To Delegate or Not to Delegate: Gender Differences in Affective Associations and Behavioral Responses to Delegation

Abstract

Effectively delegating work to others is considered critical to managerial success, as it frees up managers' time and develops subordinates' skills. We propose that female leaders are less likely than male leaders to capitalize on these benefits of delegating. Although delegation has communal (e.g., relational) and agentic (e.g., assertive) properties, we argue that female leaders, as compared to male leaders, find it more difficult to delegate tasks due to gender-role incongruence. In five studies, we draw upon social role and backlash theories to show that women imbue delegation with more agentic traits, have more negative associations with delegating, and feel greater guilt about delegating than men. These associations result in women delegating less than men and, when they do delegate, having lower-quality interactions with subordinates. We further show that reframing delegation as communal attenuates women's negative associations with delegation. These findings reveal that even when a given behavior has both agentic and communal elements, perceptions of agency can undermine women's engagement in them. However, emphasizing the communal nature of seemingly agentic acts may encourage women's engagement in such critical leadership behaviors. These findings have theoretical and practical implications for research on gender differences and leadership behavior in the workplace.

Full text

rAcademy of Management Journal
2018, Vol. 61, No. 4, 1467–1491.
https://doi.org/10.5465/amj.2016.0662
TO DELEGATE OR NOT TO DELEGATE: GENDER
DIFFERENCES IN AFFECTIVE ASSOCIATIONS AND
BEHAVIORAL RESPONSES TO DELEGATION
MODUPE AKINOLA
Columbia University
ASHLEY E. MARTIN
Stanford University
KATHERINE W. PHILLIPS
Columbia University
Effectively delegating work to others is considered critical to managerial success, as it
frees up managers’time and develops subordinates’skills. We propose that female
leaders are less likely than male leaders to capitalize on these benefits of delegating.
Although delegation has communal (e.g., relational) and agentic (e.g., assertive) prop-
erties, we argue that female leaders, as compared to male leaders, find it more difficult to
delegate tasks due to gender-role incongruence. In five studies, we draw upon social role
and backlash theories to show that women imbue delegation with more agentic traits,
have more negative associations with delegating, and feel greater guilt about delegating
than men. These associations result in women delegating less than men and, when they
do delegate, having lower-quality interactions with subordinates. We further show that
reframing delegation as communal attenuates women’s negative associations with del-
egation. These findings reveal that even when a given behavior has both agentic and
communal elements, perceptions of agency can undermine women’s engagement in
them. However, emphasizing the communal nature of seemingly agentic acts may en-
courage women’s engagement in such critical leadership behaviors. These findings have
theoretical and practical implications for research on gender differences and leadership
behavior in the workplace.
Delegation, defined as the assignment of re-
sponsibilities to subordinates and conferral of au-
thority to carry out assigned tasks, is considered an
important and effective leadership behavior. For
leaders, delegation can reduce work overload and
improve the speed and quality of decisions, while
simultaneously enabling subordinates to view
leaders as participative (De Pater, Van Vianen, &
Bechtoldt, 2010). For subordinates, delegation can
increase intrinsic motivation and provide opportu-
nities for professional development, as it empowers
subordinates, enables efficacy, and strengthens
relationship building between leaders and sub-
ordinates (Chen & Aryee, 2007; Leana, 1986, 1987;
Yukl & Fu, 1999). For organizations, delegation en-
hances task coordination, productivity, and perfor-
mance by drawing upon the specializations and
skills of different employees (Becker & Murphy,
1992; Miles, Snow, Meyer, & Coleman, 1978).
Definitions of delegation in much of the manage-
ment literature have painted this leadership behav-
ior as one that is primarily communal in nature, as
it is relational and can help leaders develop em-
ployees. Since communality is a prescriptive gender
stereotype that is rewarded when exhibited by
women (Rudman & Phelan, 2008), one might argue
that women leaders should benefit from engaging
in delegation (Amanatullah & Morris, 2010; Carli,
LaFleur, & Loeber, 1995; Heilman & Okimoto,
2007). Yet the act of delegating can also be consid-
ered agentic, as it requires assertiveness and
The first two authors contributed equally to this work.
We thank Sarah Anolik, Hayley Blunden, Zach Bucknoff,
Ashley Culver, Zach Heinemann, Yael Warach, Laura
Fitzelle, and Lucie Yang for their research assistance. We
also thank Daniel Ames, Adam Galinsky, Astrid Homan,
Jessica Kennedy, Michael Slepian, and the Social In-
teractions in Groups and Organizations Lab at Columbia
Business School for their thoughtful comments and sug-
gestions regarding this research.
1467
Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyright holder’s express
written permission. Users may print, download, or email articles for individual use only.

intentionality. Evidence has suggested that pre-
scriptive gender stereotypes that reward agentic be-
havior for men can sanction this same behavior for
women (Eagly, 2007; Heilman, 2001; Rudman &
Phelan, 2008). Thus, delegation can be considered
agender-role-paradoxical behavior for women: in
some respects, delegation is aligned with female role
stereotypes, thus presumably helping women; in
other respects, delegation is misaligned with female
role stereotypes, thus potentially hurting women
when they engage in this behavior. In this paper, we
propose that women will associate beneficial lead-
ership behaviors that are simultaneously construed
as both agentic and communal with more negative
attributes, and will engage in these behaviors less
frequently than men.
Specifically, we examine whether women view
delegation with more negative affect than men do,
and if so, why. Research on social role and backlash
theories has demonstrated that women leaders face
a double bind, as the same qualities that make them
“good leaders”(e.g., agency) also make them “bad
women”(e.g., not communal) (Tannen, 1990: 244).
In contrast, male leaders are often free to exercise
their agency without the constraint of simulta-
neously needing to engender communality (Heilman
& Okimoto, 2007; Rudman, 1998). Drawing on these
theories, we argue that when leadership behaviors
can be viewed as both agentic and communal, the
agentic nature of the behavior will be more salient for
women than the communal nature. Thus, engaging
in such behaviors will be more difficult for women
than for men.
When applied to the context of delegation, we
theorize that women are more likely to associate
delegation with agency than men are, as women are
penalized more for showing agency than they are
rewarded for showing communality (Heilman &
Chen, 2005). Since women often see themselves as,
and are expected to be, less agentic (i.e., assertive,
aggressive) than men (Bem, 1981; Rudman & Phelan,
2008), we propose that they will be more sensitive to
the act of delegating, ascribing it with more agentic
characteristics compared to men. In turn, women’s
affective associations (i.e., emotions and feelings
associated with a behavior) with delegation will be
more negative, affecting their propensity to delegate.
Accordingly, we argue that the gender-role-
paradoxical nature of delegation will lead women,
compared to men, to associate delegation with more
negative emotions, being more sensitive to punish-
ment that can ensue from engaging in gender-role
incongruent behavior (Brescoll, 2012; Moss-Racusin
& Rudman, 2010). Since affective associations with
a given behavior are powerful predictors of engage-
ment in such behavior, we expect that women’s more
negative associations with delegation will decrease
their likelihood of delegating and will negatively
affect the quality of their interactions with sub-
ordinates. We also identify ways to mitigate gender
differences in delegation. We argue that for women,
emphasizing the communal aspects of delegation
may allow its communal properties to outweigh its
agentic ones. Thus, we propose that framing dele-
gation as communal, by highlighting its benefits for
subordinates, has the potential to change women’s
negative associations with delegation.
The current research contributes to management
theory in four key ways. First, to our knowledge, we
offer the first empirical evidence demonstrating that
women and men differ in their delegation behavior.
Gender researchers have largely examined how dif-
fering gender-role stereotypes shape how women
experience and perceive leadership behaviors that
are construed as either agentic or communal, such as
self-promotion and negotiating on behalf of others
(Amanatullah & Morris, 2010; Heilman, 2001;
Rudman & Phelan, 2008). Instead, we offer insight
into how women experience and perceive gender-
role-paradoxical leadership behaviors, such as del-
egation, that could be simultaneously construed as
both agentic and communal.
Second, we provide the first empirical evidence of
the processes through which gender differences may
influence delegation behavior. We propose that
affective associations are an important predictor of
delegation behavior for women. Men’s and women’s
differing affective associations with delegation may
make women less likely to delegate and can affect
performance. Based on social role and backlash
theories (Eagly, 1987; Eagly & Karau, 2002; Rudman,
1998; Rudman & Phelan, 2008), we advance the
theoretical argument that for tasks that are both
agentic and communal, agentic qualities loom larger
than communal ones for women, dominating their
perceptions and generating greater negative affect
as compared to men.
Third, effective delegation entails having a pro-
ductive relationship with subordinates, yet little is
known about the quality of managers’interactions
with subordinates in the context of delegation.
We propose that women’s negative associations
with delegation can influence the quality of their
interactions with subordinates. Additionally, we
examine delegation in a controlled laboratory ex-
periment. In doing so, we experimentally test
1468 AugustAcademy of Management Journal

whether there are gender differences in delegation
when position, task, and employees are held
constant.
Fourth, by clarifying why women differentially
delegate, we offer insight into potential interventions
that may help address gender disparities in leader-
ship behaviors. Specifically, our research suggests
that when leadership behaviors that have both
agentic and communal properties are reframed as
communal in nature, women may view them more
favorably. Thus, this research not only offers insight
into the challenges women may face as they continue
to ascend to top leadership positions in organiza-
tions, particularly in engaging in behaviors that are
both agentic and communal, but also provides po-
tential remedies to these challenges.
In what follows, we draw upon literature on so-
cial role (Eagly, 1987, 2007) and backlash (Rudman,
1998) theory to describe the theoretical rationale for
why women may have more negative associations
with delegation relative to men, and for how these
negative associations may influence women’sdel-
egation behavior. We build a theoretical case for
how reframing delegation to be more gender-role
congruent may change women’snegativeassocia-
tions with delegation. We then describe five studies
that test our hypotheses that women and men differ
in their perceptions of, and behavior during, dele-
gation, and that women’s affective associations
can be changed by reframing how they perceive
delegation.
SOCIAL ROLE THEORY: GENDER STEREOTYPES
AND ROLE INCONGRUENCE
Gender stereotypes pervade society and influ-
ence how men and women are perceived and ex-
perience their lives (Eagly, 2007; Heilman, 2001).
According to social role theory, the roles occupied
by, and division of labor between, men and women
create expectancies about what women are like and
how they are expected to behave. As such, women
are expected to be communal, enacting behaviors
that are cooperative and other-oriented, whereas
men are expected to demonstrate agency, enacting
behaviors that are independent and action-oriented.
Both men and women internalize gender roles at
a very young age through a variety of means. Media,
parents, and peers teach children to behave in
gender-normative ways, rewarding girls for being
communal and boys for being agentic (Bem, 1983;
Bryant & Check, 2000). For example, parents en-
courage and reward gender-congruent behaviors in play
and household chores (Weisner & Wilson-Mitchell,
1990), and teachers encourage agency in boys by
calling on them more often and giving them more
talking time than girls (Kindlon & Thompson, 2000).
By adulthood, men and women tend to self-identify
with these gender stereotypes, with women often
rating themselves as more communal and behaving
more communally than men, and men rating them-
selves as more agentic and behaving more agenti-
cally (Bem, 1981; Valian, 1998).
Those who violate these gender norms are often
socially and economically penalized (Rudman,
1998; Rudman & Glick, 2001). For example,
agentic women are evaluated less favorably, receive
lower salaries, and are less likely to be hired for the
same position compared to agentic men (Heilman,
Wallen, Fuchs, & Tamkins, 2004; Rudman, 1998;
Rudman & Fairchild, 2004). As a result, both women
and men are sensitive to situations that elicit so-
cial sanctions for counter-stereotypical behavior
(Brescoll, 2012; Moss-Racusin, Phelan, & Rudman,
2010). This social sensitivity to gender-role in-
congruence may have implications for how women
perceive delegation, as the social penalties for
stereotype-incongruent behavior may loom larger
in their minds than the rewards for congruency
(Heilman & Chen, 2005; Rudman & Phelan, 2008).
Thus, although delegation has qualities of both
agency (stereotype incongruence for women) and
communality (stereotype congruence for women),
we argue women will be especially sensitive to
the agentic nature of delegation, as compared to
men, due to social costs outweighing gains (i.e.,
communality does not help women as much as
agency hurts them [Cuddy, Fiske, & Glick, 2004]).
With regard to the communal nature of delegation,
communality is often perceived more positively
than agency (Abele & Wojciszke, 2014), and people
describe themselves with more communal than
agentic traits (Bruckm ¨
uller & Abele, 2013); thus, we
do not expect to see gender differences in percep-
tions of delegation as communal. We therefore
hypothesize:
Hypothesis 1. Women will associate the act of dele-
gating (in general) with more agentic traits (e.g., aggres-
siveness, power) compared to men.
GENDER-ROLE STEREOTYPES AND AFFECTIVE
ASSOCIATIONS WITH DELEGATION
Both men and women are sensitive to situations
that evoke penalties for gender-role incongruent
2018 1469Akinola, Martin, and Phillips

behavior (Moss-Racusin & Rudman, 2010; Vandello,
Bosson, Cohen, Burnaford, & Weaver, 2008).
Awareness of these repercussions creates negative
associations with engaging in gender-role in-
congruent behaviors, where women are more likely
to associate agency with negative emotions com-
pared to men. These negative associations can take
many forms, including feeling heightened anxiety
in anticipation of engaging in agentic behavior
(Babcock & Laschever, 2003; Spencer, Steele, &
Quinn, 1999), feeling guilt for exhibiting agentic
behavior (Benetti-McQuoid & Bursik, 2005), and
fearing backlash from engaging in agentic behavior
(Brescoll, 2012).
One specific emotion related to gender-role viola-
tion is guilt. Guilt, an unpleasant emotion associated
with the recognition that one has violated a per-
sonal or social standard (Kugler & Jones, 1992),
can arise from a sense of responsibility to up-
hold cultural norms for social behavior (Izard,
1977). Women who engage in gender-role-
paradoxical behaviors that they perceive as
agentic (e.g., delegating) are likely to recognize that
their actions run counter to the female gender norm
of behaving communally. Consequently, they may
experience feelings of guilt, as engaging in situa-
tions that are incongruent with one’s gender role
can influence the emergence of guilt (Benetti-
McQuoid & Bursik, 2005). Thus, if women associ-
ate delegation with agentic traits more than men do
(as proposed in Hypothesis 1), they may have
greater negative associations with delegation com-
pared to men.
A second negative association that has been
found to co-occur with agentic behavior for women
is fear of backlash. Fear of backlash, or trepidation
about potentially incurring social and economic
penalties from others (Rudman 1998; Rudman &
Phelan, 2008), has been shown to be prevalent for
women in a variety of contexts. For example,
women experience a fear of backlash for self-
promoting, as they view this behavior as more as-
sertive and gender-role incongruent than men do
(Moss-Racusin & Rudman, 2010). Likewise, women
avoid behaving aggressively in negotiations
(Amanatullah & Morris, 2010) and minimize power
displays in political and organizational settings
(Brescoll, 2012) for fear of backlash. This research
suggests that because women may have negative
associations with engaging in agentic behaviors,
they may feel guilty for violating communality
expectations and may also fear backlash. We pro-
pose that these negative associations extend to
the domain of delegation through the following
hypotheses:
Hypothesis 2a. Women, as compared to men, will
have more negative affective associations when an-
ticipating or engaging in the act of delegation.
Hypothesis 2b. Women, as compared to men, will
have the specific negative affective association of
guilt when anticipating or engaging in the act of
delegation.
Hypothesis 2c. Women, as compared to men, will
have the specific negative affective association of
fear (of backlash) when anticipating or engaging in
the act of delegation.
GENDER-ROLE STEREOTYPES AND
DELEGATION BEHAVIOR
We argue that perceiving delegation as agentic and
having negative associations with delegation can
impact whether women choose to delegate when
given the opportunity, and influence women’s be-
havior when delegating. Previous models of decision
making have assumed that individuals engage in
a rational process of weighing the costs and benefits
of behavioral choices and selecting the course of
action with the most favorable ratio of costs to ben-
efits (Weinstein, 1993). However, significant support
has emerged for the idea that the emotions, or affect,
associated with a behavior play an important role
in decision making (Crites, Fabrigar, & Petty, 1994;
Kiviniemi, Voss-Humke, & Seifert, 2007; Simons &
Carey, 1998). Specifically, behavioral choices are
influenced by the anticipation of feeling particular
emotions (such as regret, guilt, pride, or happiness)
as a result of engaging in a specific behavior (Richard,
van der Pligt, & de Vries, 1996).
Emotions associated with a particular behavioral
practice are powerful predictors of one’s likelihood
of participating in specific behaviors, with more
positive affective associations leading to a greater
likelihood of engaging in a behavior (Kiviniemi &
Bevins, 2008). For example, more positive and less
negative associations with exercise are related to
increased motivation to engage in physical activity
(Laverie, 1998). Having a less negative association
with medical testing is related to having a greater
likelihood of partaking in preventive medical tests
(Kiviniemi, Jandorf, & Erwin, 2014). And holding
more negative associations with marijuana, nico-
tine, and alcohol is related to reduced smoking
and drinking (Simons & Carey, 1998; Trafimow &
1470 AugustAcademy of Management Journal

Sheeran, 1998). Together, these results suggest that if
women have negative associations with delegation,
including guilt and fear of backlash, then these neg-
ative affective associations should deter women
from delegating. In contrast, since for men agency is
prescribed and rewarded, they should not experi-
ence these same negative affective associations and
will be free to engage in delegating. We therefore
hypothesize:
Hypothesis 3a. Women will be less likely to delegate
as compared to men.
Hypothesis 3b. Men’s and women’s differential af-
fective associations with delegation will mediate their
likelihood of delegating.
Outside the realm of health psychology, social-
psychological research has shown that affective
associations influence behavior toward others. For
example, individuals who have more negative asso-
ciations with a particular social group are less likely
to interact with individuals in that category (Rooth,
2010) and to have more negative interactions when
they do interact with those individuals (McConnell &
Leibold, 2001). Negative associations can increase
anxiety and avoidant behavior (Plant, Devine, &
Brazy, 2003; Stephan & Stephan, 1985), depleting
cognitive capacity and heightening self-regulation in
social interactions (Apfelbaum, Sommers, & Norton,
2008). These forces can divert attention away from
the social interaction itself, resulting in poorer-
quality interactions (Eysenck, 1992) and impaired
performance (Richeson & Shelton, 2003). Thus, if
women have negative associations with delegation
then when they do delegate, the quality of their in-
teractions with subordinates should be lower than
those of male delegators, who are less likely to have
negative associations with delegation due to its
gender-role congruent nature for men. Thus, we
hypothesize:
Hypothesis 3c. Women’s interactions with sub-
ordinates when delegating will be perceived by both
subordinates and independent observers as being
poorer in quality compared to men’s behavior when
delegating.
Further, given that delegation has been found to
provide critical performance benefits for managers
(Leana, 1987; Yukl & Fu, 1999), we more broadly
hypothesize:
Hypothesis 3d. Delegators will outperform non-
delegators in the tasks they have been given to
complete.
ATTENUATING NEGATIVE ASSOCIATIONS
WITH DELEGATION
It is possible that women’s affective associations
with regard to gender-role-paradoxical behaviors may
be influenced by how these behaviors are framed.
A large body of research in judgment and decision
making (e.g., Kahneman & Tversky, 1984) has high-
lighted that using language that changes how in-
formation is framed can influence individuals’
behavior in numerous domains. For instance, framing
risks in terms of losses versus gains changes in-
dividuals’decision-making processes. Specifically,
because losses feel more painful than gains feel plea-
surable, framing risks in terms of losses fosters risk
acceptance, while framing risks as gains fosters risk
avoidance (Kahneman & Tversky, 1984; Rothman &
Salovey, 1997). These framing effects have been at-
tributed to the feelings and construals evoked by dif-
fering descriptions of the same information (Liberman,
Samuels, & Ross, 2004; McFarland & Miller, 1994;).
Gender researchers have capitalized on framing
effects in an effort to reduce gender differences in
behavior, particularly in the domain of negotiation
(Bowles, Babcock, & Lai, 2007; Bowles, Babcock, &
McGinn, 2005; Small, Gelfand, Babcock, & Gettman,
2007). Since negotiation is seen as an agentic be-
havior, women are often more reluctant to negotiate
when “given the opportunity to negotiate”compared
to men (Small et al., 2007); however, when the same
situations are described more communally (e.g., “an
opportunity to ask”) or communal properties are
highlighted (e.g., negotiating on behalf of others)
gender differences are reduced (Amanatullah &
Morris, 2010; Small et al., 2007). Highlighting the
communal properties of agentic acts aligns female
behavior with their communal prescriptions and
reduces the (fear of) social repercussions involved in
violating gender roles (Heilman & Okimoto, 2008).
Taken together, this research suggests that framing
gender-role-paradoxical behaviors as communal by
emphasizing how these behaviors benefit others has
the potential to change women’s negative associa-
tions with behaviors that are both communal and
agentic. Thus, we hypothesize:
Hypothesis 4. Framing delegation in a manner that
emphasizes its communal properties will attenuate
women’s negative associations with delegation.
OVERVIEW OF THE PRESENT RESEARCH
We conducted five studies to test our hypotheses.
In Study 1, using a survey of Master of Business
2018 1471Akinola, Martin, and Phillips

Administration (MBA) students, we tested whether
women perceive delegation to be more agentic than
men do (Hypothesis 1). In Study 2, a scenario study,
we instructed men and women to either delegate or
not delegate a set of tasks, then examined whether
men and women have differential associations with
delegation using text analysis (Hypothesis 2a). In
Study 3, a correlational study, we again used text
analysis to examine whether men and women differ
in their associations with delegation (Hypotheses 2a
and 2b) among a sample of participants with mana-
gerial experience. In Study 4, we examined the de-
gree to which men’s and women’s associations with
delegation differentially predict their actual delega-
tion behavior in an experimental setting involving
a face-to-face interaction with an ostensible sub-
ordinate (Hypotheses 3a–3d). Finally, in Study 5, we
tested whether framing delegation as communal in
nature will attenuate women’s negative associations
with delegation (Hypothesis 4).
STUDY 1
Through a survey administered to MBA students
at a large, private university, we tested the hypoth-
esis that women perceive delegation to be more
agentic than men and that these asymmetrical per-
ceptions do not extend to perceptions of delegation
as communal.
Method
Participants. We recruited 104 participants from
the MBA program of an East Coast University, as
their recent work experience likely entailed dele-
gating. Seven participants with no delegation expe-
rience were excluded, leaving a sample of 97
participants (63 men; 34 women) with an average age
of 28.05 years (SD 52.81). Fifty students were in
managerial positions prior to pursuing their MBAs,
and participants had 13.78 (SD 539.82) sub-
ordinates on average. The racial composition of the
sample was 46% White, 33% Asian, 7% Hispanic,
and 13% other.
Procedure. Participants were asked questions re-
garding delegation embedded in a decision-making
survey administered during a leadership course.
Specifically, they completed demographic questions
and were asked, “How much do you associate the
following adjectives with delegating?”on scales
ranging from 1 (not at all)to7(very much). Based on
Koenig and Eagly (2014) and Diekman and Eagly
(2000), we selected four adjectives reflecting
agency—powerful, confident, aggressive, and con-
trolling (a5.51)1—and four adjectives reflecting
communality—other-oriented, supportive, develop-
mental, and considerate (a5.79).
Results
Supporting Hypothesis 1, we found that women
associated delegation with agency (M54.57, SD 5
.84) more than men did (M54.16, SD 51.03), F(1,
95) 53.86, p5.05, h
p2
5.039. With regard to asso-
ciations of communality with delegation, there was
no difference between women (M55.24, SD 51.18)
and men (M55.17, SD 51.18), F(1, 95) 5.07, p5
.80, h
p2
,.00.
Discussion
In support of Hypothesis 1, female participants
viewed delegation as more agentic than male par-
ticipants did. Since both men and women per-
ceive communality more positively than agency
(Bruckm ¨
uller & Abele, 2013), we did not expect and
did not find gender differences in communality.
Importantly, both men and women perceived dele-
gation as more communal than agentic overall;
however, both communality and agency ratings for
delegation were above the midpoint, offering evi-
dence that delegation is indeed considered both
agentic and communal in nature. Further, these
findings support our theoretical argument that
women may be more sensitive to the gender-role
incongruence of delegation, perceiving it as more
agentic than men do. We argue that this sensitivity to
the agentic nature of delegation may lead women
to view delegation more negatively than men do,
a hypothesis we tested in Study 2.
STUDY 2
Study 2 was designed to understand women’s and
men’s differential affective responses when told ei-
ther to delegate or not delegate. Using a 2 (participant
gender: male vs. female) 32 (condition: delegate vs.
do not delegate) 32 (subordinate gender: male vs.
female) design, we examined men’s and women’s
1Low alphas have been found in samples with signifi-
cant international diversity (Zou, Tam, Morris, Lee, Lau, &
Chiu, 2009). Given the low reliability, we examined the a
values separately for U.S. and international students and
found higher values for U.S. students (n531; a5.69)
relative to international students (n566; a5.35).
1472 AugustAcademy of Management Journal

free-response affective associations with delegation.
We manipulated the gender of the subordinate, but
did not expect effects of subordinate gender on af-
fective associations with delegation. We tested Hy-
pothesis 2a, that women would have more negative
associations relative to men in the delegate condition
as compared to the do-not-delegate condition. Fur-
ther, we explored the tenor of women’s negative af-
fective associations by examining whether men and
women differentially associate delegation with guilt
(Hypothesis 2b). To explore the possibility that
women’s negative associations with delegation are
due to concerns that others will view them as less
competent (i.e., unable to handle their workload) or
their subordinates as less competent, we included
competence measures as control variables.
Method
Participants. Two hundred participants (93
women; M
age
: 36.42 years) were recruited from Am-
azon’s Mechanical Turk (MTurk). The sample was
77% Caucasian, 10% Black, 6% Hispanic, 6% Asian,
and 3% other, with an average of 15.06 (SD 58.09)
years of work experience. Participants resided in
the United States and 90% had post-secondary
education.
Procedure. The study was described as an exam-
ination of workplace experiences. Participants were
told to imagine that they were an employee at
a marketing company specializing in telecommuni-
cations, and that their position involved managing,
coordinating, and executing the company’s strategic
vision. Participants were told that they personally
had a busy week ahead as they were launching a new
marketing campaign, and that a junior employee
who worked under them (Michael or Michelle) could
help them complete their tasks.
Measures. Participants first wrote about their af-
fective associations with delegating or not delegating
and responded to measures capturing guilt and
competence on scales ranging from 1 (strongly dis-
agree)to7(strongly agree).
After reading the scenario, participants were asked
to write about how they felt about delegating or
not delegating to their employee, and to list any
thoughts, feelings, or adjectives associated with their
behavior (delegating or not delegating).
To capture guilt that may be associated with
overburdening an employee through delegation (or
the relief of not delegating and not overburdening an
employee), participants in the delegate condition
were asked to what extent they felt...“bad delegating
because your subordinate has other things to do,”
“some guilt delegating because you think your sub-
ordinate might be overburdened,”and “concerned
that your subordinate will have to stay at work late
because they are working on your tasks.”Partici-
pants in the do-not-delegate condition were asked
about the extent to which they felt... “good not del-
egating because your subordinate has other things to
do,”“some relief not delegating because you think
your subordinate might be overburdened,”and
“happy that your subordinate will not have to stay at
work late because they are working on your tasks”
(a5.93).
To capture whether women’s negative associa-
tions with delegation were influenced by the con-
cern that others would perceive them as less
competent for delegating or not delegating, partici-
pants were asked, when deciding whether to dele-
gate, the extent to which they would... “think that
others might see you as less competent?”“feel like
you cannot manage your time effectively at work?”
“feel like it seems like you can’t handle the job on
your own?”“worry that your employees might view
you less favorably?”“think that you might not get the
credit you deserve from others at work?”and “think
that you won’t get respect you want from others at
work?”(a5.95). This scale was used as a control
variable in all analyses.
We also asked three questions to capture whether
women’s negative associations with delegation
might be influenced by concerns about the compe-
tence of their subordinate: in deciding whether to
delegate, to what extent would the participant...
“think you could accomplish the tasks much faster
on your own?”“think you could accomplish the
tasks much better yourself?”and “worry about the
tasks getting done quickly and effectively?”(a5.71).
This scale was used as a control variable in all
analyses.
Participants were asked to identify the name of
their junior employee, their department, and their
role from a number of options.
Data-analysis strategy. Written responses were
analyzed using Linguistic Inquiry Word Count
(LIWC) software (Pennebaker, Francis, & Booth,
2001). LIWC computes percentages of words in
a text within specified categories. Given our interest
in affective associations with delegation, we ana-
lyzed negative and positive emotions. We analyzed
each participant’s response separately to yield the
percentage of words pertaining to negative emo-
tions and positive emotions. Due to the skewness of
the distribution of the LIWC output, data were
2018 1473Akinola, Martin, and Phillips

transformed with a square-root function to attain a
normal distribution (Tabachnick & Fidell, 2007).
Results
Participant attrition. Overall, participants accu-
rately reported the name of their junior employee,
their department, and their role; 86% passed the at-
tention checks. The final sample therefore included
171 participants (83 women; M
age
537.0 years).
Affective associations with delegation. On av-
erage, participants used 19.5 (SD 514.18) words in
their free-response text, 95% of which were recog-
nized by the LIWC program.2We first examined the
three-way interaction of participant gender, delega-
tion condition, and subordinate gender on negative
and positive affect; as expected, we found no signif-
icant effects.3We therefore collapsed across sub-
ordinate gender for all subsequent analyses. Next, we
examined the effects of gender, delegation condition,
and their interaction on negative affect. There were
no main effects of gender, F(1, 170) 5.49, p5.49, h
p2
5
.003, or delegation condition, F(1, 170) 5.19, p5
.66, h
p2
5.001, on negative affect. However, con-
sistent with our prediction, we observed a significant
gender 3delegation conditioninteraction, F(1, 170) 5
4.67, p5.03, h
p2
5.03. As shown in Figure 1a,
women who were told to delegate associated dele-
gation with greater negative affect (M51.36, SD 5
1.91) relative to men who were told to delegate (M5
.66, SD 51.38), F(1, 170) 54.47, p5.04, h
p2
5.03.
There was no difference in negative affect between
men (M51.41, SD 51.70) and women (M5.84,
SD 51.54) told not to delegate, F(1, 170) 50.86, p5
.35, h
p2
5.005. Nor were there significant differ-
ences between the delegate and do-not-delegate con-
ditions for women (p5.23) or for men (p5.07).
We had intended to examine the negative affect
experienced by women with greater granularity by
analyzing the LIWC negative emotion subscales
of anger, sadness, and anxiety. However, because
only 5%, 8%, and 20% of the sample listed words
associated with anger, sadness, and anxiety, re-
spectively, we did not have sufficient statistical
power to analyze these subscales.
We next examined the effects of gender, delegation
condition, and their interaction on positive affect.
We observed no significant effects of gender, F(1,
170) 50.24, p5.62, h
p2
,.01, delegation condition,
F(1, 170) 53.25, p5.07, h
p2
5.02, or gender 3
delegation condition interaction on positive affect,
F(1, 170) 50.30, p5.58, h
p2
,.001. (Figure 1b).
Guilt of overburdening subordinates.4Our ex-
amination of guilt associated with delegating (or
relief associated with not delegating) revealed a main
effect of delegation condition, F(1, 170) 534.86, p,
.0001, h
p2
5.16, such that those in the do-not-
delegate condition felt more relief from not having to
delegate (M55.10, SD 51.54) relative to the guilt felt
by those in the delegate condition (M53.65, SD 5
1.59). We observed no effects of gender on feelings of
guilt or relief, F(1, 170) 50.15, p5.70, h
p2
5.001,
nor was there a significant gender 3delegation
condition interaction on guilt or relief, F(1, 170) 5
0.05, p5.82, h
p2
5.0002.
Discussion
Supporting Hypothesis 2a, Study 2 showed that
women have different affective associations with
delegation compared to men, such that women asso-
ciate delegation with more negative affect compared
to men. This finding is consistent with research
demonstrating that engaging in counter-stereotypical
2We did not include word count as a covariate, but re-
sults do not change when including word count in the
analysis.
3Correlations across all study variables are available
upon request from the first author.
4We thank two anonymous reviewers for raising the
point that gender differences in delegation may be attrib-
uted to concerns about being perceived as less competent
for delegating. To rule out this alternative explanation, in
Study 2 we tested whether women were more concerned
about their own competence and their subordinate’s
competence compared to men. Counter to prevailing hy-
potheses, men were more concerned about how their
competence would be perceived (M52.68; SD 51.47)
than were women (M51.82; SD 51.09), F(1, 170) 518.71,
p,.001, h
p2
5.10, collapsing across conditions. Men were
also more concerned about their subordinate’s compe-
tence (M54.03; SD 51.30) than were women (M53.57;
SD 51.56), F(1, 170) 54.45, p,.04, h
p2
5.03, collapsing
across conditions. Further, we included competence con-
cerns (both self and subordinate) as control variables in our
analysis and saw no changes in our results. Third, in
a separate study we asked 46 female managers (M
age
5
44.76; SD 58.07) to rate the extent to which their concerns
about their competence, their subordinate’s competence,
and guilt of overburdening subordinates prevented them
from delegating. We found self-competence concerns to be
the lowest-rated reason for not delegating (M52.01; SD 5
1.45), relative to guilt (M53.43; SD 51.49) and sub-
ordinate competence (M54.32; SD 51.40). These findings
suggest that competence is not an alternative explanation
for our effects.
1474 AugustAcademy of Management Journal

(e.g., agentic) behavior can result in women feeling
concerned about appearing too agentic (Moss-
Racusin & Rudman, 2010). Importantly, controlling
for competence perceptions allowed us to rule out
the alternative explanation that women’s negative
associations with delegation are influenced by con-
cerns about being perceived as less competent for
delegating or by concerns regarding their sub-
ordinates’competence.
STUDY 3
In Study 3, we sought to test whether our findings
in Study 2 would replicate in a sample of managers
based on their daily delegation experiences. We
predicted that women delegators would have more
negative associations with delegation compared to
male delegators (Hypothesis 2a) and that women
would feel more guilt for overburdening their sub-
ordinates compared to men (Hypothesis 2b).
Method
Participants. We recruited 164 participants (79
women; M
age
532.90) from MTurk. Since holding
a managerial role was a requirement for participa-
tion, participants were asked to indicate whether
they were in a managerial role, list their job de-
scription, and report the kinds of tasks they tended to
delegate. The sample consisted of 77% Caucasian,
8% Asian, 7% Black, and 5% Hispanic participants,
with an average of 14.33 years of work experience.
FIGURE 1
Study 2: Percentage of (A) Negative Affect and (B) Positive Affect Words Used by Men and
Women by Delegation Condition
0.0
0.0
0.5
1.0
1.5
2.0
2.5
3.0
A
B3.5
Did Not Delegate
Did Not Delegate
Delegation Condition
Dele
g
ation Condition
Delegated
Delegated
% Positive Affect Words
(Square Root)
% Negative Affect Words
(Square Root)
Male
Male
Female
Female
0.5
1.0
1.5
2.0
2.5
3.0
3.5
Note: Error bars show standard errors.
2018 1475Akinola, Martin, and Phillips

Participants resided in the United States and 51%
had at least some post-secondary education. Partic-
ipants on average had worked at their organization
for 5.76 (SD 54.80) years, had 14.23 (SD 539.17)
people working under them, and worked 42.15
(SD 59.93) hours per week.
Procedure. The study was described as examining
day-to-day experiences and behaviors at work. Par-
ticipants were first asked about their job description
and tasks that they had authority to delegate. Those
with the authority to delegate (our key exclusion
criterion) were allowed to proceed and were asked
a number of questions about their self-reported del-
egation behavior, as well as their associations with
delegation.
Measures. Participants were asked to respond
to self-report measures on scales ranging from 1
(strongly disagree)to7(strongly agree). To test par-
ticipants’unprimed associations with delegation, we
asked them to list any thoughts, feelings, and emo-
tions they associated with delegation in an open-
ended format. We asked two questions designed to
capture guilt that may be associated with delegation
and overburdening an employee: “I sometimes feel
bad delegating because my employee has other
things to do”and “I feel some guilt delegating be-
cause my employee might be overburdened”(r5
.81). We also asked two questions on how much
participants delegate (r5.59): “Compared to other
people in your position, how much do you delegate
to your employees?”and “How effective do you
think you are at delegating?”As an attention check
during the study, participants received a question,
embedded within our scales, asking them to click
“strongly disagree”to show that they were paying
attention.
Data-analysis strategy. As in Study 2, we ana-
lyzed written responses using LIWC. We tabulated
the overall number of positive and negative emotion
words for each response and examined three
subscales of negative emotions: anger, sadness, and
anxiety. Due to skewness, data were transformed
with a square-root function. Additionally, we con-
ducted analyses of variance on each of the measures
listed above to examine gender differences on each
variable.
Results
Participant attritio n. Nineteen participants (12%)
were removed for lacking authority to delegate. Of
the remaining participants, 14 (10%) did not report
their work experience or failed the attention check.
The final sample was 130 participants (66 women;
M
age
534.02 years).
Affective associations with delegation. To test
whether men and women had different associations
with delegation, we examined participants’free-
response texts. On average, participants used 9.78
(SD 513.97) words, and LIWC captured 90% of the
words used. An examination of the captured nega-
tive and positive affect words associated with dele-
gation revealed that, consistent with Hypothesis 2a,
there was a significant main effect of gender on
negative affect, F(1, 128) 56.99, p5.009, h
p2
5.05,
such that women listed significantly more negative-
affect-based words (M53.31, SD 53.57) than did
men (M51.87, SD 52.55). Means by gender for all
key variables are presented in Table 1.
An examination of the three subscales of negative
affect revealed that only one participant (,1% of
sample) used “sad”or “angry”affect-based words;
as such, we did not analyze these subscales. A larger
proportion of participants (40%) used anxiety-
related words, which allowed us to analyze the
LIWC anxiety subscale. We observed a significant
main effect of gender on anxiety, F(1, 128) 510.26,
p5.002, h
p2
5.07), with women listing significantly
more anxiety-based words (M52.93, SD 53.58)
compared to men (M51.26, SD 52.16).
TABLE 1
Study 3: Means by Gender and Analysis of Variance Results for Key Dependent Variables
Men Women
MSDMSD F p h
p2
Self-Reported Delegation 4.64 1.18 4.78 1.08 0.50 0.480 0.00
Negative Affect (Sqrt) 1.87 2.55 3.31 3.57 6.99 0.009 0.05
Anxiety (Sqrt) 1.26 2.16 2.93 3.58 10.26 0.002 0.07
Positive Affect (Sqrt) 3.82 3.27 2.77 3.02 3.63 0.059 0.03
Guilt 4.02 1.53 4.55 1.42 4.07 0.046 0.03
Note:n5130.
1476 AugustAcademy of Management Journal

Finally, there was a marginally significant effect of
gender on positive affect, F(1, 128) 53.63, p,.06,
h
p2
5.03, such that women listed marginally fewer
positive-affect-based words (M52.77, SD 53.02)
compared to men (M53.82, SD 53.27).
Guilt of overburdening subordinates. There was a
significant main effect of gender on guilt, F(1,129) 5
4.07, p,.05, h
p2
5.03, such that women expressed
greater guilt about their employee being over-
burdened (M54.55, SD 51.42) compared to men
(M54.02, SD 51.53), supporting Hypothesis 2b.
Self-reported delegation. We examined whether
self-reported delegation behavior differed between
men and women. Interestingly, we observed no
gender differences in participants’reporting of how
much they delegated, F(1, 128) 5.50, p5.48, h
p2
,
.00, with men (M54.64, SD 51.18) reporting similar
levels of delegation as women (M54.78, SD 51.08).
We examined the two measures (i.e., how much
participants delegated and how effective they were
at delegating) independently and observed no sig-
nificant effects (pvalues..19).
Discussion
Consistent with Study 2, the results of Study 3
offer support for Hypothesis 2a and demonstrate
that women associate delegation with more nega-
tive affect, namely anxiety, as well as with less
positive affect, than do men. Further, we extended
these findings to show that women associate del-
egation with greater guilt compared to men with
regard to overburdening subordinates, supporting
Hypothesis 2b. Although Study 3 offers pre-
liminary evidence of gender differences in dele-
gation, we found that women’sself-reportsof
delegation indicated that they did not delegate
any less than men. Despite the lack of gender
differences in our self-reported measure, we did
find that negative affect was negatively correlated
with self-reported delegation (see Table 2); those
who had stronger negative associations with dele-
gation self-reported less delegation behavior.
Since women have more negative associations
relative to men, this finding suggests an indirect
relationship between gender and self-reported
delegation through negative associations. Given
the limitations of self-reportbehavioralmeasures
(Baumeister, Vohs, & Funder, 2007) and the sub-
jective nature of our measures, which allowed
participants to qualify the quality and quantity of
their delegation, it is possible that the items we
selected to capture delegation were not sufficient
to capture differential delegation behavior be-
tween women and men. Further, the job charac-
teristics of the sample (i.e., industry, tenure,
occupation, position) varied significantly; thus,
a variety of important employment circumstances
may have affected men’sandwomen’sself-
reported delegation. Therefore, in Study 4, we ex-
amined gender differences in actual delegation
behavior and controlled for potential confounds
present in Study 3.
STUDY 4
In Study 4, we turned our attention to the impact of
negative associations with delegation on women’s
propensity to delegate. In addition, we examined
how these negative associations affected men’s and
women’s interactions with their subordinates and
influenced their performance on a set of tasks. To this
end, we studied delegation in a controlled laboratory
experiment employing a confederate as a sub-
ordinate. We sought to replicate and extend Hy-
pothesis 2a by testing whether women feel more
TABLE 2
Study 3: Means, Standard Deviations, and Correlations
MSD 123456
1. Gender 1.51 0.50 –
2. Negative Affect (Sqrt) 2.60 3.18 .23** –
3. Anxiety (Sqrt) 2.11 3.07 .27** .91** –
4. Positive Affect (Sqrt) 3.28 3.18 –.17
†
–.36** –.29** –
5. Guilt 4.29 1.49 .18* .17
†
.18* .07 (0.81)
6. Self-Report Delegation 4.71 1.13 .06 –.21* –.12 .11 –.17* (0.59)
Notes:n5130. Gender is coded such that 1 5Male and 2 5Female. Reliabilities appear on the diagonal in parentheses (where applicable).
†p,.10
*p,.05
**p,.01
2018 1477Akinola, Martin, and Phillips

negative affect while delegating, and to provide
richness to our measures by collecting emotions
before and after the opportunity to delegate (rather
than through affective associations devoid of con-
text). Specifically, we measured anxiety before and
guilt after delegation, and included a fear-of-
backlash scale to test our argument that the anxi-
ety associated with delegation is due to a fear of
being perceived as too agentic. Finally, we exam-
ined the downstream consequences of having
negative associations with delegation, testing our
predictions that women will be less likely to del-
egate than men (Hypothesis 3a), that women will
fear backlash for delegating more than men will
(Hypothesis 2c), and that women’sinteractions
with their subordinates when delegating will be
perceived as being of lower quality than men’s
(Hypothesis 3c). We further explored whether
women’sandmen’s differential negative associa-
tions with delegation will mediate their likelihood
of delegating (Hypothesis 3b), and the implications
of delegation on actual performance in this context
(Hypothesis 3d).
Method
Participants. We recruited 148 participants (81
women; M
age
522.77) from a behavioral research lab
at a large, private university. Participants were
scheduled for a 90-minute session and paid $20 plus
a $4 performance bonus. The majority (62%) of
participants were undergraduate students, while the
remaining 38% were graduate students or commu-
nity members. The sample was 77% Caucasian and
11% Asian, while 3% identified with more than one
race.
Procedure. The study, described as being about
identity and performance, had three phases. In
Phase 1, participants first answered personality
questionnaires. In Phase 2, they engaged in an ac-
tivity with a partner (a trained confederate).5In Phase
3, participants completed a questionnaire about their
interactions and experience with their partner dur-
ing Phase 2.
Phase 1: Participants entered the lab with a person
who was ostensibly another participant (confeder-
ate); the two were escorted into separate rooms. After
providing consent, participants were asked to com-
plete an “Activity Role Assessment”and told that
this assessment would be used to determine their
roles in an upcoming leadership activity.6
Participants were then told that, based on their
assessments, they would be put into the senior role of
“Sponsorship Coordinator”for a “University Goes
Green”campus campaign and that their partner
(confederate) would be put into the junior role of
“Committee Member.”This paradigm has been
found to create status and power differences in
experimental settings (Akinola & Mendes, 2013).
Participants were told that both they and the “Com-
mittee Member”had six tasks to complete, but that
their tasks and compensation would differ based on
their leadership roles, and as the leader, they could
delegate to their partner if they would like to. The
confederate then went to a separate workspace and
participants completed questionnaires assessing
their associations with delegation, and their inten-
tions to delegate in the upcoming activity.
Phase 2: Participants were given 35 minutes to
complete an activity consisting of si x tasks in their role
of sponsorship coordinator: (1) draft an email to the
student body about the campaign, (2) enter data into an
Excel spreadsheet, (3) redact or interpret the data en-
tered into the Excel spreadsheet, (4) proofread a Pow-
erPoint presentation, (5) format a newsletter, and (6)
design a logo for the campaign (see Appendix A for
details on each task). Each participant was videotaped
during the session in an effort to capture his or her
delegation behavior. After 35 minutes had elapsed,
participants were informed that the activity was over.
Phase 3: Participants then filled out a final ques-
tionnaire and were debriefed, thanked, and paid.
Dependent measures prior to the delegation activ-
ity (Phase 1). Prior to engaging in the delegation
activity, participants completed the role assessment
questionnaire as well as questionnaires assessing
their affective associations with, and likelihood
of, delegating.
This questionnaire served as a cover story for
assigning participants to the senior role of “Sponsor-
ship Coordinator.”The questionnaire included the
Ten Item Personality Inventory (Gosling, Rentfrow, &
Swann, 2003) and three Rorschach Inkblot Tests
5We manipulated and controlled for confederate gender
in all analyses.
6While participants waited for their assessments to be
scored, they answered questions about themselves for
eight minutes either alone or with the confederate. We
were interested in whether forming a relationship with an
employee would affect women’s propensity to delegate.
We observed no effect of this manipulation on our key
dependent variables. In all analyses, we controlled for self-
reflection condition and confederate gender.
1478 AugustAcademy of Management Journal

(Karson, 2007), in which participants viewed ambig-
uous pictures and described what they saw.
To assess the changes in emotions that may ac-
company delegation, we assessed participants’self-
reported emotions using the Positive and Negative
Affect Schedule (PANAS) (Watson & Clark, 1994) at
two points: (1) upon arrival at the lab in the activity
role assessment questionnaire (Time 1) and (2) after
learning about the delegation activity and their lead-
ership role (Time 2). Participants rated themselves on
15 negative and 19 positive items. However, we were
particularly interested in the anxiety subscale of the
PANAS given our findings that women had more
negative associations, namely anxiety, with delega-
tion in Study 3. Participants rated how nervous,
afraid, and scared they were on scales ranging from 1
(not at all)to5(a great deal), resulting in an anxiety
index (alphas ranged from .71 to .85). We also exam-
ined three positive emotions (confident, proud, and
strong), as these words were most relevant based on
the open-ended content of Studies 2 and 3, resultingin
apositiveaffectsubscaleindexingagenticemotions
(alphas ranged from 78 to .82). We calculated changes
in emotions by subtracting participants’emotion rat-
ings just prior to the delegation task (Time 2) from
those collected upon arrival at the lab (Time 1).
Dependent measures during delegation activity
(Phase 2). We assessed whether participants dele-
gated, examined their behavior toward their sub-
ordinate when they did delegate, and measured their
performance. We assessed delegation by having
confederates indicate whether the participant dele-
gated any tasks to them and how many tasks they
delegated. Additionally, independent coders exam-
ined each videotape to confirm whether the partici-
pant delegated. For each task that was delegated, we
measured the number of seconds that elapsed be-
tween the time the participant entered the confed-
erate’s room to delegate a task and the time the
participant exited the room.
We assessed the quality of participants’interac-
tions with the confederate7in two ways, using items
based on the “showing concern”dimension of the
empowering leadership questionnaire (Arnold,
Arad, Rhoades, & Drasgow, 2000). First, immedi-
ately following each interaction, we asked each
confederate to rate the participant on six questions:
(1) “the participant seemed to consider how their
delegation would affect me,”(2) “the participant
kept me motivated throughout the task,”(3) “Ifelt
like the participant had confidence in me to com-
plete the tasks,”(4) “Ifeltliketheparticipanttrusted
me to complete the tasks,”(5) “the participant fol-
lowed up with me to make sure I was completing my
tasks,”and (6) “the participant checked in on me to
make sure I was completing my tasks.”Ratings were
on scales from 1 (strongly disagree)to7(strongly
agree)(a5.68).
Second, five research assistants watched each
participant’s video and rated participants on the
following four questions: To what extent did the
participant in the video... (1) “ask about the con-
federate’s to-do list to see if they had the capacity to
take on the task,”(2) “compliment the confederate on
his or her ability to accomplish the task,”(3) “express
gratitude for taking on or finishing the task,”and (4)
“offer to help, if there was extra time.”Coders rated
these items, which capture how considerate the
participant appeared, on scales ranging from 1 (not at
all)to7(very much). Reliability between the coders
was high (a5.88), as was the reliability between the
items (a5.90). Further, confederate ratings of their
interactions and coder ratings were strongly corre-
lated, r5.57, p,.001, indicating consistency in
their perceptions of interaction quality.
Each of the six tasks was scored by two research
assistants on several dimensions (see Appendix A).
We created an average performance score by stan-
dardizing the different metrics of performance across
our tasks (using a z-score). Participants who dele-
gated a task to the confederate received full points
for that task.
Dependent measures following the delegation
activity (Phase 3).8To understand the theoretical
drivers of negative associations found in previous
studies, we asked participants who had delegated to
complete a questionnaire assessing guilt and fear of
backlash.
We measured guilt associated with overburdening
with the same two items adapted from Study 3 (“I felt
bad delegating because the Committee Member had
7To ensure gender-role stereotyping did not explain the
ratings of interaction quality, we transcribed participants’
interactions and had two research assistants, blind to
participant gender, rate each transcript on the same di-
mensions on which participants were rated by confeder-
ates and video coders. Blind transcript coders’ratings
were significantly correlated with those of confederates
(r5.39, p,.001) and with those of video coders (r5.46,
p,.001), suggesting that ratings of consideration were
not unduly influenced by gender-role stereotypes.
8Since we were interested in men’s and women’s af-
fective responses to delegation, these items were only
asked of participants who delegated.
2018 1479Akinola, Martin, and Phillips

other things to do”and “I felt some guilt delegating
because I thought the Committee Member might be
overburdened”) using scales ranging from 1 (strongly
disagree)to7(strongly agree). We also added a third
item, “I was concerned that the Committee Member
may not receive their full payment because they were
working on my tasks”(a5.82).
To assess participants’concerns about backlash,
they were told to imagine that a group of people was
watching a tape of their interaction with the con-
federate and asked to what extent they would be
concerned that... “[they] might be disliked?”“peo-
ple would think [they] dominated the interaction?”
“others would perceive [them] as too controlling?”
and “[they] might be judged for being too assertive?”
(a5.83). These items were anchored at 1 (not at all)
to 7 (very much) (see Moss-Racusin & Rudman,
2010).
Results
Participant attrition. Eleven participants were
removed from the analysis due to their suspicions
about their interaction with the confederate, as dis-
covered in participant debriefing and in confederate
notes. The final sample included 137 participants
(75 women; M
age
521.97). Additionally, three
videos were not coded due to technical malfunc-
tions, and one participant did not complete the
PANAS correctly. Varying degrees of freedom reflect
this data loss.
Our primary prediction was that women would
feel greater increases in anxiety relative to men when
anticipating delegation. In line with our hypothesis,
there was no main effect of gender on changes in
anxiety, nor was there a main effect of delegation on
changes in anxiety (ps..54). However, we did ob-
serve a significant gender 3delegation interaction,
F(1, 130) 54.04, p,.05, h
p2
5.03. See Figure 2a.
Planned contrasts revealed that women who
delegated had greater increases in anxiety prior to
being given the opportunity to delegate (M5.64,
SD 5.88) than women who did not (M5.31, SD 5
.68), F(1, 130) 53.57, p5.06, h
p2
5.03. For men,
there were no differences in anticipatory anxiety
based on whether they chose (M5.28, SD 5.71) or
chose not (M5.51, SD 5.90) to delegate, F(1, 130) 5
1.07, p5.30, h
p2
5.01.
We conducted a similar analysis for our positive
emotions subscale. We found no main effect of gen-
der, F(1, 130) 5.83, p5.36, h
p2
5.01. However,
there was a main effect of delegation on positive af-
fect, F(1, 130) 54.67, p5.03, h
p2
5.04, such that
those who delegated felt more positive emotions
prior to being given the opportunity to delegate
(M5.11, SD 5.67) than did those who did not
delegate (M52.10, SD 5.67). This main effect was
qualified by a gender 3delegation interaction, F(1,
130) 53.90, p5.05, h
p2
5.03. As shown in Figure 2b,
planned contrasts reveal that for women there was
no difference in positive affect prior to being given
the opportunity to delegate, regardless of whether
they chose (M5.04, SD 5.64) or chose not (M5.02,
SD 5.65) to delegate, F(1, 130) 5.02, p5.88, h
p2
,
.01. In contrast, men who did not delegate experi-
enced significantly greater decreases in positive af-
fect prior to being given the opportunity to delegate
(M52.32, SD 5.67), and men who did delegate
experienced increases in positive affect prior to be-
ing given the opportunity to delegate (M5.17, SD 5
.70), F(1, 130) 57.28, p5.01, h
p2
,.05.
We used a binary logistic regression to test
whether men and women differentially delegated
and observed a significant main effect of gender,
such that women (51%) delegated less than did men
(68%), B52.72, SE 5.36, p,.05, Wald z54.01.
Interestingly, for those who delegated, there was no
difference between men and women in the amount of
items delegated, F(1, 76) 5.42, p5.52, h
p2
,.01.
There was a significant main effect of gender on
the time spent interacting with the confederate while
delegating, F(1, 70) 54.85, p5.03, h
p2
5.07,9with
women spending fewer seconds with the confeder-
ate when they delegated (M533.28, SD 525.98)
compared to men (M554.20, SD 559.68).
There was also a significant main effect of gender
on confederates’ratings of their interactions with
participants, F(1, 75) 511.07, p5.001, h
p2
5.13).
Confederates rated female participants as being less
considerate, trustworthy, motivating, and support-
ive (M53.25, SD 51.01) compared to male partic-
ipants (M53.96, SD 51.15). Additionally,
independent coder ratings yielded a marginally sig-
nificant effect of gender with regard to how consid-
erate the participant appeared, F(1, 73) 53.48, p,
.07, h
p2
5.05. Namely, the coders rated female par-
ticipants as less considerate (M52.94, SD 51.22),
compared to male participants (M53.44, SD 51.16)
in their interactions.
While we observed no effect of gender on total
performance score, F(1, 129) 5.02, p5.89, h
p2
,.00,
there was a significant main effect of delegation on
9We controlled for number of items delegated to the
confederate, which could have affected the amount of time
spent interacting with the confederate.
1480 AugustAcademy of Management Journal

performance, F(1, 129) 510.44, p5.002, h
p2
5.08.
Participants who chose to delegate performed sig-
nificantly better (M5.11, SD 5.45) than those who
did not delegate (M52.15, SD 5.44).
In sum, we found that women delegated less than
men when given the opportunity to do so. Further,
women spent less time and showed less consider-
ation toward their subordinate when they did
FIGURE 2
Study 4: Change in (A) Anxiety and (B) Positive Affect for Men and Women by Delegation
Change in Positive Affect (T2–T1) Change in Anxiety (T2–T1)
Did Not Delegate Delegated
Did Not Dele
g
ate Dele
g
ated
Male
Female
Male
Female
.90
.80
.70
.60
.50
.40
.30
.20
.10
.00
.60
.40
.20
.00
-.20
-.40
-.60
-.80
A
B
Note: Error bars show standard errors.
2018 1481Akinola, Martin, and Phillips

delegate. Finally, we found that delegation influ-
enced performance, as those who delegated com-
pleted more tasks.
Dependent measures following the delegation
activity were only given to those who delegated, as
the questions were specific to their interaction and
how they felt about delegating. There was a signifi-
cant main effect of gender for those participants
who delegated in the amount of guilt they felt about
overburdening their subordinate, F(1, 76) 55.24,
p,.03, h
p2
5.06, such that women who delegated
felt more guilt (M55.36, SD 51.41) compared to
men who delegated (M54.65, SD 51.49). There
was also a significant main effect of gender for par-
ticipants who delegated in terms of how much they
feared others viewing them negatively, F(1, 75) 5
4.26, p5.04, h
p2
5.05, such that women who
delegated felt more fear of backlash (M52.91,
SD 51.38) than did men who delegated (M52.34,
SD 51.10).
Given that we found significant differences in
men’sandwomen’s positive and negative affective
responses to delegation, as well as their likelihood
to delegate, and delegation behavior, it is possible
that men’sandwomen’s differential affective re-
sponses mediated their likelihood of delegating
(Hypothesis 3b) and associated behaviors. We
tested for the presence of an indirect effect of gender
on delegation and interaction quality by calculating
95% confidence intervals (CIs) using the PROCESS
macro for SPSS (Hayes, 2013) with 5,000 bootstrap
resamples. We specifically examined negative and
positive affect as mediators predicting delegation
and interaction quality, as both of these measures
were assessed prior to participants deciding
whether to delegate. We were unable to examine
guilt and fear of backlash as mediators, as they were
assessed after delegation and only for those who
delegated.
Results from mediation analyses examining neg-
ative and positive affective responses to delegation,
using binary logistic regression, indicated that neg-
ative affective responses did not mediate the effects
of gender on delegation (CI
95
52.02, .23); nor did
positive affective responses (CI
95
52.12, .18), as
both confidence intervals contained zero. Negative
affective responses also did not mediate the effects of
gender on the quality of women’s interactions with
subordinates when delegating (CI
95
52.007, .28),
nor did positive affective responses (CI
95
52.21,
.03). Despite our nonsignificant mediations, we
found a marginal relationship between gender
and interaction quality through anxiety (indirect
effect 5.07, SE 5.07, CI
90
5.002, .24). Though in-
conclusive, this finding suggests that greater anxiety
may influence women’s delegation behavior with
subordinates, even if it does not influence their de-
cision to do so.
Discussion
The goal of Study 4 was to replicate our previous
findings that women have more negative associations
with delegation, and to expand on these findings to
show that these negative associations have implica-
tions for actual (and not just self-reported) delegation
behavior. In a controlled experiment we found that
women delegated less than did men, consistent with
Hypothesis 3a, and that when women did delegate,
they spent less time with their subordinates and had
lower-quality interactions, a finding aligned with
Hypothesis 3c. Further, we offer evidence of the dif-
ferent types of negative associations women have
with gender-role-paradoxical behaviors by showing
that women experience relational guilt and fear of
backlash about delegating, supporting Hypotheses 2b
and 2c. Notably, keeping performance of the confed-
erate constant, we found that participants who did
delegate performed better on the task than those who
did not, as they had more time to accomplish their
tasks. Specifically, women delegators outperformed
women who chose not to delegate. These findings
replicate prior research showing the adaptive nature
of delegation (Leana, 1987) and offer evidence that
women’s negative associations with delegation may
have negative repercussions for workplace perfor-
mance when they choose not to delegate.
The fact that we did not observe a gender effect
or a significant gender 3delegation interaction on
performance suggests that women did not perform
worse than men on the task, even though they dele-
gated less; instead, they essentially outperformed
men through their own contributions. However, it is
unclear whether women could sustain this out-
performance across multiple tasks or days. Because
the benefits of delegating may be cumulative, easing
the superior’s workload over time, the potential
negative performance implications for women of not
delegating might be difficult to capture based on
a single point in time.
In addition, Study 4 provides insight into the role
affect plays in delegation decisions. We found that
women who delegated self-reported more anxiety
prior to delegating than did men. This finding sug-
gests that prospective anxiety does not necessarily
hinder women from delegating. On one hand, it is
1482 AugustAcademy of Management Journal

possible that the negative associations women have
with delegation prevent them from delegating, such
that women who associate delegation with anxiety
do not delegate. On the other hand, it is possible that
some women press ahead and delegate despite feel-
ing anxious about doing so. Our results in Study 4
suggest that prospective anxiety may not only in-
fluence the decision to delegate, but also has impli-
cations for the quality and quantity of women’s
interactions with their subordinates when they
do delegate. Nonetheless, we did not find
conclusive—though we did find suggestive—
evidence that prospective anxiety (i.e., negative af-
fect) mediated the effects of gender on delegation
behavior.
Given our findings in Study 1 demonstrating that
women perceive delegation as more agentic than
men, our findings in Study 4 suggest that women
internalize these social role perceptions, as evident
in their emotions before, their behavior during, and
their reflections after delegating. Women’s concerns
about violating their social role may degrade their
interactions with subordinates, leading them to rush
through delegating (evidenced by shorter interaction
times) and to become more self-focused (evidenced
by being rated as less considerate) as compared to
men. Interestingly, men who did not delegate expe-
rienced less positive affect relative to those who did
delegate. These male nondelegators violated gender
roles by not acting agentically, which influenced
their emotions. This finding supports the concept
that gender roles play an important part in responses
to delegation behavior for both men and women.
STUDY 5
In Study 5, we sought to explore one possible
factor that might attenuate gendered responses to
delegation behavior, particularly for women. Be-
cause we theorized that women have negative
associations with delegation due to its gender-role-
paradoxical nature of being communal yet agentic,
we tested whether framing delegation as more
communal would attenuate women’s negative as-
sociations with delegation. We examined whether
highlighting how delegation can empower and help
subordinates would reduce women’s negative as-
sociations with delegation.
Method
Participants. One hundred and thirty-five MBA
students (47 women; M
age
528.11), enrolled in an
introductory organizational behavior course at a
large private university on the East Coast, partici-
pated in the study. Fifty-nine students were in
a managerial position prior to pursuing their MBA,
and participants on average had 14.35 (SD 543.84)
subordinates. The sample was 46% White, 27%
Asian, 11% Hispanic, 3% Black, and 13% other.
Procedure. We told participants that they would
engage in an in-class feedback-giving and -receiv-
ingroleplaybetweenamanagerandsubordinate,
and randomly assigned them to either be the manager
(feedback giver) or the subordinate (feedback re-
ceiver). Participants first read their role materials,
which provided context for the role play. Sub-
ordinates all read the exact same role materials, in
whichtheywereinformedthattheywouldmeet
with their manager to receive performance feed-
back and that some tasks might be delegated to
them. Managers were told that they would give
their subordinate performance feedback and that
there were three tasks that could be delegated to
subordinates. After reading the role materials, but
prior to engaging in the role play, both managers
and subordinates answered questions about their
emotions.
We manipulated whether managers were told that
delegatingtasks can help subordinates develop, learn,
and grow (delegation-as-communal condition) or
were told nothing about the communal benefits that
could accrue to the subordinate from performing
delegated tasks (control condition). We chose
a manipulation that emphasized the communal
properties of delegation, and not onethat emphasized
the benefits of engaging in agentic behaviors, for two
reasons. First, much research has shown that high-
lighting communal properties of agentic behaviors
can attenuate fear and increase the likelihood of en-
gaging in such behaviors (see Amanatullah & Morris,
2010; Kray, Galinsky, & Thompson, 2002; Small et al.,
2007). Thus, we expected highlighting communality
to be beneficial here in the domain of delegation, just
as it has been shown to be beneficial for negotiations.
Second, agency is generally perceived to be adaptive
in leadership roles. Thus, it is not that women do not
know the benefits of engaging in agentic behavior;it is
more likely that gender-role prescriptions and back-
lash prevent women from engaging in agentic be-
haviors (Rudman & Phelan, 2008). We were not able to
include other control conditions but see this initial
test of intervention as a first step toward potentially
mitigating gendered concerns about delegation.
Materials and measures. Participants were
asked to read the role play (with the embedded
2018 1483Akinola, Martin, and Phillips

manipulation for managers) and respond to mea-
sures capturing their affective responses, guilt
from overburdening their subordinate (for man-
agers), and evaluations of their manager (for
subordinates).
Participants in the role of manager were ran-
domly assigned to either a control condition or a
delegation-as-communal condition prior to engag-
ing in the feedback role play. These conditions
differed only in the instructions participants were
given at the end of the exercise. All managers were
told the following:
You are planning to delegate a few tasks on your plate
to your subordinate. Among these are: (1) preparing
materials for a client meeting, (2) reconciling ex-
penses, and (3) running a training session. You rec-
ognize that [your subordinate] has been working hard
and is very busy, but you need help.
In the delegation-as-communal condition, this state-
ment was followed by:
Further, you know that by assigning these tasks, you
are being a good mentor, helping [your subordinate] to
develop, and teaching your subordinate critical skills
that are needed for the senior management position.
As in Study 4, participants rated how nervous,
afraid, and scared they were on scales ranging from 1
(not at all)to5(a great deal), resulting in an anxiety
index (a5.76). We also examined three positive
emotions (confident, proud, and strong), resulting in
a positive affect scale (a5.67).
To capture guilt associated with overburdening an
employee, we asked those in the manager role the
three questions used in Study 4 (a5.77) on scales
ranging from 1 (strongly disagree)to7(strongly
agree).
FIGURE 3
Study 5: (A) Anxiety and (B) Positive Affect for Men and Women by Delegation Condition
0.0
2.7
2.8
2.9
3.0
3.1
3.2
3.3
3.4
3.5
3.6
3.7
0.5
1.0
1.5
2.0
2.5
3.0
Delegation as Communal
Delegation as Communal
Control
Control
Delegation Condition
Dele
g
ation Condition
Male
Male
Positive Affect Anxiety
A
B
Female
Female
Note: Error bars show standard errors.
1484 AugustAcademy of Management Journal

To ensure that the manager and the subordinate
were both acclimated to their respective roles, we
asked those in the subordinate role about their
desire to take on their managers’tasks using two
questions, “I am motivated to take on my man-
ager’stasks”and “I am happy to take on my man-
ager’swork”(r5.71), on scales ranging from 1
(strongly disagree)to7(strongly agree). Since there
was no manipulation in the subordinate role, we
did not expect any gender differences in these
measures.
Results
Affective responses to delegation. Our primary
prediction was that women managers in the control
condition would feel greater anxiety than those in
the delegation-as-communal condition when antic-
ipating delegating to the subordinate. We did not
expect any differences in anxiety by delegation
condition for male managers. In line with our pre-
diction, there was no main effect of gender on anxi-
ety, F(1, 68) 50.01, p5.90, h
p2
,.00, nor was there
a main effect of delegation condition on anxiety, F(1,
68) 52.68, p5.11, h
p2
,.04. However, as can be
seen in Figure 3, we did observe a marginally sig-
nificant gender 3delegation condition interaction,
F(1, 68) 53.55, p5.06, h
p2
5.05.
Planned contrasts revealed that women in the
control condition had greater anxiety prior to dele-
gating (M52.31, SD 5.76) than women in the
delegation-as-communal condition (M51.67, SD 5
.55), F(1, 23) 54.75, p5.03, h
p2
5.07. For men, there
were no differences in anxiety by delegation condi-
tion, F(1, 44) 50.04, p5.84, h
p2
5.00.
We conducted a similar analysis for our positive
emotions subscale and found no main effects or
significant interactions (ps ..29). We also con-
ducted this analysis for subordinates’ratings of
anxiety and positive emotions. As expected, be-
cause there was no manipulation in the subordinate
role, we found no differences between men and
women (ps ..57).
Guilt of overburdening subordinate (managers
only). There was a significant main effect of delega-
tion condition in the amount of guilt managers felt
about overburdening their subordinate in anticipa-
tion of delegating, F(1, 68) 57.04, p,.01, h
p2
5.10,
such that managers in the control condition felt more
guilt (M54.03, SD 51.13) than did managers in the
delegation-as-communal condition (M53.37, SD 5
1.27). However, there was no main effect of gender,
F(1, 68) 51.46, p5.23, h
p2
5.02, nor was there
a significant gender 3delegation condition effect for
guilt of overburdening one’s subordinate, F(1, 68) 5
1.99, p5.16, h
p2
5.03.
Desire to take on manager’s tasks (subordinates
only). Since there was no manipulation in the sub-
ordinate role, we did not expect there to be differ-
ences between men and women in their desire to take
on their manager’s tasks. Consistent with our ex-
pectations, we found no gender differences in this
variable (p5.48).
Discussion
The goal of Study 5 was to demonstrate that a
simple manipulation emphasizing the communal
aspects of delegation can attenuate women’sneg-
ative affective responses to delegation. We found
that women who were told delegation could help
their subordinate develop, learn, and grow felt less
anxiety and less guilt about overburdening their
subordinate prior to delegating, compared to
women who were not given these instructions and
compared to male managers. Taken with Study 4,
in which we found that women’s negative associ-
ations had implications for their actual delegation
behavior and performance, our findings in Study 5
suggest that reducing women’s negative associa-
tions with delegation may increase their likelihood
of delegating, improving interactions with sub-
ordinates and potentially positively affecting their
performance.
GENERAL DISCUSSION
Thegoalofthecurrentresearchwastoexamine
whether men and women have different associa-
tions with delegation, and to explore how these as-
sociations may differentially affect their propensity
to delegate, behavior during delegation, and ulti-
mately performance. To this end, in Study 1 we find
that women perceive delegation as more agentic
than men do. In Studies 2 and 3, we find that wo-
men associate delegation with more negative affect
compared to men. Although women’s self-reports
of delegation indicated that they did not delegate
any less than men, our behavioral examination of
delegation in Study 4 demonstrated that women
did indeed delegate less than men. Further, when
women did delegate, they spent less time with their
subordinates and had lower-quality interactions
compared to men. Confederate ratings highlighted
the nature of these interactions by demonstrating
that women who delegated were less likely to check
2018 1485Akinola, Martin, and Phillips

in and follow up than were men, leading to lower-
quality interactions in which the subordinate felt
unsupported, less trusted, and less motivated.
Moreover, women felt greater guilt after delegating
than men did and were more likely to fear backlash
for being perceived as too agentic in their delegation
behavior. Importantly, women who did delegate
outperformed those who did not. Finally, in Study
5, we show that reframing delegation as communal
attenuates women’s negative affective associations
with delegation.
Theoretical Contributions
Collectively, our findings contribute to the litera-
ture on gender and leadership in several key ways.
First, our findings suggest that even when a leader-
ship behavior, such as delegation, could potentially
be seen as both communal and agentic in nature,
women seem to focus more on the agentic charac-
teristics than on the communal characteristics. Al-
though delegation can be considered communal, as it
can give subordinates opportunities to engage in
higher-level activities (De Pater et al., 2010), the
agentic nature of delegation, in which one exhibits
dominance and power, looms larger for women than
for men. This perspective may help to explain why
women experience greater fear of backlash and guilt
when delegating. Building on the body of research
that touts the advantages of women’s more commu-
nal, transformational style of leadership (Eagly &
Karau, 2002), our findings demonstrate that, in
certain contexts, even when a behavior could be
construed and enacted in a communal manner, tra-
ditional notions of agency associated with certain
behaviors have a powerful effect on women’s ability
to enact a communal style of leadership, ironically
resulting in them leading with a less trans-
formational and communal style. As women con-
tinue to ascend to higher levels in organizations, it
will be important to further develop our under-
standing of the role of gender-role incongruence in
women’s propensity to engage in behaviors that are
both agentic and communal in nature. Beyond del-
egation, research on managerial behaviors such as
negotiation, and even the act of leadership itself, of-
fer convergent support that women experience be-
haviors that are considered both agentic and
communal differently than men do, in part due to
gender-role incongruence. Further research is
needed to enhance our understanding of other cru-
cial leadership behaviors that fall under this category
of being gender-role paradoxical.
Second, our findings offer insight into how engag-
ing in gender-role-paradoxical behaviors may influ-
ence women’s relationships with their subordinates
and subsequent performance. Because effective del-
egation requires a productive relationship with sub-
ordinates, our finding that men and women treat their
subordinates differently when delegating has impli-
cations for the development of effective and pro-
ductive long-term work relationships. Ironically,
rather than exhibiting gender-role-congruent behav-
iors, such as showing warmth and sensitivity when
delegating, women appeared less considerate and
spent less time withtheir subordinates as compared to
men. Given that employees are affected by the mood
and behaviors of their leaders (Sy, Cote, & Saavedra,
2005), the behavioral spillover effects of having
more negative and fewer positive emotions associated
with delegating have important implications for
work-related outcomes, including performance, co-
ordination, and relationship development among
employees. Moreover, our finding that delegators
performed better on the task than did nondelegators
further emphasizes the importance of overcoming
these negative emotions associated with delegation.
Interestingly, although our focus was on the im-
plications of gender-role-paradoxical behaviors for
women’s experiences in organizations, we found
that men who did not delegate, presumably violating
their expected social role (e.g., being agentic and self-
assertive), experienced less positive affect. Given
that research on gender and leadership has mostly
examined the repercussions of gender-role conflict
for women (Rudman & Phelan, 2008), our findings
among males highlight a need to further examine
how gender-role expectations negatively influence
both women and men, potentially preventing men
from enacting more communal and less hierarchical
styles of leadership.
Finally, and critically, our research uncovers an
important remedy that can help encourage women
to engage in gender-role-paradoxical behaviors. Con-
sistent with research demonstrating the powerful ef-
fects of framing in shaping behavior (Kray et al.,
2002; Small et al., 2007), we show that women’s
affective associations with gender-role-paradoxical
behaviors can be changed by emphasizing the com-
munal properties of these behaviors.
Limitations and Future Directions
While the present studies offer implications for
our understanding of gender and delegation, it does
so with several limitations, many of which offer
1486 AugustAcademy of Management Journal

promising avenues for future research. First, the fact
that we did not observe differences in self-reported
delegation behavior between male and female man-
agers in Study 3, but did observe differences in actual
behavior in Study 4 with a student sample, makes it
important for future research to test delegation be-
haviorally across men and women in organizations
with varied work experience and across hierarchical
levels within organizations. Given that we did ob-
serve differential associations with delegation across
men and women in Studies 1 and 3, which were both
studies of managers, it is possible that these associ-
ations only affect behavior among younger and less
experienced managers.
Additionally, delegation is typically predicated on
having an ongoing relationship with one’s subordinate.
Given that our participants in Study 4 had no history
with one another, it is possible that any delegation be-
havior we observed might have been suppressed or
exaggerated relative to the behavioral responses that
would occur in actual work situations (Heilman &
Okimoto, 2008). We tried to address this concern by
including a “fast friends”manipulation (shown to cre-
ate close, intimate, and continued friendships inside
and outside the laboratory [Page-Gould, Mendoza-
Denton, & Tropp, 2008]). Although this manipulation
did affect how close participants felt to their sub-
ordinates, it did not affect how much or whether par-
ticipants delegated, suggesting that these results may
extend to close relationships outside of the laboratory.
Future research could benefit from exploring this phe-
nomenon longitudinally, inducing familiarity and
observing delegation behavior over time. More
importantly, delegation research in organizations
should examine the degree to which familiarity and the
length of the relationship may influence the likelihood
of and comfort with delegation among women.
We did not find conclusive evidence that anxiety
(i.e., negative affect) mediated the effects of gender
on delegation or delegation behavior. One reason
why we did not find the expected mediation may be
driven by the way in which we measured the emo-
tions associated with delegation in Study 4. In
Studies 2 and 3, participants were asked to list any
thoughts, feelings, and emotions they associated
with delegation in an open-ended format. This
measure captures associated anxiety with delega-
tion more broadly, whereas in Study 4 participants
completed a PANAS, which instead captures felt or
prospective anxiety. That is, anxiety in the case of
Study 4 more accurately reflects the anxiety around
prospective delegation in the moment by those who
do indeed delegate (as shown by the greater anxiety
of women who chose to delegate versus those who
chose not to delegate), and does not necessarily
prevent women from delegating. Instead, we would
expect that general anxiety associated with delega-
tion might prevent women from delegating. From
these results, it is clear that distinguishing between
these different measures of affect and examining the
effects of associated, prospective, and even retro-
spective anxiety on the decision to delegate and on
delegation behavior is an important direction for
future research. Additionally, future research should
test a broader range of negative affective associa-
tions, including anticipatory guilt, to gain greater
insight into the mediating mechanisms underlying
gender differences in delegation.
Finally, we operationalized delegation, particularly
in Studies 2 and 4, in a more agentic manner by placing
emphasis on time scarcity, as both the employee and
subordinate were told they had a very busy week ahead
and many tasks to complete. Although we chose this
manipulation because managers who are busy and
have many tasks to complete have been shown to del-
egate (Yukl & Fu, 1999), this setup may have minimized
the more communal aspects of delegation in partici-
pants’minds. Our results in Study 5 highlight the ben-
efits of framing delegation as communal for women.
Future studies should not only operationalize delega-
tion in a manner that emphasizes its communal nature,
but also explore different manipulations intended to
change how women perceive delegation. It is possible
that a noncommunal, and less prosocial, framing of
delegation could also attenuate its negative effects.
Women might respond well to a manipulation that
emphasizes the importance of delegation for getting
their own work done and for signaling who is in charge.
This framing could also legitimize women behaving
agentically, giving them confidence to do so, and thus
rescripting gender role expectations. Drawing from
negotiation research demonstrating that when situa-
tional ambiguity is reduced, women are more likely to
follow nongender-specific scripts and achieve better
negotiation outcomes (Bowles et al., 2005), we would
expect that emphasizing the communal aspects of any
gender-role-paradoxical leadership behavior, or sanc-
tioning enactment of the agentic aspects, would reduce
ambiguity about whether it is appropriate to engage in
the behavior, constraining the potential for gender to
influence its enactment.
CONCLUSIONS
Delegation is a critical managerial skill that can
enhance performance and allow leaders to effectively
2018 1487Akinola, Martin, and Phillips

manage their time. Our findings suggest that if women
have more negative associations with the gender-role-
paradoxical behavior of delegating, they may delegate
less than men, a choice thatcan hinder their ability to
complete tasks effectively and engage in adaptive and
symbiotic relationships with their subordinates. Be-
cause gender roles are evolving (Diekman & Eagly,
2000), it is possible that, over time, the gender-role
incongruence evoked through delegation may be-
come less pronounced. However, our research im-
plies that a more promising avenue for increasing
women’s propensity to delegate may be to reframe
their associations with delegation, such that they
perceive it as a more communal and selfless behavior
rather than as an agentic and self-serving one.
REFERENCES
Abele, A. E., & Wojciszke, B. 2014. Communal and agentic
content in social cognition: A dual perspective model.
Advances in Experimental Social Psychology, 50:
195–255.
Akinola,M.,&Mendes,W.B.2013.It’s good to be the king:
Neurobiological benefits of higher social standing. So-
cial Psychological & Personality Science,5:43–51.
Amanatullah, E. T., & Morris, M. W. 2010. Negotiating
gender roles: Gender differences in assertive negoti-
ating are mediated by women’s fear of backlash and
attenuated when negotiating on behalf of others.
Journal of Personality and Social Psychology, 98:
256–267.
Apfelbaum, E. P., Sommers, S. R., & Norton, M. I. 2008.
Seeing race and seeming racist? Evaluating strategic
colorblindness in social interaction. Journal of Per-
sonality and Social Psychology, 95: 918–932.
Arnold, J. A., Arad, S., Rhoades, J. A., & Drasgow, F. 2000.
The empowering leadership questionnaire: The con-
struction and validation of a new scale for measuring
leader behaviors. Journal of Organizational Behav-
ior, 21: 249–269.
Babcock, L., & Laschever, S. 2003. Women don’t ask:
Negotiation and the gender divide. New York, NY:
Princeton University Press.
Baumeister, R. F., Vohs, K. D., & Funder, D. C. 2007. Psy-
chology as the science of self-reports and finger
movements: Whatever happened to actual behavior?
Perspectives on Psychological Science, 2: 396–403.
Becker, G. S., & Murphy, K. M. 1992. The division of labor,
coordination costs, and knowledge. Quarterly Jour-
nal of Economics, 107: 1137–1160.
Bem, S. L. 1981. Gender schema theory: A cognitive ac-
count of sex typing. Psychological Review, 88: 354–
364.
Bem, S. L. 1983. Gender schema theory and its implications
for child development: Raising gender-aschematic chil-
dren in a gender-schematic society. Signs,8:598–616.
Benetti-McQuoid, J., & Bursik, K. 2005. Individual differ-
ences in experiences of and responses to guilt and
shame: Examining the lenses of gender and gender
role. Sex Roles, 53: 133–142.
Bowles, H. R., Babcock, L., & Lai, L. 2007. Social incentives
for gender differences in the propensity to initiate
negotiations: Sometimes it does hurt to ask. Organi-
zational Behavior and Human Decision Processes,
103: 84–103.
Bowles, H. R., Babcock, L., & McGinn, K. L. 2005. Con-
straints and triggers: Situational mechanics of gender
in negotiation. Journal of Personality and Social
Psychology, 89: 951–956.
Brescoll, V. L. 2012. Who takes the floor and why: Gender,
power, and volubility in organizations. Administra-
tive Science Quarterly, 56: 622–641.
Bruckm ¨
uller, S., & Abele, E. A. 2013. The density of the big
two: How are agency and communion structurally
represented? Social Psychology, 44: 63–74.
Bryant, A., & Check, E. 2000. How parents raise boys and
girls. Newsweek: Your child, 136: 64–65.
Carli,L.L.,LaFleur,S.J.,&Loeber,C.C.1995.Inter-
personal relations and group processes: Nonverbal
behavior, gender, and influence. Journal of Person-
ality and Social Psychology, 68: 1030–1041.
Chen, Z. X., & Aryee, S. 2007. Delegation and employee
work outcomes: An examination of the cultural con-
text of mediating processes in China. Academy of
Management Journal, 50: 226–238.
Crites, S. L., Fabrigar, L. R., & Petty, R. E. 1994. Measuring
the affective and cognitive properties of attitudes:
Conceptual and methodological issues. Personality
and Social Psychology Bulletin, 20: 619–634.
Cuddy, A. J. C., Fiske, S. T., & Glick, P. 2004. When pro-
fessionals become mothers, warmth doesn’t cut the
ice. The Journal of Social Issues, 60: 701–718.
De Pater, I. E., Van Vianen, A. E., & Bechtoldt, M. N. 2010.
Gender differences in job challenge: A matter of task al-
location. Gender, Work and Organization, 17: 433–453.
Diekman, A. B., & Eagly, A. H. 2000. Stereotypes as dy-
namic constructs: Women and men of the past, pres-
ent, and future. Personality and Social Psychology
Bulletin, 26: 1171–1188.
Eagly, A. H. 1987. Sex differences in social behavior: A
social-role interpretation. Hillsdale, NJ: Lawrence
Erlbaum Associates Inc.
Eagly, A. H. 2007. Female leadership advantage and dis-
advantage: Resolving the contradictions. Psychology
of Women Quarterly, 31: 1–12.
1488 AugustAcademy of Management Journal

Eagly, A. H., & Karau, S. J. 2002. Role congruity theory of
prejudice toward female leaders. Psychological Re-
view, 109: 573–598.
Eysenck, M. W. 1992. Anxiety: The cognitive perspective.
Hillsdale, NJ: Erlbaum.
Goff, K., & Torrance, E. P. 2002. Abbreviated Torrance test for
adults. Bensenville, IL: Scholastic Testing Service, Inc.
Gosling, S. D., Rentfrow, P. J., & Swann, W. B. 2003. A very
brief measure of the big-five personality domains.
Journal of Research in Personality, 37: 504–528.
Hayes, A. F. 2013. Introduction to mediation, moderation,
and conditional process analysis: A regression-
based approach. New York, NY: Guilford Press.
Heilman, M. E. 2001. Description and prescription: How gen-
der stereotypes prevent women’s ascent up the organiza-
tional ladder. The Journal of Social Issues, 57: 657–674.
Heilman, M. E., & Chen, J. J. 2005. Same behavior, different
consequences: Reactions to men’s and women’s al-
truistic citizenship behavior. The Journal of Applied
Psychology, 90: 431–441.
Heilman, M. E., & Okimoto, T. G. 2007. Why are women
penalized for success at male tasks? The implied
communality deficit. The Journal of Applied Psy-
chology, 92: 81–92.
Heilman, M. E., & Okimoto, T. G. 2008. Motherhood: A
potential source of bias in employment decisions. The
Journal of Applied Psychology, 93: 189–198.
Heilman, M. E., Wallen, A. S., Fuchs, D., & Tamkins, M. M.
2004. Penalties for success: Reactions to women who
succeed at male gender-typed tasks. The Journal of
Applied Psychology, 89: 416–427.
Izard, C. E. 1977. Human emotions. New York, NY:
Plenum.
Kahneman, D., & Tversky, A. 1984. Choices, values, and
frames. The American Psychologist, 39: 341–350.
Karson, M. 2007. Rorschach inkblot test. In N. Salkind
(Ed.), Encyclopedia of measurement and statistics:
849–850. Thousand Oaks, CA: SAGE Publications.
Kindlon, D., & Thompson, M. 2000. Raising Cain: Pro-
tecting the emotional life of boys. New York, NY:
Random House.
Kiviniemi, M. T., & Bevins, R. A. 2008. Role of affective
associations in the planning and habit systems of
decision-making related to addiction. Behavioral and
Brain Sciences, 31: 450–451.
Kiviniemi, M. T., Jandorf, L., & Erwin, D. O. 2014. Dis-
gusted, embarrassed, annoyed: Affective associations
relate to uptake of colonoscopy screening. Annals of
Behavioral Medicine, 48: 112–119.
Kiviniemi, M. T., Voss-Humke, A. M., & Seifert, A. L. 2007.
How do I feel about the behavior? The interplay of
affective associations with behaviors and cognitive
beliefs as influences on physical activity behavior.
Health Psychology, 26: 152–158.
Koenig, A. M., & Eagly, A. H. 2014. Evidence for the social
role theory of stereotype content: Observations of
groups’roles shape stereotypes. Journal of Person-
ality and Social Psychology, 107: 371–392.
Kray, L. J., Galinsky, A. D., & Thompson, L. 2002. Reversing
the gender gap in negotiations: An exploration of ste-
reotype regeneration. Organizational Behavior and
Human Decision Processes, 87: 386–409.
Kugler, K., & Jones, W. H. 1992. On conceptualizing and
assessing guilt. Journal of Social and Clinical Psy-
chology, 62: 318–327.
Laverie, D. A. 1998. Motivations for ongoing participation
in a fitness activity. Leisure Sciences, 20: 277–302.
Leana, C. R. 1986. Predictors and consequences of delega-
tion. Academy of Management Journal,29:754–774.
Leana, C. R. 1987. Power relinquishment versus power
sharing: Theoretical clarification and empirical com-
parison of delegation. The Journal of Applied Psy-
chology, 72: 228–233.
Liberman, V., Samuels, S. M., & Ross, L. 2004. The name of
the game: Predictive power of reputations versus sit-
uational labels in determining prisoner’s dilemma
game moves. Personality and Social Psychology
Bulletin, 30: 1175–1185.
McConnell, A. R., & Leibold, J. M. 2001. Relations among
the Implicit Association Test, discriminatory behav-
ior, and explicit measures of racial attitudes. Journal
of Experimental Social Psychology, 37: 435–442.
McFarland, C., & Miller, D. T. 1994. The framing of relative
performance feedback: Seeing the glass as half empty
or half full. Journal of Personality and Social Psy-
chology, 66: 1061–1073.
Miles, R., Snow, C., Meyer, A., & Coleman, H. 1978. Orga-
nizational strategy, structure, and process. Academy
of Management Review, 3: 546–562.
Moss-Racusin, C. A., Phelan, J. E., & Rudman, L. A. 2010.
When men break the gender rules: Status incongruity
and backlash against modest men. Psychology of Men
& Masculinity, 11: 140–151.
Moss-Racusin, C., & Rudman, L. A. 2010. Disruptions in
women’s self-promotion: The backlash avoidance
model. Psychology of WomenQuarterly, 34: 186–202.
Page-Gould, E., Mendoza-Denton, R., & Tropp, L. R. 2008. With
a little help from my cross-group friend: Reducing in-
tergroup anxiety through cross-group friendship. Journal
of Personality and Social Psychology, 95: 1080–1094.
Pennebaker, J. W., Francis, M. E., & Booth, R. J. 2001. Lin-
guistic inquiry and word count: LIWC 2001. Mah-
wah, NJ: Erlbaum Publishers.
2018 1489Akinola, Martin, and Phillips

Plant, E. A., Devine, P. G., & Brazy, P. C. 2003. The bogus
pipelineand motivations to respondwithout prejudice:
Revisiting the fading and faking of racial prejudice.
Group Processes & Intergroup Relations,6:187–200.
Richard, R., van der Pligt, J., & de Vries, N. 1996. Antici-
pated affect and behavioral choice. Basic and Applied
Social Psychology, 18: 111–129.
Richeson, J. A., & Shelton, J. N. 2003. When prejudice does
not pay: Effects of interracial contact on executive
function. Psychological Science, 14: 287–290.
Rooth, D. 2010. Automatic associations and discrimination
in hiring: Real world evidence. Labour Economics,
17: 523–534.
Rothman, A. J., & Salovey, P. 1997. Shaping perceptions to
motivate healthy behavior: The role of message fram-
ing. Psychological Bulletin, 121: 3–19.
Rudman, L. A. 1998. Self-promotion as a risk factor for
women: The costs and benefits of counter-stereotypical
impression management. Journal of Personality and
Social Psychology, 74: 629–645.
Rudman, L. A., & Glick, P. 2001. Prescriptive gender ste-
reotypes and backlash toward agentic women. The
Journal of Social Issues, 57: 743–762.
Rudman, L. A., & Fairchild, K. 2004. Reactions to coun-
terstereotypic behavior: The role of backlash in cul-
tural stereotype maintenance. Journal of Personality
and Social Psychology, 87: 157–176.
Rudman, L. A., & Phelan, J. E. 2008. Backlash effects for
disconfirming gender stereotypes in organiza-
tions. Research in Organizational Behavior, 28:
61–79.
Simons, J., & Carey, K. B. 1998. A structural analysis of
attitudes toward alcohol and marijuana use. Person-
ality and Social Psychology Bulletin, 24: 727–735.
Small,D.A.,Gelfand,M.,Babcock,L.,&Gettman,H.2007.
Who goes to the bargaining table? The influence of gen-
der and framing on the initiation of negotiation. Journal
of Personality and Social Psychology, 93: 600–613.
Spencer, S. J., Steele, C. M., & Quinn, D. M. 1999. Stereo-
type threat and women’s math performance. Journal
of Experimental Social Psychology, 35: 4–28.
Stephan, W. G., & Stephan, C. W. 1985. Intergroup anxiety.
The Journal of Social Issues, 41: 157–175.
Sy, T., Cote, S., & Saavedra, R. 2005. The contagious
leader: Impact of the leader’s mood on the mood of
group members, group affective tone, and group
processes. The Journal of Applied Psychology, 90:
295–305.
Tabachnick, B. G., & Fidell, L. S. 2007. Using multivariate
statistics. Boston, MA: Pearson/Allyn & Bacon.
Tannen, D. 1990. You just don’t understand: Women and
men in conversation. New York, NY: Morrow.
Trafimow, D., & Sheeran, P. 1998. Some tests of the dis-
tinction between cognitive and affective beliefs. Jour-
nal of Experimental Social Psychology,34:378–397.
Valian, V. 1998. Why so slow? The advancement of
women. Cambridge, U.K.: MIT Press.
Vandello, J. A., Bosson, J. K., Cohen, D., Burnaford, R. M., &
Weaver, J. R. 2008. Precarious manhood. Journal of
Personality and Social Psychology, 95: 1325–1339.
Watson, D., & Clark, L. A. 1994. The PANAS-X: Manual
for the positive and negative affect schedule—
Expanded Form. Iowa City, IA: University of Iowa.
Weinstein, D. D. 1993. Testing four competing theories of
health-protective behaviors. Health Psychology, 12:
324–333.
Weisner, T. S., & Wilson-Mitchell, J. E. 1990. Non-
conventional family life-styles and sex typing in six-
year-olds. Child Development, 61: 1915–1933.
Yukl, G., & Fu, P. P. 1999. Determinants of delegation and
consultation by managers. Journal of Organizational
Behavior, 20: 219–232.
Zou, X., Tam, K. P., Morris, M. W., Lee, S. L., Lau, I. Y. M., &
Chiu, C. Y. 2009. Culture as common sense: Perceived
consensus versus personal beliefs as mechanisms of
cultural influence. Journal of Personality and Social
Psychology, 97: 579–597.
Modupe Akinola (makinola@columbia.edu) is the Sanford
C. Bernstein & Co. Associate Professor of Leadership and
Ethics at Columbia Business School. She received her PhD
in organizational behavior from Harvard University. Her
research explores how stress affects workplace perfor-
mance. She also examines workforce diversity.
Ashley Martin (ashley.martin@stanford.edu) is an assis-
tant professor at the Stanford Graduate School of Business.
She received her PhD in management from Columbia
Business School. Her research focuses on the challenges
and benefits associated with gender and racial diversity in
organizations.
Katherine W. Phillips (kwphillips@columbia.edu) is the
Reuben Mark Professor of Organizational Character at Co-
lumbia Business School. She received her PhD in organiza-
tional behavior from Stanford University. Her research
focuses on the impact of diversity on performance, and re-
lationship building across racial boundaries in organizations.
1490 AugustAcademy of Management Journal

APPENDIX A
TABLE A1
Study 4 Tasks
Task Name Description Scoring
1. Influencing Email Write an email to the school to influence students to
donate to the organization. Six relevant points had
to be included in the body of the text.
This task was scored by two different raters (a5.75)
on four dimensions: length, completion, writing
quality, following instructions.
2. Organizing Donations
(Pt. 1)
Copy a list of student names, majors, school years, and
donation amounts into a spreadsheet. Participants
were given a computer and Excel file in which to
enter this information.
This task was scored based on completion (% of list
completed) and accuracy (number of spelling
errors).
3. Organizing Donations
(Pt. 2)
Participants were asked three questions about the
information accumulated from the donation list: (1)
what was the average amount donated? (2) which
major donated the most money? and (3) which year
was most represented in new memberships?
This task required the participant to have completed
the first part of the task accurately and have
calculated these numbers correctly. This metric
was scored from 023 based on how many of these
questions they answered correctly.
4. Proofreading
PowerPoint
Participants were told to proofread an eight-page
PowerPoint presentation. The presentation was
printed on 8”311”paper, with each slide on
a separate page, and contained 38 errors in total.
This task was scored based on how many of the 38
errors the participant found, with scores ranging
from 0238.
5. Formatting Newsletter Participants were given a printed copy of a newsletter,
which contained a variety of fonts, colors, and
sizing. Their goal was to format a document in
Microsoft Word, which included the unformatted
raw materials, to look like the newsletter they were
given.
Performance on this task was scored based on the 34
changes that needed to be made to the document.
Scores ranged from 0232.
6. ATTA (Creativity Task) This task was based on the Abbreviated Torrance Test
for Adults (ATTA; Goff & Torrance, 2002).
Participants were given a sheet with nine triangles
and were told to create creative and novel logos for
the environmental campaign.
This task was scored by two research assistants, blind
to the hypothesis, on (1) overall creativity, (2)
fluency, (3) flexibility, and (4) flexibility. Any
discrepancies were discussed and resolved by the
two research assistants.
2018 1491Akinola, Martin, and Phillips