DISCRIMINATION IN EMPLOYMENT AND THE LAW

O R I G I N A L A R T I C L E

Gender Harassment: Broadening Our Understanding of Sex-Based Harassment at Work

Emily A. Leskinen • Lilia M. Cortina •

Dana B. Kabat

Published online: 27 July 2010

� American Psychology-Law Society/Division 41 of the American Psychological Association 2010

Abstract This study challenges the common legal and

organizational practice of privileging sexual advance forms

of sex-based harassment, while neglecting gender harass-

ment. Survey data came from women working in two male-

dominated contexts: the military and the legal profession.

Their responses to the Sexual Experiences Questionnaire

(SEQ) revealed five typical profiles of harassment: low

victimization, gender harassment, gender harassment with

unwanted sexual attention, moderate victimization, and

high victimization. The vast majority of harassment vic-

tims fell into one of the first two groups, which described

virtually no unwanted sexual advances. When compared to

non-victims, gender-harassed women showed significant

decrements in professional and psychological well-being.

These findings underscore the seriousness of gender

harassment, which merits greater attention by both law and

social science.

Keywords Gender harassment � Sexual harassment � Working women � Well-being

In 1964, Title VII of the Civil Rights Act codified the ille-

gality of sex discrimination. However, it took another

13 years for a federal appellate court to recognize sexual

harassment as a form of sex discrimination (e.g., Barnes v.

Costle, 1977). The prevailing view of sexual harassment,

both then and now, sees unwanted sexual attention as ‘‘the

quintessential harassment’’ (Schultz, 1998, p. 1710). Among

legal scholars, a notable group has emerged who contend

that this conceptualization is too narrow, arguing that

‘‘gender harassment’’ should be included in legal under-

standings of sex-based harassment (Epstein, 1998; Franke

1995, 1997, 2004; Growe, 2007; Hébert, 2005; Shultz, 1998,

2003, 2006). Gender harassment refers to ‘‘a form of hostile

environment harassment that appears to be motivated by

hostility toward individuals who violate gender ideals rather

than by desire for those who meet them’’ (Berdahl, 2007a,

p. 425). In this article, we lend empirical support to the

assertion that gender harassment is a serious form of sex

discrimination that deserves more attention. Using survey

data from two samples of working women, we demonstrate

that most sexual harassment in traditionally male domains

entails gender harassment in the absence of sexual advan-

ces; we also show how these experiences are associated with

negative personal and professional outcomes.

Central Constructs

In its broadest sense, sex-based harassment 1

refers to

‘‘behavior that derogates, demeans, or humiliates an indi-

vidual based on that individual’s sex’’ (Berdahl, 2007b,

E. A. Leskinen (&) � D. B. Kabat Department of Psychology, University of Michigan,

530 Church Street, Ann Arbor, MI 48109, USA

e-mail: [email protected]

L. M. Cortina

Departments of Psychology and Women’s Studies,

University of Michigan, Ann Arbor, MI 48109, USA

1 Psychology and the law have most commonly referred to this

phenomenon as ‘‘sexual harassment.’’ However, Berdahl (2007b)

makes a compelling case that ‘‘sex-based harassment’’ is a better

construct label, as it emphasizes sex (in the sense of femaleness or

maleness) rather than sexuality or sexual desire. We use both terms

interchangeably—‘‘sexual harassment’’ and ‘‘sex-based harass-

ment’’—to be consistent with the terminology of other literature but

also encourage revision of that terminology.

123

Law Hum Behav (2011) 35:25–39

DOI 10.1007/s10979-010-9241-5

p. 644). Subsumed under this umbrella term are (at least)

three related categories of behavior (e.g., Fitzgerald, Gelf-

and, & Drasgow, 1995; Fitzgerald, Swan, & Magley, 1997).

First, gender harassment refers to ‘‘a broad range of verbal

and nonverbal behaviors not aimed at sexual cooperation but

that convey insulting, hostile, and degrading attitudes about

women’’ (Fitzgerald et al., 1995, p. 430). Examples of

gender harassment include anti-female jokes, comments

that women do not belong in management, and crude terms

of address that denigrate women (e.g., referring to a cow-

orker as a ‘‘dumb slut’’). By contrast, unwanted sexual

attention involves expressions of romantic or sexual interest

that are unwelcome, unreciprocated, and offensive to the

recipient (e.g., unwanted touching, pressure for dates or

sexual behavior). The third category is sexual coercion:

bribes or threats that make the conditions of the victim’s

employment contingent on her sexual cooperation (e.g.,

offering a promotion in exchange for sexual favors, threat-

ening termination unless sexual demands are met).

Lim and Cortina (2005) elaborated on the relationships

among gender harassment, unwanted sexual attention, and

sexual coercion. They explained that ‘‘unwanted sexual

attention, as the name suggests, represents unwelcomed,

unreciprocated behaviors aimed at establishing some form

of sexual relationship. One could argue that sexual coercion

is a specific, severe, rare form of unwanted sexual attention,

involving similar sexual advances coupled with bribery or

threats to force acquiescence’’ (p. 484). In stark contrast,

gender harassment communicates hostility that is devoid of

sexual interest. Gender harassment can include sexually

crude terminology or displays (for instance, calling a col-

league a ‘‘cunt’’ or telling a sexually graphic joke about her),

but these behaviors differ from unwanted sexual attention in

that they aim to insult and reject women, not pull them into a

sexual relationship. In colloquial terms, the difference

between unwanted sexual attention/coercion versus gender

harassment is analogous to the difference between a ‘‘come

on’’ versus a ‘‘put down’’ (Fitzgerald et al., 1995).

Gender harassment, unwanted sexual attention, and

sexual coercion refer to behaviors rather than legal con-

structs. That said, sexual coercion is roughly parallel to what

the law calls quid pro quo harassment, whereas unwanted

sexual attention and gender harassment together map onto

the legal category of hostile environment harassment (e.g.,

Fitzgerald et al., 1995; Fitzgerald, Drasgow, Hulin, Gelfand,

& Magley, 1997). The next section elaborates on these legal

understandings of sexual harassment.

Legal Perspectives on Sexual Harassment

Title VII prohibits employers from discriminating against

any individual with regard to ‘‘compensation, terms,

conditions, or privileges of employment, because of such

individual’s…sex’’.2 When the federal courts first recog- nized sexual harassment as a Title VII violation, they were

acting on cases in which women had lost jobs for failing to

comply with their employers’ sexual demands, termed

‘‘quid pro quo harassment’’ (beginning with Barnes v.

Costle, 1977). It took another decade before the U.S.

Supreme Court would rule that ‘‘hostile work environment

harassment’’ could constitute unlawful sexual harassment.

In Meritor Savings Bank v. Vinson (1986), and later reaf-

firmed in Harris v Forklift Systems, Inc. (1993), the Court

described hostile environment sexual harassment as

occurring ‘‘[w]hen the workplace is permeated with ‘dis-

criminatory intimidation, ridicule, and insult’… that is ‘sufficiently severe or pervasive to alter the conditions of

the victim’s employment and create an abusive working

environment.’’ (Harris v Forklift Systems, Inc., 1993,

p. 21). Hostile environment sexual harassment is the pri-

mary focus of the current article.

According to both Meritor and Harris, to create a leg-

ally actionable hostile environment, the sexually harassing

conduct must be either severe or pervasive. In Harris, the

Court provided further guidance for determining whether a

hostile work environment is present: the harassing conduct

must pass both an objective test (a ‘‘reasonable’’ person

would find it hostile or abusive) and a subjective test (the

victim must have experienced it as abusive). Referring to

the objective test, it added that ‘‘whether an environment is

‘hostile’ or ‘abusive’ can be determined only by looking at

all the circumstances’’ (Harris v Forklift Systems, Inc.,

1993, p. 22).

Five years later, in Oncale v Sundowner Offshore Ser-

vices, Inc. (1998), the Supreme Court elaborated on what

should be considered when examining ‘‘all the circum-

stances,’’ explicitly mandating attention to the larger social

context. The Oncale decision also described types of

conduct that might be considered ‘‘severe.’’ Particularly

relevant to our article, it clearly stated that a motivation of

sexual desire is not a prerequisite for establishing objective

severity:

…harassing conduct need not be motivated by sexual desire to support an inference of discrimination on

the basis of sex. A Trier of fact might reasonably find

such discrimination, for example, if a female victim

is harassed in such sex-specific and derogatory

terms…as to make it clear that the harasser is moti- vated by general hostility to the presence of women

2 In addition, Title VII prohibits discrimination based on race, color,

religion, or national origin. Due to space constraints, this article only

details how Title VII has been interpreted to prohibit sexual

harassment.

26 Law Hum Behav (2011) 35:25–39

123

in the workplace. (Oncale v Sundowner Offshore

Services, Inc., 1998, p. 80).

Despite this last development in judicial interpretations of

Title VII, the prevailing legal conception of sexual harass-

ment remains a highly sexualized one, in which sexually

advancing or threatening conduct is seen as ‘‘the essence of

harassment’’ (Schultz, 1998, p. 1716). The harassing

behaviors alleged in Oncale, although recognized as moti-

vated by hostility rather than desire, still involved sexually

predatory behavior (e.g., sodomy with a bar of soap, threa-

tened rape). The Supreme Court has never clearly stated

whether the harassing conduct itself (as opposed to the

motivation for the conduct) must involve some form of

sexual advance to violate Title VII. Some appellate decisions

have rejected this requirement of sexualized content; for

example, in Williams v. General Motors Corp. (1999), the

Sixth Circuit Court of Appeals stated that: ‘‘harassing

behavior that is not sexually explicit but is directed at women

and motivated by discriminatory animus against women

satisfies the ‘based on sex’ requirement.’’ The federal bench,

however, is far from unanimity on this issue. In the same

Williams v GMC (1999) case, one judge filed a dissenting

opinion, arguing vehemently that ‘‘…Title VII does not proscribe ‘anti-female animus’ at all’’ and ‘‘the broad new

standard my colleagues have conjured here is not a correct

application of Title VII sex discrimination law presently on

the books.’’ Some courts routinely either dismiss hostile

environment cases that do not involve sexual conduct, or

they ‘‘disaggregate’’ sexual from nonsexual conduct and

then deem the latter to be irrelevant to a hostile environment

claim (see Franke, 2004, Growe, 2007, and Schultz, 2006 for

various post-Oncale case examples). Thus, the privileging of

the sexual advance in sexual harassment law continues.

To summarize, on many occasions the federal judiciary

has indicated, implicitly and explicitly, that offensive

behavior must reference sexuality to constitute unlawful

sex-based harassment. As a result, gender harassment

involving no sexual advances routinely gets neglected by

the law. This occurs even when the behavior fits all other

characteristics of a legally actionable hostile environment:

occurring ‘‘because of’’ the victim’s sex (interpreting

‘‘sex’’ to mean femaleness rather than sexuality); being

‘‘severe or pervasive’’ enough to adversely change the

conditions of her employment; and creating a work envi-

ronment that a ‘‘reasonable’’ person would find hostile or

abusive, and that the victim herself finds as such. Promi-

nent legal scholars have critiqued the exclusively sexual

view of sex-based harassment, arguing for instance that

‘‘most harassment is not designed to achieve sexual grati-

fication. Instead, it is used to preserve the sex segregation

of jobs by claiming the most highly rewarded forms of

work as masculine in composition and content’’ (Schultz,

2006, p. 22; see also Epstein, 1998; Franke, 1995, 1997,

2004; Growe, 2007; Shultz, 1998, 2003). At the same time

that these issues have been debated in law reviews, an

empirical literature on sexual harassment has developed in

psychology. How has psychological science made sense of

harassment based on sex and gender?

Psychological Research on Sexual Harassment

In psychology, researchers have examined lay perceptions

of sexual harassment more than any other aspect of sexual

harassment. These perceptions have differed over time,

between men and women, and across cultures (Cortina &

Berdahl, 2008; Rotundo, Nguyen, & Sackett, 2001). One

finding has not changed, however: when researchers have

compared perceptions of gender harassment versus

unwanted sexual attention/coercion, participants have

consistently rated the former as less severe, less offensive,

and less likely to represent what they see as ‘‘sexual

harassment’’ (e.g., Fitzgerald & Ormerod, 1991; Loredo,

Reid, & Deaux, 1995; Tang, Yik, Cheung, & Choi, 1995).

Quite separate from studies of sexual harassment per-

ceptions within the lay public have been surveys of actual

harassment experiences among working adults. As in the

law, much of this work has concentrated on sexually

advancing behaviors. For instance, the U.S. Merit Systems

Protection Board (USMSPB) surveyed federal employees

in 1980, 1987, and 1994 about their encounters with

‘‘sexual harassment,’’ defined as ‘‘uninvited and unwel-

come sexual attention and/or behavior’’ (USMSPB, 1994,

p. vi). Employees were asked to indicate the extent that

they had experienced a list of specific acts, virtually all of

which contained some form of sexual advance or sexual

threat (from unwanted touching to pressure for dates to

sexual assault). These surveys were well-executed and

have had a major impact on the field. However, they

neglected gender harassment.

Unlike the USMSPB, some sexual harassment

researchers routinely include questions about gender

harassment in their surveys. This is true, for example, of

the Sexual Experiences Questionnaire (SEQ), which has

become the most widely used and validated measure of

sexual harassment experiences (Fitzgerald et al., 1988;

Fitzgerald, Magley, Drasgow, & Waldo, 1999; Stark,

Chernyshenko, Lancaster, Drasgow, & Fitzgerald, 2002).

Some versions of the SEQ even assess two subtypes of

gender harassment—both ‘‘sexist’’ and ‘‘crude/offensive’’

behavior (Stark et al., 2002). When SEQ researchers have

divided harassment into its various subtypes, they have

found gender harassment to be the most common (e.g.,

Fitzgerald et al., 1988, 1999; Langhout et al., 2005;

Schneider, Swan, & Fitzgerald, 1997). Little SEQ research,

Law Hum Behav (2011) 35:25–39 27

123

however, has focused on experiences of gender harassment

in isolation from other behaviors. Studies of sexual

harassment prevalence and outcomes—using the SEQ and

other instruments—typically collapse across the different

facets of behavior for an overall measure of ‘‘offensive sex-

related experiences in the workplace’’ (Fitzgerald, Swan

et al., 1997, p. 9; for examples from other research pro-

grams, see Barling, Rogers, & Kelloway, 2001; Culbertson

& Rosenfeld, 1994; Richman, Shinsako, Rospenda, Flah-

erty, & Freels, 2002). Although this practice permits

examination of sexual harassment as a holistic phenome-

non, and has yielded many important findings, it obscures

the unique experience and impact of gender harassment.

In addition to survey research, some psychologists have

approached the study of sexual harassment from an

experimental perspective. This work has also been char-

acterized by an emphasis on sexual attention. For instance,

lab studies have operationalized men’s sexually harassing

behavior only as the sexual touching of women (Perry,

Kulik, & Schmidtke, 1998; Pryor, 1987) or as sexually

suggestive questioning (Woodzicka & LaFrance, 2005). In

these examples and others, there was virtually no attention

to gender harassment. 3

While gender harassment has been neglected in research,

this behavior has also almost certainly gone unreported in

organizations. Studies have found that victims who perceive

the harassment as more severe are more likely to report their

experiences to a superior (Bergman, Langhout, Palmieri,

Cortina, & Fitzgerald, 2002). However, research into lay

perceptions of sexual harassment (described above; e.g.,

Fitzgerald & Ormerod, 1991; Loredo et al., 1995; Tang

et al., 1995) suggests that gender harassment is widely

believed to be inconsequential, or somehow less important

than unwanted sexual attention in the workplace. Victims

should, therefore, be less likely to see gender harassment as

worthy of reporting, which means that organizational

authorities should be less likely to intervene (Langhout

et al., 2005). This makes it all the more imperative that

social science bring gender harassment to the fore, so that it

may be recognized as a legitimate and serious form of sex-

based discrimination in the workplace.

If It’s ‘‘Just’’ Gender Harassment, Why Should

We Care?

Because gender harassment has no explicit, sexually

predatory component to it (unlike unwanted sexual

attention or sexual coercion), it may seem less worthy of

scientific or legal scrutiny. However, past research on

everyday sexism has found that regular sexist interactions

decrease psychological well-being and predict symptoms

of psychological trauma (Berg, 2006; Swim, Hyers, Cohen,

& Ferguson, 2001). For example, Swim and colleagues

(2001) asked participants to keep track of instances of

ordinary sexist behavior (e.g., anti-female jokes, comments

reflecting gender stereotypes) observed or experienced in

any life setting. They found that these everyday sexist

encounters were associated with greater anger, anxiety, and

depression. To explain these negative outcomes, Swim and

colleagues (2001) argued that everyday sexism triggers

feelings of stereotype threat, defined as ‘‘being at risk of

confirming, as self-characteristic, a negative stereotype

about one’s group’’ (Steele & Aronson, 1995, p. 797).

Similar to everyday sexism, gender harassment may foster

stereotype threat in women, especially those working in

traditionally masculine domains. Gender harassment may

be used to cue women that they are inadequate, out of

place, and unable to perform at the level of men. The

associated experience of stereotype threat could set off a

cascade of negative outcomes in victims.

Research has demonstrated that sexual harassment is

linked with a wide range of victim outcomes (for recent

reviews, see Berdahl & Raver 2010; Cortina & Berdahl,

2008; Foote & Goodman-Delahunty, 2005). For instance,

studies have found that sexual harassment is associated

with decreased satisfaction with one’s job and professional

relationships, loss of productivity, and increased turnover

intentions and behaviors (e.g., Barling et al., 2001;

Langhout et al., 2005; Sims, Drasgow, & Fitzgerald, 2005).

Moreover, the consequences of sexual harassment are not

constrained to the job site. Victims also report lower psy-

chological well-being, more physical health problems, and

even symptoms of traumatic stress (e.g., Culbertson &

Rosenfeld, 1994; Fitzgerald, Swan et al., 1997; Richman,

Shinsako, Rospenda, Flaherty, & Freels, 2002). By and

large, however, this research has analyzed ‘‘sexual

harassment’’ as a global phenomenon, failing to differen-

tiate among the subtypes of sex-harassing behavior. It is,

therefore, impossible to know from this work whether

gender harassment by itself would have the same adverse

implications for employee well-being.

Hypotheses

In sum, legal scholars have developed compelling theories

about the importance of gender harassment, which we

sought to test using large-scale survey research. In line with

prior studies, we hypothesized that gender harassment,

without unwanted sexual attention or coercion, would be

3 Notable exceptions exist, however, in the experimental literature on

sexual harassment. For instance Schneider, Tomaka, and Palacios

(2001) studied sexual harassment in the laboratory by exposing

women to sexist comments—a clear form of gender harassment.

28 Law Hum Behav (2011) 35:25–39

123

the most common form of sex-based harassment that

women experience (Hypothesis 1). We also hypothesized

that women would report negative professional and per-

sonal outcomes, even when they ‘‘only’’ experience gender

harassment (Hypothesis 2). We tested these hypotheses

with survey data from women working in two male-dom-

inated domains: the U.S. Military (Study 1) and federal

legal practice (Study 2). Performing jobs that are highly

nontraditional for their gender, women in these domains

blur the boundaries between stereotypically ‘‘male’’ and

‘‘female’’ behavior. This makes them particularly vulner-

able to being scorned and rejected (e.g., gender harassed)

by colleagues who value rigid and clear distinctions

between the sexes (Berdahl, 2007a, b).

Study 1: The Military Survey

Participants and Procedure

Study 1 involved secondary analysis of survey data col-

lected by the U.S. Military. This survey began with a non-

proportional stratified, single stage random sample of

active-duty members from all branches of the U.S. Military

(Army, Navy, Air Force, Marines, and Coast Guard);

women and people of color were oversampled. The initial

sample contained 60,415 individuals, of whom 53,170

were deemed ‘‘eligible’’ (reasons for ineligibility were

various, such as inability to locate the sample member).

These individuals were invited to complete surveys either

on paper or online, and 19,960 usable surveys were

returned (38% response rate). The current study focused on

the 9,725 women who responded to the survey. Just over

one-half of these women identified as White (55%), one-

quarter as Black or African American (24%), and 11% as

Hispanic or Latina. Forty-eight percent of the respondents

reported some college, and 38% reported having at least a

4-year college degree. The number of years of active ser-

vice reported by members revealed a bi-modal distribution,

with 43% reporting less than 6 years and 33% reporting

10–20 years of active duty. For more information on this

sample and procedures, see Lipari and Lancaster (2003).

Measures

All participants completed the 2002 Department of Defense

Status of the Armed Forces Survey on Workplace and

Gender Relations.4 Descriptive statistics, coefficient

alphas, and intercorrelations for all variables analyzed in

Study 1 appear in Table 1. For multi-item scales, we

reverse-coded items as needed and then summed relevant

items to create scale-scores; higher scores reflect greater

levels of the underlying construct.

Sex-Based Harassment. To assess unwanted sex-

based experiences in the military, surveys contained an

updated version of the Sexual Experiences Questionnaire-

Department of Defense (SEQ-DoD) developed by Fitzgerald

and colleagues (1999; see also Stark et al., 2002).

Participants described how often over the prior 12 months

they had experienced various forms of unwanted, uninvited

‘‘sex/gender related talk and/or behavior’’ involving military

personnel, civilian employees, or contractors. They

responded on a 5-point scale ranging from 0 = never to

4 = very often. The measure consisted of 18 questions, in

which there are four subscales: (1) gender harassment: sexist,

(2) gender harassment: crude, (3) unwanted sexual attention,

and (4) sexual coercion.

The gender harassment: sexist subscale measured

treatment that conveys explicit antipathy toward members

of one gender. The subscale consisted of four items, such

as ‘‘made offensive sexist remarks (for example, suggest-

ing that people of your gender are not suited for the kind of

work you do)’’ and ‘‘referred to people of your gender in

insulting or offensive terms.’’ Four items also assessed

experiences of gender harassment: crude behavior;

although sexual on the surface, this behavior expresses

animosity rather than attraction. Examples included:

‘‘made offensive remarks about your appearance, body, or

sexual activities’’ and ‘‘made gestures or used body lan-

guage of a sexual nature that embarrassed or offended

you.’’ The unwanted sexual attention subscale consisted of

six items, including ‘‘made unwanted attempts to establish

a romantic relationship with you despite your efforts to

discourage it’’ and ‘‘touched you in a way that made you

feel uncomfortable.’’ The sexual coercion subscale con-

tained four items, e.g., ‘‘implied faster promotions or better

treatment if you were sexually cooperative’’ For more

detail on this measure, including evidence of its high

reliability and validity (see Fitzgerald et al., 1999; Stark

et al., 2002).

Psychological Well-Being. Psychological well-being

was measured by the short, 5-item version of the Mental

Health Inventory (MHI-5) (Viet & Ware, 1983). On a scale

from 1 (little or none of the time) to 4 (all or most of the

time), survey respondents rated how often they had

experienced various psychological states over the prior

4 weeks. Examples of psychological states included in this

measure are: ‘‘felt calm and peaceful’’ and ‘‘felt so down in

the dumps that nothing could cheer you up.’’ Researchers

have found this scale to be reliable when used in the

general population (Berwick et al., 1991).

4 This is a recurring survey; prior surveys took place in 1995 and

1988.

Law Hum Behav (2011) 35:25–39 29

123

Performance Decline Due to Poor Emotional

Health. In order to determine whether participants

experienced any difficulties during daily activities or

work as a result of mental health problems, surveys

included three questions adapted from the short-form

health survey (SF-36) used in the Medical Outcomes

Study (Ware & Sherbourne, 1992). Respondents rated how

often over the past 4 weeks had they experienced a

problem (such as ‘‘didn’t do work or other activities as

carefully as usual’’) with their work or other activities ‘‘as a

result of emotional problems (such as feeling depressed or

anxious).’’ Response options ranged from 1 (little or none

of the time) to 4 (all or most of the time).

Performance Decline Due to Poor Physical

Health. Surveys assessed health effects on work via

four items adapted from the SF-36 (Ware & Sherbourne,

1992). On a scale from 1 (little or none of the time) to 4 (all

or most of the time), respondents rated how often over the

past 4 weeks they had had problems with their work or

other daily activities ‘‘as a result of their physical health.’’

Examples of problems included: ‘‘were limited in the kind

of work or other activities you do’’ and ‘‘had difficulty

performing the work or other activities you do (for

example, it took extra effort).’’

General Health. Surveys asked respondents about

their general health by having them respond to four

statements on a scale of 1 (definitely false) to 4 (definitely

true). Examples of statements included ‘‘I seem to get sick

a little easier than other people’’ (reverse coded), and ‘‘my

health is excellent.’’ These questions were adapted from the

SF-36 (Ware & Sherbourne, 1992).

Work Attitudes. Three measures tapped work

attitudes. On a scale from 1 (strongly disagree) to 5

(strongly agree), six items assessed coworker satisfaction;

examples included: ‘‘you are satisfied with the

relationships you have with your coworkers’’ and ‘‘there

is very little conflict among your coworkers.’’ This scale

was adapted from multiple sources, two items being taken

from the Job Satisfaction Survey (Spector, 1997), three

items adapted from the 1995 Armed Forces Sexual

Harassment Survey (Edwards, Elig, Edwards, & Riemer,

1997), and one item created for this survey.

Six items adapted from the 1995 Armed Forces Sexual

Harassment Survey (Edwards et al., 1997) measured work

satisfaction. These items included statements such as ‘‘you

like the kind of work you do’’ and ‘‘your work makes good

use of your skills.’’ Organizational commitment was

assessed using a modified version of Mowday, Steers, and

Porter’s (1979) Organizational Commitment Question-

naire. This scale contained four items, including ‘‘you areT a

b le

1 D

e sc

ri p

ti v

e st

a ti

st ic

s, a lp

h a

c o

e ffi

c ie

n ts

, a n

d c o

rr e la

ti o

n s—

S a m

p le

1 (M

il it

a ry

)

V a ri

a b

le s

M S

D a

1 2

3 4

5 6

7 8

9 1

0 1

1

1 .

G e n

d e r

h a ra

ss m

e n

t: S

e x

is t

2 .1

7 3

.2 4

.7 1

–

2 .

G e n

d e r

h a ra

ss m

e n

t: C

ru d

e 1

.6 1

2 .8

4 .6

7 .7

1 –

3 .

U n

w a n

te d

se x

u a l

a tt

e n

ti o

n 0

.8 4

2 .1

7 .7

5 .4

7 .6

6 –

4 .

S e x

u a l

c o

e rc

io n

0 .2

3 1

.2 0

.7 9

.3 7

.4 7

.6 8

–

5 .

P sy

c h

o lo

g ic

a l

w e ll

-b e in

g 1

5 .9

8 3

.2 2

.8 4

- .2

8 -

.2 7

- .2

1 -

.1 6

6 .

P e rf

o rm

a n

c e

d e c li

n e

d u

e to

p o

o r

e m

o ti

o n

a l

h e a lt

h 3

.9 3

1 .7

6 .9

1 .2

2 .2

2 .2

1 .1

6 -

.5 9

–

7 .

G e n

e ra

l h

e a lt

h 1

3 .3

7 2

.2 9

.7 7

- .1

2 -

.1 3

- .1

1 -

.0 9

.3 3

- .2

5 –

8 .

P e rf

o rm

a n

c e

d e c li

n e

d u

e to

p o

o r

p h

y si

c a l

h e a lt

h 5

.1 9

2 .2

9 .8

9 .1

6 .1

6 .1

5 .1

4 -

.2 5

.3 1

- .4

0 –

9 .

C o

w o

rk e r

sa ti

sf a c ti

o n

2 1

.4 9

5 .0

5 .9

2 -

.3 1

- .2

9 -

.1 9

- .1

5 .3

2 -

.2 0

.1 1

- .1

1 –

1 0

. W

o rk

S a ti

sf a c ti

o n

2 1

.3 7

5 .9

4 .9

1 -

.2 3

- .2

0 -

.1 5

- .1

1 .3

6 -

.2 5

.1 6

- .1

5 .3

6 –

1 1

. O

rg a n

iz a ti

o n

a l

c o

m m

it m

e n

t 1

6 .1

0 2

.9 2

.8 2

- .1

3 -

.1 6

- .1

3 -

.1 0

.2 9

- .2

0 .1

8 -

.1 5

.2 7

.4 1

–

1 2

. T

u rn

o v

e r

in te

n ti

o n

s 2

.9 1

1 .7

4 .8

0 .1

7 .1

5 .0

9 .0

8 -

.2 3

.1 7

- .1

5 .1

5 -

.1 5

- .2

6 -

.2 7

N o

te .

S c a le

-s c o

re s

w e re

d e ri

v e d

b y

su m

m in

g re

sp o

n se

s a c ro

ss a ll

it e m

s in

e a c h

sc a le

fo r

a ll

w o

m e n

in th

e sa

m p

le ;

h ig

h e r

sc o

re s

re fl

e c te

d g

re a te

r le

v e ls

o f

th e

u n

d e rl

y in

g c o

n st

ru c t.

A ll

c o

rr e la

ti o

n s

a re

si g

n ifi

c a n

t, p \

.0 0

1

30 Law Hum Behav (2011) 35:25–39

123

willing to make sacrifices to help your Service’’ and ‘‘you

are glad that you are a part of your Service.’’

Turnover Intentions. In order to measure

respondents’ thoughts and intentions of leaving military

employment, five items were adapted from the 1999 Survey

of Active Duty Personnel Form A (1999 ADS). Using a

dichotomous yes/no scale, respondents indicated whether

over the prior 6 months they had, for example, ‘‘Thought

seriously about leaving the military’’ or ‘‘Discussed leaving

and/or civilian opportunities with family or friends.’’

Control Variables. We controlled for race, rank, and

service branch in all outcome analyses. Respondents self-

reported their race (coded 0 = minority and 1 = white).

They also provided their rank (i.e., paygrade) at the time of

the survey. Response options ranged from E-1 to E-9 for

enlisted personnel; from W-1 to W-5 for warrant officers;

and from O-1/O1E to O-6 or above for commissioned

officers. When releasing these data to the public, the

Defense Manpower Data Center (DMDC) collapsed these

choices into five ordered categories: 1 = E1–E4; 2 = E5–

E9; 3 = W1–W5; 4 = O1–O3; and 5 = O4–O6.

Participants self-reported their service branch as either

Army, Navy, Air Force, Marines, or Coast Guard.

Results

Profiles of Sex-Based Harassment

In order to test our first hypothesis, we used k-means cluster

analysis. k-means cluster analysis groups persons who are

similar on specified variables (see Hartigan, 1975, for more

information on this analytic approach). In the present study,

the k-means analysis classified women by the type and

amount of sex-based harassment they had experienced, as

indicated on the SEQ-DoD. Included in this analysis were

all women who reported experiencing at least one behavior

on the SEQ-DoD at least one time over the previous

12 months (n = 5,698). After standardizing these women’s

scores on the four subscales of the SEQ-DoD, we requested

two-, three-, four-, five-, and six-cluster solutions, and chose

the five-cluster solution for further analysis. We based this

decision on theoretical interest; we wanted to isolate women

who had experienced primarily gender harassment without

unwanted sexual attention or coercion. Profiles of means on

the z-scored SEQ-DoD scales appear in Fig. 1.

The largest group consisted of women who reported the

lowest levels of harassment (Group 1; n = 3,933). As seen

in Fig. 1, the experiences they described almost exclu-

sively consisted of sexist behavior. The second-largest

group (n = 1,161) contained women who had encountered

both subtypes of gender harassment—sexist and crude—

but very little unwanted sexual attention or sexual coercion

(henceforth, this group will be referred to as Group 2, the

‘‘Gender Harassment’’ group). Group 3 (n = 429) dis-

closed episodes of unwanted sexual attention, in addition to

moderate levels of sexist and crude behavior. Group 4

(n = 138) had encountered moderate levels of all types of

harassment measured by the SEQ-DoD. Group 5 (n = 37),

the smallest profile group, reported the most frequent

harassment on all four subscales. In sum, 89.4% of

harassment victims fell into Group 1 or 2, which described

experiences of gender harassment but virtually no unwan-

ted sexual attention or coercion. This pattern of

victimization provides strong support for Hypothesis 1,

which had predicted gender harassment (in the absence of

sexual attention or coercion) to be the most common

manifestation of sex-based harassment.

-2

0

2

4

6

8

10

12

14

Gender Harassment:

Sexist

Gender Harassment:

Crude

Unwanted Sexual

Attention

Sexual Coercion

Z -S

c o

re s

High Victimization (0.6%) - Group 5

Moderate Victimization (2.4%) - Group 4

Gender Harassment and Unwanted Sexual Attention (7.5%) - Group 3

Gender Harassment (20.4%) - Group 2

Low Victimization (69%) - Group 1

Fig. 1 Profiles of harassment among military women

(n = 5,698)

Law Hum Behav (2011) 35:25–39 31

123

In addition to the five groups revealed through cluster

analysis, we identified a sixth ‘‘Nonvictims’’ group, which

contained all women who had responded ‘‘never’’ to all

items of the SEQ-DoD. This group (n = 4,014) served as a

comparison group in subsequent analyses.

Table 2 reports demographic information for each of

the profile groups. Using chi-square and ANOVA analy-

ses, we found significant relationships between profile

membership and race, v2 (5, N = 9711) = 55.34, p \ .001; rank, v2 (10, N = 9704) = 159.00, p \ .001; years served, F (5, 9664) = 81.90, p \ .001; education level, F (5, 9621) = 32.14, p \ .001; and service branch, v2 (20, N = 9712) = 246.04, p \ .001. As seen in Table 2, the racial make-up of each profile group generally mirrored

the overall sample. However, white respondents were

slightly more likely to be in the Nonvictims Group, Group

1 (Low Victimization), or Group 2 (Gender Harassment).

Ethnic minority respondents were more likely to appear in

Group 5 (High Victimization). Regarding military rank

and group membership, commissioned officers were dis-

proportionately likely to appear in Group 1 (Low

Victimization), whereas enlisted personnel were dispro-

portionately represented in Group 3 (Gender Harassment

with Unwanted Sexual Attention). In terms of tenure (i.e.,

years served) in the military, Nonvictims differed from all

other groups by having served the greatest average number

of years. Educational differences were most pronounced

for the Nonvictims and Group 1 (Low Victimization), who

had significantly higher educational levels than other

groups.

Outcomes of Sex-Based Harassment

For theoretical reasons, we were most interested in out-

comes for the group that had experienced primarily gender

harassment (Group 2: the Gender Harassment group 5 ),

which we compared to outcomes for women in the Non-

victims group. A comparison of these two groups addresses

the question of whether experiences of ‘‘just’’ gender

harassment are associated with adverse consequences.

In order to equalize cell sizes for this analysis, we ran-

domly selected 1,000 women from the Gender Harassment

group and 1,000 women from the Nonvictims group. Using

multivariate analysis of covariance (MANCOVA), we then

compared these two groups on the mental health, physical

health, and organizational outcome variables. Covariates in

this analysis were racial minority status, military rank, and

service branch (to account for factors that could potentially

affect harassment risk). We found a significant multivariate

main effect of gender harassment on the collection of

outcomes, Wilks’ Lambda = .81, F (8, 1822) = 52.06,

p \ .001. As Table 3 shows, women in the Gender Harassment

group scored significantly lower than the Nonvictimized

women on all work attitudes (work satisfaction, coworker

satisfaction, and organizational commitment). They also

reported greater performance decline due to both physical

and emotional health, and they described less overall

psychological well-being and health satisfaction. Further-

more, women in the Gender Harassment group disclosed

greater thoughts and intentions of leaving their jobs.

Cohen’s effect sizes (d) ranged from .22 to .79, averaging

Table 2 Descriptive statistics for profile groups—Sample 1 (Military)

Profile group Non-

White

percent

Rank percent Years

served

Education Branch of service (%)

Enlisted Warrant

officer

Commissioned

officer

Mean SD Mean SD Army Navy Marine corps

Air

Force

Coast

guard

Nonvictims (n = 4,014) 47.0 69.8 3.6 26.5 2.28 1.07 2.21 0.70 26.6 21.1 11.0 34.4 6.9

Group 1: Low Victimization

(n = 3,933) 42.1 68.2 3.3 28.4 2.07 1.06 2.21 0.70 28.9 21.4 13.8 26.8 9.1

Group 2: Gender Harassment

(n = 1,161) 39.8 75.3 3.6 20.9 1.82 1.02 2.11 0.69 27.6 24.4 17.5 19.8 10.7

Group 3: Gender Harassment

& Unwanted Sexual

Attention (n = 429)

51.3 91.6 0.9 7.5 1.47 0.81 1.87 0.63 34.5 24.7 14.7 15.4 10.7

Group 4: Moderate

Victimization (n = 138) 55.8 92.0 0.0 8.0 1.50 0.83 1.83 0.70 39.1 22.5 21.7 9.4 7.2

Group 5: High Victimization

(all types) (n = 37) 70.3 91.9 0.0 8.1 1.65 0.95 1.83 0.57 56.8 2.7 29.7 5.4 5.4

5 We did not include Group 1 (the ‘‘low victimization’’ group) in

outcome analyses, even though their experiences largely consisted of

gender harassment, for two primary reasons. First, it is unlikely that

the extremely low rates of harassing behavior described by this group

would be seen as ‘‘sufficiently severe or pervasive’’ to be actionable

under Title VII. In addition, we hope to avoid the criticism that we are

‘‘making mountains out of molehills’’ by foregrounding conduct that,

while offensive, is transient and rare.

32 Law Hum Behav (2011) 35:25–39

123

.48; the largest differences emerged for coworker satis-

faction (.79), psychological well-being (.67), and work

satisfaction (.58). These findings provided strong support

for Hypothesis 2, which had predicted that experiences of

gender harassment (alone) would be associated with neg-

ative outcomes.

To provide an additional point of comparison, we also

combined Groups 3, 4, and 5 into a ‘‘Sexual Advance

Harassment’’ group (n = 604). All of these women, unlike

those in Groups 1 or 2, reported significant experiences of

unwanted sexual advances at work, in the form of sexual

attention and/or sexual coercion. We then conducted a

second MANCOVA, comparing the outcomes of this Sex-

ual Advance Harassment group to those of the Gender

Harassment group and the Nonvictims group. There was a

significant multivariate main effect of harassment-group-

membership on the collection of outcomes, Wilks’

Lambda = .77, F (16, 4748) = 40.82, p \ .001. As Table 3 shows, outcome means for the Gender Harassment

Victims fell in between those for the Nonvictims and Sexual

Advance Victims (although closer to the means of the latter

group). According to follow-up Tukey tests, all outcomes

for the gender-harassed women were significantly worse

than those of the nonvictimized women. Moreover, Sexual

Advance Victims showed significantly worse outcomes

than Gender Harassment Victims, with two exceptions:

there were no significant differences between the two

groups on work satisfaction and turnover intentions.

Study 2: The Attorney Survey

Participants and Procedure

We sought to cross-validate the results from Study 1 with

secondary analysis of data collected from women working

in a very different context: the legal profession. Participants

were drawn from a stratified random sample of attorneys

from a large federal judicial circuit. Surveys were sent to

9,223 individuals, yielding a 53% response rate. The current

study focused on the 1,425 women who responded to the

survey. Most of these women were white (93%). They

ranged in age from 24 to 79 years (M = 39.09, SD = 7.81).

Sixty-eight percent of the women were married or part-

nered, while 18% had never been married. These women

were highly educated, all holding at least a Juris Doctor, and

some holding additional graduate degrees. Similar to the

military sample, they worked in a traditionally masculine

occupation where women remain a minority. They com-

pleted a paper-and-pencil self-report survey. More

information about these participants and procedures appears

in Cortina et al. (2002) and Lim and Cortina (2005). 6

Measures

Table 4 displays the descriptive statistics, coefficient

alphas, and intercorrelations for all variables from Study 2.

Similar to Study 1, for multi-item scales, we reverse-coded

items as needed and then summed relevant items to create

scale-scores; higher scores reflect greater levels of the

underlying construct.

Sex-Based Harassment. Similar to Study 1, we used

items from the SEQ developed by Fitzgerald et al. (1995,

1988) to assess sex-based harassment. Participants

described how often they had experienced a list of

unwanted sexual or sexist behaviors over the past 5 years

Table 3 ANCOVAs for physical health, mental health, and work outcomes—Sample 1 (Military)

Outcome measured Nonvictims Gender Harassment

Victims

Sexual Advance

Harassment Victims

df F p

Mean (SD) 95% CI Mean (SD) 95% CI Mean (SD) 95% CI

1. Psychological well-being 16.76 (2.92) 16.58, 16.94 14.65 (3.43) 14.43, 14.86 13.62 (3.93) 13.31, 13.94 2, 2550 122.07 \.001 2. Performance decline due to

poor emotional health

3.63 (1.49) 3.54, 3.73 4.47 (2.07) 4.34, 4.60 5.14 (2.51) 4.94, 5.34 2, 2562 75.91 \.001

3. Health satisfaction 13.60 (2.10) 13.47, 13.73 13.04 (2.49) 12.89, 13.20 12.53 (2.62) 12.31, 12.74 2, 2550 24.06 \.001 4. Performance decline due to

poor physical health

4.91 (2.09) 4.78, 5.04 5.62 (2.56) 5.46, 5.78 6.40 (3.16) 6.15, 6.66 2, 2555 37.90 \.001

5. Coworker satisfaction 22.96 (4.50) 22.68, 23.24 18.89 (5.43) 18.55, 19.23 18.08 (5.73) 17.61, 18.54 2, 2532 169.93 \.001 6. Work satisfaction 22.73 (5.44) 22.39, 23.07 18.96 (6.33) 18.56, 19.35 18.54 (6.62) 18.01, 19.07 2, 2554 89.16 \.001 7. Organizational commitment 16.53 (2.66) 16.36, 16.70 15.49 (3.28) 15.28, 15.69 14.78 (3.35) 14.51, 15.05 2, 2555 29.15 \.001 8. Turnover intentions 2.69 (1.80) 2.57, 2.80 3.44 (1.54) 3.35, 3.54 3.46 (1.58) 3.33, 3.58 2, 2558 53.49 \.001

6 Although drawing on the same larger dataset, these two past studies

have different foci from each other and from the current article.

Cortina et al. (2002) focused on gender differences in experiences of

incivility and harassment in the legal profession. Lim and Cortina

(2005) detailed how uncivil and harassing behaviors tend to co-occur

and jointly affect personal and professional outcomes.

Law Hum Behav (2011) 35:25–39 33

123

from judges, attorneys, trustees, marshals, court security

officers, and court personnel. They responded on a 5-point

scale ranging from 0 = never to 4 = many times. The

measure consisted of nine questions, from which there are

three subscales: gender harassment, unwanted sexual

attention, and sexual coercion.

The gender harassment subscale consisted of two items:

‘‘made offensive remarks or jokes about women in your

presence?’’ and ‘‘publicly addressed you in unprofessional

terms (e.g., ‘honey,’ ‘dear’)?’’ 7

The unwanted sexual

attention subscale contained four items, including:

‘‘attempted to establish a romantic or sexual relationship

despite your efforts to discourage it?’’ Two items com-

prised the sexual coercion subscale, such as, ‘‘implied

more favorable treatment of you or your client if you were

sexually cooperative?’’

Job-Related Outcomes. We used a three-item scale

(a = .74), developed for the purposes of this survey, to measure attorneys’ intentions to change careers (e.g., ‘‘I

often think about leaving federal litigation’’). We measured

job stress with three items (a = .75), such as ‘‘my expe- riences working in the federal court are more stressful than

I’d like.’’ Finally, in order to assess professional relationship

satisfaction, we used a three-item scale (a = .76) that consisted of items such as ‘‘in general, I am satisfied with my

professional relationships with other attorneys in federal

court.’’ For all three scales, response options ranged from 1

(strongly disagree) to 5 (strongly agree).

Control Variables. We controlled for race and job

tenure in outcome analyses. Respondents self-reported

their race (coded 0 = minority and 1 = white). They also

reported the number of years they had actively practiced

law, including clerkships. Response options were 1

(0–5 years of practice), 2 (6–10 years of practice), 3 (11–

20 years of practice), 4 (21–30 years of practice), and 5

(31 ? years of practice); we collapsed response options 4

and 5 due to the small number of women who chose the

latter option.

Results

Profiles of Sex-Based Harassment

Similar to Study 1, we standardized the attorney women’s

scores on the subscales of the SEQ, including all women

who had reported at least one experience of an SEQ

behavior (n = 491). We again performed a k-means cluster

analysis of these subscales, and again chose a 5-cluster

solution. Profiles of attorney women’s means on the z-

scored SEQ appear in Fig. 2.

Mirroring the profiles from the military sample, the

largest group consisted of women who reported minimal

experiences of harassment (Group 1; n = 320). Those who

disclosed high levels of gender harassment, with almost no

unwanted sexual attention or sexual coercion, made up the

second-largest group (Group 2; n = 134). Group 3

(n = 30) described episodes of unwanted sexual attention,

in addition to moderate levels gender harassment. Group 4

(n = 3) and Group 5 (n = 4) were quite small, and dis-

closed moderate and high levels of all types of harassment

measured by the SEQ, respectively. To summarize, over

90% of harassed women fell into one of the two groups

reporting little or no sexually advancing harassment

experiences (Group 1 or 2), further supporting Hypothesis

1. We also identified a group of Nonvictims (n = 338),

who had responded ‘‘never’’ to all items of the SEQ.

Demographic information for each of the profile groups

appears in Table 5. Using chi-square and F tests, we found

no significant relationships between profile membership

and race, v2 (5, N = 1317) = 4.24, p = .52; age, F (37, 1286) = 0.79, p = .81; or years practicing law F (5,

1333) = 0.98, p = .43.

Table 4 Descriptive statistics, alpha coefficients, and correlations—Sample 2 (Attorneys)

Variables M SD a 1 2 3 4 5

1. Gender harassment 1.74 2.14 .69 –

2. Unwanted sexual attention 0.19 0.79 .74 .42** –

3. Sexual coercion 0.01 0.12 .84 .14** .41** –

4. Intention to change careers 6.00 2.93 .74 .02 .03 .00 –

5. Professional relationship satisfaction 12.04 2.24 .76 -.06* -.04 -.02 -.36** –

6. Job stress 9.17 2.74 .75 .15** .04 .06* .32** -.34**

Note. Scale-scores were derived by summing responses across all items in each scale for all women in the sample; higher scores reflected greater levels of the underlying construct

** p \ .01, * p \ .05

7 Unlike the military survey in Study 1, the brevity of this subscale

precluded distinctions between ‘‘sexist’’ and ‘‘crude’’ gender harass-

ment. Together, the two items assessed ‘‘gender harassment’’ as a

global phenomenon.

34 Law Hum Behav (2011) 35:25–39

123

Outcomes of Sex-Based Harassment

Again, for theoretical reasons, the group that experienced

high gender harassment (without sexual attention or coer-

cion; Group 2, n = 134) was the focus of our primary

outcome analysis. To serve as a comparison group of sim-

ilar size, we randomly selected 150 women from the

Nonvictims group. We then compared these 150 Nonvic-

tims to the 134 Gender Harassment victims on the three job-

related outcomes, using MANCOVA. Racial minority sta-

tus and years practicing law served as covariates. Again,

results suggested that there was a significant multivariate

main effect of gender harassment on the collection of out-

comes, Wilks’ Lambda = .96, F (3, 238) = 3.00, p \ .05. Follow-up univariate analyses (ANCOVAs), reported in

Table 6, revealed significant effects on job stress and sat-

isfaction with professional relationships. Compared to their

non-harassed counterparts, gender-harassed women repor-

ted significantly higher levels of job stress (d = .34). They

also described less satisfaction with their relationships with

federal judges, other attorneys in the federal court, and

court personnel (d = .32). However, we did not find a

significant group difference in intention to change careers

(the means for both groups were similarly low: close to six,

on a scale that can range from 3 to 15). With the exception

of this last result, Hypothesis 2 was supported among

women attorneys.

To provide additional insight into group differences, we

again combined Groups 3, 4, and 5 into a Sexual Advance

Harassment group (n = 37); their outcome means appear in

Table 6. We compared the outcomes of this group with

those of the Nonvictims and Gender Harassment Victims

(using MANCOVA), finding a significant multivariate main

effect of harassment on outcomes, Wilks’ Lambda = .95, F

(6, 546) = 2.29, p \ .05. Follow-up Tukey tests indicated that the Sexual Advance group differed significantly from

the Nonvictims group in terms of professional relationship

satisfaction. The Gender Harassment Victims also reported

less professional relationship satisfaction, and more job

stress, than Nonvictims. Gender Harassment Victims did

not differ significantly from Sexual Advance Victims,

however, on any outcome. Put differently, we found that

gender-harassed women attorneys fared just as poorly as

those who had experienced sexual advance harassment.

General Discussion

This article draws attention to the incidence and correlates

of gender harassment in the workplace. Social science

-2

0

2

4

6

8

10

12

14

Gender Harassment

Unwanted Sexual Attention

Sexual Coercion

Z -S

c o

re s

High Victimization (0.8%) - Group 5

Moderate Victimization (0.6%) - Group 4

Gender harassment and Unwanted Sexual Attention (6.1%) - Group 3

Gender Harassment (27.3%) - Group 2

Low Victimization (65.2%) - Group 1

Fig. 2 Profiles of harassment among women attorneys

(n = 491)

Table 5 Descriptive statistics for profile groups—Sample 2

(Attorneys)

Profile group Non-White Percent Age Years practicing law

M SD M SD

Nonvictims (n = 851) 5.2 39.37 8.09 2.23 0.91

Group 1: Low Victimization (n = 320) 6.7 38.32 7.31 2.21 0.90

Group 2: Gender Harassment (n = 134) 4.5 39.45 7.36 2.37 0.87

Group 3: Gender Harassment & Unwanted

Sexual Attention (n = 30) 6.7 37.63 6.13 2.23 0.77

Group 4: Moderate Victimization (n = 3) 0.0 42.50 0.71 2.67 0.58

Group 5: High Victimization (all types) (n = 4) 25.0 35.00 2.31 1.75 0.50

Law Hum Behav (2011) 35:25–39 35

123

research has often analyzed sex-based harassment as a

global phenomenon, failing to distinguish among the dif-

ferent facets of behavior; this practice may have obscured

the experiences of many harassed women, especially those

working in male-dominated fields. At the same time, many

federal judges have evaluated potentially harassing conduct

through a (hetero)sexualized lens, in which they privilege a

‘‘top-down, male-female sexual come-on image of

harassment’’ (Schultz, 2006, p. 26). This ‘‘sexual model of

sexual harassment’’ does not provide an explanation for

gender harassment that is devoid of sexual interest, which

we found to be the norm in women’s experiences of

harassment in traditionally masculine domains.

Key Findings

The first goal of this article was to investigate the preva-

lence of different dimensions of sexually harassing

conduct. Consistent with our hypothesis, gender harass-

ment in the absence of unwanted sexual attention or sexual

coercion was the most common manifestation of harass-

ment faced by women in the military and the law

(employment contexts which, importantly, were once the

exclusive province of men). In fact, in both settings, 9 out

of every 10 victims had experienced primarily gender

harassment, with virtually no unwanted sexual overtures.

Taken together, our empirical results support the legal

theory that ‘‘much of the time, harassment assumes a form

that has little or nothing to do with sexuality but everything

to do with gender’’ (Schultz, 1998, p. 1687). This conduct

is not about misguided attempts to draw women into sexual

relationships; quite the contrary, it rejects women and

attempts to drive them out of jobs where they are seen to

have no place. One could argue that, in these instances,

‘‘sexual harassment is used both to police and discipline the

gender outlaw: the woman who dares to do a man’s job is

made to pay’’ (Franke, 1997, p. 764). Had we collapsed

across the subtypes of harassing behavior, as many psy-

chologists do, this striking pattern of results would not have

surfaced.

The second goal of this article was to understand the

correlates of gender harassment for working women. When

comparing victims of gender harassment to women who

reported no harassment experiences, we found that ‘‘just’’

gender harassment was associated with multiple negative

outcomes. Specifically, in the military context, gender-

harassed women reported lower psychological well-being,

job performance, job commitment, and satisfaction with

their employment and health; they also described more

thoughts and intentions of leaving their jobs. These results

remained significant even after controlling for the women’s

race, rank, and service branch. Among attorneys, gender-

harassed women (compared to nonharassed women)

reported lower satisfaction with professional relationships

and higher job stress, above and beyond the effects of race

and job tenure. Thus, experiences of gender harassment

alone were associated with negative personal and profes-

sional outcomes in two very different contexts of work.

Implications for Sexual Harassment Jurisprudence

This research has important legal implications. Although

popular wisdom might suggest that the legal definition of

sexual harassment is fixed, in actuality the legal under-

standing changes as courts interpret and refine precedent.

Our research underscores the need to broaden legal and

scientific conceptualizations of sexual harassment, so that

gender harassment can be recognized as a harmful and

objectionable condition of employment, even when not

paired with unwanted sexual attention. Title VII of the

Civil Rights Act of 1964 says nothing about sexual

behavior. Definitions that limit sex-based harassment to

unwanted advances emerged over time as the courts

revised their interpretations of Title VII. Our results sug-

gest that further revision is in order, to prohibit not just

sexually predatory conduct, but also behavior that creates a

hostile work environment for members of one sex but

contains no sexual advance—that is, gender harassment. As

we have shown, gender harassment does not simply pro-

vide a backdrop for other kinds of harassment; it is the

modal form of sex-based harassment faced by women at

work (at least in male-dominated domains). Moreover, it

alters the terms of employment for targeted women, being

associated with a variety of negative professional

outcomes.

Table 6 ANCOVAs for job-related outcomes—Sample 2 (Attorneys)

Outcome measured Nonvictims Gender Harassment

Victims

Sexual Advance

Harassment Victims

df F p

Mean (SD) 95% CI Mean (SD) 95% CI Mean (SD) 95% CI

1. Intention to change careers 5.81 (3.01) 5.32, 6.31 5.95 (2.93) 5.45, 6.46 6.51 (3.25) 5.43, 7.60 2, 307 0.98 .452

2. Professional relationship satisfaction 12.51 (2.14) 12.15, 12.86 11.81 (2.21) 11.43, 12.18 11.38 (2.97) 10.39, 12.37 2, 308 5.25 .006

3. Job stress 9.04 (2.84) 8.53, 9.55 10.01 (2.82) 9.51, 10.50 9.76 (3.67) 8.48, 11.04 2, 277 3.61 .032

36 Law Hum Behav (2011) 35:25–39

123

Our outcome results suggest that harassment exclusively

consisting of gender-related hostility has adverse work-

related correlates. That is, the more that women experi-

enced gender harassment, the less satisfied they were with

their jobs and colleagues, the more they experienced stress

on the job, and the more they suffered health problems that

detracted from their job performance. Moreover, these

results were not trivial in magnitude, being associated with

large effect sizes in some cases (particularly for coworker

satisfaction and psychological well-being). Findings such

as these could be relevant to legal claims of hostile envi-

ronment sexual harassment.

As noted at the outset of this article, in Harris v Forklift

Systems, Inc. (1993) the Supreme Court stated that ‘‘all the

circumstances’’ must be considered when determining

whether an environment is ‘‘hostile’’ or ‘‘abusive,’’ in

violation of Title VII. The Court went on to say that these

circumstances may include a number of factors, including

whether the discriminatory conduct ‘‘unreasonably inter-

feres with an employee’s work performance’’ Harris v

Forklift Systems, Inc. (1993, p. 23). In the current study, we

documented that gender harassment on its own is linked

with a variety of adverse work outcomes, including but not

limited to performance decline. These outcomes do not

necessarily reflect traumatization or incapacitation of the

victim, but this is not a requirement of Title VII: ‘‘…Title VII comes into play before the harassing conduct leads to a

nervous breakdown. A discriminatorily abusive work

environment, even one that does not seriously affect

employees’ psychological well-being, can and often will

detract from employees’ job performance, discourage

employees from remaining on the job, or keep them from

advancing in their careers’’ (Harris v Forklift Systems, Inc.,

1993, p. 21). Although we cannot draw definitive causal

conclusions from our correlational findings, our results are

consistent with these sorts of effects. They support the

possibility that ‘‘just’’ gender harassment can create a

hostile environment that disadvantages women.

Limitations

As with any research, our studies have their limitations. All

results were based on cross-sectional, correlational data,

precluding strong temporal or causal inferences. That said,

longitudinal studies of sexual harassment (e.g., Glomb,

Munson, Hulin, Berman, & Drasgow, 1999; Sims et al.,

2005) provide compelling evidence that our personal and

professional outcomes follow, rather than precede,

harassment experiences. These data were self-reported;

because of this, common method variance or response set

could potentially explain some of the significant findings.

Surveys were designed to minimize some of these prob-

lems: questions about mental health, physical health, and

job attitudes were asked prior to and independently of the

SEQ, so that responses about harassment did not bias

reports of health and attitudes. Also, we were only able to

use proxies for job and career turnover, with measures of

turnover intentions rather than actual turnover rates.

However, past research tells us that one of the best pre-

dictors of actual turnover is thoughts of turnover (e.g.,

Griffeth, Hom, & Gaertner, 2000). Finally, while we did

cross-validate results across two large samples that differed

by ethnicity and socio-economic status, both samples came

from male-dominated organizations. Gender parity has

increased in both industries, but as of 2008, only 34% of

lawyers were women (U.S. Department of Labor, 2009),

and as of September 2009, only 14% of active-duty mili-

tary personnel were women (Department of Defense,

2009).

We should also emphasize that our data combine the

experiences of hundreds of women, and just because gen-

der harassment correlates with negative outcomes in these

aggregate data does not mean that this behavior has neg-

ative outcomes for every individual victim. We found that

this is true on average. Whether it is true for any individual

woman is a determination that must be made on a case-by-

case basis. Our point is simply that courts and social sci-

entists should not automatically assume that ‘‘just gender

harassment’’ is, by definition, too trivial to create an abu-

sive work environment.

Future Directions

Social scientists continue to focus on sexual ‘‘come-on’’

forms of sexual harassment (e.g., de Haas, Timmerman, &

Höing, 2009; Woodzicka & LaFrance, 2005). When sur-

veys do include questions about gender harassment, during

the analysis of data these questions are often combined

with questions about unwanted sexual advances (e.g.,

Fitzgerald, Drasgow et al., 1997; Sims et al., 2005). This

happens even though there are major qualitative differ-

ences across the experiences; for example, being sexually

propositioned on one occasion is not the same as being

targeted with demeaning anti-female remarks on a daily

basis. Based on the current study, we recommend that more

research parse out experiences of gender harassment from

unwanted sexual attention/coercion, which will give rise to

new avenues of inquiry. For example, proponents of the-

ories as to why people harass others have generally looked

for one unifying explanation (e.g., Bargh, Raymond, Pryor,

& Strack, 1995; Berdahl, 2007b). It remains entirely pos-

sible that different goals motivate the different subtypes of

sex-based harassment.

Regarding construct labels, we second Berdahl’s

(2007b) recommendation that the term ‘‘sex-based harass-

ment’’ be used in lieu of ‘‘sexual harassment.’’ Research on

Law Hum Behav (2011) 35:25–39 37

123

lay perceptions suggests that the latter term conjures up

narrow notions of unwanted sexual advances; it fails to

include gender harassment in the minds of many (e.g.,

Fitzgerald & Ormerod, 1991; Loredo et al., 1995; Tang

et al., 1995). The term ‘‘sex-based harassment’’ is broader,

calling attention to both categories of behavior. It is also

closer to the original language of Title VII, which pro-

hibited harassment ‘‘based on sex.’’

The current study focused on two male-dominated pro-

fessions. Researchers have not yet determined how ‘‘just

gender harassment’’ operates in gender-balanced and

female-dominated industries. We would speculate that the

behavior might be less common in more female-integrated

contexts, where women are not perceived as ‘‘encroach-

ing’’ on ‘‘men’s territory.’’ In those settings, anti-female

hostility might also be less tolerated and more penalized,

and thus experienced by victims as less threatening. These

and other possibilities await future research.

In closing, we emphasize the need for scholars of sex-

based harassment, both in psychology and law, to continue

the interdisciplinary exchange of ideas. Social scientists

and legal theorists often struggle to converse, but consid-

erable efforts have been made to bridge the disciplinary

gap when it comes to this topic (for examples, see the

special issue on ‘‘sexual harassment’’ in Psychology,

Public Policy, and Law, 1999, and the special issue on

‘‘psychology, law, and the workplace’’ in Law and Human

Behavior, 2004). The arguments of prominent legal

scholars inspired the current psychological research. We

hope that, in turn, our results can inform further evolution

in legal thinking about harassment based on sex and

gender.

Acknowledgments We are grateful to Anna Kirkland, Abby Stewart, and members of our research lab at the University of

Michigan for their valuable feedback on this research. Thanks also to

the following individuals for their contributions to Study 2: Louise F.

Fitzgerald, Leslie V. Freeman, Vicki J. Magley, Kimberly A. Lon-

sway, Regina Day Langhout, Jill Hunter-Williams. This article is

based on the master’s thesis of the first author, who presented parts of

it in July 2008 at the Annual Meeting of the International Society for

Political Psychology in Paris, France.

References

1999 Survey of Active Duty Personnel [Data file and documentation on CD-ROM]. (2001). Arlington, VA: DMDC.

Bargh, J. A., Raymond, P., Pryor, J. B., & Strack, F. (1995).

Attractiveness of the underling: An automatic power ? sex association and its consequences for sexual harassment and

aggression. Journal of Personality and Social Psychology, 68, 768–781. doi:10.1037/0022-3514.68.5.768.

Barling, J., Rogers, A., & Kelloway, E. (2001). Behind closed doors:

In-home workers’ experience of sexual harassment and work-

place violence. Journal of Occupational Health Psychology, 6(3), 255–269. doi:10.1037/1076-8998.6.3.255.

Barnes v. Costle, 561 F.2d 983 (D.C. Cir. 1977).

Berdahl, J. L. (2007a). The sexual harassment of uppity women.

Journal of Applied Psychology, 92, 425–437. doi:10.1037/0021- 9010.92.2.425.

Berdahl, J. L. (2007b). Harassment based on sex: Protecting social

status in the context of gender hierarchy. Academy of Manage- ment Review, 32(2), 641–658.

Berdahl, J. L., & Raver, J. L. (2010). Sexual harassment. In S. Zedeck

(Ed.), Handbook of industrial/organizational psychology. Wash- ington, DC: American Psychological Association.

Berg, S. H. (2006). Everyday sexism and post traumatic stress

disorder in women: A correlational study. Violence Against Women, 12(10), 970–988. doi:10.1177/1077801206293082.

Bergman, M., Langhout, R. D., Palmieri, P., Cortina, L. M., &

Fitzgerald, L. F. (2002). The (un)reasonableness of reporting:

Antecedents and consequences of reporting sexual harassment.

Journal of Applied Psychology, 87, 230–242. doi:10.1037/0021- 9010.87.2.230.

Berwick, D. M., Murphy, J. M., Goldman, P. A., Ware, J. E., Barsky,

A. J., & Weinstein, M. C. (1991). Performance on a five-item

mental health screening test. Medical Care, 29, 169–176. doi: 10.1097/00005650-199102000-00008.

Cortina, L. M., & Berdahl, J. L. (2008). Sexual harassment in

organizations: A decade of research in review. In J. Barling & C.

L. Cooper (Eds.), The SAGE handbook of organizational behavior (pp. 469–497). Thousand Oaks, CA: Sage.

Cortina, L. M., Lonsway, K. L., Magley, V. J., Freeman, L. V.,

Collinsworth, L. L., Hunter, M., et al. (2002). What’s gender got

to do with it? Incivility in the federal courts. Law and Social Inquiry, 27, 235–270.

Culbertson, A., & Rosenfeld, P. (1994). Assessment of sexual

harassment in the active-duty Navy. Military Psychology, 6(2), 69–93. doi:10.1207/s15327876mp0602_1.

de Haas, S., Timmerman, G., & Höing, M. (2009). Sexual harassment

and health among male and female police officers. Journal of Occupational Health Psychology, 14, 390–401. doi:10.1037/ a0017046.

Department of Defense. (2009). Active duty military personnel by

service by rank/grade. Retrieved December 14, 2009, from

http://siadapp.dmdc.osd.mil/personnel/MILITARY/miltop.htm.

Edwards, J. E., Elig, T. W., Edwards, D. L., & Riemer, R. A. (1997).

The 1995 Armed Forces Sexual Harassment Survey: Adminis- tration, datasets, and codebook for Form B (report no. 95–015). Arlington, VA: DMDC.

Epstein, L. B. (1998). What is a gender norm and why should we

care? Implementing a new theory in sexual harassment law.

Stanford Law Review, 51, 161–182. Fitzgerald, L. F., Drasgow, F., Hulin, C. L., Gelfand, M. J., & Magley,

V. J. (1997). Antecedents and consequences of sexual harass-

ment in organizations: A test of an integrated model. Journal of Applied Psychology, 82(4), 578–589. doi:10.1037/0021-9010. 82.4.578.

Fitzgerald, L. F., Gelfand, M. J., & Drasgow, F. (1995). Measuring

sexual harassment: Theoretical and psychometric advances.

Basic and Applied Social Psychology, 17, 425–445. doi:10.1207/ s15324834basp1704_2.

Fitzgerald, L. F., Magley, V. J., Drasgow, F., & Waldo, C. R. (1999).

Measuring sexual harassment in the military: The Sexual

Experiences Questionnaire (SEQ-DoD). Military Psychology, 11(3), 243–263. doi:10.1207/s15327876mp1103_3.

Fitzgerald, L. F., & Ormerod, A. J. (1991). Perceptions of sexual

harassment: The influence of gender and academic context.

Psychology of Women Quarterly, 15, 281–294. doi:10.1111/j. 1471-6402.1991.tb00797.x.

Fitzgerald, L. F., Schullman, S. L., Bailey, N., Richards, M., Swecker,

J., Gold, Y., et al. (1988). The incidence and dimensions of

38 Law Hum Behav (2011) 35:25–39

123

sexual harassment in academia and the workplace. Journal of Vocational Behavior, 32, 152–175. doi:10.1016/0001-8791 (88)90012-7.

Fitzgerald, L. F., Swan, S., & Magley, V. J. (1997). But was it really

sexual harassment? Legal, behavioral, and psychological defini-

tions of the workplace victimization of women. In W.

O’Donohue (Ed.), Sexual harassment: Theory, research, and treatment (pp. 5–28). Boston: Allyn & Bacon.

Foote, W. E., & Goodman-Delahunty, J. (2005). Evaluating sexual harassment: Psychological social, and legal considerations in forensic evaluations. Washington, DC: American Psychological Association.

Franke, K. M. (1995). The central mistake of sex discrimination law:

The disaggregation of sex from gender. University of Pennsyl- vania Law Review, 144, 1–99.

Franke, K. M. (1997). What’s wrong with sexual harassment?

Stanford Law Review, 49, 691–772. Franke, K. M. (2004). What’s wrong with sexual harassment? In C. A.

MacKinnon & R. B. Siegel (Eds.), Directions in sexual harass- ment law (pp. 169–181). New Haven, CT: Yale University Press.

Glomb, T. M., Munson, L. J., Hulin, C. L., Berman, M., & Drasgow,

F. (1999). Structural equation models of sexual harassment:

Longitudinal explorations and cross-sectional generalizations.

Journal of Applied Psychology, 84, 14–28. doi:10.1037/0021- 9010.84.1.14.

Griffeth, R. W., Hom, P. W., & Gaertner, S. (2000). A meta-analysis

of antecedents and correlates of employee turnover: Update,

moderator tests, and research implications for the next millen-

nium. Journal of Management, 26, 463–488. doi:10.1177/014 920630002600305.

Growe, J. D. (2007). Reform the EEOC guidelines: Protect employees

from gender discrimination as mandated by Title VII. Journal of Law and Policy, 24, 275–300.

Harris v. Forklift Sys (1993) 114 S.Ct. 367.

Hartigan, J. A. (1975). Clustering algorithms. New York: Wiley. Hébert, C. (2005). The disparate impact of sexual harassment: Does

motive matter? Kansas Law Review, 53, 341–395. Langhout, R. D., Bergman, M. E., Cortina, L. M., Fitzgerald, L. F.,

Drasgow, F., & Williams, J. H. (2005). Sexual harassment

severity: Assessing situational and personal determinants and

outcomes. Journal of Applied Social Psychology, 35, 975–1007. doi:10.1111/j.1559-1816.2005.tb02156.x.

Lim, S., & Cortina, L. M. (2005). Interpersonal mistreatment in the

workplace: The interface and impact of general incivility and

sexual harassment. Journal of Applied Psychology, 90, 483–496. doi:10.1037/0021-9010.90.3.483.

Lipari, R. N., & Lancaster, A. R. (2003). Armed Forces 2002 Sexual Harassment Survey. DMDC Report No. 2003–026. Arlington, VA: Defense Manpower Data Center.

Loredo, C., Reid, A., & Deaux, K. (1995). Judgments and definitions

of sexual harassment by high school students. Sex Roles, 32, 29– 45. doi:10.1007/BF01544756.

Meritor Savings Bank v. Vinson, 477 U.S. 57 (1986).

Mowday, R. T., Steers, R. M., & Porter, L. W. (1979). The

measurement of organizational commitment. Journal of Voca- tional Behavior, 14, 224–247. doi:10.1016/0001-8791(79) 90072-1.

Oncale v. Sundowner Offshore Services, 523 U.S. 75 (1998).

Perry, E., Kulik, C., & Schmidtke, J. (1998). Individual differences in

the effectiveness of sexual harassment awareness training.

Journal of Applied Social Psychology, 28(8), 698–723. doi: 10.1111/j.1559-1816.1998.tb01727.x.

Pryor, J. B. (1987). Sexual harassment proclivities in men. Sex Roles, 269–290. doi:10.1007/BF00288453.

Richman, J. A., Shinsako, S. A., Rospenda, K. M., Flaherty, J. A., &

Freels, S. (2002). Workplace harassment/abuse and alcohol-

related outcomes: The mediating role of psychological distress.

Journal of Studies on Alcohol, 63(4), 412–419. Rotundo, M., Nguyen, D., & Sackett, P. R. (2001). A meta-analytic

review of gender differences in perceptions of sexual harass-

ment. Journal of Applied Psychology, 86, 914–922. doi:10.1037/ 0021-9010.86.5.914.

Schneider, K. T., Swan, S., & Fitzgerald, L. F. (1997). Job-related

psychological effects of sexual harassment in the workplace:

Empirical evidence from two organizations. Journal of Applied Psychology, 82(3), 401–415. doi:10.1037/0021-9010.82.3.401.

Schneider, K. T., Tomaka, J., & Palacois, R. (2001). Women’s

cognitive, affective, and physiological reactions to a male

coworker’s sexist behavior. Journal of Applied Social Psychol- ogy, 31, 1995–2018. doi:10.1111/j.1559-1816.2001.tb00161.x.

Schultz, V. (1998). Reconceptualizing sexual harassment. Yale Law Journal, 107, 1683–1732.

Schultz, V. (2003). The sanitized workplace. Yale Law Journal, 112(8), 2061–2197.

Schultz, V. (2006). Understanding sexual harassment law in action:

What has gone wrong and what we can do about it. Thomas Jefferson Law Review, 29, 1–53.

Sims, C. S., Drasgow, F., & Fitzgerald, L. F. (2005). The effects of

sexual harassment on turnover in the military: Time-dependent

modeling. Journal of Applied Psychology, 90(6), 1141–1152. doi:10.1037/0021-9010.90.6.1141.

Spector, P. E. (1997). Job satisfaction: Application, assessment, cause, and consequences. Thousand Oaks, CA: Sage.

Stark, S., Chernyshenko, O. S., Lancaster, A. R., Drasgow, F., &

Fitzgerald, L. F. (2002). Toward standardized measurement of

sexual harassment: Shortening the SEQ-DoD using item

response theory. Military Psychology, 14, 49–72. doi:10.1207/ S15327876MP1401_03.

Steele, C., & Aronson, J. (1995). Stereotype threat and the intellectual

test performance of African Americans. Journal of Personality and Social Psychology, 69(5), 797–811. doi:10.1037/0022- 3514.69.5.797.

Swim, J. K., Hyers, L. L., Cohen, L. L., & Ferguson, M. J. (2001).

Everyday sexism: Evidence for its incidence, nature, and

psychological impact from three daily diary studies. Journal of Social Issues, 57(1), 31–53. doi:10.1111/0022-4537.00200.

Tang, C. S.-K., Yik, M. S. M., Cheung, F. M. C., & Choi, P.-K.

(1995). How do Chinese college students define sexual harass-

ment? Journal of Interpersonal Violence, 10, 503–515. doi: 10.1177/088626095010004008.

Title VII of the Civil Rights Act of 1964, 42 U.S.C. § 2000e (1994).

Retrieved May 15, 2009, from http://www.eeoc.gov/policy/

vii.html.

U.S. Department of Labor. (2009). Women in the labor force: A

databook. Retrieved December 14, 2009, from http://www. bls.gov/cps/wlf-databook-2009.pdf.

U.S. Merit Systems Protection Board. (1994). Sexual harassment in the federal workplace: Trends, progress, continuing challenges. Washington, DC: U.S. Government Printing Office.

Viet, C. T., & Ware, J. E., Jr. (1983). The structure of psychological

distress and well-being in general populations. Journal of Consulting and Clinical Psychology, 51, 730–742. doi:10.1037/ 0022-006X.51.5.730.

Ware, J. E., & Sherbourne, C. D. (1992). The MOS 36-item short

form health survey (SF-36): I. Conceptual framework and item

selection. Medical Care, 30, 473–483. doi:10.1097/00005650- 199206000-00002.

Williams v. General Motors Corp., 187 F.3d 553 (6th Cir. 1999).

Woodzicka, J., & LaFrance, M. (2005). The effects of subtle sexual

harassment on women’s performance in a job interview. Sex Roles, 53(1), 67–77. doi:10.1007/s11199-005-4279-4.

Law Hum Behav (2011) 35:25–39 39

123

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.