Friday, June 5, 2020


A theme of protest demonstrations since late May 2020 is that police violence persists despite previous episodes of public outrage and efforts at reform. The problem has not been solved, including by the protests themselves.

Police violence was prominent in triggering the uprisings of the 1960s. The two most destructive riots were both started by police arrests: Newark in June 1967 (26  dead); Detroit in July 1967 (43 dead). In Newark 5 days of riots began after a taxi driver was arrested; in Detroit, when police attempted to raid a popular after-hours club, patrons fought back by attacking police cars; backup was called and eventually the National Guard; fighting with snipers, arson and looting lasted 4 days. The pattern continued in riots over the acquittal verdict in the Rodney King beating by the LAPD in 1992, and a long series of highly publicized cases through the Ferguson Missouri protests of 2014 and down to today.

There have been occasions where police have been adulated; notably in the public ceremonies so prominent in the months after the 9/11/2001 attacks, when police and firefighters were repeatedly honored for their sacrifices at the Twin Towers. On the other side of the ledger, there are a series of reasons why large portions of the public -- not just African-Americans-- dislike the police, and will join in protests against them.

[1] Police are used for collecting fines for municipal budgets. This has been a long-standing practice in speed traps, where heavy fines are levied on drivers, usually on highways outside of town; since locals know where the speed traps are, it falls mostly on strangers (similar to resting your budget on hotel taxes in popular tourist destinations). Cities where there is strong resistance to tax increases, or which have serious budget short-falls, often explicitly adopt the policy of increasing fines for all sorts of infractions. It then becomes the police duty to seek out offenses, however trivial; they are expected to produce at high rates, sometimes with quotas set by police officials (Moskos 2008). This was a notorious practice in Ferguson, where the protests began after police shot a young man who defied an order about walking in the street.

In Philadelphia, Alice Goffman (2014) showed how computerization of court records and police communications has intensified pressure on persons (mostly minorities in the ghetto) who have some kind of previous record. Offenses may range from drugs to violence to gang association; police stops on the street immediately run a computer check in their car, above all for outstanding warrants. These often involve failure to appear for a court hearing, or failure to pay fines, since the penalties for everything include fines. It becomes a vicious cycle as fines mount up. The courts are overburdened, and this combined with attempts to reduce over-sentencing to prison, results in most offenders being released but required to make future appearances and pay fines which they can’t afford. Persons caught up in the system no longer can get a bank account, a legitimate job, or driver’s license -- which generates further fines. Police, as the front-line enforcers of the system, are understandably unpopular. On their side, police also regard the criminal justice system as a revolving door.

[2] Police are used for enforcing unpopular regulations. A long history includes prohibition on alcohol (now mostly passé except for prohibitions on young people); prohibitions on marijuana (ditto). All of these promote counter-cultures of defiance. There have been many examples during the stay-at-home lockdowns during the coronavirus plague. Public parks have been closed, playing ball prohibited, beaches and/or their adjacent parking lots are closed; children’s playgrounds roped off. In many instances, ordinary people find these prohibitions inconsistent or irrational-- areas closed even if people maintain their distance; young people who have heard the statistics and know that their chances of surviving the coronavirus are above 99 percent. It appears that another counter-culture of defiance is building up today, likely to become exacerbated during the phase of opening up public activities under a regime of masking and social distancing. To a considerable degree, this coincides with conflict between age groups.

What many people regard as trivial offenses can escalate when officials enforce the rules. In San Diego, a black man walking his dog in a state park (actually the old Spanish settlement) was accosted by park rangers; when he refused to leave, they called police backup, who arrested the man; when exiting the police car downtown, he slipped his handcuffs, ran away, and was shot and killed. His mother said he was schizophrenic and did not understand the order to wear a face mask.  (San Diego Union-Tribune, May 6, 2020)  This is the archetype of many such events: one damn thing leads to another.

[2a] Police hypocrisy and cynicism. In both [1] and [2] police are required to carry out the dirty work of government. When this becomes the primary part of their job, it makes them cynical and hardened. They know that it doesn’t necessarily make sense to punish harmless violations, or that they are lying when they say their city-mandated increase in traffic stops are purely in the interest of public safety. In their own work lives, they are under a regime that demands hypocrisy; after a while, this unpleasant feeling turns into a bitter that’s-the-way-it-is.  Like prison guards who have to play the role of the bad guy, they embrace the tough-cop image. (Striking descriptions of this are in Jennifer Hunt’s 2010 close-up ethnography of the NYPD.)  Citizens who argue with cops about these things  increase the tension; one reaction is to be more aggressive. Taking videos of the police is felt as threatening them; and this can lead to attempts at retaliation.

[3]  Police dislike defiance. Jonathan Rubinstein (1973), a sociologist who joined the Philadelphia police in order to study their everyday life (similar to Peter Moskos in the Baltimore PD 30 years later), found that their number-one priority is to be the person in control in all encounters with civilians. For the most part, a cop is out there alone, or with a single partner; they are almost always outnumbered by civilians. Particularly in areas where they know they are unpopular, they feel it is imperative to not let things get out of control. They want to be the one who starts and ends the encounter, who sets the speaking turns (micro-sociology of conversation), who sets the rhythm of the interaction. Acts of defiance, whether micro-actions on the level of voice and gesture, or more blatant words and body movements, will cause a cop to increase their own aggressiveness in order to maintain dominance (Alpert and Dunham 2004). This a reason why trivial encounters with the police can escalate to violence far beyond what seems called for by the original issue.

[3a] Inner-city black code of the street emphasizes defiance. Elijah Anderson’s ethnography of black street life (1999; also Krupnick and Winship 2015) point out that in dangerous areas, where the police are distrusted, most people adopt a stance of being hyper-vigilant about threats and disrespect, and portray themselves as ready to use violence. Anderson says this is mostly a Goffmanian frontstage, a pretence at being tough designed to avoid being victimized. When dealing with the police, this leads to another vicious circle. Black people, particularly on their home turf, are more defiant of police than are whites; often this is no more than a confrontational way of talking, but these are micro-interactions that arouse police aggressiveness. Anderson notes that one reason people in the ghetto are wary of calling police is that they themselves may end up being arrested, because of the tone of these micro-interactions. Donald Black (1980), who pioneered observer ride-alongs in police cars, found that police arrested black suspects more than whites, but this happened when black people were defiant, which was more often than whites. Martín Sánchez-Jankowski (1991) in his gang ethnographies (including black, hispanic, and white) describes the culture of gang members as “defiant individualism.”  The pervasiveness of the street code in black lower-class areas, even among the majority who are not sympathetic with a gang life-style, hardens mutual hostility between citizens and police.

[4] Police dislike property destruction.  Anne Nassauer [2019] who studied protest demonstrations in the US and Germany by compiling videos of these events, was able to pin-point the conditions that led to a turning point where violence broke out. One of the major conditions was when police could see protestors destroying property, but were unable to do anything about it; this happened if they were under orders not to respond, or when they had relatively limited forces compared to the numbers of protestors. Normally police are concerned to prevent robbery and vandalism; it is one of their more favored duties, since they get to be the heroes protecting people. But now they are in a situation where they have to stand by and let it happen. This builds up their frustration. Although they may perceive that only a small part of the crowd is doing the destruction, they dislike the crowd for providing the opportunity to get away with it. Given further trigger events during the protest-- more on this in [5]--  police will take out their tension and anger on whoever is nearby in the crowd.

Property destruction in a mass demonstration puts police in a damned-if-you-do, damned-if-you-don’t dilemma. If they take action against looters and arsonists, they get accused of whatever violence they use and casualties they cause. If they stand by and let the destruction happen, they are accused of neglecting their duty and not caring. Eye-witnesses to such scenes are particularly likely to be outraged (see letters to the editor in recent days).

[5]  Adrenaline overload and forward-panic attacks on unresisting targets. When tension builds up, humans experience rising heart rate, driven by adrenaline. At a high level, perception narrows in, time becomes distorted, fine motor control is lost. Nassauer found that the level of tension is visible in videos: whether the police are in relaxed or tense postures, and similarly with the crowd. When tension builds up, from escalating gestures of confrontation, unexpected movements by crowd or police units, police getting surrounded and cut off, a trigger point sets both sides in action. Adrenaline is the fight-or-fight hormone; it produces generalized arousal of the large muscles of the body, but in what direction will it go? Police, like soldiers, are trained to respond to high adrenaline arousal by attacking. Most civilians, of the other hand, will run. But the one reaction feeds back on the other. The crowd suddenly running away is felt by the police as a release of their own tension into action.

In interviews (reported by Nassauer and others), police say they can see the crowd is divided between peaceful demonstrators and a small number of trouble-makers; but when the situation boils over, the crowd is infected by the violent ones. --This is how the police perceive it; what happens is that the panic of the crowd running away puts the police in an over-the-top rush of adrenaline in which their own perception is narrowed. When police rush forward, they become likely to strike those who have fallen down, or are screaming uncontrollably. The content of what people are saying is lost; all that is heard is the sounds and sights of out-of-control people. Since the police are trained to operate as a unit, officers who rush forward with their comrades tend to imitate what they do; if they are striking someone on the ground, it must be for good reason, and they will join in or protect them.

I have called this “forward panic” because it is like a panic flight where the overwhelming emotion of the crowd increases individuals’ adrenaline level; but in this case, the adrenaline is driving them forward, towards an easy target who have their backs turned, running away or falling down. 

Police who have been in shoot-outs generally report that their senses are blurred, they have tunnel-vision, can’t hear the sounds of their own guns, don’t know how much time is passing (Artwohl and Christensen 1997). They also tend to fire wildly, with poor aim, and with an overkill of bullets as they empty their magazines. It is similar with those who deliver a large number of blows with their batons, or put their full weight on a captured suspect’s neck.  It is the same in military massacres (with a higher level of casualties because of more weapons).  There is the same time-sequence: a period during which tension has built up on both sides; a sudden tipping point when the tension is released; one side becomes incapable of resisting (because they are caught in a traffic jam, fallen in the mud, turning their back, running away); the result is  hot rush, piling on, overkill.

In real-life situations, violence is usually incompetent-- in the sense that it often fails to hit its intended target, or hits the wrong target, or is disproportional to what is necessacry to prevail.  Soldiers and police are much more accurate shooters on firing ranges than they are in the emotional conditions of real-life confrontation.The clichés of military and police officials refer to “surgical strikes” and proportionate response. But the military is all too aware of “collateral damage”, especially in counter-insurgency warfare, where violent enemies hide in the civilian population.  This is a close analogy to confronting peaceful protests in which aggressive militants cover themselves.

[6] Police training for extreme situations.  Police training tends to emphasize the worst-case scenarios. Knowing that firing in real-life situations is encumbered by high adrenaline, weapons instructors tell them to aim middle-mass-- the center of the body; don’t try to shoot for extremities like arms or legs (the cowboy movie myth of shooting a gun out of someone’s hand never happens). The result is, police shootings tend to be deadly. Emphasis also is on rapid reaction; in the worst-case scenario, the suspect is armed and dangerous; you have to train your muscle memory to react as quickly as possible.

There is sometimes training in how to calm dangerous situations, but this tends to be overshadowed by the quick reaction scenario: your life or someone else’s life is in danger; train yourself to react automatically.

Another process that enhances the atmosphere of worst-case scenarios is police communications. When police call for backup, they tend to emphasize the danger of the situation. When the call is propagated more widely, the message is propagated just as rumors are: the distinctive elements are dropped out as the message is repeated. A man on a highway overpass threatening suicide by jumping, will get transformed into the cliché-- suicidal and threatening to take someone else with him --  into armed and dangerous. This is how individuals end up getting shot dozens of times by an aroused network of converging cop cars. The distortion may start when a civilian calls in, starting with an ambiguous situation, which the police dispatcher (a civilian employee), transforms into the more conventional warning. This was the case with the famous incident in 2009 when a Harvard professor, a black man, arrived home and had difficulty getting his front door open, getting the taxi driver to help un-jam it. A well-meaning Harvard secretary passing on the street phoned to say a possible burglary might taking place, but did not mention anyone’s race on the 911 recording and said: “I don’t know if they live here and they just had a hard time with their keys”. The dispatcher transformed this into a house-breaking by two black men; the cop who showed up was restrained at first but reacted to the irate professor by arresting him.

Lesson: police training needs to be drastically reformed. And training for police dispatchers, as well as from one police car to another, needs to be instructed on how rumors are formed; and procedures to avoid inflammatory worst-case clichés.

[7] Racism among police. Some cops are racists. How many are there, and what kind of racists they are, needs better analysis. What kind? There is a difference between white supremacists of the pre-1960s period; stereotyping racists who think most black people are potential criminals; situational racists who react to black people in confrontational situations with fear and hostility; casual racists who make jokes. These aren’t insoluble questions; if ethnographers followed people around in everyday life and observed what they talked about and how they behaved in different situations, we would have a good picture.  And there still remains the further question, does one or another degree of racism explain when police violence happens?

My estimate is that racism among police is less important a factor than the social conflicts and situational stresses outlined in points [1-6]. To put it another way, if we got rid of racist attitudes, but left [1-6] in place, how much would police violence be reduced? Very little, I would predict.

What can be done? And how likely is it to have effects?

Let’s go through the list.

[1] Collecting fines for municipal budgets. Getting rid of this corrupt practice would be important for reducing hostility between police and citizens; especially since it is a version of color-blind racism insofar as it targets poor black areas. But how to get municipal officials to forego money that can raised without taxpayer consent?

[2]  Enforcing unpopular regulations. A solution would be to legalize more prohibited substances. It does raise a problem of trade-offs, such as deaths from fentanyl. And there are other kinds of prohibitions being invented from time to time, as in the coronavirus period. Some conflict of this sort is going to be with us for a long time.

[2a] If police don’t have to do the dirty work enforcing unpopular policies, they’d be a lot less cynical and hard-assed, and we’d get along better with each other. This depends on what we do about [1] and [2].

[3] The code of the street, ostentatious defiance. I think this is declining already, with the growth of a black middle class. On the whole, recent protest demonstrations are more civil than those of the late 1960s.

[4] Police anger at property destruction. This is a genuine dilemma; either way, bad feelings are created. If we had fewer riots -- if some of the other conditions get better-- this would be less of a problem. Caveat: racism and police violence are not the only things riots can be about; for example, the anti-globalization riots of the past decade in the US and Europe. We may well be headed towards increased class division in the future, among other things between the computerized elite (now riding out the coronavirus working from their nice homes) and the other two-thirds of the population whose jobs are steadily being replaced by computerized robots.

[5] Forward panic violence in policing demonstrations. There are ways that police (as well as everyone else) can learn techniques to monitor their adrenaline level, and to not rush into action until they have a clear perception of the situation and have reduced their heart rate by breathing exercises. This one is solvable.

This could go along in tandem with:

[6] Reforming police training. More than reforming police departments, we need full-scale investigation and reform of police academies. They need to get away from the emphasis on worst-case scenarios and the quick-trigger, muscle-memory approach to weapons training. As noted, civilian dispatchers as well as cops need better training about rumor propagation and its tendency to revert to stereotypes as messages pass along the chain.

[7]  Police racism.  If we have enough of these kind of reforms, this will take care of itself.

As of now, most calls for reforms reiterate long-standing demands for independent review boards and stronger penalties for police misconduct. Having a reform-oriented black police chief in Minneapolis did not solve the problem. It is dubious that the top-down approach would solve it, as long as the everyday conditions of police work go unchanged.


Alexis Artwohl and  Loren Christensen. 1997. Deadly Force Encounters.
Geoffrey Alpert and Roger Dunham. 2004. Understanding Police Use of Force.
Elijah Anderson. 1999. Code of the Street.
Donald Black. 1980. The Manners and Customs of the Police.
Donald Black. 1989. Sociological Justice.
Randall Collins. 2008. Violence: A Micro-sociological Theory.
Randall Collins. "Cool-headed Cops Needed: Heart Rate Monitors can Help." [posted 10.05.16]
Alice Goffman. 2014. On the Run: Fugitive Life in an American City.
Jennifer Hunt. 2010.  Seven Shots: An NYPD Raid on a Terrorist Cell and its Aftermath.
Dave Klinger. 2004. Into the Kill Zone. A Cop’s Eye View of Deadly Force.
Joseph Krupnick and Christopher Winship. 2015. “Keeping up the front: how disadvantaged black youth avoid street violence in the inner city.” in Orlando Patterson (ed.), The Cultural Matrix.
Peter Moskos. 2008. Cop in the Hood.
Anne Nassauer. 2019. Situational Breakdowns: Understanding Protest Violence.
Jonathan Rubinstein. 1973. City Police.
Martín Sánchez Jankowski. 1991. Islands in the Street.