Content uploaded by Jon Overton
Author content
All content in this area was uploaded by Jon Overton on Jan 20, 2020
Content may be subject to copyright.
Who Wants to Lead?
Anticipated Gender
Discrimination Reduces
Women’s Leadership
Ambitions
Susan R. Fisk1and Jon Overton1
Abstract
We examine whether anticipated gender discrimination—specifically, gendered sanctions
for leadership failure—decreases women’s leadership ambitions. We find that laypeople
expect that women leaders will be punished more harshly for failure than otherwise similar
men. We also compare the leadership ambitions of women and men under conditions of
benign and costly failure and find that leadership roles with costly failure—which implic-
itly have the potential for gendered sanctions for failure—disproportionally depress wom-
en’s leadership ambitions relative to men’s. Anticipated sanctions for failure mediate this
effect, providing evidence that anticipated gender discrimination reduces women’s leader-
ship ambitions. These results illuminate microlevel foundations of the stalled revolution
by demonstrating how gendered beliefs about leadership are recreated, legitimized, and
contribute to the dearth of women leaders. These findings also suggest that organizational
responses to failure may produce gender differences in leadership ambitions and risk-
taking behavior.
Keywords
experimental social psychology, gender, gender discrimination, gender and leadership,
gendered sanctions for failure, organizations, stereotypes
Despite progress in women’s social stand-
ing and legal rights over the past century,
top leadership positions remain domi-
nated by men. Women constitute only 5
percent of Fortune 500 CEOs, 24 percent
of members of the 115th US Congress,
and 30 percent of college and university
presidents (Pew Research Center 2019).
Many factors contribute to the gender
gap in leadership. At a macro level,
women struggle to acquire and retain
leadership positions due to organizational
structures like workplace norms and cul-
tural schema connecting masculinity
and leadership (Eagly and Carli 2007).
At meso levels, women are disadvantaged
by interactional processes like the
double-bind (Phelan, Moss-Racusin, and
1Kent State University, Kent, OH, USA
Corresponding Author:
Susan Fisk, Department of Sociology, Kent State
University, PO Box 5190, Kent, OH 44242, USA.
Email: sfisk@kent.edu
Social Psychology Quarterly
2019, Vol. 82(3) 319–332
ÓAmerican Sociological Association 2019
DOI: 10.1177/0190272519863424
journals.sagepub.com/home/spq
Rudman 2008) and resistance to women’s
leadership (Vial, Napier, and Brescoll
2016). Yet despite the extent of this field
of scholarship, others argue that women
are not leaders because they do not
want to lead (Kenrick, Trost, and Sundie
2004; Reuben, Sapienza, and Zingales
2015) because there is a gender gap in
leadership ambitions (Barsh and Yee
2012; Eagan et al. 2016; Konrad et al.
2000).
In the present research, we bridge the
gap between these schools of thought.
We examine whether anticipated gender
discrimination—in the form of gen-
dered sanctions for leadership failure—
depresses women’s leadership ambitions.
While the punishments associated with
failing in a leadership role are not explic-
itly gendered, research finds that women
leaders face greater penalties and reputa-
tional losses for failure (Brescoll, Dawson,
and Uhlmann 2010; Eagly and Carli
2007; Kennedy, McDonnell, and Stephens
2016). We argue that women are aware of
these gendered sanctions for failing in
a leadership role and that this knowledge
depresses their leadership ambitions. We
also discuss how organizational struc-
tures can heighten or lessen these effects
and consider how this research adds to
our understanding of the causes of gender
differences in risk-taking behavior.
GENDER DIFFERENCES IN
LEADERSHIP AMBITIONS
Although gender differences in career
preferences have narrowed over time
with changing social conditions, a gender
gap in leadership ambitions remains
(Eagan et al. 2016). A representative sur-
vey of full-time American college fresh-
men found that from 1966 to 2015, men
have been more likely to rate their
leadership ability as high, prioritize
becoming an authority in their field,
and value becoming successfully self-
employed (Eagan et al. 2016). Surveys of
working women reach similar conclu-
sions: 36 percent of men but only 18 per-
cent of women report wanting to advance
into executive management (Barsh and
Yee 2012). Meta-analyses of psychological
research also find that men have a slightly
higher preference for leadership in their
occupations than women (Konrad et al.
2000).
Some researchers have argued that
gender differences in male-typed career
behaviors—including leadership ambi-
tions—are caused by stable sex differen-
ces in competitiveness and aggression
between women and men (Kenrick et al.
2004; Reuben et al. 2015). However, it
seems unlikely that biological sex differ-
ences are potent causes of the gender dif-
ferences in leadership ambitions given
that these differences have shrunk over
time with changing social conditions
(Eagan et al. 2016) and the half-century
of sociological research demonstrating
that gendered organizational structures
produce gendered leadership ambitions
(Cassirer and Reskin 2000; Damaske
2011; Kanter 1977).
GENDERED SANCTIONS FOR
LEADERSHIP FAILURE
There is also reason to believe that antic-
ipated gender discrimination depresses
women’s leadership ambitions. While
many gendered interactional processes
disadvantage women leaders (e.g., the
double-bind, backlash), we focus on one
particular kind of discrimination: gen-
dered sanctions for leadership failure.
We do this for two reasons.
First, women leaders face gendered
sanctions for failure as they receive
hasher punishments than otherwise sim-
ilar men. Leadership continues to be
a male-typed task (Eagly and Carli
2007), and women are punished more
severely than men for failure on male-
320 Social Psychology Quarterly 82(3)
typed tasks such as leadership (Brescoll
et al. 2010; Sarsons 2017). For instance,
women managers are formally penalized
more harshly for committing an ethical
violation than otherwise similar men
managers (Kennedy et al. 2016), women
CEOs are more likely than men CEOs to
face shareholder activism (Gupta et al.
2018) and suffer harsher consequences
(in regards to stock-based pay) when their
companies lose money (Albanesi, Olivetti,
and Prados 2015), and women surgeons
receive a sharper drop in referrals when
a patient dies than male surgeons (Sar-
sons 2017).
In addition, leaders who fail typically
suffer greater reputational losses if they
are women, further amplifying the gen-
dered sanctions associated with leader-
ship. On male-typed tasks, a woman’s
failure is more likely than a man’s failure
to be attributed to low ability rather than
to bad luck (Swim and Sanna 1996). In
this manner, women leaders in gender-
incongruent fields who make a mistake
are accorded less status and perceived
as significantly less competent than oth-
erwise similar men (Brescoll et al. 2010).
Moreover, imperfections become magni-
fied for women leaders (Eagly and Carli
2007), making women’s reputations
more tenuous than men’s (Williams and
Dempsey 2014). In sum, due to gendered
sanctions for leadership failure, women
leaders have more to lose relative to oth-
erwise similar men.
This brings us to the second reason we
focus on gendered sanctions for leader-
ship failure: potential losses are particu-
larly salient in decision-making pro-
cesses. The large literature on risk
aversion and negativity bias finds that
negative stimuli (i.e., punishments)
attract more attention and are more
influential in decision making compared
to positive stimuli (i.e., rewards; Bau-
meister et al. 2001; Kahneman and Tver-
sky 1979; Rozin and Rozyman 2001).
Thus, while important societal rewards
are attached to achieving success in
a leadership position, it is important to
look beyond likely outcomes and consider
the severity of possible bad outcomes to
predict leadership ambitions. This is
especially true for women, who are likely
to face harsher punishments for leader-
ship failure than otherwise similar men.
THE PRESENT RESEARCH
We argue that women are aware of the
gendered sanctions for leadership failure
and that this knowledge depresses their
leadership ambitions. We test our thesis
using three interconnected studies. Study
1 was a survey designed to establish that
laypeople are aware that women leaders
face gendered sanctions for failure. While
research lends support for this supposi-
tion (as people are acutely aware of norms
associated with their position in hierar-
chies; Anderson et al. 2006; Ridgeway
and Nakagawa 2017), the prevalence of
this belief (to the best of our knowledge)
has not yet been directly assessed. Thus,
Study 1 tests the following hypothesis:
Hypothesis 1: Men and women believe
that women leaders face greater sanc-
tions for failure than otherwise simi-
lar men leaders.
Study 2 was a survey-experiment that
examined how the costliness of leadership
failure (e.g., benign failure vs. costly fail-
ure) influences women’s and men’s lead-
ership ambitions. If women anticipate
gendered sanctions for leadership failure,
then leadership roles with costly failure
should disproportionally depress women’s
leadership ambitions relative to men’s
because these roles implicitly have the
potential for gendered sanctions for fail-
ure. We compare the change in leadership
ambitions of women and men under con-
ditions of benign and costly failure
Anticipated Discrimination Reduces Women’s Leadership Ambitions 321
(instead of directly comparing women’s
and men’s leadership ambitions) because
women and men hold different positions
in society due to the gender-segregated
nature of the labor force and controlling
for these differences is notoriously diffi-
cult. By focusing on change in leadership
ambitions between conditions of benign
versus costly failure, we are able to side-
step concerns about gender differences
in the types of positions held by women
and men. More formally, Study 2 tests
the following hypothesis:
Hypothesis 2: Women’s leadership ambi-
tions decrease more than men’s when
leadership failure is costly (instead of
benign).
Study 3 was a survey-experiment con-
necting the findings from Studies 1 and
2, allowing us to test whether costly fail-
ure disproportionately affects women’s
leadership ambitions because of expected
gender discrimination. Specifically, Study
3 tests the following hypothesis:
Hypothesis 3: Anticipated sanctions for
failure mediate the disproportionate
effect of costly failure on women’s
leadership ambitions.
METHODOLOGICAL APPROACH
Following prior studies (e.g., The
´baud
2015), we used survey-experimental
vignettes in Studies 2 and 3 to examine
how structural conditions shape atti-
tudes. An experimental approach was
ideal for our research question because
it allowed us to manipulate the costliness
of leadership failure and avoid potentially
gendered selection effects into organiza-
tions. Participants came from Amazon
Mechanical Turk (mTurk), an online mar-
ketplace for paid work in which a national
convenience sample of adults completes
tasks that only humans can do (e.g., com-
pleting surveys). Research finds that
mTurk samples are more diverse and rep-
resentative than samples obtained from
college campuses (Mason and Suri
2012), allowing us to understand the
leadership ambitions of everyday women
and men. This is important because lead-
ership opportunities are not limited to
white-collar professions, and vertical sex
segregation contributes to pay inequal-
ities between women and men from all
walks of life (Cohen and Huffman 2007).
STUDY 1: COMMON KNOWLEDGE
OF GENDERED SANCTIONS
FOR LEADERSHIP FAILURE
Study 1 Methods
Procedures. Participants were asked to
answer six 7-point Likert scale questions
(wherein 1 = strongly agree and 7 =
strongly disagree) about whether women
or men face greater sanctions for failing
in a leadership role.1The items were com-
bined into an index measuring expected
sanctions against women leaders, wherein
1 = women will be punished more harshly,
4 = no gender difference in treatment, and
7=menwillbepunishedmoreharshly
(a= .75). After answering these questions,
participants indicated whether they
identified as a man, woman, or neither
a man nor a woman.
Participants. Participants were recruited
from mTurk. One hundred and nine par-
ticipants were paid $0.30 to complete
a two-minute survey. No participants
failed the attention check. One respon-
dent who identified as neither a man
nor a woman was excluded from the anal-
yses because our analyses focus on com-
paring the attitudes of women and men.
The final sample was 51.4 percent women,
82.7 percent white, and 17.3 percent non-
white; 44.6 percent of participants had at
1See Supplement Part A, available with the
online version of the paper, for question wording.
322 Social Psychology Quarterly 82(3)
least a bachelor’s degree. The average age
was 34.6 years.
Study 1 Results and Discussion
Using one-sample ttests comparing the
mean values against the midpoint value
of 4, we find that women, t(55) = 26.55,
M= 3.23, SD = .12, p\.001, believe
that women leaders face greater sanc-
tions for failure than otherwise similar
men. In addition, men also hold this
belief, albeit less strongly, t(52) = 22.40,
M= 3.63, SD = .15, p\.05. This supports
Hypothesis 1: men and women believe
that women face discrimination in leader-
ship roles.
STUDY 2: COSTLINESS OF
LEADERSHIP FAILURE AND GENDER
DIFFERENCES IN LEADERSHIP
AMBITIONS
Study 2 Methods
Design. Participants were randomly
assigned to read one of two vignettes
about assuming a leadership position at
work that varied costliness of failure
(i.e., costly vs. benign failure). This led
to a 2 32 factorial design (man/woman
3costly failure/benign failure). The
vignettes were constructed to maintain
the control of experiments while also tap-
ping into ‘‘real-world’’ experiences. Spe-
cifically, all participants were told to
imagine that their company is going to
launch a new initiative, that the outcome
of this initiative was uncertain (i.e., it
might be a major failure or a great suc-
cess), that they have the opportunity to
be the leader of this new initiative, and
that leading a successful initiative would
strongly benefit their career.
In the benign failure condition, partic-
ipants were told that they could return to
their current team if the initiative failed
and that they had the full support of their
manager. For the condition to be globally
benign, it discussed both formal sanctions
(e.g., job security) and informal reputa-
tional losses (e.g., manager support). In
the costly failure condition, participants
were told that they might lose their job
if the initiative failed and that their man-
ager has given them neither direction nor
encouragement to take the position.
Moreover, a lack of explicit managerial
encouragement increases the cost of fail-
ure as there is no buffer to dampen a fail-
ing employee’s punishment: the failure
will be entirely the employee’s responsi-
bility. To maintain realism, we used
somewhat vague managerial language
as research finds that managers are
intentionally unclear in order to dodge
responsibility for failures (Jackall 1988;
Turco 2016).2We then asked participants
to write a few sentences in response to
three prompts that included: (1) the fac-
tors that would make the participant
both more or less likely to accept the lead-
ership position, (2) how strongly partici-
pants would advocate to be the leader,
and (3) how the situation would make
them feel. These questions existed to
engage participants, stimulate deep
thinking about the scenario, and serve
as an attention check.
Participants. Participants were paid $2 to
complete a 15-minute survey. No partici-
pants were excluded for inattention as
they all wrote appropriate answers to
the questions and correctly answered
the attention check. Two-hundred fifty-
seven mTurk workers participated in
the study, 56.4 percent of whom were
women.3Eighty-seven percent of the sam-
ple was white, and 55.2 percent had at
2See Supplement Part A, available with the
online version of the paper, for the full text of
the vignettes and for details about the robustness
of our manipulation.
3Binary gender categories were used in this
study even though gender is not binary
phenomenon.
Anticipated Discrimination Reduces Women’s Leadership Ambitions 323
least a bachelor’s degree. See Table 1 for
additional descriptive statistics.
Leadership Ambition (Dependent Variable).
Leadership ambition was measured as
a dummy variable, which was taken
from the response participants gave
about how strongly they would advocate
for themselves to be the leader. Respond-
ents could write ‘‘N/A’’ if they had no
interest in leading the team. Responses
were coded such that 1 = participant
wrote a response to the second open-
response question listed previously (as they
expressed some desire in becoming the
team leader) and 0 = N/A (as they expressed
no desire in becoming the team leader).
Controls. Although random assignment in
experimental methodologies typically
eliminates the need for controls, we
included controls for race, age, employ-
ment status, and education to ensure
that any observed gender effects were
caused by gender, as opposed to other dif-
ferences between women and men in the
sample. We dummy-coded race into
whites (reference category) and non-
whites and interacted race and gender
given that past research shows white
men tend to perceive situations as less
risky compared to white women, men of
color, and women of color (Finucane
et al. 2000). We also controlled for educa-
tion (with dummy categories for high
school, associate’s, bachelor’s, or advanced
degrees) because women are more likely
than men to pursue higher education
(U.S. Census Bureau 2017).
Study 2 Results and Discussion
Table 2 presents a series of nested linear
mixed probability models predicting will-
ingness to lead.4In Model A, we predict
willingness to lead using only gender
and costliness of failure, and in Model B,
we add the interaction of gender and cost-
liness of failure. Both models include con-
trols for race, education, age (linear and
squared term), and employment status
(dummy coded as employed or not).5
When we predict willingness to lead
using only gender and costliness of fail-
ure, we find that costly failure decreases
Table 1. Descriptive Statistics for Study 2 (N= 257)
Proportion/mean (SD) N
Costly failure .502 128
Leadership ambitions .743 191
Woman .564 145
Nonwhite .132 34
Education
High school diploma .300 77
Associate’s degree .148 38
Bachelor’s degree .350 90
Advanced degree .202 52
Age 32.023 (10.259) 257
Employed .872 224
4See Supplement Part B, available with the
online version of the paper, for an explanation
of our decision to use linear mixed probability
models.
5No controls significantly attenuate the inter-
action between gender and costliness of failure,
and no controls significantly predict leadership
ambitions. See Supplement Part C, available
with the online version of the paper, for control
coefficients.
324 Social Psychology Quarterly 82(3)
participants’ leadership ambitions by 28.1
percentage points (p\.001; Table 1,
Model A). However, when we add the
interaction of gender and costliness of
failure to the model (Table 1, Model B),
we find that women’s leadership ambi-
tions are disproportionally decreased (rel-
ative to men’s) when leadership failure is
costly as the interaction term between
gender and costliness of failure is statisti-
cally significant (p\.05; Table 2, Model
B). High costs of failure (relative to
benign failure) depress men’s leadership
ambitions by 16.7 percentage points
(p\.05) and depress women’s leadership
ambitions by an additional 20.3 percent-
age points (p\.05; Table 2, Model B).
This provides support for Hypothesis 2. It
isalsoworthnotingthatwefindnogender
differences in leadership ambitions when
failure is benign (b=.048,p= .512) (Table
2, Model B).6This reaches the benchmark
proposed by Frick (1995) of a pvalue of
.50 to infer statistical equivalence.
STUDY 3: DOES ANTICIPATED
DISCRIMINATION DEPRESS WOMEN’S
LEADERSHIP AMBITIONS?
Study 3 Methods
Study 3 used the same vignette, experi-
mental manipulations, and open-
response questions as Study 2. However,
Study 3 used a larger sample (to have
the statistical power to conduct a media-
tion analysis) and included a series of
questionnaire items after the vignettes to
provide support for Hypothesis 3 (antici-
pated sanctions for failure mediate the dis-
proportionate effect of costly failure on
women’s leadership ambitions) and rule
out alternative explanations for this effect.
Participants. Participants were recruited
from mTurk and were paid $1.10 to com-
plete a survey that took approximately 10
minutes to complete. Participants were
excluded for answering open-response
prompts with nonsensical answers (N =
7). A final sample of 484 mTurk workers
participated in the study, 57.9 percent of
whom were women.7Fifty-five percent
held a bachelor’s degree or higher. A
majority of the sample was white (70.2
percent non-Hispanic white). The mean
age was 35 years. See Table 3 for addi-
tional descriptive statistics.
Leadership Ambition (Dependent Variable).
Leadership ambitions were measured
using an index (a= .94) created from
three questions about the likelihood of
accepting, applying for, and advocating
for the leadership role (a= .94). All three
questions used 100-point slider bars,
where 0 = 0 percent chance and
Table 2. Linear Mixed Model Coefficients
Predicting Leadership Ambitions (N= 257)
Model A Model B
Costly failurea–.281***
(.053)
–.167*
(.078)
Woman –.046
(.056)
.048
(.074)
Costly failure 3
woman
–.203*
(.104)
Intercept .832***
(.245)
.798***
(.243)
Note: Standard errors appear in parentheses.
Additional controls are nonwhite, nonwhite 3
woman, associate’s degree, bachelor’s degree,
advanced degree, age, age2, and employed.
aReference category is benign failure.
*p\.05. ***p\.001.
6This result also provides evidence that gen-
der differences in obedience to authority do not
account for our results. Were this the case, we
would find that women would be more likely to
lead than men when they had the full support
of their manager and no gender differences in
leadership ambitions when the manager gave
the employee no direction on what to do.
7Participants received an open-response
option for gender identification, and all indicated
that they were either men or women.
Anticipated Discrimination Reduces Women’s Leadership Ambitions 325
100 = 100 percent chance.8We departed
from Study 2’s measure of leadership ambi-
tions to have a continuous measure of lead-
ership ambitions because binary dependent
variables are not well suited for mediation
analysis (MacKinnon and Dwyer 1993).
Anticipated Sanctions for Failure (Mediat-
ing Variable). To avoid creating a demand
effect wherein participants would deduce
and conform to our hypotheses, we did
not directly ask Study 3 participants
whether they believe that women leaders
faced gendered sanctions for failure or
whether they believed their gender would
influence their own sanctions for failing
in a leadership role. This is an especially
important concern with mTurk because
experienced workers are very familiar
with commonly used social research tech-
niques and can often guess the purpose of
a study (Chandler, Mueller, and Paolacci
2014). Instead, we measured participants’
anticipated sanctions for failure using
a six-item index designed to capture par-
ticipants’ beliefs about their likelihood of
experiencing formal and informal sanc-
tions (i.e., loss of respect manager and
co-workers, demotion, firing; a= .86).9
Higher likelihoods correspond to harsher
expected sanctions.
Controls. We used the same controls as in
Study 2 (race, education, age, and
employment status) and added controls
for income, relationship status, and par-
enthood status. We controlled for income
because women tend to earn less than
men (Bureau of Labor Statistics 2017)
and income may influence participants’
leadership ambitions. The average
income was $46,211 (Table 3).10 We also
Table 3. Descriptive Statistics for Study 3 (N= 484)
Proportion/average (SD) N
Costly failure .502 243
Anticipated sanctions 48.221 (21.151) 484
Leadership ambitions 64.717 (30.488) 484
Woman .579 280
Nonwhite .298 144
Education
High school diploma .265 128
Associate’s degree .186 90
Bachelor’s degree .380 184
Advanced degree .169 82
Age 35.043 (11.432) 484
Employed .862 417
Income 46,211.050 (34,282.05) 484
Log of income 10.412 (.907) 484
Relationship status
Single .355 172
Cohabiting .138 67
Married .409 198
Other .097 47
Parent .386 187
8See Supplement Part A, available with the
online version of the paper, for detailed question
wording.
9See Supplement Part A, available with the
online version of the paper, for question wording.
10See Supplement Part D, available with the
online version of the paper, for details.
326 Social Psychology Quarterly 82(3)
controlled for relationship status and the
interaction between gender and relation-
ship status, as relationship status may
differentially impact the leadership ambi-
tions of women and men due to gendered
marital expectations for husbands to be
breadwinners (Damaske 2011). Partici-
pants were coded as either single (35.5
percent; the reference category), cohabit-
ing (13.8 percent), married (40.9 percent),
or other (9.7 percent; participants were
coded as other if they indicated that
their relationship status was separated,
divorced, widowed, or something else;
Table 3). Lastly, we controlled for parent-
hood status and the interaction between
gender and parenthood status, because
gendered prescriptions for breadwinning
are especially acute for fathers (Brescoll
and Uhlmann 2005), which may increase
the leadership aspirations of fathers.
Conversely, mothers may feel pressure
to rein in their leadership ambitions to
balance work and family (Damaske
2011). Respondents were coded as either
parents (if they had one or more children;
38.6 percent) or nonparents (Table 3).
Study 3 Results and Discussion
Costliness of Failure and Leadership Ambi-
tions. As in Study 2, we find that costly
failure depresses leadership ambitions
for both men and women (p\.001) and
that women do not have significantly differ-
ent leadership ambitions than men when
failure is benign. Most importantly, Study
3 also replicates the main findings from
Study 2 as we find a statistically significant
interaction of gender and costliness of fail-
ure. Costly failure hampers the leadership
ambitions of both men and women but dis-
proportionately reduces women’s leadership
ambitions. Focusing on Model B in Table 4
(which includes racial, educational, socio-
economic, and family controls), we find
that costly failure depresses men’s leader-
ship ambitions by about 17 percentage
points (p\.001) and that the interaction
between gender and costliness of failure is
statistically significant (p\.05), reducing
women’s leadership ambitions by an addi-
tional 11 percentage points.11
We next determine whether antici-
pated sanctions statistically mediate the
interaction effect between gender and
costliness of failure on leadership ambi-
tions. We begin by predicting expected
sanctions using a linear mixed model.
As predicted, the interaction of gender
and costliness of failure positively and
significantly predicts expected sanctions
(p\.05; Model D, Table 4). This means
that costly failure disproportionately
increases the penalties women expect for
failure. In addition, anticipated sanctions
significantly and negatively predict lead-
ership ambitions (Model C, Table 4, p\
.001). Moreover, when expected sanctions
are added to the model predicting leader-
ship ambitions, the interaction effect
between gender and costliness of failure
loses statistical significance (see Model
C, Table 4). A Sobel test shows that the
mediation effect is statistically signifi-
cant: of the total effect of woman 3costly
failure on leadership ambitions, antici-
pated sanctions account for about 31 per-
cent of the decrease in women’s leader-
ship ambitions (p\.05; Figure 1).
Robustness Checks
Given existing research on gender and
leadership, we also test three additional
mechanisms that could plausibly create
gender differences in leadership ambi-
tions between the costly failure and
benign failure conditions: (1) gender dif-
ferences in the anticipated odds of failure
(which could be understood as gender
11As in Study 2, control variables did not
meaningfully attenuate the interaction term of
woman 3costly failure. Coefficients for control
variables can be found in Supplement Part E,
available with the online version of the paper.
Anticipated Discrimination Reduces Women’s Leadership Ambitions 327
differences in confidence), (2) gender dif-
ferences in anticipated resistance from
subordinate group members, and (3) gen-
dered social pressures on men to lead.12
We find no evidence that women and
men differ on their assessments of these
items or evidence of an interaction effect
between gender and costliness of failure.
Thus,wefindnoevidencethatthecostly
failure condition causes women to believe
that they have higher odds of failing, men
to feel more pressure to lead, or women to
become more concerned about resistance
from subordinate group members.
Given both the statistically significant
mediation effect as well as the lack of evi-
dence for other plausible explanations of
our results, we conclude that there is sup-
port for Hypothesis 3 (anticipated sanc-
tions for failure mediate the effect of
costly failure on gender differences in
leadership ambitions).
GENERAL DISCUSSION AND
CONCLUSIONS
The gender gap in leadership is one of the
starkest and most intractable gender
inequalities in contemporary U.S. society.
Scholars have extensively studied the
causes of this gap, uncovering an array
of gendered interactional processes that
make it harder for women to become lead-
ers and retain leadership positions
(Ridgeway 2011). Our findings provide
evidence that women are aware of gender
discrimination against women leaders
and that this knowledge depresses their
leadership ambitions.
This research illuminates macro-to-
micro processes that contribute to the
dearth of women in leadership positions.
We find that gender discrimination
against women leaders not only harms
women leaders: this discrimination also
decreases the likelihood that women pur-
sue leadership positions in the first place.
This gives women fewer opportunities
to reap the rewards associated with
Table 4. Linear Mixed Model Coefficients Predicting Leadership Ambitions (Models A–C) and
Anticipated Sanctions (Model D; N= 484)
Leadership ambitions Anticipated sanctions
Model A Model B Model C Model D
Costly failurea–23.002***
(2.539)
–16.622***
(3.888)
–8.629*
(3.803)
15.176***
(2.491)
Woman –2.810
(4.815)
3.286
(5.562)
–.027
(5.258)
–5.874
(3.556)
Costly failure 3woman –10.982*
(5.087)
–7.606
(4.812)
7.607*
(3.249)
Expected sanctions –.505***
(.065)
Intercept 55.935**
(18.655)
52.620**
(18.629)
83.235***
(17.984)
62.000***
(11.810)
Note: Standard errors appear in parentheses. Additional controls are nonwhite, nonwhite 3woman,
associate’s degree, bachelor’s degree, advanced degree, age, age2, employed, log of income, married,
cohabiting, other relationship status, parent, married 3woman, cohabiting 3woman, other relationship
status 3woman, and parent 3woman.
aReference category is benign failure.
*p\.05. **p\.01. ***p\.001.
12See Supplement F, available with the online
version of the paper, for details on these
measures.
328 Social Psychology Quarterly 82(3)
leadership, reifying existing gender
inequalities. While other scholars have
discussed how women’s leadership ambi-
tions may be decreased by gender dis-
crimination (Cassirer and Reskin 2000;
Kanter 1977), to our knowledge there
has been no experimental research dem-
onstrating the causal chain between gen-
dered sanctions for failure and gender dif-
ferences in leadership ambition. This
research also gives theoretical insight on
the microlevel foundations of the stalled
revolution (see England 2010) as we dem-
onstrate one way in which gender beliefs
about leadership are recreated and legiti-
mized. Gendered sanctions for leadership
failure dampen women’s leadership ambi-
tions, decreasing the number of women in
leadership positions and thereby affirm-
ing the gender stereotypes about leader-
ship that created the original inequality.
Our research also gives new insights
on how organizational responses to fail-
ure may produce gender differences in
behavior and decrease the willingness of
both women and men to lead. Sanctions
are not intrinsic to failure: organizations
determine the consequences for failing
in a leadership role. Organizations would
be wise to reconsider whether heavy-
handed individualized responses to fail-
ure are discouraging otherwise capable
employees—especially women—from pur-
suing leadership roles. As Pfeffer (1992)
explains, organizations are too complex
to realistically assign blame or praise to
any one person for most events. Alterna-
tive approaches to failure that assess
system-level causes for failure and evalu-
ate individuals on their process (instead
of solely focusing on outcome) may allow
organizations to minimize harm without
discouraging creativity and innovation.
A process-oriented response would also
have the benefit of decreasing gendered
blame attributions for failure in leader-
ship situations.
Lastly, these results provide sociologi-
cal insights about gender and risk taking.
Pursuit of a leadership position can be
understood as a form of risk-taking
behavior as leadership can be risky due
to the sanctions associated with failing
in a leadership role. Given that meta-
reviews find gender differences in risk-
taking behavior in intellectual, economic,
and physical domains (Byrnes, Miller,
and Schafer 1999; Croson and Gneezy
2009), some have argued that women’s
unwillingness to take risks further
Figure 1. Sobel Mediation Test Results for the Gendered Effect of Costly Failure on Anticipated
Sanctions and Leadership Ambitions (N = 484)
Note: Standard errors are shown in parentheses. Additional controls are woman, costly failure, nonwhite,
nonwhite 3woman, associate’s degree, bachelor’s degree, advanced degree, age, age2, employed, log
of income, married, cohabiting, other relationship status, parent, married 3woman, cohabiting 3woman,
other relationship status 3woman, and parent 3woman.
Anticipated Discrimination Reduces Women’s Leadership Ambitions 329
disadvantages them in leadership posi-
tions (Horowitz, Igielnik, and Parker
2018). Our results suggest that women’s
‘‘unwillingness’’ to take risks may actu-
ally be a rational response to social con-
texts characterized by gendered interac-
tional dynamics: if women suffer greater
sanctions for failing, taking on leadership
positions becomes riskier for them. Our
results show that women understand
this (Study 1) and that this knowledge
reduces their willingness to take the
risk of being a leader (Studies 2 and 3).
This demonstrates that gendered social
processes can contribute to gender differ-
ences in risk-taking behavior.
Future research should try to examine
these processes with behavioral measures
in other settings, such as laboratory
experiments and actual workplaces. In
addition, future research should also
include controls for leadership experience
and examine the effect of other forms of
gender discrimination on women’s ambi-
tions and behavior. Since these results
suggest that gendered sanctions for fail-
ure may broadly impact women’s aspira-
tions and actions, it would also be useful
to test the effect of punishments on other
outcomes and processes: for instance,
career ambitions, workplace risk taking,
and confidence.
In sum, this research provides new
insights on the microlevel foundations of
the stalled revolution and gives new
insights on how sanctions for failing in
a leadership role—a seemingly gender-
neutral organizational structure—can
create gender inequality.
ACKNOWLEDGMENTS
We are grateful for the insightful comments and
helpful feedback we received on earlier drafts of
this manuscript from Cecilia Ridgeway, Erin
Cech, Shelley Correll, Kjerstin Gruys, Will Kalkh-
off, Jenn MacLure, Christin Munsch, Jen Schradie,
Amanda Sharkey, Stanford Social Psychology
Workshop, the editors, and the anonymous
reviewers.
FUNDING
The author(s) disclosed receipt of the following
financial support for the research, authorship,
and/or publication of this article: We would like
to thank the Ric Weiland Graduate Fellowship
and Kent State University for providing the nec-
essary funds to conduct this research.
REFERENCES
Albanesi, Stefania, Claudia Olivetti, and
Maria Jose Prados. 2015. ‘‘Gender and
Dynamic Agency: Theory and Evidence on
the Compensation of Top Executives.’’ Fed-
eral Reserve Bank of New York Staff
Reports, No. 718. https://www.newyorkfed
.org/medialibrary/media/research/staff_rep
orts/sr718.pdf.
Anderson, Cameron, Sanjay Srivastava, Jen-
nifer S. Beer, Sandra E. Spataro, and Jen-
nifer A. Chatman. 2006. ‘‘Knowing Your
Place: Self-Perceptions of Status in Face-
to-Face Groups.’’ Journal of Personality
and Social Psychology 91(6):1094–110.
Barsh, Joanna, and Lareina Yee. 2012. Unlock-
ing the Full Potential of Women at Work.
New York, NY: McKinsey & Company.
Baumeister, Roy F., Ellen Bratslavsky, Catrin
Finkenauer, and Kathleen D. Vohs. 2001.
‘‘Bad Is Stronger Than Good.’’ Review of
General Psychology 5(4):323–70.
Brescoll, Victoria L., Erica Dawson, and Eric
Luis Uhlmann. 2010. ‘‘Hard Won and Eas-
ily Lost: The Fragile Status of Leaders in
Gender-Stereotype-Incongruent Occupa-
tions.’’ Psychological Science 21(11):1640–
42.
Brescoll, Victoria L., and Eric Luis Uhlmann.
2005. ‘‘Attitudes toward Traditional and
Nontraditional Parents.’’ Psychology of
Women Quarterly 29(4):436–45.
Bureau of Labor Statistics. 2017. ‘‘Highlights
of Women’s Earnings in 2016.’’ Report
1069. https://www.bls.gov/opub/reports/wo
mens-earnings/2016/home.htm.
Byrnes, James P., David C. Miller, and Wil-
liam D. Schafer. 1999. ‘‘Gender Differences
in Risk Taking: A Meta-Analysis.’’ Psycho-
logical Bulletin 125(3):367–83.
Cassirer, Naomi, and Barbara Reskin. 2000.
‘‘High Hopes: Organizational Position,
Employment Experiences, and Women’s
330 Social Psychology Quarterly 82(3)
and Men’s Promotion Aspirations.’’ Work
and Occupations 27(4):438–64.
Chandler, Jesse, Pam Mueller, and Gabriele
Paolacci. 2014. ‘‘Nonnaı¨vete
´among Amazon
Mechanical Turk Workers: Consequences
and Solutions for Behavioral Researchers.’’
Behavior Research Methods 46(1):112–30.
Cohen, Philip N., and Matt L. Huffman. 2007.
‘‘Working for the Woman? Female Manag-
ers and the Gender Wage Gap.’’ American
Sociological Review 72(5):681–704.
Croson, Rachel, and Gneezy Uri. 2009. ‘‘Gen-
der Differences in Preferences.’’ Journal of
Economic Literature 47(2):448–74.
Damaske, Sarah. 2011. For the Family?: How
Class and Gender Shape Women’s Work.
New York, NY: Oxford University Press.
Eagan, Kevin, Ellen Bara Stolzenberg, Joseph
J. Ramirez, Melissa C. Aragon, and Maria
Ramirez Suchard, et al. 2016. The Ameri-
can Freshman: Fifty-Year Trends 1966–
2015. Los Angeles, CA: Higher Education
Research Institute, UCLA.
Eagly, Alice H., and Linda L. Carli. 2007.
Through the Labyrinth: The Truth about
How Women Become Leaders. Cambridge,
MA: Harvard Business Review Press.
England, Paula. 2010. ‘‘The Gender Revolu-
tion: Uneven and Stalled.’’ Gender & Soci-
ety 24(2):149–66.
Finucane, Melissa L., Paul Slovic, C. K. Mertz,
James Flynn, and Theresa Satterfield.
2000. ‘‘Gender, Race, and Perceived Risk:
The ‘White Male’ Effect.’’ Health, Risk &
Society 2(2):159–72.
Frick, Robert W. 1995. ‘‘Accepting the Null
Hypothesis.’’ Memory & Cognition 23(1):
132–38.
Gupta, Vishal K., Seonghee Han, Sandra C.
Mortal, Sabatino (Dino) Silveri, and Daniel
B. Turban. 2018. ‘‘Do Women CEOs Face
Greater Threat of Shareholder Activism
Compared to Male CEOs? A Role Congruity
Perspective.’’ Journal of Applied Psychol-
ogy 103(2):228–36.
Horowitz, Juliana Menasce, Ruth Igielnik, and
Kim Parker. 2018. ‘‘Women and Leadership
2018.’’ Pew Research Center. http://www
.pewsocialtrends.org/2018/09/20/women-and-
leadership-2018/.
Jackall, Robert. 1988. Moral Mazes: The World
of Corporate Managers. New York, NY:
Oxford University Press.
Kahneman, Daniel, and Amos Tversky. 1979.
‘‘Prospect Theory: An Analysis of Decision
under Risk.’’ Econometrica 47(2):263–92.
Kanter, Rosabeth Moss. 1977. Men and
Women of the Corporation. New York, NY:
Basic Books.
Kennedy, Jessica A., Mary-Hunter McDonnell,
and Nicole M. Stephens. 2016. ‘‘Does Gen-
der Raise the Ethical Bar? Exploring the
Punishment of Ethical Violations at
Work.’’ Vanderbilt Owen Graduate School
of Management Research Paper No.
2770012. https://dx.doi.org/10.2139/ssrn.27
70012.
Kenrick, Douglas T., Melanie R. Trost, and Jill
M. Sundie. 2004. ‘‘Sex Roles as Adapta-
tions: An Evolutionary Perspective on Gen-
der Differences and Similarities.’’ Pp. 65–
91 in The Psychology of Gender, edited by
A. H. Eagly, A. E. Beall, and R. J. Stern-
berg. New York, NY: Guilford Press.
Konrad, Alison M., J. Edgar Ritchie, Jr.,
Pamela Lieb, and Elizabeth Corrigall.
2000. ‘‘Sex Differences and Similarities in
Job Attitude Preferences: A Meta-Analy-
sis.’’ Psychological Bulletin 126(4):593–641.
MacKinnon, David P., and James H. Dwyer.
1993. ‘‘Estimating Mediated Effects in Pre-
vention Studies.’’ Evaluation Review
17(2):144–58.
Mason, Winter, and Siddharth Suri. 2012.
‘‘Conducting Behavioral Research on Ama-
zon’s Mechanical Turk.’’ Behavior Research
Methods 44(1):1–23.
Pew Research Center. 2019. ‘‘The Data on
Women Leaders.’’ http://www.pewsocialt
rends.org/fact-sheet/the-data-on-women-lea
ders/
Phelan, Julie E., Corinne A. Moss-Racusin,
and Laurie A. Rudman. 2008. ‘‘Competent
yet Out in the Cold: Shifting Criteria for
Hiring Reflect Backlash toward Agentic
Women.’’ Psychology of Women Quarterly
32(4):406–13.
Pfeffer, Jeffrey. 1992. Managing with Power:
Politics and Influence in Organizations.
Boston, MA: Harvard Business School
Press.
Reuben, Ernesto, Paola Sapienza, and Luigi
Zingales. 2015. ‘‘Taste for Competition
and the Gender Gap among Young Busi-
ness Professionals.’’ NBER Working Paper
No. 21695.
Ridgeway, Cecilia L. 2011. Framed by Gender:
How Gender Inequality Persists in the Mod-
ern World. New York, NY: Oxford Univer-
sity Press.
Ridgeway, Cecilia L., and Sandra Nakagawa.
2017. ‘‘Is Deference the Price of Being
Anticipated Discrimination Reduces Women’s Leadership Ambitions 331
Seen as Reasonable? How Status Hierar-
chies Incentivize Acceptance of Low Status.’’
Social Psychology Quarterly 80(2):132–52.
Rozin, Paul, and Edward B. Royzman. 2001.
‘‘Negativity Bias, Negativity Dominance,
and Contagion.’’ Personality and Social
Psychology Review 5(4):296–32.
Sarsons, Heather. 2017. ‘‘Interpreting Signals
in the Labor Market: Evidence from Medi-
cal Referrals.’’ Working paper. https://sch
olar.harvard.edu/files/sarsons/files/sarsons_
jmp.pdf.
Swim, Janet, and Lawrence J. Sanna. 1996.
‘‘He’s Skilled, She’s Lucky: A Meta-Analy-
sis of Observers’ Attributions for Women’s
and Men’s Successes and Failures.’’ Person-
ality and Social Psychology Bulletin 22(5):
507–19.
The
´baud, Sarah. 2015. ‘‘Status Beliefs and the
Spirit of Capitalism: Accounting for Gender
Biases in Entrepreneurship and Innova-
tion.’’ Social Forces 94(1):61–86.
Turco, Catherine J. 2016. The Conversational
Firm: Rethinking Bureaucracy in the Age
of Social Media. New York, NY: Columbia
University Press.
U.S. Census Bureau. 2017. ‘‘Educational
Attainment in the United States: 2017.’’
https://www.census.gov/data/tables/2017/de
mo/education-attainment/cps-detailed-tables
.html.
Vial, Andrea C., Jaime L. Napier, and Victoria
L. Brescoll. 2016. ‘‘A Bed of Thorns: Female
Leaders and the Self-Reinforcing Cycle of
Illegitimacy.’’ The Leadership Quarterly
27(3):400–14.
Williams, Joan C., and Rachel Dempsey. 2014.
What Works for Women at Work: Four Pat-
terns Working Women Need to Know. New
York, NY: New York University Press.
BIOS
Susan R. Fisk is an Assistant Professor
of Sociology at Kent State University.
Her research focuses on understanding
how gendered processes influence deci-
sion making, with special focus on how
organizational practices can trigger or
dampen gender differences. She is partic-
ularly interested in studying the connec-
tion between gender, risk-taking, and
inequality.
Jon Overton is a PhD candidate in
sociology at Kent State University.
His research revolves primarily around
social psychological and organizational
approaches to status processes, especially
how people gain or lose status from their
social ties. His dissertation research
explores the consequences of this phe-
nomenon for social inequality.
332 Social Psychology Quarterly 82(3)
