Gender and Public Organization: A Quasi-Experimental Examination of Inclusion on Experiencing and Reporting Wrongful Behavior in the Workplace

Abstract

Most public organizations share values and beliefs and socially constructed patterns of action that guide the behavior of their employees, resulting in their unique organizational culture. Existing literature on police organizations describes an unmistakable culture that celebrates masculine values and a social structure that exists purposely and specifically to repress female officers. Using a survey data set of 1,114 female federal law enforcement officers, this research employs coarsened exact matching to examine perceived inclusion and its effects on women experiencing disrespect by male colleagues and incidences of sexual harassment and sexual discrimination. In addition, reporting behavior is captured for female officers who experience wrongful conduct. The study finds that women who report working in an inclusive organizational culture are less likely to experience pervasive negative attitudes from their male colleagues or occurrences of sexual harassment and sexual discrimination. However, the existence of an inclusive organizational culture did not show a significant difference in reporting sexual harassment or sexual discrimination if women experienced such wrongful behavior.

Keywords

inclusion organizational culture gender law enforcement sexual harassment sexual discrimination

Since the issuance of Executive Order 13583 (2011)—Establishing a Coordinated Government-Wide Initiative to Promote Diversity and Inclusion in the Federal Workforce—many federal agencies have created diversity and inclusion councils to identify and remove barriers to equal employment opportunity while engendering a culture of inclusion (Office of Personnel Management [OPM], 2011). The OPM (2016) defines inclusion “as a set of behaviors (culture) that encourage employees to feel valued for their unique qualities and experience a sense of belonging” (p. 6). However, fostering a culture of inclusion can be difficult for traditionally masculine organizations such as law enforcement that provoke a particular view of women and the nature of policing. As such, the topic of organizational culture—shared values and beliefs and socially constructed patterns of action that guide the behavior of its members (Jennings, 2012; Jung et al., 2009; Ott, 1989; Rice, 2004; Schein, 2010)—in public agencies that are grossly underrepresented by women is particularly relevant.

Existing literature on police organizations describes an unmistakable culture that celebrates masculine values and a social structure that exists purposely and specifically to repress female officers (Dick & Jankowicz, 2001; Franklin, 2007; Hughes, 2011; Paoline, 2003; Waddington, 1999). These values form the basis of an organizational culture detrimental to the very skill sets—communication, critical thinking, problem solving, and conflict resolution—that women are known to bring into law enforcement (Women in Federal Law Enforcement, 2015). Despite women’s contributions to the profession, organizational culture within law enforcement remains male dominated, and women who challenge these cultural norms are at risk of nonacceptance, sexual harassment, and sexual discrimination that adversely affect work assignments, training opportunities, and promotions (Brown, 1998; Dick & Jankowicz, 2001; Haarr, 1997; Keverline, 2003; Rabe-Hemp, 2008; Seklecki & Paynich, 2007; Yu, 2015, 2017). An in-depth look at organizational culture in law enforcement shows there is a need for fundamental change in the police culture; however, missing from this scholarship is the examination of inclusion and its effects on women experiencing and reporting wrongful behavior in the workplace. This article aims to fill this gap.

This research used a survey data set from sworn female officers (n = 1,114) employed by a large federal law enforcement agency. Coarsened exact matching examines perceived inclusion—an organizational culture that values and supports women and treats women fairly and equitably—and its effects on experiencing disrespect by male colleagues, sexual harassment, and sexual discrimination, as well as female officers’ actual reporting behaviors when they experience wrongful conduct. Coarsened exact matching, which is a quasi-experimental method, enables us to match female officers who perceive their organization as inclusive with those female officers who perceive their organization as noninclusive using their individual and work characteristics. By utilizing coarsened exact matching, endogeneity bias derived from individual and work characteristics is reduced to provide the effects of inclusive organizational cultures on overcoming female officers’ occupational barriers and obstacles. Although this study targets a specific federal workgroup, its implication has far-reaching consequences for other occupational workgroups promoting inclusion and gender equity in public employment.

The article is structured in the following manner. The first section offers both a theoretical and a practical framework on gender representation and organizational culture and draws hypotheses from previous literature. The next section introduces the data and method as well as empirical results. The final section offers discussion, implications of the findings, and the limitations of this study for future research.

Theoretical Framework

Representation

The theory of representative bureaucracy—passive, symbolic, and active representation—is an appropriate framework to examine the impact women have in a disproportionate work environment, likely altering an agency’s organizational culture. Passive representation contends that a workforce should have the same characteristics as the overall population it serves (Choi, 2011; Clark, Ochs, & Frazier, 2013; Meier & Nicholson-Crotty, 2006; Mosher, 1982; Smith & Monaghan, 2013). In other words, if women represent 47% of all working Americans (Bureau of Labor Statistics, 2017), then the percentage of women working in all occupations should accompany those figures. Unfortunately, the proportion of women across all occupations and organizations are not evenly distributed, as is the case for the target population of this study, resulting in a noninclusive culture commonly associated with law enforcement. According to the Bureau of Justice Statistics, women account for just 15.5% of all sworn federal law enforcement officers (Reaves, 2012). Despite their higher employment figure—for example, women represent 6.5% of all officers in state agencies, 12.2% in local police departments, and 13.9% in sheriffs’ offices (Burch, 2016; Langston, 2010; Reaves, 2015)—research has shown that the noninclusive culture commonly associated with state and municipal policing also pervades throughout federal law enforcement (Keverline, 2003; Schulz, 2009; Yu, 2015, 2017). However, recent research on symbolic representation contends that the sheer presence of passive representation in a public organization can affect how citizens assess the legitimacy of that organization (Gade & Wilkins, 2013; Riccucci, Van Ryzin, & Lavena, 2014; Theobald & Haider-Markel, 2009). For example, Riccucci et al. (2014) found that the gender composition of a police organization, in this case more women, can causally influence people’s judgment of the organization as being both fair and trustworthy in the context of handling domestic violence complaints, a crime in which women are most often the victim. As women’s passive and subsequent symbolic representation is not expected to change anytime soon in law enforcement, a noninclusive culture will likely remain. To combat the disproportion of female representation, scholars of active representation contend that, under certain settings, female members have the ability to use their discretion to foster an improvement of equity in the workplace and/or enact policy outcomes for those with similar gender origins (Meier, 1993; Mosher, 1982; Sowa & Selden, 2003). For example, Meier and Nicholson-Crotty (2006) found that an increase in female police officers is associated with an increase in reporting sexual assaults and arrests made for sexual assaults. Regardless of mode—passive, symbolic, or active—the applicability of representative bureaucracy toward inclusion and organizational culture is especially interesting because of the long history of law enforcement as a male-dominated profession.

The scholarship on proportional representation, commonly known as tokenism, also clearly demonstrates the impact of a disproportionate work environment. Kanter (1977) argues that an organization having a very low percentage of minority workers—in this case female officers—places a manifold of workplace pressure on those individuals. They encounter strong feelings of isolation and stress because they are highly visible within the organization and are given little margin for error (Belknap & Shelley, 1992). As a result, the error of one woman becomes embellished and applied to all others, resulting in the majority members of an organization to apply negative stereotypes (Martin & Jurik, 2006). Therefore, the indirect consequence of tokenism is the pressure for female officers to conform and manage their gender image, not too feminine, not too masculine (Stivers, 2001). They are required to perform at the same level as male officers while acting within the established female stereotype (Wertsch, 1998). Brown (1998) concurs that the low percentage of women functions as a powerful inhibitor in the equal treatment of women in the workforce and suggests that until women reach a 25% proportion of an organization’s membership, they will continue to endure token status, reinforcing negative stereotypes (Krimmel & Gormley, 2003). Dahlerup (1988, 2005) also expands Kanter’s analysis but argues that 30% is the crucial threshold for impact by a minority group. Again, as the percentage of women in law enforcement is not expected to change anytime soon, the theory of tokenism alone cannot fully describe the experiences of female officers (Franklin, 2007; Greene & Del Carmen, 2002).

Social identity theory also demonstrates the negative impact of a demographic group when interacting with other groups in an organization. The premise of social identity theory is that a person’s social reality is significantly determined by their group membership, such as gender and racial/ethnic affiliation, and can serve as a mechanism to promote superiority over other groups (Ashforth & Mael, 1989; Mor Barak, 2017). The meaning attached to these social categories influences how people interact with those who are both in and out of their own identity group (Mor Barak, 2017). Furthermore, we divide group memberships into “them” and “us” in an effort to create distinct out-groups and in-groups (Sabharwal, 2014, p. 198). The in-group (e.g., men) discriminates against the out-group (e.g., women) to promote a more positive group identity for themselves and to enhance its superiority. This environment of exclusivity promotes a negative working environment of inequality for the out-group and likely increases acts of sexual harassment and sexual discrimination in the workplace. Therefore, the inclusion–exclusion continuum in which individuals participate also represents an organizational culture that appears resistant to women.

Organizational Culture

Over the past couple of decades, organizational culture has been a significant topic of interest in public management research (Di Pietro & Di Virgilio, 2013; Garnett, Marlowe, & Pandey, 2008; Jennings, 2012; Jung et al., 2009; Molina, 2009; Moon, 2000; Muscalu, 2014; Rice, 2004; Taylor, 2014; Teodoro & Hughes, 2012). Although its definition has varied in the literature, many researchers agree that organizational culture consists of shared values and beliefs and socially constructed patterns of action that guide the behavior of its members (Jennings, 2012; Jung et al., 2009; Ott, 1989; Rice, 2004; Schein, 2010). These values and beliefs are both formal and informal and can be both written and unwritten (Rice, 2004). In addition, these patterns serve as a powerful social control function, limiting the range of acceptable behavior and restricting individual differences in organizations (Di Pietro & Di Virgilio, 2013; O’Reilly & Chatman, 1996). Often, organizational culture represents long-standing traditions or assumptions, and socialization of these traditions begins once a member joins the organization. Organizational culture is not, however, without problems. Important shared values and beliefs may lead members to think and act inappropriately, impeding a positive working environment or culture change required for an organization to meet community responsibilities. As organizational culture develops over time, creating a new culture is often met with great resistance (Denhardt, Denhardt, & Aristiqueta, 2012; Muscalu, 2014); however, the necessity for change, if warranted, should not be abandoned. Some public organizations have been successful in institutionalizing change in organizational culture (Starling, 2011). For example, the 2010 repeal of “Don’t Ask, Don’t Tell” within the Armed Forces and the 2015 landmark decision by (then) Defense Secretary Ashton Carter to open all combat positions to women will be recorded as two of the most historic cultural changes to occur at any public organization.

Further investigations into organizational culture identify subcultures within an organization (Taylor, 2014). Organizational subcultures can be based on differences between groups, such as geography, profession, and functional specialization (Trice & Beyer, 1993); however, they can also be based on demographic groupings such as gender. Subcultures constructed by gender are particularly common in nontraditional occupations—those that have fewer than 25% women in their membership—and arise from the shared experiences of the majority gender. According to Schein (2010), if members of an organization discover a way of managing a stakeholder group that is constantly successful, they will develop shared assumptions to maintain that success, hence the existence of the “good ole boy” network. Subcultures derived from groupings such as geography, profession, or functional specialization are mostly benign, especially in large organizations, but conflicting subcultures derived from different assumptions about gender may undermine an organization’s purpose and their capacity to change themselves in response to experience and current events (Mahler, 1997) and potentially serve as a landscape ripe with sexual harassment and sexual discrimination. Miller and Katz (1995) introduce an example of a culture detrimental to any public organization in the modern era—the monocultural organization. The monocultural organization has three key features: domination of one group over another, maintenance of its superiority, and application of exclusionary hiring and membership practices (Miller & Katz, 1995). These attributes have all the makings of a subculture resistant and exclusionary to women. An example of a perceived monocultural organization is law enforcement.

Law Enforcement Culture

There is agreement that the self-governing police culture is a unique occupational subculture that celebrates masculine values that provoke a particular view of women, the purpose of law enforcement, and the roles for which male and female officers are believed to be most appropriate (Dick & Jankowicz, 2001; Franklin, 2007; Hughes, 2011; Paoline, 2003; Skolnick, 2008; Waddington, 1999). The occupation of policing typically associates itself with physical strength, aggressive behavior, and a high degree of male cohesion and solidarity (Dick & Jankowicz, 2001; Rabe-Hemp, 2008). However, women are perceived to be weak, unable to fulfill job requirements, and a physical liability for the organization (Brown, 1998; Martin, 1999; Schulz, 2009; Wadman & Allison, 2004). This recurring image fosters a cultural tradition that characterizes women as physically inferior, despite the fact that physical altercations are infrequent in modern policing (Kurtz, Linneman, & Williams, 2012). The incursion of women into the police culture has the capacity to alter these norms, values, and customs and hence is met with great resistance (Hughes, 2011; Rabe-Hemp, 2008). Therefore, it comes as no surprise that research by Keverline (2003) and Yu (2015, 2017) identifies lack of respect by male colleagues as the biggest barrier for women in federal law enforcement; however, Yu (2015, 2017) also found that a similar or higher percentage of women did not perceive a negative attitude from their male colleagues and were treated with respect and equitability, rejecting the premise that all law enforcement agencies function as a monocultural organization. Accordingly, this study expects female officers who perceive working in an inclusive organizational culture to experience less pervasive negative attitude from their male colleagues.

Hypothesis 1: Female officers who perceive working in an inclusive organizational culture are less likely to experience pervasive negative attitudes from their male colleagues.

To achieve even a limited social acceptability, the law enforcement culture demands women who enter it assume male-like characteristics or behave within the established acceptable female stereotype (Kanter, 1977; Rabe-Hemp, 2008; Schulze, 2012; Wertsch, 1998). For example, female officers who often curse, laugh at sexually inappropriate jokes, or are exceedingly physically fit are often perceived as “one of the boys” and included into the informal networks that are essential to the police culture (Dodge, Valcore, & Klinger, 2010; Lonsway, 2003; Schulze, 2012). Female officers who confront or oppose these cultural norms and gender stereotypes are at risk of sexual harassment (Haarr, 1997; Hunt, 1990; Rabe-Hemp, 2008; Yu, 2015, 2017). The Equal Employment Opportunity Commission (2018a) defines sexual harassment as

unwelcome sexual advances, requests for sexual favors, and other verbal or physical conduct of a sexual nature when this conduct explicitly or implicitly affects an individual’s employment, unreasonably interferes with an individual’s work performance, or creates an intimidating, hostile, or offensive work environment.

In Hunt’s (1990) research on the underlying logic of police sexism, almost two thirds of the women in her study reported incidences of sexual harassment on the job. In addition, Chaiyavej and Morash (2009) report that 90.6% of the policewomen in their study experienced sexual harassment within the past 2 years. Unfortunately, to fit in, many women tolerate the unlawful behaviors as part of the law enforcement culture (Martin, 1990). As a result, most women do not officially report incidences of sexual harassment out of fear of retribution, exclusion by peers and supervisors, and the code of silence (Chaiyavej & Morash, 2009; Collins, 2004; Keverline, 2003; Yu, 2017). The code of silence is a common but unwritten norm in the police subculture that prohibits reporting misconduct against other law enforcement officers (Ivkovic, Haberfeld, & Peacock, 2018). If an officer violates this code, they are ostracized and excluded from the informal networks that are essential to the police culture. On the contrary, some women have reported incidences of sexual harassment if they perceive the act as severe or if organizational policies strongly prohibit sexual harassment, resulting in women feeling valued and more supportive to respond assertively (Chaiyavej & Morash, 2009; Gruber & Smith, 1995). Accordingly, this study expects female officers who perceive working in an inclusive organizational culture are less likely to experience sexual harassment and if so will report wrongdoing.

Hypothesis 2: Female officers who perceive working in an inclusive organizational culture are less likely to experience sexual harassment.

Hypothesis 3: When sexual harassment occurs, female officers who perceive working in an inclusive organizational culture are more likely to report sexual harassment.

Similarly, female officers who resist these cultural norms and gender stereotypes are also at risk of sexual discrimination accentuated by their token and lack of passive status (Brown, 1998; Dick & Jankowicz, 2001; Haarr, 1997; Rabe-Hemp, 2008; Yu, 2015, 2017). The Equal Employment Opportunity Commission (2018b) defines sexual discrimination as treating “any aspect of employment, including hiring, firing, pay, job assignments, promotions, layoff, training, fringe benefits, and any other term or condition of employment” unfairly due to a person’s sex. In Yu’s (2017) research on occupational barriers in federal law enforcement, 29.3% of the women in her study experienced sexual discrimination in regard to work assignments, promotions, and training opportunities. Despite these occurrences and the federal government’s efforts to encourage all employees to report unlawful behavior on the basis of sex (e.g., Notification and Federal Anti-Discrimination and Retaliation Act [No FEAR Act] of 2002), coupled with the code of silence as mentioned previously, less than a third filed a formal report out of fear of retribution and the belief that nothing would be done (Yu, 2017). On the contrary, the majority of the women in Yu’s (2017) study did not experience sexual discrimination, lending some credibility to the ongoing efforts to eliminate discrimination and promote an inclusive workforce in the federal sector. In addition, similar to sexual harassment, some women will respond assertively and report incidences of sexual discrimination if the act is extreme (Haarr & Morash, 2013). Accordingly, this study expects female officers who perceive working in an inclusive organizational culture are less likely to experience sexual discrimination and if so will report wrongdoing.

Hypothesis 4: Female officers who perceive working in an inclusive organizational culture are less likely to experience sexual discrimination.

Hypothesis 5: When sexual discrimination occurs, female officers who perceive working in an inclusive organizational culture are more likely to report sexual discrimination.

Data and Method

Data

This research draws its survey data set from sworn female officers employed by a large federal law enforcement agency. This agency was specifically asked to participate in this research project because it exercises a broad law enforcement mission and its percentage of female officers is 19.5%, almost a quarter higher than the national average. As the main objective of this research is to examine the effects of a perceived inclusive organizational culture on male colleagues’ wrongful behaviors toward female officers, as well as female officers’ actual reporting patterns when they experience sexual harassment and sexual discrimination, collecting data from only female officers is a valid and appropriate measurement of this research. According to Wooldridge (2015), a sample selection based on exogenous factors (i.e., gender) is independent of error terms and is not directly related to outcome variables, precluding any statistical problems.

In 2017, an online Qualtrics survey was sent to all female officers at this organization, using their work email accounts. To improve validity, the definitions of all measured concepts (e.g., inclusion, negative attitude, sexual harassment, and sexual discrimination) were provided with the appropriate survey question. In addition, practitioners who work in federal law enforcement had an opportunity to review the survey instrument prior to dissemination. The survey remained open for 30 days with a reminder email sent mid-study. To increase reciprocity with potential respondents, the survey link was sent by the agency’s human resource division in an email describing the study and encouraging maximum voluntary participation. In addition, all the respondents were guaranteed anonymity. The response rate for the survey was 41.2% and the final sample size was 1,114. Overall, the quantitative questions in the survey asked the respondents for their opinion of their agency’s organizational culture and their experiences with occupational behaviors, including negative attitudes from male colleagues, sexual harassment, and sexual discrimination. In addition, four qualitative questions were included to capture data on additional reporting patterns.

Variables

Treatment variable

The treatment variable refers to whether the respondent defined her agency as inclusive. The survey question used for the treatment variable was as follows: “Do you consider the agency women-friendly? Women-friendly is defined as an agency with an organizational culture that values and supports women and treats women fairly and equitably.” The response categories for this survey item included “yes,” “no,” and “unsure.” This survey item was recoded as a binary variable in which 1 indicated inclusive and 0 indicated noninclusive. That is, “yes” was coded as 1 and “no” and “unsure” were coded as 0. About 47.7% of the respondents (n = 532) defined the agency as inclusive, whereas 52.3% of the respondents (n = 582) defined the agency as noninclusive.

Outcome variables

This research employs five outcome variables to capture the degrees of wrongful behaviors that female officers received from male colleagues and female officers’ actual reporting patterns when they have experienced sexual harassment or sexual discrimination. To understand these wrongful behaviors, three survey items were used: “I experienced pervasive negative attitudes from my male colleagues. Pervasive negative attitudes are defined as beliefs and actions that women cannot handle the job physically or emotionally,” “I experienced sexual harassment in my federal law enforcement employment. Sexual harassment is defined as a sexual advance or proposition with which women must comply or forfeit an employment benefit. In addition, it can also be defined as unwanted sexual behaviors, such as touching, teasing, and making comments about a woman’s appearance or sexuality,” and “I experienced sexual discrimination in my federal law enforcement employment. Sexual discrimination is defined as the practice of letting a person’s sex unfairly become a factor when deciding who receives an initial job offer, promotion, training opportunity, job assignment, compensation, or other employment benefit.” The questions were rated using a 5-point Likert-type scale: “strongly disagree,” “disagree,” “neutral,” “agree,” and “strongly agree.” These three outcome variables were standardized before the analyses. Overall, the negative experiences were predominantly contained within a third of the population—35.8% of the women (n = 390) in this study report experiencing at least one wrongful behavior and 18.8% (n = 204) report experiencing two. Individually, almost a third (30.6%) of the women (n = 333) in this study perceive pervasive negative attitudes from their male colleagues. In addition, 30.2% of the women (n = 336) report experiencing sexual harassment in the workplace. Similarly, 33.7% of the women (n = 375) report sexual discrimination in regard to promotions, training opportunity, job assignment, or other employment benefit.

As a follow-up to those respondents who experienced sexual harassment or sexual discrimination, two additional survey items captured reporting behaviors. These survey items consisted of the following: “If you experienced sexual harassment, did you report it?” and “If you experienced sexual discrimination, did you report it?” The response categories for these two survey items were “yes,” “no,” and “not applicable.” These items were recoded as binary variables after removing respondents who reported “not applicable.” Thus, 1 indicated “yes” and 0 indicated “no.” As a follow-up to those respondents who indicated “no” and did not file a formal report, two additional survey items captured reasons why. These survey items were open-ended and consisted of the following: “If no, why?” Of those respondents who reported experiencing sexual harassment, only 19.3% (n = 65) filed a formal report. Similarly, of those respondents who reported experiencing sexual discrimination, only 22.7% (n = 85) did the same. Altogether, 26 women filed reports for both sexual harassment and sexual discrimination.

Matching covariates

When using matching methodology, researchers are required to select a list of pretreatment matching covariates to create the most comparable units that are the most similar to the treatment units in terms of pretreatment matching covariates. The condition of pretreatment matching covariates is observed characteristics, which are not affected by treatment status. In this research, 11 matching covariates are used to match the demographic and work characteristics of the respondents between the treatment and comparison units. Demographic characteristics were captured by age (26-59), education level (bachelor = 1; master = 2; Juris Doctor = 3; PhD = 4), minority status (non-White = 1; White = 0), marital status (married = 1; otherwise = 0), parental status (mother = 1; no child = 0), number of children (0-7), and single mother status (single mother = 1; otherwise = 0). Work characteristics include grade level (1-9; GS-5/7/9/11/12/13/14/15/SES), supervisory status (yes = 1; no = 0), frequency of relocation (0-10), and work years (1-30). Five variables were binary variables, and other variables were coded as numerical discrete variables (i.e., education level, grade level, relocation frequency, and work years) or a numerical continuous variable (i.e., age). The descriptive statistics of all the variables are reported in Table 1.

Table 1.

Descriptive Statistics.

Variable	n	M	SD	Minimum	Maximum
Treatment variable
Inclusion	1,114	0.47	0.49	0	1
Outcome variables
Negative attitude (standardized)	1,088	−4.30e-09	1	−1.22	2.02
Sexual harassment (standardized)	1,080	−1.02e-08	1	−0.94	2.45
Sexual discrimination (standardized)	1,080	−6.40e-09	1	−1.24	1.78
Sexual harassment report	336	0.19	0.23	0	1
Sexual discrimination report	375	0.22	0.27	0	1
Matching covariates
Age	1,106	41.55	7.17	26	59
Education^a	1,113	1.79	0.77	1	4
Minority status (1 = minority; 0 = nonminority)	1,100	0.20	0.40	0	1
Marital status (1 = married; 0 = otherwise)	1,105	0.64	0.48	0	1
Parental status (1 = mother; 0 = no child)	1,097	0.53	0.49	0	1
Single mother (1 = yes; 0 = no)	1,105	0.08	0.28	0	1
Children number	1,097	1.02	1.12	0	7
Grade level	1,100	5.91	1.05	3	9
Supervisory status (1 = yes; 0 = no)	1,100	0.21	0.40	0	1
Relocation number	1,111	1.75	1.32	0	10
Work years	1,109	12.62	7.04	1	30

1 = bachelor; 2 = master; 3 = Juris Doctor; 4 = PhD.

To check for systematic differences between the respondents who identified as “inclusive” and “noninclusive,” Table 2 shows the mean difference among matching covariate variables. Five variables (i.e., education, single mother, grade level, supervisory status, and relocation frequency) are shown to have a significant difference in means. The matching covariates enable us to determine the most comparable units with regard to the characteristics of the matching covariates; however, those comparable units did not report their workplace as inclusive.

Table 2.

Mean Difference in Two Groups’ Personal and Work Characteristics.

Variable	Inclusion	Noninclusion	Mean difference
Age	41.27 (7.43)	41.80 (6.91)	0.53 (0.43)
Education	1.73 (0.75)	1.85 (0.79)	0.12*** (0.04)
Minority status (1 = minority; 0 = nonminority)	0.19 (0.39)	0.21 (0.41)	0.01 (0.02)
Marital status (1 = married; 0 = otherwise)	0.65 (0.47)	0.62 (0.48)	−0.03 (0.02)
Parental status (1 = mother; 0 = no child)	0.51 (0.50)	0.55 (0.49)	0.03 (0.03)
Single mother (1 = yes; 0 = no)	0.06 (0.24)	0.10 (0.31)	0.04** (0.01)
Children number	0.99 (1.11)	1.04 (1.12)	0.05 (0.06)
Grade level	5.76 (0.99)	6.05 (1.08)	0.28*** (0.06)
Supervisory status (1 = yes; 0 = no)	0.14 (0.35)	0.27 (0.44)	0.12*** (0.02)
Relocation number	1.59 (1.17)	1.89 (1.43)	0.30*** (0.07)
Work years	12.52 (7.36)	12.71 (6.74)	0.18 (0.42)

Note. Standard errors are in parentheses.

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

Coarsened Exact Matching Method

The coarsened exact matching method is utilized to reduce systematic differences in personal and work characteristics between female officers who perceive inclusion and female officers who perceive noninclusion. By comparing differences in outcome variables between the two groups, we are able to examine the effects of an inclusive organizational culture on the negative behaviors that female officers received from their male colleagues, as well as female officers’ actual reporting patterns when they experienced sexual harassment or sexual discrimination. In experimental research, a randomization is conducted to remove any source of endogeneity bias with regard to the observed and unobserved characteristics between the treatment and control units; thus, the only differences, by definition, in the outcome variables are due to the treatment effects (Morgan & Winship, 2007). It is, however, difficult and somewhat unethical to randomly assign female officers to either inclusive or noninclusive workplaces to conduct this type of research.

In the absence of a randomized experiment, a quasi-experimental method called coarsened exact matching uses the “monotonic imbalance bounding” matching method to block units within the same values of coarsened covariates to create the most comparable units in terms of the matching covariates (see Iacus, King, & Porro, 2009). A number of different matching strategies exist, such as propensity score, Mahalanobis distance, and coarsened exact matching. Each matching methodology has a different approach by which to create the most comparable units. The coarsened exact matching uses the exact matching approach to “simply match a treated unit to all the control units with the same covariate values” (Blackwell, Iacus, King, & Porro, 2009, p. 526). Among the various matching strategies, previous research has shown that coarsened exact matching performs very well in terms of unbiasedness, dependence on model specification, balance in covariates, and efficiency, especially when the observed characteristics of the units are heterogeneous (King & Nielsen, 2015).

The key benefit of coarsened exact matching is that researchers are able to argue that the observed characteristics of each comparison unit are exactly the same as the matched treatment unit as “the coarsening on any variable becomes finer. . ., the bound on the maximum imbalance on the moments of that variable becomes tighter” (Blackwell et al., 2009, p. 528). Thus, the only differences in outcome variables after the matching process are due to treatment status, which is inclusive. In this study, we estimated the average treatment effects on the treated (i.e., ATT), as we compared treated units (i.e., individuals who identified their agency as inclusive) with comparable units (i.e., individuals who identified their agency as noninclusive).

STATA15 was used to run coarsened exact matching. First, the level of imbalance was determined in regard to the pretreatment matching covariates between the treated and comparison units. As explained earlier, a series of matching covariates was used to measure the overall imbalance between the treated and comparison units (see Blackwell et al., 2009). To check the imbalance before the matching process, the automatic process of estimating a global measure of imbalance was computed within the sample data set. A global measure of imbalance is referred to as $L_{1}$ , by using “the absolute difference over all the cell values” (Blackwell et al., 2009, p. 530). According to Blackwell et al. (2009), “perfect global balance is indicated by $L_{1} = 0$ and larger values indicate larger imbalances between the groups, with a maximum of $L_{1} = 1^{″}$ (p. 530). That is, a lower $L_{1}$ means a better balance, which is what we aimed for between the two groups. In Appendix A, the global measure of balance using all of the matching covariates is 0.909, which indicates a large imbalance is occurring between the treatment and comparison units when using an unmatched data set. In addition, the univariate balance of each matching covariate is shown in Appendix A. The univariate balance indicates the level of balance between two units as multivariate balance, $L_{1}$ . Furthermore, Appendix A provides the mean and 5-number summaries to show which part of each matching covariate has a larger imbalance (i.e., 0 is perfect balance).

The Stata command CEM is then used to “perform exact matching on coarsened data to determine matches and then pass on the uncoarsened data from observations that were matched to estimate the causal effects” (Blackwell et al., 2009, p. 531). After the coarsened matching process, a multivariate $L_{1}$ became 0.155, which is much lower than before the matching process (see Appendix B). In addition, the univariate $L_{1} s$ were close to 0 after the matching process. These numbers indicate that coarsened exact matching successfully performed well enough to argue that the comparison units are exactly matched with the treatment units with regard to the matching covariates.

The uniqueness of coarsened exact matching is that it uses multiple stratum to match the units within the exact same coarsened matching covariates and drops any observation that is unmatched between the treatment and comparison units (Blackwell et al., 2009; Iacus, King, & Porro, 2012). To maximize information from all observations, each stratum could get at least one matched unit based on block randomization to match treated and untreated units (Blackwell at al., 2009). Thus, the region of common support only covers those comparison units that are exactly matched with the treatment units. Before the matching process, the total sample size was 1,114, and the female officers who reported an inclusive organizational culture was 532 (i.e., treatment units). After the matching process, the remaining units in the region of common support were 116. The treatment units numbered 61, whereas the comparison units numbered 55. Although a relatively large number of observations were left out after the matching process, an argument can be made that there are no systematic differences in the observed matching covariates.

Results

The findings from coarsened exact matching are displayed in Table 3. They support Hypotheses 1, 2, and 4 that female officers within inclusive organizational cultures are less likely to experience pervasive negative attitudes (ATT = −0.966, p < .001), sexual harassment (ATT = −0.459, p < .01), and sexual discrimination (ATT = −1.005, p < .001).

Table 3.

Empirical Results of the Intensities of Negative Experience.

	Average treatment on the treated
	Negative attitudes	Sexual harassment	Sexual discrimination	Sexual harassment report	Sexual discrimination report
Inclusion
Yes vs. No	−0.966*** (0.173)	−0.459** (0.159)	−1.005*** (0.174)	−0.150 (0.172)	−0.098 (0.071)
n	115	114	114	33	33

Note. Outcome variables are standardized. Robust standard errors are in parentheses.

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

The levels of pervasive negative attitudes that female officers within inclusive organizational cultures experienced from male colleagues were 0.966 standard deviations (SD) lower than female officers within noninclusive organizational cultures. Furthermore, the levels of sexual harassment and sexual discrimination from female officers within inclusive organizational cultures were 0.459 SD and 1.005 SD lower than those individuals who reported their agencies as noninclusive. The empirical findings support the argument for the importance of promoting inclusive organizational cultures to prevent wrongful behaviors, including negative attitudes, sexual harassment, and sexual discrimination.

Surprisingly, the empirical findings did not support Hypotheses 3 and 5. Based on the empirical results, working in an inclusive organizational culture does not provide significant impact for female officers to officially report sexual harassment (ATT: −0.150, ns) or sexual discrimination (ATT: −0.098, ns). These results suggest that women employed in an inclusive organizational culture do not lead female officers to report sexual harassment or sexual discrimination. Thus, the empirical findings support that inclusive organizational cultures actually lower the negative work attitudes that female officers have toward their male colleagues, whereas organizational culture itself does not lead female officers to actually report their male colleagues for wrongful behavior. To ensure robustness, respondents who answered “unsure” to inclusive organizational culture were removed so that only female officers who answered “no” became the comparison group. These findings are consistent with the main model (see Appendix C). Female officers who identified their agency as inclusive are less likely to experience negative attitudes (ATT: −1.313, p = .001), sexual harassment (ATT: −0.787, p = .001), and sexual discrimination (ATT: −1.158, p = .001) than those who identified their agency as noninclusive. However, inclusive culture does not lead to positive reporting behaviors when experiencing sexual harassment (ATT: −0.011, ns) or sexual discrimination (ATT: −0.014, ns).

In addition, our findings may be limited if female officers who experienced sexual harassment or discrimination report their agency as noninclusive (i.e., potential reverse-causality explanation is available between the outcome and treatment variables). To reduce potential bias from the reverse-causality explanation, we conducted the instrumental variable approach (i.e., two-stage least squares [2SLS]). Two additional survey questions were utilized as the instruments: adoption of family-friendly policies and pregnancy-friendly policies. These two instruments passed the weak instrument and overidentification tests, providing a plausible argument that outcome variables were not directly affected by these instruments (i.e., exclusion restriction assumption). In other words, the instrumental variable approach only uses the variations from the treatment variable affected by these two instruments, which are not associated by the outcome variables so as to free itself from the reverse-causality explanation (Angrist & Pischke, 2008). Specifically, we can argue that the respondents’ satisfaction with both family-friendly and pregnancy-friendly policies is not directly related to the outcome variables in this study but only through the treatment variable. Appendix D shows that the instrumental variable estimations are consistent with the main model. That is, after controlling for the potential reverse-causality explanation between the outcome and treatment variables, we can still support Hypotheses 1, 2, and 4 but not Hypotheses 3 and 5. Based on these two robustness checks, we can argue that the findings are robust for the adjusted comparison group as well as a different methodological approach to support our findings from the main model.

Discussion and Conclusion

The empirical findings reveal an interesting picture of organizational culture in federal law enforcement and its impact on women experiencing wrongful behaviors in the workplace, as well as their reporting patterns. The treatment status—inclusion—on its own represents the espoused values and beliefs of an organizational group, and the findings can be interpreted in two ways. On one hand, over half (52.3%) of the respondents depict their organization as unsure or noninclusive to women and their equitable treatment, lending credibility to prior research on the continued existence of the masculine police culture resistant to female officers (Brown, 1998; Martin, 1999; Schulz, 2009; Wadman & Allison, 2004; Yu, 2017). Its impact highlights the negative attitudes of male colleagues and a landscape susceptible to sexual harassment and sexual discrimination. As women remain undervalued, they will continue to have difficulties integrating into the informal networks that are essential to the police culture. Although it appears women will tolerate certain aspects of the profession, the “good ole boy” network must be eliminated, particularly in promotion boards, and the male-dominated culture altered to ensure women feel valued and are treated equitably to reduce negative attitudes by male colleagues and incidences of sexual harassment and sexual discrimination. Changing an organizations’ culture is no easy task; however, leaders in any public organization can work to alter organizational culture through changed behavioral patterns. Federal law enforcement agencies must transform and create a gender-neutral culture that does not tolerate harassment and discrimination and implement policies that highlight women’s contributions to policing. For example, in addition to posting annual discrimination complaint data (e.g., total number of complaints filed) on an agency’s public website, as required by the No FEAR Act of 2002, agencies should also ensure they have enforceable organizational policies on sexual harassment and sexual discrimination that provide both legal and practical definitions, as well as specific reporting procedures for addressing the behavior. Prior research has shown that organizations with long-standing policies against harassment or discrimination positively impact women (Gruber & Smith, 1995), resulting in their perceived inclusion. In addition, although the No FEAR Act requires federal agencies to provide written notification and training on employee rights and remedies to federal antidiscrimination and retaliation laws, the delivery mechanism is left to the discretion of the agencies. Organizations should conduct in-person training (vice the more common web-based version) to emphasize the importance of the matter and to encourage discussion on what constitutes sexual harassment and discrimination. Furthermore, due to variations among federal agencies, executive leaders at each federal law enforcement agency should conduct their own in-depth analysis of all human resource policies and programs to remove artificial barriers that reflect nonrelevant requirements (e.g., physical strength) and incorporate new relevant skill sets such as communication, report writing, judgment and problem solving, de-escalation of violent crime, intelligence analysis, and working collaboratively across communities and cultures into all performance and promotion metrics (Women in Federal Law Enforcement, 2015). Executive leaders who prioritize these three recommendations would accelerate a transformational culture change of gender equality and inclusion and likely see a decrease in wrongful behavior.

On the contrary, almost half (47.7%) the respondents do find their agency as inclusive. The latter may suggest the police culture in federal law enforcement is changing, albeit slowly, to reflect the needs of a modern law enforcement agency and the communities they represent, thus changing the values and behavior of its members. As mentioned from the start, a major piece of executive legislation was signed into policy to promote a model of equal opportunity, diversity, and inclusion. Executive Order 13583 (2011) requires all federal agencies to develop and implement a comprehensive strategy to identify and remove barriers to equal employment opportunity that may exist in human resource policies and practices in regard to recruitment, hiring, promotion, retention, and professional development and training, as well as identify best practices to improve the effectiveness of each agency’s effort in those areas while maintaining consistency with merit system principles and applicable law. Future research should follow its progress to evaluate the true impact of Executive Order 13583.

Finally, although the existence of an inclusive workplace culture did not have a significant difference in reporting sexual harassment or sexual discrimination, the topic still deserves another look. For those respondents in the current study who chose not to file a formal report, retribution, segregation, acceptance as part of the job, and the belief that nothing would be done were common responses when asked why. Fear of retaliation and/or being blacklisted for certain work assignments and future promotional opportunities were the respondents’ primary concern. In addition, respondents did not want the reputation as someone who had to complain to get promoted, nor did they believe the hassle that came in the wake of the report was worth it. Many develop a “thick-skin” over time and learn to adapt. Finally, based on the perceived outcomes from prior reporting of sexual harassment and sexual discrimination, nothing was done to the offender, and in some cases, the offender was “promoted” to another office. Despite ongoing efforts to eliminate and report sexual harassment and sexual discrimination in the federal sector, women are still hesitant with filing a formal report, lending credibility to prior research, and the code of silence (Chaiyavej & Morash, 2009; Yu, 2017). Perhaps the current #MeToo and Times Up movements that have renewed national interest toward the chronic mistreatment of women in the workplace will alter reporting patterns in future studies. Regardless, managers need to pay more attention to their workforce and protect female officers from experiencing segregation, alienation, or any other negative side effect they may receive as a result of reporting wrongful behavior. In addition, managers must find ways to actually make change happen when female officers do report sexual harassment or sexual discrimination. For example, as opposed to “promoting” an offender to another office, real punishment should be prescribed such as reduction in grade level or loss of current or future supervisory status. Separating the alleged offender from the victim may appear to resolve the issue; however, without any punitive consequences, the appearance that nothing will be done will continue to manifest itself. Furthermore, as mentioned previously, if agencies have enforceable organizational policies that prohibit wrongful behaviors such as sexual harassment and sexual discrimination, as well as specific procedures for addressing the behavior, women will feel valued and more supportive to report the incident (Gruber & Smith, 1995), as opposed to maintaining the code of silence.

In addition to these practical implications, this research also presents one possible way to conduct a quasi-experimental design to reduce endogeneity bias in public management research. Coarsened exact matching is a useful methodology to estimate treatment effects by comparing differences in outcome variables between treatment and comparison units. The advantage of coarsened exact matching is to use coarsened pretreatment matching covariates to remove systematic differences between treatment and comparison units. Thus, empirical results from coarsened exact matching are relatively free from endogeneity coming from observed characteristics (i.e., matching covariates). In addition, matching methodology can be useful in large-N-size observational studies when a randomized experiment is infeasible (Stuart, 2010). However, relatively little empirical research has been conducted by matching methodologies in public management research. In particular, coarsened exact matching has been proven to be superior to propensity score matching, which is the most widely used matching methodology in social science, in terms of reducing imbalance level between treatment and comparison units by matching covariates (King & Nielsen, 2015). Thus, coarsened exact matching can be used to conduct a quasi-experimental design in public management research.

The present study is not unique in its research limitations. First, the sample represents those respondents from one federal organization. This limits the sampling frame and may therefore not be representative of all federal organizations; however, this sample is indicative of law enforcement organizations in general due to the male-dominated profession, although patterns of behaviors may manifest differently in a small or rural police department with a smaller proportion of women. For future research projects, this study should be replicated using different agencies and comparing the results. In addition, the inclusion of male officers would be an interesting comparison on the perception of female officers to their male colleagues, as well as respondents’ office locations to determine whether organizational structure (e.g., headquarter vs. field office) or geography (e.g., specific field office) accounts for many of the wrongful behaviors. Second, responses relied on the self-reporting of participants, and some reporting bias may be present to preserve organizational loyalty or accentuate disloyalty. Third, a single survey question was used as the treatment variable and may be insufficient to capture actual inclusion in the workplace; however, the single binary question was necessary for methodological benefits. Fourth, coarsened exact matching cannot address unobserved characteristics, such as individual motivations between treatment and comparison units, and requires a full list of matching covariates, which is nearly impossible in this kind of research. Thus, the empirical findings must be interpreted with caution, and we hope to see whether the empirical results are consistent when using more matching covariates and different organizations. Last, although the empirical findings bolster the impact of a positive organizational culture on employee work conditions and attitudes, we have not reached out to test the effects of inclusive organizational cultures on organizational- or individual-level performance, as well as individual work-related motivations such as commitment, satisfaction, or retention intentions. Further research should focus on the mechanisms of providing an inclusive culture on performance and work-related motivation to support the important roles of providing inclusive organizational cultures not only for equitable and fair treatment but also for beneficial organizational outcomes.

Overall, this research extends public management literature by exploring an underdeveloped but important topic (i.e., organizational culture in federal law enforcement). This research suggests two important lessons. First, it confirms that federal law enforcement agencies require a culture change of inclusivity that does not tolerate sexual harassment and sexual discrimination. Second, this study informs representative bureaucracy, tokenism, and social identity theory by demonstrating how a dominant male culture can create a negative working environment. As explained earlier, the findings have both theoretical and practical implications for all federal law enforcement agencies as demand for relevant knowledge provides the evidence to enact and change policies in response to a gender-neutral organizational culture. Accordingly, its implications have far-reaching consequences for other workgroups promoting gender equity in public sector employment.

Footnotes

Appendix A

Balance Check Before Matching Process.

Multivariate $L_{1}$ distance: 0.909	$L_{1}$	M	Minimum	25%	50%	75%	Maximum
Variable	$L_{1}$	M	Minimum	25%	50%	75%	Maximum
Age	0.100	−0.425	0	−1	−1	0	−2
Education	0.064	−0.109	0	0	0	0	0
Minority status (1 = minority; 0 = nonminority)	0.028	−0.028	0	0	0	0	0
Marital status (1 = married; 0 = otherwise)	0.032	0.032	0	0	0	0	0
Parental status (1 = mother; 0 = no child)	0.056	−0.049	0	0	0	0	−2
Single mother (1 = yes; 0 = no)	0.039	−0.039	0	0	0	0	0
Children number	0.042	−0.042	0	0	0	0	0
Grade level	0.132	−0.280	0	0	0	−1	0
Supervisory status (1 = yes; 0 = no)	0.132	−0.132	0	0	0	−1	0
Relocation number	0.088	−0.305	0	0	−1	−1	0
Work years	0.074	−0.022	0	0	0	1	0

Appendix B

Balance Check After Matching Process.

Multivariate $L_{1}$ distance: 0.155	$L_{1}$	M	Minimum	25%	50%	75%	Maximum
Variable	$L_{1}$	M	Minimum	25%	50%	75%	Maximum
Age	0.081	0.147	−1	0	0	1	2
Education	5.0e-16	0	0	0	0	0	0
Minority status (1 = minority; 0 = nonminority)	5.2e-18	−1.0e-17	0	0	0	0	0
Marital status (1 = married; 0 = otherwise)	4.4e-16	−5.6e-16	0	0	0	0	0
Parental status (1 = mother; 0 = no child)	3.6e-16	−6.7e-16	0	0	0	0	0
Single Mother (1 = yes; 0 = no)	4.7e-16	−5.0e-16	0	0	0	0	0
Children number	5.2e-18	−1.0e-17	0	0	0	0	0
Grade level	4.4e-16	0	0	0	0	0	0
Supervisory status (1 = yes; 0 = no)	6.6e-17	−2.1e-17	0	0	0	0	0
Relocation number	5.0e-16	−4.4e-16	0	0	0	0	0
Work years	0.032	0.016	0	0	−1	0	1

Appendix C

Robustness Check 1: Adjusted Comparison Group.

	Average treatment on the treated
	Negative attitudes	Sexual harassment	Sexual discrimination	Sexual harassment report	Sexual discrimination report
Inclusion
Yes vs. No	−1.313*** (0.177)	−0.787*** (0.185)	−1.158*** (0.197)	−0.011 (0.151)	−0.014 (0.108)
n	100	100	100	30	43

Note. Outcome variables are standardized. Robust standard errors are in parentheses.

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

Appendix D

Robustness Check 2: Instrumental Variable Estimations.

	Negative attitude	Sexual harassment	Sexual discrimination	Sexual harassment report	Sexual discrimination report
Inclusion	−1.629*** (0.176)	−1.174*** (0.179)	−2.053*** (0.172)	−0.237 (0.197)	0.572 (0.539)
Age	0.012 (0.008)	−0.012 (0.008)	0.014 (0.009)	0.011 (0.007)	0.007 (0.007)
Education	0.032 (0.038)	−0.030 (0.037)	0.069^† (0.038)	0.017 (0.030)	0.002 (0.033)
Minority	0.035 (0.068)	0.092 (0.073)	−0.013 (0.068)	−0.0009 (0.053)	0.106 (0.075)
Marital	−0.018 (0.084)	−0.203* (0.087)	0.027 (0.084)	0.015 (0.078)	0.021 (0.091)
Parental	−0.060 (0.118)	−0.248* (0.124)	−0.156 (0.119)	−0.232* (0.104)	−0.105 (0.108)
Single mother	0.035 (0.137)	0.019 (0.144)	−0.143 (0.143)	0.167 (0.115)	0.049 (0.126)
Children	−0.004 (0.046)	0.086^† (0.049)	0.056 (0.045)	0.053 (0.041)	0.068 (0.045)
Grade level	−0.065 (0.047)	0.081^† (0.048)	0.001 (0.048)	0.036 (0.039)	0.068 (0.045)
Supervisory status	−0.065 (0.100)	−0.108 (0.108)	−0.138 (0.100)	0.028 (0.090)	0.010 (0.09)
Relocation number	0.058* (0.023)	0.047^† (0.024)	−0.001 (0.022)	−0.010 (0.016)	0.018 (0.020)
Work years	0.669 (0.352)	0.000 (0.008)	−0.003 (0.009)	−0.005 (0.007)	−0.007 (0.007)
N	1,038	1,037	1,038	325	357
Wald χ²	166.47	119.21	234.46	20.98	9.57
R ²	.195	.160	.170	.130	.060

Note. Robust standard errors are in parentheses.

†

p < .1. *p < .05. **p < .01. ***p < .001.

Acknowledgements

The authors thank the blind reviewers for their constructive criticism and thoughtful feedback.

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.

ORCID iD

Helen H. Yu

Author Biographies

Helen H. Yu is an assistant professor of public administration at the University of Hawai’i at Manoa. She is a retired lieutenant colonel from the U.S. Air Force, and her research interests include the recruitment, retention, and promotion of women in federal law enforcement.

David Lee is an assistant professor of public administration at the University of Hawai’i at Manoa. His research interests include collaboration, intergovernmental management, and human resource management in the public sector.

References

Angrist

J. D.

Pischke

J. S.

(2008). Mostly harmless econometrics: An empiricist’s companion. Princeton, NJ: Princeton University Press.

Ashforth

B. E.

Mael

F. A.

(1989). Social identity theory and organization. The Academy of Management Review, 14, 20-39.

Belknap

Shelley

J. K.

(1992). The new lone ranger: Policewomen on patrol. American Journal of Police, 12(2), 47-76.

Blackwell

Iacus

King

Porro

(2009). CEM: Coarsened exact matching in Stata. The Stata Journal, 9, 524-526.

Brown

J. M.

(1998). Aspects of discriminatory treatment of women police officers serving in England and Wales. British Journal of Criminology, 38, 265-282.

Burch

A. M.

(2016, June). Sheriff’s office personnel, 1993-2013 (Bureau of Justice Statistics Bulletin, NCJ 249757). Washington, DC: U.S. Department of Justice.

Bureau of Labor Statistics. (2017, April). Women in the labor force: A databook (Report 1065). Retrieved from http://www.bls.gov/opub/reports/womens-databook/2016/pdf/home.pdf

Chaiyavej

Morash

(2009). Reasons for policewomen’s assertive and passive reaction to sexual harassment. Police Quarterly, 12, 63-85.

Choi

(2011). Diversity and representation in the U.S. federal government: Analysis of the trends of federal employment. Public Personnel Management, 40, 25-46.

10.

Clark

R. C.

Ochs

H. L.

Frazier

(2013). Representative bureaucracy: The politics of access to policy-making positions in the federal executive service. Public Personnel Management, 42, 75-89.

11.

Collins

S. C.

(2004). Sexual harassment and police discipline: Who’s policing the police? Policing: An International Journal of Police Strategies & Management, 27, 512-538.

12.

Dahlerup

(1988). From a small to a large minority: Women in Scandinavian politics. Scandinavian Political Studies, 11, 275-297.

13.

Dahlerup

(2005, September). The theory of a “critical mass” revisited. Paper presented at the annual meeting of the American Political Science Association, Washington, DC.

14.

Denhardt

R. B.

Denhardt

J. V.

Aristiqueta

M. P.

(2012). Managing human behavior in public and non-profit organizations. Los Angeles, CA: Sage.

15.

Dick

Jankowicz

(2001). A social constructionist account of police culture and its influence on the representation and progression of female officers. Policing, 24, 181-199.

16.

Di Pietro

Di Virgilio

. (2013). The role of organizational culture on informal conflict management. International Journal of Public Administration, 36, 910-921.

17.

Dodge

Valcore

Klinger

D. A.

(2010). Maintaining separate spheres in policing: Women on SWAT teams. Women & Criminal Justice, 20, 218-238.

18.

Equal Employment Opportunity Commission. (2018a). Facts about sexual harassment. Retrieved from https://www.eeoc.gov/eeoc/publications/fs-sex.cfm

19.

Equal Employment Opportunity Commission. (2018b). Discrimination by type. Retrieved from https://www.eeoc.gov/laws/types/index.cfm

20.

Franklin

C. A.

(2007). Male peer support and the police culture: Understanding the resistance and opposition of women in policing. Women & Criminal Justice, 16(3), 1-25.

21.

Gade

D. M.

Wilkins

V. M.

(2013). Where did you serve? Veteran identity, representative bureaucracy, and vocational rehabilitation. Journal of Public Administration Research and Theory, 23, 267-288.

22.

Garnett

J. L.

Marlowe

Pandey

S. K.

(2008). Penetrating the performance predicament: Communication as a mediator or moderator of organizational culture’s impact on public organizational performance. Public Administration Review, 68, 266-281.

23.

Greene

T. H.

Del Carmen

(2002). Female police officers in Texas: Perceptions of colleagues and stress. Policing: An International Journal of Police Strategies &Management, 25, 385-398.

24.

Gruber

J. E.

Smith

M. D.

(1995). Women’s responses to sexual harassment: A multivariate analysis. Basic and Applied Social Psychology, 17, 543-562.

25.

Haarr

R. N.

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

26.

Haarr

R. N.

Morash

(2013). The effect of rank on policewomen coping with discrimination and harassment. Police Quarterly, 16, 395-419.

27.

Hughes

P. J.

(2011). A new sheriff in town: The barriers of structural discrimination facing women leaders. Advancing Women in Leadership, 31(5), 8-13.

28.

Hunt

J. C.

(1990). The logic underlying police sexism. Women & Criminal Justice, 1(2), 3-30.

29.

Iacus

S. M.

King

Porro

(2009). CEM: Software for coarsened exact matching. Journal of Statistical Software, 30(9), 1-27.

30.

Iacus

S. M.

King

Porro

(2012). Causal inference without balance checking: Coarsened exact matching. Political Analysis, 20(1), 1-24.

31.

Ivkovic

S. K.

Haberfeld

Peacock

(2018). Decoding the code of silence. Criminal Justice Policy Review, 29, 172-189.

32.

Jennings

E. T.

(2012). Organizational culture and effects of performance measurement. Public Administration Review, 72(Suppl. 1), 93-94.

33.

Jung

Scott

Davies

Bower

Whalley

McNally

Mannion

(2009). Instruments for exploring organizational culture: A review of the literature. Public Administration Review, 69, 1087-1096.

34.

Kanter

R. M.

(1977). Men and women of the corporation. New York, NY: Basic Books.

35.

Keverline

(2003, August). Executive summary: In the face of challenges, women in federal law enforcement persist and excel. Paper presented at the 4th annual WIFLE® Leadership Training Conference, Washington, DC.

36.

King

Nielsen

(2015). Why propensity scores should not be used for matching. Retrieved from https://gking.harvard.edu/files/gking/files/psnot.pdf

37.

Krimmel

J. T.

Gormley

P. E.

(2003). Tokenism and job satisfaction for policewomen. American Journal of Criminal Justice, 28, 73-88.

38.

Kurtz

D. L.

Linnemann

Williams

L. S.

(2012). Reinventing the matron: The continued importance of gendered images and division of labor in modern policing. Women & Criminal Justice, 22(3), 239-263.

39.

Langston

(2010, June). Women in law enforcement, 1987-2008 (Bureau of Justice Statistics Crime Data Brief, NCJ 230521). Washington, DC: U.S. Department of Justice.

40.

Lonsway

K. A.

(2003). Tearing down the wall: Problems with consistency, validity, and adverse impact of physical agility. Police Quarterly, 6(3), 237-277.

41.

Mahler

(1997). Influences of organizational culture on learning in public agencies. Journal of Public Administration Research and Theory, 7, 519-541.

42.

Martin

S. E.

(1990). On the move: The status of women in policing. Washington, DC: Police Foundation.

43.

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.

44.

Martin

S. E.

Jurik

N. C.

(2006). Doing justice, doing gender: Women in legal and criminal justice occupations. Thousand Oaks, CA: Sage.

45.

Meier

K. J.

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

(Ed.), Research in public administration (Vol. 2, pp. 17-45). Greenwich, CT: JAI Press.

46.

Meier

K. J.

Nicholson-Crotty

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

47.

Miller

F. A.

Katz

J. H.

(1995). Cultural diversity as a developmental process: The path from a monocultural club to inclusive organization. In Pfeiffer

(Ed.), The 1995 annual, Vol. 2: consulting (pp. 267-281). San Diego, CA: Pfeiffer and Co.

48.

Molina

A. D.

(2009). Values in public administration: The role of organizational culture. International Journal of Organization Theory & Behavior, 12, 266-279.

49.

Moon

M. J.

(2000). Organizational commitment revisited in public management: Motivation, organizational culture, sector, and managerial level. Public Performance & Management Review, 24, 177-194.

50.

Mor Barak

M. E

. (2017). Managing diversity: Toward a globally inclusive workplace. Thousand Oaks, CA: Sage.

51.

Morgan

S. L.

Winship

(2007). Counterfactuals and causal inference: Methods and principles for social research. New York, NY: Cambridge University Press.

52.

Mosher

F. C.

(1982). Democracy and the public service. New York, NY: Oxford University Press.

53.

Muscalu

(2014). Organizational culture change in the organization.” Revista Academiel Fortelor Terestre, 19, 392-396.

54.

Notification and Federal Employee Antidiscrimination and Retaliation Act (2002). Pub. L. No. 107-174, 116 Stat. 566.

55.

Office of Personnel Management. (2011). Government-wide diversity and inclusion strategic plan 2011. Retrieved from https://www.opm.gov/policy-data-oversight/diversity-and-inclusion/reports/governmentwidedistrategicplan.pdf

56.

Office of Personnel Management. (2016). Government-wide diversity and inclusion strategic plan 2016. Retrieved from https://www.opm.gov/policy-data-oversight/diversity-and-inclusion/reports/governmentwide-inclusive-diversity-strategic-plan-2016.pdf

57.

O’Reilly

C. A.

Chatman

J. A.

(1996). Culture as a social control: Corporations, cults, and commitment. In Staw

B. M.

Cummings

L. L.

(Eds.), Research in organizational behavior: An annual series of analytical essays and critical reviews (Vol. 18, pp. 157-200). San Diego, CA: Elsevier Science.

58.

Ott

J. S.

(1989). The organizational culture perspective. Pacific Grove, CA: Brooks Cole.

59.

Paoline

E. A.

(2003). Taking stock: Toward a richer understanding of police culture. Journal of Criminal Justice, 31, 199-214.

60.

Rabe-Hemp

C. E.

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

61.

Reaves

B. A.

(2012, June). Federal law enforcement officers, 2008 (Bureau of Justice Statistics Bulletin, NCJ 238250). Washington, DC: U.S. Department of Justice.

62.

Reaves

B. A.

(2015, May). Local police departments, 2013: Personnel, policies, and practices (Bureau of Justice Statistics Bulletin, NCJ 248677). Washington, DC: U.S. Department of Justice.

63.

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.

64.

Rice

M. E.

(2004). Organizational culture, social equity, and diversity: Teaching public administration education in the postmodern era. Journal of Public Affairs Education, 10, 143-154.

65.

Sabharwal

(2014). Is diversity management sufficient? Organizational inclusion to further performance. Public Personnel Management, 43, 197-217.

66.

Schein

E. H.

(2010). Organizational culture and leadership. San Francisco, CA: Jossey-Bass.

67.

Schulz

D. M.

(2009). Women special agents in charge: The first generation. Policing: An International Journal of Police Strategies & Management, 32, 675-693.

68.

Schulze

(2012). The masculine yardstick of physical competence: U.S. police academy fitness tests. Women & Criminal Justice, 22, 89-107.

69.

Seklecki

Paynich

(2007). A national survey of female police officers: An overview of findings. Police Practice & Research, 8, 17-30.

70.

Skolnick

J. H.

(2008). Enduring issues of police culture and demographics. Policing and Society, 18, 35-45.

71.

Smith

A. E.

Monaghan

K. R.

(2013). Some ceilings have more cracks: Representative bureaucracy in federal regulatory agencies. The American Review of Public Administration, 43, 50-71.

72.

Sowa

J. E.

Selden

S. C.

(2003). Administrative discretion and active representation: An expansion of the theory of representative bureaucracy. Public Administration Review, 63, 700-710.

73.

Starling

(2011). Managing the public sector. Boston, MA: Wadsworth.

74.

Stivers

(2001). Gender images in public administration: Legitimacy and the administrative state. Thousand Oaks, CA: Sage.

75.

Stuart

E. A.

(2010). Matching methods for causal inference: A review and a look forward. Statistical Science, 25(1), 1-21.

76.

Taylor

(2014). Organizational culture and the paradox of performance management. Public Performance & Management Review, 38(1), 7-22.

77.

Teodoro

M. P.

Hughes

A. G.

(2012). Socializer or signal: How agency accreditation affects organizational culture. Public Administration Review, 72, 583-591.

78.

Theobald

N. A.

Haider-Markel

D. P.

(2009). Race, bureaucracy, and symbolic representation: Interactions between citizens and police. Journal of Public Administration Research and Theory, 19, 409-426.

79.

Trice

H. M.

Beyer

J. M.

(1993). The culture of work organizations. Englewood Cliffs, NJ: Prentice Hall.

80.

Waddington

P. A.

(1999). Police (canteen) subculture: An appreciation. British Journal of Criminology, 39, 287-309.

81.

Wadman

R. C.

Allison

W. T.

(2004). To protect and serve: A history of police in America. Upper Saddle River, NJ: Pearson.

82.

Wertsch

T. L.

(1998). Walking the thin blue line: Policewoman and tokenism today. Women & Criminal Justice, 9(3), 23-61.

83.

Women in Federal Law Enforcement. (2015, January 29). Testimony in response to the President’s task force on 21st century policing. Arlington, VA: Author.

84.

Wooldridge

J. M.

(2015). Introductory econometrics: A modern approach. Mason, OH: South-Western.

85.

H. H.

(2015). An examination of women in federal law enforcement: An exploratory analysis of the challenges they face in the work environment. Feminist Criminology, 10, 259-278.

86.

H. H.

(2017). Post-executive order 13583: A re-examination of occupational barriers in federal law Enforcement. Women & Criminal Justice, 27, 205-218.