European Journal of Social Psychology

Eur. J. Soc. Psychol. 35, 633–642 (2005)

Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/ejsp.270

The burden of benevolent sexism: How it contributes

to the maintenance of gender inequalities

MANUELA BARRETO* AND NAOMI ELLEMERS

Leiden University, The Netherlands

Abstract

This study (N ¼ 235) examines the responses of male and female participants to information about the
alleged endorsement of either hostile or benevolent sexist beliefs by a sample of either men or women.
We predicted that people endorsing benevolent sexist statements would be less likely to be perceived as
sexist than those endorsing hostile sexist views, and examined the judgmental process through which
people fail to recognize benevolent sexism as a form of prejudice. We argue that benevolent sexists do
not match the mental prototype of sexist perpetrators, because they are seen as likeable. Our results
confirm that because benevolent sexists are evaluated more positively than hostile sexists, they are less
likely to be seen as sexists. This judgmental process occurs relatively independently of emotional
responses to hostile vs. benevolent sexism. These results are discussed in terms of their relevance to the
maintenance of gender inequalities. Copyright # 2005 John Wiley & Sons, Ltd.

When women are portrayed as inferior to men few would hesitate in finding the source of these
opinions sexist. However, contrary to more traditional conceptualizations of prejudice, researchers
have argued that sexism does not necessarily take such overtly hostile forms (Glick & Fiske, 1996;
Jackman, 1994). Indeed, in modern societies, expressions of sexism are often quite subtle, and can
even be positive in tone (Benokraitis & Feagin, 1995; Swim, Aikin, Hall, & Hunter, 1995), for instance
because they are presented as jokes (Greenwood & Isbell, 2002; LaFrance & Woodzicka, 1998), or as
a form of flattery (Pryor & Whalen, 1996).
In this paper we focus on a specific type of ‘positive’ sexism, namely benevolent sexism, as
compared to hostile sexism (Glick & Fiske, 1996). Although social scientists have established that
benevolent sexism can be harmful for women, so far no research has examined whether it is perceived
as a form of sexism by ordinary men and women (see also Barreto & Ellemers, 2005). Given that
prejudice first needs to be perceived before it can be challenged as an illegitimate cause of social
inequalities (Tajfel & Turner, 1979; see also Jost & Major, 2001, for reviews), examining how men and
women perceive and experience benevolent sexism will significantly advance our understanding of the

*Correspondence to: Manuela Barreto, Social and Organizational Psychology, Leiden University, PO Box 9555, 2300 RB
Leiden, The Netherlands. E-mail: barreto@fsw.leidenuniv.nl
Contract/grant sponsor: Dutch Science Foundation (NWO).

Received 31 November 2004

Copyright # 2005 John Wiley & Sons, Ltd. Accepted 11 February 2005
634 Manuela Barreto and Naomi Ellemers

processes that contribute to the promotion and maintenance of gender inequalities. In the present
research we address this important gap in existing knowledge by directly examining whether
benevolent sexism is perceived as sexist, both by men and by women, and by studying the
psychological process that mediates these perceptions.

WHAT IS BENEVOLENT SEXISM AND WHY IS IT A PROBLEM?

It has long been known that gender stereotypes can be positive in tone (Eagly & Mladinic, 1989).
However, in parallel to what has been documented within research on social identity processes in other
inter-group contexts (Tajfel & Turner, 1979; see e.g. Ellemers, Van Rijswijk, Roefs, & Simons, 1997;
Mummendey & Schreiber, 1984; Spears & Manstead, 1989), women are only seen as superior to men
on dimensions which are either inconsequential for the status relation between men and women, or
that imply their dependence on men (Fiske, Xu, Cuddy, & Glick, 1999; Glick & Fiske, 1996; Jackman,
1994). Following this thought, Glick and Fiske (1996) distinguished ‘hostile’ from ‘benevolent’ forms
of sexism. Hostile sexism communicates a clear antipathy towards women, whereas benevolent
sexism takes the form of seemingly positive but in fact patronizing beliefs about women. Benevolent
sexism encompasses three components: protective paternalism (e.g. the belief that women should be
protected by men), complementary gender differentiation (e.g. the belief that women have—typically
domestic—qualities that few men possess), and heterosexual intimacy (e.g. the belief that women
fulfil men’s romantic needs). It is similar to hostile sexism because it relies on gender stereotypes (e.g.
women are dependent on men), but it clearly differs from hostile sexism in the sense that it conveys
these beliefs in a positive tone.
Research has demonstrated that, despite its positive tone, and despite stemming from a genuine
positive feeling towards women, benevolent sexism clearly has negative implications. For instance, like
hostile sexism, benevolent sexism predicts endorsement of gender stereotypes and of old-fashioned and
modern sexist beliefs (Glick & Fiske, 1996, 2001a). It is related to sexual harassment (Fiske & Glick,
1995; Pryor, Geidd, & Williams, 1995), as well as to negative reactions to rape victims (Abrams, Vicky,
Masser, & Bohner, 2003; Viki & Abrams, 2002). Moreover, although benevolent sexism encompasses
the conviction that women should receive male protection, research has shown that benevolent sexism
is positively associated with attitudes that legitimize domestic violence (Glick, Sakalli-Ugurlu,
Ferreira, & Souza, 2002; Sakalli, 2002). Finally, in international comparisons, the national level of
endorsement of benevolent sexism is related to United Nations indicators of gender inequality such as
the participation of women in the economy and in politics (Glick et al., 2000; Glick & Fiske, 2001a).

HOW BENEVOLENT SEXISM CONTRIBUTES TO THE

MAINTENANCE OF GENDER INEQUALITIES

These findings demonstrate that to successfully address existing gender inequalities it is essential to
combat not only hostile but also benevolent forms of sexism. Although people are generally more
likely to endorse benevolent than hostile sexism (Glick & Fiske, 1996), it has been suggested (Glick &
Fiske, 1996, 2001b; Jackman, 1994) that benevolent sexism might be more difficult to combat than
hostile sexism. However, there is as yet no direct evidence that this is true, or why this might be the
case. To complement previous research in this area, we argue that benevolent sexism contributes to the
maintenance of social inequalities because it passes unnoticed as a form of prejudice. Given that

Copyright # 2005 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 35, 633–642 (2005)
 Benevolent sexism 635

perceptions of prejudice are an important determinant of people’s types of responses to social

inequalities (see Schmitt & Branscombe, 2002, for a review), demonstrating that these perceptions are
impaired by benevolent sexism constitutes an important contribution to the understanding of how
gender inequalities are created and maintained. Thus, our first prediction is that compared to hostile
sexists, those who endorse benevolent sexism are less likely to be seen as holding sexist views
(Hypothesis 1).
We examine how people perceive benevolent and hostile sexism when it is endorsed by a group of
others, rather than when it is voiced by a single individual (see also Barreto & Ellemers, 2005). Past
research examining the circumstances under which people perceive subtle expressions of discrimina-
tion as a form of prejudice has mainly focused on the attributions people make regarding negative
treatment received by a single individual (see Crocker, Major, & Steele, 1998, for a review). However,
it has been argued that victims can see these perpetrators as exceptions to the rule, and hence discount
the prejudice they express (Abelson, Dasgupta, Park, & Banaji, 1998), whereas prejudice which is
seen as more pervasive has been found to have a greater impact on its victims (Schmitt, Branscombe,
& Postmes, 2003). We therefore chose to employ this procedure when examining our hypotheses.
Additionally, we examine the psychological process responsible for the failure to recognize
benevolent sexism for what it is. We propose that people who express benevolent sexist views do
not match the mental prototype of sexists. Seeing the source of a statement as sexist involves
the comparison between the observed exemplar (i.e. the source of sexism) and mental prototypes of
sexist perpetrators (Baron, Burgess, & Kao, 1991; Feldman-Barrett & Swim, 1998; Inman & Baron,
1996). If the source fits the mental prototype it is seen as sexist, but if it does not this characterization is
hindered.
Prior research found that people evaluate a benevolent sexist perpetrator more positively than a
hostile sexist (Killianski & Rudman, 1998). We argue that this positive evaluation is inconsistent with
the prototype of sexist perpetrators, who are seen as unlikeable people, typically expressing antipathy
towards women (e.g. Dion, 1975; Locksley, Borgida, Brekke, & Hepburn, 1980). Thus, our second
prediction is that the evaluation of the source that expresses hostile or benevolent sexism mediates the
extent to which this source is perceived as sexist (Hypothesis 2). Importantly, the empirical connection
between the evaluation of a source and the recognition of the views endorsed by this source as
prejudicial has not been made in previous research.
To further support the validity of our reasoning that evaluative judgments of the source affect the
perception of sexism, we aim to exclude the alternative possibility that emotional responses to sexist
statements cause people to designate the source as sexist. Exposure to overt discrimination induces
more anger than subtle expressions of prejudice (e.g. Barreto & Ellemers, 2005). Hence, it could be
argued that the greater experience of anger induced by exposure to hostile sexism (as opposed to the
flattery implied by benevolent sexism, Glick & Fiske, 1995) causes people to perceive a source of
hostile sexism as more sexist than a benevolent sexist. Although we acknowledge that exposure to
hostile sexism is likely to evoke more anger than benevolent sexism, we predict that these emotional
responses occur relatively independently of people’s evaluative judgments of the source. That is, if our
reasoning in terms of prototypes is correct, the experience of anger should not mediate the perception
of the source as sexist.
An additional aspect of the prototype of sexist events that should be taken into account in this
analysis is the gender of the perpetrator. Prior research found that people are more likely to perceive a
sexist statement as sexist when it is voiced by a male perpetrator (Baron et al., 1991). It is unclear
whether this is also true in the case of benevolent sexism, since it can be interpreted as ‘gentlemanly’
behaviour when endorsed by a man (e.g. Glick & Fiske, 2001a). Therefore, it is important to examine
whether the gender of the source of a benevolent sexist statement affects how the source is evaluated,
and whether it is seen as sexist.

Copyright # 2005 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 35, 633–642 (2005)
636 Manuela Barreto and Naomi Ellemers

GENDER DIFFERENCES IN REACTIONS TO BENEVOLENT SEXISM

Prior research suggests that benevolent sexism may be perceived differently by men and by women.
Specifically, in a study by Greenwood and Isbell (2002), benevolent sexism was associated with
enjoyment of sexist jokes among men, but not among women. Although the authors explained this
finding by suggesting that benevolent sexism is likely to mean different things to men and to women,
no direct evidence has been provided so far as to whether or not this is the case. Some evidence does
suggest that this may be the case: research examining endorsement of hostile and benevolent sexist
attitudes shows that men tend to agree more with hostile sexist statements than women do (Glick &
Fiske, 2001a, for a review). This is in line with the blatant threat that hostile sexism represents to
women, and the fact that as a form of intergroup bias it is more likely to be held by men. However,
while some samples also reveal differential agreement with benevolent sexist statements, others fail to
show this gender difference (see Glick & Fiske, 2001a, for a review). Because of these contradictory
findings, in this study we will directly examine whether men and women differ in their perceptions of
benevolent sexism as prejudicial.

METHOD

Design and Participants

The study consisted of a 2 (Type of sexism: Hostile sexism, Benevolent sexism) X 2 (Gender
of participant: Male, Female) X 2 (Gender of the source: Male, Female) between-participants design.
A total of 79 male and 156 female students at a Dutch University voluntarily took part in this study and
were randomly assigned to the experimental conditions.

Procedure

The majority of participants completed the questionnaire at the start of one class, taught by another
member of staff at the University where the authors are affiliated. These participants were debriefed
and provided with the results of the study in a class that took place 2 weeks later. The remaining
participants (approximately 50) completed the questionnaire in our laboratory, and were debriefed by
the experimenter.

Manipulations

Participants read a short description purportedly summarizing the results of a prior study concerning
opinions about the position of women in Dutch society. Gender of source was manipulated by varying
how the sample of the prior study was described. Half of the participants were told that the study
examined the opinions of a sample of women (Female Source), and the remaining participants were
told that the previous study had investigated the opinions of a sample of men (Male Source). It is
important to note that we did not describe this sample as representative of the opinions of all men or all
women.

Copyright # 2005 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 35, 633–642 (2005)
 Benevolent sexism 637

Subsequently, the ostensible results of this study were summarized, in which the Type of sexism
was manipulated. In the Hostile condition participants read that the women/men in the sample agreed
that women are too easily offended and interpret innocent remarks quickly as sexist, that women seek
to gain power by getting control over men, that women exaggerate problems they get at work, that
when women lose to men in a fair competition they typically complain of being discriminated, and that
most women do not appreciate fully what men do for them. These opinions were modelled after the
items of Glick and Fiske’s (1996) scale of Hostile sexism. In the Benevolent sexism condition,
participants read that the women/men in the sample agreed with opinions that correspond to each of
the three components in the Benevolent sexism scale developed by Glick and Fiske (1996): that many
women have a quality of purity that few men possess, and that women—compared to men—tend to
have superior moral sensibility (complementary gender differentiation), that no matter how accom-
plished he is, a man is not truly complete as a person unless he has the love of a woman (heterosexual
intimacy), that women should be cherished and protected by men, and that men should be willing to
sacrifice themselves in order to provide financially for the women in their lives (protective
paternalism).

Dependent Measures

Subsequently, participants were asked to indicate their reactions to these opinions on 7 point rating
scales ranging from (1) ‘not at all’ to (7) ‘very much’. To assess participants’ evaluation of the source
of the sexist statements, participants indicated the extent to which they expected that they would like
people with these (sexist) views, and would be willing to collaborate with people who endorse such
views (r ¼ 0.55, p < 0.001). Perceived sexism was assessed by asking to what extent participants
thought that people who held those opinions were prejudiced against women. Participants indicated to
what extent they experienced anger by stating to what extent they experienced each of five negative
emotions after reading the (sexist) message: angry, indignant, irritated, disappointed, and frustrated
(alpha ¼ 0.83).

RESULTS

Analyses of Variance

All variables were entered in one 2 Type of sexism (Hostile, Benevolent sexism) X 2 Gender of
participant (Male, Female) X 2 Gender of the source (Male, Female) between-participants multivariate
analysis of variance (MANOVA). This analysis revealed a significant multivariate main effect of Type
of sexism, F(3, 222) ¼ 9.16, p < 0.001 as well as a significant interaction between Type of sexism and
Gender of participant, F(3, 222) ¼ 2.91, p < 0.05. At the univariate level, the main effect of Type of
sexism was significant for all dependent variables: evaluation of source, F(1, 224) ¼ 25.56, p < 0.001,
perceived sexism, F(1, 224) ¼ 5.35, p < 0.05, and anger, F(1, 224) ¼ 4.21, p < 0.05. Participants
evaluated the Hostile sexist source less positively (M ¼ 3.29, SD ¼ 1.04) than the source of a
Benevolent sexist message (M ¼ 4.01, SD ¼ 1.04). They saw the Hostile sexist source as more
prejudiced (M ¼ 4.88, SD ¼ 1.43) than the Benevolent sexist source (M ¼ 4.44, SD ¼ 1.35). Finally,
participants were more angry at the Hostile sexist source (M ¼ 3.04, SD ¼ 1.35) than at the Benevolent
sexist source (M ¼ 2.49, SD ¼ 1.29). These results corroborate Hypothesis 1, as they indicate that a
source that expresses benevolent sexism is less likely to be seen as sexist and accordingly elicits a less
negative evaluation and less anger than a source of hostile sexism.

Copyright # 2005 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 35, 633–642 (2005)
638 Manuela Barreto and Naomi Ellemers

Table 1 Means and standard deviations of evaluation of source, perceived sexism, and anger as a function of
Type of sexism and Gender of participant
Gender of participant

Male Female

Type of sexism Hostile Benevolent Hostile Benevolent

Evaluation of the source 3.41b 4.24a 3.24b 3.89a

(0.91) (1.07) (1.11) (1.01)
Perceived sexism 4.89a 4.41b 4.88a 4.45b
(1.47) (1.36) (1.41) (1.35)
Anger 2.61b 2.72b 3.25a 2.39b
(1.25) (1.39) (1.35) (1.25)
Note: Standard deviations are presented within parentheses. Means with different subscripts differ significantly at p < 0.05.
Comparisons are made within rows.

The interaction between Type of sexism and Participant’s gender was significant only for anger,
F(1, 226) ¼ 7.28, p < 0.01 (see Table 1). Women felt more angry when facing hostile sexism than
when facing benevolent sexism, F(1, 231) ¼ 17.05, p < 0.001, while men reported similar levels of
anger in both conditions, F(1, 231) ¼ 0.10, ns. As a result, women clearly felt more angry than men
when facing hostile sexism F(1, 231) ¼ 9.26, p < 0.005, whereas in the benevolent condition, there
was a marginally significant gender difference in the opposite direction F(1, 231) ¼ 3.32, p ¼ 0.07,
indicating that women tended to feel less angry than men when facing benevolent sexism. This latter
effect is likely caused by some degree of flattery women experienced as a result of benevolent sexism
(Glick & Fiske, 1996).

Mediation Analyses

The mediation of the effect of Type of sexism on perceived sexism by evaluation of source (Hypothesis
2) was tested with the regression approach. Following Baron and Kenny (1986), we tested whether 1)
the manipulation of Type of sexism reliably affects perceived sexism (the dependent variable), 2) the
manipulation of Type of sexism reliably affects evaluation of the source (the mediator), and 3) when
evaluation of the source is controlled for, the effect of Type of sexism on perceived sexism decreases
reliably. Type of sexism reliably predicted perceived sexism (beta ¼ 0.14), F(1, 232) ¼ 4.5, p < 0.05,
and type of sexism reliably predicted evaluation of the source (beta ¼
0.33), F(1, 232) ¼ 27.78,
p < 0.001. Evaluation of the source also predicted perceived sexism (beta ¼
0.29, p < 0.001),
F(1, 231) ¼ 21.26, p < 0.001. When Type of sexism and evaluation of the source were entered
simultaneously as predictors, the effect of Type of sexism on perceived sexism was no longer reliable
(beta ¼ 0.06, ns), while the effect of evaluation of source on perceived sexism remained significant
(beta ¼
0.27, p < 0.001), F(2, 230) ¼ 10.99, p < 0.001. The reduction in the effect of Type of sexism
after accounting for mediation (Baron & Kenny, 1986) is significant (Sobel test: z ¼ 2.25, p < 0.05).
Moreover, following the method proposed by MacKinnon and Dwyer (1993) we found that evaluation
of source mediates 63% of the total effect of Type of sexism on perceived sexism, constituting a ratio
of indirect effect to direct effect of 1.7.
To exclude alternative possibilities, we first tested the inverted model (i.e. whether perceived
sexism mediated the effect of Type of sexism on evaluation of the source). This turned out not to be the

Copyright # 2005 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 35, 633–642 (2005)
 Benevolent sexism 639

case. When Type of sexism and perceived sexism were simultaneously entered as predictors both Type
of sexism (beta ¼
0.29, p < 0.001) and perceived sexism (beta ¼
0.25, p < 0.001) remained
negatively and reliably associated with evaluation of the source, F(2, 230) ¼ 23.01, p < 0.001. Next,
we tested whether, instead of evaluation of the source, anger could function as a mediator of the effect
of Type of sexism on perceived sexism. Although Type of sexism significantly predicted reported
anger (beta ¼ 0.20, p < 0.005), and the effect of Type of sexism is slightly reduced (beta ¼ 0.11,
p ¼ 0.08) when anger is also included as a predictor of perceived sexism (beta ¼ 0.19, p < 0.01), the
Sobel test for this mediation is not significant (z ¼ 1.09, p ¼ 0.28), and the percentage of mediation is
considerably smaller (27% of the total effect or 0.37 ratio) than when evaluation of source is taken as a
mediator of this effect. These analyses further support our reasoning, and confirm Hypothesis 2 that it
is the negative evaluation of the hostile sexist source (and not the emotional response people
experience) that contributes to perceiving this source as sexist, while this perception is hindered by
a positive evaluation of the benevolent sexist source.

DISCUSSION

This study sheds light on a psychological process that contributes to the maintenance of gender
inequalities in modern societies. Our results are the first to demonstrate that when people express
benevolent sexism, they are less likely to be recognized as holding sexist views than when expressing
hostile sexism. That is, this study revealed that both men and women perceived people endorsing
hostile sexist views as significantly more sexist than those who voiced benevolent sexism. Addition-
ally, our results extend previous research as they clarify the process through which this happens. Our
mediational analyses clearly showed that benevolent sexism is not recognized as sexist because its
source is relatively positively evaluated and therefore deviates from the prototype of a sexist
perpetrator (see also Baron et al., 1991; Dion, 1975; Locksley et al., 1980). In sum, whereas prior
research demonstrated that endorsement of benevolent sexism is associated with a range of negative
attitudes towards women, the present results illustrate another way in which benevolent sexism might
have adverse consequences, namely because it tends not to be recognized as sexism by those who are
exposed to it and therefore is likely to remain unchallenged.
Our results also point to the pertinence of examining not only the responses of those more directly
targeted by sexism (women), but also to include the perspective of men, who tend to be in the position
to either sanction the occurrence of sexism or to offer social support to those who challenge it.
Although women (who are more directly targeted by sexist views) experienced more anger than did
men when exposed to hostile sexism, it is important to stress that—as predicted—this did not translate
into gender differences in the core judgmental process (relating evaluation of the source to perception
of sexism). Thus, our data indicate that this core judgmental process takes place in similar ways for
men and for women, and is relatively independent of affective reactions (which were mainly elicited
among women). Although there may be some degree of variability in how people experience this
process, and a sample with different characteristics (e.g. people with more conservative ideologies)
may be more likely to accept hostile as well as benevolent sexist views, there is no reason to assume
that this will uncover gender differences which were not found in the present study.
It is important to stress that, unlike prior research, we did not subject our participants to negative
treatment that affected them directly (see Crocker et al., 1998, for a review). One important
consequence of this feature of our study is that it provides some indications about the likelihood
that the actual victims of prejudice may expect to receive social support from others when they recount
their experiences or challenge those who hold sexist views (see Stangor, et al., 2003 for a review).

Copyright # 2005 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 35, 633–642 (2005)
640 Manuela Barreto and Naomi Ellemers

Given that social support is often a pre-requisite for protest against illegitimate treatment, these
findings suggest still another way through which benevolent sexism may inhibit restorative action and
thereby promote gender inequalities.
Some limitations of this research should also be noted. Whereas previous research focusing on
other forms of sexism revealed that the gender of the sexist perpetrator tends to affect the perception of
sexism (e.g. Baron et al., 1991), in our examination of hostile vs. benevolent sexism this was not the
case. This is an intriguing finding, and future research should further investigate the process that
causes this effect. In addition, although our results indicate that women express less anger in response
to benevolent rather than hostile sexist views, this does not imply that it is less harmful for them to be
subjected to benevolent rather than hostile sexism. In fact, the evidence we reviewed to show that
benevolent sexism is related to gender inequality suggests that this is not the case. Therefore, future
research should examine whether exposure to benevolent sexism might have other negative affective
or behavioural consequences not examined here.
Future research should also examine the relationship between benevolent sexism and other
phenomena that contribute to the creation and promotion of gender differences in less obvious
ways. Some examples are the queen bee phenomenon (Ellemers, van den Heuvel, de Gilder, Maass, &
Bonvini, 2004), and the glass cliff phenomenon (Ryan & Haslam, in press). These phenomena are all
based on the belief that current gender differences are legitimate in the sense that they reflect objective
differences between men and women rather than being the result of a system in which men and women
are not given the same opportunities. Exposure to, or the endorsement of, benevolent sexism
contributes to such beliefs. Additionally, it supports the notion that women are better suited than
men to deal with crises and conflictual situations, which causes women to be assigned to more difficult
leadership situations than men, and sets them up for failure (the glass cliff phenomena). Likewise,
benevolent sexism supports the idea that women as a group are better suited for domestic and nurturing
tasks than for professional achievement, so that women who are successful in their career are inclined
to see themselves (and are seen by others) as different from other women (the queen bee effect). In this
sense, the failure to recognize benevolent sexism as a form of prejudice might be seen as a central
mechanism that lies at the heart of a number of phenomena through which seemingly positive beliefs
about women indirectly induce and sustain gender discrimination.
In sum, although benevolent sexist expressions appear to be positive, they are far from benign (see
also Abrams et al., 2003; Glick et al., 2002). The present research shows that in some sense,
benevolent sexists are more (rather than less) harmful than hostile sexists, since by promoting
acceptance of prejudicial attitudes they contribute to the perpetuation of gender inequalities (Glick &
Fiske, 2001a; Jackman, 1994).

AUTHORS’ NOTE

This research was financed by a grant from the Dutch Science Foundation (NWO) awarded to the first
author.

ACKNOWLEDGEMENTS

We wish to thank Ken Dion, Alex Haslam, Michelle Ryan and four anonymous reviewers for their
valuable comments on a prior version of this manuscript.

Copyright # 2005 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 35, 633–642 (2005)
 Benevolent sexism 641

REFERENCES

Abelson, R. P., Dasgupta, N., Park, J., & Banaji, M. R. (1998). Perceptions of the collective other. Personality and
Social Psychology Review, 2, 243–250.
Abrams, D., Vicky, G. T., Masser, B., & Bohner, G. (2003). Perceptions of stranger and acquaintance rape: The
role of benevolent and hostile sexism in victim blame and rape proclivity. Journal of Personality and Social
Psychology, 84, 111–125.
Baron, R. H., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological
research: Conceptual, strategic and statistic considerations. Journal of Personality and Social Psychology, 51,
1173–1182.
Baron, R. S., Burgess, M. L., & Kao, C. F. (1991). Detecting and labeling prejudice: Do female perpetrators go
undetected? Personality and Social Psychology Bulletin, 17, 115–123.
Barreto, M., & Ellemers, N. (2005). The perils of political correctness: Responses of men and women to old-
fashioned and modern sexist views. Social Psychology Quarterly, 68, 75–88.
Benokraitis, N. V., & Feagin, J. R. (1995). Modern sexism. Englewood Cliffs: Prentice Hall.
Crocker, J., Major, B., & Steele, C. (1998). Social stigma. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The
handbook of social psychology (pp. 504–553). Boston: McGraw-Hill.
Dion, K. L. (1975). Women’s reactions to discrimination from members of the same or opposite sex. Journal of
Research in Personality, 9, 294–306.
Eagly, A. H., & Mladinic, A. (1989). Gender stereotypes and attitudes toward women and men. Personality and
Social Psychology Bulletin, 15, 543–558.
Ellemers, N., Van Rijswijk, W., Roefs, M., & Simons, C. (1997). Bias in intergroup perceptions: Balancing group
identity with social reality. Personality and Social Psychology Bulletin, 23, 186–198.
Ellemers, N., Van den Heuvel, H., De Gilder, D., Maass, A., & Bonvini, A. (2004). The underrepresentation of
women in science: Differential commitment or the Queen-bee syndrome? British Journal of Social Psychology,
43, 315–338.
Feldman-Barrett, L., & Swim, J. K. (1998). Appraisals of prejudice and discrimination. In J. K. Swim, &
C. Stangor (Eds.), Prejudice: The target’s perspective (pp. 12–37). San Diego: Academic Press.
Fiske, S. T., & Glick, P. (1995). Ambivalence and stereotypes cause sexual harassment: A theory with implications
for organizational change. Journal of Social Issues, 51, 97–115.
Fiske, S. T., Xu, J., Cuddy, A. C., & Glick, P. (1999). (Dis)respecting versus (dis)liking: Status and interdependence
predict ambivalent stereotypes of competence and warmth. Journal of Social Issues, 55, 473–489.
Glick, P., & Fiske, S. T. (1996). The ambivalent sexism inventory: Differentiating hostile and benevolent sexism.
Journal of Personality and Social Psychology, 70, 491–512.
Glick, P., & Fiske, S. T. (2001a). Ambivalent sexism. In M. P. Zanna (Ed.), Advances in experimental social
psychology. San Diego: Academic Press.
Glick, P., & Fiske, S. T. (2001b). An ambivalent alliance: Hostile and benevolent sexism as complementary
justifications for gender inequality. American Psychologist, 56, 109–118.
Glick, P., Fiske, S. T., Mladinic, A., Sainz, J., Abrams. D., et al. (2000). Beyond prejudice as simple antipathy:
Hostile and benevolent sexism across cultures. Journal of Personality and Social Psychology, 79, 763–775.
Glick, P., Sakalli-Ugurlu, N., Ferreira, M. C., Souza, M. A. (2002). Ambivalent sexism and attitudes toward wife
abuse in Turkey and Brazil. Psychology of Women Quarterly, 26, 292–297.
Greenwood, D., & Isbell, L. M. (2002). Ambivalent sexism and the dumb blonde: Men’s and women’s reactions to
sexist jokes. Psychology of Women Quarterly, 26, 341–350.
Inman, M. L., & Baron, R. S. (1996). Influence of prototypes on perceptions of prejudice. Journal of Personality
and Social Psychology, 70, 727–739.
Jackman, M. R. (1994). The velvet glove: Paternalism and conflict in gender, class, and race relations. Berkeley:
University of California Press.
Jost, J. T., & Major, B. (Eds.). (2001). The psychology of legitimacy: Emerging perspectives on ideology, justice,
and intergroup relations. Cambridge: Cambridge University Press.
Killianski, S. E., & Rudman, L. A. (1998).Wanting it both ways: Do women approve of benevolent sexism? Sex
Roles, 39, 333–352.
LaFrance, M., & Woodzicka, J. A. (1998). No laughing matter: Women’s verbal and nonverbal reactions to sexist
humor. In. J. Swim, & C. Stangor (Eds.), Prejudice: The target’s perspective (62–79). San Diego: Academic
Press.

Copyright # 2005 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 35, 633–642 (2005)
642 Manuela Barreto and Naomi Ellemers

Locksley, A., Borgida, E., Brekke, N., & Hepburn, C. (1980). Sex stereotypes and social judgment. Journal of
Personality and Social Psychology, 39, 821–831.
MacKinnon, D. P., & Dwyer, J. H. (1993). Estimating mediated effects in prevention studies. Evaluation Review,
17, 144–158.
Mummendey, A., & Schreiber, H. J. (1984). ‘Different’ just means ‘Better’: Some obvious and some hidden
pathways to in-group favouritism. British Journal of Social Psychology, 23, 363–368.
Pryor, J. B., & Whalen, N. J. (1996). A typology of sexual harassment: Characteristics of harassers and social
circumstances under which sexual harassment occurs. In W. O’Donohue (Ed.), Sexual harassment: Theory,
research, and treatment. Needham Heights: Allyn, & Bacon.
Pryor, J. B., Geidd, J. L., & Williams, K. B. (1995). A social-psychological model for predicting sexual
harassment. Journal of Social Issues, 51, 69–84.
Ryan, M. E., & Haslam, S. A. (In press). The glass cliff: Evidence that women are over-represented in precarious
leadership positions. British Journal of Management.
Sakalli, N. (2002). Beliefs about wife beating among Turkish college students: The effects of patriarchy, sexism,
and sex differences. Sex Roles, 44, 599–610.
Schmitt, M. T., & Branscombe, N. R. (2002). The meaning and consequences of perceived discrimination in
disadvantaged and privileged social groups. European Review of Social Psychology, 12, 167–199.
Schmitt, M. T., Branscombe, N. R., & Postmes, T. (2003). Women’s emotional responses to the pervasiveness of
gender discrimination. European Journal of Social Psychology, 33, 297–312. DOI:10.1002/ejsp.147
Spears, R., & Manstead, A. S. R. (1989). The Social Context of Stereotyping and Differentiation. European
Journal of Social Psychology, 19, 101–121.
Stangor, C., Swim, J. K., Sechrist, G. B., DeCoster, J., Van Allen, K. L., & Ottenbreit, A. (2003). Ask, answer, and
announce: The stages in perceiving and responding to discrimination. European Review of Social Psychology,
14, 277–310.
Swim, J. K., Aikin, K. J., Hall, W. S., & Hunter, B. A. (1995). Sexism and racism: Old-fashioned and modern
prejudices. Journal of Personality and Social Psychology, 68, 199–214.
Tajfel, H., & Turner, J. C. (1979). An integrative theory of intergroup conflict. In W. G. Austin, & S. Worchel
(Eds.), The social psychology of intergroup relations (pp. 33–47). Monterey, CA: Brooks/Cole.
Viki, G. T., & Abrams, D. (2002). But she was unfaithful: Benevolent sexism and reactions to rape victims who
violate traditional gender role expectations. Sex Roles, 47, 289–293.

Copyright # 2005 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 35, 633–642 (2005)