Police Occupational Culture: Testing the Monolithic Model^*

* Points of view are those of the authors and do not necessarily represent the official position or policies of the U.S. Department of Justice. An earlier version of this manuscript was presented at the Southern Criminal Justice Association Meeting in Savannah, Georgia, in September 2016.

Abstract

Since foundational work in the 1950s, researchers have described a variety of dimensions of the occupational culture of police. In an effort to integrate the disparate works, a theoretical model has been constructed depicting the ways in which the stressful features of the police work environment produce coping mechanisms and outcomes. While this model usefully organizes the vast literature on police culture, it has yet to be empirically tested. The current study addresses this void. Path analyses of officer survey data reveal support for several of the propositions set forth by the monolithic model, although the magnitude of the statistical associations was not very powerful and overall model fit was marginal. The implications of these findings are especially relevant given recent concerns over police-community relations and the renewed interest in the police occupational culture expressed by the President’s Task Force on 21st Century Policing.

Keywords:

• police
• police culture
• police attitudes
• path analysis

View correction statement:

Erratum

Notes

* Points of view are those of the authors and do not necessarily represent the official position or policies of the U.S. Department of Justice. An earlier version of this manuscript was presented at the Southern Criminal Justice Association Meeting in Savannah, Georgia, in September 2016.

¹ By contrast, organizational cultures, which are often erroneously used synonymously with occupational culture, represent the values and missions of particular police agencies. As opposed to culture developing from the front line, organizational cultures are created by police leaders (who take into account the various needs and expectations of the external community), in a top-down manner to be espoused by all departmental members (see Wilson, 1968). The occupational version of a monolithic culture is the one that dominates discussions about police and serves as the focus of the current study. This conceptualization of culture is assumed to transcend time and place. As Crank (1998, p. 26) once asserted “street cops everywhere tend to share a common culture because they respond to similar audiences everywhere.”

² These studies were often grounded in prior typology research that illuminated a variety of working styles that were used by officers to deal with the internal and external demands of the occupation (see, for example, Broderick, 1977; Brown, 1988; Muir, 1977; White, 1972).

³ Subsequent analyses connecting these orientations to street-level behavior revealed that officers whose attitudes were congruent with the monolithic culture perspective searched citizens during traffic stops at higher rates (Paoline & Terrill, 2005) and used force more frequently (and at higher levels; Terrill, Paoline, & Manning, 2003) than their culturally divergent counterparts.

⁴ Throughout the twentieth century, independent reviews of police performance were critical of the narrowly focused crime fighting tactics being employed across American police agencies that shaped the occupational culture (Walker, 1977). For example, both the President’s Commission on Law Enforcement and the Administration of Justice (1967) and the National Advisory (Kerner) Commission on Civil Disorders (1968) identified the detached aggressive crime fighting policing approaches, where officers engaged in frequent citizen stops, searches, and coercion as a principal source of community tension and separation. In a similar manner, the Christopher Commission’s (1991, p. 98) highly publicized review of the Los Angeles Police Department (LAPD), following the Rodney King beating, identified LAPD’s “seize mentality” and “hardnosed” crime fighting culture as a principle cause of the coercive (mis)treatment of citizens.

⁵ Because police culture is said to originate at the entry point of the occupation (Paoline, 2014), as well as the recognition that culture varies by rank (see Manning, 2007; Reuss-Ianni, 1983), we focus our efforts on patrol officers with street-level assignments. This ensures that all of the surveyed officers are exposed to the internal (i.e. interactions with supervisors in the police organization) and external (i.e. interactions with citizens on the street) facets of the police occupation.

⁶ The eight agencies were of varying organizational size and geographic locales, and include: Columbus, Ohio; Charlotte-Mecklenburg, North Carolina; Portland, Oregon; Albuquerque, New Mexico; Colorado Springs, Colorado; St. Petersburg, Florida; Knoxville, Tennessee; and Fort Wayne, Indiana. The St. Petersburg Police Department declined to participate in the section of the survey that focused on occupational culture, and thus seven of the eight are included in the current study.

⁷ The Fort Wayne Police Department did not use a roll call system, so we coordinated with the department’s annual in-service training and administered the survey during these training sessions.

⁸ Survey periods varied across each of the seven agencies, ranging from September, 2007 to June, 2008.

⁹ The monolithic model has built-in mediation mechanisms, and we therefore model these relationships despite recognition of the fact that mediation is sometimes recommended only for longitudinal data. The danger in using cross-sectional data lies in potential overestimation of regression paths between hypothesized predictors and outcomes (Cole & Maxwell, 2003; Maxwell & Cole, 2007; Maxwell, Cole, & Mitchell, 2011). On the other hand, our mediator variables do meet Baron and Kenny’s (1986, p. 1,176) requirements in that “(a) variations in levels of the independent variable significantly account for variations in the presumed mediator (i.e. Path a), (b) variations in the mediator significantly account for variations in the dependent variable (i.e. Path b), and (c) when Paths a and b are controlled, a previously significant relation between the independent and dependent variables [Path c] is no longer significant … [although] a more realistic goal may be to seek mediators that significantly decrease Path c rather than eliminating the relation between the independent and dependent variables altogether.” Hayes (2013) argues that nothing forbids the use of mediation analysis with cross-sectional data when no claims of causation are being made, as is the case in this study. An additional reason why this technique is justified here is that we are testing for the absence of full or partial mediation, not the presence of it. The monolithic model implies full mediation, and we are putting that hypothesis to empirical test. The use of cross-sectional data within this theoretical framework makes for a conservative test of the model. If an analysis using longitudinal data were to generate even weaker effects than those seen on the basis of cross-sectional data, this finding would affirm our ultimate conclusion that the monolithic model is not properly specified.

¹⁰ Weisburd et al.’s (2000) national survey of police (n = 905) across 121 departments, from which our “code of silence” measure was adopted, also found that vast majority of respondents reported that they “disagreed” (66%) or “strongly disagreed” (18%) that the code was necessary for good policing. A few (speculative) factors might be operating to explain the difference in modern occupational cohesion compared to that noted by researchers during the professional reform era (e.g. Skolnick, 1966; Van Maanen, 1974; Westley, 1970). First, the similarities in officers’ backgrounds, which might have contributed to the previously described unity, has changed as police departments are more diversified in terms of sex, race, and education (Paoline et al., 2000; Sklansky, 2006). Second, contemporary officers are trained and socialized with (at a minimum) an exposure to accountability awareness and individual responsibility (i.e. culpability for other officer’s indiscretions if they are aware and do not report) that did not equally operate during the time before (or just after) the 1960s Supreme Court’s due process standards when foundational studies of police culture were conducted (Paoline, 2003). Finally, as we note for our social isolation measure, the majority of officers (59%) responded that “they would rather hang around with non-police than other police officers” when they were not working. This suggests that the days of officers having “choir practice” (Wambaugh, 1975) in the park after their shifts may be a thing of the past. Cumulatively, these features of contemporary policing could be working to interrupt the “friendship” and “family” features that Kleinig (2001) identified as the core of the “blue wall of silence.” We thank an anonymous reviewer for stimulating this thought.

¹¹ Although this is a path model and thus all scales were treated as observed (not latent) variables, we nevertheless checked for the validity of the underlying measurement model. The only constructs with three items (the minimum needed for confirmatory factor analysis [CFA]) were danger, role ambiguity, and stress. A three-factor CFA produced statistically significant factor loadings ranging from acceptable to high (.267 to .912), low between-factor correlations, and acceptable fit values (CFI = .891, SRMR = .054, RMSEA = .080). The underlying measurement structure of these three scales was sufficiently sound to proceed to path analysis.

¹² The seven agencies included in this study were chosen as part of a larger project examining the relevance of use of force policy variation on police outcomes (e.g. force usage, citizen complaints, perceptions of policy guidance) versus that of variation in organizational mission (or culture). We include a measure of agency size (i.e. small, medium, and large), which typically correlates with the level of bureaucratic structure of the internal environment (Maguire, 2003). Again though, the monolithic model that is being tested in the current study, and part of the contemporary pleas for reform, is assumed to be applicable to all American police departments. Future works interested in exploring potential sources of variation in officers’ adherence to the monolithic model could include, among other factors, the relevance of organizational culture (i.e. internal missions and expectations vis-à-vis community expectations).

¹³ Van Maanen’s (1974) seminal longitudinal study of police tracked the metamorphosis of officer occupational attitudes from the training academy through the first 2.5 years of their professional career.

Additional information

Notes on contributors

Eugene A. Paoline

Eugene A. Paoline III is a professor and Graduate Director in the Department of Criminal Justice at University of Central Florida. His research interests include police culture, police use of force, and occupational attitudes of criminal justice practitioners. He is the author of Rethinking Police Culture (2001, LFB Scholarly Publishing) and Police Culture: Adapting to the Strains of the Job (2014, Carolina Academic Press), and is currently serving as Co-Principal Investigator on a National Institute of Justice grant geared toward examining the structure, operation, and effectiveness of police Early Intervention (EI) systems.

Jacinta M. Gau

Jacinta M. Gau is an associate professor in the Department of Criminal Justice at the University of Central Florida. Her primary research interests are in policing, with an emphasis on police-community relations, racial issues, and procedural justice and police legitimacy. She has also written about quantitative methods and criminal justice policy. Her work has appeared in multiple journals. She has published the books Statistics for Criminology and Criminal Justice (Sage Publications; 3rd edition 2017) and Criminal Justice Policy: Origins and Effectiveness (Oxford University Press; forthcoming).