Women in Policing and the Response to Rape: Representative Bureaucracy and Organizational Change

Abstract

The purpose of this study is to evaluate the theory of representative bureaucracy. Cross-sectional and longitudinal data, covering a period from 1997 to 2013, suggest that a greater number of women in the policing workforce is related to higher reporting rates for rape incidents and higher clearance rates for rape cases. Furthermore, the findings suggest that the relationship between female representation and women-centered outcomes is moderated by the presence of a union and mediated by resources for victims and community policing. Looking ahead, more research is needed to identify how female representation in policing influences organizational processes and outcomes for citizens.

Keywords

policing rape quantitative research sexual assault women women as professionals in the CJS longitudinal study women and social policy

A major concern for the administration of justice is the extent to which criminal justice organizations adhere to the principles valued by people in liberal democracies. Employees in justice-focused organizations have a tremendous amount of power and are prodigiously involved in the introduction, design, and enforcement of officially sanctioned policies on behalf of the state. These agents of formal social control exercise discretion regarding which policies should be enforced and who should be punished. Because of a lack of transparency and electoral accountability, the amount of power wielded by street-level bureaucrats sometimes results in a conflict between those who lead and manage justice-focused organizations and the people whom such agencies have been designed to serve.

Increasing female representation has been identified as one possible strategy for creating justice-focused organizations that are more responsive to public needs (President’s Task Force on 21st Century Policing, 2015; Silvestri, 2015). Historically, the occupation of policing has been dominated by White men, and although female representation in the United States has increased to approximately 12% (United States Department of Justice, Federal Bureau of Investigation, 2016), most law enforcement departments remain highly gendered organizations in which stereotypes about men and women drive norms and practices in the agency (Shelley, Morabito, & Tobin-Gurley, 2011). Drawing from the theory of representative bureaucracy, scholars argue that sociodemographic representation consistent with the characteristics of the constituency helps ensure that all citizens’ interests are represented in the formulation and implementation of public policy (Coleman-Selden, 1998; Oberfield, 2014). In addition, representation is hypothesized to strengthen citizens’ participation in the decision-making process and to increase their cooperation in interactions with specific bureaucratic agents (Riccucci, Van Ryzin, & Lavena, 2014).

The purpose of this study is to explore the effect of female representation in law enforcement on the reporting of rape incidents and the clearing of rape cases in the United States. Police officers are some of the most important street-level bureaucrats in the processing of sexual violence crimes. Officers function as the gatekeepers of the justice system (Frazier & Haney, 1996) and exercise a significant amount of discretion when deciding which cases are worthy of further review and how much organizational capital should be devoted to the investigation. Some evidence indicates that agencies with a greater number of female officers are more effective in identifying and clearing sexual violence cases compared with agencies with fewer female officers (Meier & Nicholson-Crotty, 2006; Walfield, 2016). However, research on the topic is limited, and other scholars have not found support for the representative bureaucracy hypothesis (Morabito, Pattavina, & Williams, 2017).

The Theory of Representative Bureaucracy

According to Mosher (1968), passive representation is defined as the extent to which bureaucratic organizations diversify their workforce by employing individuals from a variety of sociodemographic backgrounds. In theory, for an organization to be truly representative, the diversity of the people whom it serves must be adequately reproduced in its personnel structure. Passive representation can be interpreted as an indicator of society’s “commitment to equal opportunity and access to power and can promote legitimacy of public bureaucracies” (Meier & Nicholson-Crotty, 2006, p. 851). Although passive representation is largely symbolic, it is hypothesized to be associated with a citizenry that is more receptive to the bureaucracy because citizens feel connected to the organization and involved in the process of developing policies that affect the public (Riccucci et al., 2014).

Implicit in the theory of representative bureaucracy is the expectation that passive representation will lead to active representation in which the beliefs and values of individuals with common demographic characteristics are translated into decisions that benefit those with similar demographic characteristics (Pitkin, 1967). Public officials are hypothesized to use their discretion to promote equality and to improve outcomes for those whom public programs have traditionally underserved. In theory, for the presence of women (i.e., passive representation) to be translated into benefits for women (i.e., active representation), two conditions must be present: bureaucrats must have a significant amount of discretion in decision making (Meier, 1993), and the policy area must be gendered (Keiser, Wilkins, Meier, & Holland, 2002). According to Keiser et al. (2002), a policy area becomes gendered if (a) the political process defines the matter as a women’s issue, (b) the relationship between the bureaucrat and the client is changed because of gender, or (c) practices in the area directly benefit women as a group. Some examples of a gendered policy area include pay equality, reproductive rights, sexual assault, child care issues, and the underrepresentation of women in math and science occupations (Keiser et al., 2002).

Law Enforcement and the Reporting and Processing of Rape Cases

In the process of applying the theory of representative bureaucracy to women in policing, scholars have hypothesized several possible pathways by which passive representation in law enforcement may be translated into active representation for female victims. In theory, female officers may be more empathic and responsive to sexual assault victims than male officers, which in turn encourages victims to solicit police services and to be more cooperative in the investigative process (Martin, 1999; Rabe-Hemp, 2008; Schuck, 2014). Female officers may also sensitize male officers to the trauma experienced by sexual assault survivors, such that male officers become more sympathetic (Haarr, 1997). These propositions are rooted in the belief that women share a common set of values regarding the causes and consequences of victimization and the appropriateness of types of interpersonal and legal responses.

In the context of policing, most research on sexual assault focuses on officer decision making, specifically how legal and extralegal factors influence officers’ reactions to the victim and their willingness to do the necessary work to document the incident and make an arrest. In the general population, the research typically shows that men are more accepting of rape myths than women (Anderson, Cooper, & Okamura, 1997; Suarez & Gadalla, 2010) and that men tend to view the perpetrator as less culpable than women do (Alicke & Yurak, 1995; Gerber, Cronin, & Steigman, 2004). Some evidence indicates that female officers are less accepting of rape myths and are more likely to believe victims compared with their male colleagues (Brown & King, 1998; Page, 2007; Schuller & Stewart, 2000). However, other studies suggest that women officers may be more hostile and less willing than their male counterparts to exercise their discretion to benefit female victims (Alderden & Ullman, 2012; Sleath & Bull, 2012; Wentz & Archbold, 2012). The inconsistent empirical support for gender differences in officers’ attitudes and behaviors suggest that other factors may be important in understanding the active translation process.

Moderators and Mediators

Police Unions as Moderators

Situational factors such as environmental and organizational characteristics may moderate the relationship between female representation and the reporting and processing of rape cases by changing the way discretion is managed and the resources upon which frontline officers can draw to support their discretionary practices. One potential moderator in the active representation process is the presence of a police union. Officers began forming unions around the turn of the 20th century, and today, many of the rank and file see collective bargaining as a way to preserve officer autonomy and protect themselves from arbitrary and inconstant disciplinary practices by management (Fisk & Richardson, 2017). Scholars and police executives generally agree that unions have a significant impact on policies and practices internal to the agency, as well as accountability regarding police misconduct and the discipline of individual officers (Walker, 2008). While most of the literature focuses on the negative consequences of collective bargaining agreements, there is some evidence that unions can facilitate reform by protecting officers who question the fairness or effectiveness of departmental policies. For example, both the New York and New Jersey unions publicly sided with officers regarding the negative consequences of imposing unofficial quotas (Fisk & Richardson, 2017). For female officers, the misconduct and discipline principles specified in collective bargaining agreements may provide them with a context where they feel more enabled to exercise their discretion in ways that are consistent with the active representation proposition.

Resources to Assist Victims and Community Policing as Mediators

The disruption thesis has been proposed as one possible explanation for how increasing female representation serves as a catalyst for organizational change (Schuck, 2014, 2017; Schuck & Rabe-Hemp, 2016). In theory, increasing the number of women in the organization creates relational social conflict (Jehn, Northcraft, & Neale, 1999), diversifies associations to include more connections with women and groups representing women (Sklansky, 2006), and strengthens the peripheral values surrounding gender (i.e., the values that members can accept or reject and still function effectively in the organization; see Schein, 2003). In response to the social conflict created by greater female representation in the department, police leaders and organizational gatekeepers begin to formalize cultural beliefs and create institutional rules and organizational structures that reaffirm key values and principles within the agency. For example, in a study of use of force, Schuck and Rabe-Hemp (2016) found that the relationship between the percentage of women in the workforce and the amount of citizens’ complaints was a function of the number of rules, policies, and structures designed to identify and quantify complaints.

Leaders are an important part of the disruption process. Police executives and informal gatekeepers, regardless of their gender or personal beliefs, are under pressure to frame issues related to gender in ways that are consistent with the dominant beliefs of their workforce. To maintain their legitimacy, leaders in bureaucratic organizations need to reflect the value system of those whom they lead. Research suggests that law enforcement executives have limited power and are restricted in their ability to make policy, to exercise authority over hiring and firing personnel, and to bring about major changes in the operating philosophy of the department; instead, police reformers tend to focus on creating partnerships, inside and outside the organization, to realize meaningful change (Mastrofski, 2002). In theory, leaders from agencies with higher levels of female representation will be more likely to grant sovereign status to representatives who frame gendered issues in ways that benefit women because gender diversity within the organization is associated with less group solidarity and stronger peripheral values around gendered topics. These factors contribute to an organizational climate that allows leaders to maintain or even increase their legitimacy while supporting institutional rules and organizational structures that are supportive of women.

One organizational structure that may affect the reporting and processing of rape offenses is the dedication of full-time personnel to investigate incidents and provide services to victims. Scholars hypothesize that victim-focused community responses are associated with more positive criminal justice outcomes (DePrince, Belknap, Labus, Buckingham, & Gover, 2012; Lord & Rassel, 2000). Although considerable research focuses on the positive effects of services on quality-of-life outcomes for victims of violence, including mental health and physical and emotional functioning (Bybee & Sullivan, 2002), little research examines the impact of police-centered units (Epstein & Langenbahn, 1994). Nevertheless, in theory, viewing victims’ participation in the process as important and dedicating resources to working with victims should improve the levels of reporting, case processing, and justice-related outcomes. Furthermore, research suggests that experience investigating sexual assault cases and additional training may improve justice-related outcomes by changing officers’ attitudes toward victims and by better preparing them for the emotional labor involved in rape investigations (Brown & King, 1998; Kinney, Bruns, Bradley, Dantzler, & Weist, 2008; Lonsway, Welch, & Fitzgerald, 2001; Page, 2007; Schuller & Stewart, 2000).

Another potentially important organizational factor in the active translation process may be the implementation of community policing practices. Although no universal definition exists for community policing, most scholars agree that community policing consists of three interconnecting elements: (a) the implementation of problem solving that targets the underlying causes of crime, (b) an emphasis on citizens’ roles as partners in the coproduction of community safety, and (c) organizational restructuring to perform the tasks of problem solving and citizen engagement (Bayley, 1994; Rosenbaum & Lurigio, 1994). The community policing philosophy requires giving residents a greater voice in setting the department’s priorities and in shifting the police’s focus from responding to individual events to engaging in proactive prevention strategies. It emphasizes community building by focusing on crime prevention, building rapport with residents, and providing victims with assistance and connecting them to services in their community (Mastrofski, Worden, & Snipes, 1995).

Hypotheses

Drawing from the theory of representative bureaucracy and the moderator and mediator framework articulated above, the following hypotheses will be tested:

Hypothesis 1: Greater female representation in police organizations will be associated with higher reporting rates for rape offenses and higher clearance rates for rape cases.

Hypothesis 2: The effect of female representation on the reporting and clearing of rape cases will be stronger in agencies where the officers are represented by a police union than in agencies where they are not represented by a union.

Hypothesis 3: Greater female representation in police organizations will be associated with higher rape reporting and clearance rates through the mechanisms of more resources to assist victims and greater implementation of community policing practices.

Method

Data

Information on female officer representation, organizational moderators and mediators, and other agency characteristics were collected from the Law Enforcement Management and Administrative Statistics (LEMAS) survey for the years 1997, 2000, 2003, 2007, and 2013 (U.S. Department, of Justice, 2006, 2008a, 2008b, 2011, 2015). The LEMAS project is a nationally representative survey of U.S. law enforcement agencies that is conducted every 3 to 4 years by the Bureau of Justice Statistics (BJS). A variety of agency data are collected, including information on personnel, pay and benefits, budget and fiscal issues, technology and information systems, use of force and officer safety, and policing initiatives. The sampling strategy used for the LEMAS data collection system produces two samples, one self-representing (SR) sample of agencies with 100 or more sworn officers who receive an invitation to participate in every data collection cycle and one non-self-representing (NSR) sample of agencies with fewer than 100 sworn personnel who only receive an invitation to participate if selected through the random selection process implemented by BJS representatives. Because program managers use this sampling strategy, agencies in the SR sample can be matched over the series to produce a longitudinal data set of police organizations.

Using both the NSR and the SR samples is important because it allows the hypotheses to be tested with a variety of sized agencies. While the longitudinal structure of the SR sample facilitates a more methodologically rigorous test of the theory of representative bureaucracy than the cross-sectional data, the SR sample is restricted to large departments, which only represent about 5% of all local police agencies in the United States (Reaves, 2011). Furthermore, because of missing data, especially missing information about rape incidents across the five waves, the sample size is relatively small (i.e., 315 for the hypotheses tests vs. 542 SR agencies in the 2013 LEMAS sample).¹ The cross-sectional sample, which includes all local police agencies (both NSR and SR agencies), allows the hypotheses to be tested with small-, medium-, and large-sized organizations. Furthermore, the addition of the five waves of cross-sectional data facilitates the examination of trends in the data over time for the larger and more heterogeneous set of law enforcement organizations.

Community data were gathered from the American Community Survey (ACS; U.S. Census Bureau, 2016) and the 2000 and 1990 Decennial Census (Minnesota Population Center, 2016). Departments were matched to their respective communities using census place codes or county codes if place codes were not available. The ACS 5-year estimates for 2011-2015 were matched to the 2013 LEMAS sample, the ACS 5-year estimates for 2005-2009 were matched to the 2007 LEMAS sample, the average between the ACS 5-year estimates for 2005-2009 and the 2000 Decennial Census data was matched to the 2003 LEMAS, the 2000 Decennial Census was matched to the 2000 LEMAS sample, and the weighted average of the 2000 and 1990 Decennial Census was matched to the 1997 LEMAS sample (i.e., .75 for 2000 and .25 for 1990).

Information regarding the number of rape cases that were reported and cleared was taken from the Uniform Crime Reports (UCR) for the years 1997, 2000, 2003, 2007, and 2012 (Federal Bureau of Investigation, 2002, 2005, 2009, 2014b, 2016). The year 2012 was used because agencies began reporting statistics for the new definition of rape starting in 2013. The UCR provides information on the number of women in each law enforcement jurisdiction, regardless of age, who were victims of a forcible rape, an attempted forcible rape, or an assault with the intent to rape. Sexual attacks on men are not included. The UCR also provides information on the number of rape incidents that were cleared by arrest or exceptional means. For a rape incident to be cleared by arrest, three conditions must be met: (a) a person must be arrested; (b) the person must be charged with the offense; and (c) the case must be turned over to the courts for prosecution. Cases cleared by exceptional means involve situations in which an offender has been identified, but, for reasons outside the agency’s control, the suspect cannot be arrested, charged, or prosecuted.

Samples

Using the data described above, five cross-sectional samples and one longitudinal sample consisting of five waves of data were created. The samples were restricted to local municipal police departments that had at least one full-time sworn officer and reported UCR data to the FBI. The final sample sizes for the cross-sectional samples were 1,664 for 2013, 1,504 for 2007, 1,507 for 2003, 1,472 for 2000 and 1,309 for 1997. The final sample size for the longitudinal sample was 315 agencies.

Measures

Female representation was measured as the percentage of full-time (i.e., working 40 or more hours weekly) sworn female officers employed by the agency (see Table 1 for descriptive statistics). The police union variable indicated whether a collective bargaining unit existed to negotiate on behalf of sworn personnel (coded 0 = no and 1 = yes). The victim assistance variable measured how much attention and resources the organization gives to assisting victims (coded 0 = problem not addressed, 1 = no dedicated personnel, 2 = some dedicated personnel or a special unit with part-time personnel, and 3 = a special unit with full-time personnel). Items for the community policing scale were taken directly from the community policing section of the LEMAS survey and are consistent with other research on the topic (Morabito, 2010). Because there were changes to the section over the series, the variable was computed as a percentage and represents the topics in that section for that year.² The following items were used (coded 0 = no and 1 = yes): (a) created or maintained a written community policing plan or statement; (b) trained all in-service personnel in community policing; (c) trained citizens in community policing; (d) ran a citizens’ academy; (e) assigned patrol officer to geographic areas or beats; (f) assigned detectives responsibility for geographic areas or beats; (g) encouraged officers to engage in scanning, analysis, response, and assessment (SARA)-type problem solving; (h) used problem-solving project in the evaluation criteria for patrol officers; (i) formed problem-solving partnerships with organizations in the community; (j) upgraded technology to support community policing; and (k) conducted, sponsored, or utilized a survey of residents. The items were summed for each wave to create a scale, with higher values indicating greater implementation of community policing. Using the Kuder–Richardson Formula 20 method, the internal consistency alphas for the community policing scales are .60 for 2013, .82 for 2007, .82 for 2003, .80 for 2000, and .76 for 1997.

Table 1.

Descriptive Statistics.

	1997		2000		2003		2007		2013
	n	M (SD)	n	M (SD)	n	M (SD)	n	M (SD)	n	M (SD)
Cross-sectional data
Female officers	1,309	6.58 (5.70)	1,472	6.93 (6.35)	1,507	7.75 (6.69)	1,504	8.57 (6.68)	1,664	8.54 (6.99)
Black officers	1,309	6.34 (10.69)	1,472	6.19 (11.28)	1,507	6.04 (10.51)	1,504	6.13 (11.25)	1,664	5.58 (10.49)
Size	1,309	23.91 (17.16)	1,472	23.67 (16.77)	1,507	23.01 (12.47)	1,504	26.64 (144.53)	1,664	21.85 (13.50)
Vertical differentiation	1,309	1.59 (1.69)	1,472	1.49 (0.76)	1,507	1.59 (.84)	1,504	1.67 (0.83)	1,664	1.58 (0.90)
Union	1,309	0.63 (0.48)	1,472	0.62 (0.49)	1,507	0.45 (0.50)	1,504	0.57 (0.49)	1,664	0.88 (0.32)
Economic disadvantage	1,309	0.05 (2.63)	1,472	0.06 (2.56)	1,507	0.10 (2.57)	1,504	−0.01 (2.39)	1,664	−0.03 (2.40)
Community policing	428	58.28 (24.72)	673	49.34 (26.14)	480	59.17 (23.51)	497	60.22 (23.34)	1,457	49.50 (25.67)
Victim assistance	428	1.64 (1.13)	673	1.58 (1.10)	480	1.73 (1.05)	497	1.84 (1.06)	1,457	1.39 (0.92)
Rape reports	1,309	34.14 (108.57)	1,472	28.68 (91.18)	1,507	27.91 (86.37)	1,504	27.63 (76.22)	1,664	21.67 (67.81)
Clearance rape cases	1,059	20.21 (64.79)	1,151	16.02 (51.65)	1,192	14.32 (43.88)	1,211	12.87 (38.32)	1,290	9.97 (33.07)
Longitudinal data
Female officers	315	9.25 (4.59)	315	9.71 (4.72)	315	10.21 (4.63)	315	10.92 (4.98)	315	11.20 (4.80)
Black officers	315	9.63 (11.14)	315	9.90 (11.66)	315	10.07 (12.08)	315	10.12 (12.14)	315	10.38 (12.97)
Size	315	22.38 (8.19)	315	21.95 (8.45)	315	21.69 (8.14)	315	21.78 (8.10)	315	20.49 (8.06)
Vertical differentiation	315	2.20 (2.29)	315	2.09 (0.71)	315	2.20 (0.73)	315	2.30 (0.79)	315	2.30 (.95)
Union	315	0.71 (0.45)	315	0.72 (0.45)	315	0.63 (0.48)	315	0.68 (0.47)	315	0.92 (0.27)
Economic disadvantage	315	0.42 (2.36)	315	0.55 (2.33)	315	0.78 (2.45)	315	0.50 (2.14)	315	0.70 (2.21)
Community policing	253	60.91 (25.69)	253	59.25 (22.89)	253	61.77 (22.97)	253	61.03 (22.93)	253	64.20 (22.29)
Victim assistance	253	1.62 (1.16)	253	1.84 (1.03)	253	1.81 (1.13)	253	1.93 (1.02)	253	1.89 (1.01)
Rape reports	315	97.06 (192.08)	315	87.63 (170.03)	315	86.64 (160.71)	315	78.42 (133.69)	315	74.03 (133.53)
Clearance rape cases	299	48.76 (105.11)	299	42.28 (90.98)	299	36.45 (74.09)	299	31.62 (67.33)	299	29.07 (62.24)

Drawing from the literature, four additional measures were included as control variables. Black officer representation was measured as the percentage of full-time sworn Black officers, and size of the organization was measured as the number of full-time officers in the department per 10,000 residents. Vertical differentiation was calculated by subtracting the minimum patrol officer’s salary from the senior executive’s salary and then dividing it by the minimum patrol officer’s salary. Economic disadvantage was measured as the sum of three standardized variables—percent of residents in poverty, percent of unemployed residents, and the percent of residents receiving public aid. The community measures were drawn from prior research on rape (Mustaine, Tewksbury, Corzine, & Huff-Corzine, 2013) and from the application of social disorganization theory to victimization (Sampson, Raudenbush, & Earls, 1997). A principal components factor analysis with varimax rotation was estimated, and the results suggested that the construct was unidimensional, with all the variables loading high (i.e., >.70) on one component for all samples. The variable was created by standardizing the three items and adding them together.

Analytic Techniques

Structural equation models (SEM) were used to test the hypotheses with the cross-sectional data and growth curve models (GCM) were used to test the hypotheses with the longitudinal data. All models were estimated with Stata 14 (StataCorp, 2015). Because the dependent variables were counts with low arithmetic means, and agencies vary in terms of the at-risk population, the Poisson distribution with an offset variable was used to provide sufficient statistical power while maintaining the appropriate Type I error rate (Rabe-Hesketh & Skrondal, 2012). The interpretation of the coefficient for the Poisson regression is for every one-unit change in the independent variable, the coefficient is the predicted change in the logarithm of the count in the dependent variable (Allison, 2012). Another way to interpret Poisson regression coefficients is in terms of incident rate ratios (IRR). The IRR is obtained by exponentiating the Poisson regression coefficient. The interpretation for the IRR is for every one-unit change in the independent variable, the rate for the dependent variable is predicted to change by a factor of the IRR value. For the GCM analyses, the intercept represents the average value of the rape variable in 1997, and the time variable represents the average rate of growth in the rape variable for each wave. The coefficient for female representation is the estimated effect of more female officers in 1997, and the interaction between female representation and time is the effect of more female officers on the average growth rate in the rape variable for each wave. While there is no direct analogue to R² in Poisson regression, the maximum likelihood method produces a deviance value that can be used to assess the fit of the model. This pseudo-R² value represents the reduction in deviance due to the predictors and does not indicate the absolute fit of the model like in ordinary least squares regression (Rabe-Hesketh & Skrondal, 2012). Prior to testing the hypotheses, the data were screened for problems using the procedures outlined by Tabachnick and Fidell (2007).³

Results

Main Effects

The results for the first hypothesis are presented in Table 2. All models include controls for Black officers, size of organization, vertical differentiation, collective bargaining, and economic disadvantage. The number of women in the community was used as the exposure variable for the rape reports, and the number of rape reports was used for the clearance models. Clearance models only include agencies that had at least one reported rape. In general, the findings are supportive of the first hypothesis that female representation is associated with the reporting and processing of rape cases. The pseudo-R² values for the SEM models are as follows: .41 for 1997, .42 for 2000, .35 for 2003, .24 for 2007, and .21 for 2013. The GCM findings from the longitudinal sample suggest that, over the five waves, data agencies experienced an increase in rape reports and that agencies with more female officers experienced a greater increase in reported rape incidents than departments with fewer female officers (i.e., for every 1% increase in female officers, the upward trend in reporting is expected to increase by a factor of 1.002).

Table 2.

Poisson Regression Results for Female Officer Representation on Rape Report and Clearance Rates.

	Rape reports			Clearance rape cases
	n	Coefficient	SE	n	Coefficient	SE
Cross-sectional data
1997 FR	1,309	.060***	.001	1,059	−.002	.002
2000 FR	1,472	.065***	.001	1,151	.005**	.002
2003 FR	1,507	.062***	.001	1,192	.004*	.002
2007 FR	1,504	.042***	.001	1,211	.018***	.002
2013 FR	1,664	.050***	.001	1,290	.018***	.002
Longitudinal data
Intercept		−7.634***	.077		−1.243***	.061
FR		−.004*	.002		−.011***	.003
T		.018*	.009		−.052***	.014
FR × T		.002***	.001		.008***	.001
	Agency n = 315Observations over time n = 1,575			Agency n = 299Observations over time n = 1,496

Note. All models include percent Black officers, size of organization, vertical differentiation, collective bargaining, and economic disadvantage. FR = female representation; T = time.

p < .05. **p < .01. ***p < .001.

Turning our attention to the clearance of rape reports, the SEM findings from four of the five cross-sectional samples suggest that female representation is associated with higher clearance rates. The results from the 1997 sample are not significant. The increase in the size of the effect over the other four waves suggests that the idea of a tipping point may be relevant. The pseudo-R² values for the SEM clearance models are as follows: .01 for 1997, .01 for 2000, .01 for 2003, .04 for 2007, and .03 for 2013. The results from the longitudinal sample suggest that over time, clearance rates decreased, but that agencies with more female officers had a less steep decline than agencies with fewer female officers.⁴

Police Unions as Moderators

Overall, the findings are consistent with the hypothesis that the relationship between female representation and the reporting and processing of rape cases is moderated by the presence of a police union (see Table 3). The SEM interactions for female representation and police union for all cross-sectional samples are significant and positive, indicating that when officers are represented by a collective bargaining unit, the effect of female representation is stronger than when officers are not represented by a collective bargaining unit. The pseudo-R² values for the SEM moderation models were similar to those reported above for the main effects models. The three-way interaction between female representation, collective bargaining, and time is also significant in the longitudinal sample for rape reports, which suggests that agencies with more female officers and a police union experienced an increase in rape reports over the five waves compared with departments with fewer female officers and no police union. For clearance rates, the three-way interaction was not significant, but the two-way interaction between female representation and union was indicating that there was a strong interaction effect in 1997 that did not change over the five-wave period.⁵

Table 3.

Poisson Regression Results for Moderation Effect of Police Union.

	Rape reports			Clearance rape cases
	n	Coefficient	SE	n	Coefficient	SE
Cross-sectional data
1997 FR	1,309	.035***	.002	1,059	−.025***	.003
1997 U		−1.093***	.027		−.358***	.041
1997 FR × U		.027***	.002		.029***	.003
2000 FR	1,472	.031***	.002	1,151	−.010**	.003
2000 U		−1.337***	.025		−.177***	.042
2000 FR × U		.045***	.002		.020***	.003
2003 FR	1,507	.060***	.001	1,192	−.016***	.003
2003 U		−.606***	.019		−.380***	.042
2003 FR × U		.005***	.001		.029***	.003
2007 FR	1,504	.017***	.001	1,211	.006*	.003
2007 U		−1.393***	.022		−.359***	.042
2007 FR × U		.049***	.002		.020***	.003
2013 FR	1,664	.007*	.003	1,290	−.009^†	.006
2013 U		−1.388***	.040		−.560***	.074
2013 FR × U		.049***	.003		.031***	.006
Longitudinal data
Intercept		−7.714***	.085		−1.171***	.084
FR		.001	.003		−.018**	.005
U		.113	.047		−.052	.071
T		.066***	.017		−.018	.026
FR × T		−.001	.001		.005**	.002
U × T		−.088***	.017		−.062*	.026
FR × U		−.003	.004		.016**	.005
FR × U × T		.004**	.001		.002	.002
	Agency n = 315Observations over time N = 1,575			Agency n = 299Observations over timeN = 1,496

Note. All models include percent Black officers, size of organization, vertical differentiation, and economic disadvantage. FR = female representation; U = union; T = time.

†

p < .10. *p < .05. **p < .01. ***p < .001.

Resources to Assist Victims and Community Policing as Mediators

The results for the mediator SEM models are presented in Figure 1. Because the victim’s assistance question was only asked on the long-form version of the survey prior to 2013, the sample sizes for the earlier years are smaller. The results are generally supportive of the third hypothesis, with a few notable exceptions. The effect of community policing changes over time, with little or no effect in 1997 and a stronger positive effect in 2013. The results for the longitudinal sample suggest that community policing is significantly related to clearance rates but not reporting rates (see Table 4). More resources for assistance to victims is related to an increase in both reporting and clearance rates over time. It is important to note that the direct effects between female officers and the reporting and clearing of rape cases are significant, suggesting that there are additional pathways by which female representation affects women-centered outcomes. The pseudo-R² values for the rape reports mediator SEM models are as follows: .32 for 1997, .37 for 2000, .28 for 2003, .29 for 2007, and .26 for 2013. The pseudo-R² values for clearance mediator SEM models are as follows: .02 for 1997, .02 for 2000, .03 for 2003, .06 for 2007, and .06 for 2013.

Figure 1.

Results for mediation effects of victims’ assistance and community policing.

Table 4.

Growth Curve Model Poisson Results for Mediation Effects of Victims’ Assistance and Community Policing.

	Rape reports		Clearance rape cases
	Coefficient	SE	Coefficient	SE
Intercept	−7.487***	.087	−.578***	.073
FR	−.006*	.002	−.011**	.003
BO	.002	.002	.001	.003
S	.001	.002	−.003	.002
VD	−.008	.010	−.019	.014
U	.054**	.016	.107***	.024
ED	.051***	.006	−.019*	.008
CP	−.001	.001	−.001	.001
VA	−.013*	.006	−.001	.009
T	−.015	.012	−.261***	.018
FR × T	.003***	.001	.007***	.001
BO × T	−.003***	.001	−.001*	.001
S × T	.001	.001	.001*	.001
VD × T	−.012***	.003	.010*	.004
U × T	−.033***	.006	−.016	.009
ED × T	−.002	.001	−.011***	.002
CP × T	−.001	.001	.001***	.001
VA × T	.009***	.002	.011**	.004
	Agency n = 253Observations over timeN = 1,265		Agency n = 241Observations over timeN = 1,205

Note. FR = female representation; BO = Black officers; S = size; VD = vertical differentiation; U = union; ED = economic disadvantage; CP = community policing; VA = victim assistance; T = time.

p < .05. **p < .01. ***p < .001.

Discussion and Conclusion

The topic of female representation in policing is important because the concept of demographic representation in public institutions is consistent with the principles proposed in democratic societies. Among the public, there is “widespread acceptance of the notions that in a democracy public bureaucracies ought to be representative in meaningful ways of the citizen-clients they serve” (Hale & Kelly, 1989, p. 9). The representations of different groups have become powerful symbols of access, equality, and influence. Participation in the occupation of law enforcement symbolizes equality and socioeconomic opportunities for female police officers, but more importantly, women’s participation in the institution of policing symbolizes women’s right to stand beside, instead of behind, men in exercising state-sanctioned acts of formal social control.

Research on how female representation affects the victim’s experiences with the justice system also has important policy implications. According to the data from the National Intimate Partner and Sexual Violence Survey, nearly 22 million women in the United States, or 1 out of 5, have been a victim of rape (Black et al., 2011). Research suggests that the majority of women do not report their victimization to the police (Planty et al., 2013) and that women who do often have negative experiences and suffer harmful psychological stress from their interactions with officers and other criminal justice personnel (Bachman, 1998; Campbell & Raja, 1999). Understanding why some law enforcement agencies are more sensitive and responsive to victims and more effective in processing rape cases is important for improving victims’ experiences with the justice system and increasing the legitimacy of the institution of policing.

The results from this study are consistent with the active translation proposition derived from the theory of representative bureaucracy. The findings suggest that greater female representation in law enforcement is associated with higher reporting rates for rape incidents and higher clearance rates for rape cases. The inconsistent findings among current studies (cf. Meier & Nicholson-Crotty, 2006; Morabito et al., 2017; Walfield, 2016) may be attributed to differences in sample composition and statistical power, as well as disagreements regarding the theoretical role that organizational factors play in women-centered services.⁶ The current study expands on prior research by utilizing a large number of diverse law enforcement agencies over a long period of time, employing rigorous statistical techniques and a longitudinal design, and incorporating organizational characteristics as moderators and mediators in the process of active translation.

Moving forward, more research is needed. Research is needed to understand how female representation influences the reporting of rape cases, given that many victims may not know how many female officers are in the department or whether a female officer will be processing their case. Are departments with greater gender diversity better at facilitating the reporting of rape incidents as an organization, or is it that female officers themselves are more likely to convince women to follow through with an incident report? Either way, given the results of this study, departments should publicize female officers’ achievements in an effort to inform community members about gender diversity within the agency. Research is also needed to examine whether the theory of representative bureaucracy applies to the new definition of rape. Starting in 2013, the UCR’s definition of rape was changed to include a wider range of offenses against both women and men (Federal Bureau of Investigation, 2014a). Questions remain as to whether female representation in policing will be associated with changes in the reporting and processing of a more diverse set of sexual assault cases, as well as sexual assault cases involving male victims.

More research is also needed to identify specific mechanisms involved in the theory of representative bureaucracy. Understanding the causal mechanisms is not only important for theoretical reasons but is also necessary to inform public policy. If one of the goals is to leverage concepts from theory to provide better services to female victims who have been historically underserved by police agencies, identifying the mechanisms that are responsible for the observed correlations between female representation and women-centered services is crucial. This is important because each mechanism may be associated with a different managerial policy for addressing the problem related to inadequate services for female sexual assault victims.

Currently, there are five general categories of mechanisms that may be worthy of further review. First, drawing from the more traditional literature on representative bureaucracy, public officials are hypothesized to act on behalf of citizens who match their demographic characteristics (Meier & Nicholson-Crotty, 2006). From this perspective, female officers act on behalf of women in their community out of a sense of moral obligation or feelings of empathy toward individuals like themselves. One of the biggest challenges regarding this perspective is the presence of heterogeneity with the population of female officers and whether gender serves as the master status in the translation process (Rabe-Hemp, 2009; Schuck & Rabe-Hemp, 2017).

Second is the idea that greater female officer representation is associated with more positive services for women because female officers have better interpersonal and emotional labor skills than male officers and are more effective at using these skills when interacting with victims of sexual violence (Martin, 1999; Rabe-Hemp, 2008; Schuck, 2014). Some scholars suggest that female police executives are likely to adopt a transformative leadership style characterized by participation, cooperation, and inclusion, which is a better fit with modern theories of police management and more likely to promote progressive views of services for victims of sexual violence (Silvestri, 2003). Another potentially important mechanism focuses on community members and suggests that greater symbolic representation is associated with citizens who are more willing to work with bureaucrats who have similar demographic characteristics (Riccucci et al., 2014).

Fifth and finally, there is some evidence that increasing female representation creates changes in the police agency by way of affecting organizational structures and institutional rules (Schuck & Rabe-Hemp, 2016). In theory, greater female representation in policing is associated with greater reporting of rape cases and higher clearance rates through the pathway of formalizing rules and structures that are beneficial to female victims. The relationship is rooted in the assumption that the presence of more women in a police agency is associated with relational social conflict, new associations, and strengthened peripheral values related to women. These changes create an environment in which forces inside and outside the organization have a greater likelihood of successfully implementing strategies that represent women and that reflect the needs of female citizens. Consistent with this idea, the findings from the cross-sectional SEM analyses suggest that greater female representation is associated with more resources dedicated to victims and greater implementation of community policing activities, both of which were subsequently related to higher reporting and clearing rates for rape cases. Findings from the longitudinal GCM analyses support the disruption proposition regarding clearance rates, but not the relationship between community policing and reporting rates.

In addition to studying mechanisms, it is also crucial to identify key moderators. The results from this study suggest that the process of active translation is moderated by larger situational factors. The effect of female representation on the reporting and clearing of rape cases was stronger in organizations where the workforce was represented by a collective bargaining agreement than in departments where there was no collective bargaining agreement. These findings are consistent with the idea that the processes of active translation may be contingent on the resources on which female officers can draw to support their discretionary decisions. Moving forward, more research is needed to identify other organizational or environmental factors that may moderate the relationship between female representation and women-centered outcomes.

This study is not without limitations. Knowledge regarding the quality of data from the UCR and from the LEMAS survey is limited, and the accuracy of the information that is provided most likely varies across agencies. The sample was restricted to municipal police agencies that completed the LEMAS survey, reported crime statistics to the UCR, and had available estimates regarding environmental characteristics from the American Communities Survey. More robust qualitative measures regarding how agencies implement victim assistance units and actualize community policing initiatives would allow for a more complete understanding of the process. Finally, there is controversy surrounding the use of exceptional means to clear rape cases (Spohn & Tellis, 2012), and because this study did not measure quality of services directly or outcomes associated with legal intervention, more research is needed to understand how different types of case processing affect female victims.

Reform is an important part of the history of American policing. Over time, the public has sought to create law enforcement organizations that are more fair and effective, as well as more responsive and accountable to a broader cross section of the people whom they serve. Looking ahead, the findings from this study suggest a need to examine more closely the processual dynamics around women in policing. Evaluating the intricacies of women’s expanding role in the exercise of social control across time and location has the potential to expatiate our understanding of gendered dynamics in law enforcement. Furthermore, identifying the mechanisms by which police practices change in relation to increasing numbers of women entering the profession may help to close the gap between the ideals of policing and the bureaucratic practices of law enforcement officials.

Footnotes

Declaration of Conflicting Interests

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

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biography

Amie M. Schuck is an associate professor in the Department of Criminology, Law and Justice at the University of Illinois at Chicago. Her research interests include: the administration of justice, communities and crime, and quantitative methodology.

References

Alderden

Ullman

(2012). Gender difference or indifference? Detective decision making in sexual assault cases. Journal of Interpersonal Violence, 27(1), 3-22.

Alicke

Yurak

(1995). Perpetrator personality and judgments of acquaintance rape. Journal of Applied Social Psychology, 25, 1900-2100.

Allison

P. D.

(2012). Logistic regression using SAS: Theory and application (2nd ed.). Cary, NC: SAS Institute.

Anderson

Cooper

Okamura

(1997). Individual differences and attitudes toward Rape: A meta-analytic review. Personality and Social Psychology Bulletin, 23, 295-315.

Bachman

(1998). The factors related to rape reporting behavior and arrest new evidence from the national crime victimization survey. Criminal Justice and Behavior, 25(1), 8-29.

Bayley

(1994). Police for the future. New York, NY: Oxford University Press.

Black

M.C.

Basile

K.C.

Breiding

M.J.

Smith

S.G.

Walters

M.L.

Merrick

M.T.

Chen

Stevens

M.R.

(2011). The National Intimate Partner and Sexual Violence Survey (NISVS): 2010 summary report. Atlanta, GA: Centers for Disease Control and Prevention.

Brown

King

(1998). Gender differences in police officers attitudes towards rape: Results of an exploratory study. Psychology, Crime and Law, 4, 265-279.

Bybee

Sullivan

(2002). The process through which an advocacy intervention resulted in positive change for battered women over time. American Journal of Community Psychology, 30, 103-132.

10.

Campbell

Raja

(1999). Secondary victimization of rape victims: Insights from mental health professionals who treat survivors of violence. Violence and Victims, 14, 261-275.

11.

Coleman-Selden

(1998). The promise of representative bureaucracy: Diversity and responsiveness in a government agency. Armonk, NY: M. E. Sharpe.

12.

DePrince

Belknap

Labus

J. S.

Buckingham

S. E.

Gover

A. R.

(2012). The impact of victim-focused outreach on criminal legal system outcomes following police-reported intimate partner abuse. Violence against Women, 18, 861-881.

13.

Epstein

Langenbahn

(1994). The criminal justice and community response to rape. Collingdale, PA: Diane Publishing.

14.

Federal Bureau of Investigation. (2002). Uniform crime reporting program data [United States]: Offenses known and clearances by arrest, 2000 (ICPSR03447. v1). Ann Arbor, MI: Inter-university Consortium for Political and Social Research.

15.

Federal Bureau of Investigation. (2005). Uniform crime reporting program data [United States]: Offenses known and clearances by arrest, 2003 (ICPSR04124. v1). Ann Arbor, MI: Inter-university Consortium for Political and Social Research.

16.

Federal Bureau of Investigation. (2009). Uniform crime reporting program data [United States]: Offenses known and clearances by arrest, 2007 (ICPSR25101. v2). Ann Arbor, MI: Inter-university Consortium for Political and Social Research.

17.

Federal Bureau of Investigation. (2014a). Frequently asked questions about the change in the UCR definition of rape. Retrieved from https://ucr.fbi.gov/recent-program-updates/new-rape-definition-frequently-asked-questions

18.

Federal Bureau of Investigation. (2014b). Uniform crime reporting program data: Offenses known and clearances by arrest, 2012 (ICPSR35021. v1). Ann Arbor, MI: Inter-university Consortium for Political and Social Research.

19.

Federal Bureau of Investigation. (2016). Uniform crime reporting program data [United States]: 1975-1997 (ICPSR09028. v7). Ann Arbor, MI: Inter-university Consortium for Political and Social Research.

20.

Fisk

Richardson

L. S.

(2017). Police unions. George Washington Law Review, 85, 712-799.

21.

Frazier

Haney

(1996). Sexual assault cases in the legal system: Police, prosecutor, and victim perspectives. Law and Human Behavior, 20, 607-628.

22.

Gerber

Cronin

J. M.

Steigman

H. J.

(2004). Attributions of blame in sexual assault to perpetrators and victims of both genders. Journal of Applied Social Psychology, 34, 2149-2165.

23.

Haarr

R. N.

(1997). Patterns of interaction in a police patrol bureau: Race and gender barriers to integration. Justice Quarterly, 14(1), 53-85.

24.

Hale

Kelly

(1989). Gender, democracy, and representative bureaucracies. In Hale

M. M.

Kelly

R. M.

(Eds.), Gender, bureaucracy, and democracy: Careers and equal opportunity in the public sector (pp. 3-17). New York, NY: Praeger.

25.

Jehn

Northcraft

Neale

(1999). Why differences make a difference: A field study of diversity, conflict and performance in workgroups. Administrative Science Quarterly, 44, 741-763.

26.

Keiser

Wilkins

Meier

Holland

(2002). Lipstick and logarithms: Gender, institutional context, and representative bureaucracy. American Political Science Review, 96, 553-564.

27.

Kinney

Bruns

Bradley

Dantzler

Weist

(2008). Sexual assault training of law enforcement officers: Results of a statewide survey. Women & Criminal Justice, 18(3), 81-100.

28.

Lonsway

Welch

Fitzgerald

(2001). Police training in sexual assault response process, outcomes, and elements of change. Criminal Justice and Behavior, 28, 695-730.

29.

Lord

Rassel

(2000). Law enforcement’s response to sexual assault: A comparative study of nine counties in North Carolina. Women & Criminal Justice, 11(1), 67-88.

30.

Maltz

M. D.

Targonski

(2002). A note on the use of county-level UCR data. Journal of Quantitative Criminology, 18, 297-318.

31.

Martin

S. E.

(1999). Police force or police service? Gender and emotional labor. The Annals of the American Academy of Political and Social Science, 561, 111-126.

32.

Mastrofski

S. D.

(2002). The romance of police leadership. In Waring & Weisburd

(Eds.), Crime & Social Organization (pp. 153-196). New Brunswick: Transaction Publishers.

33.

Mastrofski

S. D.

Worden

Snipes

(1995). Law enforcement in a time of community policing. Criminology, 33, 539-563.

34.

Meier

(1993). Representative bureaucracy: A theoretical and empirical exposition. In Perry

J. L.

(Ed.) Research in Public Administration (pp. 1-35). Greenwich, CT: JAI Press.

35.

Meier

Nicholson-Crotty

(2006). Gender, representative bureaucracy, and law enforcement: The case of sexual assault. Public Administration Review, 66, 850-860.

36.

Minnesota Population Center. (2016). National historical geographic information system: Version 11.0 [Database]. Minneapolis: University of Minnesota.

37.

Morabito

M. S.

Pattavina

Williams

L. M.

(2017). Active representation and police response to sexual assault complaints. Journal of Crime and Justice, 40, 20-33.

38.

Morabito

S. M.

(2010). Understanding community policing as an innovation: Patterns of adoption. Crime & Delinquency, 56, 564-587.

39.

Mosher

(1968). Democracy and the public service. New York, NY: Oxford University Press.

40.

Mustaine

Tewksbury

Corzine

Huff-Corzine

(2013). Social disorganization and unfounded sexual assault case clearances. Violence and Victims, 28, 90-102.

41.

Oberfield

Z. W.

(2014). Becoming bureaucrats: Socialization at the front lines of government service. Philadelphia: University of Pennsylvania Press.

42.

Page

(2007). Behind the blue line: Investigating police officers’ attitudes toward rape. Journal of Police and Criminal Psychology, 22(1), 22-32.

43.

Pitkin

(1967). The concept of representation. Berkeley: University of California Press.

44.

Planty

Langton

Krebs

Berzofsky

Smiley-McDonald

Loscocco

(2013). Female victims of sexual violence, 1994-2010 (Special Report NCJ. 240655). U.S. Department of Justice, Bureau of Justice Statistics.

45.

President’s Task Force on 21st Century Policing. (2015). Final report of the president’s task force on 21st century policing. Washington, DC: Office of Community Oriented Policing Services.

46.

Rabe-Hemp

(2008). Survival in an “all boys club”: Policewomen and their fight for acceptance. Policing: An International Journal of Police Strategies & Management, 31, 251-270.

47.

Rabe-Hemp

C. E.

(2009). POLICEwomen or policeWOMEN? Doing gender and police work. Feminist Criminology, 4, 114-129.

48.

Rabe-Hesketh

Skrondal

(2012). Multilevel and longitudinal modeling using Stata. College Station, TX: Stata Press.

49.

Reaves

B. A.

(2011). Census of state and local law enforcement agencies, 2008 (NCJ 233982). Washington, DC: Bureau of Justice Statistics. Available from https://www.bjs.gov/

50.

Riccucci

N. M.

Van Ryzin

G. G.

Lavena

C. F.

(2014). Representative bureaucracy in policing: Does it increase perceived legitimacy? Journal of Public Administration Research and Theory, 24, 537-551.

51.

Rosenbaum

Lurigio

(1994). An inside look at community policing reform: Definitions, organizational changes, and evaluation findings. Crime & Delinquency, 40, 299-314.

52.

Sampson

Raudenbush

Earls

(1997). Neighborhoods and violent crime: A multilevel study of collective efficacy. Science, 277, 918-924.

53.

Schein

(2003). Organizational socialization and the profession of management. In Porter

L. W.

Angle

H. L.

Allen

R. W.

(Eds.), Organizational influence processes (pp. 283-294). Armonk, NY: M.E. Sharpe.

54.

Schuck

A. M.

(2014). Gender differences in policing: Testing hypotheses from the performance and disruption perspectives. Feminist Criminology, 9, 160-185.

55.

Schuck

A. M.

(2017). Female officers and community policing: Examining the connection between gender diversity and organizational change. Women & Criminal Justice, 27, 341-362.

56.

Schuck

A. M.

Rabe-Hemp

(2016). Citizen complaints and gender diversity in police organisations. Policing and Society, 26, 859-874.

57.

Schuck

A. M.

Rabe-Hemp

(2017). Inequalities regimes in policing: Examining the connection between social exclusion and order maintenance strategies. Race and Justice. Advance online publication. doi:10.1177/2153368716689491

58.

Schuller

Stewart

(2000). Police responses to sexual assault complaints: The role of perpetrator/complainant intoxication. Law and Human Behavior, 24, 535-551.

59.

Shelley

T. O. C.

Morabito

M. S.

Tobin-Gurley

(2011). Gendered institutions and gender roles: Understanding the experiences of women in policing. Criminal Justice Studies, 24, 351-367.

60.

Silvestri

(2003). Women in charge: Policing, gender and leadership. Portland, OR: Willan.

61.

Silvestri

(2015). Gender diversity: Two steps forward, one step back . . . . Policing, 9(1), 56-64.

62.

Sklansky

(2006). Not your father’s police department: Making sense of the new demographics of law enforcement. The Journal of Criminal Law and Criminology, 96, 1209-1243.

63.

Sleath

Bull

(2012). Comparing rape victim and perpetrator blaming in a police officer sample differences between police officers with and without special training. Criminal Justice and Behavior, 39, 646-665.

64.

Spohn

Tellis

(2012). The criminal justice system’s response to sexual violence. Violence Against women, 18, 169-192.

65.

StataCorp. (2015). Stata statistical software: Release 14. College Station, TX: Author.

66.

Suarez

Gadalla

(2010). Stop blaming the victim: A meta-analysis on rape myths. Journal of Interpersonal Violence, 25, 2010-2035.

67.

Tabachnick

Fidell

(2007). Using multivariate statistics. Boston, MA: Pearson.

68.

United States Department of Justice, Federal Bureau of Investigation. (September 2016). Table 74: Full-time law enforcement employees, by population group, percent male and female. Crime in the United States, 2015. Retrieved from https://ucr.fbi.gov/crime-in-the-u.s/2015/crime-in-the-u.s.-2015

69.

U.S. Census Bureau. (2016). American community survey: 2011-2015 ACS 5 year estimates and 2005-2009 ACS 5 Year Estimates (generated by author using American Fact Finder). Available from https://factfinder.census.gov/

70.

U.S. Department of Justice. (2006). Law Enforcement Management and Administrative Statistics (LEMAS): 2003 sample survey of law enforcement agencies (ICPSR04411. v1). Ann Arbor, MI: Inter-university Consortium for Political and Social Research.

71.

U.S. Department of Justice. (2008a). Law Enforcement Management and Administrative Statistics (LEMAS): 2000 sample survey of law enforcement agencies (ICPSR03565. v2). Ann Arbor, MI: Inter-university Consortium for Political and Social Research.

72.

U.S. Department of Justice. (2008b). Law Enforcement Management and Administrative Statistics (LEMAS): 1997 sample survey of law enforcement agencies (ICPSR02700. v1). Ann Arbor, MI: Inter-university Consortium for Political and Social Research.

73.

U.S. Department of Justice. (2011). Law Enforcement Management and Administrative Statistics (LEMAS), 2007 (ICPSR31161. v1). Ann Arbor, MI: Inter-university Consortium for Political and Social Research.

74.

U.S. Department of Justice. (2015). Law Enforcement Management and Administrative Statistics (LEMAS), 2013 (ICPSR36164. v2). Ann Arbor, MI: Inter-university Consortium for Political and Social Research.

75.

Walfield

S. M.

(2016). When a cleared rape is not cleared: A multilevel study of arrest and exceptional clearance. Journal of Interpersonal Violence, 31, 1767-1792.

76.

Walker

(2008). The neglect of police unions: Exploring one of the most important areas of American policing. Police Practice and Research: An International Journal, 9, 95-112.

77.

Wentz

Archbold

(2012). Police perceptions of sexual assault victims exploring the intra-female gender hostility thesis. Police Quarterly, 15(1), 25-44.