We train police to be warriors — and then send them out to be social workers
The fatal mismatch at the heart of American policing.
Richard Nixon called police forces “the real front-line soldiers in the war on crime.” Bill Clinton, in his signing ceremony for the 1994 crime bill, called them “the brave men and women who put their lives on the line for us every day.” In 2018, Donald Trump described their job as follows: “Every day, our police officers race into darkened alleys and deserted streets, and onto the doorsteps of the most hardened criminals … the worst of humanity.”
For decades, the warrior cop has been the popular image of police in America, reinforced by TV shows, movies, media, police recruitment videos, police leaders, and public officials.
This image is largely misleading. Police do fight crime, to be sure — but they are mainly called upon to be social workers, conflict mediators, traffic directors, mental health counselors, detailed report writers, neighborhood patrollers, and low-level law enforcers, sometimes all in the span of a single shift. In fact, the overwhelming majority of officers spend only a small fraction of their time responding to violent crime.
However, the institution of policing in America does not reflect that reality. We prepare police officers for a job we imagine them to have rather than the role they actually perform. Police are hired disproportionately from the military, trained in military-style academies that focus largely on the deployment of force and law, and equipped with lethal weapons at all times, and they operate within a culture that takes pride in warriorship, combat, and violence.
This mismatch can have troubling — even fatal — consequences. Situations that begin with civilians selling loose cigarettes, attempting to use possibly counterfeit currency, sleeping intoxicated in their cars, recreationally selling or using low-level drugs, violating minor traffic laws, or calling the police themselves because they are experiencing a mental health crisis end with those same civilians, disproportionately Black Americans, unnecessarily killed at the hands of a police force primed for violent encounters and ill-equipped for interventions that demand mediation, deescalation, and social work.
“Cops are very equipped to be the hammer and enforce the law,” says Arthur Rizer, a former police officer and 21-year veteran of the US Army who heads the criminal justice program at the center-right R Street Institute. “They know how to use those tools forcefully and effectively; for everything else, they are lacking. Of course that’s going to end badly.”
There is considerable disagreement about the best way to change policing. But as my colleague Aaron Ross Coleman points out, a cross-factional coalition is emerging, centered on the idea that America relies far too heavily on police to address problems that have nothing to do with what they are trained, hired, and equipped to handle.
“The spectrum of skill sets we are currently asking police to embody is simply not realistic,” says Christy E. Lopez, a legal scholar at Georgetown Law who investigated police misconduct as an attorney for the Obama administration’s Justice Department. “It’s not realistic to ask any profession to do that much.”
In recent weeks, I’ve spoken to a dozen current and former police officers, police reformers, legal scholars, and criminologists to better understand this fatal mismatch at the heart of American policing — and what it would take to fix it.
How police officers spend their time on the job
The best information on how police officers spend their time comes from “calls for service” data made publicly available by individual police agencies. These are often defined as calls to emergency operators, 911 calls, alarms, and police radio and non-emergency calls. Most calls for service are initiated by citizens, but the data I draw on here captures the officer’s final categorization of the incident.
The data overwhelmingly finds that police officers in aggregate spend the vast majority of their time responding to non-criminal calls, traffic-related incidents, and low-level crimes — and only a tiny fraction on violent crimes.
My favorite visualization of this data comes from former UK police officer and Temple University criminologist Jerry Ratcliffe, who used 2015 data from Philadelphia, a city with relatively high crime rates, to construct this graphic. The area of each box represents the proportion of reported incidents within that category:
If you squint a bit, you can see that violent crimes like rape, homicide, and aggravated assault are tucked away in the bottom right-hand corner. Less serious crimes like petty theft, drug use, and vandalism take up slightly more space but not all that much. The vast majority of calls have nothing to do with crime. Instead, they involve disorderly crowds, domestic disputes, traffic accidents, minor disturbances, and a whole array of “unfounded” calls where the officer arrived on the scene only to discover nothing was happening.
Of course, the exact incident breakdown will vary by place, but this general picture holds for a number of police departments in major cities. In a June article for the New York Times, crime analysts Jeff Asher and Ben Horwitz dug through the call data for the 10 police agencies that had made such data available, including in places with relatively high violent crime rates like Baltimore and New Orleans. They found that incidents that met the FBI Uniform Crime Report definition of violent crime made up only around 1 percent of calls for service.
Then, for the handful of police agencies that also provided data on when a given call for service was first reported and when that incident was closed, Asher and Horwitz used the difference between those two numbers to gauge the time officers actually spent on different types of policing activities.
Across these departments, the biggest category of time spent by police was on “responding to noncriminal calls,” which took up around a third or more of total on-call time. The next biggest categories were “traffic” (mostly car accidents) and “other crime” (low-level crimes like drug use, truancy, disorderly conduct, etc.). Almost 10 percent of police time was spent on “medical” calls, which involve non-crime-related physical emergencies. Meanwhile, police spent only around 4 percent of their time responding to violent crime and even less time (closer to 0.1 percent) on homicides.
“When I was an officer, I got calls about dead animals, ungovernable children who refused to go to school, people who hadn’t gotten their welfare checks, adults who hadn’t heard from their elderly relatives, families who needed to be informed of a death, broken-down cars, you name it,” says Seth Stoughton, a legal scholar at the University of South Carolina and former Tallahassee police officer. “Everything that isn’t dealt with by some other institution automatically defaults to the police to take care of.”
Calls for service data do not include what police often refer to as “unassigned” time — the hours police officers spend between calls patrolling neighborhoods, taking a meal break, or filling out paperwork. Observational studies of patrol officers have found that anywhere from 46 percent to 81 percent of their time is spent on unassigned activities. That means the total percentage of time police spend responding to crime could well be far less than even the call data indicates (the main exception being members of specialized units in major departments like homicide and SWAT whose activities aren’t captured by observational studies).
Numerous academic studies confirm these basic patterns in the data. They find that patrol officers — even in suburban and rural communities for which public data is often lacking — spend the overwhelming majority of their time writing reports, driving around neighborhoods, and responding to non-criminal calls.
“The job is 99 percent boredom and 1 percent sheer panic,” says Matthew Bostrom, a criminologist at the University of Oxford who spent more than 30 years as a police officer, commander, and sheriff in St. Paul, Minnesota. “Most of what you deal with is fairly routine.”
In his recent paper “Disaggregating the Policing Function,” Barry Friedman, the director of the Policing Project at New York University’s School of Law, breaks down this dizzying array of tasks and responsibilities into a handful of distinct roles:
- The traffic cop: The majority of police-civilian interactions take place on the road. Police help stranded motorists with broken-down cars, take reports in car accidents, direct traffic around serious incidents in which other responders are needed, set and staff speed “traps,” and issue citations. And when police are off-call, they spend much of their time performing routine street patrol.
- The mediator cop: A huge number of calls to the police involve relatively minor interpersonal disputes: disputes over noise levels, trespassing, misbehaving pets, or rowdiness; disputes between spouses, family members, roommates, or neighbors. In these situations, police are called to calm things down, deescalate, and act as counsel.
- The social worker cop: Police work often involves populations like the homeless, intoxicated people, people with substance use issues, or those with mental illness. This role isn’t often captured well in the aggregate data, but police spend a huge chunk of their time on these functions.
- The first responder: In most jurisdictions, the only government entities that respond to problems 24 hours a day, seven days a week are police, fire, and emergency medical services. That means for the vast majority of social problems, police are often the default institution for people to call. This is how cops get stuck chasing runaway dogs, tracking down welfare checks, dealing with noise complaints, and a whole host of other issues that appear to have nothing to do with policing.
- The crime-fighting, law enforcement cop: There is something to be said for rapid response by force- and law-trained individuals to situations involving serious criminal activity. However, studies find that this time is mostly spent interviewing witnesses, gathering evidence, advising victims, and writing reports. “Often cops are just there to pick up the pieces after the fact,” says Peter Moskos, a former Baltimore police officer and criminologist at John Jay College. “By the time you arrive, the crime is usually no longer in progress.”
The time a given officer spends on each of these roles varies greatly. In bigger cities, police work tends to involve dealing with a lot of substance abuse, mental illness, and homelessness. In suburban areas, domestic and other interpersonal disputes take up a larger portion of police time. In rural communities, police deal with a huge number of unique, one-off tasks.
What remains true in each of these cases is that police officers aren’t primarily crime fighters and law enforcers; instead, they fill a huge range of other social functions, often ones that other social services and institutions don’t have the ability to respond to quickly or at all.
“As a society we’ve decided to sweep these problems aside rather than to deal with them,” Friedman tells me. “And the police are the broom. They don’t want to be the broom, but that’s exactly what they are.”
The job we prepare police for
This all adds up to a fundamental problem with policing in America: We prepare police for a role vastly different from the one they actually play in society.
A 2016 national study of the training of 135,000 recruits across 664 local police academies found that, on average, officers each received 168 hours of training in firearm skills, self-defense, and use of force out of 840 total hours. Another 42 hours were spent on criminal investigations, 38 on operating an emergency vehicle, 86 on legal education aimed primarily at force amendment law, and hundreds more on basic operations and self-improvement. Topics like domestic violence (13 hours), mental illness (10 hours), and mediation and conflict management (9 hours) received a fraction of trainee time. Others, like homelessness and substance abuse, were so rare they didn’t make the data set.
Those averages mask an even more worrying reality. Almost half of American police academies utilize what is called the “military model” of instruction — a high-stress, physically and psychologically excruciating approach traditionally used to train soldiers for battle. Another third use a hybrid approach that draws heavily on the military model.
In many major-city police departments where this military model is prevalent, training is even more skewed toward force and law enforcement. At Nashville’s police academy, for instance, officers spent two-thirds of their training time on law enforcement and use of force and less than 10 percent of their time on “social work/mediation” issues like interpersonal communication and human relations.
“The amount of firearms and use of force training in our academies is completely at odds with the problem we most often ask police to deal with,” says Ratcliffe, the former UK police officer turned Temple criminologist. “Police training is simply not reflective of the role of police in our society.”
In the field, this trend continues. Despite the fact that American police deal with a vast array of different situations, they are equipped with the exact same tools for each one: handcuffs and a firearm. Increasingly, that tool basket also includes assault rifles, camouflage, and armored vehicles, even for routine tasks.
The structure of police agencies, too, reflects a commitment to force. Glance at the organization chart of any major police department and you’ll see specialized departments like SWAT, bomb squad, narcotics, vice, street crimes, gang unit, criminal intelligence, and counterterrorism. What you won’t see, with a handful of exceptions, are departments focused on conflict mediation or social work.
The emphasis on force, law, and crime fighting is undergirded by a powerful ideological ecosystem. As my colleague Zack Beauchamp writes, “The ideology [of policing] holds that the world is a profoundly dangerous place: Officers are conditioned to see themselves as constantly in danger and that the only way to guarantee survival is to dominate the citizens they’re supposed to protect.” That ideology is baked into the culture of policing at all levels.
Crime fighting and deployment of force are also culturally valorized. Take the International Association of Chiefs of Police’s “Police Officer of the Year” award, which “symbolizes the highest level of achievement among police officers,” and selects those who can stand as models for the profession — it’s a big deal in the policing world. In the 30-year period from 1986 to 2015, 25 recipients of the award were honored for actions they took in combat conditions while under attack.
Or just look up any police department recruitment video, where you’re likely to see police officers battering down doors, firing assault rifles, engaging in high-speed freeway chases, and running after suspects through alleyways — sometimes with a few brief shots of community outreach sprinkled in.
As for in-person recruiting efforts, police agencies concentrate primarily on military bases and, to a lesser degree, sports facilities and private security companies. The result is that military veterans — who are more likely to generate excessive force complaints and be involved in unjustified police shootings than non-military cops — represent almost 20 percent of police officers despite being just six percent of the US population. Men more generally make up almost 90 percent of all police officers; they are considerably more likely to use force and aggressive tactics than female officers.
“What excites police is action, and that means ultimately applying violence,” says Rizer. “The people attracted to police work want that type of action — they are giddy about it. The people who don’t want that type of action either never make it in the first place or are ridiculed for it if they do.”
A mismatch with devastating consequences
Police officers are functionally generalists responsible for dealing with a vast array of our society’s most sensitive situations; yet we’ve recruited, hired, trained, equipped, and deployed them to be specialists in force. And we’ve done it all using an often disproportionately white police force with a well-documented racial bias problem entering Black and brown communities that historically distrust the police.
Would it surprise anyone if this occasionally resulted in unnecessary violence?
“Often what these situations require is someone to calm things down, cool things off, and deescalate,” says Tom Tyler, a legal scholar at Yale Law School and a founding director of Yale’s Justice Collaboratory. “But police tend to manage all the problems they face through the threat or use of coercive force. This amplifies the level of emotion and anger in a given situation and can create a spiral of conflict that ends tragically.”
Take the case of Rayshard Brooks. On June 12, Atlanta police officers were sent to respond to a complaint that Brooks was sleeping in his vehicle in a Wendy’s drive-through. Video evidence shows the interaction starts out calm. Brooks repeatedly asks the arresting officer, Garrett Rolfe, if he can leave his car parked and walk to his sister’s home, which he says is nearby. But Rolfe insists Brooks take a field sobriety test, which reveals that Brooks had a blood alcohol level slightly above the legal limit. Rolfe attempts to handcuff Brooks, Brooks resists, and a struggle ensues. Brooks grabs Rolfe’s Taser, begins running away, and turns to fire it. Rolfe shoots Brooks three times.
Brooks died in the hospital.
There are numerous points at which this interaction could have gone differently. If Atlanta had delegated certain responsibilities to non-police agencies, they could have sent an unarmed civilian to drive Brooks home. If the officers on the scene had the mindset of solving a problem without the use of force, they probably wouldn’t have escalated the situation by trying to forcefully handcuff Brooks. If the arresting officer didn’t have a Taser, Brooks would never have taken control of his weapon. If that same officer weren’t armed — or perhaps had stricter use of force requirements — he wouldn’t have shot and killed someone holding a less lethal weapon.
You can do the same kind of analysis for the deaths of George Floyd, Eric Garner, Breonna Taylor, Philando Castile, Euree Martin, Tony Timpa, Erik Salgado, and countless others. In each situation, the mismatch is crystal clear: Officers trained primarily in the deployment of force and law, armed with lethal weapons, and told to think of themselves as warriors were the chosen first responders to situations that demand anything but. And each situation ended with someone killed at the hands of the people ostensibly tasked to protect and serve them.
Police killings of unarmed civilians in the United States are magnitudes higher than those in peer countries. Using 2015 data, Franklin Zimring, a UC Berkeley criminologist and author of When Police Kill, calculates that the chance of an unarmed civilian being killed by police in the US is three times higher than the chance of any civilian, armed or unarmed, being killed by police in Germany and more than 10 times higher than in the UK (and that’s using a very conservative estimate of unarmed shootings in the US). A separate analysis found that in almost half of police killings of unarmed civilians in the US, the person killed was revealed to be or suspected of experiencing either a mental health crisis or narcotic intoxication.
Even when civilians are armed, that doesn’t necessarily mean police killings are justified. Upon extensively analyzing the 1,100 total fatal police killings in the US in 2015, Zimring concluded that “almost half the cases … were confrontations where the police were not at objective risk of a deadly attack.” And, of course, it is impossible to quantify how many of those confrontations would not have escalated to the point of potential violence in the first place if not for police presence and tactics.
The unnecessary use of deadly force isn’t the only, or even the most likely, consequence of this mismatch. It also leads routinely to the overcriminalization of issues like drug use, mental illness, and homelessness; it causes predominantly Black and brown communities to live in constant fear of their own police departments; it destroys trust between police officers and the people they are supposed to protect; and it places a major financial burden on local government budgets (armed police officers are an expensive way to address social problems) that leads to the underfunding of key social services. All the while, it fails to solve the underlying problems that lead to police being called in the first place.
“The definition of failure is that what we’re doing isn’t solving the problem and is actually causing harm in the process,” says Friedman, the Policing Project director. “That basically describes the state of policing today.”
Reimagining public safety
When it comes to addressing the mismatch between the nature of our police forces and the roles we ask them to perform, there are two broad paths that stand out.
The first is to transform our police forces — to change how officers are recruited, hired, trained, and equipped to meet the actual demands of their role.
Hiring and recruiting practices can be reformed to increase the diversity of police forces in terms of gender, race, and non-military backgrounds. Training can be refocused to include a stronger emphasis on procedural justice principles, conflict deescalation, and crisis intervention. Use of force policies can be made much stricter. Tactics like chokeholds, shooting at moving vehicles, and shooting without warning can be banned, as many departments have already done. Military-grade weapons can be taken off the streets. Legal protections like qualified immunity can be revoked.
On a structural level, police agencies can create an entire department focused on crisis response with specialized units focused on homeless outreach, mental illness, substance abuse, and conflict mediation (as some progressive departments have already done). Those officers can be recruited from fields like social work and psychology, hired based on their capacity to calmly handle highly stressful situations, trained primarily in crisis response, and rewarded not for arrests or stops but for peaceably resolving issues and handing them over to the appropriate social services institution.
The challenges associated with this approach aren’t difficult to imagine. Reform would have to take place on numerous levels: training, hiring, recruitment, agency structure, weaponry. You’d have to get buy-in not only from state and local public officials and police chiefs but from rank-and-file officers. You’d have to fight police unions for even an inch of reform. And even if you fixed one or two of these areas (which could take years or decades), sending armed officers to deal with social problems will always leave open the possibility of unnecessary violence. Cities like Minneapolis, Atlanta, and Tucson — all of which have experienced high-profile police killings recently despite reform efforts — have learned that lesson the hard way.
“It’s impossible to point to one specific problem and say, ‘That’s it — that’s the issue,’” says Tracey Meares, a legal scholar and founding director of Yale University’s Justice Collaboratory. “This is about the system of policing itself. Our communities lack the resources to deal with their social problems. And our response has been to deploy armed first responders to address the issue way down the chain from the source.”
That leads us to a second approach: to transform how we address public safety such that police play a smaller, more targeted role altogether. This would involve communities designating a certain subset of current police duties that don’t require armed police response, delegating those responsibilities — along with requisite funding — to an institution that could better handle the issue, and designing systems for service delivery (like a 911 call diversion program) and coordination (like a silent alert system that unarmed first responders could use to quickly summon police backup).
Models for this approach have been implemented successfully in some places in the US and across the globe. In the UK, certain traffic functions have been designated to unarmed, non-police public servants. In cities across the US, “violence interruption” programs run by community nonprofits have been largely successful in mediating conflict and reducing violence. The much-applauded Cahoots program in Eugene, Oregon, sends a team of unarmed crisis specialists to address many non-criminal 911 calls without having to involve police.
There’s public support for such an approach. A recent poll found that 68 percent of voters support the creation of a “new agency of first responders” (although just a quarter of Americans say they support “reducing funding” for police departments).
The challenge is that designing an entirely new approach to public safety, rather than merely reforming an existing one, means stepping into relatively uncharted territory.
“There is no single, definitive answer to what will work in a given place,” Megan Quattlebaum, director of the Council of State Governments Justice Center, tells me. “Anything we do is going to be in the space of experimentation with different models.”
That means things are bound to go wrong. Some programs might not scale. Others will not receive adequate funding. Crime may temporarily increase in some places. Occasionally, a violence interrupter or mobile crisis worker will be seriously injured or killed. And when those things happen, it will take an incredible amount of political will and community solidarity to persist.
These two approaches are not mutually exclusive. There is general agreement that armed officers should still respond to violent crimes, like an active shooter, and definitively non-criminal, nonviolent activity should be delegated to alternative institutions. There are also a handful of hybrid solutions that combine the approaches — for instance, collaborative models between police and other agencies or nonprofits that co-respond to issues like homelessness or mental health. Or the “civilianization” of police departments: hiring unarmed professionals without arrest powers to fulfill certain police responsibilities, as many European countries have done.
But once you get into the details, difficult trade-offs emerge. There are plenty of cases where there is legitimate ambiguity about whether a situation will escalate to violence: like when a 911 caller isn’t sure whether what she is seeing is a man at a playground with a lethal weapon or a young teenager playing with a toy gun, or when a woman experiencing a severe mental health crisis is threatening others with a knife. In cases like those, do we send unarmed first responders and risk putting them, and others, in harm’s way? Or do we send armed police officers and risk the use of unnecessary state force against civilians?
“This is a conversation that needs to be had with communities,” says Tracie L. Keesee, a former Denver police officer and the co-founder of the Center for Policing Equity. “Where do you want police and where do you not want them? Who would you rather have show up? What kinds of qualities would you like your police officers to have?”
Reimagining the role police play in our society is far from being anti-police. Plenty of police officers recognize that our current one-size-fits-all approach to public safety is fundamentally broken. They lament the fact that we ask police to solve far too many of our social problems and don’t give them the training or resources they need to do so — and then point the finger at them when they inevitably come up short.
“The reason I think we need to rethink policing is because I care about police,” says Rizer, the former officer and R Street researcher. “I want to make policing prestigious again — not the prestige of power, but the prestige of respect. But in order to do that, we need to stop underfunding everything else and leaving the police holding a bag of shit.”
Support Vox’s explanatory journalism
Every day at Vox, we aim to answer your most important questions and provide you, and our audience around the world, with information that has the power to save lives. Our mission has never been more vital than it is in this moment: to empower you through understanding. Vox’s work is reaching more people than ever, but our distinctive brand of explanatory journalism takes resources — particularly during a pandemic and an economic downturn. Your financial contribution will not constitute a donation, but it will enable our staff to continue to offer free articles, videos, and podcasts at the quality and volume that this moment requires. Please consider making a contribution to Vox today.