(A note on terminology. The term “professional” here is slightly technical and refers not to a person who performs an activity for compensation, but rather an activity guided by specialized knowledge, discretionary decision-making, and assessed, in part, by the professional judgement of others. Professions are distinct from occupations on this use of the term. “Legitimacy” here refers to the permissible use of political power. Typically, political power must be democratically authorized to be legitimate, though as we’ll see, policing complicates this requirement.)

Policing, especially in the United States, has a sordid past. U.S. police agencies in the south grew out of slave patrols. In the north they grew out of neighborhood watch groups or private agencies that were used to break labor strikes. In cities, policing was at first largely a civilian affair. As the country urbanized, police agencies were formalized, and police services were brought under state control. Officers began to wear uniforms and badges and carry clubs and revolvers. Despite the new veneer of respectability, those agencies were often seriously corrupt and played a major role in political patronage systems with jobs going to ethnic groups in return for electoral support of local candidates. Eventually, policing underwent reforms aimed at professionalizing officers and their agencies.

“In some ways, professionalization and the military-inspired hierarchy were an improvement. In others, they left a legacy that encourages bad policing and bad thinking about the police.”

This entailed isolating them to prevent the rampant corruption associated with political patronage. Professionalization meant adopting new technology. So-called “father of modern policing” August Vollmer introduced bicycles, the police cruiser, the radio, fingerprinting, and other technologies while in charge of the Berkeley Police Department. Professionalization also meant adopting a military-inspired hierarchy, in no small part because Vollmer and other reformers had military backgrounds (Coyne and Hall 2018, 102, 145). Minneapolis’s police department adopted a military-hierarchy in the 1880s, but the civilian influence was strong enough that the city abandoned that reform shortly after (Fogelson 1977, 16). In time, however, police reformers and administrators almost unanimously adopted a military analogy for the police, with then-NYPD Commissioner Teddy Roosevelt calling the police a “quasi-military” force (Fogelson 1977, 54). In some ways, professionalization and the military-inspired hierarchy were an improvement. In others, they left a legacy that encourages bad policing and bad thinking about the police.

One problem with the military analogy and the hierarchy that it generated is that the nature of military and police work is importantly different. That both are asked to enforce political decisions with violence and coercion or its threat can be misleading. One obvious difference is that policing ought not to be as aggressive as waging a war; police are often called upon to defuse situations, not crush combatants. That difference underlies the complaints discussed above. Another important, though less obvious difference, concerns the appropriate scope of discretion in the two roles.

Soldiers are typically allowed very little discretion in the course of their work. Policing, especially in the patrol division, is characterized by significant amounts of discretion. The military is asked to defeat a threat; military activity is defensive. That task is broken down into components and delegated down the hierarchy. This requires extensive specialization, knowledge, and skill. But the task is a discrete one, and this seriously narrows the scope of discretionary judgement in the military. Police, on the other hand, have a role in governing. They are asked to maintain order, to deter crime, and more generally to help create a certain kind of society. This is a continuous, ongoing task, constrained by a wider variety of normative considerations. Of course, some military activity, such as when military forces were kept in Iraq and Afghanistan long after the initial conflicts, or the deployment of military forces to quell riots, shades into policing when its goal is order maintenance. Though the boundaries are vague, paradigmatic police and military activity differ in that the former helps to govern and the latter defends against foreign threats.

Embed from Getty Images

One interesting thing to note here is the relationship between one’s place in the hierarchy and the amount of discretion that characterizes one’s role. Military activity, like most kinds of work, allows for discretion to track one’s place in the hierarchy. Commanding officers have more discretion in carrying out a mission than do enlisted soldiers. Police activity inverts the typical relationship of hierarchy and discretion. The patrol officer at the bottom of the hierarchy has more discretion than those further up the hierarchy. Detectives have little discretion because their work responds to a well-defined and reported crime. While detectives exercise discretionary judgment about how to investigate a crime, they lack the discretion that patrol officers have about when and whether to initiate police activity. Administrators at the top of the hierarchy are under much tighter political control.

The patrol officer’s discretion is in many ways simply a result of the practical realities of policing. Administrators can’t follow officers into the field. Nor are the criminal codes officers are asked to enforce fully determinate; what counts as loitering or disorderly conduct requires police judgement, as does determining how far above the speed limit is grounds for a traffic stop and ticket. Even when the code is determinate, as for (e.g.) drug laws, officers need to use considerations of justice and efficiency in deciding when and against whom to enforce drug laws.

“Police activity inverts the typical relationship of hierarchy and discretion. The patrol officer at the bottom of the hierarchy has more discretion than those further up the hierarchy.”

This difference is significant for thinking about legitimacy. Discretion in the military tracks the democratic process for legitimating military force. Voters elect politicians who make decisions about military force; as long as the rules of jus in bello are met, voters tend not to have strong preferences about military strategy; generals draw up strategies to best achieve military goals; commanders further down the hierarchy issue orders to soldiers; soldiers carry them out, sometimes in the dark about the larger strategy. Legitimate military activity is large-scale collective action to achieve a discrete goal authorized by democracies.

Discretion in policing, on the other hand, breaks the flow of legitimacy down the hierarchy. Voters elect politicians who create and pass legislation or who staff enforcement agencies and make high-level decisions about enforcement priorities; police administrators clarify agency expectations through systems like CompStat; detectives and patrol officers answer calls for service and engage in either investigation or patrol. Once we get below the command level in which administrators make decisions about how to allocate their resources, patrol officers largely determine how they will spend their time. These decisions almost always involve politically charged questions about when and how to enforce the law.

But there is very little in the way of democratic oversight over these decisions. To use one example, one officer in Baltimore frequently arrested cyclists riding at night without a light. If they had identification, they were written a citation; if not, they were arrested, and their bike confiscated. One purported motivation for the practice was that drug “runners” and “lookouts” often use bikes a night. Another was to give the administrators “stats” and look productive (Moskos 2008, 140). I imagine most people would find this kind of policing objectionable, especially considering that for many, their bicycle is their main form of transportation. Further, this instrumental enforcement does nothing to disrupt the drug trade. Some of the officers’ colleagues found the strategy objectionable, though administrators approved of it. The point of this example is that the criminal code furnishes police officers with tools but legitimizing the criminal code doesn’t obviously amount to legitimizing the particular uses of those tools. Of course, illegitimate laws allow for illegitimate policing. But a generally legitimate criminal code doesn’t ensure legitimate policing.

Embed from Getty Images

This claim generalizes. There is no practical way for voters to authorize, or withhold authorization from, many on-the-ground decisions about policing. This is true of military activity to an extent, but if you accept the foregoing argument about the difference between military and police activity, democratic authorization of the decision to go to war carries further down the military hierarchy than the democratic authorization of a criminal code carries down the police hierarchy. And whereas the decisions left up to soldiers are significantly constrained by rules of engagement crafted for the situation, police administrators can’t similarly constrain discretionary police decisions. The law requiring cyclists to have lights on their bike can be legitimately enforced in some circumstances; guidelines from administrators can’t fully determine what those circumstances are any more than the criminal code can. Thus, there are special questions about the nature of legitimate law enforcement that go beyond the question of the criminal code’s legitimacy or whether an arrest respects procedural rights.

Political philosophy, therefore, needs a theory of legitimate police discretion. The military analogy that generated the military-inspired hierarchy for police agencies obscures this fact. It encourages us to ignore the inevitability of discretion in favor of a conception of policing in which police simply enforce democratic decisions. In turn, that encourages a skepticism about critically evaluating police performance. Even police officers tend not to criticize the performance of other officers, typically because “they weren’t there.” Professionals can and should criticize the work of other professionals. Discretionary judgement can and should be honed through critical evaluation by other professionals.

Reflecting on the military disanalogy fixes the conceptual problem by encouraging us to re-conceive of police officers as genuine professionals who should be exercising discretion by drawing on a body of specialized knowledge. It encourages us to develop an evaluative framework in light of that. It also motivates a practical solution: reforming the organizational structure of police agencies to support that kind of role. One way to achieve that is to specialize the patrol division. In some ways this will make policing more like the military, given the extensive specialization of soldiers. But the crucial difference is that specializing the patrol division is aimed at improving, rather than limiting, what I take to be inevitable police discretion.

“There is no practical way for voters to authorize, or withhold authorization from, many on-the-ground decisions about policing.“

As things are now, patrol officers spend the bulk of their time responding to 911 calls that are frequently, in the words of a former officer, bullshit (Moskos 2008, 97). They are also mostly generalists. There is a kind of zero-sum problem with police training. Increasing the amount of training an officer requires to respond to (e.g.) calls involving mental health crises amounts to decreasing training in other areas. Professional discretion requires narrow expertise; as in other professions, specialization is required to develop this expertise. It is worth emphasizing that I am not suggesting simply that more training would solve the problems of American policing. Rather, reforming organizational structure will allow for fundamentally different kinds of officers and police agencies. Just as SWAT teams help to select for people who want to throw smoke grenades through a suspected drug dealer’s window, specialized teams within patrol divisions could select for people more interested in protecting and serving than fighting crime or arresting people for riding their bike without a light.

Of course, saying anything substantive about what legitimate discretion looks like, or how the structure of police agencies ought to be reformed, is a large project. My goal in this post has been to make the case for why that project is worth taking up by offering some reasons for thinking that the military analogy has predictably generated some of the problems in policing that are once again at the front of conversations about criminal justice. Not only has it generated over-used SWAT teams and riot squads staffed by police officers pretending to be Rambo but, more fundamentally, it has misled us about the right questions to ask about legitimate policing.

Coyne, Christopher J., and Abigail R. Hall. 2018. Tyranny Comes Home: The Domestic Fate of U.S. Militarism. Stanford University Press.

Fogelson, Robert M. 1977. Big-City Police. Harvard University Press.

Moskos, Peter. 2008. Cop in the Hood: My Year Policing Baltimore’s Eastern District. Princeton University Press.