Policing and Economic Calculation

Police departments are bureaucracies tasked with enforcing the laws within a jurisdiction and responding to requests for service. They have limited resources to do so and many competing demands for how those resources are allocated. Since individuals pay a zero price for calls for police service, there is greater quantity demanded than supplied. As a result, police have to find ways of rationing resources, leaving many demands unsatisfied. Because of their wide mandate and lack of a reliable measure of output in terms of consumer satisfaction, it is impossible to determine whether police resources are allocated to their most highly valued uses. Even though police services are nominally to be provided equally to everyone within a given jurisdiction, the reality of scarcity keeps such a goal from being obtained. The decisions police make on how to allocate resources are therefore necessarily controversial.

In recent years in the United States, well-publicized killings by police of minority citizens have brought greater attention to the issue of police conduct, particularly regarding minority communities and the use of force. Some have argued that more aggressive policing tactics, such as the “broken windows” style adopted by the New York City Police Department in the early 1990s, have reduced crime (Mac Donald 2016). Consequently, they conclude that it is important to have measures in place that protect police officers from being punished for doing their jobs in good faith so that they are not hesitant to engage in proactive policing. Others argue that some of these tactics, such as stop-and-frisk policies that are mostly carried out on law-abiding citizens, strain relationships between police and the community and delegitimize the police (White and Fradella 2016). Similarly, there are controversies over the level of employee protections that should be afforded to police officers. Organizations such as Check the Police (Check the Police 2016), an outgrowth of the Black Lives Matter movement, argue that these protections protect officers that engage in misconduct and should all be abolished. These protections also serve as a form of compensation to police officers, reducing the amount taxpayers have to spend on this public service (Fegley 2020). To maintain the same level of service and reduce the level of protections police officers enjoy would require that taxpayers pay for higher wages. It is unclear how much greater of a burden taxpayers are willing to bear in order to reduce police misconduct.

The purpose of this article is to show how these controversies in contemporary policing are fundamentally due to the inability of police departments to engage in economic calculation. Since law enforcement agencies are provided bureaucratically and without market prices, they are unable to engage in economic calculation, leading to an inability to rationally allocate police resources to their most highly valued uses. This issue manifests itself in controversies over how police resources should be allocated, what constitutional protections individuals should enjoy from unreasonable searches and seizures, what the rules governing an officer’s use of force should be, and what employment protections police ought to have. It is argued that these problems are inherent to the institution of state-provided policing and can only be ameliorated when security services are provided through competitive markets.

This analysis contributes to three strands of literature. The first is the literature on evaluating police performance. There is general agreement that police departments have focused their attention on what is measured: crime rates, arrests, citations issued, response times (Bayley 1994; Chandek 2000; Cordner 2014; E. Ostrom 1971, 1973). These measures have been criticized on a number of grounds, such as their validity (E. Ostrom 1971), their measurement of outcomes largely outside of police control (Allen and Maxfield 1983), and their failure to accurately reflect what police spend most of their time doing (Oettmeier and Wycoff 1999). As an alternative, scholars have proposed systems that measure performance based on multiple dimensions (Davis et al. 2010; Maguire 2005; Milligan, Fridell, and Taylor 2006; Whitaker et al. 1982), though very few departments have implemented such systems (Uchida 2014). Some have explicitly contrasted the methods of measurement used by police bureaucracies and for-profit firms (Cordner 2014; Mises 2007; E. Ostrom 1973). However, they treat the inability of policing organizations to engage in economic calculation as inherent to policing rather than as a result of the institutional framework in which it is provided.^[1] This article fills a gap in this literature by explaining how the ability for policing to be subject to calculation is determined by whether the outcomes of policing can be evaluated through a consumer’s willingness to pay for them or by marginal contributions to a firm’s revenue or capital value. As elaborated upon below, this depends on the institutional arrangements in which security is provided.

A second strand of literature to which this analysis contributes is that on the epistemic properties of institutional arrangements and their relationship to governance (Boettke, Tarko, and Aligica 2017; Hayek 1945, 1989; V. Ostrom 1993). According to Aligica, Boettke, and Tarko (2019, 17–18), the fundamental issue in the relationship between public choice, collective action, and the scope of the public domain is that public administration begins where economic calculation ends. To the extent that the provision of policing is subject to public administration, profit-and-loss accounting cannot use market-generated prices. There is no possibility of aggregating individual preferences into a social welfare function that can be used as an uncontroversial basis for public policy. The problem this presents for public administration may be less of an issue for tasks with more clearly assessable goals and that can be performed according to detailed rules that constrain the discretion of the bureaucrat. This is not the case for the current policing role, which includes a wide range of tasks (Bittner 1970) requiring a considerable degree of discretion (Breitel 1960). Thus, if a public police department is to be subject to strict rules, its role must necessarily be narrow. It cannot both have a wide mandate and be effectively bureaucratically administered.

The third strand of literature to which this article contributes is that on constitutional enforcement under different institutional arrangements. Coyne (2018) analyzes how the operations of a constitutionally limited “protective state,” which includes interpreting the appropriate role of the protective state, can eventually undermine the liberties they are meant to uphold. Leeson (2011) argues that a system of clubs has institutional features that encourage constitutional compliance that a system of government lacks. One of these features is the actual, rather than implicit, agreement to the provisions of the constitutional contract. Voluntarily joining a club, rather than being born into a constitutional contract, is a way of achieving this. Further, the “constitutional moment” of a club is not necessarily the result of the deliberations of the governed, as in The Calculus of Consent (Buchanan and Tullock 1962), but can be the result of the entrepreneurial judgment of developers selling memberships in constitutionally governed residential communities (Boudreaux and Holcombe 1989). This article explores the implications of determining the constraints on security providers by competition rather than through the constitutional compliance of a monopoly provider.

The article proceeds as follows. Section 2 discusses the theoretical framework of economic calculation and how it applies to policing. Section 3 argues that a number of issues in modern policing—including controversies over what constitutional protections individuals should enjoy from unreasonable searches and seizures, what the rules governing an officer’s use of force should be, and what employment protections police ought to have—are due to police bureaucracies’ inability to engage in economic calculation and how these problems could be resolved if police agencies were able to engage in calculation.

THEORETICAL FRAMEWORK

Bureaucratic Management versus Profit-and-Loss Management

The theoretical framework applied here relies on Ludwig von Mises’s analysis regarding the feasibility of economic calculation in the socialist commonwealth (Mises 1949, 1990). Mises argued that socialism—that is, an advanced industrial economy based on the extended division of labor and state ownership and direction of the means of production—is impossible. Mises (2007) extended this analysis to bureaucratic organizations operating within a market economy. While a bureaucracy such as a police department is able to use market prices to calculate the cost of its inputs (such as labor, vehicles, radios, weapons, and other capital goods) because they are bought on the market, it is unable to calculate any market value for its output. Since policing services are provided free of charge and a department’s revenue comes primarily from taxes, it cannot be determined whether a chosen allocation of resources is more highly valued than alternative uses of the money used to pay for them. This means that not only are police departments unable to determine the socially preferred amount of resources to be allocated to policing generally but they also lack the ability to measure their performance in terms of consumers’ evaluations of their output.

The primary goal of a police department is to produce public safety. The production process for this output primarily consists of responding to calls for service and, when an officer’s time is not otherwise allocated, engaging in vehicular patrol while waiting for calls for service (Frank, Brandl, and Watkins 1997; Parks et al. 1999; Scott 2000; Sherman 1983; Smith, Novak, and Frank 2001). Police departments have worked under the assumptions that increased officer visibility on the streets deters crime and that having officers patrol widely in vehicles (rather than on foot) allows for rapid response to calls for service and thus can aid in dealing with crimes in progress. Police departments also allocate a substantial amount of resources to investigating crimes after they occur in order to catch offenders and deter crime. While much of police activity is reactive in that it involves waiting for calls for service and crimes to occur, some of it is proactive, particularly in dealing with “victimless” crimes that are unlikely to be reported, such as drug trafficking and prostitution. In deciding how to allocate police resources among these various uses, police have no objective means of evaluating the value of marginal returns among them and there are too many infractions for police to be able to address all of them. Therefore, they must find some way of rationing resources and deciding how to allocate them among various uses.

One of the primary means by which they allocate resources involves a first-come-first-served principle. For the most part, police, when not otherwise engaged in patrol, respond to calls for service as they receive them. This leads to the next decisions they have to make, which are how to respond to these calls for service and how much of their resources to allocate to each call. Since police are a common-pool resource for which a zero price is charged to call for service, they cannot prioritize calls based on the client’s willingness to pay (Benson 1994).

There are a number of ways in which police have decided how to ration resources in response to calls for service. The Leicestershire Police, for example, could not respond to all the burglary reports they received, so they decided to only respond to calls from houses with even numbers (Evans, Kirk, and McCann 2018). The London Metropolitan Police have decided as a matter of policy to not pursue property crimes such as vandalism, vehicle crime, and fuel theft if the cost of the loss is less than fifty pounds (Hamilton 2018). These actions free up officer time for other uses, but without having some means of comparing the value of these alternatives, police are in the dark about whether these alternative pursuits are worth forgoing the investigation of burglaries and petty theft.

Since they are unable to engage in profit-and-loss calculation, they must find other ways of evaluating their performance. Police have traditionally focused on measures like crime rates, clearance rates, and arrests. These figures are deemed appropriate because the police are agents of crime control, and they are collected as a matter of course. Elinor Ostrom (1971, 1973) and Parks (1971), however, describe the limitations of such figures as measures of police performance: since many factors affect reported crime, clearance rates may be more of a function of the type of crime than police efforts,^[2] and arrests do not necessarily measure the ultimate end of public safety. Much of the research attempting to measure the effectiveness of police strategies (see Sherman et al. 1998) focuses on how they affect rates of reported crime. Even if these studies are methodologically sound and are fully able to determine the efficacy of different strategies in reducing crime, police still have no accurate way of determining whether the costs of allocating additional resources into a certain strategy are worth the benefit of crime reduction. Some studies of police performance—such as E. Ostrom, Parks, and Whitaker (1973) and E. Ostrom and Whitaker (1973, 1974)—attempt to account for costs and the value of policing services by measuring both citizen satisfaction and the cost per citizen in comparable jurisdictions served either by large metropolitan police agencies or by small, neighborhood agencies. But citizen surveys do not overcome the calculation problem, because even if citizens judge the police services they receive to be preferable to those in a comparable neighborhood and the cost per person is lower, this does not mean that police allocated resources to uses more valued than the alternative ways in which taxpayers would have used them. Ideally, public goods would be provided by a level of government corresponding to the scale of the externality generated by their provision (V. Ostrom, Tiebout, and Warren 1961) and financed according to the benefit principle (that individual consumers pay according to the benefit they receive from the good). Absent the institutions that allow individuals to demonstrate their preference for such a good to alternative uses of their individual contribution, whether they actually prefer that such a good be provided is undetermined (Rothbard 1956).^[3]

However, under certain institutional conditions that allow for voluntary exchange, the provision of security can be subject to economic calculation—that is, when it is part of a production process in which the value of output can be measured either ordinally against the cost of producing or purchasing it (the evaluator of the output also is able to evaluate alternative uses of the inputs) or cardinally (the purchaser of security services can compare the cost of production with its contribution to revenue, losses prevented, or changes in capital value of the asset protected).^[4] Examples of the former include individuals purchasing security systems for their household, LoJack for their vehicle, or pepper spray for personal protection and participating in neighborhood watch. In these cases, it is possible to determine whether there is a “psychic profit” because the individual weighs the subjective value of the safety provided by the good or service against the opportunity cost of purchasing or producing it. Examples of the latter include commercial and industrial enterprises purchasing contract security and providing security in-house. Clients of contract security agencies include recreational facilities, hospitals, community colleges, universities, office buildings, warehouses, industrial plants, shopping centers, transport companies, financial institutions, construction companies, apartment complexes, hotels, private homes, computer companies, and insurance companies (Business Round Table 1994; Jones and Newburn 1998; Kandt 1974; Shearing, Farnell, and Stenning 1980). These organizations are able to compare the cost of security with a benefit whose value can be measured monetarily, whether it is additional revenue, higher capital value, or losses prevented. As such, under such institutional arrangements, economic calculation is possible.

It should be noted that the relative size of the security industry, as compared to public policing, is significant. According to the US Census Bureau’s 2016 County Business Patterns and Nonemployer Statistics datasets, there were 142,093 business establishments employing 1,049,451 individuals providing investigation and security services.^[5] This, however, only includes firms that specialize in providing these services and not firms that produce them internally. By comparison, the Bureau of Justice Statistics estimates that in 2016 there were 701,169 full-time sworn officers among 15,328 general-purpose law enforcement agencies (i.e., municipal, county, and regional police departments, most sheriff’s offices, and primary state and highway patrol agencies) (Hyland 2018). Shearing and Stenning (1981, 1983) attribute the growth in private security relative to public policing^[6] to the rise of mass private property in which economies of scale of manned security can be exploited and which are areas unlikely to be monitored by public police. To the extent that alternative property arrangements can internalize the benefits of policing to prevent free riding, nonexcludability issues associated with policing are mitigated.^[7]

What about Tiebout Competition?

In the model presented in Tiebout (1956)—where consumer-voters can costlessly move to jurisdictions that provide their most preferred bundle of services at a price they are willing to pay, they have full knowledge of differences in revenue and expenditure patterns, and communities are the optimum size for the provision of those services—municipal providers are able to engage in economic calculation. The fiscal equivalence and benefit principles of public finance are satisfied. There are, however, a number of reasons to doubt that the conditions in the real world closely enough resemble the assumptions of the Tiebout model such that municipal revenue serves as a reliable feedback mechanism for how well police perform in satisfying consumer-voter preferences.

Boettke, Lemke, and Palagashvili (2016) argue that federal subsidies to local police departments have distorted police priorities away from the demands of community residents and toward federal initiatives. This includes incentivizing the allocation of more resources to drug enforcement through the civil asset forfeiture provisions contained in the Comprehensive Crime Act of 1984 (Benson, Rasmussen, and Sollars 1995), which allows local law enforcement to bypass state restrictions on how the forfeiture proceeds are spent. It also includes the militarization of the police to a greater extent than would have otherwise occurred, through initiatives such as the 1033 Program, through which the federal government provides law enforcement agencies with “body armor, aircraft, armored vehicles, weapons, riot gear, watercraft, and surveillance equipment” (Hall and Coyne 2013, 497). By softening budget constraints and moving the fiscal attention of local law enforcement to federal priorities, federal grants and other programs undermine Tiebout competition (Boettke, Palagashvili, and Piano 2017). Leeson (2011, 305) describes two limitations to the Tiebout mechanism. The first is that, in a federal system, the ability to develop additional subgovernments is limited, both in terms of number and variety, which undermines the competitiveness and assertiveness of governments compared to a system of clubs. The second is that individuals’ constitutional contract with the central government is not self-enforcing. In the US, despite the fiscal disadvantages associated with the creation of clubs, such as community associations, as compared to subgovernments, such as municipalities,^[8] about half of the new housing built between 1980 and 2000 was subject to the private governance of a community association (Nelson 2009, 345). The fact that, overall, almost 40 percent of local government revenue in 2005 came from the states and federal governments (Nelson 2009, 346) suggests that the perceived advantages to community associations, which do not receive such contributions, as a form of governance must be large. It also suggests that the tie between where people choose to reside and the direct effect that has on local revenue as a consumer-voter feedback mechanism for local government is not as strong as in the Tiebout model.

Calculation Issues in Policing

Search and Seizure and Constitutional Effectiveness

A free society requires limitations on the discretion of security providers to engage in searches and seizures of citizens’ persons, houses, papers, and effects. In the US, federal, state, and local law enforcement are prohibited from engaging in “unreasonable searches and seizures,” as stated in the Fourth Amendment to the US Constitution and similar provisions in state constitutions. Of course, what constitutes an “unreasonable” search or seizure is subject to interpretation. Local law enforcement agencies currently employ surveillance technologies such as automated license plate recognition (which allows law enforcement to track the location data of thousands of cars per minute), stingrays (which simulate cell phone towers, allowing law enforcement to access the communications content and location of cell phones that connect to them), and surveillance drones. The framers of the Fourth Amendment could hardly have anticipated such technologies. While Coyne (2018) is correct in that constitutional interpretation in such an open-ended system can result in antiliberty outcomes, lack of contractual completeness and clarity in constitutional language are not the only issues. If Hasnas (1995) is correct and the possibility of objective rules of justice being applied by judges neutrally capturing the plain meaning immanent within the law is a myth, then an additional underlying issue is the interpretation of law as a state monopoly. Further still, even if there is no dispute over the plain meaning of words in a constitutional contract, the system of constitutionally limited government is not self-enforcing (Leeson 2011).

Part of the reason constitutional restrictions on the ability of police to engage in searches and seizures are not self-enforcing is that there can exist a trade-off between security and less invasive policing techniques. For example, a number of scholars credit the invasive tactics used in the New York City Police Department, such as stop-and-frisk, as playing a crucial role in the trend of decreasing crime in New York City starting in the early 1990s (Bellin 2014; Bratton and Kelling 2015; Mac Donald 2016). Others question whether such tactics, even if they do reduce crime, are worth their costs in terms of the insecurity of individuals in their persons, the indignity of being frisked, the racial disparities in its application, and the strained relationships with the police. There is a conflict in how NYPD resources should be used. The NYPD cannot measure citizens’ relative willingness to pay for crime reduction caused by prevalent stop-and-frisk versus their willingness to pay for an environment where individuals would be less likely to be stopped and frisked at an officer’s discretion.

Enterprises that can engage in profit-and-loss calculation also have to make decisions regarding how invasive their security procedures are (Fegley 2021). Organizations make different decisions than others based on their particular circumstances, and some of these decisions are controversial. For example, a Kroger grocery store in the Atlanta metro area has recently experimented with putting in a new security installation that puts walls around a number of shopping aisles, creating only one entry and exit to them. Customers reported that the installation “feels like shopping in a prison just to buy toiletries or laundry detergent” (CBS46 News 2019). Other shoppers reported that many people use products in the bathroom without paying for them, “but feel a less intimidating approach should be taken.” Similarly, many urban convenience stores employ security measures that some consider unsightly or insulting, such as bulletproof Plexiglas barriers that completely separate the shopkeeper and merchandise from customers. Although individuals may state a preference for less intrusive or imposing security, this raises the question of what keeps a competitor from offering a similar service but without such security. One potential answer is that a profit opportunity is being missed. Another answer is that the costs entailed by less stringent security are greater than consumers’ willingness to pay to compensate for greater risk of loss or personal injury. But this can only be known through a competitive process in which consumers demonstrate their preferences regarding the trade-offs between convenience, security, and price through their buying and abstention from buying. Although there is no formal agreement between the consumer and seller regarding the security procedures to which the latter may subject the former (as there is theorized to be between the citizen and government), there are expectations and tacit understandings regarding what is appropriate and what is not. Such contractual incompleteness may be optimal, because fully specifying the security procedures to which a customer may be subject is costly.^[9] Competition economizes on contractual specification by allowing consumers to exit and patronize another provider offering a preferable bundle. As such, competition serves as a substitute for other means by which constitutional compliance can be encouraged, such as greater ex ante contractual (or constitutional) specification or ex post litigation over what the contract or constitution specified.

The Use of Force

In recent years, the issues of police misconduct and brutality have received increased media attention. Some argue that police officers are too well protected such that they cannot properly be held accountable for brutality. Others argue that such protections are necessary for police officers to effectively do their jobs. One of these protections is qualified immunity, which protects police officers from civil liability “insofar as their conduct does not violate clearly established statutory or constitutional rights of which a reasonable person would have known” (Harlow v. Fitzgerald, 457 U.S. 800 (1982)). Advocates of this protection argue that if police do not have qualified immunity, they will be indecisive in potentially life-threatening situations because they will be concerned about their use of force decisions being scrutinized by juries who do not understand the difficulty of police work. Thus, to ensure that officers will not be afraid to use deadly force in emergency circumstances, protections like qualified immunity are warranted. On the other hand, some consider qualified immunity an unjustified protection for police officers that increases their propensity to use deadly force, even in situations where it is unnecessary. By lowering the costs of making a bad decision, qualified immunity can increase the number of bad decisions that are made.

The trade-off faced in deciding the level of protections provided to police officers tries to minimize excessive force, on one hand, while also minimizing depolicing,^[10] on the other. The inability to properly balance this trade-off stems from the inability of public police to engage in economic calculation. The use of force in the course of policing, for purposes such as maintaining order or apprehending a suspect, has both benefits and costs, and these can only be compared against one another in the contexts where profit-and-loss calculation is possible. For example, a business owner will only want to use force against patrons when the benefits outweigh the costs. Wakefield’s (2003) observations of private police found that when they observed patrons engaging in some disorderly behavior, they would ask them to refrain and warned that refusal to do so would result in their being asked to leave. In most cases, patrons would comply when asked to leave, but in cases where they refused and physical force was required to remove them from the premises, the public police were called. Such outsourcing of physical force, however, does not fully offload the costs of using force.

This is what happened in a Philadelphia Starbucks in 2018. Two African American men had asked to use the restroom and were told it was only for paying customers. After having sat in the Starbucks for some time, they were asked to either buy something or leave. When they did neither, the police were eventually called and the men were arrested. The incident received substantial negative media attention and resulted in the Starbucks CEO personally meeting with the men and closing eight thousand Starbucks locations for a day in order that employees could receive racial bias training (Siegel and Horton 2018). A similar incident occurred in 2017 when United Airlines had overbooked a flight and was unable to persuade enough customers to voluntarily deboard and wait for the next flight. The Chicago Police were called to forcibly remove a man from the flight. The man’s bloodied face was captured on a cell phone video that went viral, resulting in substantial negative press for United (Victor and Stevens 2017). In each of these cases, even though public police carried out the enforcement, they did so at the behest of private businesses that would bear the costs of using force. It is in the interest of business owners to tolerate nuisances until the cost of toleration becomes greater than the cost of using force. In both of these cases, an entrepreneurial error was made: it turned out that the cost of using force was much greater than the cost of bearing the nuisance. If businesses engage in a type of security that uses force in a way that consumers consider illegitimate and overbearing to the extent that they choose to take their business elsewhere, economic calculation alerts the business owner that they have strayed too far into the “excessive force” side of the balance. Conversely, if security is lax and consumers feel unsafe or required to bear an intolerable level of disorderliness such that they discontinue their relationship with the business, losses in revenue suggest that security could be more stringent.^[11] Competition between businesses encourages the survival of only those firms that manage to find the balance consumers most prefer.

But outside of such contexts that enable economic calculation, police are in the dark regarding this trade-off. Public police agencies that engage in excessive force (excessive in the sense of consumer evaluation rather than in a legal sense) will likely not be able to detect losses in revenue due to it, nor will agencies that disengage from proactive policing. The feedback mechanisms that public police receive are muted at best and lack information about trade-offs. To the extent that Tiebout (1956) competition between departments exists, it is inadequate as a knowledge substitute. People move in and out of jurisdictions for a combination of reasons, and the quality of police service is unlikely to be a deciding factor. Even if one lives in a high crime area, due in part to depolicing, such an area tends to be more localized than the jurisdiction of a police department—that is, one can leave the high crime neighborhood without leaving town. As such, “voting with one’s feet” will only be a weak indicator of consumer evaluation of police services.

Complaints and civil rights lawsuits can be used as indicators of excessive force; however, there is less corresponding feedback against the overcorrection of depolicing. While some research has indicated that higher crime is a trade-off of depolicing (Rushin and Edwards 2017), the direct connection between crime and a particular officer’s disengagement is much harder to establish than the connection between complaints and an officer’s proactive policing. From an individual officer’s perspective, unless proactive policing is sufficiently rewarded (for example, through increasing the likelihood of promotion for making high-value arrests), the optimal strategy for an individual officer will be biased toward depolicing. This is suggested by Oliver’s (2019, 68) interviews with police officers about the phenomenon of depolicing. One officer reported:

When I first got on the department, I would see the older officers and think, “Man, they are lazy!” Then I realized the majority of them got in trouble for no fault of their own and were depolicing. They had good intentions, and whether it was a bad decision at the time or someone just flat out lied, they suffered the consequences. . . . The funny thing about police work is that if you don’t do anything—depolicing—you can’t get in trouble. Its only when you do something, whether it be proactive or just regular work, you can get in trouble for it.

Public police agencies’ feedback mechanisms do not reveal whether changes in policy regarding officers’ autonomy in using force bring them closer to or further from the optimal level of use of force. Without the ability to engage in economic calculation, police agencies will navigate the trade-off between excessive force and depolicing based on criteria other than what consumers desire.

Employee Protections

In addition to employment protections, police officers enjoy a number of due process protections beyond those afforded to nonpolice. While many of these protections are negotiated in contracts between municipalities and police unions (Rushin and Edwards 2017), they are also included in state and municipal codes (Walker 2006) as well as department policy and procedure manuals. These protections include delays between the occurrence of an officer-involved shooting and when the officer must provide a statement, officers’ being provided all evidence against them prior to interrogations, officers’ disciplinary records being expunged after a certain amount of time, the prohibition of anonymous complaints against police officers, and officers’ ability to appeal disciplinary actions against them to independent arbitration.

These protections have benefits for the taxpayer since, ceteris paribus, they allow municipalities to pay officers lower monetary wages, just as offering professors tenure allows universities to pay lower wages (Alchian 1959). This is illustrated by negotiations between the City of Chicago and the Chicago Police union. At a time when the city was facing budget shortfalls, Chicago’s Fraternal Order of Police, which engages in collective bargaining on behalf of the officers of the Chicago Police Department, was willing to accept greater protections for officers in lieu of wage increases (Chase and Heinzmann 2016). These protections, just like tenure, are not costless. One of the costs of these protections is the greater difficulty of holding problem officers accountable. Chicago’s Police Accountability Task Force has since recommended that these types of protections be removed or revised (Emmanuel 2017; Police Accountability Task Force 2016). From the police officers’ perspective, this is equivalent to reducing their total compensation. This is why, when Dean Angelo Sr., the president of Chicago’s Fraternal Order of Police, was asked by the city’s lawyers what it would take to remove the protections identified by the task force, he replied, “Bring the checkbook” (Newman 2017).

These protections, particularly the ability to appeal to binding arbitration, have allowed officers who would otherwise have been fired to keep their jobs. Pittsburgh Officer Paul Abel, for example, was reinstated after being fired for pistol-whipping and accidentally shooting an innocent person in retaliation for being battered by someone else. A Minneapolis police officer who was fired after being convicted of an off-duty misdemeanor sex offense was later reinstated after the arbitrator decided dismissal was too harsh (deFiebre 1993). That officer resigned a year after reinstatement due to being convicted of another sex offense (Michael Alan Kveen, License No. 10639, Docket No. 12-2402-10724-2 (Minnesota Office of Admin. Hearings Nov. 1, 1996)). Two Cincinnati officers were fired for having sex with an intoxicated woman while on duty and later lying about it during the consequent internal investigation. The arbitrator later reduced their punishment to suspensions of five days for one officer and three days for the other (Queen City Lodge No. 69, Fraternal Order of Police, Case No. 52 390 0007 03 (American Arb. Ass’n n.d.)). Thus, while these protections allow municipalities to pay lower monetary wages, they involve a shifting of costs from taxpayers to the victims of police misconduct.

How to optimally negotiate this trade-off cannot be determined outside the institutional contexts that enable economic calculation. For a public police department, the negotiation of this trade-off will be the result of the interaction of the relative influence of police officers, taxpayers, and victims of police misconduct. The latter two groups are neither one and the same, nor are they mutually exclusive. The burden of taxation is not evenly distributed but tends to fall more heavily on wealthier individuals. Conversely, victimization by police is disproportionately borne by less wealthy individuals. The composition of the compensation package, in terms of monetary wages and protections, will not be made on the basis of consumers’ willingness to pay but as a result of a political process.

The calculation problem faced by police departments is whether the composition of these compensation packages (as well as their amount) is economically justified by the value of the services provided by officers; however, because public police departments are unable to measure the market value of their output, they are also unable to impute value to the factors they use in the production process, including labor. In long-run competitive equilibrium, a laborer is paid his discounted marginal product, but this is not necessarily true of laborers in the public sector. One thing we can say for certain is that a public sector employee is paid at least his opportunity cost, but this has no necessary relationship to his marginal productivity in the public sector. Unable to measure the value of output, police managers are also unable to determine whether the wages paid to police officers reflect their marginal productivity.

The calculation problem resulting in the inability to optimally negotiate the trade-off between wages and protections does not exist in an environment of competitively provided policing services.^[12] Part of the reason these protections exist is to protect police officers from arbitrary disciplinary actions by management. The competitive process penalizes firms that fire qualified employees, because it is determined by consumer preferences. Public sector organizations are not subject to the same competitive forces as for-profit firms, so public sector managers have more leeway to make personnel decisions based on personal preferences. Managers of firms in a competitive environment attempting to do the same thing would be at a disadvantage. The compensation packages of employees would have to reflect their marginal contributions to overall output. Competition among firms for employees would lead to compensation packages that are commensurate with employee preferences and consumer desires. Only in a competitive system can the optimal compensation arrangements be determined. Compensation would have to be on par with the value of services provided to consumers, and job security would depend on the ability to continually provide that value, not on an arbitrator’s idea of what is “fair.”

Conclusion

This analysis has a number of implications. First—regarding the theory of a system of clubs versus a system of government, particularly the protective state—the security function of the protective state is underspecified. Buchanan’s (1975, 95–97) protective state protects the core rights of citizens via internal security, contract enforcement, and defense against external threats. Under a system of government, any attempt to fully specify at the constitutional stage the limits to the protective state’s power to achieve its tasks will be extremely costly (and likely impossible). The other means of ensuring constitutional compliance in such a system is through postconstitutional litigation of any perceived violation of the highly specified constitutional contract. As mentioned earlier, a system of clubs, by allowing competition in the provision of governance, economizes on contractual specification and litigation (which will only be worth doing if the expected costs of changing clubs are too high). Future research could look at the implications the literature on incomplete contracts has for the economic theory of constitutions.

A second implication is that the issues that arise due to police bureaucracies’ inability to calculate cannot be resolved through reforms that maintain the bureaucratic process through which policing is provided. Without the ability of consumers to express their willingness to pay for various policing services in a market, it is impossible to determine the optimal points of the various trade-offs mentioned. A corollary implication is that when consumers differ about what ends they would like the police to pursue and cannot express those preferences through a market, there will necessarily be political conflict over what the police do. When policing is provided in a one-size-fits-all fashion for an entire city, individuals that prefer less (or more) aggressive policing but are in the political minority are forced to live under a policing regime decided by others. This problem is exacerbated when the political minorities live in neighborhoods that are policed differently from neighborhoods occupied by the political majority.

Another related implication is that the optimal scale for service provision cannot be calculated when police are bureaucratically provided. There is little reason to believe that existing political jurisdictions are the optimal sizes for single service providers (Fegley and Growette Bostaph 2018; E. Ostrom, Parks, and Whitaker 1973; E. Ostrom and Whitaker 1973, 1974; Southwick 2005). The prevalence of private security for mass private property (Shearing and Stenning 1981) suggests that there are scale economies for some security functions. Under competition and calculation, firms can determine a more appropriate size and style of service than is possible bureaucratically. Individual neighborhoods can decide what style of policing they prefer without the need for a separate neighborhood in the same jurisdiction to agree (or to be overridden politically).

Finally, it should be noted that without changing certain institutional structures, some aspects of policing cannot be subject to economic calculation. For example, some of the costs of the decision to arrest (the use of the court system and, potentially, prison space), involve common-pool resources while individual decision-makers may only consider private costs and benefits. By the same token, private inputs for such common-access benefits (such as general deterrence through imprisonment of offenders) will be smaller when criminal justice institutions direct the benefits toward the commons. When victims have property rights to restitution, for example, they will be more willing to invest their time in costly activities such as reporting crime to police and testifying in court than when the ultimate result of the arrest and prosecution process is imprisonment (Barnett 2014; Benson 1990, 1992). It must be determined what public policy goals that can be better achieved through public administration are worth the costs of being unable to engage in calculation, if such a thing can be determined.