests, well, subjects just might be given a first hand trial to see if they’ll intervene.

    • Seth KingNo Gravatar says:

      Good point. The mall isn’t the best location.

    • Davi BarkerNo Gravatar says:

      I’m not convinced a mall manager wouldn’t allow this, or that we can’t tell them the intention of the study. No harm in asking. But I’m not comfortable misrepresenting the purpose of renting the space to the property owner. So, if it turns out to be impossible we’ll have to figure something else out. Any suggestions?

      Zimbardo consulted with the police in the Stanford Prison Experiment. The subjects were arrested and booked by real police. I’d be curious to see what the cops say about the test, because it’s not a test of their actions, it’s a test of police brutality. If they tried to construe it as a test of their actions, they’d have to admit that brutality was part of their acceptable arsenal of actions. It seems like it’s be easy to trap a cop into appearing to defend brutality if you interviewed them on camera about the experiment.

      • Seth KingNo Gravatar says:

        The problem I see is that the mall manager is going to ask for your permission slip. I’m willing to bet you need permission to do any sort of psychological testing.

        • Don DuncanNo Gravatar says:

          If there is a law that you need permission or special credentials for psychological testing, then call it a sociological study or better yet, a poll.

  8. Just reading alongNo Gravatar says:

    I applaud your efforts and believe in the importance of the study, but I’m also a realist. Millgram and Stanford were both designed and conducted by trained psychologists, who ended up scorned by their peers for their work, even though the tests were done in a highly controlled manner under the protocols of their time. In the end, all they really did was confirm that humans are cruel and vicious creatures.

    I spent many years working in retail, no mall manager is going to let anyone show violent videos to his patrons in his/her mall. Barney videos and free Cinnabons maybe, police brutality, no. People are there to shop, not be traumatized.

    Maybe this would fly in a collegial setting, or by approaching groups like the Elks or VFW for partnership. I’m hermitic, so I don’t know where large groups gather anymore, but that would be a start.

    Remember, Stanford wasn’t cops, it was prison guards. Prison guards are not in the public like cops. Unfortunately, most of the public probably sees nothing wrong with the harsh treatment of prisoners, they’re not in prison for their good deeds after all.
    Any time you do anything that might somehow show the police in a bad light, whether its their “accepted” practices or not, you’re going to run into trouble with them. Just ask anyone who has ever tried to film them.

    • Don DuncanNo Gravatar says:

      A “realist” would know that negative reviews of Millgram and Stanford said more about the reviewers than the studies. Read the studies and evaluate them for yourself.

      Cruelty exists. No confirmation was needed or sought. That was not the object of the studies or more importantly, not “…all they really did…”. The studies showed normal people’s reaction in an authoritarian environment. As such, it put authoritarianism in the spotlight. It exposed the inhuman actions authoritarian control provokes. It showed that average healthy people can act like monsters when they defer to authority. This undermines the moral & practical argument for authority. Since we live in an authoritarian political society the pimps for coercion would certainly attack the studies.

      You warn of “trouble” for exposing the problems with being ruled. Is that “trouble” worse than doing nothing? Could anything be worse than self censorship? Isn’t that cowardice?

      Davi does not seek trouble. He seeks to expose the trouble most people are ignoring. He seeks to enlighten. He seeks to end “the sanction of the victim”.

  9. Michael PriceNo Gravatar says:

    Excellent idea. If nothing else it will make people aware of their subservience. This of course is the most painful thing you can do to them.

  10. Greg LilleyNo Gravatar says:

    I love this idea and I think this design is very clever. I have two thoughts:

    1) Phones. It strikes me that instead of intervening directly, some people might react by trying to call someone like 911 or security. You don’t want a swat team descending on your operation, so you might want to use some sort of cell phone jamming device. You might also have a dummy land-line phone available on the table with a directory prominently displaying the numbers for security and mall management. If the participant uses the phone, that would be recorded as an indirect intervention.

    2) Guns. Your section on risk assessment indicates you’re already aware of this problem. Policemen always have guns, so this is probably perceived as normal. Civilians on the other hand rarely have guns, and it may be perceived as abnormal and possibly more threatening, which might lead to participants being less willing to intervene when an armed civilian is the one delivering the abuse. I can’t think of a better way to handle this problem than asking about it in the exit interview as you suggest.

    You might enjoy seeing “The Wave”, which is about a charismatic high school teacher who conducts an experiment and transforms his class into a group-thinking mob with a logo, a uniform, and a special salute. Two girls rebel at seeing their individuality suppressed. The film is in German with subtitles.

  11. AgoristTeen1994No Gravatar says:

    Hey Mr. Barker, is there any way I can help with this experiment? Because this is something, I would love to help with…since it combines my love of research and science, with my desire to spread anarchistic ideals.

    @Mr. Lilley, that’s interesting that you mentioned “The Wave” since when I in my sophmore year at the public high school I went too, the special ed teacher I had (I was placed in special ed because I “had ADHD” and was “smart but undisciplined/unmotivated” but that aside Mr. Stevens was a pretty cool teacher) had the class read the book version of “The Wave”.

  12. RagnarNo Gravatar says:

    I have some ethical problems with your experiment, but will table those for now and point out a couple of concrete flaws I see.

    1) Disparity of threat isn’t something you can eliminate by giving the plain-clothed attacker a gun and a club. Cops aren’t dangerous because they carry guns, shit, I carry a gun, too. So what? Cops are dangerous because each and every one of them is armed with the single most dangerous weapon an individual infantryman can carry… A radio. You’re never fighting “just one” cop. It’s always all of them. Additionally, if I wind up shooting a mugger, I have a very good chance of getting off. If I wind up shooting a cop, my life is over. You cannot adequately compensate for that disparity of threat in any way I can think of offhand.

    2) Severe sample bias. Your subjects are self-selected, because you asked if they wanted to view a movie, and they said “yes”. I would say “No, leave me alone.” I would also be very unlikely to be in a mall in the first place. Therefore, your study won’t include anyone like me, and I believe it is predisposed to exclude those most likely to resist authority and refuse to comply, as I suspect they are the exact same people who will tell you “no” when you want them to volunteer their time.

    3) Your “exit interview” process is, frankly, total shit. You’re going to ask people who didn’t intervene WHY they didn’t intervene, and expect them to be honest?

    Interviewer: “Why did you take the exit door?”
    Subject: “Oh, I was, um, going to get help.”

    Those answers’ veracity is basically non-existent, as basically ALL of your subjects will be attempting to give answers OTHER THAN:

    “I was scared and only cared about myself, I’m a coward with no social conscience and avoid confrontation at all costs.”

  13. RagnarNo Gravatar says:

    Oh, yeah, and…better figure a way to deprive your subjects of cellphones, or some will inevitably make false calls to 911, subjecting you and everyone working with you to significant civil and criminal liability.

  14. RagnarNo Gravatar says:

    Better have paramedics on standby, too. Dollars to doughnuts someone has a heart attack, or a panic attack they THINK is a heart attack, or runs away so fast they tumble down some steps and break their head open. You might also look into some kind of massive, umbrella, liability coverage. You’re gonna get sued. A lot.

  15. KeithNo Gravatar says:


    It isn’t only a cop’s intangible “authoriteh” which would inhibit individuals from intervening.

    Cops have a proven ability to seriously trash the lives of any who cross them. Arguably more so than any private mafiosi.

    They’re trained to lie, and they stick together; any who speak out against injustice and abuse are ousted and usually victimized too.

    Crossing a cop is likely to cost tens of $k along with possible beatings, cagings and worse.

    I agree that the state exists only in the brains of its believers, but those delusional believers can still cause some serious and very real problems, which most people are well aware of and will want to steer clear of.

    How to control for those?

    I don’t know – perhaps avoid the cop altogether and go for some bogus alphabet soup agency, which is detaining your volunteer for the “very serious offence” of participating in an “unlicensed activity”

    though again, with anything to do with the state, there is a real possibility of years of harassment, expense and hassle for not giving the appearance of complying.