Sexual Harassment of Law Enforcement Officers: Findings From a Nationally Representative Survey

Abstract

Sexual harassment continues to be a consistent destructive feature of American life and workplaces, especially in fields for which women are under-represented, such as law enforcement. We use one of the first nationally representative cross-sectional surveys (n = 2,867) of female and male law enforcement officers (LEOs) to assess the prevalence of workplace sexual harassment victimization. Next, we assess how risk factors are related to this harassment through multivariable modeling. We observed large differences between rates of sexual harassment for female compared to male officers on both our measures of non-physical and physical workplace sexual harassment (sexual assault). Our combined measure of non-physical sexual harassment and sexual assault of female officers (71%) was in the range found in prior research and our 41% rate for male officers is also not trivial and requires attention from law enforcement leaders. We tested two competing hypotheses on whether female officers (and possibly some male officers not meeting certain definitions of masculinity) viewed as the most threatening by virtue of their job role in the male-dominated hierarchy will have the highest probability of being a victim of workplace sexual harassment (power-threat model) or whether those viewed as the most vulnerable officers will have the highest probability of sexual harassment victimization (vulnerable-victim model). We found greater support for the vulnerable-victim model. We discuss the implications of these results for guiding training and prevention strategies to address sexual harassment in the law enforcement workplace.

Keywords

policing workplace environment sexual harassment sexual assault victimization

Introduction

Despite violating the law, sexual harassment continues to be a consistent destructive feature of American life (Kearl et al., 2019), including American workplaces where it is greatly under-reported based on officially reported incidences (Feldblum & Lipnic, 2016), especially in fields for which women are under-represented (Berdahl, 2007; Fitzgerald et al., 1997), such as law enforcement (Lonsway et al., 2013). Sexual harassment is part of a broader group of deleterious workplace behaviors, including general bullying, mobbing, and racial harassment (Burn, 2019; McDonald, 2012; Pina et al., 2009). However, in contrast to these other abusive workplace behaviors, sexual harassment in the workplace has an explicitly sexual element (McDonald, 2012). While definitions of sexual harassment often vary across studies, most include unwanted or unwelcome sexual advances, requests for sexual favors, and other verbal or physical conduct of a sexual nature (such as sexual assaults) which has the purpose or effect of being intimidating, hostile, degrading, humiliating or offensive (McDonald, 2012). In this study, we recognize that sexual harassment can take the form of sexual assault, and thus we use an inclusive measure of both non-physical forms and physical forms as “sexual harassment/assault” (SHA).

Workplace SHA is associated with a full range of negative sequela. First, SHA in the workplace can obstruct careers in concrete ways (McLaughlin et al., 2017), such as higher levels of job stress, work withdrawal, intentions to quit, and lower job satisfaction (Holland et al., 2016). It also places a high financial toll on organizations in terms of absenteeism, lost productivity, and turnover. For example, one analysis of Fortune 500 companies estimates that these companies would have lost $6.7 million annually due to SHA in 1998 (ERC, 2017) or over $14 million in 2017 (Parramore, 2018). Second, SHA in the workplace is associated with poor outcomes outside the work sphere such as lower levels of emotional and physical well-being (Holland et al., 2016), increased cardiovascular reactivity (Schneider et al., 2001), depression and anxiety disorders (Ho et al., 2012), higher rates of abuse of alcohol (Rospenda et al., 2008), higher risk for eating disorders (Harned & Fitzgerald, 2002), and symptoms of post-traumatic stress disorder (Ho et al., 2012; Holcomb & Holcomb, 2008).

While there are still no “official” U.S. governmental or benchmark US national estimates of the extent of SHA in the workplace and estimates vary depending on sampling methodology and question-wording (Fitzgerald & Cortina, 2018), the available survey data suggest it is very common across a variety of work settings (Langer, 2017). Over the past few decades, more than 200 studies have documented the nature and severity of workplace SHA (Fitzgerald & Cortina, 2018; Pina & Gannon, 2012). In the most recent published nationally representative survey, conducted in 2019, with 1,182 women and 1,037 men, ages 18 and up, 81% of women and 43% of men reported experiencing some form of SHA in their lifetime across all settings (Kearl et al., 2019). On the lower estimate side, a 2009 study estimates that 41% of women and 32% of men experiencing SHA at some point in their work lives (Das, 2009). However, other studies have found even higher estimates. In a large nationally representative of U.S. workers, random-digit-dial phone survey conducted in 2003–2004, Rospenda et al. (2009) found that 52% of women and 43% of men in their sample had been sexually harassed (inclusive of sexual assault) in the previous year.

Law Enforcement Workplace and Sexual Harassment/Assault

There are several reasons to believe that law enforcement (LE) is a workplace particularly susceptible to SHA. Researchers have identified features of work environment that are highly associated with workplace SHA (McDonald, 2012), including: An organizational culture that is at least somewhat tolerant of aggression/harassment (Fitzgerald et al., 1997), a workforce that is disproportionately male, an industry characterized by traditionally masculine behaviors (Chamberlain et al., 2008), organizational power is concentrated almost exclusively at higher levels of the organization (Ilies et al., 2003) by gender in the senior staff of the agency, and demographic vulnerability (Hyland & Davis, 2019; e.g., gender, age, sexuality, and ethnicity). These risk factors are common characteristics of law enforcement agencies (LEAs). For example, policing is an industry characterized by traditionally masculine behaviors (Brown & Heidensohn, 2000), women comprise about 12-15% of the sworn LEOs in the United States (Brown & Heidensohn, 2000; Seklecki & Paynich, 2007; Somvadee & Morash, 2008), only about 10% of the police leadership are at the rank of an intermediate supervisor or above are females (Hyland & Davis, 2019), and Blacks are under-represented in law enforcement (Sklansky, 2006; e.g., 11% of full-time sworn officers in local police departments were Black in 2016 compared to the U.S. census of about 14% Black in 2016; Hyland & Davis, 2019). Despite the high-risk nature of the LE environment for SHA, relatively little empirical research has been conducted on SHA prevalence rates or risk factors for SHA in the LE environment.

In this article, we use one of the first nationally representative samples (n = 2,867) of female and male law enforcement officers (LEOs) to assess the prevalence of this problem of SHA victimization and consider if rates of SHA among LEOs rivals levels of SHA found in the literature for other workplace settings. Next, we identify a set of risk factors for SHA and assess how they are related to LEO workplace SHA through multivariable modeling.

Conceptual Framework

We explore two main competing conceptual frameworks to help explain the occurrence of SHA. The first framework is the power-threat perspective of motivation for SHA (Chamberlain et al., 2008). While this model has been applied to female workers, organizational theory extends these types of power models by removing the assumption that power is gender-specific (Pina et al., 2009). This perspective suggests that some men feel threatened by women who deviate from traditionally female roles to inhabit traditionally male roles (De Coster et al., 1999), perpetrating SHA in an effort to control a female worker deviating from gender norms (Walby, 1990). Male workers who do not meet certain male stereotypes of masculinity could also potentially also fall victim to SHA under the power-threat model. SHA becomes a tool to enforce appropriate ways of “doing gender” in the workplace and to penalize gender nonconformity for women (and possibly men too; West & Zimmerman, 1987). This framework suggests that the motivation to harass comes from the instrumental desire to protect male privilege (De Coster et al., 1999). Under this framework, men (or women) who harass are more likely to choose as their victims women (or men) who challenge traditional gender roles and who are perceived as competitors (De Coster et al., 1999). SHA serves to manage interactions between women and men according to accepted sex status norms, and therefore, historically, serves to maintain male dominance occupationally, by intimidating, and discouraging women from advancing in their work (Pina et al., 2009). For example, women who are perceived to be a threat to the existing male hierarchical power structure such as females in supervisory positions, women with higher levels of education, and women with longer job tenures (independent of age) can be targeted for SHA (De Coster et al., 1999).

The second framework, the vulnerable-victim model, explaining SHA suggests that offenders prey on vulnerable victims to express their structural power over less powerful victims (De Coster et al., 1999). From this perspective, the female or male workers who are most attractive as targets are likely to have other characteristics indicative of perceived low structural power (De Coster et al., 1999). Those who are young, unmarried, and/or minority status, and with lower levels of organizational power are highly attractive targets from a vulnerable-victims perspective (De Coster et al., 1999). The vulnerable-victim framework suggests that people of color, workers with lower status within the organization (line workers as opposed to supervisors), those with less experience on the job, less education, more stressed, more adverse childhood events (such as child abuse), experiencing alcohol abuse, or not living with a romantic partner (and who thus may appear to be less protected than their peers cohabiting with a partner) will be targeted more frequently for SHA in the workplace.

Unlike some workplaces, the LE profession by definition, in serving public safety, is interacting with individuals who are not on the police force, as well as fellow officers. Thus, to be comprehensive, we extend these conceptual frameworks to take into consideration that officers may experience SHA by residents in the community. There may be circumstances in which a male civilian would seek to defend their own power when faced with a female officer whom the civilian felt was inappropriately inhabiting a role that should be reserved for men. Likewise, there may be civilians who try to take advantage of a perceived vulnerability in an officer in order to shift the terms of an interaction. While the power imbalance between officers and civilians is such that one would expect less victimization of officers by civilians than by fellow officers, both explanatory models could apply to civilian perpetrators.

Prevalence Rates for Sexual Harassment/Assault in Law Enforcement

There is considerable variation in estimates for the prevalence of SHA among LEOs: from 24% for at least one act of SHA in female LEOs’ careers (Timmins & Hainsworth, 1989) to 100% of female LEOs experiencing SHA (Haarr, 1997). Other studies with female LEOs have fallen somewhere in between from half to about three-quarters experiencing SHA (Bartol et al., 1992; Christopher et al., 1991; Martin, 1994; Nichols, 1995). However, most of these studies explicitly asked respondents whether they have been “sexually harassed,” which investigates the labeling of behaviors rather than their objectively defined experiences (Lonsway et al., 2013). For example, in a 2007 national representative survey of U.S. female LEOs (27% response rate) by Seklecki and Paynich (2007), a total of 531 female LEOs responded to a survey with questions on SHA. When asked about whether they had experienced a range of acts (e.g., being called homosexual, someone trying to have a sexual relationship with the respondent despite their objections, someone making sexually suggestive remarks at or about the respondent), every female LEO in the sample answered they had experienced at least one of these acts (Seklecki & Paynich, 2007). However, only 27% of female LEOs indicated they had ever been “sexually harassed” (Seklecki & Paynich, 2007).

Further, studies from the 1990s and earlier did not measure whether male LEOs experienced SHA. A number of more recent studies have addressed both male and female LEO experience with SHA. Lonsway et al. (2013) explored the characteristics of SHA among LEOs in two study components: (Study 1) A survey of female and male sworn personnel within a single large LEA (n = 679 LEOs completing a survey) and (Study 2) a small national sample of only female LEOs (n = 531; Lonsway et al., 2013). About 84% of the male and female LEOs in Study 1 experienced SHA in the past year (82.6% of the males and 92.5% of the females), while an even higher percentage (93.8%) of the female LEOs indicated in Study 2 that they experienced SHA at least one of the SHA behaviors during the course of their LE career (Lonsway et al., 2013). The other nationally representative study of SHA in law enforcement, discussed above, was also limited to female officers (n = 531; Seklecki & Paynich, 2007).

Studies of LE SHA in other cultures contribute to our understanding of the incidence SHA in the police workplace. In a 2009 national Dutch study (3,001 male LEOs and 1,295 female LEOs; overall response rate of 15%), researchers found that 64% of female LEOs and 48% of male LEOs experienced one or more forms of SHA limiting recall to the past 24-month reference period (De Haas et al., 2009). In a 2018 study of a large provincial Japanese department (1,592 male and 273 female; overall response rate of 93%), about 35% of the female LEOs indicated that they had experienced some type of SHA at their workplaces at least once during the course of their career in the police, compared to 1.8% of the male LEOs (Kobayashi, 2018). The researchers also concluded that the SHA victimization rate of female LEOs in Japan was much lower than that for female workers in ordinary Japanese workplaces (50-70%; Kobayashi, 2018). Finally, in a small 2018 national study with Icelandic LEOs (n = 347; 50% response rate), researchers found that 31% of female and 4% of male LEOs had experienced a behavior that they labeled as “sexual harassment” (SteinÞórsdóttir & Pétursdóttir, 2018).

Risk Factors for Sexual Harassment in the Workplace

To develop programs and policies to prevent SHA, it is important to examine risk factors for SHA. Outside demographic factors, there is little exploration of risk factors for SHA in the police workplace literature. Therefore, we review the broader literature on risk factors for SHA across a variety of work sectors to build our selection of covariates for our multivariable models.

Studies consistently show that women are more likely to encounter SHA than men across numerous fields (Das, 2009; Rospenda et al., 2008), including in LE (De Haas et al., 2009; Kobayashi, 2018; Lonsway et al., 2013; SteinÞórsdóttir & Pétursdóttir, 2018). Another personal risk factor is ethnicity. Individuals of non-White ethnicity are more likely to experience SHA than Whites (Chamberlain et al., 2008; Gettman & Gelfand, 2007). For example, perpetrators of SHA against African American women and Asian American women may use racist stereotypes to attempt to justify their actions (Buchanan & Ormerod, 2002; Ho et al., 2018). Researchers have explored other covariates related to having supervisory responsibilities, level of education, and length of job tenure (De Coster et al., 1999). Some research shows those with lower status within the organization (line workers) and less power tend to be most vulnerable to SHA, such as women, those with less tenure in the organization, temporary workers, and people of color (Gettman & Gelfand, 2007; Uggen & Blackstone, 2004). Other covariates relate to other potential vulnerabilities of victims. Researchers have found an association between SHA and problem drinking (Gradus et al., 2008; Richman et al., 1996; Richman et al., 1999; Rospenda et al., 2008). These data provide some support for a self-medication hypothesis where alcohol is misused as way to cope with the SHA or possibly as a dangerous routine that puts individuals at greater risk for SHA.

Other vulnerabilities explored in the literature are stress and adverse childhood experiences (ACEs). In a study of SHA among employees of an urban American university, the research team found a strong relationship between SHA and measure of stress for both genders (Richman et al., 1999). It is reported that the nurses who experienced SHA suffered from various forms of stress more than nurses who were not harassed (Houle et al., 2011). Likewise, exposure to ACEs has been found to be related to SHA victimization in the workplace (Gaska & Kimerling, 2018). Higher ACE scores and measures of child abuse are associated with a greater likelihood of being the victim of SHA and other forms of sexual victimization in adulthood (Jones, 2019).

The Present Study

As part of a broader research program on LEO health, we collected survey data from a nationally representative sample of LEOs and queried them about their SHA victimization experiences in their policing careers and the hypothesized risk factors. The present study design provides data to answer our first research question: What is the extent of workplace SHA experienced by male and female officers in the U.S.? Our first hypothesis was that we would find higher rates of SHA for female LEOs compared to male LEOs and that the rates for workplace SHA would be broadly consistent with rates detected in prior research of between 50% and 94% (Bartol et al., 1992; Christopher et al., 1991; Lonsway et al., 2013; Martin, 1994; Nichols, 1995).

Our second research question relates to competing hypotheses regarding risk factors for workplace SHA against LEOs. That is, whether those LEOs viewed as the most threatening by virtue of their job role in the male-dominated hierarchy would have the highest probability of having been a victim of workplace SHA or whether those viewed as the most vulnerable LEOs would have the highest probability of having been a victim of workplace SHA. Our second hypothesis is that female LEOs (and possibly some male LEOs not meeting certain definitions of masculinity) in supervisory positions, with higher levels of education, and with longer job tenures would be more likely to be victims of workplace SHA. Our third hypothesis is that those LEOs with lower status within the LEA (line workers as opposed to supervisors), less experience on the job, less education, more stressed, more ACEs (such as child abuse), experiencing alcohol abuse, not living with a romantic partner, female, and non-White will be targeted more frequently for workplace SHA.

Method

Participants and Procedures

For the LEO Safety and Wellness (OSAW) study, we used a two-stage probabilistic sampling approach. We started with randomly selecting and surveying publicly funded civilian LEAs representing municipal, county, state police, and highway patrol or Bureau of Indian Affairs LEAs (no federal LEAs were included in the study) that employ at least one full-time sworn LEO with general arrest powers. The agency survey was fielded over the period of August 2017 to February 2019 (Stage 1). The sampling frame for the OSAW Initiative was the 2017 National Directory of Law Enforcement Administrators (NDLEA), Correctional Institutions, and Related Agencies (53rd ed.; National Public Safety Information Bureau, 2017). As described by Mumford et al. (2020), we achieved a response rate of nearly 40%, with 2,867 LEOs completing the survey out of 8,060 invited eligible LEOs, a response rate that exceeds that of other established panels in other fields (Fontes et al., 2015). Statistical weights to assure national representativeness were applied to all results. Our power analyses revealed that we had 1.65% precision at the 95% confidence interval level to detect the most conservative rate of 50% non-physical sexual harassment or sexual assault prevalence over respondents’ policing careers (allowing for a finite population correction factor). Even after allowing for missing data, our precision level is still no greater than 1.8%. The margin of error is smaller for proportionate estimates less than or greater than 50% (e.g., for a 10% prevalence, we would have a narrower ± 1% margin of error).

The research team made a large number of attempts (often up to 12 attempts) via U.S. Postal mail, email, and phone. We provided the respondents with a user-friendly survey (about 30 minutes long) through a secure online survey. The project was approved by the authors’ Institutional Review Board (IRB), who holds a multiple project assurance with Health and Human Services (HHS) and conforms to HHS standards for IRB review. Our team collected these survey data from September 2017 to November 2018.

Measures

Dependent variables.

Sexual harassment and sexual assault.

As noted earlier, we recognize that sexual harassment can take the form of sexual assault, and we therefore measure both sexual harassment in its non-physical forms and physical forms as sexual assault. First, we asked about non-physical sexual harassment. LEOs were asked whether they experienced unwanted sexual behavior (excluding physical contact) “in their professional career as a police officer.” Examples provided in the question included “unwelcome sexual advances, requests for sexual favors, direct or indirect threats or bribes for sexual activity, sexual innuendos and comments, sexually suggestive jokes, pervasive displays of materials with sexually illicit or graphic content.”

Next, we asked about physical sexual harassment in the form of sexual assaults. LEOs were asked whether they experienced any type of sexual contact or behavior that occurred without their explicit consent during their professional career as a police officer, with examples provided in the question text including “sexual activities as forced sexual intercourse, forcible sodomy, child molestation, incest, fondling, and attempted rape,” as referenced in other research (McDonald, 2012). Response categories for both measures were “never,” “once or twice” and “more than twice.” We created separate binary variables for non-physical sexual harassment and sexual assault by combining “once or twice” and “more than twice” to code responses of “more than once” as 1.

Covariates.

LEO rank was coded as a three-level categorical variable (line officer as the reference, supervisor, and commander). The number of year’s sworn LE experience (excluding military LE experience) was coded as a five-level categorical variable (0–5 years, 6–10 years, 11–15 years, 16–20 years, and 21 or more years). An indicator was included for whether the respondent reported risky drinking or not (0 = No and 1 = Yes). An indicator of living with a romantic partner was coded as 1 if the LEO reported a co-resident partner (married/remarried/lived together with a partner) and zero otherwise (widowed/divorced/separated/never married/in a relationship, but not living together). We included 11 items (e.g., living with anyone who had a drinking problem, whether parents separated or divorced, whether they were touched sexually, made to touch or forced to have sex with an adult or anyone who was at least 5 years older than them; Ford et al., 2014) to assess LEO exposure to ACEs before the age of 18. Factor analysis was carried out to create three dichotomous sub-scale variables: Family instability, childhood physical/emotional abuse, and childhood sexual abuse. Cronbach’s alpha scores for all three factors were above .8. LEOs reported feelings of stress over the past month using the Perceived Stress Scale (PSS; Cohen et al., 1983), with higher scores (range 0 to 4) representing more stress. The Cronbach’s alpha for the PSS was .8. Sociodemographic measures included biological sex (female coded as zero and male coded as 1), race/ethnicity (White coded 1 and all other races coded as 0), and educational attainment (bachelor degree or more coded 1; 0 otherwise).

Analytic Plan

Analyses were conducted in R (version 3.6.2). Of the 2,867 responding officers, 420 (14.6%) did not answer at least one of the two outcome questions: 407 (14.2%) were missing data on non-physical sexual harassment and 414 (14.4%) were missing data on physical SHA. Also, 647 (22.6%) observations were missing for the covariates (the total missing percentage was 23.4, when we include all the covariates and two outcomes together).To mitigate the bias that may be introduced by only considering complete cases, following best practices in the field of statistics (van Buuren, 2018), we conducted analyses with multiple imputed datasets (100 iterations and a combined 25 datasets, each with analytic sample of 2,867) and present pooled results (van Buuren, 2018). To answer the first research question, we first examined the descriptive statistics for sexual harassment and sexual assault, respectively. Then, to facilitate comparisons with other studies reporting a single total measure of physical and non-physical harassment/assault, we estimated the rate of total harassment/assault for the full sample and by gender. For the second research question, we fit four multivariable models (logistic regression for binary outcomes) separately to test relationships between the non-physical sexual harassment measures (female and male models) and sexual assault measures (female and male models) and the independent variables.

Results

Descriptive Results

Table 1 presents the weighted distribution of all study variables. Given that LEAs demographically are composed of mostly White men (Hyland & Davis, 2019), our sample also was 87% male and 78% White. However, we did have some diversity in our sample with 13% women and mostly African-American and Hispanic LEOs making up the other 22% of the sample who were not White. Our sample of LEOs was on average 41 years old. A total of 65.6% of the LEOs were line LEOs, 17.7% were supervisors and the rest were commanders. Close to half of the sample (47%) reported having at least a bachelor’s degree. About one in five (19%) of the LEOs had been in LE (whether in their current department or another department) for less than 5 years, 14% for 6–10 years, 16% for 11–15 years, 19% for 16–20 and 31% had worked in LE for more than 20 years. One in five LEOs reported living alone and one in three (36%) reported having alcohol abuse issues. A little more than half (52%) of the sampled LEOs noted that they had experienced family instability/dysfunction, 27% said they had experienced physical/emotional abuse as a child and 7% noted they experienced sexual abuse as a child. The average perceived stress mean score was 1.2 (on a scale of 0 to 4) representing fairly low general stress.

Table 1.

Sample Description: Proportion/Mean (SD).

	Female	Male	Total
	(N = 671)	(N = 2,174)	(N = 2,867)
Dependent variables
Non-physical sexual harassment	70.3%	40.0%	44.2%
Physical sexual harassment/assault	8.3%	2.2%	3.0%
Covariates
Age	40.9 (9.7)	41.5 (9.6)	41.4 (9.6)
Gender
Male			87.0%
Female			13.0%
Race*
White	73.0%	78.8%	78.0%
Non-White	27.0%	21.2%	22.0%
Education: Associate degree or higher*	56.8%	45.8%	47.2%
Rank
Line officer	69.5%	65.2%	65.6%
Supervisor	18.4%	17.6%	17.7%
Commander	12.1%	17.2%	16.7%
Years sworn
0–5	22.9%	18.5%	19.0%
6–10	13.1%	14.3%	14.1%
11–15	17.8%	15.9%	16.1%
16–20	18.1%	19.7%	19.5%
21+	28.0%	31.5%	31.4%
Alcohol abuse*	46.9%	35.1%	36.6%
Married/living with partner*	64.0%	83.1%	80.7%
Adverse childhood experiences
Family instability	56.0%	51.6%	52.3%
Childhood physical/emotional abuse*	37.2%	25.9%	27.5%
Childhood sexual abuse*	15.9%	5.3%	6.7%
Perceived Stress Scale*	1.4 (.6)	1.2 (.6)	1.2 (.6)

Note. *Indicates there is a statistically significant difference (<.05) between genders.

We observed that 44% of the LEOs said that they had experienced some form of non-physical sexual harassment at least once during their tenure as an LEO, with a rate of 62% for female LEOs and 35% for male LEOs (X² = 122.4 [df = 1], p <.001). Sexual assault was rarer, with 3% indicating that they had experienced this at least once in their career, with a rate of 7.2% for female LEOs and 1.9% for male LEOs (X² = 40.9 [df = 1], p <.001). We hypothesized that males and females will have a different experience of SHA at work. We carried out additional bivariate analyses to test this hypothesis. We found that male LEOs were at a much lower risk (AOR = .28, p < .001) of experiencing non-physical sexual harassment than their female counterparts. Similarly, they were also at a much lower risk (AOR = .24, p < .001) of experiencing sexual assault than the female LEOs.

For comparability with studies that used a single measure, the combined total prevalence of SHA was 44.5% (up from 44% for non-physical sexual harassment). Among female officers, the rate was 71% and for male officers, the rate was 40.5%. This gender difference was statistically significant (X² = 141.1 [df = 1], p <.001). If we calculate a simple bivariate logistic regression (not shown in the tables), we found that male LEOs were at a much lower risk (AOR = .81, p < .001) of experiencing our combined measure of SHA than their female counterparts.

Multivariable Results

Based on the differences we observed for female and male SHA rates, we used stratified logistic regression models for female and male LEOs separately for non-physical sexual harassment and sexual assault, totaling four multivariable models. In Table 2, among female LEOs, we found sworn LE years to be positively associated with non-physical sexual harassment. Those in LE for 11–15 years had about 350% higher odds (AOR = 3.5, p = .007) of experiencing non-physical sexual harassment, compared to LEOs who were in LE for less than 5 years. This increased to 424% (AOR = 4.2, p = .001) when we look at LEOs with 16–20 years of experience, as compared to those relatively newly sworn (less than 5 years). The odds of experiencing non-physical sexual harassment reduced to 3.3 (p = .009) for LEOs with 20 years of experience, compared to LEOs with 5 or less years of experience. We also found that higher perceived stress (AOR = 1.7, p = .026) was associated with a higher risk of non-physical sexual harassment.

Table 2.

Logistic Regression Model of Law Enforcement Officer Characteristics and Non-physical Sexual Harassment.

Female Officers			Male Officers
	AOR^a	SE^b	AOR^a	SE^b
Race: White	1.40	.34	1.15	.16
Education: Associate degree or higher	1.49	.27	.88	.12
Rank
Supervisor	.92	.60	1.06	.18
Commander	1.19	.56	.94	.18
Years sworn
6–10	1.70	.38	1.55*	.22
11–15	3.50**	.46	1.67*	.23
16–20	4.24**	.43	1.96**	.22
21+	3.29**	.45	1.77**	.21
Alcohol abuse	.66	.31	.90	1.14
Married/Living with partner	1.00	.30	1.19	.17
Adverse childhood experiences
Family instability	1.27	.32	1.05	.14
Childhood physical/emotional abuse	1.12	.37	1.18	.16
Childhood sexual abuse	.82	.41	1.26	.25
Perceived Stress Scale	1.66*	.22	1.28*	.11

Notes. Reference groups: Non-White, Less than bachelor’s degree, Line officer, Sworn for 5 years or less, No alcohol abuse and Not married/Living with partner.

^aAdjusted odds ratio. ^bSE are of βs.

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

Table 2 presents similar results for male LEOs, in that increasing years as a sworn officer was significantly associated with non-physical sexual harassment, but the odds were much lower than their female counterparts. LEOs who had been in LE for 11–15 years, 16–20 years and more than 20 years had about 167% (AOR = 1.7, p = .028), 196% (AOR = 2.0, p = .003) and 177% (AOR = 1.8, p = .007) higher odds, respectively, of experiencing non-physical sexual harassment compared to LEOs who were in LE for less than 5 years. We further found that male LEOs with 6–10 years of experience had a statistically higher odds (AOR = 1.5, p = .048) of experiencing non-physical sexual harassment compared to male LEOs who had 5 or less years of LE experience. Perceived stress (AOR = 1.3, p = .021) was found to be positively associated with non-physical sexual harassment.

In Table 3, for sexual assault of female LEOs, we did not find any significant associations for any of the covariates at the 5% critical level of significance. We found marginally significant association of childhood sexual abuse and experiencing sexual assault as an officer. Female LEOs who experienced sexual abuse as a child had a 237% higher odds (AOR = 2.4, p = .06) of experiencing sexual assault than female LEOs who did not experience sexual abuse as a child. Notably, childhood sexual abuse was a statistically significant risk factor at the .05 significant level (p = .047) controlling for age, race/ethnicity, years sworn, rank, and romantic partnership cohabitation (i.e., excluding the PSS and alcohol abuse measures from the model).

In Table 3, for sexual assault for male LEOs, we found that White LEOs had a 55% lower odds (AOR = .45, p = .04) of sexual assault than non-White male LEOs. We also found that male LEOs with higher perceived stressed score (AOR = 2, p = .004) had a larger risk of sexual assault than male LEOs with lower perceived stressed scores.

Table 3.

Logistic Regression Model of Law Enforcement Officer Characteristics and Sexual Assault.

Female Officers			Male Officers
	AOR^a	SE^b	AOR^a	SE^b
Race: White	.68	.42	.45*	.39*
Education: Associate degree or higher	1.08	.40	1.05	.40
Rank
Supervisor	1.85	.53	.38	.59
Commander	.70	.81	.71	.49
Years sworn
6–10	2.36	.71	.85	.81
11–15	1.78	.74	.59	.72
16–20	1.66	.78	1.17	.70
21+	3.29	.73	1.17	.63
Alcohol abuse	.87	.46	.47	.49
Married/living with partner	1.25	.38	.80	.50
Adverse childhood experiences
Family instability	1.42	.40	1.00	.43
Childhood physical/emotional abuse	1.16	.44	1.19	.42
Childhood sexual abuse	2.37	.46	.80	.75
Perceived Stress Scale	1.38	.29	2.00**	.24**

Notes. Reference groups: Non-White, Less than bachelor’s degree, Line officer, Sworn for 5 years or less, No alcohol abuse and Not married/Living with partner.

^aAdjusted odds ratio. ^bSE are of βs.

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

Discussion

The current study is the first nationally representative survey on sexual harassment and sexual assault in U.S. law enforcement inclusive of both female and male officers. We observed large statistically significant differences between rates of SHA for female LEOs compared to male LEOs. For both our non-physical and our physical workplace sexual harassment measures female LEOs experienced substantially higher rates than male LEOs, consistent with our first hypothesis. Also consistent with our expectations, the combined measure of non-physical sexual harassment and sexual assault of female LEOs (71%) was in the range found in prior research (50% to 94%; Bartol et al., 1992; Christopher et al., 1991; Lonsway et al., 2013; Martin, 1994; Nichols, 1995). The total SHA prevalence rate for male LEOs (41%) is also not trivial and requires attention from LE leaders. For the general population, the National Intimate Partner and Sexual Violence Survey (NISVS), reports lifetime sexual assault rates of 44% for female LEOs and 25% for male LEOs (Smith et al., 2016). While not as high as lifetime rates of sexual assault found by NISVS in the general population, OSAW sexual assault rates (7.2% for female LEOs and 1.9% for male LEOs) only cover officers’ careers and not their entire lives. These reported rates are still suggestive of a major problem that officers may need services, training, and counseling to address.

Next, we move to our two competing hypotheses on whether those female LEOs (and possibly some male LEOs not meeting certain definitions of masculinity) viewed as the most threatening by virtue of their job role in the male-dominated hierarchy will have the highest probability of being a victim of workplace SHA (Hypothesis 2, power-threat model) or whether those viewed as the most vulnerable LEOs will have the highest probability of being a victim of workplace SHA (Hypothesis 3, vulnerable-victim model). We found only some support for our second hypothesis. That is, female LEOs with longer job tenures are at an increased risk of being a victim of workplace SHA, as evidenced by escalating risk for non-physical sexual harassment for those female LEOs with more years of LE experience compared to less LE experience. It is possible that those females LEOs with the accumulated experience of more years on the job are a threat to the power structure of this male-dominated profession and are thus sexually harassed as an attempt to retain power in the workplace. However, we did not detect a similar statistically significant finding for sexual assault for female LEOs. Also, the other variables of our power-threat model of female LEOs being in supervisory positions and having higher levels of education were also not statistically significant in our non-physical sexual harassment and sexual assault models. In addition, more years of experience were also associated with escalated risk of non-physical sexual harassment for male LEOs. Rather, than reflecting a threat to the male hierarchy, perhaps female and male LEOs with more experience in LE may simply be at greater risk of SHA given the longer period of exposure on the job.

Next, we found some support for our third hypothesis. Based on our available measures, an argument could be made that those more vulnerable in the LE work sector are more likely to be victims of SHA workplace. We observed large bivariate differences in female LEOs being much more likely to be victims of non-physical sexual harassment and sexual assault compared to male LEOs. While not shown, this was also the case in a multivariable model using the total sample, that is, including both female and male LEOs. Moreover, non-White male LEOs were more likely to be victims of workplace sexual assault than White male LEOs. We did not find a statistically significant difference for White female LEOs compared to non-White female LEOs. Among female and male LEOs, we found that higher perceived stress was associated with a higher risk of non-physical sexual harassment. For male LEOs, we also found that higher perceived stress was associated with a higher risk of sexual assault. These data suggest that officers suffering from higher life stress may be at an increased risk of victimization for SHA. Prevention education programs for highly stressed officers could alert them to this increased vulnerability to SHA. Alternatively, the experience of SHA in their policing careers could be generating the excess stress reported by these male officers, as the cross-sectional data cannot determine directionality.

Next, we found that female LEOs who experienced sexual abuse as a child had a 237% higher odds (p-value = .06) of experiencing sexual assault than female LEOs who did not experience sexual abuse as a child. Given the smaller size of the female sample in our study, this finding may be more of a statistical power issue than a substantive issue (as noted in the results, in sensitivity analyses that excluded stress and risking drinking from the model, the association was statistically significant at the .05 level). However, we found no statistically significant covariates for our separate female and male models for our measures of line workers (patrol LEOs as opposed to supervisors), education, living with a romantic partner, most of our measures of adverse childhood events (with the one exception noted above), alcohol abuse, and those LEOs with less experience on the job actually had a lower likelihood of being a victim of non-physical sexual harassment. Therefore, only partial support is provided for Hypothesis 3.

Limitations

First, the current study includes a reliance on self-reported survey data and potential recall bias and/or social desirability biases. However, the collection of survey data with employees is a common accepted practice for collecting reliable data on SHA in the workplace (McDonald, 2012). Also, to mitigate any concerns LEOs might have about the confidentiality of the survey results, LEOs were assured (through a legal Privacy Certificate) that their agency supervisors/leadership would not have data about whether they completed a survey nor any information on their specific survey responses. Second, the interpretation of these results must take into account the cross-sectional nature of the data. Our data show associations but not casual links. For example, where we found an association between stress and SHA it is not possible to disentangle which came first and caused the other. Prospective longitudinal designs can begin to sort out these time order issues. Next, OSAW is focused on a broad set of health-related issues and we were limited to two global measures of non-physical sexual harassment and sexual assault. A more comprehensive measure of SHA, such as the Sexual Experiences Questionnaire (SEQ; Fitzgerald et al., 1988), would yield greater detail on the nature of the SHA experienced. Nevertheless, given that the rate of SHA we detected in our study was consistent with the prior literature, our global measures worked well for a prevalence study and associated risk factors.

Another limitation of our study was around some of our measurement choices with SHA. We measured SHA as a dichotomous measure to reduce the cognitive burden on respondents as opposed to measuring SHA as a continuous variable. This measurement choice is likely associated with a loss of precision in our SHA measure, for it is not the same to have been a victim of SHA once or twice compared to daily or weekly abuse. We also did not include measures of the source of the SHA in terms of the organizational standing and other characteristics of the perpetrator(s). This is a limitation of the study that we did not collect data on the characteristics of the perpetrators. Future research on officer SHA should include measures of the characteristics and other features of the perpetrators. Next, our measure of SHA did not make a distinction between SHA from other officers than from civilians and this could have altered the results for the statistical significance of some of our covariates. As an illustration, higher-ranked officers typically have less interaction with the general public (e.g., making a traffic stop) than line officers of lower rank. Therefore, even if higher rank officers were harassed more because they challenge traditional roles, they might be harassed less by civilians due to less frequent contact with the general public. Future research should make this distinction between SHA from other officers and from civilians in a longer survey instrument.

Implications

Given the high rates of SHA among female LEOs, and the elevated rates of SHA among male LEOs—especially non-White male LEOs—LEAs need to consider implementing strategies to reduce SHA victimization of their officers. The strategies would highlight the higher rates of SHA among LEOs who are female, those with more LE experience, those perceived to be more stressed, non-White males, and possibly those LEOs with a history of childhood sexual abuse. Because the current study does not distinguish whether the perpetrators are fellow officers or civilians, officer education should address both potential scenarios.

Research suggests that workplace SHA is less prevalent in organizations with proactive policies on SHA and rigorous enforcement of those policies (Gruber, 1998), with official reporting of SHA also higher in these types of organizational settings (Fitzgerald et al., 1997). Also, there are a number of evidence-based training strategies LEAs can use that have reduced the occurrence of SHA across a variety of workplaces (Antecol & Cobb-Clark, 2003; Reese & Lindenberg, 2003; Roehling & Huang, 2018). In their systematic review, McDonald et al. (2015) identified four main elements for effective SHA prevention training. First, training should be developed from data on the extent and risk factors for SHA in the LEA, using human resource records on SHA and anonymous LEO surveys on these topics (McDonald et al., 2015). Many LEAs already conduct officer surveys and could add items from the SEQ scale or our global measures to those surveys to measure workplace SHA. Having the more detailed SEQ measure might be helpful to get at the forms of SHA that can be taught in the trainings, but the global measures are shorter and easier to add to an existing survey and have done a good job in this study capturing the prevalence of SHA. Second, SHA training should raise awareness and clarify misconceptions about what constitutes SHA while highlighting and reinforcing acceptable behavioral norms (McDonald et al., 2015). Third, SHA training should be implemented for all sworn and non-sworn personnel, and integrated into employee orientation (Howald et al., 2018), with additional training for LEA managers to include conflict management, the managing of emotions, and facilitation techniques as a way to prepare managers for handling SHA cases appropriately (McDonald et al., 2015). Fourth, training for LEOs should challenge gendered organizational cultures which tend to be associated with SHA in the workplace (McDonald et al., 2015). While agencies might not be able to change problematic aspects of their local culture in the short-term, it is useful to raise awareness about the potential effects of this kind of culture on increasing SHA in LEAs.

Conclusion

All in all, the prevalence and nature of officer workplace SHA share many similarities with workplace SHA across other professions (Kearl et al., 2019; Rospenda et al., 2009). LEOs need to recognize that there is nothing in particular about being a police officer (with a firearm, body armor, radio, etc.) that shields them from being victims of SHA. Also, there needs to be a recognition that SHA not only has serious consequences for the victim’s mental, physical health and productivity but also affects the public’s confidence in the police (Brown et al., 2018). LEAs also need to make a clear statement about their intent to prevent SHA (Brown et al., 2018). In each agency, there is a need for a unified stance from leadership and through the ranks that SHA is not just a result of a few “bad apples” in the LEA but is a broader agency problem that requires attention to the organizational climate that might foster SHA. Addressing SHA will bring LEAs closer to an organizational culture that is characterized by inclusion and equality and puts an end to SHA.

Footnotes

Acknowledgments

We would like to express our sincere gratitude to the agencies and dedicated officers who participated in this research. Further thanks are due to the members of our Expert Panel who collaborated in the design and review of the survey instrument: Timothy Baysinger, Sarah Creighton, Stephen James, Tara Kelley-Baker, Sandra Ramey, Elizabeth Stanley, Darrel Stephens, Bryan Vila, and John Violanti. We would like to thank PERF Staff Nathan Ballard and Adam Kemerer and NORC Staff Meghan O’Leary, Amanda O’Keeffe, Jackie Sheridan, Katie Archambeau, Julie Banks, and Steven Pedlow.

Declaration of Conflicting Interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Points of views in this article are those of the authors and do not necessarily represent the official position or policies of the U.S. Department of Justice or any other organization.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by the National Institute of Justice (Grant # 2016-IJ-CX-0021) and we are grateful for this funding.

ORCID iD

Bruce G. Taylor

Author Biographies

Bruce G. Taylor, PhD, is a senior fellow at NORC at the University of Chicago. He is a Criminologist, has conducted over 50 policing studies, and is Co-PI on the OSAW study.

Poulami Maitra, PhD, is a statistician III, in Statistics and Data Science at NORC at the University of Chicago and is an Expert in regression and multiple imputation modeling.

Elizabeth Mumford, PhD, is a principal research scientist at NORC at the University of Chicago, a Social Epidemiologist, and Co-PI on the OSAW study.

Weiwei Liu, PhD, is a senior research scientist at NORC at the University of Chicago and is an expert in Latent Class Analysis (LCA).

References

Antecol

, & Cobb-Clark

(2003). Does sexual harassment training change attitudes? A view from the federal level. Social Science Quarterly , 84(4), 826–842.

Bartol

C. R.

, Bergen

G. T.

, Volckens

J. S.

, & Knoras

K. M.

(1992). Women in small-town policing: Job performance and stress. Criminal Justice and Behavior , 19(3), 240–259.

Berdahl

J. L.

(2007). The sexual harassment of uppity women. Journal of Applied Psychology , 92(2), 425.

Brown

, Gouseti

, & Fife-Schaw

(2018). Sexual harassment experienced by police staff serving in England, Wales and Scotland: A descriptive exploration of incidence, antecedents and harm. The Police Journal , 91(4), 356–374.

Brown

, & Heidensohn

(2000). Gender and policing: Comparative perspectives . Palgrave Macmillan.

Buchanan

N. T.

, & Ormerod

A. J.

(2002). Racialized sexual harassment in the lives of African American women. Women & Therapy , 25(3–4), 107–124.

Burn

S. M.

(2019). The Psychology of Sexual Harassment. Teaching of Psychology , 46(1), 96–103. https://doi.org/10.1177/0098628318816183

Buuren

S. van

. (2018). Flexible imputation of missing data (2nd ed.). Chapman & Hall/CRC.

Chamberlain

L. J.

, Crowley

, Tope

, & Hodson

(2008). Sexual harassment in organizational context. Work and Occupations , 35(3), 262–295.

10.

Christopher

, Arquellas

, Anderson

, Barnes

, Estrada

, Kantor

, & Tranquada

R. E.

(1991). Report of the Independent Commission on the Los Angeles Police Department . Independent Commission on the Los Angeles Police Department. http://www.parc.info/client_files/Special%20Reports/1%20-%20Chistopher%20Commision.pdf

11.

Cohen

, Kamarck

, & Mermelstein

(1983). A global measure of perceived stress. Journal of Health and Social Behavior , 24(4), 385–396.

12.

Das

(2009). Sexual harassment at work in the United States. Archives of Sexual Behavior , 38(6), 909–921.

13.

De Coster

, Estes

S. B.

, & Mueller

C. W.

(1999). Routine activities and sexual harassment in the workplace. Work and Occupations , 26(1), 21–49.

14.

De Haas

, Timmerman

, & Höing

(2009). Sexual harassment and health among male and female police officers. Journal of Occupational Health Psychology , 14(4), 390.

15.

ERC. (2017). The cost of sexual harassment in the workplace . https://www.yourerc.com/blog/post/the-cost-of-sexual-harassment-in-the-workplace.aspx

16.

Feldblum

C. R.

, & Lipnic

V. A.

(2016). Select task force on the study of harassment in the workplace . US Equal Employment Opportunity Commission. https://www.eeoc.gov/eeoc/task_force/harassment/upload/report.pdf

17.

Fitzgerald

L. F.

, & Cortina

L. M.

(2018). Sexual harassment in work organizations: A view from the 21st century. In Travis

J. W. W. C. B

, Rutherford

, Williams

W. S.

, Cook

S. L.

& Wyche

K. F.

(Eds.), APA handbook of the psychology of women: Perspectives on women’s private and public lives , Vol. 2 (pp. 215–234). American Psychological Association.

18.

Fitzgerald

L. F.

, Drasgow

, Hulin

C. L.

, Gelfand

M. J.

, & Magley

V. J.

(1997). Antecedents and consequences of sexual harassment in organizations: A test of an integrated model. Journal of Applied Psychology , 82, 578–589.

19.

Fitzgerald

, Shullman

, Bailey

, Richards

, Swecker

, Gold

, Ormerod

, & Weitzman

(1988). The dimensions and extent of sexual harassment in higher education and the workplace. Journal of Vocational Behavior , 32(2), 152–175.

20.

Fontes

, Chew

, & Layrakas

(2015). Boosting probability-based web survey response rates via nonresponse follow-up [Paper presentation]. The Conference of the American Association of Public Opinion Research, Hollywood, FL.

21.

Ford

D. C.

, Merrick

M. T.

, Parks

S. E.

, Breiding

M. J.

, Gilbert

L. K.

, Edwards

V. J.

, Dhingra

S. S.

, Barile

J. P.

, & Thompson

W. W.

(2014). Examination of the Factorial Structure of Adverse Childhood Experiences and Recommendations for Three f Scores. Psychol Violence , 4(4), 432–444. https://doi.org/10.1037/a0037723

22.

Gaska

K. A.

, & Kimerling

(2018). Patterns of adverse experiences and health outcomes among women veterans. American Journal of Preventive Medicine , 55(6), 803–811.

23.

Gettman

H. J.

, & Gelfand

M. J.

(2007). When the customer shouldn’t be king: Antecedents and consequences of sexual harassment by clients and customers. Journal of Applied Psychology , 92(3), 757.

24.

Gradus

J. L.

, Street

A. E.

, Kelly

, & Stafford

(2008). Sexual harassment experiences and harmful alcohol use in a military sample: Differences in gender and the mediating role of depression. Journal of Studies on Alcohol and Drugs , 69(3), 348–351.

25.

Gruber

J. E.

(1998). The impact of male work environments and organizational policies on women’s experiences of sexual harassment. Gender & Society , 12(3), 301–320.

26.

Haarr

R. N.

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

27.

Harned

M. S.

, & Fitzgerald

L. F.

(2002). Understanding a link between sexual harassment and eating disorder symptoms: A mediational analysis. Journal of Consulting and Clinical Psychology , 70(5), 1170.

28.

I. K.

, Dinh

K. T.

, Bellefontaine

S. A.

, & Irving

A. L.

(2012). Sexual harassment and posttraumatic stress symptoms among Asian and White women. Journal of Aggression, Maltreatment & Trauma , 21(1), 95–113.

29.

I. K.

, Dinh

K. T.

, Bellefontaine

S. M.

, & Irving

A. L.

(2018). Cultural adaptation and sexual harassment in the lives of Asian American women. Women & Therapy , 41(3–4), 281–297.

30.

Holcomb

W. R.

, & Holcomb

M. G.

(2008). A victim typology of sexual harassment. Psychological Reports , 103(3), 819–826.

31.

Holland

K. J.

, Rabelo

V. C.

, Gustafson

A. M.

, Seabrook

R. C.

, & Cortina

L. M.

(2016). Sexual harassment against men: Examining the roles of feminist activism, sexuality, and organizational context. Psychology of Men & Masculinity , 17(1), 17.

32.

Houle

J. N.

, Staff

, Mortimer

J. T.

, Uggen

, & Blackstone

(2011). The impact of sexual harassment on depressive symptoms during the early occupational career. Society and Mental Health , 1(2), 89–105.

33.

Howald

, Walker

, Melick

, Albert

, & Huang

(2018). Addressing sexual harassment in the workplace . SIOP White Paper Series. http://www.infocoponline.es/pdf/ACOSOSEXUALLABORAL.pdf

34.

Hyland

S. S.

, & Davis

(2019). Local police departments, 2016: Personnel . Bureau of Justice Statistics, U.S. Dept of Justice Office of Justice Programs. https://www.bjs.gov/content/pub/pdf/lpd16p.pdf

35.

Ilies

, Hauserman

, Schwochau

, & Stibal

(2003). Reported incidence rates of work-related sexual harassment in the United States: Using meta-analysis to explain reported rate disparities. Personnel Psychology , 56(3), 607–631.

36.

Jones

(2019). An examination of the relationship between moral disengagement, adverse childhood experiences, and sexual misconduct victimization and perpetration among men and women on the college campus in the United States . Middle Tennessee State University.

37.

Kearl

, Johns

N. E.

, & Raj

(2019). Measuring #MeToo: A national study on sexual harassment and assault . Center on Gender Equity and Health (GEH). http://geh.ucsd.edu/wp-content/uploads/2019/05/2019-metoo-national-sexual-harassment-and-assault-report.pdf

38.

Kobayashi

(2018). Exploring sexual harassment in the police force in Japan: A comparative study between Japan and the United States. Journal of Asia-Pacific Studies . https://core.ac.uk/download/pdf/286958621.pdf

39.

Langer

(2017, October 17). Unwanted sexual advances not just a Hollywood, Weinstein story, poll finds . ABC News. https://abcnews.go.com/Politics/unwanted-sexual-advances-hollywood-weinstein-story-poll/story?id=50521721

40.

Lonsway

K. A.

, Paynich

, & Hall

J. N.

(2013). Sexual harassment in law enforcement: Incidence, impact, and perception. Police Quarterly , 16(2), 177–210.

41.

Martin

S. E.

(1994). “Outsider within” the station house: The impact of race and gender on Black women police. Social Problems , 41(3), 383–400.

42.

McDonald

(2012). Workplace sexual harassment 30 years on: A review of the literature. International Journal of Management Reviews , 14(1), 1–17.

43.

McDonald

, Charlesworth

, & Graham

(2015). Developing a framework of effective prevention and response strategies in workplace sexual harassment. Asia Pacific Journal of Human Resources , 53(1), 41–58.

44.

McLaughlin

, Uggen

, & Blackstone

(2017). The economic and career effects of sexual harassment on working women. Gender & Society , 31(3), 333–358.

45.

Mumford

E. A.

, Taylor

B. G.

, Liu

, Barnum

, & Goodison

(2020). Final Summary Overview, Law Enforcement Officers Safety and Wellness: A Multi-Level Study. Chicago: IL:

NORC at the University of Chicago.

46.

National Public Safety Information Bureau. (2017). National directory of law enforcement administrators 2013: Correctional institutions and related agencies . National Public Safety Information.

47.

Nichols

(1995). The brotherhood: Sexual harassment in police agencies. Women Police , 29(2), 10–12.

48.

Parramore

(2018). $MeToo: The economic cost of sexual harassment . Institute for New Economic Thinking. https://www.ineteconomics.org/research/research-papers/metoo-the-economic-cost-of-sexual-harassment

49.

Pina

, & Gannon

T. A.

(2012). An overview of the literature on antecedents, perceptions and behavioural consequences of sexual harassment. Journal of Sexual Aggression , 18(2), 209–232.

50.

Pina

, Gannon

T. A.

, & Saunders

(2009). An overview of the literature on sexual harassment: Perpetrator, theory, and treatment issues. Aggression and Violent Behavior , 14(2), 126–138.

51.

Reese

L. A.

, & Lindenberg

K. E.

(2003). The importance of training on sexual harassment policy outcomes. Review of Public Personnel Administration , 23(3), 175–191.

52.

Richman

J. A.

, Flaherty

J. A.

, & Rospenda

K. M.

(1996). Perceived workplace harassment experiences and problem drinking among physicians: Broadening the stress/alienation paradigm. Addiction , 91(3), 391–403.

53.

Richman

J. A.

, Rospenda

K. M.

, Nawyn

S. J.

, Flaherty

J. A.

, Fendrich

, Drum

M. L.

, & Johnson

T. P.

(1999). Sexual harassment and generalized workplace abuse among university employees: Prevalence and mental health correlates. American Journal of Public Health , 89(3), 358–363.

54.

Roehling

M. V.

, & Huang

(2018). Sexual harassment training effectiveness: An interdisciplinary review and call for research. Journal of Organizational Behavior , 39(2), 134–150.

55.

Rospenda

K. M.

, Fujishiro

, Shannon

C. A.

, & Richman

J. A.

(2008). Workplace harassment, stress, and drinking behavior over time: Gender differences in a national sample. Addictive Behaviors , 33(7), 964–967.

56.

Rospenda

K. M.

, Richman

J. A.

, & Shannon

C. A.

(2009). Prevalence and mental health correlates of harassment and discrimination in the workplace: Results from a national study. Journal of Interpersonal Violence , 24(5), 819–843.

57.

Schneider

K. T.

, Tomaka

, & Palacios

(2001). Women’s Cognitive, Affective, and Physiological Reactions to a Male Coworker’s Sexist Behavior 1. Journal of Applied Social Psychology , 31(10), 1995–2018.

58.

Seklecki

, & Paynich

(2007). A national survey of female police officers: An overview of findings. Police Practice and Research , 8(1), 17–30.

59.

Sklansky

D. A.

(2006). Not your father’s police department: Making sense of law enforcement’s new demographics. Journal of Criminal Law and Criminology , 96(3), 1209–1243.

60.

Smith

S. G.

, Zhang

, Basile

K. C.

, Merrick

M. T.

, Wang

, Kresnow

M. J.

, & Chen

(2016). National intimate partner and sexual violence survey: 2015 data brief . https://www.cdc.gov/violenceprevention/pdf/2015data-brief508.pdf

61.

Somvadee

, & Morash

(2008). Dynamics of sexual harassment for policewomen working alongside men. Policing: An International Journal of Police Strategy & Management , 31(3), 485–498.

62.

SteinÞórsdóttir

F. S.

, & Pétursdóttir

G. M.

(2018). Preserving masculine dominance in the police force with gendered bullying and sexual harassment. Policing: A Journal of Policy and Practice , 12(2), 165–176.

63.

Timmins

W. M.

, & Hainsworth

B. E.

(1989). Attracting and retaining females in law enforcement: Sex-based problems of women cops in 1988. International Journal of Offender Therapy and Comparative Criminology , 33(3), 197–205.

64.

Uggen

, & Blackstone

(2004). Sexual harassment as a gendered expression of power. American Sociological Review , 69(1), 64–92.

65.

Walby

(1990). Theorizing patriarchy . Basil Blackwell.

66.

West

, & Zimmerman

D. H.

(1987). Doing gender. Gender & Society , 1(2), 125–151.