Agency Correlates of Police Militarization: The Case of MRAPs

Abstract

In 2014, protests in Ferguson, Missouri (MO), and the subsequent law enforcement response, shined a light on police militarization—the adoption of military styles, equipment, and tactics within law enforcement. Since 1990, the U.S. Department of Defense has transferred excess military equipment to domestic law enforcement agencies via the federal 1033 program. This article examines transfers of mine-resistant ambush-protected vehicles or MRAPs. Designed to withstand explosive blasts during U.S. military occupations in Iraq and Afghanistan, surplus MRAPs have been shipped to more than 800 domestic law enforcement agencies. This article uses national data on law enforcement agencies and on 1033 program transfers to analyze the pattern of MRAP distribution. The results show that MRAPs are disproportionately acquired by agencies that have warrior tendencies and rely on asset forfeiture to generate revenue. This pattern of militarization is consistent with a model of governance that views citizens as both opportunities and threats.

Keywords

militarization 1033 program SWAT asset forfeiture

On August 9, 2014, Darren Wilson, a White police officer, fatally shot Michael Brown, an 18-year-old Black man, in Ferguson, MO. Initial accounts of the incident varied. Wilson claimed that Brown reached for the officer’s gun. Many bystanders claimed that Brown was in retreat, hands in the air. What was uncontested is that Brown’s body lay in the street for some 4 hours after his death. Local residents gathered and the level of discontent grew. Thus began 2 weeks of protest, which included property damage and violence. Local police forces responded to the protests with a show of physical and symbolic strength. Large, imposing military vehicles arrived. Snipers took positions above crowds. Officers wearing gas masks and riot gear deployed canisters of tear gas toward groups of citizens. Armored police with long guns confronted residents (Institute for Intergovernmental Research, 2015). For many Americans, this was the first awareness of police militarization: the adoption of military styles, equipment, and tactics within law enforcement.

Police militarization calls attention to the relationship between police and civilians (Balko, 2013; Kappeler & Kraska, 2013; Kraska & Cubellis, 1997; Stoughton, 2016; Wester, 2016). In adopting styles, equipment, and tactics from the military, domestic law enforcement risk alienating the citizens they are charged to serve. Public opinion polls reveal that U.S. residents are split in their perceptions of police militarization. Several national polls conducted since the events in Ferguson have found that roughly 40% to 50% of Americans oppose police use of military equipment (Ekins, 2016; Pew Research Center, 2014; YouGov, 2014). Ultimately, demonstrations of military force may contribute to a decline in the perceived legitimacy of local police forces (Bieler, 2016, p. 596; Kochel, 2017; Morin, Parker, Stepler, & Mercer, 2017; President’s Task Force on 21st Century Policing, 2015).

This study analyzes militarization as manifest in the federal 1033 program. The 1033 program transfers surplus military equipment—including weapons, vehicles, and armor, among other items—to law enforcement agencies (LEAs) across the country. Since its inception, in 1990, the 1033 program has transferred over $6 billion worth of equipment from the Department of Defense (DOD) to more than 8,000 LEAs (Defense Logistics Agency, 2017). Prior to Ferguson, the 1033 program attracted little attention. Recently, several studies (reviewed later) have begun to examine the causes and consequences of militarization via 1033.

The present study makes two contributions to this literature. First, it considers agency-level correlates of police militarization. By merging 1033 acquisition data with other agency-level data, this analysis offers new insights into the organizational factors—including budgets, staffing, and missions—that are associated with militarization. Second, this study focuses on a specific piece of military equipment: mine-resistant ambush-protected vehicles or MRAPs. As described later, modern MRAPs were developed for and deployed in foreign counter-insurgency contexts where they were expected to resist attacks by roadside bombs and permit soldiers to fight their way out of coordinated ambushes. Unlike other pieces of military equipment (e.g., handguns, rifles, utility trucks), MRAPs have no obvious role to play in domestic law enforcement due to their singular design. As such, they represent a unique contribution of the 1033 program to police militarization. A series of regression analyses reveal that MRAPs are disproportionately acquired by agencies that have warrior tendencies and rely on asset forfeiture to generate revenue. This pattern of militarization is consistent with a vision of predatory policing in which police view citizens as both opportunities and threats.

The article has five sections. The first situates the analysis in the context of policing in a democratic society. The second section surveys the history of police militarization in the United States and describes the 1033 program. It also introduces a series of hypotheses regarding MRAP acquisitions. The third and fourth sections detail data and methods and the results of the analysis. A concluding section discusses the implications of our findings for modern policing and its role in a democratic society.

Policing in a Democratic Society

A police force plays a dual role in modern societies. On one hand, the police are the manifestation of the coercive state (Neocleous, 2014), but on the other hand, they are public servants empowered to protect and safeguard the citizenry. Stoughton (2016) characterizes this dual nature in terms of guardian officers and warrior cops. Guardian officers value public service, respect, and empathy. The guardian role underpins community policing initiatives, which aim to cultivate robust and collaborative relationships with citizens and community organizations (President’s Task Force on 21st Century Policing, 2015). In contrast, warrior cops value honor, duty, resolve, and the “willingness to engage in righteous violence” (Stoughton, 2016, p. 632) against citizens deemed deserving of it. Warrior cops value experience and dismiss input from civilians, politicians, and other would-be critics (Stoughton, 2016).

The role of warrior cop encourages police to view citizens as potentially hostile and in turn encourages citizens to view police as an occupying force (Hayes, 2017). Surveys show that trust in the police is low, especially among minority populations (Morin et al., 2017; Peffley & Hurwitz, 2010). Moreover, survey research suggests that police perceive the populace as highly volatile and threatening (Marenin, 2016; Nix, Wolfe, & Campbell, 2018). Shows of overt military-style force threaten to amplify these antagonistic perceptions. Ferguson serves as a case in point. A postincident review of the protests noted that “deployment of an elevated, military-type police response may actually prompt unintended consequences such as were apparent in Ferguson, where initial responses appeared to galvanize a negative perspective and aggravate community concerns about police and the justice system in general” (Institute for Intergovernmental Research, 2015, p. 53). In the post-Ferguson era, police militarization stands as a potential impediment to police–community relations.

Police Militarization

Following Balko (2013), we consider police militarization to be the adoption of military styles, equipment, and tactics within law enforcement. Modern forms of police militarization arose in the 1960s in response to fears of urban unrest, racial violence, drug-related crime, and rising gun violence. Police forces around the country formed SWAT (Special Weapons And Tactics) teams which would use military-style equipment and tactics to combat domestic crime (Balko, 2013). The growth of SWAT teams was such that in March 1968, the Associated Press wrote that “in city after city across America, the police are stockpiling armored vehicles, helicopters, and high-powered rifles … they are preparing for summer and the riots they hope will not occur” (Balko, 2013, p. 63).

The militarization of police forces continued in subsequent decades and has contributed to a blurring of policing and soldiering (Baumgart, 2016; Campbell & Campbell, 2010). In a survey of more than 500 U.S. police departments, Kraska and Kappeler (1997) found that the share of departments with paramilitary units grew from less than 10% in 1970 to nearly 60% in 1982. A key driver of militarization was the “war on drugs,” which employed military equipment and training to execute drug raids (Balko, 2013). A series of executive and legislative actions (e.g., the 1981 Military Cooperation with Law Enforcement Act, the 1986 National Security Directive 221, FY 1987 National Defense Appropriate Act) furthered cooperation and collaboration between domestic law enforcement and national defense (Alexander, 2010; Balko, 2013). However, it was not until the early 1990s that the U.S. government began to systematically transfer military equipment to local LEAs.

The 1033 Program

The 1990/1991 National Defense Authorization Act (NDAA; codified in Pub. L. 101–189, 103 Stat 1566) authorized the DOD to transfer military equipment to federal and state agencies for counter-drug activities. The 1997 NDAA expanded the program to include all law enforcement activities, with priority given to counter-drug and (now) counter-terrorism operations. (The revised surplus transfer program was codified at section 1033 of the statute [Pub. L. 104–201, 110 Stat 2422], giving the program its common name.) Since 1995, the 1033 program has been administered by the Law Enforcement Support Office (LESO), within the Defense Logistics Agency. Under the program, LEAs request equipment from LESO via their state 1033 coordinator. If approved, LESO then releases the equipment to the LEA. The LEAs receive the equipment at no cost, but they must pay for transportation, training, and maintenance (American Civil Liberties Union [ACLU], 2014a, Appendix B; Defense Logistics Agency, n.d.).

The 1033 program distributes a wide variety of equipment, ranging from mundane items (e.g., furniture, luggage, and boots) to lethal equipment (e.g., rifles, firearm sights, and armored vehicles). The program enjoys broad popularity among police officers and executives (Turner & Fox, 2017). Since its inception, more than 8,000 LEAs have participated in the 1033 program (Defense Logistics Agency, n.d.). The degree of 1033 utilization varies greatly, with acquisition values ranging from $0 to over $200 million in U.S. counties (Radil, Dezzani, & McAden, 2017).

Several recent studies have considered the impact of the 1033 program on law enforcement effectiveness at the county level. A pair of studies employs an instrumental variables framework and concludes that the 1033 program is effective in reducing crime. In particular, Harris, Park, Bruce, and Murray (2017) found that the 1033 program yielded lower crime rates (robbery, assault, and motor vehicle theft), more drug-related arrests, and fewer citizen complaints against police. Moreover, they found no evidence that the 1033 program made officers less safe (as measured by officer deaths). In a similar vein, Bove and Gavrilova (2017) found that acquisitions of nonlethal equipment, vehicles, and miscellaneous items (but not weapons) caused crime rates to fall in subsequent years in counties from 2006 through 2012. The authors argue that this crime reduction effect operates through a deterrence mechanism, as additional equipment allows for a higher degree of visibility and more effective deployment of officers. However, research by Delehanty, Mewhirter, Welch, and Wilks (2017) and Lawson (2018) suggests that U.S. counties and agencies that acquired more 1033 items witnessed a subsequent increase in the number of civilians killed by the police.

The Ferguson protests brought the 1033 program and police militarization into the national spotlight. Google searches of “1033 program” and “police militarization” spiked immediately after Michael Brown’s death, as shown in Figure 1. (Search activity surged again after a grand jury failed to indict Darren Wilson for the shooting, a result which prompted a new round of public protests.) President Obama ordered an initial review of the program on August 23, 2014, shortly after Brown’s death. This culminated in Executive Order 13688 (January 2015), which directed a working group to draft new regulations for LEA acquisition of military equipment. The resulting regulations were released in May 2015 (Law Enforcement Equipment Working Group, 2015). Among other things, they prohibited the transfer of certain items (including weaponized vehicles, tracked armored vehicles, bayonets, and grenade launchers), mandated additional training for LEAs acquiring “controlled” equipment (including aircraft, wheeled armored vehicles, explosives, and battering rams), and imposed stricter accounting and monitoring practices for transferred equipment. These restrictions were later rescinded by President Trump’s Executive Order 13809 in August 2017.

Figure 1.

Search interest in police militarization and 1033 program.

Mine-Resistant Ambush-Protected Vehicles

Much of the equipment transferred through the 1033 program is mundane (tools, office supplies, street-legal vehicles, clothing) and is easily incorporated into a police force. Other items are more exotic and have a unique military nature. One example is mine-resistant ambush-protected vehicles or MRAPs. MRAPs are physically imposing vehicles weighing up to 24 tons. A V-shaped hull protects personnel from roadside bomb blasts and their armored facade protects against small arms fire (U.S. Army, 2018). Unlike other pieces of surplus military equipment (e.g., handguns, rifles, utility trucks), MRAPs have no obvious application in domestic law enforcement. In fact, some of the MRAPs’ design elements may hinder its use on American streets (e.g., its imposing nature and relative immobility).

The U.S. military made extensive use of MRAPs during the conflicts in Afghanistan (2001) and Iraq (2003), and by 2013, there were 21,000 MRAPs in U.S. service. However, in 2013, the U.S. military stopped purchasing new MRAPs and determined that some 13,000 existing MRAPs were excess to requirements (Vanden Brook, 2014). This freed considerable numbers of MRAPs for release into the 1033 program. Large-scale release of MRAPs through the 1033 program began in 2013, as shown in Figure 2.¹ Hundreds of MRAPs were shipped to LEAs before the Ferguson protests shined a light on police militarization. With only a few exceptions, MRAP transfers ceased following the Ferguson protests and through the federal review process. Transfers did not resume until January 2016, after the revised regulations on 1033 transfers were implemented. Geographically, the distribution of MRAPs is heaviest in Sun Belt and upper Midwest states, with particularly high per capita concentrations in Arkansas, Alabama, New Mexico, and Oklahoma (see Figure 3).

Figure 2.

Timeline of MRAP shipments. MRAP = mine-resistant ambush-protected vehicle.

Figure 3.

MRAP acquisitions by state. MRAP = mine-resistant ambush-protected vehicle.

Relying on prior research and theoretical expectations, we derive a series of hypotheses that attempt to explain the distribution of MRAPs to LEAs.

Hypothesis 1: Agency mission

We use the phrase agency mission to refer to the policies of a police force, which may reflect the agency’s support for the warrior or guardian roles of policing (Stoughton, 2016). A prototypical warrior mission would involve the use of tactical military equipment (e.g., body armor) and special operations forces (e.g., SWAT) for physically subduing potential threats from the populace. In contrast, a prototypical guardian mission would entail collaboration with citizen stakeholders and emphasize the embeddedness of police within the community. Because MRAPs are designed to insulate their operators from the external environment, we expect that they will be more common in agencies with a warrior mission and less likely in agencies with a guardian mission.

Hypothesis 2: Administrative capacity

Acquiring 1033 equipment requires staff. Someone must search the online 1033 database and submit a proposal to LESO. In addition, LEAs must meet federal accounting, training, and monitoring standards for all “controlled” equipment, which includes MRAPs. Smaller agencies may not have enough staff to undertake these efforts. As such, we expect that larger agencies will be more likely to acquire an MRAP. However, this type of administrative work may be a low priority for sworn officers who are responsible for making arrests and interacting with citizens. Agencies with a relatively large number of civilian (nonsworn) personnel may have a greater administrative capacity to pursue new equipment. Thus, we expect that agencies with more civilian staff will be more likely to acquire an MRAP.

Hypothesis 3: Fiscal strain

The 1033 program may be used by resource-poor agencies to acquire coveted equipment, including MRAPs (Johnson & Hansen, 2016, p. 800). We therefore expect higher rates of MRAP acquisition among LEAs with smaller budgets. We also expect MRAP acquisitions to be concentrated among agencies that rely on external revenue streams to supplement (perhaps meager) internal budgets. Many LEAs, particularly at the local level, depend on revenues extracted through the criminal justice system (ACLU, 2010). This can take the form of fines, user fees (Beckett & Harris, 2011; U.S. Department of Justice, 2015), or civil asset forfeiture (in which police seize property alleged to be involved in criminal activity; see Baicker & Jacobson, 2007; Holcomb, Kovandzic, & Williams, 2011). The use of external revenue sources may indicate budget strain, and we expect this to be associated with MRAP acquisition.

Hypothesis 4: Agency demographics

We expect that MRAP acquisitions will be associated with agency demographics: education, veteran status, race, and gender. Previous research has found that officers with more education are less likely to use force against suspects (Rydberg & Terrill, 2010; Terrill & Mastrofski, 2002). In addition, recent public opinion research has found that people with less education are more likely to support police use of military equipment (Lockwood, Doyle, & Comiskey, 2018). This leads to our expectation that agencies with higher levels of education will be less likely to acquire an MRAP. We expect agencies with many military veterans to be more likely to acquire MRAPs through the 1033 program. Many veterans work in law enforcement (Weichselbaum & Schwartzapfel, 2017), and they may bring a familiarity with military gear into the force (Turner & Fox, 2017). Gender may also influence MRAP acquisitions. Prior research has found that men are more likely than women to endorse police use of force (Carter & Corra, 2016) and more likely to support police militarization (Moule, Fox, & Parry, 2018). Moreover, men may be more accustomed to military equipment because they are overrepresented in the military (Office of the Under Secretary of Defense, n.d.). Therefore, we expect that agencies with more men are more likely to acquire an MRAP.

We also expect that agencies with more Black officers are less likely to acquire an MRAP. Trust of police officers is low among African Americans (Peffley & Hurwitz, 2010; Wozniak, 2016). We expect that Black officers will be attuned to this police–community tension and will be reluctant to engage in policing practices that risk further alienating communities—especially African American communities. Therefore, we expect that LEAs with more Black officers will be less likely to acquire an MRAP.

Hypothesis 5: Community demographics

The United States has substantial racial segregation, and this segregation contributes to racially disparate experiences with the criminal justice system, including local law enforcement (King & Johnson, 2016; Kochel, 2017; Kochel, Wilson, & Mastrofski, 2011). We expect that the racial composition of the community will influence LEAs’ MRAP acquisitions in three ways. First, we expect agencies in communities with relatively large Black populations to have fewer MRAPs. Moule et al. (2018) and Lockwood et al. (2018) have recently presented evidence that support for police militarization is lower for Black Americans than for White Americans (although this finding depends on model specification). It is plausible to think that agencies in communities with large Black populations are sensitive to race-based citizen–police relations and are attuned to the potentially troubling optics of military gear deployed in Black neighborhoods. This is borne out in the empirical literature (Ajilore, 2015). Second, agencies in severely racially segregated communities may be more likely to acquire MRAPs due to the spatial concentration of a presumed antagonistic population (Ajilore, 2015). Third, MRAP acquisition may be highest where there is a racial mismatch between the agency and the community. Racial mismatch between citizens and police exists in many cities across the country (Berube & Holmes, 2016). We expect MRAP acquisitions to be highest when agencies are predominantly White, but the surrounding community is predominantly non-White.

Data and Analysis

The primary source of data is the 2013 Law Enforcement Management and Administrative Statistics (LEMAS) survey of state and local LEAs, conducted by the Bureau of Justice Statistics. The survey was sent to all agencies with at least 100 sworn officers and a representative sample of smaller agencies stratified by agency size. The response rate was 86%, and the resulting sample consists of 2,059 local police departments, 717 sheriff’s offices, and 46 state LEAs. Data on MRAP acquisitions come from LESO, which maintains an itemized list of all property transferred to participating LEAs from 1990 to 2016 (Defense Logistics Agency, 2017). Data on community racial demographics come from the U.S. Census’ American Community Survey, 2006–2010 (U.S. Census Bureau, n.d.).

The final analytic sample omits several types of agencies. Agencies with only one full-time officer are omitted as a single officer who could not maintain and operate an MRAP.² State agencies are also omitted because their covered jurisdiction is much larger and they often have a different mission from local LEAs (e.g., highway patrol). Finally, the Chicago Police Department is omitted because it was an outlier in the LESO data, having received an MRAP in 2009, years before any other agencies (see Figure 2) and predating the LEMAS data. After removing these observations, the (unweighted) analytic sample comprises 2,740 agencies. Of these, 942 are missing data on at least one covariate.³ Multiple imputation (MI) is used to impute missing values (Acock, 2005). In this case, MI creates 30 data sets with missing values imputed through an imputation model. Regression analyses are conducted separately on each data set, after which the individual coefficients are pooled and the variances are adjusted for the uncertainty due to the imputation (StataCorp LP, 2013).⁴

Dependent Variable

The dependent variable is a binary variable indicating whether an LEA acquired an MRAP through the 1033 program between 2009 and 2016. Five percent of LEAs received at least one MRAP in this period.⁵

Independent Variables

We model MRAP acquisition as a function of variables representing agency mission, administrative capacity, fiscal strain, agency demographics, and community demographics. All independent variables come from LEMAS 2013 unless noted otherwise.

Agency mission

Agency mission is measured by considering various warrior and guardian aspects of departmental mission statements and protocols. An agency’s warrior tendencies are captured by its policies on body armor and the use of special units or personnel dedicated to particular issues. Agencies are coded as having one of three types of body armor policy: a written policy that requires all uniformed officers in the field to wear body armor (72%), a written policy that does not impose this blanket requirement (16%), and agencies that do not have a written policy on body armor (12%). The use of specialized units is captured in three dummy variables, which indicate whether an agency has dedicated personnel for “Special Operations” activities (e.g., SWAT; 40%), anti-gang activities (25%), or anti-terrorism activities (21%). An agency’s guardian tendencies are measured using a scale on community policing activities. The scale is constructed from eight questions about agencies’ community policing practices on the LEMAS survey (α = 0.70; mean = −0.18).⁶ Because these mission variables are derived from secondary survey data, the measures cannot capture nuanced features of each agency’s mission. Nonetheless, these variables permit general classification of LEAs’ warrior or guardian tendencies.

Administrative capacity

Agency administrative capacity is measured using two variables. The first is the number of full-time sworn officers per capita in an agency’s jurisdiction (×100) (mean = 0.21). The second is the number of full-time nonsworn employees per capita (×100) (mean = 0.05).

Fiscal strain

We measure fiscal strain using several variables from LEMAS. One measure of fiscal strain is the agency operating budget per officer (full-time sworn). We apply the natural log to reduce skew. The mean value is 11.54 (or US $102,744). We also measure fiscal strain with three dummy variables indicating difficult budget times: whether the agency experienced a hiring freeze (15% had), salary reduction (3% had), or employee furloughs (6% had) in the three previous years. We also consider external sources of revenue. Agencies that collect money from citizens through asset forfeiture or fees may be doing so to offset declining revenue from a central budget. We include dummy variables indicating whether an agency reported using asset forfeiture or fees as sources of revenue for operational expenses in FY2013. Roughly 35% and 18% of agencies reported these sources of revenue, respectively.

Agency demographics

We consider several agency-level demographic categories: education, veteran status, gender, and race. Education is measured with a dummy variable indicating whether an agency requires some college for new full-time sworn hires (mean = 0.13). The level of veterans on the force is measured with a categorical variable that indicates whether the agency allows a military service alternative to stated educational requirements for new hires. Three types of agencies exist in the data: those that do offer a military service alternative (15%), those that do not (58%), and those that do not have any education requirements and thus no stated policy on alternatives (27%). We measure the gender composition of agencies at three levels: the percent of all full-time sworn officers who are male (% male officers; mean = 92%), the percent of intermediate supervisors (next to chief) who are male (% male intermediate supervisors; mean = 94%), and a dummy variable representing agencies with a male chief executive (mean = 0.98). The race of agency staff is measured as the percentage of full-time sworn officers who are Black (% Black officers). The average LEA reported that just over 4% of its officers were Black.

Community demographics

The size of the Black population is measured as the Black share of the population in the county housing the LEA (% Black county). The average LEA operated in a county in which 9% of residents were Black. To assess the racial mismatch between the community and the police force, we include an interaction term between % Black officers and % Black county. Community racial segregation is measured using a Black–White dissimilarity index (B:W dissimilarity). The resulting measure captures dissimilarity across tracts within a county. It can be interpreted as “the percentage of all individuals who would have to transfer among units [tracts, in this case] to equalize the group proportions across units, divided by the percentage who would have to transfer if the system started in a state of complete segregation” (Reardon & Firebaugh, 2002, p. 42). The dissimilarity index ranges from 0 (no segregation) to 1 (complete segregation) and has a mean of 0.48. It was calculated in Stata with the seg command (Reardon, 2002).

Control variables

The models given later control for the natural log of the 2012 jurisdiction population size and whether the LEA was a county agency (vs. municipal). The models additionally control for the violent crime rate using agency-reported Uniform Crime Reports data from 2009 (U.S. Department of Justice, 2011).^7,8

Analysis

Next, we present results from linear probability models estimating agencies’ MRAP acquisition in the 2013 to 2016 period.⁹ We employ state fixed effects in order to account for unobserved state-to-state heterogeneity in propensity to acquire an MRAP, which is likely to exist because each state has its own LESO coordinator and because entire states may be restricted in their 1033 acquisitions due to malfeasance (Else, 2014). Sampling weights are applied to account for the stratified sampling design of the LEMAS survey. Estimates are generated using the imputed data (described earlier), with coefficients averaged across 30 imputed data sets and standard errors calculated using Rubin’s rules, which incorporate uncertainty due to the imputation procedure (see StataCorp LP, 2013).¹⁰

Results

The regression model provides support for several hypotheses (see Table 1). Agency mission matters, with warrior agencies being significantly more likely to acquire MRAPs. A mandate to wear body armor at all times in the field and, especially, the existence of a special operations (e.g., SWAT) team were significantly associated with MRAP acquisitions. Agencies with a special operations team were 4.6 percentage points more likely to acquire an MRAP, while agencies with a stringent body armor requirement were 1.9 percentage points more likely to acquire an MRAP. The existence of a dedicated anti-terrorism or anti-gang unit was unrelated to MRAP acquisitions, which is perhaps surprising given that the 1033 program names terrorism (and drug) mitigation activities as priorities for equipment transfer. Guardian characteristics—measured with a community policing scale–were unrelated to MRAP acquisitions. This null finding suggests perhaps that agencies may view MRAPs as specialized, limited-use forms of equipment, which are unrelated to day-to-day interactions with the community.

Table 1.

Results From Linear Probability Models With State Fixed Effects.

Results from linear probability model with state fixed effects	b	[SE]
Agency mission
Armor required all times (vs. armor not required all times)?	0.019*	[0.009]
No armor policy (vs. armor not required all times)?	0.01	[0.011]
Special operations (SWAT) unit?	0.046**	[0.010]
Anti-gang unit?	0.022	[0.014]
Anti-terrorism unit?	−0.004	[0.013]
Community policing scale	−0.009	[0.007]
Administrative capacity
Sworn officers (per 100 pop.)	0.01	[0.017]
Non-sworn employees (per 100 pop.)	−0.002	[0.057]
Fiscal strain
Budget per officer (ln)	−0.013	[0.012]
Asset forfeiture as revenue?	0.020*	[0.009]
Fees as revenue?	−0.008	[0.011]
Hiring freeze?	−0.011	[0.010]
Salary reductions?	−0.003	[0.023]
Furloughs?	0.002	[0.017]
Agency demographics
Require some college?	−0.011	[0.012]
Military alternative to educ. (vs. no military alternative)?	−0.018	[0.010]
No education requirement (vs. no military alternative)?	−0.014	[0.008]
% male officers	0.001	[0.000]
% male intermediate supervisors	0	[0.000]
Male chief executive?	0.025	[0.020]
% Black officers	−0.002*	[0.001]
Community demographics
% Black county	−0.001*	[0.000]
% Black officers × % Black county	0	[0.000]
B:W dissimilarity in county	0.027	[0.023]
State dummies included?	Yes
N (unweighted)	2,740
N (weighted)	13,961.197

Note. Model includes an intercept and control for sheriff (vs. municipal police), violent crime rate in 2009 (ln), and jurisdiction population in 2012 (ln). SWAT = Special Weapons And Tactics.

*p < .05. **p < .001.

The administrative capacity hypothesis received no support; MRAP acquisition was not associated with the number of sworn officers or civilian employees per capita. The fiscal strain hypothesis received limited support. Agencies using asset forfeiture for revenue were two percentage points more likely to acquire an MRAP. However, MRAP acquisition was not associated with operating budgets per sworn officer, the use of fees for revenue, hiring freezes, employee furloughs, or budget reductions.

There is some evidence that agency demographics were related to MRAP acquisitions. Specifically, agencies with more Black officers were less likely to receive an MRAP. This relationship is small—a ten percentage point increase in Black officers equated to a two percentage point decrease in the probability of getting an MRAP—but significant (p < .05). None of the gender variables—in sworn officers, intermediate supervisors, or chief executives—was associated with MRAP acquisition. In addition, education (proxied by a college experience requirement for new hires) and veteran status (proxied by a stated preference for applicants with military experience) were also not significant.

There is some evidence that racial demographics in the surrounding community were related to MRAP acquisition. Consistent with prior work (Ajilore, 2015), agencies in counties with larger shares of Black residents were less likely to receive an MRAP. As with the Black share of sworn officers, this Black community effect is statistically significant but substantively small—a ten percentage point increase in the Black share of a county corresponded to a roughly one percentage point drop in the probability of getting an MRAP. This association was not mediated by the Black share of the sworn officer population, as indicated by the nonsignificant interaction term. Contrary to Ajilore (2015), the present analysis fails to detect evidence of a relationship between racial segregation and MRAP acquisition. The discrepancy may be due to differences in the data or measures. Whereas Ajilore used counties as the unit of analysis and examined 9 months in 2013 and 2014, this study uses LEAs as the unit of analysis and examines 4 years from 2013 through 2016. In addition, this study incorporates variables not included in Ajilore’s study (namely, agency-level variables), the omission of which may have biased the earlier results.

Sensitivity Tests

We assess the robustness of these findings through several additional analyses (see Table 2). First, racial threat theories often posit a curvilinear relationship between minority population size and government response, such that the perceived racial threat increases with minority population size but then declines once the group has become large enough to exert influence (Keen & Jacobs, 2009; Preuhs, 2001; Stults & Baumer, 2007). To examine this possibility in the context of MRAP acquisitions, we include a quadratic term for the Black share of the county. We expect that larger Black populations will be positively associated with MRAP acquisitions, but only up to a point. After this point, the relationship will become negative, perhaps due to a greater Black political voice to oppose police militarization. (To ease interpretation, this model removes the interaction term between Black shares in the county and in the police force, which was found to be nonsignificant in the previous model.) The results of this model are substantively similar to the main model (see Model 1 in Table 2). The one exception is that the (already small) association between Black officers and MRAP acquisitions is no longer statistically significant.

Table 2.

Results From Sensitivity Tests.

	(1)		(2)		(3)		(4)
	Any MRAP?		Any MRAP?		Any MRAP before Ferguson?		Any MRAP after Ferguson?
Results from sensitivity tests	b	SE	b	SE	b	SE	b	SE
Agency mission
Armor required all times (vs. armor not required all times)?	0.019*	[0.009]	0.019*	[0.009]	0.011	[0.008]	0.007	[0.004]
No armor policy (vs. armor not required all times)?	0.011	[0.011]	0.009	[0.012]	−0.005	[0.009]	0.014	[0.008]
Special operations (SWAT) unit?	0.046**	[0.010]	0.047**	[0.010]	0.034**	[0.008]	0.013*	[0.005]
Anti-gang unit?	0.022	[0.014]	0.024	[0.015]	0.023	[0.012]	−0.003	[0.008]
Anti-terrorism unit?	−0.004	[0.013]	−0.002	[0.015]	−0.009	[0.012]	0.011	[0.008]
Other special units scale			−0.009	[0.021]
Received MRAP 2013–2014?							0.032	[0.022]
Community policing scale	−0.009	[0.007]	−0.008	[0.007]	−0.002	[0.006]	−0.008	[0.004]
Administrative capacity
Sworn officers (per 100 pop.)	0.008	[0.018]	0.011	[0.017]	0.003	[0.015]	0.007	[0.008]
Non-sworn employees (per 100 pop.)	0.001	[0.059]	−0.002	[0.057]	−0.032	[0.048]	0.026	[0.033]
Fiscal strain
Budget per officer (ln)	−0.013	[0.012]	−0.013	[0.012]	−0.014	[0.011]	−0.001	[0.007]
Asset forfeiture as revenue?	0.020*	[0.009]	0.020*	[0.009]	0.019*	[0.008]	0.003	[0.005]
Fees as revenue?	−0.008	[0.011]	−0.008	[0.011]	−0.008	[0.010]	0	[0.007]
Hiring freeze?	−0.011	[0.010]	−0.011	[0.010]	−0.018*	[0.009]	0.006	[0.007]
Salary reductions?	−0.003	[0.023]	−0.003	[0.023]	−0.024	[0.020]	0.017	[0.013]
Furloughs?	0.003	[0.017]	0.002	[0.017]	0.008	[0.016]	−0.009	[0.008]
Agency demographics
Require some college?	−0.011	[0.012]	−0.011	[0.012]	−0.008	[0.011]	0	[0.006]
Military alternative to education (vs. no military alternative)?	−0.018	[0.010]	−0.018	[0.010]	−0.016	[0.009]	−0.001	[0.006]
No educ. requirement (vs. no military alternative)?	−0.014	[0.008]	−0.014	[0.008]	−0.01	[0.007]	−0.003	[0.004]
% male officers	0	[0.000]	0.001	[0.000]	0	[0.000]	0	[0.000]
% male intermediate supervisors	0	[0.000]	0	[0.000]	0	[0.000]	0	[0.000]
Male chief executive?	0.024	[0.020]	0.026	[0.020]	0.021	[0.019]	0.003	[0.008]
% Black officers	−0.001	[0.000]	−0.002*	[0.001]	−0.002*	[0.001]	0	[0.000]
Community demographics
% Black county	−0.003*	[0.001]	−0.001*	[0.000]	−0.001	[0.000]	0	[0.000]
% Black county²	0	[0.000]
% Black officers × % Black county			0	[0.000]	0	[0.000]	0	[0.000]
B:W dissimilarity in county	0.029	[0.023]	0.027	[0.023]	0.025	[0.020]	0.007	[0.012]
State dummies included?	Yes		Yes		Yes		Yes
N (unweighted)	2,740		2,740		2,740		2,740
N (weighted)	13,961.197		13,961.197		13,961.197		13,961.197

Note. Results are from linear probability models with state fixed effects. All models include an intercept and control for sheriff (vs. municipal police), violent crime rate in 2009 (ln), and jurisdiction population in 2012 (ln). SWAT = Special Weapons And Tactics; MRAP = mine-resistant ambush-protected vehicle.

²denotes that the measure (% Black county) has been squared.

*p < .05. **p < .001.

Although the main model indicates that MRAPs are associated with SWAT teams, it is possible that this relationship says more about an agency’s capacity to form special issue task forces than about its warrior tendencies. LEMAS contains data on a number of task forces addressing other special topics that might not fall under the warrior umbrella. To adjudicate between the warrior aspect and the capacity aspect, we present a supplemental analysis that controls for other special issue task forces reported by the agency. We create a scale based on whether the agency reported having a special unit for the following issues: bias or hate crimes, bomb disposal, child abuse, cybercrime, intimate partner violence, human trafficking, impaired driving, juvenile crime, reentry surveillance, fugitives and warrants, and victim assistance (Cronbach’s α = .87). This new “other special units scale” is not itself a significant predictor of MRAP acquisition, and it does not alter the main results in any meaningful way (see Model 2 in Table 2).

Two additional analyses examine changes in the pattern of MRAP acquisitions over time. The 2014 Ferguson protests and the law enforcement response marked a turning point in police militarization. Military equipment and tactics received widespread public attention, which prompted President Obama to review the 1033 program. As seen in Figure 2, MRAP shipments stopped in September 2014 and did not resume until January 2016. Given this de facto moratorium, it is possible that LEAs reconsidered their need for MRAPs in light of recent protests in Ferguson and other cities. To assess whether factors associated with MRAP acquisition changed in the wake of Ferguson, we reran the main model on two new binary dependent variables: MRAP acquisition before Ferguson (2013–2014) and after Ferguson (2015–2016). The models were estimated using the same specifications as earlier, with the exception that the latter model controls for whether an agency received an MRAP before the Ferguson protests.

Model 3 shows that findings from the main analysis are partially replicated in the analysis of MRAP acquisitions before Ferguson: Agencies with special operations units, with fewer Black officers, and those using asset forfeiture to generate revenue were significantly more likely to acquire an MRAP. Unlike the main analysis, this one suggests that MRAP acquisitions before Ferguson were associated with agency hiring freezes but not with body armor policies or Black county populations. After Ferguson (Model 4), the prior pattern of MRAP acquisitions disappears. The only substantive factor associated with MRAP acquisition in this period was having a special operations unit. Even this reliably significant variable decreased in magnitude and statistical significance (b = 0.013, p < .05). There are two explanations for the unpredictable quality of MRAP acquisition after Ferguson. First, the number of MRAP shipments after Ferguson was not large enough for the model to detect a pattern. Only 77 agencies in the data received an MRAP after Ferguson (compared to 212 before Ferguson). The second possibility is that agencies changed their stance toward MRAPs. With the 1033 program facing increased public scrutiny, agencies with a warrior mission may have reconsidered their adoption of military gear, including MRAPs. Moreover, the new accountability and training standards for “controlled” equipment (including MRAPs) pursuant to Executive Order 13688 may have deterred some agencies from acquiring militarized equipment (Law Enforcement Equipment Working Group, 2015).

Discussion

MRAPs were initially deployed for foreign occupations in Iraq and Afghanistan. Now they are deployed for domestic law enforcement in places like Idaho and Alabama. This analysis identifies a number of factors associated with LEAs’ acquisition of MRAPs. In brief, MRAPs are most likely to be acquired by agencies with warrior tendencies (namely, a SWAT team and robust body armor policies) and those who rely on seized assets for agency revenue. The analysis also finds some evidence of an association between MRAPs and agencies with fewer Black officers and Black residents, although these results are less consistent across models.

The significance of warrior mission and asset forfeiture suggests a symbiotic relationship between police militarization and revenue generation. Prior research has shown that SWAT teams, which originated as tactical teams responding to immediate threats to public order, are now used primarily to execute search warrants (ACLU, 2014a; Klinger & Rojek, 2008; Kraska & Cubellis, 1997). These searches have the effect (and, in many cases, the intention) of collecting guns, drugs, and money. Using lucrative asset forfeiture laws, the LEA can direct the monetary value of these seized items into their coffers (Holcomb et al., 2011). An agency acquiring MRAPs through the 1033 program can thus be doubly rewarded: The MRAP costs nothing to acquire, and it can be used to bolster a SWAT team that can seize valuable assets to fund the agency. A limitation of our analysis is that we do not know how LEAs deployed their MRAPs. For example, one alternative use of MRAPs is rescue missions. Anecdotal evidence indicates that several LEAs in the Houston, Texas, area used 1033-acquired MRAPs (and other vehicles) to rescue citizens trapped by flood waters after Hurricane Harvey (Blakinger, 2018). Thus, while we cannot definitively say that MRAPs are being deployed for the purpose of asset seizure and revenue generation, the relationship between MRAP acquisition and both SWAT teams and asset forfeiture in a nationally representative sample of LEAs is consistent with that explanation.

Although MRAPs disproportionately went to LEAs with SWAT teams, they were no more (or less) likely to go to LEAs with counter-terrorism units. This is surprising, as the 1033 program (as authorized in the 1997 NDAA) was explicitly intended to support counter-terror activities. One explanation for this null finding is that LEAs do not view MRAPs as a critical component in their fight against terrorism. Previous research on terrorism preparedness suggests that information sharing, community engagement, and organization restructuring—but not MRAPs or other vehicle acquisitions—are common terrorism preparedness actions among local LEAs (Ortiz, Hendricks, & Sugie, 2007; Roberts, Roberts, & Liedka, 2012). A supplemental analysis (available upon request) of a sample of 1033 applications collected by Redden (2015) shows that terrorism was only mentioned by 7% of LEAs when requesting MRAPs. The low number suggests that terrorism is not a salient issue when demonstrating a need for an MRAP. Future work should investigate the extent to which terrorism preparedness motivates LEAs to acquire any other type of equipment through the 1033 program.

Our findings offer no evidence that MRAPs are concentrated in areas with large Black populations. In fact, we find some evidence of a negative relationship between MRAPs and Black populations. This counters the striking images from Ferguson, in which militarized police and armored vehicles aggressively asserted their presence in a majority-Black city. Prior research demonstrates that high-profile, violent police actions in minority communities can have detrimental effects on minority trust in and cooperation with police (Desmond, Papachristos, & Kirk, 2016; Kochel, 2017). One explanation of our findings is that police executives in largely Black jurisdictions, on average, wish to avoid confrontational and controversial tactics that might threaten (perhaps already fraught) citizen–police relations. These relations are a focal point of concern for police executives throughout the country (Police Executive Research Foundation, 2016). Foregoing the acquisition (and use) of military vehicles may be one discrete and relatively easy way of avoiding needless controversy. Future research would benefit from a qualitative investigation of how police chiefs’ conceive of MRAPs and other military equipment in law enforcement operations.

The disjuncture between MRAP acquisitions before and after Ferguson also warrants further attention from researchers. After Ferguson, the familiar pattern of MRAP distributions (to warrior LEAs using asset seizure) largely vanished, and MRAP distribution followed a relatively unpredictable pattern. As noted earlier, police managers in warrior agencies may have altered their MRAP acquisition behavior in response to the Ferguson protests (and the attendant public scrutiny of police militarization) or the new accounting and training practices mandated by Executive Order 13688. On the first view, senior police officers became increasingly concerned over the public perception of their forces and worried that acquisition of MRAPs would further undercut police legitimacy in a post-Ferguson world. On the second view, managers balked at the prospect of new training and accounting standards, which could impose additional monetary or opportunity costs for the agencies.¹¹ Qualitative data from police executives could reveal the extent to which either of these issues has forced agencies to reconsider their use of MRAPs.

Overall, the pattern of MRAP acquisitions by LEAs observed here is consistent with a model of governance that is both oppositional and predatory (Page & Soss, 2016). Citizens are viewed as threats to be mitigated (with body armor and SWAT teams) and in terms of opportunities to be seized (through asset forfeiture). Resource-poor agencies may turn to citizens to replenish their coffers (Baicker & Jacobson, 2007; Harvard Law Review, 2015). People convicted (or even suspected) of a crime are uniquely susceptible to this resource extraction, as they are deemed to have violated the social contract (Page & Soss, 2018). The effects of this extraction may even extend to family and friends, who offer financial assistance to those in legal jeopardy (Katzenstein & Waller, 2015).

The present article shows how militarization coexists with predatory, revenue-driven policing (ACLU, 2014b; Balko, 2013; Kraska & Cubellis, 1997). Militarized and predatory policing risks eroding the legitimacy of police and reducing the willingness of citizens to cooperate in policing. Ferguson represented the confluence of these trends, with municipal agencies relying on revenue from disenfranchised citizens and a heavily weaponized response to public protest (Page & Soss, 2016; U.S. Department of Justice, 2015). Moreover, predatory policing may be counterproductive to the public safety mission of law enforcement. For example, recent work has found that LEAs that rely heavily on fines and fees from citizens have lower violent crime clearance rates (Goldstein, Sances, & You, 2018).

Reforming these practices will require more than simply a prohibition on the use of MRAPs (or other military equipment) in law enforcement. Broader reforms that entail adequate funding for LEAs would negate the need for resource extraction and would allow police to focus on public safety. Ultimately, the primary concern of domestic law enforcement should be public service. A return to this focus—and away from militarization and revenue generation—would go some way toward repairing frayed relations between police and citizens.

Footnotes

Appendix

Authors’ Note

Elizabeth Schroeder and Paul Thompson provided useful advice on the analysis. Earlier versions of this article were presented at the American Society of Criminology in 2017 and the Oregon State University School of Public Policy in 2017. All remaining errors are our own.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biographies

Brett C. Burkhardt is an associate professor of sociology in the School of Public Policy at Oregon State University. His research examines the exercise of coercive power by legal and quasi-legal authorities. He is currently involved in four projects: (a) prison privatization, (b) policing mental illness, (c) police militarization, and (d) police use of social media. He holds a PhD degree in sociology from the University of Wisconsin-Madison.

Keith Baker is an associate professor in the Department of Public Administration, The College at Brockport, State University of New York. Keith’s research focuses on the organization and governance of state/society relationships. Keith has previously held positions at Oregon State University and Northumbria University.

References

Acock

A. C.

(2005). Working with missing values. Journal of Marriage and Family, 67, 1012–1028.

Acock

A. C.

(2016). A gentle introduction to Stata (5th ed.). College Station, TX: Stata Press.

Ajilore

(2015). The militarization of local law enforcement: Is race a factor? Applied Economics Letters, 22(13), 1089–1093. doi:10.1080/13504851.2014.1002884

Alexander

(2010). The new Jim Crow: Mass incarceration in the age of colorblindness. New York, NY: New Press.

American Civil Liberties Union. (2010). In for a Penny: The Rise of America's New Debtors' Prisons. Retrieved from https://www.aclu.org/report/penny-rise-americas-new-debtors-prisons

American Civil Liberties Union. (2014b). Warehoused and forgotten: Immigrants trapped in our shadow private prison system. Retrieved from https://www.aclu.org/sites/default/files/assets/060614-aclu-car-reportonline.pdf

Baicker

Jacobson

(2007). Finders keepers: Forfeiture laws, policing incentives, and local budgets. Journal of Public Economics, 91(11–12), 2113–2136. doi:10.1016/j.jpubeco.2007.03.009

Balko

(2013). Rise of the warrior cop: The militarization of America’s police forces. New York, NY: Public Affairs.

Baumgart

(2016). From copper to steel: Police militarization at the end of the 20th century. Open Science Framework doi:10.17605/OSF.IO/M7J3V

10.

Beckett

Harris

(2011). On cash and conviction. Criminology & Public Policy, 10(3), 509–537. doi:10.1111/j.1745-9133.2011.00726.x

11.

Berube

Holmes

(2016). Minority under-representation in city and suburban policing. Retrieved from https://www.brookings.edu/blog/the-avenue/2016/07/14/minority-under-representation-in-city-and-suburban-policing/

12.

Bieler

(2016). Police militarization in the USA: The state of the field. Policing: An International Journal of Police Strategies & Management, 39(4), 586–600. doi:10.1108/PIJPSM-03-2016-0042

13.

Blakinger

(2018). Military surplus for U.S. cops “saved lives” after hurricane: Houston chief. Retrieved from https://thecrimereport.org/2018/01/12/military-surplus-for-u-s-cops-saved-lives-after-hurricane-says-houston-chief/

14.

Bove

Gavrilova

(2017). Policeman on the frontline or a soldier? The effect of police militarization on crime. American Economic Journal: Economic Policy, 9(3), 1–18. doi:10.2139/ssrn.2678967

15.

Campbell

D. J.

Campbell

K. M.

(2010). Soldiers as police officers/police officers as soldiers: Role evolution and revolution in the United States. Armed Forces & Society, 36(2), 327–350. doi:10.1177/0095327X09335945

16.

Carter

J. S.

Corra

(2016). Racial resentment and attitudes toward the use of force by police: An over-time trend analysis. Sociological Inquiry, 86(4), 492–511. doi:10.1111/soin.12136

17.

Defense Logistics Agency. (n.d.). 1033 program FAQs. Retrieved from http://www.dla.mil/DispositionServices/Offers/Reutilization/LawEnforcement/ProgramFAQs.aspx

18.

Defense Logistics Agency. (2017). Law Enforcement Support Office. Retrieved from http://www.dla.mil/DispositionServices/Offers/Reutilization/LawEnforcement.aspx

19.

Delehanty

Mewhirter

Welch

Wilks

(2017). Militarization and police violence: The case of the 1033 program. Research & Politics, 4(2), 1–7. doi:10.1177/2053168017712885

20.

Desmond

Papachristos

A. V.

Kirk

D. S.

(2016). Police violence and citizen crime reporting in the Black community. American Sociological Review, 81(5), 857–876. doi:10.1177/0003122416663494

21.

Ekins

(2016). Policing in America: Understanding public attitudes toward the police. Results from a National Survey. Retrieved from https://www.cato.org/publications/working-paper/policing-america-understanding-public-attitudes-toward-police-results

22.

Else

D. H.

(2014). The “1033 Program,” Department of Defense Support to Law Enforcement. Washington, DC: Congressional Research Service.

23.

Goldstein

Sances

M. W.

You

H. Y.

(2018). Exploitative revenues, law enforcement, and the quality of government service. Urban Affairs Review. doi: 10.1177/1078087418791775

24.

Harris

M. C.

Park

Bruce

D. J.

Murray

M. N.

(2017). Peacekeeping force: Effects of providing tactical equipment to local law enforcement. American Economic Journal: Economic Policy, 9(3), 291–313. doi:10.1257/pol.20150525

25.

Harvard Law Review. (2015). Policing and profit. Harvard Law Review, 128, 1723–1746.

26.

Hayes

(2017). A colony in a nation. New York, NY: W.W. Norton & Company.

27.

Holcomb

J. E.

Kovandzic

T. V.

Williams

M. R.

(2011). Civil asset forfeiture, equitable sharing, and policing for profit in the United States. Journal of Criminal Justice, 39(3), 273–285. doi:10.1016/j.jcrimjus.2011.02.010

28.

Institute for Intergovernmental Research. (2015). After-action assessment of the police response to the August 2014 Demonstrations in Ferguson, Missouri. Retrieved from https://ric-zai-inc.com/ric.php?page=detail&id=COPS-P317

29.

Johnson

T. C.

Hansen

J. A.

(2016). Law enforcement agencies’ participation in the military surplus equipment program Thomas. Policing: An International Journal of Police Strategies & Management, 39(4), 791–806. doi:10.1007/s10499-012-9533-5

30.

Kappeler

V. E.

Kraska

P. B.

(2013). Normalising police militarisation, living in denial. Policing and Society, 25(3), 268–275. doi:10.1080/10439463.2013.864655

31.

Katzenstein

M. F.

Waller

M. R.

(2015). Taxing the poor: Incarceration, poverty governance, and the seizure of family resources. Perspectives on Politics, 13(3), 638–656. doi:10.1017/S153759271500122X

32.

Keen

Jacobs

(2009). Racial threat, partisan politics, and racial disparities in prison admissions: A panel analysis. Criminology, 47(1), 209–238. doi:10.1111/j.1745-9125.2009.00143.x

33.

King

R. D.

Johnson

B. D.

(2016). A punishing look: Skin tone and afrocentric features in the Halls of Justice. American Journal of Sociology, 122(1), 90–124.

34.

Klein

(2013, February 24). MIVIF: Stata module to calculate variance inflation factors after mi estimate regress. Retrieved from https://ideas.repec.org/c/boc/bocode/s457254.html

35.

Klinger

D. A.

Rojek

(2008). Multi-method study of Special Weapons And Tactics teams. Washington, DC: National Institute of Justice.

36.

Kochel

T. R.

(2017). Explaining racial differences in Ferguson’s impact on local residents’ trust and perceived legitimacy. Criminal Justice Policy Review. doi:10.1177/0887403416684923

37.

Kochel

T. R.

Wilson

D. B.

Mastrofski

S. D.

(2011). Effect of suspect race on officers’ arrest decisions. Criminology, 49(2), 473–512. doi:10.1111/j.1745-9125.2011.00230.x

38.

Kraska

P. B.

Cubellis

L. J.

(1997). Militarizing mayberry and beyond: Making sense of American paramilitary policing. Justice Quarterly, 14(4), 607–629. doi:10.1080/07418829700093521

39.

Kraska

P. B.

Kappeler

V. E.

(1997). Militarizing American police: The rise and normalization of paramilitary units. Social Problems, 44(1), 1–18. doi:10.1525/sp.1997.44.1.03x0209a

40.

Law Enforcement Equipment Working Group. (2015). Recommendations pursuant to executive order 13688: Federal support for local law enforcement equipment acquisition. Retrieved from https://ojp.gov/docs/LE-Equipment-WG-Final-Report.pdf

41.

Lawson

(2018). Police militarization and the use of lethal force. Political Research Quarterly. doi:10.1177/1065912918784209

42.

Lockwood

Doyle

M. D.

Comiskey

J. G.

(2018). Armed, but too dangerous? Factors associated with citizen support for the militarization of the police. Criminal Justice Studies. doi:10.1080/1478601x.2017.1420652

43.

Long

J. S.

Freese

(2014). Regression models for categorical dependent variables using Stata (3rd ed.). College Station, TX: Stata Press.

44.

Marenin

(2016). Cheapening death: Danger, police street culture, and the use of deadly force. Police Quarterly, 19(4), 461–487. doi:10.1177/1098611116652983

45.

Morin

Parker

Stepler

Mercer

(2017). Behind the badge. Retrieved from http://www.pewsocialtrends.org/2017/01/11/behind-the-badge/

46.

Moule

R. K.

Fox

B. H.

Parry

M. M.

(2018). The long shadow of Ferguson: Legitimacy, legal cynicism, and public perceptions of police militarization. Crime & Delinquency. doi:10.1177/0011128718770689

47.

Neocleous

(2014). War power, police power. Edinburgh, England: Edinburgh University Press.

48.

Nix

Wolfe

S. E.

Campbell

B. A.

(2018). Command-level police officers’ perceptions of the “war on cops” and de-policing. Justice Quarterly, 35(1), 33–54. doi:10.1080/07418825.2017.1338743

49.

Office of the Under Secretary of Defense. (n.d.). Population representation in the military services: Fiscal year 2014 summary report. Retrieved from http://www.people.mil/Portals/56/Documents/2014Summary.pdf?ver=2016-09-14-154051-563

50.

Ortiz

C. W.

Hendricks

N. J.

Sugie

N. F.

(2007). Policing terrorism the response of local police agencies to homeland security concerns. Criminal Justice Studies, 22(2), 91–109. doi:10.1080/14786010701396830

51.

Page

Soss

(2016). The predator state: Race, class and the new era of indentured citizenship. In 2015–16 Neoliberalism Seminar at CUNY – Hunter College. Retrieved from http://urban.hunter.cuny.edu/∼schram/predatorstate.pdf

52.

Page

Soss

(2018). Criminal justice predation and neoliberal governance. In Schram

S. F.

Pavlovskaya

(Eds.), Rethinking Neoliberalism: Resisting the Disciplinary Regime (pp. 141–161). New York, NY: Routledge.

53.

Peffley

Hurwitz

(2010). Justice in America: The separate realities of Blacks and Whites. New York, NY: Cambridge University Press.

54.

Pew Research Center. (2014). Few say police forces nationally do well in treating races equally. Retrieved from http://www.people-press.org/2014/08/25/few-say-police-forces-nationally-do-well-in-treating-races-equally/

55.

Police Executive Research Foundation. (2016). Advice from police chiefs and community leaders on building trust: “Ask for Help, Work Together, and Show Respect.” Retrieved from http://www.policeforum.org/assets/policecommunitytrust.pdf

56.

President’s Task Force on 21st Century Policing. (2015). Final report of the President’s Task Force on 21st Century Policing. Washington, DC: U.S. Department of Justice, Office of Community Oriented Policing Services. doi:10.1007/s13398-014-0173-7.2

57.

Preuhs

R. R.

(2001). State felon disenfranchisement policy. Social Science Quarterly, 82(4), 733–748.

58.

Radil

S. M.

Dezzani

R. J.

McAden

(2017). The road to Ferguson: Geographies of U.S. police militarization and the role of the 1033 program. The Professional Geographer, 69(2), 203–213. doi:10.13140/rg.2.1.2589.3604

59.

Reardon

S. F.

(2002). SEG: Stata module to compute multiple-group diversity and segregation indices. Retrieved from http://econpapers.repec.org/software/bocbocode/s375001.htm

60.

Reardon

S. F.

Firebaugh

(2002). Measures of multigroup segregation. Sociological Methodology, 32(1), 33–67. doi:10.1111/1467-9531.00110

61.

Redden

, (2015). Documents reveal the fearmongering local cops use to score military gear from the pentagon. Retrieved from https://www.motherjones.com/politics/2015/08/new-documents-reveal-fearmongering-local-cops-use-score-military-gear-pentagon/

62.

Roberts

J. M.

Liedka

R. V.

(2012). Elements of terrorism preparedness in local police agencies, 2003-2007. Crime & Delinquency, 58(5), 720–747. doi:10.1177/0011128712452960

63.

Rydberg

Terrill

(2010). The effect of higher education on police behavior. Police Quarterly, 13(1), 92–120. doi:10.1177/1098611109357325

64.

StataCorp LP. (2013). Stata multiple-imputation reference manual: Release 13. College Station, TX: Author.

65.

StataCorp

LP.

(2016). Stata: Data analysis and statistical software. Retrieved from http://www.stata.com/

66.

Stoughton

S. W.

(2016). Principled policing: Warrior cops and guardian officers. Wake Forest Law Review, 51, 611–676.

67.

Stults

B. J.

Baumer

E. P.

(2007). Racial context and police force size: Evaluating the empirical validity of the minority threat perspective. American Journal of Sociology, 113(2), 507–546. doi:10.1086/518906

68.

Terrill

Mastrofski

S. D.

(2002). Situational and officer-based determinants of police coercion. Justice Quarterly, 19(2), 215–248. doi:10.1080/07418820200095221

69.

Turner

F. W.

Fox

B. H.

(2017). Public servants or police soldiers? An analysis of opinions on the militarization of policing from police executives, law enforcement, and members of the 114th Congress U.S. House of Representatives. Police Practice and Research. doi:10.1080/15614263.2017.1371600

70.

Vanden Brook, T. (2014). MRAP sale to Emirates may enable regional military force. Retrieved from https://www.usatoday.com/story/news/world/2014/10/13/uae-mrap-peacekeeping/16965925/

71.

U.S. Army. (2018). Mine resistant ambush protected vehicles (MRAP). Retrieved from https://asc.army.mil/web/portfolio-item/cs-css-mine-resistant-ambush-protected-mrap-vehicle-mrap/

72.

U.S. Census Bureau. (n.d.). American community survey (ACS). Retrieved from https://www.census.gov/programs-surveys/acs/

73.

U.S. Department of Justice. (2011). Uniform crime reporting program data: Offenses known and clearances by arrest, 2009. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor]. Retrieved from doi.org/10.3886/ICPSR30766.v1

74.

U.S. Department of Justice. (2015). Investigation of the Ferguson police department. Washington, DC: Author.

75.

Weichselbaum

Schwartzapfel

(2017). When warriors put on the badge. Retrieved from https://www.themarshallproject.org/2017/03/30/when-warriors-put-on-the-badge#.7yzjyxHbn

76.

Wester

M. T.

(2016). Drawing a line between Rambo & Barney Fife: Overhauling the Department of Defense’s excess property program in order to halt the overmilitarization of America’s police forces. University of Illinois Law Review, 2016(2), 729–760.

77.

Wozniak

K. H.

(2016). Ontological insecurity, racial tension, and confidence in the police in the shadow of urban unrest. Sociological Forum, 31(4), 1063–1082. doi:10.1111/socf.12296

78.

YouGov. (2014). Poll results: Police. Retrieved from https://today.yougov.com/topics/politics/articles-reports/2014/08/19/poll-results-police