Reference TylerTyler's (2006) process-based model of policing is now a staple of American research on public perceptions and community relations with law enforcement. This is best illustrated by the prominent place of the model in the President's Task Force on 21st Century Policing's (2015) final report. This task force, initiated by President Obama in the aftermath of civil unrest in major American cities, emphasized the importance of trust and civil relationships between police and the communities they serve. The core of the process-based model holds that the actions of the police influence the behaviors and perceptions of the public in two stages (Reference MazerolleMazerolle et al. 2013). First, procedurally just treatment of the citizenry by authorities enhances the legitimacy of those authorities. Second, legitimacy should encourage voluntary citizen compliance with the law as well as their cooperation with police (Reference TylerTyler 2006; Reference Tyler and HuoTyler and Huo 2002). A robust body of research spanning psychology, management, and criminology has assessed the underlying assumptions of the process-based model, finding widespread support for the theory (Reference TylerTyler 2017). Indeed, this body of evidence has led some researchers to characterize the “front end” of the process-based model—the links between procedural justice and legitimacy—as a “well-trodden path” (Reference MazerolleMazerolle et al. 2013: 34).
The problem, however, is that while the core components of the process-based model have been assessed and supported among Americans (see Reference TylerTyler 2003, Reference Tyler2004, Reference Tyler2006, Reference Tyler2017), other aspects of the theory remain under-examined. This is especially true regarding the consequences of legitimacy. In addition to cooperation and compliance with the law and its agents, Reference Sunshine and TylerSunshine and Tyler (2003) proposed an “empowerment hypothesis,” where the public exhibits a greater willingness to grant police more discretion to enforce the law. According to this hypothesis, willingness is a function of legitimacy; as perceptions of legitimacy increase, so too should this willingness. Understanding public willingness to empower the police is particularly timely, given the increasingly contentious and visible divide in the United States between police and the communities they serve (President's Task Force on 21st Century Policing 2015). Chief among these issues is the militarization of the police (Reference BalkoBalko 2013; Reference KraskaKraska 2007; Reference PagePage 2014).
Police militarization refers to the process by which police agencies take on more and more characteristics of the military, including appearance, behavior, and use of surplus military equipment (Reference BalkoBalko 2013; Reference KraskaKraska 2007). In the wake of civil unrest, and the subsequent police response, seen in Ferguson, Missouri, and elsewhere over the past few years, debates about police militarization have arisen and persisted. These debates emphasized the possible consequences of militarization. Assessments of these possible consequences reflect rational concerns on the part of stakeholders and practitioners (Reference Turner and FoxTurner and Fox 2017). Some have argued militarization is a necessary development that will increase the effectiveness of law enforcement, helping officers fight crime and maintain public safety (Reference MadhaniMadhani 2014). Others still have expressed concern about the harmful implications of militarization for civil liberties (ACLU 2014; Reference LynchLynch 2014). Such concerns are likely to be echoed by members of the American public. We argue perceptions of these possible consequences of police militarization are anchored by legitimacy and, in turn, also influence public willingness to empower police (e.g., Reference Tversky and KahnemanTversky and Kahneman 1974).
To that end, the current study examines the influence of police legitimacy on the American public's willingness to empower the police to become more militarized. Drawing on insights from cognitive psychology and rational choice theory, we then consider whether legitimacy influences the perceived consequences of militarization, and whether these perceived consequences act as indirect pathways linking legitimacy and empowerment. Using a national sample of 702 American adults, and a structural equation modeling (SEM) strategy (Reference BollenBollen 1989; Reference Bowen and GuoBowen and Guo 2011), we seek to address two questions: (1) does legitimacy have a direct effect on public empowerment of police? And (2) does legitimacy exert indirect effects on empowerment through instrumental concerns about the potential consequences of militarization? Our overall goal is to elaborate on how legitimacy influences the American publics' willingness to empower the police. We begin by discussing the process-based model of policing and its consequences.
The Process-Based Model of Policing and its Consequences
Legitimacy refers to the public's views toward legal authorities (Reference Lind and TylerLind and Tyler 1988; Reference Trinkner and CohnTrinkner and Cohn 2014; Reference TylerTyler 2006). Specifically, legitimacy is “a property of an authority or institution that leads people to feel that that authority or institution is entitled to be deferred to and obeyed” (Reference Sunshine and TylerSunshine and Tyler 2003: 514; see also Reference WeberWeber 1968). Legitimacy is necessary, as institutions of governance function with the consent of the governed (Reference SabineSabine 1937; see also Reference LockeLocke 1988). Police legitimacy is specifically an individual normative orientation toward the police, reflecting a combination of trust in law enforcement as well as feelings of obligation to obey the police (Reference ParryParry et al. 2017; Reference ReisigReisig et al. 2007; Reference TylerTyler 2006; Reference WolfeWolfe et al. 2016; but see Reference BarbaletBarbalet 2009; Reference Bottoms and TankebeBottoms and Tankebe 2012; Reference JohnsonJohnson et al. 2014; Reference KainaKaina 2008).
Legitimacy is cultivated by the police in a number of ways, including demonstrating their effectiveness (Reference KochelKochel et al. 2013; Reference TankebeTankebe 2009, Reference Tankebe2013; Reference TaylorTaylor et al. 2015), distributing resources in a fair and equitable way (e.g., Reference EppEpp et al. 2014; Reference TankebeTankebe 2013), respecting the bounds of their lawful authority (Reference HuqHuq et al. 2017), and treating the citizens they come in contact with in a fair and just manner (e.g., Reference Lind and TylerLind and Tyler 1988; Reference TylerTyler 2006, Reference Tyler2017). In Western countries, including America, legitimacy is commonly considered to be generated by police engaging in procedurally just practices (Reference Hinds and MurphyHinds and Murphy 2007; Reference MazerolleMazerolle et al. 2013; Reference TylerTyler, 2006, Reference Tyler2017). When police are perceived as behaving in a procedurally just fashion—being fair, respectful, and courteous toward citizens during interactions they may have, making decisions based on the facts of a given situation, and allowing citizens to have a say in their decision-making process—they are seen as more legitimate by the public (Reference MazerolleMazerolle et al. 2013; Reference TylerTyler 2006, Reference Tyler2017; Reference Tyler and HuoTyler and Huo 2002; Reference WolfeWolfe et al. 2016; Reference Worden and McLeanWorden and McLean 2017a, Reference Worden and McLean2017b).
Legitimacy is particularly important for the police, as law enforcement relies on the voluntary assistance of citizens to maintaining order and public safety (Reference DeckerDecker 1981; Reference Frank, Smith and NovakFrank et al. 2005; Reference Huang, Vaughn, Flanagan and LongmireHuang and Vaughn 1996; Reference ReissReiss 1971). Indeed, legitimacy influences a number of citizen beliefs and behaviors related to these outcomes. When police are seen as more legitimate, citizens are more likely to cooperate with them and comply with the law (e.g., Reference DonnerDonner et al. 2015; Reference JacksonJackson et al. 2012; Reference ReisigReisig et al. 2011). Reference Sunshine and TylerSunshine and Tyler (2003) suggest an additional consequence of legitimacy: public willingness to empower police. This “empowerment hypothesis” holds that as the perceived legitimacy of law enforcement increases, the public is more willing to grant discretion, or a wider latitude, to police to execute their duties (Reference PrycePryce 2016; Reference Sunshine and TylerSunshine and Tyler 2003). This enhanced discretion can involve practices that may be seen controversial by members of the public (e.g., Reference Gau and BrunsonGau and Brunson 2010; Reference White and FradellaWhite and Fradella 2016; see Reference Sunshine and TylerSunshine and Tyler 2003).
While legitimacy and procedural justice have received much attention by researchers, the empowerment hypothesis remains largely unexamined. To our knowledge, only three studies have examined the influence of legitimacy on empowerment, finding tentative support for the relationship. Reference Sunshine and TylerSunshine and Tyler (2003) examined willingness to give police greater autonomy (e.g., conducting “stop and question” stops with members of the public; having the ability to do whatever they feel is necessary to fight crime). Using SEM, and two samples of New York City residents, the pair demonstrated that legitimacy had a moderate positive effect on police empowerment. Among Ghanaian immigrants in Washington, DC, Reference PrycePryce (2016) found obligation to obey the police to have similar effects. Reference Metcalfe and HodgeMetcalfe and Hodge (2017) also found, among Israeli adults, that elements of legitimacy were robust correlates of public willingness to empower police to fight terrorism. Despite this evidence, additional research into this relationship is necessary, as these studies did not fully elaborate on why legitimacy might influence empowerment. In the following section, we consider possible indirect paths through which legitimacy may affect public empowerment of the police.
Indirect Paths Linking Legitimacy and Empowerment: Anchoring and the Perceived Consequences of Empowerment
The crux of the process-based model is twofold. First, procedural justice underpins the legitimacy of law enforcement. Second, legitimacy has consequences for the relationship between the public and legal institutions (Reference Sunshine and TylerSunshine and Tyler 2003); specifically, legitimacy is a key contributor to how individuals think, feel, and act toward police (e.g., compliance and cooperation; Reference TylerTyler 2006). These normative orientations toward the police extend to the public willingness to empowerment of police and evaluations of police practices (Reference PrycePryce 2016; Reference Sunshine and TylerSunshine and Tyler 2003). Evaluations of these practices are likely to involve the possible consequences, good or bad, associated with them (e.g., Reference FeldmanFeldman 1988). To better understand these perceived consequences, and sources of these perceptions, we draw from cognitive psychology and theories of rational choice (Reference BeckerBecker 1968; Reference BlankenshipBlankenship et al. 2008; Reference Hechter and KanazawaHechter and Kanazawa 1997; Reference Nagin and PaternosterNagin and Paternoster 1993; Reference Tversky and KahnemanTversky and Kahneman 1974).
Individuals' normative beliefs and orientations serve as “psychological anchors” for cognitive evaluation and perception of the world. Anchoring is a cognitive heuristic where individuals rely on a preconceived notion to shape their beliefs, and adjust their views only marginally from that point (Reference BlankenshipBlankenship et al. 2008; Reference Meub and ProegerMeub and Proeger 2015; Reference Tversky and KahnemanTversky and Kahneman 1974). Indeed, anchoring plays a considerable role in the perceptions and opinions that an individual possesses, particularly when little information is known on the topic (Reference Chapman and JohnsonChapman and Johnson 1999). For instance, anchoring is commonly used as a means of estimating risk and uncertainty (Reference PlousPlous 1989; Reference Wright and AndersonWright and Anderson 1989), and predicting future performance (Reference Switzer and SniezekSwitzer and Sniezek 1991). Legitimacy appears to function as an anchor for individual evaluations of police (Reference FoxFox et al. 2018; Reference PrycePryce 2016; Reference Sunshine and TylerSunshine and Tyler 2003), with recent research demonstrating that legitimacy colors individuals' subjective perceptions of police appearance and behavior (Reference Moule and KMoule et al. 2018). As Reference Sunshine and TylerSunshine and Tyler (2003: 517) noted, “when [police] are not viewed as legitimate, their actions are subject to challenge, their decisions are not accepted, and their directives are ignored.” This recognition provides a theoretical basis for the direct effects of legitimacy on empowerment.
With respect to indirect effects connecting legitimacy and empowerment, we draw on rational choice theories to formulate these linkages. Rational choice theories contend that individuals evaluate the potential costs and benefits of possible actions (Reference BeccariaBeccaria [1764] 1963; Reference BenthamBentham [1789] 1948). These evaluations, in turn, shape individual behavior (Reference BeckerBecker 1968; Reference Nagin and PaternosterNagin and Paternoster 1993). Assessments of these possible consequences are not random; rather, paralleling the notion of psychological anchors, they vary as a function of individual dispositions and normative orientations (e.g., Reference AgnewAgnew 2011; Reference MatsuedaMatsueda et al. 2006; Reference Piquero and TibbettsPiquero and Tibbetts 1996; Reference PogarskyPogarsky et al. 2017; Reference StanovichStanovich 1999). Specifically, the perceived consequences of a behavior should provide indirect pathways between individual dispositions and behaviors. In this case, the perceived consequences of empowerment should link legitimacy and the willingness to empower police. Consistent with our interpretation of legitimacy as an anchor, we argue that it is one such influential disposition, directly influencing public willingness to empower police (Reference Sunshine and TylerSunshine and Tyler 2003). Legitimacy should then also shape the perceived consequences of police behavior, because members of the public are anchored to the belief that the police will act in a legitimate (e.g., fair, effective, and just) fashion (Reference MacCounMacCoun 2005; Reference TylerTyler 2006).
The perceived consequences of police behavior should also influence public willingness to empower the police (see, generally, Reference Levi and StokerLevi and Stoker 2000). Two perceived consequences, one cost and one potential benefit, seem especially relevant with regard to public empowerment of police: improved police effectiveness and harm to civil rights. These possible consequences of empowering the police reflect the natural tensions between the government and the citizenry (Reference ChongChong 1993; Reference McClosky and BrillMcClosky and Brill 1983; Reference PackerPacker 1964; Reference SnidermanSniderman et al. 1996) and are persisting dimensions of the public discourse on law enforcement (President's Task Force on 21st Century Policing 2015).
First, a possible benefit of any police practice is that it assists law enforcement in being more effective at preventing and fighting crime. Police effectiveness, and perceptions of this effectiveness, is partially dependent upon the legitimacy of law enforcement (Reference Engel and SmithEngel and Smith 2009; Reference SmithSmith 1994; see also Reference Weitzer and TuchWeitzer and Tuch 2005). Indeed, trust—a prominent component of legitimacy—in social organizations increases perceptions of the effectiveness of those institutions (Reference Sunshine and TylerSunshine and Tyler 2003; Reference WaltersWalters 2004). Citizen perceptions of program effectiveness should, consequently, shape their support for those programmatic efforts (see Reference GouldGould 2002; Reference LockLock 1999). If police practices are perceived as being likely to improve police effectiveness, support for those programs and police behaviors would also be expected to increase.
The second potential consequence of any police practice is that it may infringe on individuals' civil liberties (Reference GouldGould 2002). Past research suggests that key aspects of legitimacy, such as trust, color the perceptions of possible threats to personal civil liberties posed by social institutions. These perceptions similarly shape assessments of institutional practices (Reference BorchersBorchers 2001; Reference SiegristSiegrist et al. 2000). Assessments of law enforcement are no exception. Higher levels of trust in the police correspond with lower levels of concern about the loss of, or potential infringement upon, civil liberties by the government (see Reference Davis and SilverDavis and Silver 2004). Lower levels of concern about the possible loss of civil liberties, in turn, should correspond with more support for police practices. Taken together, these perceived costs and benefits associated with empowerment offer potential indirect paths for legitimacy to influence public willingness to empower law enforcement.
Understanding public willingness to empower the police is particularly timely, given that a number of contentious issues involving law enforcement that have gained prominence in recent years across the United States (President's Task Force on 21st Century Policing 2015; Reference WeitzerWeitzer 2017). One such issue involves the militarization of local law enforcement (President's Task Force on 21st Century Policing 2015). We consider empowerment in the context of police militarization, a controversial set of practices brought to light by the events that occurred in Ferguson in 2014, and around which debate persists today. We discuss the militarization of American police, and the role that legitimacy and the perceived consequences of militarization may play in shaping public willingness to empower police to become more militarized, in the following section.
A Context to Examine the Empowerment Hypothesis: Police Militarization
Writing in the mid-1990s, Kraska and colleagues (Reference Kraska and CubellisKraska and Cubellis 1997; Reference Kraska and KappelerKraska and Kappeler 1997) argued that American police were slowly beginning to look more and more like the Armed Forces. The authors elaborated on the growing resemblance of the police to the military through the use of the concepts of militarism and militarization (see also Reference KraskaKraska 2007). Militarism is the foundation for militarization (Reference BerghahnBerghahn 1982; Reference Eide and TheeEide and Thee 1980); it involves “beliefs, values, and assumptions that stress the use of force and threat of violence as the most appropriate and efficacious means to solve problems” (Reference KraskaKraska 2007: 503).Footnote 1 Police militarization, in turn, refers to the “process whereby civilian police increasingly draw from, and pattern themselves around, the tenets of militarism and the military model” (Reference KraskaKraska 2007: 503).Footnote 2 This patterning includes the acquisition of surplus military weapons, equipment, and vehicles (Reference BalkoBalko 2013; Reference Campbell and CampbellCampbell and Campbell 2010; Reference KraskaKraska 2007), and has been occurring for some time.Footnote 3
American police have slowly been becoming more militarized since the 1960s (Reference Maguire and KingMaguire and King 2004). Kraska and colleagues (Reference Kraska and CubellisKraska and Cubellis 1997; Reference Kraska and KappelerKraska and Kappeler 1997), for example, documented the growth of paramilitary policing units, such as Special Weapons and Tactics (S.W.A.T.) teams. These units are closely associated with militarization, as they commonly use surplus military weapons and equipment and were originally developed to respond to dangerous criminal events, such as terrorist attacks and hostage situations (Reference BeckBeck 1972; Reference Kraska and KappelerKraska and Kappeler 1997). Reference Kraska and CubellisKraska and Cubellis (1997) found a sharp uptick in the number of agencies harnessing paramilitary units throughout the 1980s. As these units became more commonplace, they were increasingly harnessed for additional police activities. Indeed, paramilitary units were increasingly used for the serving of warrants and proactive patrols (e.g., Reference BalkoBalko 2013; Reference Kraska and CubellisKraska and Cubellis 1997), practices that continue through to the present day.
Kraska and colleagues (Reference Kraska and CubellisKraska and Cubellis 1997; Reference Kraska and KappelerKraska and Kappeler 1997) were unable to explain why paramilitary policing units were becoming more common, finding factors such as crime rates did not predict the use of S.W.A.T. teams by police agencies. They suggested that this growth was a result of federal funding initiatives providing local police access to surplus military equipment (Reference BalkoBalko 2013; Reference Pennella and NacciPennella and Nacci 1997). These initiatives were typified by the 1033 Program, a military-law enforcement equipment exchange program enacted as part of the National Defense Authorization Act of Fiscal Year 1997 (US House Committee on Armed Services 2014). In the wake of September 11, 2001 terror attacks, additional funding streams allowed departments to acquire new technologies and equipment, further promoting the militarization of local law enforcement (Reference BalkoBalko 2013; Reference Chaffetz and CumminsChaffetz and Cummins 2016). Reference RadilRadil et al. (2017: 208) and Reference DelehantyDelehanty et al. (2017) noted that 80 percent of U.S. counties had received equipment through the 1033 Program between 2006 and 2013.Footnote 4 To date, over 8000 law enforcement agencies, representing roughly 45 percent of all police agencies in the United States,Footnote 5 have participated in the Program (Defense Logistics Agency 2018).
The militarization of American police continued largely without controversy throughout the 2000s (Reference BalkoBalko 2013). However, the civil unrest, and police response, seen in Ferguson in 2014 served as a flashpoint for the public discourse surrounding police militarization. Nightly news broadcasts depicted heavily armored police clad in camouflage and combat gear, pointing military-style weapons at civilians (Reference Kesling and ShallwaniKesling and Shallwani 2014). In the wake of these events, policy makers and stakeholders debated the merits of militarized police, expressing concerns about the consequences of militarization for citizens' civil rights and police effectiveness (Reference MadhaniMadhani 2014; Reference PaulPaul 2014; President's Task Force on 21st Century Policing 2015). These concerns lead President Obama to issue an executive order curtailing access to some surplus military equipment and weapons available through the 1033 Program (Reference KorteKorte 2015).Footnote 6 In August of 2017, President Trump rescinded the Obama-era executive order. Rescinding the Obama-era executive order was done due to concerns about officer and public safety and police effectiveness (Reference GoldmanGoldman 2017). For example, current Attorney General Jefferson Sessions criticized the Obama executive order, claiming “Those restrictions went too far. We will not put superficial concerns above public safety” (Reference EbertEbert 2017: para 9). These events, and the arguments they have provoked, provide the backdrop for the current study.
Current Study
A substantial body of research has examined the process-based model of policing, finding support for the linkages between procedural justice, legitimacy, compliance, and cooperation (Reference MazerolleMazerolle et al. 2013; Reference Nagin and TelepNagin and Telep 2017; Reference ReisigReisig et al. 2007; Reference TylerTyler 2006, Reference Tyler2017). Limited research has examined the empowerment hypothesis, by which legitimacy engenders public willingness to empower law enforcement (Reference PrycePryce 2016; Reference Sunshine and TylerSunshine and Tyler 2003). We first argue for and assess whether legitimacy exhibits direct effects on empowerment, within the context of police militarization. Second, drawing on insights from rational choice theory, we argue that legitimacy forms the basis for assessing the perceived consequence of militarization. These perceived consequences should also influence public willingness to empower police to become more militarized. We thus also consider the indirect effects of legitimacy on empowerment through these perceived consequences. We specify and test a total of seven hypotheses regarding legitimacy and empowerment.
Hypothesis 1:Higher levels of police legitimacy will increase support for police use of military surplus equipment.
Hypothesis 2: Higher perceptions that militarization increases police effectiveness will increase support for the police use of military surplus equipment.
Hypothesis 3: Higher perceptions that militarization will increase violations of citizens' rights will decrease support for police use of military surplus equipment.
Hypothesis 4a: Higher levels of police legitimacy will increase perceptions that militarization will make the police more effective.
Hypothesis 4b:Higher levels of police legitimacy will reduce perceptions that militarization will result in increased violation of citizens' rights.
Hypothesis 5a: Legitimacy will show an indirect effect on support for police use of surplus military equipment through perceptions that militarization increases police effectiveness.
Hypothesis 5b: Legitimacy will show an indirect effect on support for police use of surplus military equipment through perceptions that militarization increases violations of citizen's rights.
In the following section, we detail the data and methods used to answer our research questions and better elaborate on the legitimacy-empowerment link.
Methods
Data
Data used in the current study consist of a national sample of 702 American adults surveyed about their perceptions of law enforcement, particularly issues relating to police militarization (see Reference FoxFox et al. 2018; Reference Moule and KMoule et al. 2018). Data were collected in the Spring of 2017 using Qualtrics' online survey service.Footnote 7 The service contains over 13 million diverse users who are solicited to participate in survey research through multiple methods, and is increasingly being used in social science research (Reference BushmanBushman et al. 2012; Reference Wright and SkagerbergWright and Skagerberg 2012). Respondents were selected from Qualtrics' list of survey participants using stratified random sampling procedures. Participants were stratified on gender, race, and household income to mirror the composition of American adults from the 2010 U.S. Census.Footnote 8
A total of 705 individuals originally completed the instrument; three individuals were removed for failing attention checks (Reference Oppenheimer and Nicolas DavidenkoOppenheimer, Meyvis and Davidenko 2009). The remaining 702 participants entered and completed the survey in a satisfactory manner. Responses were required for all questions, resulting in no missing data. Surveys were completed in an average of 18 minutes, and respondents were compensated above industry standards ($3 upon completion) to encourage high response and completion rates in the study. Overall, the data are well suited for addressing the empowerment hypothesis, and examining the direct and indirect effects of legitimacy on public empowerment of the police.
Dependent Variable
Support for Police Use of Surplus Military Equipment was measured as a latent factor using four items capturing sentiments toward police use of equipment commonly associated with militarization (e.g., surplus military equipment, style of dress, or “material militarization”; see Reference KraskaKraska 2007; Reference LockwoodLockwood et al. 2018). Respondents were asked whether law enforcement should be able to (1) use surplus military weapons (e.g., assault weapons, AR-15/M4; submachine guns, MP5), (2) use surplus military vehicles (e.g., BearCat armored personnel carrier, mine resistant ambush protected vehicle-MRAP), (3) use surplus military equipment (e.g., computers, tools, generators, etc.; see, e.g., Reference RadilRadil et al. 2017), and (4) wear military style uniforms. Respondents indicated agreement with each statement on a Likert scale ranging from 1 (Strongly disagree) to 4 (Strongly agree), with higher values indicating more support for police use of surplus military equipment. We note that the phrasing of these questions explicitly concerns respondent views about whether or not police should be allowed to engage in these behaviors, consistent with the discretionary aspect of the empowerment hypothesis.
Independent Variables
Increased Police Effectiveness Associated with Militarization is a single item indicating the extent to which respondents believe providing police with surplus military equipment will make them better able to fight crime. Respondents were specifically asked the extent to which they agreed with following statement: “The use of surplus military weapons, vehicles, equipment, and technology will make law enforcement more effective at fighting crime.” Respondents indicated agreement with this statement on a Likert scale ranging from 1 (Strongly disagree) to 4 (Strongly agree).
Increased Violations of Citizens' Rights Associated with Militarization is a single item indicating the extent to which respondents believe that providing surplus military equipment to law enforcement will result in the erosion of civil rights. Respondents were asked the extent to which they agreed with the following statement: “The use of surplus military weapons, vehicles, equipment, and technology will make law enforcement more likely to violate citizens' rights.” Respondents indicated agreement with this statement, with response categories ranging from 1 (Strongly disagree) to 4 (Strongly agree).
Police Legitimacy is measured using eight items capturing respondents' feelings toward the police, an important foundation from which willingness to empower the police emerges. Measures of trust in police and obligation to obey were adapted from prior research (e.g., Reference ParryParry et al. 2017; Reference ReisigReisig et al. 2007; Reference Sunshine and TylerSunshine and Tyler 2003). Respondents were asked the extent to which they agreed with the following: (1) Most police officers do their job well; (2) I respect the police and their authority; (3) I trust police officers; (4) I am confident in law enforcement; (5) In general, police officers act professionally; (6) The police can be trusted to make decisions that are best for my community; (7) You should accept police decisions, even if you think they are wrong; and (8) You should do what the police tell you even if you disagree. Responses to each item ranged from 1 (Strongly disagree) to 4 (Strongly agree), with higher scores indicating stronger perceptions of police legitimacy. Legitimacy was also measured as a first-order latent factor (assessment of the measurement model is reported below).
We also account for demographic and socioeconomic characteristics that might influence willingness to empower police (Reference Moule and KMoule et al. 2018). Political conservatism is a two-item scale (r = .80), with respondents indicating how fiscally and socially conservative they are. Respondents were asked “How would you describe your political views on [social/fiscal] issues?” Responses for each item ranged from 1 (Extremely Liberal) to 4 (Extremely Conservative). Higher scores indicate stronger conservative beliefs. Veteran (=1, no military service = 0) is a dummy variable indicating whether the respondent reported being a current or former member of the US Armed Forces, including the National Guard or Reserves (Reference WhiteWhite et al. 2012). We account for respondent age (in years) and whether the respondent is male (=1, female = 0). With regard to race, we account for whether the respondent is white (=1, non-white = 0). Respondent residence is measured as a dummy variable, urban (=1, suburban or rural = 0). Respondent education is measured as a dummy variable indicating whether they possessed a college degree (=1, no college degree = 0). Finally, we account for whether the respondent is currently married (=1, not married = 0) and whether they are a parent (=1, no children = 0).
Analytic Strategy
To test hypotheses about public support for police use of military surplus equipment, we used SEM (Reference BollenBollen 1989; Reference Bowen and GuoBowen and Guo 2011). SEM offers several benefits for analysis including the simultaneous evaluation of multiple dependent variables, the measure of observed variables that purges error variance, and the estimation of indirect effects in tests of mediation. Two of the concepts we included in the analysis were modeled as latent factors: police legitimacy and support for police use of surplus military equipment. These latent factors were first evaluated as measurement models. Then, when determined to fit the data, these latent factors were included in the structural model.
The hypotheses were evaluated through the regression coefficients' predicted direction, statistical significance, and model r-square. Furthermore, the measurement and structural models were assessed through four goodness-of-fit indexes that help determine if the model implied covariance matrix fit the data. The chi-square test should be nonsignificant, although it is prone to detect small differences in large samples. The Tucker-Lewis index (TLI) and comparative fit index (CFI) should both be above 0.950 for a good fit or at least above 0.900 for a reasonable fit (Reference BollenBollen 1989). The root mean error of approximation (RMSEA) should be below 0.050 for a good fit and less than 0.100 for a reasonable fit (Reference Hu and BentlerHu and Bentler 1999). Also, for the measurement model, the factor loading coefficients should be above 0.300 (Reference KlineKline 2005). Taken together, these goodness-of-fit indexes suggest there is concordance between the theoretical model and the data.
Because latent factors included ordinal observed variables, the models were estimated through weighted least squares and variance adjusted (WLSMV) estimator in the software R's Lavaan package (Reference RosseelRosseel 2012). When employed, the WLSMV returns probit coefficients for the structural paths and estimates the observed categorical variables as latent continuous distributions. In addition, significance of the indirect effects was estimated by bootstrapped standard errors and confidence intervals.
Results
The descriptive statistics for the sample are reported in Table 1, and the correlation matrix is reported in Table 2. The average respondent was female (51 percent), lived in a rural or suburban location (70.9 percent), white (64 percent), not married (53 percent), a parent (68 percent), and not a veteran (89 percent). The respondents were about in the middle of the conservative scale (m = 5.08 on a scale that ranged from 2 to 8). As for views about the use of surplus military gear will make law enforcement more effective, the mean was 2.77 and the modal agreement category was 3 (44 percent) on a 4-point scale. For views on whether the acquisition and use of military surplus equipment will make law enforcement likely to violate rights, the mean was 2.42 and the modal agreement category was 2 (37 percent) on a 4-point scale.
Table 1. Descriptive Statistics (N = 702)
Notes: Standard deviation is s.d.
Table 2. Correlation Matrix (N = 702)
Note: All correlations as polychoric, except for correlations with age, which are polyserial.
Bivariate correlations present in Table 2 provide preliminary evidence in favor of our theoretical model. The components of legitimacy—measures of trust and obligation to obey the police—positively correlate with measures of support for police use of military surplus equipment. Measures of legitimacy also positively correlate with belief that police use of surplus military equipment will increase police effectiveness, while negatively correlating with belief that police use of surplus military equipment will make law enforcement likely to violate civil rights. These perceived consequences of militarization correlate as expected with support for police militarization. With preliminary evidence supporting the linkages between legitimacy, the perceived consequences of militarization, and support for militarization, we next turn to the measurement model.
The fit of the measurement model is reported in Table 3.Footnote 9 The measurement model was a good fit, although the model chi-square was significant (chi-square p < .000; TLI = 0.999; CFI = 0.999; RMSEA = 0.043, 95 percent c.i. = 0.032, 0.053). All of the factor loadings for both latent factors were greater than 0.300 and statistically significant (range: 0.460 to 0.914). The error terms for two observed variables in the legitimacy latent factor were correlated—obey a and obey b. This was done because these two variables were different from the trust variables; thus, correlating the errors made sense statistically and conceptually. Finally, the correlation between the two latent factors was moderately strong (r = 0.587) and significant (p < .000).
Table 3. Measurement Model (N = 702)
Notes: The error terms for the variables Obey a and Obey b were correlated.
After confirming that the measurement model was a good fit to the data, we estimated the structural model. The results of the SEM analysis are reported in Table 4. The model was a good fit to the data: chi-square p < .000; CFI = 0.993; TLI = 0.991; RMSEA = 0.060, 95 percent c.i. = 0.054, 0.065. Although the model chi-square was significant, the sample size is sufficiently large to expect minor deviations would result in a significant difference between the data and model implied covariance matrices. The RMSEA was in the reasonable fit range, and the TLI and CFI indicated a good fit. Therefore, we concluded the structural model was a good fit as well.
Table 4. Structural Model (N = 702)
Notes: The fit indicators were: model chi-square p value <.000, CFI = 0.993, TLI = 0.991, RMSEA = 0.056. The column labeled β reports the standardized regression coefficients. The column labeled “estimate” reports the unstandardized regression coefficients; for the military surplus model, the regression coefficients are probit.
Hypothesis 1 was confirmed: the effect of legitimacy on support for police using military surplus was positive and significant (b = 0.216; p < .050). Both of the militarization attitude variables had a significant impact of support for police use of surplus military equipment (p < .010). Perceptions that militarization would increase police effectiveness was positively related to willingness to empower the police (b = 0.489), while perceptions that militarization would correspond with increased civil rights violations was negatively related to this willingness (b = −0.070), to support for police use of surplus military equipment. Thus, Hypotheses 2 and 3 were also supported. Among the control variables, being male, white, and conservative had a positive significant impact on support for police use of military surplus. The r-square for the endogenous latent factor was 0.666. Perceptions that militarization would increase police effectiveness had the strongest effect (β = 0.543) followed by legitimacy (β = 0.199).
As for the hypotheses regarding the perceived consequences of militarization (Hypotheses 4a and 4b), legitimacy had a significant effect on both militarization variables and in the predicted directions. Legitimacy increased perceptions that militarization would increase police effectiveness (b = 0.546; p < .000), and it decreased perceptions that militarization would result in more civil rights violations (b = −0.454; p < .000). The r-squares for these outcomes were acceptable (.336 for increased effectiveness, 0.331 for increased violations of civil rights). Legitimacy had the strongest standardized effect on each outcome (β = 0.453 for increased effectiveness, β = −0.366 for increased violations of civil rights). Being conservative was the only control variable to have a significant effect on perceptions that militarization would increase police effectiveness, which was positive (b = 0.197). Regarding perceptions that militarization would increase civil rights violations, age, urbanicity, and conservative were significant. Age and conservative had a negative impact on these perceptions (b = −0.181 and −0.011, respectively), while residing in an urban area (b = 0.308) increased these perceptions. Hypotheses 4a and 4b are thus confirmed.
To test whether the two perceived consequences of police militarization mediated the direct effect of legitimacy on support for police use of surplus military equipment (Hypotheses 5a and 5b), we first examined a model with only legitimacy and the controls regressed on support for police use of surplus military equipment (not reported for space considerations). Legitimacy was positively (b = 0.507) and significantly (p < .000) related to this support. The r-square for the legitimacy-only model of support for police use of military surplus equipment was 0.426.
The complete path model assessing direct and indirect effects is shown in Figure 1. Comparing the legitimacy-only model to the full model (see Table 4), legitimacy was not entirely mediated by the inclusion of the two perceived consequences of militarization; that is, legitimacy remained statistically significant in both models. The indirect effects of legitimacy through the two perceived consequences of militarization measures were also positive and significant (p < .010): increased effectiveness (b = 0.267), and violation of rights (b = 0.248). The standardized indirect effect was greater through perceptions of increased effectiveness (β = 0.246) than perceptions of increased violations of civil rights (β = 0.029). Adding the two measures of the perceived consequences of militarization to the model improved the model fit (ΔR 2 = 0.240). Thus, Hypotheses 5a and 5b are supported.
Figure 1. Structural Path Model (N = 702). Notes: The path coefficients are standardized. Military surplus and legitimacy are latent factors. The indirect effect paths are dashed and the coefficients are in italics. For simplicity, the error terms are not shown in the diagram. The correlation between Military Effective and Militarization Violates Rights is not shown.
In sum, legitimacy exerts a direct effect on support for police use of surplus military equipment, but also works indirectly through the perceived consequences of militarization. These perceived consequences, which are influenced by legitimacy, also exert their own direct effects on support for police use of surplus military equipment. Taken together, our findings conform to theoretical expectations. We discuss these findings in the following section.
Discussion
Reference TylerTyler's (1990, Reference Tyler2003, Reference Tyler2006) process-based model of policing occupies a central place in contemporary American policing research (President's Task Force on 21st Century Policing 2015). Decades of research spanning multiple disciplines have found support for the core contentions of the model: procedurally just treatment of citizens increases the legitimacy of the police, and higher levels of police legitimacy correspond with more compliance and cooperation from citizens (Reference TylerTyler 2017). A third consequence of legitimacy, as suggested by Reference Sunshine and TylerSunshine and Tyler (2003), involves public empowerment of the police. As legitimacy increases, so too should public willingness to grant police more discretion to execute their mandate. Compared to other aspects of the process-based model, the empowerment hypothesis remains under-examined. Using a national sample of 702 American adults, and a SEM strategy, we sought to examine the direct and indirect pathways linking legitimacy and empowerment. Drawing from cognitive psychology and rational choice theories, particular attention was paid to the perceived consequences of empowering the police, including the ramifications of militarization for police effectiveness and citizens' civil rights. These perceived consequences were tested as being potential indirect pathways between legitimacy and empowerment. Our results warrant three broader points of discussion.
First, perceptions of legitimacy matter for public willingness to empower police. Consistent with findings by Reference Sunshine and TylerSunshine and Tyler (2003), Reference PrycePryce (2016), and Reference Metcalfe and HodgeMetcalfe and Hodge (2017), as citizen perceptions of police legitimacy increase, so too does their willingness to empower the police. This finding builds on prior empowerment research in two important ways. This effect occurs in a demographically and geographically diverse sample of American adults, suggesting the generalizability of the legitimacy-empowerment link beyond New York City and Washington, DC. We extended prior research by linking general perceptions of police legitimacy with assessments of specific controversial police practices associated with militarization (e.g., President's Task Force on 21st Century Policing 2015). In this case, the willingness to empower law enforcement involved allowing police to become more militarized through the acquisition and use of surplus military equipment. This focus extends prior work examining empowerment on other potentially controversial practices, such as stop and frisk (Reference Sunshine and TylerSunshine and Tyler 2003). Future research should consider whether the influence of legitimacy on public empowerment of police is invariant across categories of individuals, and whether it influences empowerment across other police practices.
To be sure, the public is not the only group whose values and beliefs are influential for understanding police policies and practices. Other stakeholders, including police chiefs and politicians, also shape the ability of law enforcement to become militarized.Footnote 10 For example, Reference NixNix (2017) demonstrated that police executives believe the public cares more about effective, rather than procedurally just, policing. This would, presumably increase the likelihood of agencies becoming more militarized. Political stakeholders also express support for police militarization (Reference Turner and FoxTurner and Fox 2017). Understanding decision-making processes among these stakeholders would be useful for understanding the militarization of police. Reference Bottoms and TankebeBottoms and Tankebe (2012, Reference Bottoms, Tankebe, Tankebe and Liebling2013), for example, suggested a dialogic model, where the path to legitimacy is a fluid process filled with calls and responses between the police, public, and other stakeholders. Others, drawing from institutional theory (Reference Crank and LangworthyCrank and Langworthy 1992; Reference KatzKatz 2001), have made similar suggestions. A more holistic approach to empowerment would take these multiple perspectives into account, and consider whether and how militarization influences stakeholder relationships with their communities. We are unable to address these considerations with our data, but they are certainly worthy of future study.
Second, consistent with notions of cognitive anchoring and rational choice (e.g., Reference BlankenshipBlankenship et al. 2008; Reference Nagin and PaternosterNagin and Paternoster 1993), police legitimacy also influences public assessments of the possible consequences of empowering the police. As perceptions of police legitimacy increase, so too do beliefs that becoming more militarized will increase police effectiveness. Alternatively, and comporting with prior research (e.g., Reference Davis and SilverDavis and Silver 2004), higher perceptions of legitimacy reduce concerns about civil rights violations due to militarization. In other words, individuals who view the police as more legitimate tend to focus on the positive potential consequences of empowering the police, while downplaying (or not seeing) the negative possibilities. These perceived consequences act as indirect pathways between legitimacy and empowerment. Notably, legitimacy continues to exert direct effects on empowerment once these indirect paths are incorporated into the structural equation model (see Figure 1), reinforcing the robust relationship between legitimacy and empowerment. Taken together, this latter set of findings supports both normative and instrumental explanations for empowering police, and the linkages between these explanations. Future research should continue to explore the interplay between perceptions of legitimacy and willingness to empower police. One important avenue for this research would involve longitudinal designs and assessments of the relationship between negative media coverage and changes in legitimacy and support for police militarization. A second avenue would be assessing the perceived effects of militarization on effectiveness in relation to specific crimes or crime types (e.g., Reference TankebeTankebe 2013)
Third, our findings have implications for police policy. As our results demonstrate, legitimacy is a strong predictor of public support for police acquiring and using surplus military equipment, net of their perceptions about the possible consequences of police militarization. This finding suggests that police administrations should work toward improving citizens' overall perceptions of legitimacy (Reference TylerTyler 2006) through engaging in procedurally just policing, demonstrating effectiveness, and respecting the boundaries of their authority (Reference HuqHuq et al. 2017; Reference TankebeTankebe 2009). Procedurally just policing especially will pay dividends if law enforcement agencies are considering implementing specific policies and practices that might be seen as controversial by the public. If the public generally views the police as legitimate, then they will not only obey the law, but also defer to those practices, programs, and policies. Conversely, in jurisdictions where the community views the police as lacking legitimacy, such practices may be seen as oppressive (Reference BalkoBalko 2013; but see Reference Kochel and WeisburdKochel and Weisburd 2017). As demonstrated here, any practice perceived to be ineffective or infringing citizens' rights will likely have less support from the public, but legitimacy is a global lens through which these practices are also evaluated.
As with any study, the current study is not without limitations. First, the data are cross sectional, carrying with this recognition of the usual concerns about causality. Longitudinal data would be ideal for assessing empowerment, as legitimacy and empowerment may have reciprocal effects. Likewise, the perceived consequences of militarization may vary over time (Reference PogarskyPogarsky et al. 2017; Reference SmithSmith et al. 1990; Reference WilsonWilson et al. 2017). Nonetheless, prior tests of the empowerment hypothesis have been cross sectional, and our findings operate in theoretically expected directions. Second, we concentrated on only two potential consequences of militarization as pathways between legitimacy and empowerment; there are likely other potential consequences, such as the financial cost of programs or accountability associated with a program (e.g., Reference Latessa and HolsingerLatessa and Holsinger 1998), that individuals consider in their willingness to empower police. Accountability may be especially influential for understanding empowerment and militarization, as the 1033 Program was scrutinized by President Obama due to a lack of oversight (Reference ShermanSherman 2014). Assessing these alternative pathways should be a priority for future research. Third, other individual characteristics might also influence willingness to empower police. For instance, general assessments of police effectiveness, or perceptions of individual risk, might play a role in fostering empowerment. Integrating these characteristics into assessments of empowerment should be a priority of future research (Reference PrycePryce 2016).
Further, in examining empowerment and police militarization, we have drawn primarily from American research on the process-based model. Nonetheless, in recent years, other countries have also begun to wrestle with issues of empowerment and militarization, including Germany, Brazil, Canada, and the United Kingdom (e.g., Reference JeffersonJefferson 1990; The Week 2017). For example, in the United Kingdom, counter-terrorism police units have been developed that resemble the Special Air Service (SAS), and are much more heavily armed than traditional British police (Reference EvansEvans 2016). Discussions of police militarization in other countries, especially those located in continental Europe, tend to focus on the issue within the context of specific problems (e.g., border security and high profile crimes), rather than the general question of public willingness to allow police to address these problems in a particular manner (Reference EastonEaston et al. 2010; Reference LinkeLinke 2010; Reference Weber and KadarWeber 2001).Footnote 11 To the extent that other countries are experiencing militarization, or already maintain a militarized police force (Reference BeedeBeede 2008; Reference LutterbeckLutterbeck 2004; Reference McCullochMcCulloch 2004), questions about empowerment persist and should be examined. We are aware of only one study examining public empowerment of police in a non-American context (see Reference Metcalfe and HodgeMetcalfe and Hodge 2017). In non-Western or developing countries, where police effectiveness, rather than procedural justice, appears to drive legitimacy (Reference TankebeTankebe 2009, Reference Tankebe2013), perceived effectiveness may be more salient for public willingness to continue empowering police (Reference PrycePryce 2016). Cross-cultural examinations of empowerment, whether general or more specifically in regard to militarization and its correlates, seem an especially fruitful avenue for future research (Reference JacksonJackson 2018).
In additional to these considerations, methodological concerns regarding the use of online samples must also be noted. Online samples are both a time- and cost-effective way for conducting social science research. While convenience samples from online platforms are known to underrepresent or exclude certain categories of individuals (racial/ethnic minorities, impoverished individuals, the illiterate, the elderly, and non-internet users), internet use in America among historically marginalized groups is increasing (e.g., Reference MossbergerMossberger et al. 2008; Reference Moule and KMoule et al. 2013; Pew 2018). One means of ensuring that online surveys capture members of historically marginalized groups, as we do here, is the use of survey panels where specific sample characteristics can be requested by researchers (Heen et al. 2014). In their comparison of online sampling platforms, Heen et al. (2014) found these platforms were able to provide samples within a 10 percent range of the actual population. Future research should work to replicate the results of the current manuscript using alternative sampling strategies. An additional concern of online surveys is that they may be vulnerable to respondents hurrying through the survey (Oppenheimer et al. 2009). Instructional manipulation checks were included to identify suspect responses and exclude respondents who failed any checks.
In the end, the process-based model of policing continues to have relevance for practitioners, policy makers, and members of the public. Despite a robust body of research into the process-based model, neglected aspects of the theory, including the empowerment hypothesis, warrant greater attention. Our results lend further support to a small body of literature supporting the link between legitimacy and public empowerment of police. We extended this work by also considering indirect paths through which legitimacy might also operate. In sum, legitimacy appears to be a robust foundation for understanding how the public evaluates the police generally, as well as specific police practices. We encourage future research to continue exploring factors influencing public empowerment of law enforcement.
Acknowledgements
The authors thank the editors and the anonymous reviewers for their comments on earlier versions of the manuscript. An earlier version of the paper was presented at the 2017 annual meeting of the American Society of Criminology in Philadelphia. We thank Rick Trinkner for his useful feedback at the meeting.
