2
Organizational Structures for Policing
Police institutions in different countries are structured in a variety of ways, with different remits of responsibility and directives, typically informed by the government and realized by police leadership, regarding how they engage with their community. For the purpose of this report, police structures have been defined as the arrangements of responsibilities and authority invested in a policing agency to include appropriate oversight mechanisms. The previous chapter introduced the committee’s endorsement of an evidence-based approach to policing, as well as defining the rule of law (ROL) and protection of the public, with an overview of existing metrics and tools used to measure both concepts. Here we focus on the question, “what organizational structures enable a police service to promote the ROL and protect the population?”
The discussion here affirms that of Bayley (2006) and Neyroud (2021); there is no evidence that any single form of organization is more effective than any other in promoting the ROL and public protection. The effort invested in reorganizing police agencies, as many Global North nations have done in recent years, offers no clear link to any changes in policing outputs or outcomes. Nevertheless, there are certain features of police agencies in some countries that attract the support of experienced observers, including the discussants in our workshop. There are also major issues in organizational patterns in Global South countries, largely concerning the relationships between responsibilities and authority of military institutions in relation to policing tasks.
The distinction between a militarized and a civilian policing service will be investigated next, followed by an examination of governance structures
for policing agencies, which are theoretically linked to supporting the ROL and protecting the population. Finally, the chapter concludes with consideration of the ways in which police misconduct is investigated to ensure that improper and/or suspect behavior by police officers is counted and addressed.
MILITARY VS. CIVILIAN FORCES
As Neyroud (2021) observes, contemporary democracies tend to separate the roles of police, devoted to public safety, and the military, devoted to national security. This distinction is a central element in civil-military relations that is conducive to civilian control over the military (Dammert and Bailey, 2005). In theory, civilian control over the military is important to help consolidate democracy, reduce the threat of a coup d’état, and assert the primacy of the ROL over the rule of might.
The separation between civilian police and military forces, however, has become increasingly blurred across many countries, especially those in Latin America and in parts of Africa (González, 2020). This boundary blurring is in part due to the increased militarization of the police. Flores-Macías and Zarkin (2020) define militarization as the processes by which police agencies adopt the weapons, organizational structure, and training typical of the military forces. The boundary is further eroded in countries where soldiers are increasingly replacing civilian police in policing tasks. Flores-Macías and Zarkin (2020) refer to this latter process, by which armed forces take on the responsibilities of civilian police agencies, as the constabularization of the military.
At the committee’s public workshop, Gustavo Flores-Macías described a continuum of militarization, which extends from nonmilitarized police to the constabularization of the military. There can be varying levels of militarization within countries, as the organizational structure of policing shifts from low degrees of centralization and hierarchy to higher degrees of centralization and hierarchy. Training may also shift from a focus on maintaining public order with minimal and nonlethal use of force to one focused on establishing order through specific tactical formations with stronger use of force in challenging environments. According to Flores-Macías, these variations in organizational structures influence the physical and psychological distance between police and the population. In Latin America, this distance is widening, and it is viewed as contributing to police indifference to escalating conflict and human rights violations (see Box 2-1 for additional information on the militarization of the police in Latin America).
Militarization of Police
One aspect of the militarization of police entails providing policing organizations with military-grade equipment. The evidence on the effects of such equipment is mixed and inconclusive. In the United States, a small number of studies have examined the impact of low-cost transfers of surplus military equipment (ranging from computers to armored vehicles) to police departments through the federal 1033 program. Bove and Gavrilova (2017), Harris and colleagues (2017), and Masera (2021) all find that violent crime fell when departments received equipment via the 1033 program, using data on equipment availability (the proximity of a county to an army distribution center) and shifting expenditures as well as crime reports and citizen complaints. However, these findings are contested; Gunderson and colleagues (2021) identify inaccuracies in the 1033 data used in Bove and Gavrilova (2017) and Harris and colleagues (2017), and find that correcting these errors, and disaggregating the unit of observation, attenuates both the magnitude and statistical precision of their estimates. Evidence on the impact of the 1033 program on police violence against citizens and community satisfaction is even more uncertain: Bove and Gavrilova (2017) and Harris and colleagues (2017) find little change, but Masera (2019) finds an increase in both killings by police and assaults against police officers. Lowande (2021) identifies no change in reported violent crime after the 1033 program was terminated and much of the acquired military equipment was returned.
In addition to issues of measurement and aggregation, none of these studies is able to estimate the “first stage impact” of these transfers on the actual use of military-style equipment by the department; the methods are limited to what economists call “reduced form” analyses, without a clear measure of exactly how much the 1033 program actually increased police militarization. Stronger conclusions would require a measure of the amount of military-grade equipment used by different agencies, data that are not regularly collected, in order to determine the extent to which the 1033 program increased militarization, or simply crowded out the purchase of military-grade gear that would have occurred anyway.
Another aspect of the militarization of police concerns the use of military-style force units (e.g., SWAT teams). Mummolo (2018) compares crime and officer safety across departments that mobilize military-style force (SWAT teams) more or less often, and finds that departments that frequently deploy SWAT teams work in areas with higher crime rates and more violence against officers. The policy interpretation of this correlation is unclear.
Constabularization of the Military
The nascent literature on the constabularization of the military focuses on two related questions: (1) What factors contribute to the constabularization of the military? and (2) What is the effect of constabularization on crime and violence as well as on the ROL more generally (including human rights, democratic stability, and legal order)? Is there evidence that the military is more (or less) effective than the police in enhancing public safety? The committee finds the evidence on these questions to be inconclusive.
Two key factors are commonly cited as an impetus for constabularization. One is increased insecurity due to dramatic increases in gang-related violence. In Latin America, this is tied to the war on drugs and governmental crackdowns on ungoverned spaces such as favelas (e.g., Dell, 2015). The second driver is theorized as the popularity of constabularization with the public. Public opinion supporting use of the military for police tasks has been tied to low esteem for police in many developing countries. Studies analyzing public opinion data from the Latinobarómetro and the Afrobarometer show high levels of citizen distrust in the civilian police force “to do its job in a successful, transparent, and humane manner” (p. 5). Police are often viewed as “inept, corrupt, outnumbered, and outgunned by lethal criminal syndicates with sufficient resources to purchase police docility or connivance” (Pion-Berlin and Carreras, 2017, p. 7). At the same time, popular trust in the armed forces is both high in absolute terms and much higher than trust in civilian police forces. Indeed, Pion-Berlin and Carreras (2017) find that citizens’ support and requests for military crime fighting becomes higher when satisfaction with the police declines. Yet while popular support for a particular type of policing is an important component of the ROL, systems of majority rule do not necessarily contain safeguards against the exploitation or persecution of minority groups.
There is ongoing debate about the effect of constabularization on crime and violence and its relative effectiveness compared to outcomes from civilian police operations. Advocates view constabularization as necessary if temporary measures to curb rampant crime, especially given the militarization of organized crime (Pion-Berlin and Carreras, 2017). By contrast, critics argue that by using excessive force, constabularization causes an increase in violence and human rights violations (Flores-Macías and Zarkin, 2019) without necessarily deterring or incapacitating criminals (Muggah et al., 2018). Constabularization can also be used as a political strategy. In 2010, the South African Police Service engaged military ranks and established a rhetoric of force, in order to portray an anticorruption image that would appeal to the public’s desire for protection and retribution (Faull, 2010).
Unfortunately, much of the debate is anecdotal (see critique by Blair and Weintraub, 2021), with arguments generally based on trends in crime—for example on rising or falling homicide rates before and after constabularization policies are implemented (Muggah et al., 2018)—that may not capture underlying causal processes, and may in fact obscure them. For example, if the military is sent to the hardest places to police, or where the level of violence is already trending upwards, then a positive association between constabularization and measures like homicide may simply reflect outcomes of other causes.
Among the handful of studies paying closer attention to causal identification, there is some evidence that constabularization in a fight against highly organized drug cartels has led to increases in levels of violence. By comparing places where the military was constabularized to other places with similar pre-trends in homicide rates, based on either a simple or weighted mean of homicide rates over time, Flores-Macías (2018) finds that constabularization in Mexico led to violence escalation relative to the proposed counterfactual places. Similarly, employing a matching estimation, Espinosa and Rubin (2015) find that military interventions in the Mexican drug war led to an increase in the average homicide rate. The committee notes, however, that the estimated effect in these studies reflects the impact of constabularization as part of an escalation of conflict between the national government and powerful drug cartels. It is unclear whether the effects estimated in these studies are generalizable beyond that specific context, namely where constabularization is part of a crackdown on highly organized and violent drug cartels.
On the other hand, there is suggestive evidence that using military forces to patrol the streets reduces crime. In a randomized controlled trial (or Level 5 on the Maryland Scale) conducted in collaboration with the Cali municipal government in Colombia and colleagues (2020) find that using military forces to patrol the streets reduced crime, but only on days and at times when soldiers were physically present on the streets. Crime rates remained unchanged when soldiers were not physically present, thereby failing to generate the “residual deterrence” effects of civilian policing, as reported by Sherman (1990) and his colleagues (Barnes et al., 2020; Koper, 1995). Despite these small or null effects of military patrols on crime, Blair and Weintraub (2020) find evidence that the militarized policing program stimulated public demand for more aggressive military involvement in policing.
In sum, much more work is needed to fully understand the tradeoffs involved in using military forces for police activities. Of note, much of the research has focused on the outcomes of militarization, that is, on the increase or decrease in levels of crime and violence; there has been very little
research on the impact of militarization on the organizational structures, policies, and practices of policing.
The body of research presents mixed findings regarding the role a militarized police service plays in protecting the population and raises concerns about the consequences for the ROL. Militarization in the Mexican police has been associated with greater homicidal violence and human rights violations, but further study is needed to see if these findings hold outside the Mexican context. Studies in the United States on the militarization of police (whether through transfer of military-grade weapons or SWAT-style teams) have not consistently found crime reductions in the areas where this has been deployed. In a randomized controlled trial in Cali, Colombia, found that crime went down in the areas patrolled by the military. Notably, researchers in Colombia and Mexico have found that militarization appears to enjoy considerable public support. However, additional research is required to determine if a militarized police force has a positive impact in promoting the ROL and protecting the population.
GOVERNANCE OF POLICE
External Governance and Regulation of Police
The qualitative evidence on systems of police governance suggests that their effectiveness strongly depends on the national form of government (Neyroud, 2021). The relevant dimensions include their structures of checks and balances, as well as the location of policing under national control, local control, both forms of control, or even more layers.
Police Fragmentation and National vs. Local Governmental Control
There is considerable variation around the world in how countries have chosen to organize their police forces, including the level of government that has jurisdiction over policing (national, state, or local) and the degree of fragmentation of police forces (i.e., the number of police forces charged with providing internal security). There is no clear evidence about the effects of placing decision-making authority at higher or lower levels of government. These choices may have inherent tradeoffs as it relates to policing practices. On the one hand, commanders who are focused on smaller regions may be better able to identify and respond to highly localized “hot spots” that cause the majority of crime and disorder. On the other hand, to the extent that the actions of police in one neighborhood affect the reputation of police in other areas, particularly concerning police legitimacy, higher-level governments may be better able to internalize any
costs of benefits that accrue to one neighborhood based on the decisions made in another. What difference these tradeoffs make, if any, is unknown.
Neither is there any clear evidence concerning the consequences of national or subnational organization of police forces or of the fragmentation of police forces on outcomes related to public protection or the ROL. One recent study of the organization of police forces in 110 developing countries (including 75 postconflict countries) finds that greater fragmentation of police forces can lead to increased police violence against the population and greater risk of armed conflict, though a greater number of local or specialized police forces had no such effect (Arriola et al., 2021). Along these lines, there is some evidence to suggest that policies that seek greater coordination and information sharing across fragmented police forces can improve police ability to protect the population. Soares and Viveiros (2017) estimate that greater coordination (through information sharing and operational coordination) of military and civil police forces in the Brazilian state of Minas Gerais resulted in a 24 percent reduction in property crime and a 13 percent reduction in violent crime, yet this finding may be primarily about numbers of people in uniforms on patrol, not about organization per se. There is little available evidence on the effects of formal consolidation of police jurisdictions, as occasionally occurs in the United States when small jurisdictions cede policing authority (as a cost-saving measure) to a more centralized government agency, typically a county sheriff or metropolitan police department. The committee is unaware of any evaluation of this type of government consolidation on aggregate crime levels, in part due to limitations on the way crime data are collected in the United States.
What does seem to matter substantially, in the view of the practitioner experts who gave evidence to the committee, is the extent to which police are operationally independent from direct political control (Neyroud, 2021; O’Connor, 2021; White, 2021). While they accepted that there is a fine balance between too little and too much police autonomy, these practitioners stressed the need for police to make decisions based on law, training, and expertise despite the transient demands of political leaders. Further detail is found in the box on operational independence. See Box 2-2.
Private vs. Public Governance
Policing scholar David Bayley defines police as “people authorized by a group to regulate interpersonal relations within the group through the application of physical force” (1985, p. 7). As this definition suggests, the function of policing is not always conducted by a public police force administered by a governmental entity. The provision of security and the maintenance of order around the world varies greatly, along a continuum ranging from public police forces and legalized private police forces to
traditional or customary forces among indigenous communities and private militias with varying degrees of state authorization. In the United States, many universities and other private entities have legally constituted private police forces with considerable authority to conduct arrests and other policing duties. Meanwhile, governments often engage private militias or irregular armed groups in security provision, such as the use of private militias as auxiliary police in Afghanistan and Mexico’s efforts to legalize and regulate autodefensas (armed vigilante groups) in the fight against drug cartels. Analysis by human rights groups and security experts suggest such experiments pose serious human rights and safety concerns1 as well as risks for further violence.2 Moreover, public safety provision may also involve a range of private actors, from the proliferation of private security companies (which typically do not have police powers such as the authority to arrest) to cooperation and support from the business sector.
In the United States, researchers have found that “Business Improvement Districts,” which grant business groups the authority to levy taxes and administer a particular area, can lead to public safety improvements (Brooks, 2008; Cook and MacDonald, 2011). In a study of a unique program carried out in the city of New Orleans, Cheng and Long (2018) find that a privately managed police service proved effective at reducing crime “as the consequence of using more effective monitoring and incentive strategies” (p. 2). There is reason to be skeptical, however, about the extent to which such findings from the United States can be applied to many countries in the Global South. Qualitative research on security and policing in Colombian cities has shown that, depending on local context, business groups can both help to support security policies that promote protection of the population, as occurred in Medellin, and form alliances with police forces that undermine the ROL, as occurred in Cali (Moncada, 2016).
INVESTIGATING POLICE MISCONDUCT
A key structural question in policing is which agencies have the power to investigate police misconduct. Such investigations seem essential to ensure that all police activity is undertaken in a fair and impartial manner and that officers are ultimately accountable to the communities they serve. In countries with a high ROL ranking, it is generally possible for a third-party agency to be assigned to investigate potential misconduct in policing—especially criminal or corrupt misconduct. In the U.K., one police force may
___________________
1 See https://www.hrw.org/report/2011/09/12/just-dont-call-it-militia/impunity-militias-andafghan-local-police.
2 See https://www.wilsoncenter.org/sites/default/files/media/documents/publication/horton_michoacan.pdf.
be assigned by the central government to investigate another, with arrests of even chief constables resulting from such assignments (examples from the U.K.’s implementation of police oversight bodies are explored in Box 2-3). Other models include the Central Bureau of Investigation in India, which can investigate state-run police agencies; the Bureau of Criminal Apprehension in Minnesota, which has investigated numerous police killings of citizens; and the Federal Police of Brazil, which collaborated in the
investigation, prosecution, and (by court sentence) imprisonment of presidents of that country.
In the United States, the elected local district attorneys in each county have the legal power to investigate the police, but may rarely do so because of the symbiotic relationships between police and prosecutors in convicting serious criminals. Because of concern about under-investigation due to this relationship, in 1972 New York State Governor Nelson Rockefeller appointed a deputy state attorney general as a special prosecutor for corruption in the New York City criminal justice system.
Sherman (1978) analyzed six years of internal investigations and intelligence on corruption before and after the appointment of the New York State special prosecutor, in order to ascertain whether there was a progressive decline in the organizational scale of ongoing corruption arrangements. Other anticorruption strategies at the same time included police internal affairs testing of officers’ propensity to steal or accept bribes. Internal informants recruited at the Police Academy were also cultivated by internal affairs officers, leading to further intelligence about and prosecutions of corrupt police networks. How much of the observed reduction in organizational complexity of police corruption was due to the external prosecutor—versus other structural changes like increased proactive policing of the police—remains unknown. However, this multiagency response to public demand for more honest policing was associated with an apparent end to widespread corruption in the New York City Police Department.
In addition to internal investigations, police misconduct may be investigated under international legal standards and local independent oversight and complaint bodies.3 These standards, which aim to promote a culture of accountability, require that members of the public must be able to report and file complaints against officers or entire agencies, either directly with the station or with the prosecutor’s office or an independent investigating agency (the latter two options might be preferred to ensure lack of intimidation or harassment from the officers accused of impropriety). A fair and anonymous system for submitting complaints, the standards state, is integral to the investigation process, as without a formal complaint a formal investigation is unlikely (see footnote 3). Effective independent investigative agencies may foster public trust and increase government legitimacy by improving citizens’ access to justice for crimes committed by state agents, while ensuring due process for the accused.4 Additional mechanisms for
___________________
3 See https://www.unodc.org/pdf/criminal_justice/Handbook_on_police_Accountability_Oversight_and_Integrity.pdf.
4 See https://www.justiceinitiative.org/uploads/888326d8-77c9-4bfc-b0f2-db908fc345fe/osjiwho-polices-the-police-5-7-2021.pdf.
promoting an accountable policing agency from within the organization are examined in Chapter 3.
CONCLUSION
This chapter has examined the limited evidence, primarily from case studies and the experience of police reform efforts, on whether and how organizational structures of policing agencies affect the agencies’ capability to both promote the ROL and protect the population. The information in this chapter is offered as guidance for consideration in building or supporting the reform of policing structures. In the next chapter, internal policies for improving the ability of police to behave in an accountable manner will be investigated, including officer recruitment and retention, the use of technology, and internal governance.
This page intentionally left blank.
