Wo r k-rela t ed St ress in R ela t i o n to G ender - b ased and S e x ual H arass m en t Amo ng a G r o u p o f Sw edis h P o lice O fficers

In this cross-sectional study, we investigated work-related stress, gender-based and sexual harassment among police officers working invulnerable areas in Stockholm. Datawere collected from 152 police officers using a setof questionnaires. The results indicated that job stress in “impact on significant others” and “operational stress” were reported as the two most highly rated work-related stress in police officers. Female police officers rated higher stress in “impact on significant others” and “operational stress” compared to their male counterparts. Job experience was correlated with two stress subscales only in female officers. Working as patrol officers and more than one shift were negatively associatedwithsome stress subscales among male officers. The male officers who had experience of sexual harassment reported higher “self-image” stress and “operational stress”, whereas the female officers had higher “self-image” and “confrontationwith death” stress. After controlling for gender and job experience, sexual harassment had a significant relationship with “operational stress” and “self-image stress” in police officers.


Introduction
There is a large amount of research focusing on work-related stress in police occupation as a demanding and high-stress job, but most of the research has paid little attention to genderbased harassment (GH) and sexual harassment (SH) as a crucial source of job stress among police officers (Thompson et al., 2006;Violanti et al., 2016).Despite abundant literature on job stress, SH and GH and gender role in the police force in other western countries, there is a lack of research on these subjects in police officers in Sweden.However, as part of the #metoo movement in Sweden during the fall of 2017, a Facebook campaign by Swedish female police officers named "self-defence" (#nödvärn) indicated the existence and importance of GH and SH problems inside the police organisation and highlighted the need of changes in this area (Nödvärn, 2017).
In 2016 the Swedish police authority in Stockholm region started an initiative named Mareld to reduce crime and increase security for authorities, businesses, as well as citizens living in vulnerable areas (The Swedish Police Authority, 2016).Vulnerable areas have been described as geographical areas with low socioeconomic status and high criminal rate and insecurity, which negatively affect the inhabitants and society (The Swedish Police Authority, 2017), and make it difficult for the police to perform their mission.According to the latest report of the police authority (2021), there are 19 areas in Sweden classified as particularly vulnerable areas, and six of those areas are in the Stockholm police region.The main characteristics of particularly vulnerable areas are the existence of parallel social structures, extremism, and a high concentration of criminals.According to a report from the Swedish Crime Prevention Council (Brå, 2019), the rate of lethal violence is 3.5 times higher in vulnerable areas compared to the other areas and about 22% of fatal violence against men occurs in these areas.In this study, we focused on the particularly vulnerable areas in the Stockholm police region, and "particularly vulnerable areas" and "vulnerable areas" have been used synonymously.One of the sub-studies in the Mareld project was designed to improve the working environment of the police officers.The results of the sub-study, indicated that GH and SH were frequently reported as a source of stress by police officers (Sundqvist et al., 2021).Accordingly, in the current study we attempted to contribute to the Mareld project by taking into account SH and GH together with other work-related stressors.The study seeks to answer the following questions: a) Are there gender differences in working environment stressor subscales in police officers?b) What are the gender stratified associations between stress subscales and socio-demographic variables, gender-based and sexual harassment?c) Which socio-demographic variables, gender-based and sexual harassment can explain variations in work-related stress subscales?
1.1 Theoretical framework

Work-related stress in police officers
Work-related stress is the response that occurs when the work demands and requirements are not matched to the individual capabilities and resources (WHO, 2021).According to transactional theory by Lazarus and Folkman (1984), stress is defined as a relationship between a person and the environment in which an incident is appraised as taxing his/her resources and threatening his/her well-being.Police work is considered to be a stressful profession, involving demanding situations such as human suffering, instant life or death decisions and also societal responsibilities and strict legal norms in their job (Violanti et al., 2017).
Studies often categorize different sources of police work stressors as operational, organisational and external stressors from the socio-political context (Patterson, 2001;Violanti, 2014).Operational stress implies the inherent aspect of police work.These stressors are due to the unique nature of police work and distinguish this profession from others.Police officers are exposed to varied types of traumatic events in their work.Dealing with critical incidents in the line of duty, fatal accidents, natural disasters, shooting and losing colleagues, sensitive events such as crimes against children, and exposure to physical/verbal attacks, contagious disease and safety-threatening material (e.g.toxic or explosive substances) are some of the operational stressors in police work (Collins & Gibbs, 2003;Johnson et al., 2019;Violanti & Aron, 1995).
Organisational stress refers to problems inside the police organisation that can influence police officers' performance and health.These stressors are mostly shared in common with other occupations and are part of the everyday work life of officers.Organisational stress includes but is not limited to organisational culture and characteristics, administrative practices, management, supervision, organisational support, officers' interactions, resources and training (Houdmont, 2017;Shane, 2010).
As Saunders et al. ( 2019) mentioned external socio-political factors could also produce stress in police officers.These stressors can come from other organisations, political changes, media, citizens and officer's family and friends.Accountability to these groups can work as a source of stress for police officers different from operational and organisational stress.Citizens, for example, expect police officers to perform perfectly without any stress or error.These high, super-heroic expectations can put a lot of pressure on police officers and raise stress.In addition, new policies, regulations and organisational reforms based on political desires can impose extra pressure on police organisation and officers, especially when these changes clash with the police officers' personal ideology and political desires (Kara et al., 2015;Stogner et al., 2020).According to the multisystem approach in police work, introduced by Ghazinour and Rostami (2021), police work should be seen as a complex system in which the parts may not always complement each other (e.g. policy versus police organisation; the police organisation versus police officer) and this can be a source of stress for the police organisation and police officers.

Significance of gender in police work
The police profession is recognised as a practical profession involved in combating crime and applying policing methods.Therefore, police officers are expected to have characteristics associated with interventions requiring physical strength, risk-taking behaviour and aggressiveness, norms and expectations that have been defined as masculine (Lander, 2014).Such a male-dominated system of masculine culture is obvious in police work, with a gendered division of work where working as patrol officer is defined as masculine and real police work, while working in the station as an internal officer is viewed as feminine and unreal police work (Hunt, 1990).Consequently, an "appropriate" police officer is expected to fulfil masculine norms in order to perform police work in the "right" way.As a result, women and non-masculine officers, such as gay police officers (Bernstein & Swartwout, 2012), may face dilemmas at work.
Although women have worked alongside men for several decades, gender issues are still obvious in many aspects of police work.The Swedish police organisation has an almost equal gender distribution, with women representing about 45% of police employees.However, gender distribution according to job categories inside the police organisation is still unequal, where women constitute 33% of police officers and 67% of civil servants (The Swedish Police Authority, 2019).As Dahlgren (2007) explained, an emphasis on the principles of equal terms forced the Swedish police organisation to solve the problem of women's presence in police work in practice.Therefore, female police officers were recruited mainly to work in office jobs or on police patrol work such as helping women and children in domestic violence interventions.As a result, the work of police patrol officers on other cases remained a male area.Thus, instead of removing females from the police organisation, they were employed in a few special areas.A qualitative study in Norway among women who had completed their basic police education and were training as emergency police officers (Jon, 2021) indicated that following the shift in gender equality within society and police work, gender is not perceived as a problem by women police officers.On the other hand, a study on conceptions of gender and competencies among 1346 young and new Scandinavian police officers (Bloksgaard, Fekjaer, & Møberg, 2020) showed between 19% to 26% of the female and 12% to 47% of the male police officers reported a type of gender-stereotypical perceptions on police competences after finishing police training courses.Also, a study (Haake, 2018) on conditions for gender equality in police leadership among 28 police leaders in Sweden demonstrates that both men and women do gender in a gender-stereotypical way through their talks on police organisation, police work, and leadership.It seems that although gendered perceptions have been partly reduced in police work following societal and organizational changes (especially in Scandinavian countries), gender stereotypes have not faded and are still alive and obvious in this context.
Research has revealed a variety experience in police work related to gender (Morash et al., 2006;Rabe-Hemp, 2008).Inequality in career promotion, workplace relationships, negative attitudes of male officers toward female officers, lack of support, underestimation of women's ability to perform police work in society, SH from male colleagues and citizens (Burke & Mikkelsen, 2005;Thompson et al., 2006) are some of the challenges female police officers deal with.According to Dillard (2019) female police trainees were concerned about negative perception of colleagues, supervisors, society, family and friends regarding police work as a woman.Therefore, as female police officers explained in Rabe-Hemp's study (2008), they must be more competent and prove themselves every day to their co-workers, the police organisation, and society.Such perceived demands can increase stress and job pressure for them.

Sexual and gender-based harassment
SH and GH are defined in different ways in research as well as in international and national laws and policies.It has been suggested in research that one reason behind the perceived difficulty in defining SH precisely is that its structure and meaning vary with age and gender (Uggen & Blackstone, 2004).In this study, we take our point of departure from the definition in the European Union context.According to the European Parliament and the European Council (2006) (Directive 2006/54/EC) SH is defined as when "any form of unwanted verbal, non-verbal or physical conduct of a sexual nature occurs, with the purpose or effect of violating the dignity of a person, in particular when creating an intimidating, hostile, degrading, humiliating or offensive environment".GH is another type of discrimination related to gender and defined in a similar way, but which does not involve conduct of an explicitly sexual nature (Fitzgerald et al., 1995).
SH and GH are associated to notions of gender role differences and perceptions of nonmale genders in a male-dominated gender system.In fact, in SH, men objectify femininity and other non-male expressions of gender and sexuality (Baptist & Coburn, 2019).The underlying factor of SH is not only sexual desire, but also notions of women being sexually accessible to men and a wish to dominate, control, humiliate and degrade a person where sexual behaviours are applied as a tool to express or gain power (Palmer, 1988;WHO, 2003).
As Mellon (2013) explained, some men conduct SH as a punishment, especially when they feel that their masculine identity or male privileges have been threatened or challenged by women.
In male-dominated jobs such as police work, in which masculine behaviours are considered valued and appropriate norms, women and other groups with non-normative expressions of gender and sexuality who are not accepted on the basis of masculine values, deal with more discrimination and harassment (Di Martino et al., 2003;Sainio et al., 2007).
SH and GH are offensive and can be destructive and lead to symptoms of stress in police officers (Dowler & Arai, 2008).In their study of 947 US police officers, Morash et al. (2006) indicated that dealing with language harassment and gender, racial and ethnic bias by colleagues was stress predictor in both male and female officers.Thompson et al. (2006), in their three-factor model of stress, defined gender discrimination and GH along with some other items as interpersonal stress in police officers, and emphasised the importance of these stressors in studying work-context stress in the police organisation.Also, Violanti and colleagues (2016), in their study on the most frequent stressors among police officers, mentioned the failure to assess stressors such as SH and other stressors that could affect female police officers as a limitation of the stress questionnaires and consequently the studies in this field.This investigation aims to increase our understanding and fill a knowledge gap in work-related stress and SH and GH.

Methods
This cross-sectional study was performed in three vulnerable areas in Stockholm (Botkyrka, Rinkeby and Södertälje).All police employees working in these areas during the spring of 2020 were able to attend the study.Accordingly, 510 questionnaires were distributed among the police employees working in those areas and 275 police employees completed the questionnaires (about 54% response rate).We had to exclude 89 questionnaires during the process of matching with their front pages and with the related Mareld questionnaires (containing the sociodemographic information and the Police Stress Identification Questionnaire).As the nature of work is different between the employees who work as police officers and those employees in the police who work as civil servants, and the low number of the latter group, we decided to study only the police officers group.After excluding 37 civil servants (from all 189 retrieved questionnaires), 152 questionnaires of police officers (including police patrol officers, investigators, and other internal services officers) were included in the study.As the questionnaire contained some personal and sensitive information about police officers, we decided to gather data through a paper questionnaire in order to protect data security.The COVID-19 situation during the spring of 2020 made it impossible for the researchers to distribute and collect the surveys themselves.Therefore, contact persons from the three local police districts were engaged to distribute and collect the questionnaires.After receiving an explanation of the research and its aims from the contact persons, the respondents were asked to complete two sets of questionnaires.
The questionnaires were completed during regular working hours and the process took approximately one hour.Participation in the investigation was voluntary and the respondents could withdraw at any time during investigation.Informed consent was obtained from all individual participants included in the study.In order to ensure confidentiality, a special code number was provided for each questionnaire; the respondents were asked to detach the front page of the questionnaire containing the personal identification number and personal sociodemographic characteristics after completing the questionnaire.The question-naire and front page were sent in separate sealed envelopes to the research team, and the principal researcher matched the questionnaire and front page using each questionnaire's specific code.

Instruments
Data collection was performed using two sets of questionnaires encompassing three parts: the sociodemographic questions and the Police Stress Identification questionnaire from the Mareld project, and SH and GH questions as a complementary part that was added as the second set of questions to the Mareld questionnaires.
The sociodemographic questions were designed to obtain socio-demographic information such as respondent's gender, age, education, marital status, number of children, type of work, job position and some other sociodemographic questions.
The Police Stress Identification Questionnaire (PSIQ) is a self-report measurement of police work stress developed by Ghazinour et al. (2021).The questionnaire consisted of 42 items placed in five subscales (organisational stress, operational stress, impact on significant others, self-image, confrontation with death) and two single items to measure different types of stressors in police work.Each item was scored on a nine-point scale (ranging from "no stress" to "the most stressful level").In the current study, as we analysed the data based on the total scores of stress subscales (mean score of items in each subscales), the single items were not included in the analysis.
The SH and GH questions were designed by the research team to obtain the respondents' SH and GH experiences in police work.The questions were based on the existing questionnaires in the field of SH, e.g. the Sexual Experiences Questionnaire (Fitzgerald et al., 1988) and the Sexual Harassment Inventory (Murdoch & McGovern, 1998).Also, according to the literature and testimonies of the Nödvärn campaign, we attempted to take police work environment, source of sexual harassment, and gender of perpetrators into account.The questionnaire included 36 questions about experiences of SH and GH from public citizens, supervisors and colleagues during the last 12 months.Each question was scored on a fivepoint scale between "never" (0) to "several times in a week" (5).The respondents were also asked to indicate the gender of the perpetrator for each item.In this study, the respondents' answers to the questions were categorized into two groups based on "had at least one experience" (1) or "had no experience" (0) of SH and GH.

Statistical analysis
Descriptive measures of the variables are reported in terms of numbers and percentages for categorical socio-demographic variables, SH and GH, and mean scores and standard deviations for continuous socio-demographic variables (age and job experience) and different dimensions of job stress.As the dependent variables were not normally distributed (tested by the Kolmogorov-Smirnov test), non-parametric tests were applied to analyse the data.Therefore, the Mann-Whitney test and Kruskal-Wallis test were applied to assess group differences for continuous variables, and c2-test was applied for categorical variables.Also, Spearman's correlation coefficient was used to indicate associations between continuous variables.Multiple linear regression analysis was performed with each work stress subscale as dependent variable and gender, job experience, and type of harassment as predictor factors.Two sets of multiple regression analyses were separately performed for SH and GH for each of the dependent variables.The variances were homogenous and the residuals showed normal distribution, suggesting that the assumptions (Barton & Peat, 2014) for multiple regression models were met.The analysis was performed applying SPSS 26.

Results
As descriptive statistics in Table 1 shows, of the 152 participants in this investigation, 30% were female, 70% of the respondents were male, and no one mentioned any other type of gender.The average age of participants in male officers was 39.24±9.41years old (range 25-64 years) and in female officers was 41.61±9.58years old (range 26-63 years).The vast majority of police officers were born in Sweden.About 86.8% of male and 77.8% of female respondents were in a type of intimate relationship (marriage or partnership) and about 38.8% of the male and 22.5% of the female officers were childless.Most of the respondents in both genders had university or college education, and there was no meaningful difference between male and female with regard to police education degree.The average period of job experience was 11.95±10.88years (range 0.5-45 years) in the male and 14.60±11.11years (range 1-41 years) in the female officers.The majority of female respondents (64.5%) were working as internal services (investigation, reception and other types of internal work), whereas 60.18% of male officers were working as police patrol officers (external services).Furthermore, most of the participants in managerial positions were male officers and the percentage distribution between the males and females in managerial positions was significantly different (36.4% of male officers compared to 20% of female officers).The data showed most of the female participants were working day shifts, and there were more men working in two-shift, three-shift and mixed plans.The collected data show that 20.8% of the male and 44.2% of the female police officers experienced one to several instances of genderbased harassment (GH) (Range 1-8 harassments), for which the difference was statistically significant.Also about 49% of the male police officers and 44% of the female police officers reported sexual harassment (SH) (Range 1-8 harassments), but the difference was not statistically significant.
In this study the subscales "impact on significant others" and "operational stress" were reported as the two most highly rated work-related stresses in both the male and female police officers.The female police officers scored higher than the males on all work stress subscales, which was statistically significant in "impact on significant others" and "operational stress" (Table 2).
Spearman's correlation between the stress subscales and age showed the younger the female police officers, the more stress reported in "self-image", whereas the negative correlation was significant in "operational stress" among the male police officers.Also having more job experience were correlated with less stress in "self-image" and "confrontation with death" in the female officers (Table 3).
There were no differences between the stress subscales and categories of birthplace and managerial position, having children and educational level, but the score of "confrontation with death" was statistically higher among male police officers who were married or had any other type of intimate relationship (Table 4).Applying the Kruskal-Wallis chi-square test to compare stress between three work type groups of police officers -patrol police officers, investigators and other police officers (internal work) -indicated the score of "operational stress" was significantly different in different work type groups only in male police officers.Using the Man-Whitney test for each pair of work type groups showed the difference is significant between the male police patrol officers and those who worked as investigator.There were no significant differences in stress subscales in different work types in female police officers (Table 4).Also, the Mann-Whitney test indicated that the stress in "impact on significant others" and "operational stress" were significantly different in different categories of work shift plan (day shift and two-/threeshift plans) only among the male police officers (Table 4).
The female police officers who had experience of GH reported higher scores in "confrontation with death" stress than female police officers without experience of GH.This difference was not significant in male police officers.Both male and female police officers with experience of SH reported higher stress in the "self-image" subscale.In addition, the male police officers with experience of SH mentioned higher "operational stress" than male officers without experience of SH.The female officers who had experience of SH had higher stress in "confrontation with death" compared with female officers without experience of SH (Table 4).
The multiple regression analysis for each subscale of work-related stress with stress subscale as a dependent variable and gender, job experience, and SH as independent variables showed that the variance in the dependent variables could explain between 4% (confrontation with death stress) and 13% (self-image stress) of the variance in the different stress subscales (except for organizational stress) and gender significantly contributes to this explained amount of variance (except for confrontation with death stress).After controlling for gender and job satisfaction, SH still had a significant relationship with operational stress and selfimage stress in police officers (Table 5).However, the multiple regression analysis for each subscale of work-related stress with stress subscale as a dependent variable and gender, job experience, and GH as independent variables, indicated that the variance in the dependent variables could explain between 4% (confrontation with death stress) and 9% (self-image stress) of the variance in the different stress subscales (except for organizational stress), and gender and job experience can mostly predict these variations.The model showed no significant association between GH and stress subscales after controlling for gender and job experience (Table 6).

Discussion
The results showed the subscales "impact on significant others" and "operational stress" were the highest scored stress in police officers.This result confirms the findings of Violanti and Aron (1995) and Violanti and colleagues (2016) that indicated some stressors such as "Exposure to battered or dead children", "Killing someone in the line of duty", "Fellow officer was killed in the line of duty", and "Situations requiring use of force" as the most stressful events in police officers.Houdmont (2017) reported that many police studies showed organisational stressors (context stressors) as the main perceived stress in police officers.Collins and Gibbs (2003) indicated that the stressors with the highest scores were lack of control on workload, inadequate support and consultation and extra workload.Also Shane (2010), in a study of 461 patrol police officers from two urban police departments, showed that organisational stressors were reported to be higher than operational stressors.The different findings may be due to different police organisational cultures.Organisational stressors are mostly in relation to organisation culture, bureaucratic attributes and organisation budget (adequate number of personnel, equipment and salary) (Paais, 2018;Peterson & Wilson, 2002) which can vary among different police organisations in different countries.Furthermore, what is crucial is whether the instruments asked police officers to score the stressors based on their experiences as perceived stress or just as probable stressors that can occur in their work.As police officers face organisational stressors more frequently than critical events in police operation, those stressors may be reported as the most perceived stressors.Also, applying different instruments to measure stress makes it difficult to compare the results.For instance, some stressor items, which were categorised as "impact on significant others" in the questionnaire applied in this study (e.g.colleague injury or death in the line of duty, family life threat) were considered as operational stress in other instruments.Police officers working in vulnerable areas are involved in more serious police interventions than other officers; this constant exposure might be another explanation for reporting higher operational stress by them.
Our results indicated female police officers rated higher than male officers on all work stress subscales, which was statistically significant in "impact on significant others" and "operational stress".This is consistent with the findings of Violanti et al. (2016) that female officers reported a higher rate of stress than males to events involving tragedy or danger.Also, Bartol et al. (1992) found in their study that the stressors experienced were the same in the male and female officers except for those stressors related to police operation, where the women reported more stress in facing tragedy and the occasions associated with a responsibility for the safety of the citizens and their colleagues.This finding is in line with our results on higher stress in relation to significant others in female police officers.In addition, female officers working alongside male police officers in a male-dominated environment with masculine values have to prove their capabilities and competencies in policing to their colleagues, supervisors, the police organisation and even citizens (Morash & Haarr, 2012;Rabe-Hemp, 2008).Female police officers who are more involved in police interventions -which has been defined as masculine work in police culture -have this extra pressure, which likely explains the higher stress in this group.
Regarding age and job experience, our findings confirm other studies that showed younger and less experienced police officers dealing with higher stress in operational and organisational items (Balakrishnamurthy & Shankar, 2009;Queirós et al, 2020).As in other professions, getting older and dealing with different stressors helps individuals to manage stress and find the best coping strategy for various forms of stress.In addition, more experience in police interventions improves their job expertise, self-esteem, self-image and work performance, and consequently lowers feelings of stress (Bannerman, 1996;Lester, 1986).
Male police officers who work two and three shifts reported higher stress in "impact on significant others" and "operational stress" than those who work day shifts.Ma and colleagues (2015), in their study among 365 police officers in New York, revealed work-related stressful events and police stress were reported higher in officers working the afternoon and night shifts.In another context, researchers indicated that shift workers dealt with less supervisor support, job control, and discretion, and more harassment than those working a fixed day shift (Nabe-Nielsen et al., 2009).Involvement in more stressful events and higher job demands with less job support and control may explain the higher score of stress in police officers who worked two and three shifts.As the majority of police officers who worked two and three shifts were male patrol officers, the significant association is clearer in male police officers.
A comparison of stress between three groups of police officers (patrol police officers, investigators, and other police officers -internal workers) showed that "operational stress" in the male patrol police officers was higher than males in the other two groups.Constant and direct exposure to threats and dangers of police interventions and stressful events may be an explanation for the higher operational stress in this group (Abdollahi, 2002;Acquadro Maran et al., 2015).Furthermore, as females encompass only 21% of patrol officers, the relationship is not as evident for them as it is for male police officers.
Our findings demonstrate that both male and female police officers experienced GH and SH, but that a statistically higher percentage of female officers reported exposure to GH than male officers.Few studies have assessed GH and SH separately; most of them considered GH as part of SH and reported higher harassment in women than in male police officers (de Haas et al., 2009;Lonsway et al., 2013).The finding of the current study is congruent with the study by Savicki et al. (2003) that showed higher GH in female police officers.In the process of gendered socialisation and defining of gender roles and norms, women are mostly expected to be passive, caring, and accepting (Nicolson, 2015).Working as female police officers is at odds with the predefined feminine roles, disturbs the dominant dichotomy of masculine and feminine roles, and is perceived as interfering in the realm of men.Therefore, female officers, especially those who work as patrol officers, are considered inappropriate for police work and face more discrimination and harassment from citizens and colleagues.Unlike GH, there was no significant difference in the experience of SH between male and female police officers that is not in line with previous studies which reported higher harassment in female police officers than male officers.More analysis on SH items in our study clarified that the police officers reported SH from citizens as the most frequent form of SH.Police patrol officers are more frequently exposed to hostile citizens and criminals during patrolling and police interventions.Since more male police officers worked as patrol police officers and in contact with citizens, male police officers reported a trivially higher percentage of SH.Also, as most of the perpetrators (colleagues and citizens) were men, another explanation for this finding could be masculinity in police work and overall society.It is possible that the findings mirror a male use of sexual language as a way to position other men as subordinate in a masculine hierarchy (Vojdik, 2014).There is a need for more research in this field to investigate GH and SH from the inside (colleagues and supervisors) and outside (citizens) of police organization in female and male police officers.
Assessing association between stress subscales and SH and GH showed that the male and female officers who had experience of SH reported higher "self-image" stress.Police officers have legal authority and power to intervene in citizens' lives and control them in order to maintain safety and security and prevent crime.This legal discretion gives them a unique position of power that is construed by citizens as a restriction of their freedom and as police misuse of authority in conflict with citizens (Cao & Huang, 2000), and leads them to oppose the police's legal power by using SH.In other words, citizens apply SH and GH to degrade and humiliate police officers and balance the power differences between them and police officers in their own way (McLaughlin et al., 2012).Therefore, SH as negative and destructive comments or acts can target police officers' self-image and causes higher stress in both male and female officers.As more male officers are involved in performing police interventions as police patrol officers, they face more sexual threats and harassment from citizens that can explain the positive relationship between experiencing SH and "operational stress" in the male police officers.The significant association between SH and "self-image" stress and "operational stress" after controlling for gender and job experience confirmed the role of SH in predicting higher stress in the mentioned subscales.
In addition, dealing with stressful situations related to "confrontation with death" was associated with the experience of both GH and SH in female police officers.They received degrading gendered and sexual comments or behaviours mostly from citizens and colleagues to disparage their competence at police work, especially in dangerous interventions and threatening situations in confrontation to death.The more threatening the situation female police officers face, the more obvious the picture of male dominance -with masculine norms and expectations -can be revealed in police work.

Conclusion
In conclusion, the results indicated that police officers of both genders reported "impact on significant others" stress and "operational stress" as the two most highly rated work-related stress.Female police officers rated higher stress in "impact on significant others" and "operational stress" compared to their male counterparts (question a).Among socio-demographic variables, job experience was correlated with two stress subscales ("self-image" stress and "confrontation with death" stress) only in female officers.The younger age female police officers reported more stress in "self-image" whereas the younger male police officers rated higher "operational stress".Working as a patrol officer and on more than one shift were negatively associated with "operational stress" and "impact on significant others" stress among male officers.The male officers who had experience of sexual harassment reported higher "self-image" stress and "operational stress" whereas the female officers had higher "self-image" and "confrontation with death" stress (question b).Multiple regressions between each stress subscale and gender, job experience, and SH/GH indicated gender and job experience could significantly predict variations of stress subscales (except organizational stress).After controlling for gender and job experience, sexual harassment had a significant relationship with operational stress and self-image stress in police officers (question c).
The main limitation of this study was the restrictions due to the COVID-19 situation, which limited the active role of the research team members in the data collection process and resulted in an inadequate response rate (about 50%) of police respondents.Furthermore, some respondents' data were lost in the process of matching each questionnaire with its front page, and matching that with the data obtained from the Mareld project.Therefore, the small number of participants limited the analysis based on the type of work of police employees (police patrol officers, investigators, and receptionists) and the analysis of multiple regression models.In addition, the GH and SH results were based on the frequency of the harassments experienced by the participants and it was impossible to assess the severity of the problem in the respondents.The sample used in this study is not representative of all Swedish police officers.Further studies on job-related stress in a larger and representative population of police employees can help to assess job stress subscales in police work (in both vulnerable and non-vulnerable areas),and clarify the role of sexual and gender-based harassment as an important stressor in police work.In addition, further studies applying an intersectional perspective (considering gender, ethnicity, type of police work, and work area) can provide a more comprehensive picture of work stress in police employees.
Despite the limitations, the findings should be considered by the police authority to improve the workplace environment for police officers.The results should also be helpful for policymakers and researchers in this field to take into account GH and SH besides other work-related stressors.

Table 1 .
Characteristics of the participants by gender

Table 2 .
Work stress subscales by gender in police officers

Table 3 .
Correlation between work stress subscales and age, job experience by gender

Table 4 .
Association between work stress subscales and marital status, shift plan, type of work, gender-based harassment, and sexual harassment by gender

Table 5 .
Multiple regressions with work-related stress subscales as independent variables in association with socio-demographic (gender and job experience) and sexual harassment

Table 6 .
Multiple regressions with Work-related stress subscales as independent variables in association with socio-demographic (gender and Job experience) and gender-based harassment