ArticlePDF Available

Abstract

For 30 years, sexual harassment has been recognized as a serious organizational problem and a violation of US law. The Navy Tailhook scandal and Clarence Thomas hearings in l99l launched sexual harassment to the forefront of public attention. This was followed by a virtual explosion ofresearch on the topic, leading to the estimate that one out of every two women is harassed at some point in her working life. We review this scholarship in the current chapter, concentrating on the last decade ol work. Our principal locus is sexual harassment in the workplace. Although sexual harassment also occurs elsewhere, other domains are beyond the scope of this chapter. This chapter also primarily covers research since the mid-1990s (except for brief historical overviews). Sexual harassment scholarship began in the late I 970s, and several large-scale surveys in the 1980s (Gutek, 1985;USMSPB, 1981, 1988; Martindale, 1990) influenced work that followed. Since then, however, the workforce has become more educated about sexual harassment, organizational methods of combating sexual harassment have evolved, and sexual-harassment research methodolo-gies have become increasingly advanced. We therefore concentrate on the most recent, methodologically sophisticated work. Finally, resea¡ch on this topic largely addresses men's harassment of women, so this will be the main focus of our review. We organize this chapter around the following questions: What is sexual harassment? Why does it happen? Who harasses whom? What are its effects? Finally, how do and how should individuals and organizations respond to sexual harassment? Each of these sections provides a brief historical recap of early work on the topic, followed by a detailed review of recent scholarship. Throughout, we address ¡elevant issues in US law but maintain a focus on theory and finclings tiom social science (particularly psychology). The chapter will close with a discussion of the future of sexual harassment scholarship.
25
Sexual
Harassment
in
Organizations:
A
Decade
of
Research
in
Review
Lilia
M.
Cortina
and
Jennifer
L. Berdahl
For 30
years, sexual
harassment
has
been
recognized
as a serious
organizational
prob-
lem and
a violation
of
US
law.
The Navy
Tailhook scandal
and
Clarence
Thomas
hear-
ings
in l99l
launched
sexual
harassment
to
the
forefront
of
public attention.
This was
followed
by a
virtual explosion
ofresearch
on
the topic,
leading
to the
estimate
that
one out
of
every two
women
is
harassed at
some
point
in her
working
life. We
review
this scholarship
in the current
chapter,
concentrating
on the
last
decade
ol
work.
Our
principal locus
is sexual
harassment
in
the
workplace.
Although
sexual
harassment
also
occurs
elsewhere,
other
domains
are
beyond
the
scope
of this
chapter.
This
chapter
also
primarily
covers
research
since
the
mid-1990s
(except
for
brief
historical
overviews).
Sexual
harassment
scholarship
began
in the
late
I
970s,
and several
large-scale
surveys
in the 1980s
(Gutek,
1985;USMSPB,
1981, 1988;
Martindale,
1990)
influenced
work that
followed.
Since
then,
however,
the
workforce
has become
more
educated
about
sexual
harassment,
organizational
methods
of
combating
sexual
harassment
have evolved,
and
sexual-harassment
research
methodolo-
gies have become
increasingly
advanced.
We
therefore
concentrate
on
the
most
recent,
methodologically
sophisticated
work.
Finally,
resea¡ch
on this
topic
largely
addresses
men's
harassment
of
women, so
this
will be the
main
focus
of our
review.
We organize
this chapter
around
the fol-
lowing
questions:
What is sexual
harassment?
Why
does
it
happen?
Who
harasses
whom?
What
are
its effects?
Finally,
how do
and
how should
individuals
and
organizations
respond
to sexual
harassment?
Each of these
sections
provides a
brief
historical
recap
of early
work on
the topic,
followed
by
a
detailed
review
of
recent
scholarship.
Throughout,
we address
¡elevant
issues
in
US
law
but
maintain
a
focus on
theory
and
finclings
tiom social
science
(particularly
psychology).
The chapter
will close
with
a
discussion
of
the
future of
sexual
harassment
scholarship.
47fJ
DEFIN¡NG SEXUAT HARASSMENT
There are two main approaches to defining
sexual harassment: One from a
legal
perspec-
tive and the other from a social-psychological
perspective.
In
general,
social-psychological
definitions are
broader
than legal ones, though
recent
exceptions exist.
A third
perspective
on sexual harassment
-
the
public,
or lay
perspective
-
preceded
legal
and social-
psychological
ones but now lags well behind
each
in
understanding the scope,
nature, and
impact of the
phenomenon.
We review each
in turn
below.
Legal
definitions
According to historical writings, sexually
harassing behavior has long been a
problem
(e.g.,
Segrave, 1994). The term
'sexual
harassment,' however,
only emerged
in the
1970s, when feminists argued that sexual
threats, bribes, and objectification
presented
odious conditions of employment often
faced
by
women,
but
rarely
by
men,
and
therefore constitute unlawful sex discrimina-
tion
(Farley,
1978; MacKinnon, 1979). The
historical
pervasiveness
of this behavior made
it so taken-for-granted that courts initially
balked at the
idea
of calling
it
discrimination,
and early cases were denied or decided
in favor
of defendants. Organizations saw
sexual harassment as a
'private'issue
between
the
harasser
and
victim,
beyond the scope
of organizational responsibility
(MacKinnon,
r979).
This changed in the late 1970s when
US courts finally
decided that
women who
lost
jobs
for failing to comply with their
employers' sexual
demands
were
discrimi-
nated against based on sex
(beginning
with
Williams v. Saxbe, 1976). Courts used
Title VII
of the
1964
Civil
Rights Act to
reason that
quid
pro quo
sexual harassment
(the
loss/denial
of a
job-related
benefit
fbr
refusal to cooperate sexually) was illegal
sex discrimination.
The legal
definition of
sexual harassment was expanded in the I 980s
to include hostile
environment
harassment:
Unwanted sexual attention and requests that
THE
SAGE
HANDBOOK
OF
ORGANIZATIONAL BEHAVIOR
do not necessarily come
from a
supervisor or
result in the loss/denial of aj ob-related benefi t,
but that create a
hostile work
environment
(Bundy
v. Jackson, l98l; Meritor Savings
Bank v. Vinson, 1986).
Unlike
quid pro
quo
harassment, which typically involves
one
perpetrator
and
victim, hostile
envi-
ronment harassment can involve multiple
perpetrators
and
victims.
Some acts
(e.g.,
posting pornography,
telling sexist
jokes)
may
be experienced by
many employees but
create a hostile environment for only a few.
Recognizing hostile
environment
harassment
meant recognizing that sexual behavior
itself
can be
hostile
and
demeaning,
particularly
to women, who constitute the main targets
of sexual objectification, exploitation, and
violence in the world. Sexual behavior at
work
can therefore
remind men and women of their
unequal status
in
society
more
broadly
and
reinforce
their
inequality
at
work.
In 1980 the
US
Equal Employment Oppor-
tunity Commission
(EEOC;
the legal entity
charged with enforcing federal sex discrimi-
nation law)
developed the
ibllowing
defìnition
of sexual harassment, still used today:
Unwelcome sexual advances,
requests for sexual
favors,
and other
verbal
or
physical
conduct of a
sexual
nature
const¡tute sexual
harassment when
this conduct expìicitly or
implicitly affects an
individual's employment, unreasonably
interferes
with
an
individual's work
performance,
or
cre-
ates an
intimidating, hostile,
or offensive
work
environment.
(p.
7 4617)
The EEOC has
since
offered more specific
guidelines
for identifying sexual harassment.
Prompted
by court
rulings,
these
include that
the victim and harasser can be ofthe same sex,
that the
harasser need not
be
employed by the
victim's organization, and that the victim can
be anyone affected
by the conduct
(including
those not directly targeted).
As
awareness of sexual
harassment and
the breadth of behavior
covered by law
have
grown,
so too have the number of
gricvancos
filed.
In 1980,
the
EEOC rcceived
one sexual harassment complaint.
By 1989,
nearly
6,000
new cases had been fìled, and
between 1990 and 1999 this number soared
SEXUAL
HARASSMENT
To 37,725.
The number
of
new complaints
filed
annually
peaked at 5,332
in the
year
2000, and
has declined
slightly
each
year
since
(http://www.eeoc. gov/stats/trarassment.html).
Sociahpsych
ol o
gi
cal d efi
niti o ns
Unlike
legal deûnitions,
social-psychological
perspectives on sexual
harassment
do
not
require
negative
work outcomes
and
therefore
tend to be
broader.
The focus
instead
is on
speciflc
behaviors
and
the victim's
subjective
experience
of those
behaviors.
Illustrating
this
perspective, Fitzgerald
et al.
(1997:
15)
define
sexual harassment
as
'unwanted
sex-
related behavior
at
work that
is appraised
by the
recipient
as offensive,
exceeding
her
resources, or
threatening
her
well-being.'
Psychologists
have concentrated
on
devel-
oping
operational defi
nitions
of
sexual
harass-
ment.
In a
now-classic
study,
Till
(1980)
collected descriptive
anecdotes
and classi-
fied sexually
harassing
conduct
into five
categories:
(1)
generalized sexist remarks
or behavior;
(2)
inappropriate
and
offensive
(but
essentially
sanction-free)
sexual advances;
(3)
solicitation
of sexual activity
orothersex'linked
behavior
by
promise
of rewards;
(4)
coercion
of sexual
act¡vity
by threat
of
punishments; and
(5)
sexual
assaults.
Fitzgerald and
colleagues
(1988)
developed
a
list of behaviors
to reflect
these five categories
and asked
women
students and
employees
how
often they
experienced
each.
Factor-
analysis
revealed a
three-factor
structure:
(1)
gender
harassment
(Till's
category
1, sex¡st
remarks
and behavior);
(2)
unwanted
sexual
attention
(Till's
categories
2
and
5, sexual
attention and
force); and
(3)
sexual
coercion
(Till'scategories3
and4,
threats
and bribes).
Gender
harassment
and
unwanted
sexual
attention
correspond
to the
legal definition
of
hostile environment
harassment,
while
sexual coercion
parallels
illegal
quid
pro
quo
harassment
(Fitzgerald
et al.,
1988;
Fitzgerald et
al.,
1995a).
Based on
this
work,
Fitzgerald and
colleagues
developed
the Sexual
Experiences
Questionnaire
(SEQ),
the most
widely-used
and
validated
measure
of sexual
harassment
to
date.
Lay definitíons
It
is important
to consider
opinions
about
sexual harassment
in the
general
public.
Lay
perceptions have
a
profound influence
on managerial
policy
and
employee
ideas
about
what constitutes'appropriate'
behavior
at work
and
what
justifies
a
complaint.
More
research
has examined
lay
perceptions
than any
other
aspect of
sexual
harassment
(over
300
studies to
date, according
to the
PsycINFO
database),
perhaps
due
to the ease
and speed
with
which such
research
can be
conducted.
Not surprisingly,
lay
perceptions of
sexual
harassment
have differed
over time,
between
men and
women,
and across
cultures.
The US
Merit
Systems
Protection
Board
(USMSPB)
asked
federal employees
in 1980,
1987, and
1994 to
indicate
whether they
thought
each
of six
different
types of
behavior
(from
sexual
teasing
to sexual
pressure) constituted
sexual
harassment.
In each
successive
survey,
a
greater
proportion of employees
judged
each type
of behavior
as harassing.
Other
studies
have consistently
shown
that
women
are
more likely
than
men to
view sexual
behaviors
as
harassing.
The
gap
between
men's and
women's
perceptions
is
quite
small
for sexual
pressure
and
coercion,
especially
from a supervisor,
but is
greater for
gender
harassment
(USMSPB,
1994;
Blumenthal,
1998;
Rotundo,
Nguyen, and
Sackett,
2001).
It
is important
to
note, however,
that a
majority
of men and
women
consider
gender-
harassing
behaviors
to
be sexual
harassment
(77-88
per
cent of
women,
and 64-70
per
cent
of men,
USMSPB,
1994). In studies
asking
participants to evaluate
how
offended
or bothered
they
would
be
(or
have been)
by
specific
behaviors,
the
gender
gap
widens:
Men
often
report not being
upset
by, and
even
enjoying,
a variety
of behaviors
that
women
THE SAGE HANDBOOK
OF ORGANIZATIONAL
BEHAVIOR
find
harassing
(Berdahl,2007a; Berdahl et
al.,
1996; Gutek,
1985).
Emerging
developments
and
debates
ahout
definitions
of
sexual
harassment
Although
sexual
harassment
is
now a
well-
established
construct
in both
law and
psychol-
ogy,
qucstions rcmain about
how bcst
to define
and assess this behavior.
Which
definition
should
researchers
adopt?
There
has been some
debate
about
whether
researchers
should
use legal
or
social-
psychological definitions
of
sexual
harass-
ment
to study
its
prevalence
(Fitzgerald
et al.,
1997; Gutek
et
al., 2004).
Because
social-psychological
definitions
are broader
than
legal ones,
measuring
sexual
harassment
according
to the
former
yields
higher
preva-
lence estimates.
This
may
pose
a
problem in
legal contexts
if the
focus of
the assessment
is
strictly
limited to
unlawful
behavior.
On
the other
hand,
if researchers
are
inter-
ested
in studying
and
understanding
sexual
harassment
as a
social
and
psychological
phenomenon, using definitions
derived
from
social-psychological
theories
makes
the most
sense.
Confining
measures
to current
legal
definitions
risks
studying
a
narrow and
mov-
ing target.
It would
make cross-temporal
and
cross-cultural
comparisons
difficult
becausc
sexual
harassment
law
has evolved
over
time and
differs
widely
across countries
(some
of which
have
no laws against
sexual
harassment).
Restricting
studies of
sexual
harassment to
legal definitions
implicitly
argues
that sexual
harassment
should
not
have
been
studied
prior
to the
late
1970s in the
US
and should
not be studied
in many
parts
of the
world today.
As social
scientists,
our charge
is
to shed
light on social
phenomena, not to
limit
our
attention to
phenomena currently
deemed
illegal.
There is also
debate
about
whether
lay
perceptions of sexual
harassment
should
be
used
to infbrm
definitions
of
the construct.
This
is a complicated
issue. On
the one
hand, the
general
public
is usually
much
less
informed about
sexual
harassment
than
the
lawyers,
judges,
policy
experts,
and
social
scientists
who study
it. During
some
eras,
and
among
some
people
and
cultures,
behaviors
that we
now consider
to
be the
most heinous
examples
of
harassment
were
considered
justified
(Segrave,
1994).
Had the
courts and
social
scientists
relied
on
majority
public
opinion
polls,
sexual
harassment
probably
would
have
never been
identified
as
a
form of sex
discrimination.
After
all, these
behaviors
were tolerated
-
even
condoned
-
for centuries.
On
the other
hand,
if sexual
harassment
is
partially
defined
by the subjective
experience
of
its victim,
fhen
how the
victim
experi-
ences
these behaviors
must be
taken
into
account.
If someone
reports
enioying
sexual
attention
at
work
-
even
uninvited
sexual
attention
-
then
it is
probably inappropriate
to
label
that
person's
experience
'harassing.'
Research
shows
that
many
victims do
not
label their
own
experiences
as
harassment
(Arvey
and
Cavanaugh,
1995;
Magley
et al.,
1999a).
Thus, rather
than
having
respondents
categoúze
it as such,
researchers
should
instead define
sexual
harassment
and use
those
guidelines
to
measure
harassment.
Considering
non-sexual
forms
of
harassment
Rcscarch
has incrcasingly
identificd
forms
of
harassment
that discriminate
based
on
sex
but do
not
necessarily
entail sexual
advances.
For example,
professional
women compared
to men
report signifìcantly
more incivility
and
aggression
-
behaviors
that
alienate
the
victim
rather than
approach
them
sexually
(Berdahl,
2007c; Cortina,
2008).
Moreover,
when
men are
harassed,
it often
involves
punishment for deviating
from traditional
masculine
gender
roles
(Berdahl
et al.,
1996;
Oncale
v. Sundowner
Offshore
Services,
1998;
Waldo et
al., 1998).
Examples
include
teasing
a
man about
his
role in the
home;
deriding
him
for failing
to
participate in
the
objectification
of
womenl and
calling
him derogatory
names
that
challenge
his
masculinity.
Consistent
with this,
legal theo-
rists argue
that
sex-based
harassment
often
SEXUAL HARASSMENT
entails
behaviors that undermine
the victim
but make
no
explicit
reference to
sexuality
(Franke,
1997; Schultz,
1998). That
is,
'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: 1687). Capturing
the notion
sexual harassment
can be based
on sex
but
not necessarily sental,
Berdahl
(2007b)
offers a new definition
of sexual
harass-
ment as
'behavior
that derogates,
demeans,
or humiliates an
individual based
on that
individual's sex.'
Co ns i deri ng
perspect¡ves
beyon d
mainstream
White Am eri ca
Despite the fact that
some of the
most
prominent
sexual harassment cases
in the US
have involved ethnic-minority
victims
(e.g.,
Anita Hill in the Senate confirmation
hearings
of Clarence
Thomas; Mechelle
Vinson in
Meritor Savings Bank
v. Vinson, 1986), the
most
prominent
sexual harassment
research
has focused on
White/European American
women.
Questions
remain about
whether and
how models of
sexual harassment extend
to
women from other ethnic
and cultural
backgrounds.
One
manifestation of
harassment that may
be more salient
to ethnic minority
women
is
sexual
racism. This refers
to harassment
that combines sexism
and racism to
creaie a
simultaneous
manifestation
of sex and
race
discrimination.
These are'...forms of
sexual
aggression
fthat]
are embedded
in a system of
interlocking
race,
gender,
ethnicity,
and class
oppression'
(Munell,
1996: 56). Behaviors
falling
into
this
category
include not only
those that disproportionately
target
minority
women, but also conduct
that reflccts and
perpetuates
stereotypes
about
particular
gen-
ders in
particular
ethnic
groups
(e.g.,
Adams,
1 997;
Buchanan and Ormerod,
2002; Cortina,
2001). To date, this concept
has
primarily
been the
focus of theory and commentary,
so
it remains unclear
how sexual racism
might
fit into empirical
models. Given
that sexual
harassment transcends
boundaries of
race,
class, and
country
(e.g.,
Barak, 1997), defi-
nitions and assessments
of this
phenomenon
must begin
considering
perspectives
beyond
mainstream
White America.
Evolving measures of
sexual harassment
Operational
definitions of sexual
harassment
have varied over the
past
20
years, for
good
reason: Criteria
for what constitutes
sexual
harassment have expanded
(e.9.,
to
include
same-sex
harassment); research
has shown
that
groups
differ on
which behaviors they
consider to be
harassing
(e.g.,
unlike women,
many men do not experience
uninvited
sexual
attention as
harassing); and
scholars have
come to recognize
that sexual
harassment
involves
different language,
insinuation, and
reference by context
(e.g.,
the
military vs.
a
law firm; one culture
vs. another).
This
poses
a
problem
for comparative
research
purposes
(Gutek
et a1.,2004).
The USMSPB
studies spanning
14
years
used
the same
six
items assessing
'socio-sexual behavior;'
this facilitated
comparisons across
survey
administrations,
but
ignored developments
in
understanding
sexual
harassment, such as
its
frequent non-sexual
forms and the
important
qualification
that it be unpleasant
or offensive
to the victim.
The
SEQ
has been adapted
over time to
reflect
the
particular
styles
of sexual
harass-
ment in different contexts
and against differ-
ent
groups,
such
as the military
(Fitzgerald
et al.,
1999b), Latinas
(Cortina,
2001),Turkish
women
(Wasti
et a1., 2000), and
men
(Waldo
et al.,
1998;
Berdahl and Moore,
2006).
At
the
same time, the
factor structure of
the
SEQ
(gender
harassment,
unwanted sexual
attention,
and sexual
coercion)
has remained
stable across
time, culture,
and occupational
sector, despite
variations in the specifìc
items
assessing
each construct
(Gelfand
et al.,
1995; Lee and Ormerod,
2003).1
Lee and
Ormerod
(2003:
6) argued
that,
'similar
to
aptitude
testing ...
it is the dimensions,
rather than
any
particular items, that form
the core construct
... The dimensions
are
considered
finite, whereas
infinite
items
can
be
sampled as
needed for the assessment
of
particular populations.' In addition to this
structural
robustness,
the SEQ
consistently
predicts various
professional,
psychological,
474
and
physical
health outcomes
(see
Hershcovis
and
Barling, under
review). Moreover,
all
SEQ
items were
developed to
meet the
highest
psychometric
standards
(e.g.,
using
clear
behavioral
language; avoiding
'double-
barreled'items
or those
with multiple compo-
nents; including
multiple items to assess
each
latent construct;
e.g.,
Dillman,2000).
As
such,
the
SEQ
presents
a
flexible but
hi-ehly reliable
and valid approach
to assessing
unwanted
sex-related
behavior
at work.
THEORIZING
SEXUAT
HARASSMENT
Why does
sexual harassment
occur?
Below
we discuss
four viewpoints:
(1)
The
'nature' perspective,
which sees
sexual
harassment
as the result of
biological sex
differences;
(2)
The
'nurture' perspective,
which conceptual-
izes sexual
harassment as
a consequence of
socialized
sex roles and stereotypes;
(3)
The
'power'
perspective,
which
views sexual
harassment
as emerging from
sex differences
in
power;
and
(4)
the
'nurture
x
power'
perspective,
which regards
sexual harassment
as a means of
protecting
valued
social identities
(for
other reviews,
see
Tangri and Hayes,
1 997; Welsh,
1 999).
fhe nature
perspective:
Physical
design
Within the
nature
perspective,
sexual
harass-
ment is viewed as
the inevitable
and natural
result of biological
sexual
urges. The
most
common
pattern
of
male
perpetrators harass-
ing
female victims
is
attributed
to assumed
sex
differences
in
sexual
drive and
function
(Studd
and
Gattiker, 1991).
This explanation
fails to
predict
most sexual
harassment,
however,
which constitutes
hostile acts aimed
not at sexual
intimacy but
rather at degra-
dation and
alienation of
the victim.
It also
fails to explain
why sexual
harassment
is
usually targeted
at individuals
who
violate
gender
ideals
¡ather than those
who meet
them
(Berdahl,2007b).
Thus, despite
its ready
acceptance
among
the lay
public,
most sexual
THE SAGE HANDBOOK OF ORGANIZATIONAL
BEHAVIOR
harassment scholars
have dismissed
the
nature
perspective.
Some
have also
rejected it
for
its
pessimistic implications.
As one scholar
noted,
'linking sexual
harassment
with libido
laid the
groundwork for excusing, accept-
ing, and
forgiving
male violence
against
women ...
If it is libido, then
nature
is
the
culprit,
and what can
be done about
nature?'
(Segrave,
1994: 2).
The
nurture
perspective:
Cognitive design
Within the
nurture
perspective, sexual
harass-
ment is viewed as
the result
of sex
roles
and stereotypes.
One
version of this
theory
views cognitive
biases as the
main cause
of sexual
harassment.
A second accords
this
role to negative
attitudes toward
women,
or
misogyny.
A
third
version
considers
both cognitive
and attitudinal
biases to
play
important
roles.
Representing
the cognitive-bias
perspec-
tive
is sex-role spillover
theory
(Gutek,
1985;
Gutek and
Morasch,
1982), which
regards
sexual
harassment
as behavior
guided
by
socialized
roles of
men as sexual agents
and
women as sexual
objects.
When the ratio of
men to
women in an occupational
context
is
highly skewed,
these sex roles
are confounded
with the
job.
Thus, secretaries,
elementary
school teachers,
and nurses
are viewed
as
sexual objects,
whereas
construction
workers,
fire fighters,
and engineers
are seen
as
sexual agents.
Sex-role
spillover theory
pre-
dicts that women
should
experience equally
high
levels of sexual
harassment
in both
male- and
female-dominated
occupations.
However, research
shows that
women are sex-
ually harassed
more in
male-dominated
than
female-dominated
work contexts
(Berdahl,
2007a;Fitzgerald
et al.,
1997; Glomb et
al.,
1999; Grubcr,
1998; Mansfield
et al.,
l99l).
It might
be the amount
of contact
a woman
has with
men, rather than
occupational
sex
ratios,
that best
predict
women's
likelihood
to be sexually
harassed
(Gutek
et al.,
1990;
Gruber,
1998).
A second
perspective is that negative
atti-
tudes
toward women
drive sexual
harassment.
SEXUAL
HARASSMENT
Theorists
have long argued
that sexual
harassment
is a
form ofhostility
and
aggres-
sion
toward
women in the
worþlace
(e.g.,
Farley,
1978; Franke,
1997;
MacKinnon,
1979; Schultz,
1998).
In a now-classic
study,
Pryor
(
1987) demonstrated
that
men who
held
negative
attitudes
toward
women,
and who
admitted
to being
likely to
rape a
woman if
they
could
get
away with
it, were
more likely
to sexually
harass a
woman
when
given
the
chance.
A third
perspective offers
a combination
of
the
first two:
Sexual
harassers
are motivated
by sex
roles and
sexist
hostility
(Fiske
and
Glick,
1995).
Based on
their theory
of
ambivalent
sexism,
Fiske
and Glick
(1995)
suggested
that:
(1)
unwanted
sexual attention
is mainly
motivated
by romant¡c
interest and
'benevolent'
sexist
beliefs
(i.e.,
those assuming
heterosexual
interdependence
and
complementarity);
(2)
gender
harassment
is mainly
motivated
by inter-
gender
competition
and hostile
sexist beliefs
(i.e.,
those assuming
female
malevolence
and
inferiority);
and
(3)
mostepisodes
of sexual
harassment
includesome
combination of
these behaviors
and
motives.
The
fact that sexual
coercion,
unwanted
sexual
attention,
and
gender harassment are
highly
correlated
(e.g.,
Fitzgerald
et al.,
1995a;
Schneider
et
al., 1997)
supports
this assertion
as
well as the
possibility that all
forms of
sexual
harassment
share
a common
root.
The
power
perspective:
Structural
design
The
power perspective views
sexual harass-
ment
as the
result of
power inequality
that enables
harassers
to sexually
coerce
and objectify
those
'beneath' them
in a
hierarchy
(e.g.,
Farley,
1978;
MacKinnon,
1979).
Power
inequality
facilitates
sexual
harassment,
and
sexual
harassment
reinforces
power inequality.
Advocates
of this
perspec-
tive
rarely
articulate
the direct
motives
of
harassers,
but
usually assume
that
harassers
are
motivated
by sexual
desire,
a desire
to
dominate
the
victim, orboth.As
Farley
(1978:
207) argued,
[F]emale
oppression
at
work is the
result of
nearly
universal
male
power
to
hire and
fire. Men control
the
means of economic
survival.
This control,
however,
is
also
used to
coerce
working
women
sexually.
Institutionalized
male
power
has thus
created
its
own
means
of
maintain¡ng
its superior
position.
Different types
of
power may enable
sexual
harassment
(Berdahl
et
al., 1996;
Cleveland
and
Kerst,
1993; Farley,
1978; MacKinnon,
1979).
A frequent
argument
is
that harassers
use
their
organizational
power
to
impose
their
sexual
will on
victims, as
in
quid pro
quo
harassment.
This
is a limited
view of
power,
however.
A
broade¡
view considers
power relationships
outside
the organization.
Women are usually
moreeconomically
depen-
dent
on men than
vice-versa.
Thus,
if a woman
does
not
please
her
male boss
(sexually
or
otherwise),
she
is less able
to support
herself
and must
depend
on a
man at
home; if a
woman
does
not
please
a
man at
home,
she is
made more
dependent
on
her male
boss,
and so
on.
This
pervasive economic
power
yields
another
type
that enables
men
to
sexually
harass
women: Social
power,
upheld
by societal
values
and beliels
about
men
and
women's appropriate
status,
roles,
and
inherent
worthiness.
With social
power,
a
man can
act as
a sexual
agent and
treat a
woman as
a sexual object
even
when
he lacks
organizational
or economic
power over
her.
Finally,
physical
power, or the ability
to
physically
intimidate and
dominate
someone,
enables
men to
sexually
harass
women.
Physical
power
often
seems so
obvious
that
it
gets
overlooked,
but
it may be the original
source
of men's
economic
and
social
power
over women
(Engels,
I 884),
and clearly
plays
a
role in sexual
violence.
Nurture
x
power:
Íhe social
identity
perspective
Combining
the
nurture and
power
perspec-
tives,
social
identity theories
of
sexual
harass-
ment emphasize
prescriptive stereotypes
THE SAGE HANDBOOK
OF
ORGANIZATIONAL
BEHAVIOR
(beliefs
about how
men and
women shottld
differ, rather
than how
they do differ)
and
their
motives. According
to this
viewpoint,
sexual
harassment
is a
mechanism
for
punishing
those
who threaten
a harasser's
gender
identity
and the benefits
derived
from
it
(Berdahl,
2007b;
Berdahl et
a1.,
1996;
Dall'Ara and
Maass,
1999; Franke,
1997;
Maass
et al., 2003;
Schultz,
1998).
Berdahl
(2007b)
proposes
that sexual
harassment
is
triggered
by the
harasser's desire
to
protect or
enhance
his
or
her sex-based
social
status in a
system of
gender hierarchy. Sexual
harassers
are more
likely to be
men, because
men
compared
to women
have
more to
gain
lrom
protecting their sex-based
status.
Harassers
can
protect or
define
their status
by derogating
another's
in a variety
of
ways, including
sexual
and non-sexual
harassment
targeted
at
members
of both sexes.
We
have considered
different
explanations
for what motivates
harassment,
from
nature to
nurture
to
power to social
identity. Different
explanations
have different
implications
for
who harasses whom.
WHO
HARASSES
WHOM?
Becausc
the fìrst
coutt c¿tscs
of' sexual
harassment
involved male
bosses
making
sexual cooperation
a condition
of women's
employment,
this became
the
prototype
of
sexual
harassment.
We now
know, howevet
that
this scenario
represents
a small
minor-
ity of
incidents: Co-workers,
subordinates,
customers,
and
clients
are often the
harass-
ment
perpetrators; men are
harassed based
on
sex; and same-sex
harassment
is
surprisingly
common.
Gender
Most sexual
harassment
is targeted
against
women. Only
10-14
per
cent
of sexual
harassment
cases
filed with the
EEOC are
filed
by men.
The 1994 USMSPB
study of
lederal
workers
lound that
more
women
(44
per
cent)
than men
(19
per
cent)
had experienced
any of
seven types
of sexual
behavior
in
the
past
two
years
at
work. Perpetrators
of
these sexual
behaviors
toward
women
were
almost
exclusively
men
(93
per
cent).
In
contrast,
men were
targeted by
both women
(65
per
cent)
and
men
(21
per
cent).
In
more recent
research
assessing
notjust sexual
behavior
but sexual
behavior
that
is offensive,
or unwanted,
a similar
pattern
has emerged.
Women are sexually
harassed
more than
men
(e.g.,
Cortina
et al.,
2OO2; Magley
et
al., I 999), especially
when researchers
only
count
negatively-appraised
behaviors
(e.g.,
Berdahl,
2O07a). Without
such adjustments,
some
studies
have found
men and
women
report similar
amounts
of sexual
experiences
at
work
(e.g.,
Berdahl,
2007a;
Konik and
Cortina,
in
press).
Systematic
research
into the
gender
of
both harassers
and
victims
has been
rare,
however.
Many studies
only
investigate
women's experiences,
and until the
late 1990s,
most surveys
only
asked about
behaviors
instigated by
men.
More research
is needed
to understand
same-sex
sexual
harassment
as
well as
non-sexual
forms
of
gender
harassment.
'Not
man enough'
harassment,
fbr example,
first
identifred
in male samples,
has
recently been
studied
in women as
well.
It appears
that women
too are teased
for being
'not tough enough'or
'overly
sensitive'
in
male-dominated
jobs (Berdahl
and Moore,
2006). Berdahl
(2007b)
predicts that when
different
forms ofharassment
based on sex
are
considered,
the
most common
pattern
should
be men
harassing
women,
followed by
men
harassing
men and
women harassing
women;
women
harassing
men should
be the
least
common.
These and other
questions
require
further
exploration.
Status
Early
research
into sexual
harassment
focused
on
sexual attention
and
coercion
from bosses
and supervisors.
Co-workers
were
included
when hostile
environment
harassment
was
recognized.
The stem
used
in the SEQ
was
limited to
'supervisors
or co-workers'
until
recently,
when
researchers
included subor-
dinates,
customers,
and
anyone
else in the
SEXUAL
HARASSMENT
477
work
environment
(e.g.,
Berdahl, 2003; Konik
and Cortina,
in
press).
In service-oriented
jobs
and
organizations, customers
and clients
are common sources
of sexual
harassment
(Barling
et al., 2001;
Berdahl, 2003; Gettman
and Gelfand,
2007). Research
has demon-
strated that
subordinates sometimes
sexually
harass their superiors,
though
rarely. This
type of
'contrapower'harassment
has mainly
been studied among
female
professors
who
experience
it from their
male students
(e.g.,
DeSouza and
Fansler, 2003). More
research is
needed
on harassment
from subordinates and
those outside
the organization.
Racelethnicity
To
date,
sexual harassment
research
has
paid
only limited attention
to issues of
race
and
ethnicity.
Ethnic
stereotypes,
numeri-
cal minority status,
cultural marginality,
and
economic
vulnerability should
theoretically
increase the
risk of sexual harassment
for
ethnic
minority women
(e.g.,
MacKinnon,
1979;
Murrell, 1996). That said,
most
empirical
research on
sexual harassment
has
focused on White/European
American
women. When ethnic differences
have been
examined,
findings have been
mixed. Earlier
large-scale studies
yielded
no differences
in the
harassment rates of
White and
non-
White women
(Gutek,
1985; USMBSP,
1987).
More recent research
reports higher rates
of
sexual harassment
against ethnic
minority
women compared
to White
women
(Berdahl
and Moore,
2006; Bergman and
Drasgow,
2003; Cortina et al.,
1998; Mansfield
et al.,
1991) and
men
(Berdahl
and
Moore, 2006).
In contrast, some surveys
found
Latina and
Black women to
report
significantly
lower
rates of harassment than their
non-Latina
White counterparts
(Shupe
et
a7., 2002;
Wy att
and Riederle,
1995). Each of these
studies
followed a different approach
to assessing
sexual
harassment, making comparisons
and
conclusions
dil'ficult.
The empirical
literature is virtually
silent
about
the race and
ethnicity of sexual
harassers,
possibly
due to the
intricacy of
this
issue. Perpetrators can
be members of
the
victim's own ethnic
group,
or
numerous
other
groups
of varying social
class and
organizational
power,
which can change
the
victim's subjective
experience
For example,
from the
perspective
of
a Black woman,
the
experience
of
being harassed
likely
'feels'
very different,
depending
on whether it comes
from White
men in
power, Black men
in
power,
Black men of
lower organizational
status,
or male
members of other
low-status
ethnic
groups.
More studies
are warranted
to
disentangle the
complexity of
race, class, and
power in sexual harassment.
OUTCOMES
OF SEXUAI
HARASSMENT
In the early
1990s,
researchers lamented
the
'appalling' lack of
systematic empirical
atten-
tion to sexual
harassment
outcomes
(Gutek
and
Koss, 1993:43).
This situation changed
dramatically
over
the decade that
followed, as
scientists
documented
myriad links between
sexual harassment
and victims'
occupational
functioning,
psychologicaUbehavioral
health,
and
physical health. Such outcome
relation-
ships
remained significant
even
when con-
trolling
for the experience of
other stressors
(e.g.,
general
job
stress,
trauma outside
of the
workplace),
other features
of the
job
(occupa-
tional
level, organizational
tenure,
workload),
personality
(negative
affectivity,
neuroticism,
narcissism), and
other demographic
factors
(age,
education
level,
race). Table
I
summa-
rizes findings of
this scholarship,
reviewed
below.
Occupational
outcomes
Not
surprisingly,
the organizational
psychology literature has
focused
primarily
on associations
between
sexual
ha¡assment
and
victims' occupational
well-being
(see
Table
1). In
particular,
over
20 articles
report
that sexual
harassment is associated
with
job
dissatisfaction
(for
a
meta-analytic
review,
see
Lapierre et al.,
2005). This
finding applies
to not
only White
American civilians,
but
also
US military
personnel,
ethnic
minority
women
in the US, and
women in other
nations
Table I Summary of
research
(from
the
mid-f 990s to the
present)
on outcomes
of sexual harassment.
An'X'indicates
that a sig_nificant
relationship
was f-ound between
sexuat harassment
and thàt outcome.
When the
harassment-outcome
relationship was
indirect
(i'e',
mediated
through
other variables),
th¡s ¡s noted in
parentheses
Key
Job outcomes
A Job satisfaction
E Actual and intended
turnover
C Work withdrawal
or neglect
D organizationalcommitment
E Productivity
or
performance
F Job stress
G other
Psychological outcomes
H Depression, anxiety,
or
general
distress
I
Post
lraumatic
stress disorder
I Other
psychological outcomes
Study
Job outcomes
Psvtholoo ¡ ca I outcom es Health
A
B
c
D E F
G Ht
)
Health
Barling
et al.
(1
996)
Earling et al.
(2001
)
Bergman and
Drasgow
(2003)
Bond et
al.
(2004)
Chan
et al.
(l
999)
Cortina, Fi?gerald
and Drasgow
(2002)
Cortina,
Lonsway
et al.
(2002)
Culbertson and
Rosenfeld
(1
994)
X
(indirect)
X
(indirea)
X
(indirect)
X
(indirect)
X
(indirect)
X
X
(indirect)
justice
perceptions
and cognitive
difficulties
X
(indirea)
X
(indirect)
X
(indirea)
negative mood
fear
(direct)
and
negative
mood
(indirect)
life
satisfaction
(indirect)
ange¡ disgust,
fear,
self-blame,
and low
self-esteem
X
(indirect)
X
X
X
Dansky
and
Kilpatrick
(1
997)
Fitzgerald
et al.
(1ee7)
Fitzgerald et
al.
(1
ege)
Fontana and
Rosenheck
(1
998)
Freels
et al.
(2005)
Glomb et al.
(1999)
Harned and
Fiugerald
(2002)
Harned et al
(2002)
Langhout et al.
(200s)
Lim and Cortina
(200s)
Magley et al.
(2005)
Magley
et al.
(1
999)
Monow et al.
(1
ee4)
Munson et al.
(2ooo)
0'Connell and
Korabik
(2000)
Parker and
Griffin
(2002)
Piotrkowski
('1998)
Ragins and
Scandura
(1
995)
X
(indirect)
X
(indirect)
X
(indirect)
X
(indirect)
X
(indirectl
X
(indirect)
X
(indirect)
X
(indirect)
X
(indirect)
X
(indirect)
X
X
X
life satisfaction
problem
drinking
life satisfaction
disordered
eating
(indirect),
self-esteem
and
self-blame
(direct)
life satisfaction
life
satisfaction
negative mood
X
(indirect)
X
(indirect
and direct)
X
(indirect)
X
(indirect
and
direct)
X
X
X
X
X
job
burnout
role ambiguity,
role
conflict
over-performance
demands
X
X
X
X
X
x
Table I €ont'd
Study
lob outcomes
Psych o I og i cal
o utcom
es
Health
A
B
c D
E
F
G H
Health
Raver
and Gelfand
(200s)
Richman et al.
(1
ege)
Richman et al.
(2002)
Richman et al.
(2006)
Rospenda
et al.
(2oos)
Schneider et
al.
(1
ee7)
Schneider
et al.
(2001
)
Shaffer et al.
(2000)
5hupe
et al.
(2002)
sims et al.
(2005)
usMsPB
f994)
Vogt
et
al.
(2005)
Wasti
et
al.
(2000)
Wislar
et al.
(2002)
Wolfe
et al.
(1998)
Woodzicka
and
LaFrance
(2005)
impaired team
relationships
and
cohesion; increased
team conflict
X
X
X
X
X
(indirect)
X
(indirea)
prescription
drug
use;
problem
drinking
problem
drinking
problem
drinking
XX
X
life satisfaction
XX
XX
X
life satisfaction
problem
drinking
X
X
X
X
(indirect)
SEXUAL
HARASSMENT
(e.g.,
Canada,
Mainland
China,
Hong
Kong,
Turkey).
Over
15 studies
have addressed
orga-
nizational
withdrawal
as
an
outcome
of
sexual
harassment.
Some
harassed
person-
nel engage
in work
withdrawal,
remaining
in
the organization
but
disengaging
from
work
(e.g.,
through
absenteeism,
tardiness,
work
neglect).
Others
manifest
more
com-
plete forms of
withdrawal,
through
turnover
thoughts
and
intentions
or actual
tumover.
Organizational
withdrawal
is often
conceptu-
alized
as
a way
of avoiding
further
exposure
to sexual
harassment
at
work.
Sexual
harassment
is also
associated
with
decrements
in employees'
organizational
commitment,
performance,
and
productivity.
Other
job-related
corelates
include
impaired
team
relationships,
incrcascd
tcam conflicts,
lowered
team
financial
performance,
lowercd
.iustice
perceptions, cognitive
di
Í'fìculties
(c.g',
distraction),
and over-perforrnance
demands
(i.e.,
the'need
to overperform
to
gain
accep-
tance
and
recognition
within the
workplace';
Parker
and
Grif'fìn,
2002).
These
studies
often
include
job
stress
as a covariate;
when
researchers
instead
conceptualize
job
stress
as
an outcome
in
its own
right, they
invariably
uncover
significant
direct
relationships
with
sexual
harassment.
Organizations
pay
a
price
for
these
out-
comes.
The USMSPB
used
a
'behavioral
costing
approach'
to attach
a
dollar
value
to sexual
harassment,
based
on
its large-
scale
surveys
of
federal
employees.
The
most recent
figures,
extrapolated
to
the entire
federal
workforce,
estimated
the
annual
cost
of
sexual
harassment
for the
US
government
to be
$327
million
(in
1994
dollars).
This
includes
costs
related
to employee
tumover,
employees'
self-reported
use of
sick
leave
due
to
harassment,
self-reported
individual
productivity
losses,
and estimated
workgroup
productivity
losses
(USMSPB, 1994). Costs
related
to
the
harasser's
lost
time
or
pro-
ductivity,
complaint
processing,
litigation,
or medical
and counseling
services
for the
victim are
excluded
fiom this
figure,
thereby
underestimating
the
cost of
sexual
harassment
to
the federal
government.
Psychological
and
physical
health
outcomes
Many
studies
(detailed
in
Table
1) of sexual
harassment
outcomes
have appeared
in
the
clinical
and
psychiatric literatures.
The more
thatemployees
experience
sexual
harassment,
the
more
that they
report
symptoms
of depres-
sion,
general stress
and anxiety,
posttraumatic
stress
disorder,
and overall
impaired
psycho-
logical
well-being.
In a series
ofarticles
based
on a4-wave
longitudinal
survey,
Richman
and
colleagues
documented
associations
between
ea¡lier
sexual
harassment
and
later alcohol
use
and
misuse.
Other
psychological
and
behavioral
correlates
include
negative
mood,
disordered
eating,
self-blame,
lowered
self-
esteem,
increased
prescription
drug
use,
anger,
disgust,
and
lowered satisfaction
with
life in
general.
Less
research
has
addressed
relationships
between
sexual
harassment
and
physical
health.
Such
effects
are
often
indirect,
mediated
through
mental
health.
Some
research
has
documented
links
to
overall
health
perceptions or
satisfaction.
Others
have
identified
specific
somatic
complaints
(headaches,
exhaustion,
sleep
problems,
gas-
tric
problems, nausea,
respiratory
complaints,
musculoskeletal
pain, and
weight
loss/gain)
associated
with experiencing
harassment.
In
the only
experiment
of its
kind,
Schneider
et
al.
(2001)
showed
that
exposure
to
mild
gender
harassment
leads to
increased
cardiovascular
reactivitY.
What
mitigates
or exacerhates
the
harm?
Employees
report
considerable
variability
in
the
outcomes
they
experience
from
sexual
harassment,
prompting
research
on
person
and
situation
factors
that
moderate
these
outcomes.
Searching
for
moderators
of this
relationship
has both
theoretical
and
applied
signifìcance,
isolating
which
populations
arc
most
at
risk for
harm,
under
which
circum-
stances,
and
what
might
be done
to
reduce
that
harm.
THE SAGE HANDBOOK OF ORGANIZATIONAL
BEHAVIOR
Victim
gender
The moderator
that
has received the most
empirical attention
is
gender:
when sexually
harassed, do women and
men experience
comparable consequences? Male targets of
unwanted sex-related behaviors
often report
that
these experiences were not anxiety-
provoking
(Berdahl,
Magley, and Waldo,
1996),'bothersome,''stressful'
(Berdahl,
z00ib), or
'upsetting'(Cochran et al., 1997).
In fact, some men describe these behaviors
as
'welcomed'
and even
'tun
and flattering'
(Berdahl,
2007b; Gutek, 1985). Moreover,
studies
find harassed women vs. men to report
worse outcomes, in terms of
negative mood
and turnover
intentions
(Barling
et al., 1996),
disordered eating
(Harned
and
Fitzgerald,
2002),
over-performance
demands
(Parker
and Griffin, 2OO2), and
longitudinal efl-ects
on anxiety,
problem
drinking,
job
stress,
job
burnout,
and
tumover intentions
(Freels
et al.,
2005, Magley, Cortina and
Kath, 2005). In
stark contrast with this
prior
work,
Vogt
et
al.
(2005)
reported
sexual
harassment to be
a stronger depression and anxiety
risk factor
for
men compared
to women.
Other research
reports
that,
when women
and
men
experience
similar rates of sexual
harassment, the impact
is
comparable.
For
instance, no sex differences were found
in
the relationship between sexual
harassment
and
various
job
outcomes
(Cortina
et a1.,
2002; Monow et al.,
1994),
psychological
and
physical
health outcomes
(Magley
et al.,
1999), and longitudinal
links to depression,
anxiety, hostility,
prescription
drug use, and
problem
drinking
(Richman
et a1., 1999,
2002, 2006). Morrow et al.
(1994)
also
described sex similarities
in relationships
between
supervisor harassment and
victims'
occupational stress and satisfaction.
Despite these mixed findings, the
weight
of the research evidence suggests
that women
face
greater
harm from sexual harassment than
men. Even studies that
report sex similarities
acknowledge
that women are far more
likely
than men to be sexually
harassed,
'thus
making
sexual
harassment a bigger and
more
harmful
problem
for women as a
group'
(Magley
et
al., 1999: 299). Other research
(e.g.,
Berdahl et al., 1996; Waldo et
al., 1998)
shows sexual
harassment to be a
qualitatively
different
phenomenon
for women and
men,
questioning
whether sex
comparisons in
outcomes should be conducted at all.
Victim race, ethnicity, and culture
Various writers
have
suggested
that minority
ethnicity
should amplify the negative
impact
of sexual harassment
(e.g.,
Fitzgerald et al.,
1995b; Munell, 1996; Shupe et a1.,2002).
The rationale
for
this
expectation is that
minorities face additional stressors beyond
sexual
harassment, such as racism and
racial
harassment, economic
hardship, and
(for
recent
immigrants) lack of adequate support
networks.
To
date,
however, little research
has
directly
compared outcomes
for minority and non-
minority women, and ûndings
have
been
mixed.
In
comparing the
experiences of Latina
and
non-Latina White women, Shupe et
al.
(2002)
reported that Latinas fared worse
in terms of work and coworker
satisfac-
tion,
whereas
ellects
on turnover intentions
were stronger for
Whites;
culture
did not
moderate relationship between harassment
and supervisor satisfaction,
psychological
dis-
tress, or
psychological
well-being.
Bergman
and Drasgow
(2@3)
compared
harassment
outcomes across memhers
of five
ditïerent
ethnic
groups in
the
US Armed Forces,
finding no evidencc that ethnicity
moderates
relationships between sexual
harassment and
occupational,
psychological,
or health-related
outcomes.
Likewise,
Piotrkowski
(1998)
reportetl no
moderating influence ol minority
status
on
effects of
gender
harassment on
job
satislaction and
'di.stress.'It
is di{ficult to draw
definitive
conclusions
fiom
these
limited,
divergent
findings
on
ethnic difïerences
in
sexual harassment outcomes.
Even
less is known about cross-national
differences
(or
similarities)
in
the
experi-
ence of
sexual harassment outcomes.
Wasti
and colleagues
(2000)
compared
employed
women in
Turkey and the US, reporting
a similar
pattern
of
harassment outcomes
in both
populations.
Shaffer et 41.
(2000)
found no differences among US,
Chinese
SEXUAL
HARASSMENT
443
Mainland, and Hong
Kong Chinese
women
in
the impact
of harassment
on
job
satisfaction
and
tumover
intentions, but culture
did
moderate
relationships with
organizational
commitment.
Beyond these two studies,
we
could locate
no
other
recent,
rigorous research
that directly
compares
harassment-outcome
relationships
across nations,
so this remains
an area
ripe for further
research.
Victim selflabeling
Among
women who report
unwanted sex-
related behavior
in
the
workplace,
fewer than
20
per
cent
typically
label their experiences
as
'sexual
harassment'
per
se
(Magley
et al.,
1999).
Nevertheless,
regardless of
whether
victims self-label or
not, they
report a similar
pattern
of
negative occupational,
psycholog-
ical, and
physical
outcomes
(Magley
et a1.,
1999; Munson et al.,
200 I
).
In an experimental
study, Woodzicka
and
LaFrance
(2005)
demonstrated
that even brief,
subtly sexually
harassing behaviors
lead to
impaired
perfor-
mance in victims,
irrespective of
what they
call these behaviors.
These studies
suggest
that
labeling does
not moderate
the link
between sexual
harassment
and outcomes.
Perpetrator
power
Although sexual
harassment
has adverse
effects
whether
perpetrated
by
peers
or
superiors
(e.g.
Morrow etal.,1994),
research
suggests that
harassment
'from
above'is
more
harmful. Cortina
eL al.
(2002)
and
Langh-
out
and colleagucs
(2005)
fbund signifìcant
correlations
between
perpetrator
power/status
and victim
perceptions
that the
harassment
was severe,
upsetting,
and frightening.
In
addition, O'Connell
and
Korabik
(2000)
and
Morrow et al.
(1994)
analyzed outcomes
of harassment
from higher- and
equal-level
perpetrators
separately;
they reported
more
numerous
negative effects of
the former.
O'Connell
and Korabik
(2000)
also
investi-
gated
women's
experiences of sexual
harass-
ment
from lower-status
men
('contrapower
harassment'),
finding no negative outcomes
at
all.
Explanations
for
the
greater
consequences
associated with
top-down sexual
harassment
emphasize
the victim's
heightened experience
ol helplessness
and fear
(e.g.,
Cortina et
al.,
2002; Langhout et al.,
2005).
Support
of the victim
To
generate
practical
recommendations
for
organizations,
some studies
have
investigated
whether social and
organizational
supports
mitigate
the impact of
sexual
harassment.
Several
studies
have found
that military
women's
perceptions of leaders
as fair,
supportive, trustworthy,
and
intolerant of
sexual
harassment
were related to
higherjob
satisfaction
and
organizational
commitment
and
lower tumover
intentions
(Muny
et al.,
2001
;
Offerman
and Malamut,
2002; Williams
et al.,
1999). Likewise,
Bond and
colleagues
(2004)
and
Cortina
(2004)
reported that
pos-
itive social support
from leaders,
co-workers,
friends, and
family attenuates
effects ofsexual
harassment on
women's
job
satisfaction.
Mixed
results
have emerged
regarding
the
benefits
of social support
for victims'
mental
health. Cortina
(2004)
found no
moderating
impact of
positive
support
on
victims'
anxiety and
depression,
whereas
Bond
and colleagucs
(2004)
did
find such
¿rn cffect.
Thc divergent
flndings could
result
from disparate
methodologies
(e.g.,
sample
composition,
measurement
of harassment,
operationalization
of support).
More research
is clearly
needed to understand
what types
of
suppoi-.t
can benefit
which harassment
vicLims,
and under
what conditions.
COPING
WITH SEXUAT
HARASSMENT
Compared
to
the 1980s, studies
ol coping
with sexual
harassment
have been
relatively
scarce
in the
past
decade.
We summarize
this research
below and
in Table
2. Excluded
from this
review are
'analogue'
studies,
in
which
participants
(often
college
students
with
limited
work experience)
read brief
harassing
scenarios
and report
how they
would
respond
r/ confronted
with such a
situation.
This method
is known to be
highly
flawed:
how
individuals say
they would think,
feel or behave
in
response to
hypothetical
sexual
harassment
fhils to
reflect the reality
THE SAGE
HANDBOOK
OF ORGANIZATIONAL
BEHAVIOR
of
how sexually
harassed
individuals
actu-
ally
respond in
real life.
In
particular, the
analogue
method
yields inflated estimates
of assertive
or confrontational
coping
(e.g.,
Fitzgerald et al.,
1995c;
Gutek and
Koss,
1993;
Woodzicka and
LaFrance,
2001).
The nature and
antecedents
of
harassment
coping
Reporting
Of
all
potential
responses
to
sexual
harassment,
intra-organizational
reporting
has
received
most research
attention,
reflecting
increasing
emphases
by American
employers
and courts on
organizationalreportingas
the
key mechanism
for eliminating
workplace
sexual harassment
(Burlington
Industries
v.
Ellereth,
1998; Faragher
v. City
of Boca
Raton,
1998). Some
social scientists
allege
that
reporting
is the most
appropriate
or
effective
means of
coping
with sexual
harassment
(e.g.,
Knapp
et al.,
1997; Reese
and
Lindenberg,199'l).
This claim,
however,
has little empirical
basis.
On the
contrary,
various
studies
(described
below)
have
revealed that
harassment
reporting can
give
rise
to additional
problems that exacerbate
the situation
for the
victim.
According to
victims'accounts
of how
they
responded
to
previous experiences
of sexual
harassment,
fewer than one-third
of
victims
inlormally
discuss sexual
harassment
with
supcrvisors,
and less than
25
per
ccnt
fìlc
tbr-
mal sexual
harassment complaints
with their
employers
(see
Table
2). Moreover,
only a tiny
minority of
victims take
their complaints
to
court.
Victims typically
only
turn to
formal
reporting
(internal
or external
to the organi-
zation)
after they
have exhausted
all other
response options.
Employees'
reluctance
to
report experiences
of sexual
harassment
is
primarily
attributed
to
fear
-
fear of blame,
disbeliel
inaction,
retaliation,
humiliation,
ostracism,
and
damage to
one's career
and
reputation
(e.g.,
Cortina,
2004;
Fitzgetald
et
al.,
1995c; Wasti
and Cortina,
2003).
Victims'
fears of
reporting
are
well-
founded. Cortina
and Magley
(2003)
learned
that two-thirds
of employees
who spoke
out
against
workplace mistreatment
then
faced
some form
of retaliation.
Others
have
found
that
sexual
harassment
reporting
is often
followed
by organizational
indifference
or
trivialization
of
the harassment
complaint
as
well as
hostility and
reprisals
against
the
victim
(Bergman
et al.,2002;Lee
et
al., 2004).
Perhaps
it comes as
no surprise,
then,
that
victims
often
leave the complaint
process
with
a
greater
perception
of
organizational
injustice
(Adams-Roy
and
Barling,
1998).
Confrontation
Confronting
the
harasser
is less common
that
popular wisdom suggests.
Woodzicka
and
LaFrance
(2001)
conducted
an experiment
in
which 50job
applicants
were asked
questions
by a
male
interviewe¡ that
were
mildly
harassing
and clearly
inappropriate
in
a
job
interview
(e.g., 'Do
you have a boyfriend?',
'Do
pcople find
you
desirablc?').
Not tr single
woman
challenged
the
interviewer
about the
inappropriate
questions
or
refused to answer
them.
Among the
harassed
women
in
civilian
organizations
described
in Table2,
an average
of39
per
cent
had confronted
their harassers
in
some
way.
This coping
strategy
may be
more
prevalent
in
the
military,
where Culbertson
and
Rosenfeld
(1994)
found
72
per
cent
of
enlisted
women
and 54
per
cent of
lemale
offi cers confronting
their
perpetrators.
S oci a
I
-s
upport
seeki
n
g
A
more typical
response
to sexual
harassment
is to
rely on
informal social
support
from
colleagues,
friends, and
family
members.
In
the research
detailed
in Table
2, an avenge
of
one-third
ol victims
had discussed
the
harassing
situation
with family
members,
and
approximately
5G-70
per
cent
had sought
support
from friends
or trusted
others.
Avoidance,
denial,
and endurance
Illustrating
what might
be the
most
prevalent
response
to sexual
harassment
in the
work-
place
(e.g.,
Fitzgerald et
al.,1995; Gutek
and
Koss,
1993), harassed
employees
frequently
try
to avoid
the
perpetrator or the
harassing
context,
deny or downplay
the seriousness
Table
2 Summary
of research
(from
the mid-l990s
to the
present)
on the
prevalence
of
specific harassment
coping
strategies
Key
Coping
strategy
Filed
formal complaint
or
grievance
Talked w¡th supervisor,
manager
or union representative
'Reported'
the harassment
Confronted
the harasser in
some way
5 Tâlked with friend or trusted other
6 Talked
with family
7
Avoided harasser
I Denied
or downplayed
gravily
of situation
9
Attempted to ignore,
forget or endure
Note on
populat¡on
of study
Cochran
et al.
(1997)
-
University staff
and students
(male
and female)
Cortina
(2004)
-
Working Latinas
(different
companies)
Culbertson
and Rosenfield
(1
994)
-
Navy women
Schneider
et al.
(1
997)
-
Working women
(different
companies)
USMSPB
(1
994)
-
Federal workers
(male
and female)
1
3
Coping strategY
r
---l---t
2
61718e
Cochran
et al.
(1
ee7)
Cortina
(2004)
Culbertson
and
Rosenfeld
(1994)
Schneider et al.
(1ee7)
usMsPB
(1994)
2o/o
17-200/o
260/o
210/o
6-8%
24Yo ofenlisted,
1 9% of officers
6-130/0
17160/0
l3%
ofwomen,
8% of men
25%
72% of enlisted,
54%
of officers
33-57o/o
41
%
of
women,
23% of men
49-640/0
2717o/o
53-700/o
45o/o
75% of enlisted,
54% of officers
54-l
4o/o
33%
of women,
20% of men
49-1\Yo
45% ofwomen,
44% of men
THE SACE
HANDBOOK
OF ORGANIZATIONAL
BEHAVIOR
of
the situation,
or
simply
ignore or endure
it
if
possible
(see
Table
2). Women
in these
circumstances
often
hope that
ifthey evade
the
harasser
or fail to
show any
reaction,
he will
lose interest and leave
them
alone. Note
that
this
research only
represents
individuals
who
remain in their
jobs
despite
being
harassed,
typically
excluding
those
who
quit
or retire
due
to the
harassment.
Some
writers
have criticized
avoidance,
denial, and endurance
as
passive, unassertive,
or otherwise
undesirable
reactions
to sexual
harassment
(e.g.,
Knapp et al.,
1997; Gruber
and Smith,
1995). Judges,
juries,
and
the
media and
lay
public often
interpret the
lack of
vocal
protest
(preferably,
a formal
complaint)
as evidence
that
the woman
consented
to,
'welcomed,'
or fabricated
the
inappropriate
conduct.
This was
clear in the
case of
Anita
Hill, whose
credibility
was assailed
because
she had
not formally
complained
about
Clarence
Thomas
at the time
of the
alleged
harassment.
If the
harassment
had
really
happened,
Senators
reasoned,
she
would
have
reported
it.
However,
avoidance,
denial
and
'doing
nothing'
often
reflect deliberate
attempts
to
extinguish
the
harasser's
behavior
by refusing
to reinforce
it. Such
reactions
appear
quite
reasonable
when a
woman
fears for
hersell
or her
job,
has
no other effective
response
options
available,
or
seeks to bring
an
end
to the
harassment
without
'rocking
the boat'
(Fitzgerald
et
al., 1995c;
Magley,
2002).
Thus,
these behaviors
represent
a common,
albeit
quiet
method
of resisting
sexual
harassment;
they are
the only
coping
responses
available
to some women in some
situations.
Antecedenß
of coping
Generally
speaking,
as
sexual
harassment
becomes
more severe
(i.e.,
upsetting,
disrup-
tive,
enduring,
frequent),
attempts
to ignore
it decrease
while efforts
to avoid,
confront,
report,
and seek
social
support
increase
(Bergman
et a1.,2002;
Cochran
et al.,
1997:'
Cortina,
2004;
Malamut
and Offermann,
2001; Wasti
and Cortina,
2003). Victims
are
also
more likely
to report
the situation
and
seek social
support
when being
harassed
by
an authority
frgure
(Bergman
et
al., 2002',
Cochran
etal.,
1997;
Conina,
2004;
Malamut
and Offermann,
2001).
Reporting
and con-
lrontation
are
more common
among
victims
who are
lower
in occupational
status,
female,
or
WhiteÆuropean
American
(Malamut
and
Offermann,
2001;
Rudman et
al., 1995;
Wasti
and Cortina,
2003),
whereas avoidance
and
denial
are
more
frequent among
women
from
traditional,
patriarchal, collectivist
cultures
(Wasti
and
Cortina,
2003).
ETIMINATING
SEXUAL
HARASSMENT
IN ORGANIZATIONS
Given
the negative
consequences
-
personal,
legal, antl
financial
-
that
accompany
scxual
harassment,
many organizations
have
taken
steps
to eliminate
harassment
where
possible,
and
correct
it
where necessary.
We now
review
and critique
harassment
policies,
pro-
cedures, and
training
programs implemented
by organizations.
Sexual
harassment
policies
Today,
most
large US organizations
and
many smaller
ones
have
policies
prohibiting
sexual
harassment
and specifying
reporting
procedures.
The
policies
vary
with
respect
to content,
but
often include
language
from
thc EEOC's
(1980)
definition
of scxual
harassment,
quoted
earlier.
This dcflnition,
however,
has been
criticized
as overly
vague
(Gutek,
1991), and
it
privileges
sexualizecl
actions
while
neglecting
the more
common,
s¿*isr
forms
of hostility
(Schultz,
1998;
2003).
Experts agree
that sexual
harassment
poli-
cies
should
clearly delineate
grievance
proce-
dures.
Some add
that
policies
should
explain
disciplinary
actions
that harassers
might face;
prohibit
retaliation
against
complainants;
and
provide
safeguards
for the
confidentiality
of all
parties. Several
scholars
emphasize
that
supervisors
and
top
management
must
be committed
to and
well-trained
on these
policies for them to
be effective
(Gutek,
1997;
Gruber,
1998;
Reese and
Lindenberg,1997;
Riger, 1991
;
Rowe,
1996; Stokes
et al.,
2000).
SEXUAL HARASSMENT
4a7
Few empirical
studies
have
evaluated
the
impact of
such
policy-making. One notable
exception
is Gruber
(1998),
who
found
that
employees
reported
the
lowest
rates
of sexual
harassment
when
they
worked
for
organizations
that
proactively
developed,
disseminated,
and enforced
the sexual
harass-
ment
policy
(e.g.,
by training
all employees,
creating official
complaint
procedures, des-
ignating
a specialist
to
receive complaints).
Significantly
more
harassment
was
reported
by
personnel whose companies
only used
informational
approaches
to
policy
dissem-
ination
(e.g.,
posting it in the
workplace
or employee
handbook), and
individuals
in
workplaces
with no
policy
at all
described
the
most sexual
harassment.
In
a
unique
study of
factors related
to men's
reports
of
harassment
perpetration, Dekker and
Barling
(1998)
found that
men who
perceived
strong
sanctions
against sexual
harassment
in their
organization
reported engaging
in less fre-
quent harassment of
others.
S exual
h arassm
e nt
co m
pl
ai
nt
procedures
In some
workplaces,
the only
procedures
available
for reporting sexual
harassment
are
formal,
requiring
victims
to lodge
written,
signed complaints
against
their
harassers;
thc
organization
then typically
notifies the
harasser
of the complaint
and
conducts
an
investigation.
Some
companies
have spe-
cialized
personnel for these
investigations;
many do
not. Organizations
also
differ
in the
standard
of
proof
used to determine
whether
sexual
harassment
has taken
place.
Some
rely on the civil
standard
of
'preponderance
of evidence'
(i.e.,
is it
'more likely
than
not'that
harassment
occurred),
the
standard
used by the
courts
in Title
VII cases.
Other
companies
use
the more
stringent criminal
standard of
'beyond
a
reasonable
doubt,'
the
highest level
of
proof required to
win a case
in court;
because sexual
harassment
is not
considered
a criminal
offense
under
US law,
it is
peculiar
to apply
this criminal
standard
to
these investigations.
Penalties
imposed on
employees
lound
guilty
of sexual
harassment
also vary
widely
across organizations.
Some
grievance
procedures offer
the
possibility of
appeal,
others do
not. Companies
typically
maintain
formal
records of the
complaint
and outcome,
including
names of
all
parties
involved
(Gutek,
1997;
Riger,
1991;
Rowe,
r996).
Formal
grievance mechanisms
have
dis-
tinct advantages,
allowing
for otficial
sanc-
tions
to be
imposed,
repeat offenders
to
be
tracked,
and managers
to be
held accountable
(Rowe,
1996).
However,
these
procedures
are often
adversarial,
with
the complainant's
perspective
potentially competing
against
that
of the accused,
his union
representatives,
and
management.
Such
procedures frequently
fail
to end
the harassment,
sometimes
worsen
the
situation,
and rarely
protect the complainant's
privacy
(e.g.,
Cortina
and Magle¡
2003;
Gutek,
1997; Rieer,
1991).
Because
of these
drawbacks,
some
experts
recommend
that informal
dispute
resolution
also be available
to sexually
harassed
employ-
ees
(Gutek,
1997;
Rigea l99l;
Rowe,
1996).
Many
victims do
not want
to lodge
a formal
complaint,
set
an
investigation
into
motion, or
see their
harasser
punished; they simply
want
the
offensive
behavior
to end
(Fitzgerald
et
al.,
1995c).
Informal dispute
resolution
could
involve,
for example,
someone
speaking
with
the offender
on
behalf of
the complainant
or
a
neutral
third
party
mediating
discussions
between
them.
The
goal
is
generally not
to
determine
guilt
or impose
punishment,
but
rather to
restore
peaceful co-existence
between
the
parties.
Often
no formal
records
are
kept, and
participation is
purely
voluntary
(Riger,
l99l; Rowe,
1996).
Outcomes
can
include an agreement
to change
behavior,
an
apology,
a voluntary
transfer
or resignation
of
either
party,
or
nothing
at all
(Gutek,
1997).
Opinions
of
informal
methods
for handling
sexual
harassment
have
been
mixed.
Informal
processes
provide a more
accessible
and
realistic
option
for
harassment
victims
who
wish to
avoid
formal
investigation
and
adjudication
(Riger,
l99l; Rowe,
1996),
and
they tend
to be
less
public,
confrontational,
and
litigious
(Gutek,
1997).
However,
if the
aim of
informal
procedures is not to
establish
THE SAGE
HANDBOOK OF ORGANIZATIONAL
BEHAVIOR
guilt
or
punish
the offender, this
process
will
not necessarily
deter would-be
harassers. If a
third
party
mediator
is involved, that
person
typically
must remain neutral, lacks authority
over
the harasser, and cannot
protect
the
victim from
retaliation
(Riger,
1991).
A
general
recommendation about
harass-
ment complaint
procedures
is that choices
be
available
to complainants,
including a choice
among multiple
procedures
(both
informal
and formal), and choices
among multiple
'complaint
handlers' with different ethnici-
ties,
sexes, and
positions in
the
organization
(Gutek,
1997; Reese and Lindenberg,2004;
Rigea l99l; Rowe,
1996; Stokes et al., 2000).
Sexual
harassment training
As with reporting
procedures,
sexual harass-
ment
training
initiatives also
vary tremen-
dously. Some organizations
mandate training
for
all employees;
others train only
man-
agers, supervisors,
complaint handlers, or
employees
found
guilty
ofsexual
harassment;
still others offer
no training at all.
The
trainer may be a manager,
HR employee,
compliance
officer,
EEO specialist,
sexual
harassment expert, or attorney.
Different
training
programs
use different
formats,
including lectures, speeches
from organiza-
tional
leaders, skits or
plays
with
professional
actors, behavioral
modeling, role-plays and
other experiential
exercises, computer-based
programs,
fllms, readings, and casc-studies.
The length of these diffe¡ent
programs ranges
from minutes to hours to days
(e.g.,
Bingham
and Scherer,
2001;
Gutek,
1997). In a
recent
survey
of 1,277 working adults
in
the
US,
Magley and colleagues
(2004)
found that only
46
per
cent of respondents
had received any
training at all on sexual
harassment, which
was more common in larger organizations
and
lasted an average of
1.5 hours.
Content
also differs across
sexual
harassment training
programs. Many
experts
agree that training should
include
education about what constitutes
harassing
conduct
and how employees can
report such
conduct. Some
programs
are oriented around
awareness-raising
or sensitivity+raining,
whereas others
focus more on legal issues
and
penalties
for harassers
(Bingham
and
Scherer,
2001; Gutek,
1997; Magley et al.,
1997; Reese and
Lindenberg,
1997). Some
training uses'rational-empirical'
techniques
(assuming
that
people
are more
likely to
change
behavior when
given
a rational
justifrcation),
while others use
'power-
coercive'
strategies
(assuming
that a threat
from a legitimate
authority will
promote
behavior
change).
Yet another approach
is
'normative
re-educative,' the assumption
being
that substantive
change requires
the development
of
new norms, shared
meanings, and transformations
in attitudes,
values, skills,
and relational styles
(Bingham
and Scherer,
2001). The
primary
goal
of
such
training is
generally prevention,
but
little empirical evidence
shows that training
actually
deters would-be
harassers
from
abusing
others. Another
goal
of training is to
encourage
employees
to come forward
with
intemal complaints of
sexual harassment,
but as
noted earlier, such complaints
are
rare. One longitudinal
study, however, did
document
an increase
in internal sexual
harassment complaints
following company-
wide training
(Magley
et al.,
1997). What
else do sexual
harassment training
programs
accomplish?
Researchers
have found that recipients
of sexual
harassment training
(particularly
men) report
increased knowledge of
sex-
ual
harassment definitions,
legal regulations,
and
organizational
policies
(Antecol
and
Cobb-Clark,
2003;
Bingham and Scherer,
2001). Other outcomes
include satisfaction
with
the organization's
harassment
policy
or
complaint
procedures
(Magley
et
al., 2004;
Reese
and
Lindenberg,
2004), lowered victim-
blaming or
harassment{rivializing
attitudes
(Lonsway
et a1.,
2008; Magley et al.,2004),
and
greater
belief
that sexual behavior
is
inappropriate
in the
worþlace
(Bingham
and Scherer,
2001). The USMSPB
(1994)
found that employees
working in lederal
agencies
providing
sexual harassment training
described
less
'uninvited
and unwanted sexual
attention.'
This effect was strongest
lor
agencies that
trained ø/l employees.
SEXUAL
HARASSMENT
We could
locate
only one
study
providing
direct
evidence
that sexual
harassment
train-
ing affects
men's behavior
toward
women,
at
least in the
short run.
Perry et al.
(1998)
showed
male
participants
a training
video on
either
sexual
harassment
or sign
language,
followed by
a
golftraining video.
Participants
then
trained
a female confederate
on how to
putt.
The
researchers
found that
harassment
training
increased
knowledge
and
reduced
inappropriate
touching
for
men with a
prior
propensity
to sexually
harass women.
Some
research,
however,
has lound
sexual
harassment
training
programs
to
have either
null effects
(e.g.,
Magley
et al.,
1997,
2004) or adverse
effects
on employees.
Magley
and colleagues
(1997)
and
Antecol
and
Cobb-Clark
(2003)
reported
that
some
trained employees
became
more cynical
about
their
organization's
ability or
commitment
to
prevent sexual
harassment,
and
Bingham
and Scherer
(2001)
found male
trainees
to report
greater victim-blaming
attitudes
and
/¿ss willingness
to
file a complaint
of
sexual
harassment, compared
to women
and
non-trained
men. Moreover,
the
short-term
attitudinal
changes
reported
by Perry
et al.
(
1998) did
not
persist
over
the long
term.
It is
important to
note,
however, that
each
of
these studies
evaluated
different
sexual
harassment
training
programs. Moreover,
the
methodological
quality
of
this
research
varied
(Magley
et al.,
1997), often
lacking
control
groups,
utilizing
small samples,
failing to take
into account
pre-training
assessment
effects,
or
lacking
pre-training assessment
altogether.
Most
post-training assessments
were con-
ducted
immediately
after
the training,
making
it
impossible to
know whether
the training
has
any
lasting effects.
More research
is clearly
warranted
to demonstrate
empirically
the
effectiveness
of sexual
harassment
training
interventions.
THE
FUTURE OF
SEXUAT
HARASSMENT
RESEARCH
As this
chapter
makes clear, sexual
harassment
remains a serious
problem
that takes
a toll
on
employees,
workgroups,
and their
organiza-
tions.
Notabl¡
the bulk of
research
on this
topic
has appeared
outside
of the
mainstream
organizational
literature. Conducting
searches
on
the keyword
'sexual
harassment'
in all
issues
published through
2006
of the
top five
OB and
VO
journals,
we
found only
31 sexual
harassment
articles:
.
Journal
of Applied
Psychology
(l
8 articles);
.
Organizat¡onal
Behavior
and Human
Decision
Processes
(3
articles);
.
Acadeny
of Management
tournal
(6
articles);
.
Acadenyof
Management
Revíew
(3
articles);
and
c
Administrative
Sciences
Quarterly
(l
article).
By contrast,
I 83 articles
on
workplace
sexual
harassment
havc appearcd
in
fivc social
science
journals
that do
not specialize
in
organizational
behavior:
.
Sex Roles
(100
articles);
.
Psychology
of Women
Quarterly
(31
articles);
o
lournal of
Social lssues
(24
articles);
.
Gender
and Society
(19
articles);
and
.
Signs:
The lournal
of Women
in
Culture and
Society
(9
articles).
The
topic of
workplace
sexual
harassment
clearly
deserves
greater
attention
from orga-
nizational
scientists.
Below, we
propose three
questions that warrant
further
research,
and
conclude
with methodological
suggestions
to
enrich this
literature.
What constitutes'sexual
harassment'7
After
three
decades
of legal
decisions
and
social science
on sexual
harassment,
an
explicit
definition
remains elusive.
We con-
tinue to
discover
'new'
forms of
harassment,
which
have
long occurred
but
seldom
been
studied.
One such
form is extra-organizational
sexual
harassment,
i.e.,
harassment
lrom
customers,
clients,
or members
of
the
public.
These
organizational
outsiders
can
easily
target an employee
with sex-based
disparage-
ment
and unwanted
sexual
advances,
interfer-
ing with
that
person's work and
well-being
490
(Barling
et al.,
2001; Gettman
and Gelfand,
2007).
With the
recent expansion
of the ser-
vice
prolessions,'outsider
sexual
harassment'
deserves
more empirical
attention.
Non-sexual
forms of
harassment
are
also
being
increasingly
recognized.
Sexist
behav-
iors directed
at
women,
such as
jokes
about
women's
intelligence
and
comments
about
women
'not
belonging'
in certain
jobs,
represent
the
most common
manifestations
of sex-based
harassment;
these
behaviors,
however, are
often
overlooked
in
research,
the
law, and organizational
policy.
Likewise,
sexist
conduct
directed
against
men
(e.9.,
'not
man enough'
harassment)
is
rarely studied.
Not
only
should
science
focus
more
on
these
sexist but
non-sexualized
behaviors,
it
should also examine
sex-based
harassment
that
makes
no explicit
reference
to
gender,
such as
incivility,
sabotage,
and
th¡eats
directed
disproportionately
at
women
(and
some
men)
in the workplace
(Berdahl,2007b;
Berdahl,
2007c;
Cortina,
2008).
More sexual
harassment
research
should
also consider
how
gender intersects
with
other
social
identities.
Having
a
low-status
identity,
such
as being
ethnic
minority,
poor,
or
gay, may increase
an
individual's
risk
for sex-based
harassment
(e.g.,
Berdahl
and
Moore,
2006;
Konik and
Cortina,
in
press).
These
identities
may
also aflect
the
type
of
harassment
experienced,
because
sex-
based
disparagements
can take
different
forms
and
meanings
specific
to one's
ethnicity,
sexual
orientation,
or other
social
dimension.
Research
addressing
these
intersections
can
further
our
understanding
of
sexual
harass-
ment as
a tool that
reinforces
social
hierarchies
in the
workplace.
What
motivates
(or
inhihits)
sexual
harassment?
Sexual
harassers
were originally
assumed
to
be driven
by
a desire
for sexual
expression
or
gratification.
Men's
wish to dominate
or
control
women
was
later
proposed to
motivate
sexual
harassment.
Research
now considers
a
more basic
motive:
The desire
to
retain a
valued social
identity
and
attendant
benefits
in
THE SAGE
HANDBOOK
OF ORGANIZATIONAL
BEHAVIOR
a system
of
gender hierarchy.
Given
that this
theorizing
has continued
for several
decades,
it
is striking
to see
how
little
empirical
research
has focused
explicitly
on
harassers
(exceptions
include
Bargh et
al., 1995;
Dekker
and
Barling,
1998;
Perry
et al.,
1998;
Pryor,
1987;
Pryor et
al., 1993).
This is an
important
direction
for future
studies.
How can
organizations
eliminate
sexual
harassment?
Sexual
harassment
grievance mechanisms
have
limited efTþctiveness
and
efficiency,
as
they
attempt
to correct
harassment
by rooting
out
and
punishing individual
harassers,
and
place
the burden
of managing
misbehavior
on individual
victims.
Moreover,
grievance
procedures typically
fail
to address
broader
problems that
fuel hostile
wo¡k
environments.
Experts therefore
emphasize
the
futility
of
relying
primarily or solely
on
formal
victim
complaints
to coffect
workplace
harassment
(e.g.,
Fitzgerald
et
al.,
1995c;
Magley,
2002; Shultz,
2003).
Instead,
innovative
harassment-prevention
and
control
mecha-
nisms are
badly
needed;
below
are
several
examples.
One
novel approach
to managing
sexual
harassment
is bystander
intervention
(Bowes-
Sperry
and O'Leary-Kelly,
2005;
Rowe'
1996).
As
Bowes-sperry
and O'Leary-Kelly
(2005)
explain,
this
can take
a
variety of
forms.
Employees
who
witness
the
harass-
ment of
a co-worker
can
redirect
the
harasser,
remove
the victim,
or otherwise
interrupt
the situation.
Further, bystanders
can
provide
support
to
the victim,
bolstering
the victim's
resources
and
sense
of clarity
and
control.
Bystanders
can also
take
it upon
themselves
to confront
or
report
the
harasser,
as such
responses
may
be easier
for
nonvictims.
These
possibilities suggest
that organizations
should
train employees
on
how to
respond
not only
when they
personally experience
harassment,
but also
when
they
witness
the
harassment
of others.
Sexual
harassment
interventions
might
also
be embedded
in broaderinitiatives
to establish
a civil,
respectful
workplace.
For
instance,
SEXUAL
HARASSMENT
491
to
promote
civility,
experts
(e.g.,
Pearson
and Porath,
2004)
recommend that
senior
management
model appropriate,
respectful
workplacebehavior;
clearly
state expectations
of civility
in
mission statements
or
policy
manuals; and
educate
all employees
on
civility expectations.
Cortina
(2008)
adds that
organizational
practices
to
set norms of
civil-
ity should
explicitly
discuss equitable
respect
toward
women
and men
(and
Whites,
gays,
ethnic
minorities, etc.).
Leaders
should also
emphasize
that
unacceptable
conduct
includes
notjust overt
acts ofmisogyny,
obscenity,
and
sexual
aggression,
but also
subtle devaluation
and derision
of
members of either
sex.
This
integrated strategy
of embedding
harassment-
prevention
efforts
into
larger civility-
promotion
programs would attract
broader
audiences,
being
relevant to all
employees
(female
and
male) and avoiding
the
resistance
met by
interventions
that exclusively
target
sexual
harassment
(Cortina,
2008; Cortina
et a1.,2002;
Lim and
Cortina,
2005).
A fìnal intervention
aims to
prevent sexual
harassment
by overhauling
the structures
that
support
it.
The recommendation
itself
is
quite
simple:
employ
more women,
promote more
women, and
integrate more
women
into every
level of
the organization.
The
goal
should
be a'well-integrated,
structurally
egalitarian
workplace,'in
which
women and
men equally
share
power and authority
(Schultz,
2003:
2071). Supporting
this recommendation
is
empirical
research linking
male-skewed
sex
ratios to sexual
harassment
(e.g.,
Berdahl,
2OO7a;
Fitzgerald et
al., 1997;
Gruber,
1998), stereotyping,
and discrimination
(see
Kanter,
1977; Whitley
and Kite,
2006).
Organizational'desegregation'
may not
eradicate sexual
harassment
entirely,
but
it
can
reduce the culture
of
hypermasculinity
that
promotes
objectification,
devaluation,
and
aggression
toward
women
and
gender-nonconforming
men.
M eth odo
I ogical
reco m me nd ati
o ns
for
sexual harassment
research
Most
sexual
harassment
research
relies
on cross-sectional,
self-report
surveys,
with
findings restricted
to a single
level ofanalysis.
The importance
of
this work
cannot
be
overstated,
but
it is time to broaden
the
range
of
methods
employed
in this domain.
With some
notable exceptions
(Glomb
et
al.,
1999 Magley et
a7.,2005;
Rospenda
et al.,
2006),
few have
investigated
sex-
ual harassment
longitudinally.
Longitudi-
nal
methods can
address
questions about
causality, temporal
patteming,
and
the
persis-
tence
ofnegative
outcomes.
These approaches
also allow
researchers
to conceptualize
harassment
as
a dynamic
process that
'unfolds'
or has
'cascading effects'
over
time.
Moreover,
when self-report
research
temporally
separates
assessment
of the
pre-
dictor
(harassment)
and
criterion
(outcome)
variables,
concerns
about
mono-method
bias
are
lessened.
We also encourage
more experimental
studies
of sexual
harassment.
However,
we
do
not advocate
vignette-based
paradigms
in which
students
imagine
how they
might
interpret or
respond to
a hypothetical
scenario
(if
anything,
such
studies
should be
dis-
continued,
given
that
their data
have
ques-
tionable
validity).
Instead,
more
fruitful and
interesting
possibilities
lie in simulations
of
harassing behavior
in
the laboratory.
There
are ethical
limits to
such studies,
but
they
are
possible
(see
Maass et
al., 2003;
Pryor,
1987; Schneider
et
al.,
2001; Woodzicka
and
LaFrance,
2001) and
can
yield
unique
contributions
to the
field.
Sexual harassment
research should
also
incorporate
multiple
levels of analysis.
Most studies
have
revolved around
indi-
vidual
self-reports
of
perceptions,
experi-
ences,
responses,
etc.
Not only
does this
approach
have
potential
problems
with
mono-
method bias,
it typically
misses
group-
or
organizationallevel
antecedents
(e.g.,
work-
group gender attitudes)
and outcomes
(e.g.,
declines
in unit
productivity;
see
Raver
and
Gelfand,
2005, for a
notable exception).
We
therefore
recommend
that
multilevel methods
become
more customary
in
sexual
harassment
research,
addressing
processes at the
level
of
the individual,
team,
organization,
and
society.
THE SAGE
HANDBOOK OF ORGANIZATIONAL
BEHAVIOR
coNcrusloN
The
past
decade
has witnessed
great
strides
in research on sexual
harassment.
Much
has been
learned about
its
different
forms,
perpetrators,
and
victims, the contexts that
promote it,
and
its effects on
individuals
and
organizations.
No longer seen as
'just'
a
'women's
issue,' sexual
harassment is now
recognized as illegal and
immoral behavior
that harms
women, men, and
the
'bottom
line.' Despite these
knowledge
gains
and the
organizational
changes that
have accompa-
nied them, harassment
based on sex
remains
all too
common. More
research is clearly
needed to better understand
and
prevent
sexual
harassment,
helping
organizations
to
foster vibrant,
healthy, and
respectful work
environments.
NOTES
1 Some might
wonder about the division between
sexist
and sexual hostility
(Cortina,
2001;
Fitzgerald
et al.,
1999b), or the
gendeÊrole-deviation harass-
ment
(e.9.,
Berdahl