Allies Against Sexism: The Impact of Men’s Egalitarian Versus Paternalistic Confrontation on Women’s Empowerment and Well-Being

Abstract

Men as advantaged group members can be involved in actions against inequality. But how do women experience men’s confrontation of sexism? We examine how women perceive men’s egalitarian versus paternalistic confrontation of sexism. We hypothesized that women would be more likely to report empowerment and well-being (i.e., more happiness and less anger) after egalitarian confrontation than after paternalistic confrontation, which should increase their future intention to confront sexism. Using hypothetical scenarios, the results of three studies conducted in Spain, Germany, and Mexico confirmed our hypotheses. They also highlighted that empowerment (but not happiness) triggered by egalitarian confrontation, as well as anger triggered by paternalistic confrontation, lead women to express greater future intention to confront sexism. Our findings suggest that male confronters motivated by egalitarian reasons are more likely perceived as allies of women because they not only make women feel better but also empower them to keep fighting. Further, women may react against men motivated by paternalistic reasons (especially if they are strongly identified as feminist or endorse low benevolent sexist beliefs). Implications for activists, policymakers, and practitioners who are interested in involving men in fighting gender inequality are discussed.

You’re a woman partying with your girlfriends. A stranger starts flirting with you in an insistent and annoying way. Another man witnessing the scene decides to confront him and says: “Don’t be such a male chauvinist! Men should respect women and fight against inequality.” A third guy also gets involved and says, “Hey! Stop being rude! Men should treat women more delicately.”

How would you feel? Would you feel happy and grateful to those who intervened or annoyed because they assumed that you needed their protection? Both men confronted the perpetrator, but in a different manner: The first confronter labeled the perpetrator’s behavior as sexist whereas the second confronter failed to do so. From a social identity approach (Tajfel and Turner 1979; Turner et al. 1987), we assume that confrontation enacted by an advantaged group member can be considered as a form of intergroup behavior. The present paper contributes to the growing literature on allyship by examining the consequences of actions by advantaged group members against inequality on targets of discrimination. Specifically, we test the effects of two forms of confrontation against sexist behavior by advantaged group members (i.e., egalitarian vs. paternalistic) on women’s empowerment, well-being, and future intention to confront sexism.

Allies Against Sexism

Individual actions against inequality, such as confrontation, contribute to social change because they can reduce future sexist behaviors (Mallett and Wagner 2011), and they are associated with more competence, self-esteem, and empowerment among women (Gervais et al. 2010; Hyers 2007). However, explicit and public confrontation of sexism by women is infrequent (Hyers 2007; Mallett and Melchiori 2014; Swim and Hyers 1999), and those who confront risk being disliked by advantaged and disadvantaged group members (Dodd et al. 2002; Eliezer and Major 2012). In fact, many women consider confrontation unhelpful and aversive (Czopp and Monteith 2003).

Some studies suggest that men may be more effective than women in confronting sexist behavior because their actions are taken more seriously and they are less likely to experience social costs (Drury and Kaiser 2014; see also Kutlaca et al. 2019). Moreover, men’s acknowledgement of sexism may also empower women. For instance, women increased their self-confidence, showed less stereotype confirmation, and were more likely to file a complaint against a perpetrator when the discriminatory experience was confirmed by a male rather than by a female colleague (Cihangir et al. 2014). However, male confrontation of sexism may also have potential costs for women. Advantaged group member’s actions can contribute to normalizing power relations between groups (Hasan-Aslih et al. in press) and reinforce inequality by fostering the disadvantaged group member’s dependence on the advantaged group.

According to the model of intergroup helping as a status relation (Nadler 2002), there are two types of outgroup helping: dependency-oriented help (which perpetuates social hierarchies) and autonomy-oriented help (which challenges them). Autonomy-oriented help implies providing the tools for the disadvantaged group members to resolve their problems by themselves. Similarly, intergroup contact literature has highlighted that positive contact may undermine collective action by the disadvantaged group (Saguy et al. 2009; Wright and Lubensky 2008). In contrast, when advantaged group members explicitly recognize inequality as illegitimate, it does not reduce the disadvantaged groups’ support for social change (Becker et al. 2013). In fact, in opposition to positive contact, supportive contact (a specific positive intergroup contact characterized by recognizing inequality as illegitimate and by opposition to it) may increase engagement in collective action by the disadvantaged group (Droogendyk et al. 2016a). Thus, autonomy-oriented help and supportive contact might represent two forms in which advantaged group members can be allies for social change (Droogendyk et al. 2016b; Radke et al. 2020).

However, to understand whether advantaged group members’ actions contribute to social change or perpetuate social hierarchies, we must consider their underlying motivations (Broido 2000; Edwards 2006; Estevan-Reina et al. 2020; Louis et al. 2019; Radke et al. 2020). We propose that confrontation might have a different impact on women depending on the motivations underlying advantaged group members’ actions, or the way targets perceive these motivations. Specifically, we distinguish between two types of confrontation (egalitarian vs. paternalistic), depending on whether they aim to promote social change or perpetuate the status quo.

Egalitarian or Paternalistic Confrontation

Egalitarian or paternalistic reasons might motivate advantaged group members’ actions (Estevan-Reina et al. 2020). Egalitarian motives are linked to feminist identity—a form of politicized collective identity aimed at ending gender inequality (Simon and Klandermans 2001). In contrast, paternalism and sexism reinforce power asymmetries in intergroup relations (Glick and Fiske 1996; Jackman 1994). Specifically, literature has highlighted the pernicious effect of benevolent sexism in perpetuating gender inequality (Barreto and Ellemers 2005; Becker and Wright 2011; Jost and Kay 2005). One of the core aspects of benevolent sexism is the belief that men have a duty to protect women (i.e., protective paternalism: Glick and Fiske 1996), and it promotes dependency-oriented help (Shnabel et al. 2016). Importantly, the duty to protect women predicts the frequency of confronting sexism on behalf of socially close women, but not on behalf of distant ones (Good et al. 2018).

Estevan-Reina et al. (2020) found two distinct paths explaining men’s intention to confront sexism: a feminist path and a paternalistic one. Men’s endorsement of feminist identification led them to confront sexism through egalitarian motivation, whereas benevolent sexism leads men to confront sexism through paternalistic motivation. Moreover, only the feminist path leads men to express greater collective action intentions and actual engagement in social movements designed to question male societal privileges. Consistently, Radke et al. (2018) found that benevolent sexism in men (but not in women) was positively related to protective actions (e.g. behavior designed to guard women against male violence), but not to feminist collective actions (i.e., behaviors that challenge gender inequality). In contrast, feminist identification predicted willingness to engage in feminist actions for both genders.

Still, little is known about the consequences of men’s sexism confrontation on women’s empowerment and well-being (i.e., happiness and anger). We define egalitarian confrontation as a behavior triggered by beliefs about gender equality that push men to act against discriminatory situations; paternalistic confrontation, as a behavior triggered by beliefs about the duty to protect women that push men to act against discriminatory situations. Moreover, women’s reactions to male confrontation might be contingent on the extent to which women endorse feminist identification or benevolent attitudes. Finally, we examined whether egalitarian and paternalistic confrontation might motivate women to confront sexism.

Women’s Empowerment and Well-Being

Empowerment is a multifaceted concept that includes personal, relational and societal dimensions (Huis et al. 2017). From a feminist perspective, empowerment can be understood as “power-to,” which is close to the concept of self-efficacy (Bandura 1995), in opposition to “power-over” (Yoder and Kahn 1992). According to Zimmerman (1995), empowerment is at the same time both an outcome and a process. In this line, being empowered is a state in which one’s goals can be fulfilled (Pratto 2016). Previous literature has shown that confrontation is positively associated with competence, self-esteem, and empowerment among women (Gervais et al. 2010). We propose that men’s confrontation of sexism might also empower women. Some indirect evidence for this argument has been provided by previous literature (Cihangir et al. 2014; Droogendyk et al. 2016a). Egalitarian confrontation can be seen as a form of supportive contact (Droogendyk et al. 2016b) that may empower women because confrontation signals that one is supportive of social change. Thus, we hypothesize that egalitarian confrontation will empower women more than paternalistic confrontation (Hypothesis 1).

Furthermore, we expect egalitarian confrontation to have positive effects on women’s well-being (i.e., increased happiness—Hypothesis 2; decreased anger—Hypothesis 3) compared to paternalistic confrontation. Subjective well-being has been positively associated with pleasant and positive emotions (popularly referred to as “happiness”) and negatively associated with unpleasant and negative emotions (Diener et al. 2018). Disadvantaged group members who do not perceive the hierarchy as legitimate or stable might reject dependency-oriented help, such as paternalistic confrontation, and only accept autonomy-oriented help, which underlies more egalitarian relationships (Nadler 2002). In other words, if women perceive men’s confrontation as a form of sexist behavior because it is motivated by paternalistic beliefs, they might feel negatively about it and thus experience decreased empowerment and well-being. This effect should be most pronounced for women who identify as feminists and reject benevolently sexist beliefs. Recent research has shown that women strongly identified as feminist perceive a feminist man who offers autonomy-oriented help as a better ally than a man who offers dependency-oriented help (Wiley and Dunne 2019).

Empowerment, Anger, and Women’s Intention to Confront

We also investigate the roles empowerment, happiness, and anger play in motivating women to engage in social change. Intergroup conflict literature has pointed out the role of subjective power (labeled “efficacy”; Drury et al. 2015, p. 95) in motivating social change (Hornsey et al. 2006; van Zomeren et al. 2012; van Zomeren et al. 2008; van Zomeren et al. 2004). Thus, the expected positive effects of egalitarian confrontation on empowerment may enhance women’s future intention to confront. This linkage is consistent with the positive effect of efficacy on collective action (Social Identity Model of Collective Action: SIMCA; van Zomeren et al. 2008) and the needs-based model of reconciliation (Shnabel et al. 2009). According to the latter model extended to intergroup contexts, when advantaged group members restore disadvantaged group members’ sense of agency through their empowerment, this prevents passive acceptance of inequality and increases disadvantaged group members’ readiness to act for change (Shnabel and Nadler 2015). In fact, the perception of #MeToo movement as empowering for women is positively associated with their campaign support (Kende et al. 2020). In contrast, the role of positive emotions in promoting social change has been questioned. Self-directed positive emotions do not play an important role in predicting collective actions (Becker et al. 2011), and hope for harmony in intergroup conflicts is negatively associated with the disadvantaged group members’ motivation for collective action (among the low identifiers; Hasan-Aslih et al. 2019). Thus, we hypothesize that the empowerment (but not happiness) experienced after egalitarian confrontation will predict women’s future intention to confront (Hypothesis 4).

Anger triggered by perceived injustice also motivates participation in social change actions (Iyer et al. 2007; van Zomeren et al. 2008). In our work, however, we focused on the role of anger triggered by men’s paternalistic confrontation. We argue that confrontation based on paternalistic arguments might trigger more opposition than egalitarian confrontation because paternalism maintains the status quo and reinforces social hierarchies (Becker and Wright 2011; Jost and Kay 2005). Recent literature has shown that even subtle discrimination cues can trigger resistance responses in women, which include reporting more anger (de Lemus et al. 2018). Thus, we hypothesize that paternalistic confrontation might trigger anger in women as a form of resistance against a sexist man, which might increase their future intention to confront (Hypothesis 5).

The Current Studies

Based on the social identity approach (Tajfel and Turner 1979; Turner et al. 1987) and extending previous research that has proposed the distinction between autonomy-oriented and dependency-oriented help (Nadler 2002), as well as cross-group positive and supportive contact (Droogendyk et al. 2016a), we examine the impact of egalitarian and paternalistic confrontation of sexism by men. Specifically, the aim of our research is to examine the effects of men’s confrontation on women’s empowerment, well-being, and future intention to confront sexism (Studies 1, 2, and 3). We hypothesize that egalitarian confrontation will lead to more empowerment and happiness but less anger among women than paternalistic confrontation. We also expect that the empowerment (but not happiness) triggered by egalitarian confrontation, as well as anger triggered by paternalistic confrontation, will predict women’s future intention to confront.

In Studies 2 and 3, we included a target-confrontation condition, in which the woman confronts sexism herself, in order to be able to compare the effects of target versus advantaged group member’s confrontation on women’s empowerment and well-being. We also wanted to analyze whether these processes were consistent across different cultural contexts. For this reason, we conducted our studies in the following countries: Spain (Study 1), Germany (Study 2: a preregistered study), and Mexico (Study 3: a preregistered study). According to the Gender Inequality Index of the United Nations Development Programme (2017), Germany and Spain have similar levels of gender inequality, and both countries have a lower level of gender inequality than Mexico. In less egalitarian countries, women endorse more benevolent sexist beliefs (Glick et al. 2000); thus, paternalism might be more accepted in Mexico than in Spain and Germany. Finally, we conducted an integrative data analysis by pooling the three datasets into one (Curran and Hussong 2009), which allowed us not only to test the differences among countries but also to check the main results of Studies 1–3 with more statistical power and sample heterogeneity. A larger sample size also allowed us to explore feminist identification and benevolent sexism as possible moderators. All data collections were reviewed and approved by university Institutional Review Boards.

Pilot Study

We recruited 60 participants to take part in our pilot study on the campuses of a Spanish university (n = 30) and a German university (n = 30) in exchange for a chocolate bar. Half the participants in each country were randomly assigned to read the egalitarian confrontation scenario and the other half read the paternalistic one. They then completed 14 items that included questions about the confronter. Four items measured the extent to which they perceived the confronter as paternalistic (e.g., “he is protecting women”; after excluding one of them with a total-item correlation under .10, α = .77); three items measured the extent to which they perceived the confronter as sexist (e.g., “he is macho”; α = .80); and the other seven items measured the extent to which they perceived the confronter as egalitarian (e.g., “he is fighting against gender inequality”; α = .91). The participants rated their opinions from 1 (totally disagree) to 7 (totally agree).

We conducted a Multivariate Analysis of Variance (MANOVA) on perceptions of the confronter with type of confrontation (egalitarian vs. paternalistic) and country (Spain vs. Germany) as between-subject factors. That analysis revealed that the manipulation had a significant multivariate effect, Wilks’s Λ = .700, F(3, 54) = 22.42 p < .001, η_p² = .300. As expected, women perceived the egalitarian confronter as more egalitarian (M = 4.55, SE = .26) than the paternalistic confronter (M = 3.10, SE = .26), F(1, 56) = 15.11, p < .001, η_p² = .213. In contrast, they perceived the paternalistic confronter as more sexist (M = 4.02, SE = .33) than the egalitarian confronter (M = 2.82, SE = .33), F(1, 56) = 6.70, p = .012, η_p² = .107. Women perceived the paternalistic (M = 4.49, SE = .31) and the egalitarian (M = 4.3, SE = .31) confronters as similarly paternalistic, F(1, 56) = .18, p < .673, η_p² = .003. Neither a univariate main effect of country (F < .60, p = .441) nor an interaction between the type of confrontation and country (F < .98, p = .327) was found. Despite this similar perception in terms of paternalism, the man who confronted in a blatantly paternalistic way was perceived as more sexist than the egalitarian one. These ratings provide empirical support that the paternalistic confrontation is qualitatively different from the condition in which the man expresses egalitarian reasons to confront (although this can still be perceived as paternalistic from the perspective of women to the extent that it implies acting on their behalf). From the perspective of the advantaged group’s motivations to confront, we label the two conditions as egalitarian and paternalistic. However, women’s perceptions of the two confronters may differ depending on their interpretation of men’s motivations and actions. We will address this point in the general discussion.

Study 1

We tested whether imagined men’s egalitarian or paternalistic confrontation had different consequences for women. We hypothesized that after men’s egalitarian confrontation, women would be more likely to feel empowered (Hypothesis 1) and experience more happiness (Hypothesis 2) and less anger (Hypothesis 3) than after men’s paternalistic confrontation. Furthermore, we analyzed the implications of empowerment, happiness, and anger for women’s future intention to confront.

Method

Participants

A total of 200 Spanish women took part in the study. One participant was excluded because she did not finish the questionnaire. The final sample consisted of 199 women. The participants’ ages ranged from 18 to 33 years-old, with a mean age of 22.03 years (SD = 2.73, mdn = 21). Of the total number of participants 193 (97%) were students from a university in the south of Spain and 192 (97.5%) were Spanish citizens. We conducted a sensitivity analysis using G*power (Faul et al. 2007) to determine the effect size that the current study could detect. The results showed that with this sample size (n = 199) and with α = .05 and 1-β (power) = .80, the minimum effect size that we could detect for an ANOVA unifactorial analysis was f = .20, and the minimum effect size we could detect for a multiple regression with two predictors was f ² = .05.

Procedure and Measures

We approached students at the university library to encourage them to take part in a 15-min paper-and-pencil survey. We first recorded participants’ ages, nationality, and occupation. The rest of the measures are described here in the same order as they appeared in the survey unless otherwise specified. At the end, participants were debriefed and rewarded with chocolate bars to thank them for their contributions.

Men’s Confrontation Manipulation

All participants saw a hypothetical scenario presented in the style of a comic that represented a social interaction in which a man makes a sexist comment to a woman. We asked participants to imagine that they were the targets of the sexist comment. The first picture depicted a woman asking two men on the street for a lighter. The second picture depicted the perpetrator saying: “Of course, I‘ll lend it to you, gorgeous. But only if in return you’ll come to sleep with me tonight, because I don‘t want to sleep alone.” A third picture included the confrontation manipulation depending on the experimental condition. In the egalitarian condition, the male bystander says, “Hey! What’s up? That comment is sexist. I don’t think that it’s fair to treat women like that. Men should fight against gender inequality.” In the paternalistic condition, a male bystander confronts the sexist comment by saying, “Hey! What’s up? That comment is rude. I don’t think that it’s appropriate to treat women like that. Men should take care of and protect women.” The comics are provided in the online supplement.

Empowerment

We measured empowerment with eight items adapted from Moya-Garófano et al. (2018), namely “powerful,” “full of energy,” “stimulated,” “empowered,” “without control of the situation,” “weak,” “inferior,” and “defenseless.” We assessed participants’ happiness and anger, asking them how they would feel after hearing the confronter’s comment. Responses were recorded on a scale ranging from 0 (nothing) to 10 (very much). Scores on the items designed to measure low empowerment were reversed, and a total score was calculated, with higher scores indicating greater empowerment (α = .83).

Emotions

We used the Escala de Valoración del Estado de Ánimo (EVEA) (Scale for Mood Assessment; Sanz 2001), which measures the following emotions: happiness (happy, optimistic, joyful, and cheerful), hostility (irritated, angry, annoyed, and displeased), sadness, and anxiety. (More information can be found in the online supplement.) Additionally, based on literature that highlights the role of anger in promoting collective actions (van Zomeren et al. 2004; van Zomeren et al. 2012), we decided to include five anger-related items (“with rage,” “outraged,” “insulted,” “offended,” and “humiliated”). It is important to note that these adjectives measure emotions toward the confronter’s rather than the perpetrator’s comment, which is why we evaluated interpersonal rather than intergroup anger. We also included four items measuring the feeling of gratitude (“respected,” “comfortable,” “relaxed,” and “grateful”). Responses were recorded on a scale ranging from 0 (nothing) to 10 (very much). We conducted a principal components analysis with varimax rotation. (Factor loadings can be found in the online supplement.) It extracted four factors with eigenvalues higher than 1 that explained 68.84% of the variance. Anger items were loaded together with the EVEA hostility items, whereas the gratitude items were loaded on the happiness factor. Therefore, all these items were averaged across two dimensions (anger, 9 items: α = .96; happiness, 8 items: α = .90).

Confrontation Intentions

We asked the participants how they would behave if they experienced a similar sexist situation. We selected two items (“I would tell him that he has no right to treat women like this” and “I would let him know that I don’t think it’s right to have this kind of attitude toward women”) from a broader set of items used in previous studies (Estevan-Reina et al. 2020). The Pearson correlation between both items was adequate (r = .74). We included additional items to assess aggressive confrontation, denigratory confrontation, and avoidance responses. (More information about these items can be found in the online supplement.)

Manipulation Check

We used the same items as in the pilot study to measure to what extent the confronter was perceived by women as egalitarian (8 items, α = .93) and paternalistic (3 items, α = .83). Evaluations of both the perpetrator’s and the confronter’s comments were measured with two items (“To what extent do you consider the comment of the [white/black shirt] guy to be sexist?” and “To what extent do you consider the comment of the [white/black shirt] guy to be very negative/very positive?”). The format of responses was from −3 to +3.

In addition, participants rated their political orientation, endorsement of benevolent sexism, feminist identification, postural measure of submission or dominance, self-description as agentic or communal, and awareness of gender inequality. (These additional measures are described in detail in the online supplement.)

Results

Manipulation Check

We conducted a MANOVA, including the type of confrontation (egalitarian vs. paternalistic) as the independent variable and perceptions of the confronter as egalitarian or paternalistic as dependent variables, revealing a significant multivariate effect, Wilks’s Λ = .518, F(2, 196) = 91.27, p < .001, η_p² = .482. A significant univariate effect of the type of confrontation emerged on perceived egalitarianism, F(1, 197) = 141.25, p < .001, η_p² = .418. Women perceived the confronter in the egalitarian condition (M = 5.03, SD = 1.35) as more egalitarian than the confronter in the paternalistic condition (M = 2.71, SD = 1.41). There was no significant effect on perceptions of paternalism, F(1, 197) = .96, p = .329, η_p² = .005.

We conducted a second MANOVA, including the type of confrontation (egalitarian vs. paternalistic) as the independent variable and women’s perceptions of the perpetrator’s and confronter’s comments as dependent variables, uncovering a significant multivariate effect. Wilks’s Λ = .682, F(4, 192) = 22.42, p < .001, η_p² = .318. As we expected, univariate analyses showed no significant differences in how women evaluated the perpetrator’s comment (perceived sexism: F(1, 195) = .02, p = .89, η_p² < .001; negative/positive valence: F(1, 195) = .34, p = .56, η_p² = .002) but significant differences in how they perceived the confronter’s comment (perceived sexism: F(1, 195) = 81.96, p < .001, η_p² = .296; negative/positive valence: F(1, 195) = 63.87, p < .001, η_p² = .247). Specifically, women perceived the paternalistic confronter as more sexist (M = 1.15, SE = .18) and negative (M = .26, SE = .18) than the egalitarian confronter (M = −1.23, SE = .19 and M = 1.80, SE = .18, respectively).

Women’s Empowerment and Well-Being

To test Hypotheses 1 through 3, we conducted a univariate MANOVA, including the type of confrontation (egalitarian vs. paternalistic) as the independent variable. The empowerment and the two emotions representing well-being (anger and happiness) were dependent variables, finding a significant multivariate effect of the type of confrontation, Wilks’s Λ = .844, F(3, 195) = 11.98, p < .001, η_p² = .156. As predicted, the type of confrontation had a significant effect on empowerment, F(1, 197) = 12.52, p = .001, η_p² = .060; happiness, F(1, 197) = 29.50, p < .001, η_p² = .130; and anger, F(1, 197) = 31.91, p < .001, η_p² = .139. The results showed that women reported more empowerment and happiness as well as less anger after the imagined male egalitarian confrontation than after the male paternalistic confrontation (see Table 1a). Thus, Hypotheses 1, 2, and 3 were supported.

Table 1 Main effects of type of confrontation on Women’s empowerment, well-being, and future intention to confront by country

Women’s Future Intention to Confront Via Empowerment and Anger

To test whether the empowerment and anger that women experienced after being exposed to a hypothetical scenario of confrontation would lead them to express greater future intention to confront sexism (Hypotheses 4 and 5), as well as to explore the role of happiness in predicting future confrontation intentions, we conducted a multiple mediation model with the macro PROCESS (Hayes 2013), using 5000 bootstrap samples to estimate bias-corrected standard errors and 95% confidence intervals. We performed a parallel mediational model (Model 4 in PROCESS) that included empowerment, happiness, and anger as mediators (see Fig. 1). The total effect of type of confrontation on women’s future intention to confront sexism was not significant (b = .07, 95% CI [−.39, .54], p = .739). Means and standard errors are shown in Table 1a. The indirect effects of type of confrontation through empowerment (b = .22, 95% CI [.07, .46]) and anger (b = −.30, 95% CI [−.63, −.07]) were significant, but not the indirect effect through happiness (b = .02, 95% CI [−.23, .29]). The direct effect remained nonsignificant when the mediators were included in the model (b = .13, 95% CI [−.36, .62], p = .550). In line with Hypotheses 4 and 5, these results showed that higher levels of empowerment and anger (but not of happiness) predicted higher intention to confront sexism.

Fig. 1

Parallel mediation model for the relationship between type of confrontation and women’s future intentions to confront. Bs are reported. Dashed line indicates a nonsignificant pathway. S1 = study 1 (Spain); S2 = study 2 (Germany); S3 = study 3 (Mexico); P = pooled analyses. *p < .05. **p < .01. ***p < .001.

Discussion

Study 1 supports the idea that women react more positively after witnessing an egalitarian confronter than a paternalistic confronter. First, men’s egalitarian confrontation made women report feeling more empowered and happy. Second, the results indicate that increased empowerment (but not happiness) motivates women to express greater intention to act against sexism in the future. Additionally, we found that women reported experiencing more anger after paternalistic rather than after egalitarian confrontation. This effect may be due to male egalitarian confrontation reducing women’s anger (increasing their well-being), as well as to negative reactions of female participants to the paternalistic confrontation. Consistent with previous literature about the role of anger in predicting action, the results suggest that increases in anger lead women to express greater future intention to confront sexism.

Although in Study 1 and in the pilot study the man who confronts in an egalitarian way was perceived by women as more egalitarian and less sexist than the paternalistic confronter, both were perceived as paternalistic to the same extent. These results suggest that when a man confronts sexism on a woman’s behalf, even if he is guided by egalitarian attitudes, he may still be perceived as paternalistic because he is not allowing the woman to act by herself. Therefore, it is important to compare male egalitarian confrontation with a situation confronted by a female target of sexism, which to our knowledge has not been done before. We incorporated target confrontation in Studies 2 and 3. In the months prior to data collection, massive demonstrations took place demanding gender equality in Spain (Gómez 2019; Grodira et al. 2018). Therefore, to be able to generalize our findings beyond the Spanish context, we decided to run two new studies in different cultural contexts (Germany and Mexico).

Studies 2 and 3

In these two studies we included a new experimental condition (target confrontation). As in Study 1, we hypothesized that women would be more likely to feel empowered (Hypothesis 1a) and experience more well-being (more happiness—Hypothesis 2a; less anger—Hypothesis 3a) after a male egalitarian confrontation than after a male paternalistic one. We further hypothesized that women would feel more empowered after imagining themselves as confronters (target confrontation) than after a male egalitarian (Hypothesis 1b) or paternalistic confrontation (Hypothesis 1c) because confrontation by women is positively associated with their sense of competence, self-esteem, and empowerment (Gervais et al. 2010; Hyers 2007). Because previous literature has documented that confrontation includes important emotional costs for women (Czopp and Monteith 2003; Dodd et al. 2002; Eliezer and Major 2012; Gervais and Hillard 2014), we also hypothesized that women would experience less well-being after imagining themselves confronting (target confrontation) than after male egalitarian and paternalistic confrontation. Thus, after target confrontation, women would feel less happiness (Hypothesis 2b) and more anger than after male egalitarian (Hypothesis 3b) or paternalistic confrontation (happiness—Hypothesis 2c; anger—Hypothesis 3c). However, it is important to note that the emotions experienced by women after imagining their own confrontation in contrast to a male confrontation reflect different processes. Emotions that women experience after male confrontation may reflect agreement or disagreement with the male confronter, whereas emotions experienced after taking the perspective of a disadvantaged group member’s confrontation may project facing a threatening situation by themselves.

In relation to the indirect effects of male confrontation on women’s future intention to confront sexism, we expected to replicate the results found in Study 1 regarding men’s types of confrontation (egalitarian vs. paternalistic) on women’s future intention to confront sexism via empowerment (Hypothesis 4) and anger (Hypothesis 5). Although in Study 1 this indirect effect through happiness was not significant, we explored it again in Studies 2 and 3 in different cultural contexts.

Method

Participants

In Study 2, 315 German women started the online survey. However, 79 were excluded because they did not finish it, eight because they did not answer the manipulation check correctly, three because they self-identified as men, and two because the time they spent answering the survey exceeded the total average time by more than two standard deviations. The final sample comprised 223 women. Participants’ ages ranged between 17 and 45 years-old, with a mean age of 23.59 years (SD = 4.30, mdn = 23). Of the total number of participants, 218 (97.3%) were students from a northern university in Germany, and 217 (97.3%) were German citizens.

In Study 3, 180 Mexican women answered the questionnaire. Four participants were excluded because they did not answer the manipulation check, another four because they failed the manipulation check question, and one more because she did not complete the questionnaire. The final sample consisted of 171 women. Participants’ ages ranged between 18 and 36 years-old, with a mean age of 21.26 years (SD = 2.65, mdn = 21). All were Mexican students from a southeast university in Mexico. An univariate ANOVA showed significant differences across samples in age, F(2, 590) = 24.65, p < .001, ηp2 = .077, with German participants being older (M = 23.59, SD = 4.29) than Spanish (M = 22.02, SD = 2.73) and Mexican ones (M = 21.26, SD = 2.65).

According to effect sizes detected in Study 1 for ANOVA (f = .25; medium effect) using G*Power, we estimated a minimum sample of 154 participants to obtain a power (1 − β) = .80. For the same power standard, a minimum sample of 156 participants was needed, according to Monte Carlo simulation for indirect effects.

Procedure and Measures

To collect the data for Study 2, three research assistants approached students who were on the university campus and invited them to take part in the study, offering sweets as an incentive. If they accepted, the students provided their e-mail addresses and were later sent an e-mail with a link to the 15-min online survey. At the end of the survey, participants were debriefed and asked again for their e-mail addresses (stored separately from their answers) in case they wanted to participate in a raffle for one of five €20 Amazon vouchers. Participants in Study 3 were approached by one female researcher, who asked them to take part in a 15-min paper-and-pencil survey. At the end, participants were thanked and debriefed.

The measures used in Studies 2 and 3 were the same as those used in Study 1 with the exceptions that in Study 2 we employed scales validated in German (or translated to German when no validations were available) and in Study 3 we adapted some items to the Mexican context. Both Study 2 (https://osf.io/nfg8z ) and Study 3 (https://osf.io/m4rqh) were preregistered in the Open Science Framework platform.

Confrontation Manipulation

In Studies 2 and 3, we used the same vignettes described in Study 1. A third experimental condition was incorporated in which the woman herself confronted the sexist comment. The content of the target confrontation was the same as in the male egalitarian condition, but in this case the woman gave the egalitarian argument.

Empowerment

We measured empowerment with the same eight items as in Study 1, either translated into German (Study 2: α = .84) or adapted to the Mexican context (Study 3: α = .78). In Study 3, we culturally adapted one item, replacing estimulada (i.e., stimulated) with activada (i.e., activated).

Emotions

In Studies 2 and 3, we measured happiness and anger with the same items used in Study 1. In Study 2, for translation reasons, we included four items to measure anger instead of five because we did not find distinctive equivalent words for all of them. In Study 3, one item, alicaída (i.e., downcast), was culturally adapted, replaced by desanimada (i.e., disheartened). The main components of factor analysis with varimax rotation extracted two factors with eigenvalues larger than 1, which explained 67.32% of the variance in Study 2 and 71.55% of the variance in Study 3. The reliability coefficients were strong for happiness (Study 2: α = .93; Study 3: α = .90) and anger (Study 2: α = .92; Study 3: α = .96).

Confrontation Intentions

They were measured with the same two items as in Study 1, with the addition of two more items (“I would try to make the guy see that his attitude is offensive” and “I would try to explain to the guy that his comment bothered me”). The reliability coefficient for the set of four items was acceptable in Study 2 (α = .85) and in Study 3 (α = .75).

Manipulation Checks

We asked participants to remember the social interaction described in the vignettes and select the option that best summarized it (attention check). We offered them four possible options, one for each experimental condition and one additional in case they did not remember well what they had previously read. Because materials for the experimental manipulation had not been validated previously in a Mexican context, we also included in Study 3 the items used to validate the scenarios in the pilot study: four items to measure the perception of the confrontation as paternalistic (α = .67) and seven items to measure the perception of the confronter as egalitarian (α = .89).

In addition, participants rated their political orientation, endorsement of benevolent sexism, feminist identification, and self-description as agentic or communal, as well as answered a modern sexism scale. (These additional measures are described in detail in the online supplement.)

Results

Manipulation Check

Most participants selected the correct attention check options in Study 2 (78, 97.5% in the target confrontation condition; 76, 96.2% in the male feminist confrontation; and 69, 95.8% in the male paternalistic confrontation) and in Study 3 (57, 100% in the target confrontation condition; 55, 93.2% in the male feminist confrontation; and 59, 100% in the male paternalistic confrontation).

Because materials had not been piloted in Study 3, we conducted a MANOVA to check that women perceived the confronter in an egalitarian or a paternalistic way, documenting a significant multivariate effect, Wilks’s Λ = .659, F(2, 111) = 28.68, p < .001, η_p² = .341. A significant univariate effect of condition emerged on the set of feminist items, F(1, 112) = 34.33, p < .001, η_p² = .235. Women perceived the confronter in the egalitarian condition as more egalitarian (M = 4.84, SE = .19) than the confronter in the paternalistic condition (M = 3.27, SE = .19). However, we again did not find an effect of condition on paternalistic items, F(1, 112) = .63, p = .43, η_p² = .006. Thus, these results replicate the findings in the Spanish and German pilot studies.

Women’s Empowerment and Well-Being

As in Study 1, we conducted a MANOVA to test whether there were differences in the empowerment and well-being (happiness and anger) that women experienced as a function of the scenario that they had previously read (target confrontation vs. egalitarian confrontation by man vs. paternalistic confrontation by man).

In Study 2, we found a significant multivariate main effect of type of confrontation, Wilks’s Λ = .373, F(6, 436) = 46.30, p < .001, η_p² = .389. A significant univariate effect of confrontation emerged on empowerment, F(2, 220) = 19.94, p < .001, η_p² = .153; happiness, F(2, 220) = 57.53, p < .001, η_p² = .343; and anger, F(2, 220) = 43.97, p < .001, η_p² = .286. Post hoc analyses (Sidak) revealed that participants reported feeling more empowered after a man’s egalitarian confrontation than after a man’s paternalistic confrontation, as in Study 1 (see Table 1b). Moreover, they experienced even more empowerment after target confrontation than after both types of men’s confrontations. Concerning well-being, participants felt more happiness and less anger when men confronted in an egalitarian versus paternalistic way, as we found in Study 1. Additionally, participants felt more anger and less happiness after target confrontation than after men’s (egalitarian and paternalistic) confrontations (see Table 1b). Thus, in Germany, Hypotheses 1a–c, 2a–c, and 3a–c were supported.

In Study 3, we found a significant multivariate main effect of type of confrontation, Wilks’s Λ = .367, F(6, 332) = 36.06 p < .001, η_p² = .395. A significant univariate effect of confrontation emerged on empowerment, F(2, 168) = 8.81, p < .001, η_p² = .095; happiness, F(2, 168) = 53.44, p < .001, η_p² = .389; and anger, F(2, 168) = 68.79, p < .001, η_p² = .450. Post hoc (Sidak) analyses revealed that participants reported feeling more empowered after men’s egalitarian rather than men’s paternalistic confrontations and more empowered after target confrontation than after men’s paternalistic confrontation (see Table 1c). There were no significant differences between target confrontation and men’s egalitarian confrontation on empowerment, contrary to Study 2. With regard to well-being, as in Studies 1 and 2, participants felt more happiness and less anger when men confronted in an egalitarian versus paternalistic way. Also, as in Study 2, participants felt more anger and less happiness after target confrontation than after men’s (egalitarian and paternalistic) confrontations (see Table 1c). Thus, in Mexico, Hypotheses 1a and 1c, 2a–c, and 3a–c were supported, but Hypothesis 1b was not.

Women’s Future Intention to Confront Via Empowerment and Anger

As in Study 1, to know whether empowerment, anger, and happiness induced by the manipulation led women to express greater future intention to confront, we conducted process analyses (Hayes 2013) using 5000 bootstrap samples to estimate bias-corrected standard errors and 95% percentile confidence intervals. We used a parallel mediational model (Model 4 in PROCESS) including empowerment, happiness, and anger as mediators (see Fig. 1). Because the independent variable had three levels, to run these analyses we created two contrasts. To replicate the results of Study 1, in Contrast 1 we compared men’s egalitarian confrontation (coded 1) versus men’s paternalistic confrontation (coded −1; target confrontation coded 0). In Contrast 2, we compared target confrontation (coded 2) to men’s confrontations (egalitarian −1; paternalistic = −1). All the analyses were conducted including Contrast 1 as the main predictor and Contrast 2 as a covariate to control for it.

In Study 2 (Germany), the total effect of Contrast 1 (egalitarian vs. paternalistic confrontation) on future intention to confront was significant (b = −.47, 95% CI [−.88, −.06], p = .024), as well as the indirect effect through empowerment (b = .16, 95% CI [.02, .37]) and anger (b = −.21, 95% CI [−.44, −.06]), but not through happiness (b = .01, 95% CI [−.11, .16]) (see Table 2b). The direct effect was significant (b = −.44, 95% CI [−.86, −.01], p = .046). However, in Study 3 (Mexico), the total effect of this contrast was not significant (b = −.11, 95% CI [−.35, .13], p = .353), but the indirect effect via empowerment was (b = .09, 95% CI [.02, .22]) (see Table 2c). No other indirect effects were found in Study 3 (anger: b = −.05, 95% CI [−.17, .03]; happiness: b = .02, 95% CI [−.09, .13]). The direct effect was not significant (b = −.18, 95% CI [−.43, .07], p = .167). Thus, Hypothesis 4 was supported in Germany and Mexico, whereas Hypothesis 5 was supported in Germany but not in Mexico.

Table 2 Total, Direct, and Indirect Effects of Type of Confrontation (Egalitarian or Paternalistic) on Women’s Intention to Confront via Feeling of Power, Happiness, and Anger

Summary of the Results across Studies

The effects of type of men’s confrontation on women’s empowerment and well-being found in Study 1 were replicated in two different cultural contexts (Study 2: Germany and Study 3: Mexico). Men’s egalitarian confrontation had beneficial effects on women compared to paternalistic confrontation because it made women feel more empowered, happier, and less angry. Concerning the expected differences between target confrontation and men’s (egalitarian and paternalistic) confrontations, in Germany and Mexico, participants felt more empowered after target confrontation than after paternalistic men’s confrontation, in line with our hypotheses. However, whereas in Germany participants also felt more empowered after target confrontation than after male egalitarian confrontation, this was not the case in Mexico. That is, Mexican women were equally empowered by target confrontation and men’s egalitarian confrontation. Both in Germany and in Mexico, we found that when women imagined that they were the confronters (target confrontation condition), they experienced less happiness and more anger than after men’s egalitarian and paternalistic confrontations. This pattern is consistent with the fact that women consider confrontation aversive (Czopp and Monteith 2003; Dodd et al. 2002; Eliezer and Major 2012).

Regarding the indirect effects of type of confrontation on women’s future intention to confront, the results in Germany and Mexico confirmed that empowerment experienced after men’s egalitarian (vs. paternalistic) confrontation led women to express greater future intention to confront. However, the more anger women experienced after paternalistic (vs. egalitarian) confrontation also pushed them to confront in Germany (but not in Mexico). Thus, in Study 2, we replicated the results of Study 1 in Spain with a German sample, but some differences emerged in Mexico (Study 3). To check the stability of the results with a larger sample, we decided to conduct an integrative data analysis with the three datasets pooled into one (Curran and Hussong 2009), taking into consideration only the two experimental conditions present in the three studies (men’s egalitarian vs. paternalistic confrontations).

Pooled Analyses of Studies 1, 2, and 3

Across studies, there was evidence that women react differently to paternalistic and egalitarian confrontation. To provide insight into the robustness of the central effect, we pooled the data following an integrative data analysis approach (Curran and Hussong 2009), which allowed us not only to test the possible differences among countries but to check the main results of Studies 1–3 with more statistical power and sample heterogeneity. First, we tested whether men’s egalitarian confrontation increased women’s empowerment and happiness (Hypotheses 1 and 2) and decreased anger (Hypothesis 3) compared to men’s paternalistic confrontation. Further, the data pooled from Studies 1–3 provide stronger statistical power to explore the role of feminist identification and endorsement of benevolent sexism as possible moderators of the effects of type of confrontation on women’s empowerment and emotions. According to previous literature, we consider that the effects of type of confrontation might be most pronounced for women highly identified as feminists and those who endorse less benevolently sexist beliefs. Finally, we conducted a parallel mediation model (Model 4 in PROCESS; Hayes 2013) to test the effect of male egalitarian confrontation in predicting women’s future intention to confront via empowerment and anger (Hypothesis 4 & 5), and we also explored the role of happiness.

Method

Participants

The total sample included 457 participants (n₁ = 198; n₂ = 145; n₃ = 114). Note that the difference in sample size of Study 2 (n = 223) and Study 3 (n = 171) is due to the fact that, in the pooled analyses, we did not include the target confrontation condition. We conducted a sensitivity analysis using G*power (Faul et al. 2007) to determine the effect size the current study could detect. Results showed that with α = .05 and 1-β (power) = .80, for a sample size of 457 participants, the minimum effect size that we could detect for a unifactorial ANOVA was f = .13, and for a multiple regression with four predictors it was f ² = .02.

Measures

Beyond the measures described in the corresponding sections of Studies 1–3, participants reported their gender and feminist identification as well as their endorsement of benevolent sexist beliefs before the manipulation.

Gender and Feminist Identification

These were measured with two items: “To what extent do you identify with your gender/ feminists?” (adapted from Doosje et al. 1998) and “To what extent do you feel a bond with other members of your gender/ feminist people?” (adapted from Leach et al. 2008), scored from 1 (not at all) to 7 (very much). In Studies 1 and 3 items were written in Spanish, whereas in Study 2 they were written in German. The Pearson correlation between both items was good for feminist identification in all the studies (Study 1: r = .80, M = 5.95, SD = 1.27; Study 2: r = .83, M = 4.14, SD = 1.64; Study 3: r = .89, M = 4.59, SD = 1.48) but not for gender identification (Study 1: r = .13; Study 2: r = .45; Study 3: r = .29). Gender identification thus was not included in our analyses. The two items for Feminist Identification were averaged so that higher scores indicated stronger identification.

Benevolent sexism

This was measured using the six items of the short version (Rollero et al. 2014) of the Benevolent Sexism subscale of the Ambivalent Sexism Inventory (Glick and Fiske 1996; Spanish version by Expósito et al. 1998; German version by Eckes and Six-Materna 1999), which showed it had good psychometric properties in all studies (Study 1: α = .80, M = .88, SD = .88; Study 2: α = .78, M = 1.50, SD = .98; Study 3: α = .75, M = 1.37, SD = .92). Items were averaged so that higher scores indicate stronger endorsement of benevolent sexism.

Results

Women’s Empowerment and Well-Being

We conducted a MANOVA to compare whether there were differences in empowerment, anger, and happiness that women experienced based on type of confrontation (men’s egalitarian vs. paternalistic) by country (Spain vs. Germany vs. Mexico). We found significant multivariate main effects of type of confrontation, Wilks’s Λ = .859, F(4, 447) = 18.41, p < .001, η_p² = .141, and country, Wilks’s Λ = .944, F(8, 894) = 3.25, p = .001, η_p² = .028, but interaction between type of confrontation by country was not significant, Wilks’s Λ = .970, F(8,894) = 1.69, p < .096, η_p² = .015.

A significant univariate effects of type of confrontation emerged on empowerment, F(1, 450) = 42.72, p < .001, η_p² = .087; happiness, F(1, 450) = 44.16, p < .001, η_p² = .089; and anger, F(1, 450) = 53.27, p < .001, η_p² = .106. Participants reported feeling significantly more empowered after male egalitarian confrontation than after paternalistic confrontation (see Table 1d). Likewise, participants felt more happiness and less anger after male egalitarian versus paternalistic confrontation. Thus, Hypotheses 1, 2, and 3 were supported.

A significant univariate effect of country also emerged on anger, F(2, 450) = 7.13, p = .001, η_p² = .031. Post hoc (Sidak) analyses revealed that in Spain (M = 4.75, SE = .19) and Germany (M = 4.62, SE = .22), women experienced significantly more anger than in Mexico (M = 3.62, SE = .25). No other significant differences among countries were found, Fs < 1.98, ps > .140, nor was an interaction effect between type of confrontation and country found, Fs < 1.03, p > .358.

Women’s Empowerment and Well-Being as a Function of Benevolent Sexism and Feminist Identification

To check whether the results were contingent on participants’ feminist identification and benevolent sexism, we conducted a moderation analysis through Hayes’ (2013) PROCESS command (Model 1) using 5000 bootstrap samples to estimate bias-corrected standard errors and 95% percentile confidence intervals. We reported these analyses with pooled data from Studies 1–3 rather than each study separately to increase statistical power, which allows us to detect small effect sizes. We found an interaction of confrontation (egalitarian vs. paternalistic) (a) with feminist identification on empowerment (b = .22, 95% CI [.02, .42], p = .034), happiness (b = .45, 95% CI [.21, .69], p < .001), and anger (b = −.48, 95% CI [−.77, −.19], p = .001), and (b) with benevolent sexism on empowerment (b = −.46, 95% CI [−.82, −.11], p = .010), happiness (b = −.80, 95% CI [−.1.21, −.39], p < .001), and anger (b = 1.05, 95% CI [.55, 1.55], p < .001). The more women identify as feminists, the less happiness and more anger they experienced after men’s paternalistic confrontation. Likewise, the lower the benevolent sexism, the less empowerment and happiness and the more anger they experienced after paternalistic confrontation. Conditional effects are reported in Table 3.

Table 3 Conditional effects of feminist identification and benevolent sexism on Women’s empowerment, happiness, and anger under two types of confrontation

An example of this pattern of results using anger as an outcome variable is represented in Fig. 2. It is important to note that even though the interactions reported are significant, the interaction between type of confrontation with both feminist identification and benevolent sexism on empowerment was still underpowered, so it must be interpreted with caution. In fact, although the interaction effect between feminist identification and type of confrontation on empowerment is significant, conditional effects are not (see Table 3).

Fig. 2

Interaction between feminist identification and type of confrontation and on women’s anger (pooled data). Lower and higher feminist identification represent −1 SD and + 1 SD from the mean, respectively

Women’s Future Intention to Confront Via Empowerment and Anger

The total effect of type of confrontation on women’s future intention to confront was not significant (b = −.32, 95% CI [−.66, .03], p = .072). The indirect effect through empowerment was significant (b = .14, 95% CI [.03, .28]), as well as through anger (b = −.21, 95% CI [−.39, −.06]), but not through happiness (b = .06, 95% CI [−.10, .22]). The direct effect was not significant (b = −.31, 95% CI [−.68, .06], p = .115). These results confirmed that the more empowerment women reported after egalitarian (vs. paternalistic) confrontation and the more anger they felt after paternalistic (vs. egalitarian) confrontation, the more they expressed greater future intention to confront. However, the more happiness women experienced after egalitarian (vs. paternalistic) confrontation did not lead them to express greater future intention to confront (see Table 2d). Thus, when we pooled the data of Studies 1–3, the results confirmed Hypotheses 4 and 5.

General Discussion

Our primary aims were to investigate the effects of men’s egalitarian versus paternalistic confrontation of sexism on women and to analyze their implications for women’s willingness to confront sexism. We conducted three studies in three different cultural contexts (Spain, Germany, and Mexico) to replicate and test the generalizability of our findings. Beyond some small differences found between studies (see discussion of Study 1 and summary results section of Studies 2 & 3), the results of integrative data analyses (Curran and Hussong 2009) confirmed that male egalitarian confrontation made women report feeling more empowered (Hypothesis 1), happier (Hypothesis 2), and less angry (Hypothesis 3) compared to paternalistic confrontation. The results highlight that men’s confrontation not only affects women’s emotions and attitudes but also indirectly influences their future intention to confront. Interestingly, the results showed two pathways. If men confront sexism for feminist reasons, women report more empowerment and happiness, but only empowerment makes women more willing to engage in sexism confrontation (Hypothesis 4). But if men confront sexism for paternalistic reasons, women experience anger, which increases their interest in confronting as well (Hypothesis 5). Thus, our results suggest that to consider men as genuine allies in fighting inequality, it is important that their actions promote women’s empowerment because increasing women’s happiness does not guarantee their engagement in future sexism confrontation. However, women can also experience anger as a reaction against paternalistic advantaged group members, and this anger may encourage women to confront sexism even more, especially if they identify with being feminist and weakly endorse benevolently sexist beliefs.

Positive Consequences of Egalitarian Confrontation

Male confrontation of sexism may create an anti-sexist atmosphere where men might be seen as allies against sexism (Cihangir et al. 2014). Social support is a key factor in promoting social change (van Zomeren et al. 2004); thus, men’s confrontation of sexism could be interpreted as a form of supportive intergroup contact (Droogendyk et al. 2016b). However, paternalistic or egalitarian motives might drive men’s confrontation (Estevan-Reina et al. 2020), and our results suggest that the motivations underlying advantaged group members’ actions determine the extent to which their actions may be beneficial, but also harmful, for disadvantaged group members. These findings support the need to consider underlying motivations not only when we analyze advantaged group members’ actions against inequality (Estevan-Reina et al. 2020; Radke et al. 2020), but also when we try to understand the impact of these actions on disadvantaged groups. Importantly, men’s egalitarian confrontation of sexism can be as empowering as when women themselves confront sexism, as our results from Mexico show. However, target confrontation made women report more empowerment than men’s egalitarian confrontation in Germany. These results are consistent with literature that shows that women’s confrontation increases their sense of competence, self-esteem, and empowerment (Gervais et al. 2010; Hyers 2007).

The harmful effects of paternalistic confrontation were mostly evident on well-being. When the target confrontation condition was included (Studies 2 and 3), this was the most aversive type of confrontation (i.e., it made women report more anger and less happiness than male confrontation) both in Germany and Mexico. This result is consistent with previous literature which showed the costs of confrontation for targets of prejudice (Kaiser and Miller 2001) and for women in particular (Czopp and Monteith 2003; Dodd et al. 2002; Eliezer and Major 2012; Gervais and Hillard 2014). However, although male confrontation reduces women’s well-being, this does not justify preventing women from confronting sexism themselves, as our results on empowerment show.

The effects of confrontation on empowerment and well-being also depended on women’s feminist identification and endorsement of benevolent sexism. The more women identify as feminists (and the less they endorse benevolently sexist beliefs), the more anger but less happiness they experienced after paternalistic confrontation. Also, the less benevolently sexist they were, the less empowerment they experienced after paternalistic confrontation. Unlike Wiley and Dunne (2019), we did not find that the positive effects of egalitarian confrontation occurred only for strongly feminist-identified women. It is important to notice that, unlike the work by these prior authors, in the current work we do not use the “feminist” label to describe any of the confronters. A man labeled as a feminist who acts in a condescending way (such as offering dependency-oriented help) is not perceived positively by women who are more motivated to challenge gender inequality. This might explain why Wiley and Dunne’s participants viewed feminist men who offered autonomy-oriented help as better allies. Perhaps differences between both works concerning independent variables (sexism confrontation—ours vs. helping behavior—Wiley and Dunne) and dependent variables (empowerment and emotions —ours − vs. perception of allies—Wiley and Dunne) may also explain the different findings. Despite differences, the two works are complementary because they place emphasis on women’s feminist identification to understand both the rejection of male condescending treatment and the acceptance of egalitarian treatment. These results are consistent with the predictions of intergroup helping relations as status relations (Nadler 2002), confirming that highly identified disadvantaged group members may reject dependency-oriented help or seek and accept autonomy-oriented help if they believe that they can succeed by themselves as capable actors.

Empowerment (not Happiness) Encourages Women to Keep Fighting

The positive effects of men’s confrontation on women’s well-being are no guarantee that these will translate into future actions to resist sexism. Literature on prejudice reduction has evidenced positive effects of intergroup contact on attitudes and emotions toward the outgroup on an interpersonal level (Pettigrew and Tropp 2006), whereas collective action literature has shown that this improvement in intergroup relations may undermine social change (Hasan-Aslih et al. 2019; Saguy et al. 2009; Wright and Lubensky 2008). In line with this argument, our results showed that improved happiness after egalitarian confrontation did not increase women’s future intention to confront, whether in Spain, Germany, or Mexico.

However, in the three countries, our results showed that improved empowerment encouraged women to keep fighting against sexism. This result is consistent with literature that points out that advantaged group members’ actions do not undermine social change if they recognize the inequality as illegitimate (Becker et al. 2013), and they can even promote change if they offer disadvantaged group members supportive contact (Droogendyk et al. 2016a). But our study goes one step further in uncovering the underlying mechanism of this positive effect by highlighting the role of empowerment in promoting social change, over and above positive emotions. In a similar line, a very recent work found that satisfying the need for empowerment of disadvantaged groups during intergroup contact is related with their support for social change (Hässler et al. 2020).

Thus, subtyping advantaged group members who show a commitment to fighting inequality as allies (or not) might be a useful strategy to manage positive intergroup relations without undermining social change (Wright and Lubensky 2008). To become allies, advantaged group members must have a genuine interest in improving the status of the disadvantaged group (outgroup focused motivation: Radke et al. 2020; egalitarian motivation: Estevan-Reina et al. 2020) and not override women’s agency, but empower them to keep fighting.

Women’s Resistance to Paternalistic Confrontation

Women are not passive recipients of discrimination (Swim and Hyers 1999), and recent work showed that women oppose men’s actions when these are motivated by paternalistic reasons (Estevan-Reina et al. 2020). Previous research showed that college-educated men try to appear non-prejudiced and progressive, caring, and respectful of women (Lamont 2015), but still many of them may perpetuate inequality when they do not challenge gender power asymmetries in society. Our results are consistent with research showing that even subtle forms of discrimination can trigger resistance responses in women (de Lemus et al. 2018), even when they are not aware of it, if they have internalized egalitarian norms (van Breen et al. 2018). That feminist identification moderates these effects supports this resistance interpretation. The more women identify with feminists, the more anger they reported in response to paternalistic confrontation. This is also in line with findings from the helping relations as power relations model with regard to the idea that highly identified in-group members may reject dependency-oriented help (Nadler 2002). We found the increase in anger after paternalistic confrontation not only in more egalitarian countries (Germany and Spain) but also in less egalitarian ones (Mexico), where support for benevolent sexism is higher (Glick et al. 2000).

When we pooled the datasets, we found that paternalistic (vs. egalitarian) confrontation leads women to express greater future intention to confront via anger. We can interpret these findings as resistance to paternalism. Sexism threatens women’s freedom, and male paternalistic confrontation may strengthen this threat, activating the idea that women cannot stand up for themselves. This reasoning would explain why the women across our studies reported not only feeling more anger after paternalistic rather than egalitarian confrontation, but also that their enhanced anger leads them to express greater future intention to confront to restore their agency.

Limitations and Future Research Directions

The measures used in our work might have triggered responses influenced by task demand characteristics. To address this point, future research could compare egalitarian and paternalistic confrontation with a sexist situation in which there is no confrontation at all, or even with some neutral event like non-sexist bullying, as well as include behavioral measures to increase ecological validity. Adding a control condition would also help us explain women’s resistance toward paternalistic confrontation. Perhaps paternalistic confrontation is still more empowering than no confrontation, or perhaps it is equally annoying. In addition, combining the confrontation motivation (egalitarian vs. paternalistic) with the gender of the source (women vs. men) might contribute to understanding whether both women’s and men’s paternalistic confrontation have the same negative effects on women.

Furthermore, although we collected data in three countries, our college samples are not sufficiently heterogeneous. Furthermore, the sexist situation is always the same (i.e., an episode of street sexual harassment). More diversity in sample composition (in terms of age, political orientation, cultural backgrounds, etc.) and in the scenarios described would contribute to making our findings more solid. In addition, more research would help us to know whether we can generalize our results to other prosocial behaviors beyond confrontation and to other intergroup relations beyond gender inequality.

An interesting direction for future research would be to differentiate group emotions (against the perpetrator of the sexist comment or toward gender inequality itself) and interpersonal emotions (toward the confronter). For instance, it is possible that women experience positive emotions toward egalitarian confronters (interpersonal happiness) and, at the same time, that egalitarian confrontation triggers more anger toward gender inequality (intergroup anger). This possibility may help us to understand why positive cross-group contact in interpersonal relations, if supportive, can contribute to social change.

Future research should also explore whether paternalistic confrontation might have a cumulative effect that makes women perceive the sexist comment not as an isolated act but as a pervasive reality (i.e., a double threat) (Garcia et al. 2009). Furthermore, in our study we did not directly assess the motivations that women attribute to confronters and perhaps women may still doubt the sincerity of advantaged group members’ expressed motivations. In all studies, women perceived both confronters as equally paternalistic (although different in terms of sexism), which suggests that women may not be entirely convinced that the egalitarian confronter is truly egalitarian.

We conceptualized confrontation of discrimination as intergroup behavior that is close to helping behavior; however, confrontation can also be seen as an act of moral courage when it is aimed at restoring a violated moral standard (Halmburger et al. 2015). The two conceptualizations overlap in the case of the “egalitarian confronter,” which is when the confrontation is motivated by moral or equality concerns. In contrast, when the confrontation is motivated by paternalistic concerns, it cannot be seen as a moral courage because it does not aim to address the violated norm (Niesta Kayser et al. 2010). Importantly, behaviors that are considered as morally courageous also involve (potential) risks for those who engage in it (Halmburger et al. 2015). From this perspective, a paternalistic confronter may face less backlash from other advantaged group members because he reaffirms and does not challenge the existing hierarchies. Future research could examine whether women respect an egalitarian confronter more than a paternalistic one because they assume that expressing support for equality is more likely to be punished by other advantaged group members.

Practice Implications

Over the last few years, because of the rise of feminist claims (e.g., #MeToo movement, women’s marches, feminist strikes), the role of men in fighting gender inequality has become a relevant issue. Although men can be involved in change toward gender inequality (Subašić et al. 2018; Wiley et al. 2012), our findings show that not every male confrontation of sexism has positive consequences for women. This information can be useful for policymakers and activists who develop both social interventions and campaigns aimed at involving men in fighting gender inequality. Furthermore, our research can inspire those men who want to become true allies of women to do it in a way that promotes social change. We encourage men to act against sexism and endorse egalitarian (instead of paternalistic) values—that is, to identify the comment as discriminatory (sexist; Cihangir et al. 2014) and illegitimate (unfair; Becker et al. 2013) and to oppose the notion that women are inferior to men (Droogendyk et al. 2016a). In this way, male sexism confrontation will not only make women experience more well-being but also empower them to keep fighting.

Conclusions

The rise of women’s movements for gender equality in the last years has been accompanied by an increase (although still modest) in support by men in this endeavor. However, whereas some men have a real egalitarian motivation, others may be motivated by paternalistic reasons. Our research conveyed that advantaged group members’ actions motivated by genuine egalitarian reasons empower women, which encourages women to keep fighting. However, confrontation motivated by paternalistic reasons makes women feel anger (especially among those who identify more as feminist and endorse less benevolently sexist beliefs), which pushes them to not keep quiet, perhaps as resistance against acts that may still be reinforcing gender hierarchies.

From a theoretical point of view, our research contributes to understanding the impact of confrontation on targets of discrimination in intergroup relations. Following the distinction between dependency- and autonomy-oriented help (Nadler 2002) and positive and supportive contact (Droogendyk et al. 2016a), the distinction between egalitarian and paternalistic confrontation allows the identification of two existing ways of confronting discrimination with different implications for women. Further, beyond sexism confrontation, our current work highlights the role of empowerment and anger as mechanisms to understand in which cases advantaged group members’ actions promote social change or reinforce social hierarchies, at least from the perspective of a disadvantaged group.