European Journal of Work and Organizational ...

7 downloads 0 Views 179KB Size Report
Mar 1, 2010 - Keywords: Double male dominance; Police; Sexual harassment. Correspondence ..... divisions and the Division National Police Services were represented in the sample. ..... Cleveland, J. N., & Kerst, M. E. (1993). Sexual ...
This article was downloaded by: [Timmerman, Greetje][University of Groningen] On: 25 November 2010 Access details: Access Details: [subscription number 907173570] Publisher Psychology Press Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 3741 Mortimer Street, London W1T 3JH, UK

European Journal of Work and Organizational Psychology

Publication details, including instructions for authors and subscription information: http://www.informaworld.com/smpp/title~content=t713684945

Sexual harassment in the context of double male dominance

Stans de Haasa; Greetje Timmermanb a Rutgers Nisso Group, Dutch Expert Centre on Sexuality, Utrecht, The Netherlands b Behavioural and Social Sciences, University of Groningen, Groningen, The Netherlands First published on: 01 March 2010

To cite this Article de Haas, Stans and Timmerman, Greetje(2010) 'Sexual harassment in the context of double male

dominance', European Journal of Work and Organizational Psychology, 19: 6, 717 — 734, First published on: 01 March 2010 (iFirst) To link to this Article: DOI: 10.1080/09541440903160492 URL: http://dx.doi.org/10.1080/09541440903160492

PLEASE SCROLL DOWN FOR ARTICLE Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf This article may be used for research, teaching and private study purposes. Any substantial or systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.

EUROPEAN JOURNAL OF WORK AND ORGANIZATIONAL PSYCHOLOGY 2010, 19 (6), 717–734

Sexual harassment in the context of double male dominance

Downloaded By: [Timmerman, Greetje][University of Groningen] At: 12:30 25 November 2010

Stans de Haas Rutgers Nisso Group, Dutch Expert Centre on Sexuality, Utrecht, The Netherlands

Greetje Timmerman Behavioural and Social Sciences, University of Groningen, Groningen, The Netherlands

It has been suggested that numerically male-dominated workplaces propagate cultural norms that support sexual bravado, sexual posturing, and the denigration of feminine behaviour (Sbraga & O’Donohue, 2000). These cultural norms are features of normative male dominance, which have been shown to increase the risk of sexual harassment. This implies that the effect of numerical male dominance on sexual harassment may be mediated by the level of normative male dominance in the work environment. The aim of this study was to test this assumption. Our sample consisted of 1295 police women, who filled out an Internet questionnaire. The results suggest that normative male dominance indeed mediates the relation between numerical male dominance and sexual harassment. We add to the sexual harassment literature by building on Gruber’s concept of double dominance. Our study helps to make clear why women are at greater risk of sexual harassment in work situations where men outnumber women than in more gender-balanced workplaces. Keywords: Double male dominance; Police; Sexual harassment. Correspondence should be addressed to Ms. S. de Haas, Rutgers Nisso Groep, PO Box 9022, 3506 GA Utrecht, The Netherlands. E-mail: [email protected] The data reported were gathered by the Rutgers Nisso Groep, the Dutch Expert Centre on Sexuality, as part of a larger study, which was commissioned by the Dutch Police Union (NPB) and financed by the Grant Committee Labour Market and Educations Police (SAOP). During the research process the researchers (Ms. De Haas, MSc, Ms. Zaagsma, MSc, Ms. Ho¨ing MSc, Ms. Van Berlo, MSc, and Ms. Vanwesenbeeck, PhD) were advised by a committee with representatives of the NPB (Mr. Van Duijn and Ms. Van de Putte) and the SAOP (Ms. Hooijenga), as well as the Police Trade Union ACP (Mr. Van de Kamp and Mr. Van der Pal), the Ministry of the Interior and Kingdom Relations (Ms. Broek, Ms. Hovius, and Mr. Marsman) and the Dutch Police Institute (Mr. Diepenbach). Data were gathered with support of the chiefs of the police divisions. Ó 2010 Psychology Press, an imprint of the Taylor & Francis Group, an Informa business http://www.psypress.com/ejwop

DOI: 10.1080/09541440903160492

Downloaded By: [Timmerman, Greetje][University of Groningen] At: 12:30 25 November 2010

718

DE HAAS AND TIMMERMAN

Since the mid-1970s, sexual harassment has received more and more research attention. Meta-analytic data from the United States revealed that one out of every four women has experienced sexual harassment at work (Ilies, Hauserman, Schwochau, & Stibal, 2003). Furthermore, research on the effects of sexual harassment has shown that sexual harassment is a workplace danger that is not only detrimental to the target’s well-being (Willness, Steel, & Lee, 2007), but also brings about organizational costs, such as lowered productivity and morale, higher absenteeism and turnover, and legal fees (Sbraga & O’Donohue, 2000). It has been shown that sexual harassment is associated with male dominance (Gruber & Morgan, 2005). However, it is not yet clear which mechanism underlies the relation between sexual harassment and male dominance. We aim to study why women are at greater risk of sexual harassment in work situations where men outnumber women than in more gender-balanced workplaces. Sexual harassment can be defined as offensive behaviour that has a sexual dimension (O’Donohue, Downs, & Yeater, 1998). A frequently used psychological definition was formulated by Fitzgerald, Swan, and Magley (1997) as ‘‘unwanted sex-related behaviour at work that is appraised by the recipient as offensive, exceeding her resources or threatening well-being’’ (p. 15). In previous research three forms of sexual harassment were distinguished: gender harassment (nonsexual gender-based experiences, such as sexist comments), unwanted sexual attention (uninvited sex-based comments, gestures, or attempts at physical contact), and sexual coercion (workrelated intimidation or rewards used to induce sexual cooperation; Fitzgerald et al., 1988; Fitzgerald, Gelfand, & Drasgow, 1995). Gutek, Murphy, and Douma (2004) have noted that the gender harassment does not refer to sexual behaviour and as such may not be considered as sexual harassment. In our study we focused on unwanted sexual attention and sexual coercion. With regard to the causes and aetiology of sexual harassment, different models have been proposed. Most of these models share an important feature: Power differences between victims and perpetrators are considered to be the core of the problem (McKinney & Maroules, 1991; Wilson & Thompson, 2001). Male dominance and masculine hegemony are regarded the root causes of sexual harassment, which means that sexual harassment is used to accomplish or maintain masculinity and status differences between men and women (O’Donohue et al., 1998; Stockdale, 2005). So the objective of sexual harassment is not necessarily the pursuit of sex, but rather intimidation and the satisfaction of power needs (Pryor, Giedd, & Williams, 1995). Recently, the study of sexual harassment in organizations addresses the relation between sexual harassment and male dominance from a more complex perspective (Gruber & Morgan, 2005). These studies not only focused on how women experience male dominance, they also investigated

Downloaded By: [Timmerman, Greetje][University of Groningen] At: 12:30 25 November 2010

MALE DOMINANCE AND SEXUAL HARASSMENT

719

the place of men in sexual harassment, as perpetrators as well as victims. Building on Gruber’s concept of double dominance, Gruber and Morgan (2005) have proposed to study male dominance as a multifaceted concept, consisting of at least two dimensions: numerical and normative male dominance. Organizations that are numerically dominated by men have many more male than female employees. In the past, most research has explored male dominance from a numerical perspective, e.g., sex-ratio studies. However, over the last two decades normative male dominance has gained more attention. Normative male dominance refers to an organizational or occupational culture that rewards traditional masculine values, such as support of the heterosexual norm, the devaluation of women, sexual bravado, aggression, emotional self-regulation, risk taking, and technological competence. The relation between both numerical and normative male dominance and sexual harassment has been supported by many recent empirical studies (Gruber & Morgan, 2005). Normative male dominance can be put into operation in several ways, for example as sexist behaviour. In organizations where sexist behaviour was tolerated or part of the job, women were denigrated and are not accepted as equally capable as men (O’Donohue et al., 1998). It has been found that sexist behaviour is related to sexual harassment; women who perceived sexist attitudes and beliefs among male co-workers were more likely to have experienced sexual harassment (O’Hare & O’Donohue, 1998). In contrast, respondents who perceived their department as positive towards the equal treatment of men and women experienced less unwanted sexual behaviour (Timmerman & Bajema, 2000). Another manifestation of male dominance is the sexualized organization. Sexualized work environments are defined by frequent sexual talking and joking (O’Donohue, 1997). Sexual harassment was more prevalent in sexualized companies than in professional companies (Welsh, 1993). For example, research in a sample of male university faculty and staff revealed that men who report sexualized workplace relations and interactions between co-workers are more likely to engage in sexual harassment of women (Dekker & Barling, 1998). Furthermore, in male-dominated organizations harassing behaviours are likely to be tolerated, which in turn contributes to the prevalence of sexual harassment. If managers tolerate harassing behaviours, potential harassers may perceive that they are free to harass. Research by Hulin, Fitzgerald, and Drasgow (1996) has confirmed this assumption. The perceived risk to the victim of complaining, the likelihood that their complaints will not be taken seriously, and the probability that offenders will not be sanctioned in any meaningful way increased the incidence of sexual harassment. Further, women who believed their organization is tolerant of sexual harassment experienced higher levels of sexual harassment (Fitzgerald, Drasgow, Hullin,

Downloaded By: [Timmerman, Greetje][University of Groningen] At: 12:30 25 November 2010

720

DE HAAS AND TIMMERMAN

Gelfand, & Magley, 1997). Similarly, in organizations where more men saw the manager as indifferent or neutral with regard to sexual harassment, more women were actually harassed than in organizations where fewer men perceive the manager as indifferent or neutral with regard to sexual harassment (Pryor et al., 1995). In numerically male-dominated organizations it has been found that sexual harassment occurs relatively often, which means that women who work in predominately male workplaces experience more instances of sexual harassment than women in more gender-balanced workplaces (Gutek, Cohen, & Konrad, 1990; Welsh, 1999). Fitzgerald, Drasgow, et al. (1997) included sex ratio in their integrated model of antecedents and consequences of sexual harassment as an aspect of job gender context. Support for this model was found in several empirical studies, for example in the army (Fitzgerald, Drasgow, & Magley, 1999). In the federal workplace it has also been found that as the ratio of male co-workers increases women are more likely to have experienced sexual harassment (Jackson & Newman, 2004). Also, in meta-analytic studies the link between sex ratio and sexual harassment had been found (Willness et al., 2007). The effect of double male dominance has been studied too. According to Baigent (2005), firefighting is a doubly male-dominated occupation. It is numerically dominated by men as male firefighters outnumber female firefighters by far. In addition, the cultural norms of firefighting are extremely masculinized. Qualitative research among firefighters in the USA has revealed that for aspiring firefighters conquering a place into the existing male-dominated culture is their first priority. In this process senior firefighters acted as gatekeepers to an informal hierarchy of men. It appeared that to socialize female aspirants these gatekeepers had little means but sexually harassing these women (Baigent, 2005). It seems that numerically male-dominated workplaces propagate cultural norms that support sexual bravado, sexual posturing, and the denigration of feminine behaviour (Sbraga & O’Donohue, 2000). These cultural norms are features of normative male dominance, which have been shown to increase the risk of sexual harassment. This implies that the effect of numerical male dominance on sexual harassment may be mediated by the level of normative male dominance in the work environment. However, to our knowledge this has not been studied empirically. Building on the model of Fitzgerald, Drasgow, et al. (1997), many researchers studied the effect of job gender context and organizational tolerance on sexual harassment (Cantisano, Domı´ nguez, & Depolo, 2008; Willness et al., 2007). However, mechanisms that explain why numerical male dominance is a risk factor have not received sufficient research attention. As such it is not yet clear which mechanism underlies the relation between sexual harassment and numerical male dominance. Previous research has suggested that numerical male dominance

Downloaded By: [Timmerman, Greetje][University of Groningen] At: 12:30 25 November 2010

MALE DOMINANCE AND SEXUAL HARASSMENT

721

reinforces normative male dominance, which in turn is related to an increased risk of sexual harassment. To test this assumption, normative male dominance needs to be studied as a mediator. The aim of the present study was to test whether normative male dominance mediates the relation between numerical male dominance and sexual harassment. By studying normative male dominance as a mediator, our study helps to make clear why women are at greater risk of sexual harassment in work situations where more men than women are employed. We add to the sexual harassment literature by providing more insight in the effect of male dominance as a multifaceted concept. As such we build on Gruber’s concept of double dominance. Based on previous research, we hypothesized that the effect of numerical male dominance on sexual harassment is mediated by normative male dominance. We tested our hypothesis with survey data from a sample of police officers working in 26 Dutch police divisions. As a whole the police organization is numerically dominated by men. The total police population across The Netherlands consists of 72% men and 28% women. However, sex ratios vary in different work situations. Although men also experience sexual harassment at the workplace, it has been consistently found that women are more at risk of becoming a victim of sexual harassment than men (Sbraga & O’Donohue, 2000). Therefore, women were the focus of this study. Further, it should be noticed that young and single women are relatively more often victimized. Young and unmarried women have less sociocultural power than older or married women (Kauppinen-Toropainen & Gruber, 1993) and low sociocultural power is related to a high risk of sexual harassment (Harned, Ormerod, Palmieri, Collinsworth, & Reed, 2002). So, age and partner status are related to power issues and sexual harassment; therefore, we controlled for age and partner status in our analyses. We did not control for organizational power, because that type of power is closely related to age (MacKinnon, 1979). Young women usually do not hold the powerful organizational positions, and organizational power can be viewed as an extension of sociocultural power (Ragins & Sundstrom, 1989).

METHOD Procedure The study was part of a project on sexual harassment, bullying, and diversity among Dutch male and female police officers. For the present study we used the sexual harassment data of the female officers. On behalf of a police labour union, the Rutgers Nisso Groep, Dutch Expert Centre on sexuality, sent out an introduction letter to a representative group of police officers at their home addresses. Respondents were invited to anonymously fill out an

722

DE HAAS AND TIMMERMAN

Internet questionnaire either at home or at work. The Internet questionnaire has been placed on the Internet by the Dutch Police Academy. As an incentive, one out of every hundred respondents received a book token of 25 Euros. After 4 weeks all respondents received a written reminder to fill out the questionnaire.

Downloaded By: [Timmerman, Greetje][University of Groningen] At: 12:30 25 November 2010

Participants The sample consisted of 1295 police women. All 25 Dutch regional police divisions and the Division National Police Services were represented in the sample. The average age was 39 years (SD ¼ 9.40). Of the women 51% (n ¼ 652) worked as executives (graduated from the police academy and qualified for the core business of police work and law enforcement) and 8% (n ¼ 102) were part of the managerial staff (in charge of a team of police officers). Of the police women 24% (n ¼ 307) were single. Six per cent of the police women (n ¼ 75) were from ethnic minority groups, such as Surinamese, Turkish, or Moroccan. Two per cent of the women worked exclusively with men, 49% worked with more men than women, 35% worked with as many men as women, and 14% worked with more women than men or exclusively with women. Although most survey studies in The Netherlands reach a response rate of 25%, the response rate of our sample was 15%. Low response rate was possibly due to questionnaire overload and technical problems with getting the Internet questionnaire started. In spite of the low response rate, the total sample of men and women was representative for age and post (executive or nonexecutive and managerial or nonmanagerial; Table 1).

Measurements Besides questions relating to demographic aspects, the questionnaire included instruments measuring sexual harassment and male dominance. Demographic variables were age and being single or in a relationship. The latter is used as a dummy variable with 0 referring to having a partner and 1 referring to being single. Sexual harassment. The Dutch adaptation of the Sexual Experience Questionnaire (SEQ; Fitzgerald et al., 1988; Gelfand, Fitzgerald, & Drasgow, 1995) was used to measure sexual harassment. Three items measuring gender harassment were deleted from the original scale, and replaced by other items that have been found to be applicable to the police context in previous research (Sandfort & Vanwesenbeeck, 2000). The translation from English into Dutch made some minor changes in formulation necessary. The Dutch version presents 18 sexually harassing

MALE DOMINANCE AND SEXUAL HARASSMENT

723

Downloaded By: [Timmerman, Greetje][University of Groningen] At: 12:30 25 November 2010

TABLE 1 Sample and population characteristics Sample (N ¼ 4296)d

Population

Age 18–30 31–40 41–50 50þ

15% 23% 35% 26%

15% 27% 37% 21%

Job status Manageriala Executiveb Nonexecutivec

18% 57% 25%

20% 57% 23%

a Respondent is in charge of a team of police officers. bRespondent does not have a managerial position and is graduated from the police academy (qualified for the core business of police work and law enforcement). cRespondent does not have a managerial position and did not graduate from the police academy (e.g., secretarial employee). dThe study was part of a larger project. For the present study we only used the data on women. As the population characteristics are not available for men and women separately, we presented the total sample here.

behaviours and asks respondents to indicate the frequency of having experienced these personally from superiors or co-workers during the previous 24 months. Scores are on a 5-point scale, with 1 ¼ ‘‘never’’, 2 ¼ ‘‘a few times’’, 3 ¼ ‘‘several times’’, 4 ¼ ‘‘regularly’’, and 5 ¼ ‘‘many times’’. We conducted a Principal Component Analysis with oblique rotation on our sexual harassment data. The Scree plot revealed two factors. The Eigenvalues associated with these factors are 5.6 and 2.0, respectively. The first, Unwanted Sexual Attention and Advances, consists of 12 items referring to forms of verbal and behavioural harassment. The first factor explained 30.9% of the variance. The loadings on the first factor ranged from .27 to .74. The second factor, Sexual Coercion, consists of six items referring to behaviours in which resistance, promises, or retaliation are into play. The second factor explained 11.3% of the variance and the loadings on the second factor ranged from .20 to .73. As the sample exceeds 1000 all loading can be considered statistically meaningful. Two items: ‘‘Unwanted attempts to stroke or fondle you’’ and ‘‘Subtly bribed into sexual intimacy with promise or reward’’ loaded on both factors. Factor loadings for the former were .42 and .41, whereas factor loadings for the latter were .27 and .21. Based on the content of the items, these items were allocated to a subscale (Table 2). The two-factor structure was inconsistent with the three-factor structure that was found with the revision of the SEQ of 1995 (Fitzgerald et al., 1995). Replacing the three

724

DE HAAS AND TIMMERMAN

Downloaded By: [Timmerman, Greetje][University of Groningen] At: 12:30 25 November 2010

gender harassment items by three other items measuring unwanted sexual attention and advances may explain this inconsistency. In the analyses sum scores of the two scales were used. Cronbach’s alpha was .89 for unwanted sexual attention and advances and .68 for sexual coercion. Item-total correlations ranged from .30 to .76 for the first scale and from .27 to .62 for the second scale. Numerical male dominance. We operationalized numerical male dominance as the ratio of men and women in the daily work situation. Numerical male dominance was measured with one question ‘‘Do you have more contact with male or with female colleagues in your daily work?’’ Answers could be given on a 5-point scale, with 1 ¼ ‘‘exclusively women’’, 2 ¼ ‘‘more women than men’’, 3 ¼ ‘‘as many men as women’’, 4 ¼ ‘‘more men than women’’, and 5 ¼ ‘‘exclusively men’’. Normative male dominance. Normative male dominance was measured with nine propositions which were selected from a work and organization culture scale created by Sandfort and Vanwesenbeeck (2000). These propositions measured several aspects of male dominance such as TABLE 2 Percentages of reported sexual harassment (experienced at least once) % Unwanted sexual attention and advances Unwanted attempts to draw you into a discussion of personal or sexual matters Crude or offensive sexual remarks made to you personally Offensive remarks about your appearance or body Unwanted sexual attention Staring, leering, or ogling you in a way that made you feel uncomfortable Attempt to establish a romantic or sexual relationship despite discourage Send sexually tainted messages by e-mail Touched you in a way that made you feel uncomfortable Physical contact (e.g., larking about) that makes you feel uncomfortable Subtly testing how far they can go in sexual matters Unwanted attempts to stroke or fondle you Display, use, or distribution of sexually tainted photos or films Minimally one of the above at least once Sexual coercion Subtly bribed into sexual intimacy with promise or reward Threatened with some sort of retaliation for not being sexually cooperative Unwanted attempts to have sex resulting in you pleading or struggling Implied faster promotions or better treatment if you were sexually cooperative Forcing sexual activity upon you that for some reason you cannot refuse Treated you badly for refusing to have sex Minimally one of the above at least once

21 37 34 21 26 12 16 27 8 20 5 11 64 1 1 1 0.2 0.3 2 3

Downloaded By: [Timmerman, Greetje][University of Groningen] At: 12:30 25 November 2010

MALE DOMINANCE AND SEXUAL HARASSMENT

725

tolerance of sexual harassment, support of the heterosexual norm, the denigration of women and of nonmasculine behaviour, the exclusion of women, and sexism. For the first six propositions respondents could indicate whether these applied to their own work environment. The first six propositions were as follows: (1) ‘‘If you are being harassed at work you take a risk when complaining about it’’, (2) ‘‘Perpetrators of sexual harassment are not likely to be punished at our work unit’’, (3) ‘‘My colleagues would prefer not to work together with homosexual colleagues’’, (4) ‘‘Most of my colleagues feel that gay men and women are not suitable for work as police officers’’, (5) ‘‘In our work unit we attribute more authority to male superiors than to female superiors’’, and (6) ‘‘In our work unit, women’s words carry less authority than men’s words’’. Answers could be given on a 5-point scale, with 1 ¼ ‘‘strongly disagree’’, 2 ¼ ‘‘disagree’’, 3 ¼ ‘‘neither disagree nor agree’’, 4 ¼ ‘‘agree’’, and 5 ¼ ‘‘strongly agree’’. For the following three propositions the respondents could indicate how often these applied to their own work environment: (7) ‘‘Colleagues joke about gay men and women’’, (8) ‘‘In my work unit people make indecent remarks about women’’, and (9) ‘‘In my work unit women are sexually harassed by their male colleagues’’. Answers could be given on a 5-point scale, with 1 ¼ ‘‘never’’, 2 ¼ ‘‘sometimes’’, 3 ¼ ‘‘regularly’’, 4 ¼ ‘‘often’’, and 5 ¼ ‘‘very often’’. The Scree plot of the Principal Component Analysis with oblique rotation that we conducted on our data revealed one factor that explained 39.9% of the variance. Cronbach’s alpha was .79. Item-total correlations ranged from .33 to .56. Mean scores of the nine propositions were used in the analyses.

Statistical analyses To test whether normative male dominance is a mediator between numerical male dominance and both sexual harassment scales, we conducted two series of regression analyses, for the two sexual harassment scales separately. First, we regressed numerical male dominance against sexual harassment. Second, the association between numerical and normative male dominance was tested. Third, we tested whether normative male dominance and sexual harassment were related. Fourth, we again tested the relation between numerical male dominance and sexual harassment while controlling for normative male dominance. If the relation between numerical male dominance and sexual harassment would be no longer significant, then normative male dominance may have been a mediator. If only the first three steps were met then normative male dominance may have partly mediated the relation (Baron & Kenny, 1986). In all regression analyses we controlled for age and partner status, because these variables are both related to power issues and sexual harassment.

726

DE HAAS AND TIMMERMAN

RESULTS

Downloaded By: [Timmerman, Greetje][University of Groningen] At: 12:30 25 November 2010

Descriptive statistics Sixty-four per cent of the women experienced one or more forms of sexual harassment once or more frequently. The most common types of harassment were offensive remarks about body or appearance, and crude and offensive sexual remarks made either publicly or privately (Table 2). For the police women, Pearson’s correlations between age, being single, numerical and normative male dominance, and sexual harassment are presented in Table 3. Also means and standard deviations for these variables are presented in Table 3. It appeared that unwanted sexual attention and advances, and sexual coercion were related. Both types of sexual harassment decreased with age. Furthermore, single women reported more sexual harassment than women with a partner. Numerical male dominance and normative male dominance were both positively related to unwanted sexual attention and advances. This indicated that women who had more contact with men in their daily work experienced more unwanted sexual attention and advances, and women who had less contact with men, experienced this less. In addition, women who perceived the work culture as dominated by men experienced more unwanted sexual attention and advances than women who perceived the work culture as less masculinized. Normative male dominance was also related to sexual coercion, whereas there was no significant relation between numerical male dominance and sexual coercion. Further, younger women worked more often in numerically maledominated workplaces and were more often single than older women. Single women perceived more normative male dominance than women with a

TABLE 3 Means and standard deviations (SD) for sexual harassment age, partner status, and numerical and normative male dominance, and Pearson’s correlations between these variables 1 1. Unwanted sexual attention and advances 2. Sexual coercion 3. Age 4. Being single 5. Numerical male dominance 6. Normative male dominance a

.46** 7.18** .20** .09* .47**

2

7.06* .13** .03 .24**

3

7.17** 7.11** 7.02

4

.05 .12**

5

.16**

Mean

SD

15.01a

4.44

6.06b 38.53 0.24 3.38 1.66

0.44 9.40 0.43 0.75 0.52

**p 5 .01, *p 5 .05. Based on the sum score of 12 items; minimum score was 12, maximum was 60. bBased on the sum score of 6 items; minimum score was 6, maximum was 36.

MALE DOMINANCE AND SEXUAL HARASSMENT

727

partner. Women who worked in numerically male-dominated workplaces also perceived more normative male dominance.

Downloaded By: [Timmerman, Greetje][University of Groningen] At: 12:30 25 November 2010

Normative male dominance as a mediator The aim of this study was to test whether normative male dominance mediates the relation between numerical male dominance and sexual harassment. Having controlled for age and partner status, it appeared that numerical male dominance was related to unwanted sexual attention and advances, albeit to a limited extent, but not to sexual coercion (see Step 1 in Table 4). In addition, when age and partner status were also controlled for, numerical and normative male dominance were associated to each other (see Step 2 in Table 4). Furthermore, normative male dominance was related to both unwanted sexual attention and advances, and sexual coercion (see Step 3 in Table 4). The relation between numerical male dominance and unwanted sexual attention and advances was no longer significant when normative male dominance was added as a control variable (see Step 4 in Table 4). This indicates that the relation between numerical male dominance and unwanted sexual attention and advances was fully mediated by normative male dominance. Because numerical male dominance was not related to sexual coercion, normative male dominance could not be a mediator in the model with sexual coercion. However, normative male dominance was related to sexual coercion. Adjusted R2 showed that 27% of the variance in unwanted sexual attention and advances was explained by age, partner status, and numerical and normative male dominance. This is a medium effect according the criteria of Cohen (1992). The effect on sexual coercion was smaller: 7% of the variance in sexual coercion was explained by age, partner status, and numerical and normative male dominance.

DISCUSSION Our hypothesis that the effect of numerical male dominance on sexual harassment is mediated by normative male dominance was partly supported by the results. It seems that numerically male-dominated workplaces propagate male-dominated cultural norms such as tolerance of sexual harassment, the denigration of women and of nonmasculine behaviour, the exclusion of women, and sexism, which in turn increase the risk of unwanted sexual attention and advances. So normative male dominance seems to be a mechanism that helps to explain why women are at greater risk of this type of sexual harassment in work situations where men outnumber women. Sexual coercion was related to normative male dominance but not to numerical male dominance.

728

70.08 1.22 3.84

70.08 1.22 3.84 0.02

Step 3 Age Being single Normative male dominance

Step 4 Age Being single Normative male dominance Numerical male dominance 0.01 0.28 0.23 0.16

0.01 0.28 0.22

0.01 0.31 0.18

SE B

7.17* .12** .46** .00

7.17** .12** .46**

7.15** .18** .07*

b

.27** (.00)

.27** (.20)

.07**

Adj. R2 (DR2)

0.00 0.09 0.20 0.01

0.00 0.09 0.20

0.00 0.13 0.01

B

0.00 0.03 0.02 0.02

0.00 0.03 0.02

0.00 0.03 0.02

SE B

7.05 .09* .23** .02

7.04 .09* .23**

7.04 .12** .02

b

Sexual coercion

.07** (.00)

.07** (.05)

.02**

Adj. R2 (DR2)

0.00 0.14 0.11

B

0.00 0.03 0.02

SE B

.02 .11** .16**

b

.04**

Adj. R2

Normative male dominance

**p 5 .01, *p 5 .05, B ¼ regression coefficient, SE B ¼ standard error, b ¼ standardized regression coefficient, Adj. R2 ¼ Adjusted R2.

70.07 1.81 0.44

Steps 1 & 2 Age Being single Numerical male dominance

B

Unwanted sexual attention and advances

TABLE 4 Normative male dominance as a mediator between numerical male dominance and sexual harassment

Downloaded By: [Timmerman, Greetje][University of Groningen] At: 12:30 25 November 2010

Downloaded By: [Timmerman, Greetje][University of Groningen] At: 12:30 25 November 2010

MALE DOMINANCE AND SEXUAL HARASSMENT

729

Our results are in line with earlier findings of a qualitative study in the insurance sales industry, which revealed that skewed sex ratios at work seemed to increase the likelihood of sexual harassment, because the presence of women in the workplace reinforced male hegemony and masculinity (Collinson & Collinson, 1996). In their theoretical paper, Wilson and Thompson (2001) followed a similar reasoning. They have suggested that when women enter the workplace, men fear loss of their power, which results in hostility towards women and an increased risk of sexual harassment. So sexual harassment is not purely a matter of women’s contact with men on the job, as was suggested by Gutek et al. (1990). The perception of what the presence of women on the workplace means for men is also important (Chamberlain, Crowley, Tope, & Hodson, 2008). Traditional power relations are usually not threatened by the entry of a few women. Harassment is most likely to take place when gender balance is not reached, but when a significant number of women enter the workplace (Timmerman, 2005).

Strengths and future research directions By studying normative male dominance as a mediator, our study helps to make clear why women are at greater risk of sexual harassment in workplaces that are numerically dominated by men, which has not been studied previously. Our study adds to the literature because our mediation model provides more insight in the effect of the multifaceted concept of male dominance on sexual harassment. Rather than studying both numerical and normative male dominance as predictors of sexual harassment, we study the relation between sexual harassment and male dominance from a more complex perspective. In addition, our study provides more insight into the risk factors of sexual harassment, which may be of great interest for the development of preventive efforts by policy-makers in this field. However, the relation between normative male dominance and sexual harassment could be biased because victims may perceive the climate differently due to their experience with sexual harassment. One group of researchers has addressed this issue by investigating whether women whose co-workers believe that their organization is tolerant of sexual harassment are more likely to be victimized. It was found that this workgroup estimate of organizational tolerance of sexual harassment is related to sexual harassment as well (Fitzgerald, Drasgow, et al., 1997). In addition, men’s perceptions of their manager with regard to sexual harassment were also found to be related to the incidence of sexual harassment of women (Pryor et al., 1995). Thus, some researchers have addressed the problematic issue of relying on victim’s perceptions for organizational variables, but to our knowledge normative male dominance has not previously been studied from an

Downloaded By: [Timmerman, Greetje][University of Groningen] At: 12:30 25 November 2010

730

DE HAAS AND TIMMERMAN

organizational perspective. However, normative male dominance can be considered an aspect of the cultural climate of the organization. Organizational climate can be viewed as the shared perception that co-workers have about the kinds of behaviours that are tolerated, extenuated, or accepted in their workgroup, which reflects the subjective mood or feeling of the work environment as a whole (Stockdale, Visio, & Batra, 1999). So, organizational climate refers to the outcome of aggregating individuals’ perception of their work environments (James et al., 2008). In future research the shared perceptions of employees working in the same work environment should be used to predict sexual harassment of individuals, and we need to study sexual harassment in multilevel theory terms. Without taking multilevel structure into account, standard errors will be underestimated, which may result in spurious significant correlations (Hox, 2002), which means that a relation could be mistakenly labelled as significant. Unfortunately, as we did not know which respondents were co-workers in the same work unit, we could not analyse the data with multilevel techniques. In addition, in future research the actual number of men and women in the workplace, and the amount of contact that women have with men, should be distinguished. Possibly, when a female officer is experiencing harassment from her male colleagues, she could try to avoid them and try to have more contact with her female colleagues. On the other hand, it is also possible that harassed women increase their efforts to fit in and to get along with their male colleagues. For example, a qualitative study in an insurance company revealed that harassed women did not try to bond with other women. Rather, they even tried harder to enter the male-dominated ingroup (Collinson & Collinson, 1996). Interviews of harassed women could shed more light on their coping strategies and whether these strategies include avoiding contact with men.

Limitations Because of the cross-sectional nature of our data, no claims about causality can be made. In future studies, prospective methods are necessary to shed more light on the direction of effects (O’Hare & O’Donohue, 1998). Moreover, the sample could be biased due to the low response rate. Because no systematic non-response data are available, reasons for the low response rate are unknown. Low response rate was possibly due to questionnaire overload. Respondents also experienced technical problems in getting the Internet questionnaire started. Although the response rate was low, the sample was representative for age and post. The representativeness of the sample with regard to the outcome measure of the study, sexual harassment, is also crucial. The low response rates could have biased the report of sexual harassment. In studies with low response rates reported,

Downloaded By: [Timmerman, Greetje][University of Groningen] At: 12:30 25 November 2010

MALE DOMINANCE AND SEXUAL HARASSMENT

731

prevalence of sexual harassment is usually higher than in studies with higher response rates (Gruber, 1990). However, the prevalence rate in our sample was comparable to other similar samples in which sexual harassment was measured with the same instrument. For example, In the British army, the prevalence rate was 68% (Rutherford, Schneider, & Walmsley, 2006). All in all, bias caused by the low response rate seems limited. Nevertheless, the results of our sample might not be representative of the population of Dutch police officers. In addition, the use of retrospective self-report measures may result in bias due to memory problems, because respondents have to think back over quite a lengthy period of time. The alternative for retrospective self-report measures would be observational methods. However, it is clear that such methods are not suitable for workplace harassment research. Therefore, selfreport instruments are the most likely measures of choice in studying sexual harassment (Arvey & Cavanaugh, 1995). Furthermore, there are two limitations as regards the measure of numerical male dominance. First, we used numerical male dominance as an interval variable in the regression analyses. This could have biased the results, because this variable was measured as an ordinal scale. Second, the validity of the item that measures this concept might not be sufficient. We operationalized numerical male dominance as the ratio of men and women in the daily work situation. Numerical male dominance was measured with one question: ‘‘Do you have more contact with male or with female colleagues in your daily work?’’ With this question respondents could indicate whether there are more women or more men in their work unit. However, it is also possible that respondents did not indicate the sex ratio in their work situation, but rather indicated with whom they interact. If many respondents did interpret this question like this, this item may not be valid. However, police officers’ possibilities to choose with whom they interact are limited. For example, officers are not free to choose with whom they patrol the streets. Rather they are placed in work units by the planning department. So, for our sample the frequency of males in a work situation and the number of males with whom they interact may be similar. Finally, it should be noticed that male dominance is not the only power issue at work in organizations. Other levels of power are organizational or hierarchical power and interpersonal or personal power. Power is a multifaceted phenomenon and all levels of power may be abused to sexually harass. Furthermore, sexual harassment is a complex problem and besides power differences between men and women other factors play an important role (Cleveland & Kerst, 1993). The goal of this study was not to provide a full explanation of sexual harassment. Rather, we aimed to test whether normative male dominance mediates the relation between numerical male dominance and sexual harassment.

732

DE HAAS AND TIMMERMAN

Downloaded By: [Timmerman, Greetje][University of Groningen] At: 12:30 25 November 2010

CONCLUSIONS In the Dutch police force numerical male dominance seemed to reinforce normative male dominance, which was in turn related to an increased risk of sexual harassment. This may suggest that sexual harassment could be used to accomplish or maintain masculinity as well as status differences between men and women. It seems that the maintenance of power differences between men and women is especially needed when women enter numerically male-dominated workplaces. As such, numerical and normative male dominance are important predictors of sexual harassment. This finding provides more insight in the effect of the multifaceted concept of male dominance on sexual harassment, which may help policy-makers in their efforts to prevent this workplace hazard.

REFERENCES Arvey, R. D., & Cavanaugh, M. A. (1995). Using surveys to assess the prevalence of sexual harassment: Some methodological problems. Journal of Social Issues, 51, 39–52. Baigent, D. (2005). Fitting in: The conflation of firefighting, male domination, and harassment. In J. E. Gruber & P. Morgan (Eds.), In the company of men: Male dominance and sexual harassment (pp. 45–64). York, PA: Maple Press. Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182. Cantisano, G. T., Domı´ nguez, J. F. M., & Depolo, M. (2008). Perceived sexual harassment at work: Meta-analysis and structural model of antecedents and consequences. Spanish Journal of Psychology, 11, 207–218. Chamberlain, L. J., Crowley, M., Tope, D., & Hodson, R. (2008). Sexual harassment in organizational context. Work and Occupations, 35, 262–295. Cleveland, J. N., & Kerst, M. E. (1993). Sexual harassment and perceptions of power: An underarticulated relationship. Journal of Vocational Behavior, 42, 49–67. Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155–159. Collinson, M., & Collinson, D. (1996). ‘‘It’s only Dick’’: The sexual harassment of women managers in insurance sales. Work, Employment, and Society, 10, 29–56. Dekker, I., & Barling, J. (1998). Personal and organizational predictors of workplace sexual harassment of women by men. Journal of Occupational Health Psychology, 3, 7–18. Fitzgerald, L. F., Drasgow, F., 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. Fitzgerald, L. F., Drasgow, F., & Magley, V. J. (1999). Sexual harassment in the armed forces: A test of an integrated model. Military Psychology, 11, 329–343. Fitzgerald, L. F., Gelfand, M. J., & Drasgow, F. (1995). Measuring sexual harassment: Theoretical and psychometric advances. Basic and Applied Psychology, 17, 425–445. Fitzgerald, L. F., Shullman, S., Bailey, N., Richards, M., Swecker, J., Ormerod, A. J., & Weitzman, L. (1988). The incidence and dimensions of sexual harassment in academia and the workplace. Journal of Vocational Behavior, 32, 152–175.

Downloaded By: [Timmerman, Greetje][University of Groningen] At: 12:30 25 November 2010

MALE DOMINANCE AND SEXUAL HARASSMENT

733

Fitzgerald, L. F., Swan, S., & Magley, V. J. (1997). But was it really sexual harassment? Legal, behavioral, and psychological definitions of workplace victimization of women. In W. O’Donohue (Ed.), Sexual harassment: Theory, research, and treatment (pp. 5–28). Boston: Allyn & Bacon. Gelfand, M. J., Fitzgerald, L. F., & Drasgow, F. (1995). The structure of sexual harassment: A confirmatory factor analysis across cultures and settings. Journal of Vocational Behavior, 47, 164–177. Gruber, J. E. (1990). Methodological problems and policy implications in sexual harassment research. Population Research and Policy Review, 9, 271–298. Gruber, J. E., & Morgan, P. (2005). In the company of men: Male dominance and sexual harassment. York, PA: Maple Press. Gutek, B. A., Cohen, A. G., & Konrad, A. M. (1990). Predicting social-sexual behavior at work: A contact hypothesis. Academy of Management Journal, 33, 560–577. Gutek, B. A., Murphy, R. O., & Douma, B. (2004). A review and critique of the sexual experiences questionnaire (SEQ). Law and Human Behavior, 28, 457–482. Harned, M. S., Ormerod, A. J., Palmieri, P. A., Collinsworth, L. L., & Reed, M. (2002). Sexual assault and other types of sexual harassment by workplace personnel: A comparison of antecedents and consequences. Journal of Occupational Health Psychology, 7, 174–188. Hox, J. (2002). Multilevel analysis: Techniques and applications. London: Lawrence Erlbaum Associates Ltd. Hulin, C. L., Fitzgerald, L. F., & Drasgow, F. (1996). Organizational influences on sexual harassment. In M. S. Stockdale (Ed.), Sexual harassment in the workplace (pp. 127–151). Thousand Oaks, CA: Sage Ilies, R., Hauserman, N., Schwochau, S., & Stibal, J. (2003). Reported incidence rates of workrelated sexual harassment in the United States: Using meta-analysis to explain reported rate disparities. Personnel Psychology, 56, 607–631. Jackson, R. A., & Newman, M. A. (2004). Sexual harassment in the federal workplace: Influences on sexual harassment by gender. Public Administration Review, 64, 705– 717. James, L. R., Choi, C. C., Emily Ko, C. H., McNeil, P. K., Minton, M. K., Wright, M. A., & Kim, K. (2008). Organizational and psychological climate: A review of theory and research. European Journal of Work and Organizational Psychology, 17, 5–32. Kauppinen-Toropainen, K., & Gruber, J. E. (1993). Antecedents and outcomes of womenunfriendly experiences: A study of Scandinavian, former Soviet, and American women. Psychology of Women Quarterly, 17, 431–456. MacKinnon, C. (1979). Sexual harassment of working women: A case of sex discrimination. New Haven, CT: Yale University Press. McKinney, K., & Maroules, N. (1991). Sexual harassment. In E. Grauerholz & M. A. Koralewski (Eds.), Sexual coercion: A sourcebook on its nature, causes, and prevention (pp. 29–41). Lexington, MA: Lexington Books. O’Donohue, W. (Ed.). (1997). Sexual harassment: Theory, research, and treatment. Boston: Allyn & Bacon. O’Donohue, W., Downs, K., & Yeater, E.A. (1998). Sexual harassment: A review of the literature. Aggression and Violent Behavior, 3, 111–128. O’Hare, E. A., & O’Donohue, W. (1998). Sexual harassment: Identifying risk factors. Archives of Sexual Behavior, 27, 561–580. Pryor, J. B., Giedd, J. L., & Williams, K. B. (1995). A Social Psychological model for predicting sexual harassment. Journal of Social Issues, 51, 69–84. Ragins, B. R., & Sundstrom, E. (1989). Gender and power in organizations: A longitudinal perspective. Psychological Bulletin, 105, 51–88.

Downloaded By: [Timmerman, Greetje][University of Groningen] At: 12:30 25 November 2010

734

DE HAAS AND TIMMERMAN

Rutherford, S., Schneider, R., & Walmsley, A. (2006). Quantitative and qualitative research into sexual harassment in the armed forces. Andover, UK: Schneider-Ross. Sandfort, T., & Vanwesenbeeck, I. (2000). Omgangsvormen, werkbeleving en diversiteit bij de Nederlandse politie [Sexual harassment and bullying, experience in work, and diversity in the Dutch police force]. Delft, The Netherlands: Eburon. Sbraga, T. P., & O’Donohue, W. (2000). Sexual harassment. Annual Review of Sex Research, 11, 258–285. Stockdale, M. S. (2005). The sexual harassment of men: Articulating the approach-rejection theory of sexual harassment In J. E. Gruber & P. Morgan (Eds.), In the company of men: Male dominance and sexual harassment (pp. 117–142). York, PA: Maple Press. Stockdale, M. S., Visio, M., & Batra, L. (1999). The sexual harassment of men: Evidence for a broader theory of sexual harassment and sex discrimination. Psychology, Public Policy, and Law, 5, 630–664. Timmerman, G. (2005). The impact of male domination on the prevalence of sexual harassment: An analysis of European Union surveys. In J. E. Gruber & P. Morgan (Eds.), In the company of men: Male dominance and sexual harassment (pp. 171–194). York, PA: Maple Press. Timmerman, G., & Bajema, C. (2000). The impact of organizational culture on perceptions and experiences of sexual harassment. Journal of Vocational Behavior, 57, 188–205. Welsh, S. (1993). He said, she said, the workplace made it happen: Organizational determinants of sexual harassment. Paper presented at the annual meeting of SASH, Miami, FL. Welsh, S. (1999). Gender and sexual harassment. Annual Review of Sociology, 15, 169–190. Willness, C. R., Steel, P., & Lee, K. (2007). A meta-analysis of the antecedents and consequences of workplace sexual harassment. Personnel Psychology, 60, 127–162. Wilson, F., & Thompson, P. (2001). Sexual harassment as an exercise of power. Gender, Work and Organization, 8, 61–83. Original manuscript received October 2008 Revised manuscript received June 2009 First published online March 2010