Nothing Changes, Really: Why Women Who Break Through the Glass Ceiling End Up Reinforcing It.

Two correlational studies conducted in Switzerland ( N = 222) and Albania ( N = 156) explained the opposition of female managers to gender quotas by examining the origins and consequences of the "Queen Bee (QB)-phenomenon," whereby women who have been successful in male-dominated organizations do not support the advancement of junior women. Results disconfirm previous accounts of the QB-phenomenon as indicating competitiveness among women. Instead, the tendency of women managers to consider themselves as different from other women, and their opposition to gender quotas, emerged when junior women were addressed but not when they considered their direct competitors, other women managers. Personal sacrifices women managers reported having made for career success predicted self-distancing from junior women and opposition to gender quotas targeting these women. We provide a more nuanced picture of what the QB-response is really about, explaining why women managers oppose quotas for junior women, while supporting quotas for women in the same rank.

I’m against gender quotas. I can’t understand why we need to roll out the red carpet for young women and to make their life easy while I made a lot of sacrifices for my career success and I didn’t have such a privilege.

To address gender discrimination and promote the inclusion of women in areas in which they have typically been underrepresented, affirmative action policies such as quotas have been introduced whereby a certain number of positions is reserved for women. Typically, these policies meet considerable opposition, not only from men but, as exemplified by the opening quote taken from a participant in this research project, also from women. Previous research has shown that, compared with women with a low educational level, those with a high educational level are less supportive of affirmative action policies that benefit women as a group (Faniko, 2015; Faniko, Lorenzi-Cioldi, Buschini, & Chatard, 2008, 2012). Similarly, women working at higher levels in male-dominated organizations have been found to deny that gender discrimination exists and to oppose actions that would improve opportunities for junior women (Derks, Van Laar, Ellemers, & De Groot, 2011). We follow up on this initial work, reporting two studies, aiming to examine why women might not support measures that might benefit themselves and other women and under which conditions this is particularly likely to occur.

This contribution has two goals. First, we aim to show that women’s reluctance to support gender quotas (prescribing that 30% of leadership positions in organizations are reserved for women, as has been proposed in the national contexts examined here[1]) is not due to a generic tendency among women to compete with each other. Instead, we expect this to be part of a more general psychological phenomenon documented among some highly successful women. This is referred to as the Queen Bee (QB)-phenomenon. This indicates that female leaders have critical attitudes and distance themselves from junior women to overcome gender stereotypical expectations at work (see Derks, Van Laar, & Ellemers, 2016; Ellemers, Rink, Derks, & Ryan, 2012). Second, we aim to investigate in more detail why exactly women in managerial positions are less supportive of gender quotas. We argue that QB-responses emerge in part due to the personal sacrifices that women have made to overcome gender stereotypical expectations, leading them to see themselves as different from other women at work who have not made similar sacrifices.

Resistance to Affirmative Action Policies

Affirmative action policies such as gender quotas aim at supporting women’s careers and reducing gender imbalance in the workplace. Nevertheless, paradoxically, such measures also tend to be resisted by women for a variety of reasons (for reviews, see Barreto & Ellemers, 2015; Ellemers & Barreto, 2009; 2013). For instance, affirmative action may harm its beneficiaries (for reviews, see Crosby, Sabattini, & Aizawa, 2013; Leslie, Mayer, & Kravitz, 2014) by promoting the stereotype that those who benefit from these policies are not sufficiently qualified (Heilman & Alcott, 2001; Heilman, Simon, & Repper, 1987), or they could not succeed on their own (Sowell, 2004). Other research suggests that women may oppose AAPs because of system-justifying tendencies. That is, women may oppose AAPs because such opposition may help them to promote the notion that their group is not suffering from unequal treatment and their current status is valid and fair (Jost & Banaji, 1994). This may prevent them from taking action, and makes them devalue those who do (Garcia, Schmitt, Branscombe, & Ellemers, 2010).

Thus, we are not the first to note that women might resist affirmative action policies aiming to benefit women. Yet, we address a specific subgroup of women who tend to be seen as important change agents. That is, the women who have been successful and, now, occupy senior positions in the organization are usually expected to play an important role in improving the opportunities for other women. We consider the possibility that this expectation is false, as these women are particularly unlikely to do so, and connect this to the QB-phenomenon. We argue it is important to understand different sources of resistance toward programs aiming to benefit women, as a first step to be able to develop alternative strategies that are more effective in creating truly gender inclusive organizations.

The QB-Phenomenon

The reluctance of some women toward affirmative action resonates with a growing body of research on the so-called QB-phenomenon. This is the tendency of some women who have invested in their personal career success to be more critical toward junior female colleagues, and less inclined to endorse measures that support women as a group (Derks, Ellemers, Van Laar, & de Groot, 2011; Derks, Van Laar, et al., 2011; Ellemers et al., 2012; Ellemers, van den Heuvel, de Gilder, Maass, & Bonvini, 2004; Faniko, Ellemers, & Derks, in press). Previous accounts of this phenomenon (Derks et al., 2016; Derks, Van Laar, Ellemers, & Raghoe, 2015) have described three classes of characteristic behaviors: (a) distancing from junior women, for example, by emphasizing a higher career commitment of the self compared with this subgroup of women; (b) assimilation into the higher status group (in this case, men), for example, by highly masculine self-presentations; and (c) legitimizing the current status quo, for example, by opposing policies that would remedy gender inequalities.

Some have explained this by arguing that there is more criticism and competition between women than between men, and that women will not allow other women to be successful (e.g., Sheppard & Aquino, 2013). By contrast, empirical work provides initial evidence that this response is not specific to women. A recent study revealed that both female and male managers rated their own masculinity as higher than that of same-gender junior colleagues (Faniko et al., in press). Another study among Hindustani workers in the Netherlands revealed similar self-group distancing after exposure to ethnic stereotypes (Derks et al., 2015). A study among senior police women related QB-responses to experimentally induced recall of prior career-related discrimination experiences (Derks, Van Laar, et al., 2011). So, rather than revealing increased competition among women, these results suggest that relevant career experiences of women (and other minorities at work) induce them to distance themselves from others in their group to overcome negative expectations and improve individual opportunities. Thus, the first goal of the current studies is to examine whether the QB-phenomenon reflects a negative attitude toward all women, or only targets a specific subgroup of women. If the QB-phenomenon implies a general sense of competitiveness among women, we should observe a negative attitude toward gender quotas regardless of which subgroup of women will benefit. If, however, the QB-phenomenon stems from a negative attitude of women managers toward more junior women, we should find that they are only less supportive of gender quotas when they target this particular subgroup of women.

Personal Sacrifices for Career Success

There is, by now, converging evidence to suggest that prior career experiences of successful women play a role in the emergence of the QB-phenomenon. Nevertheless, it is as yet unclear what it is about these experiences that make some individually successful women feel different from other women and how this may help explain their resistance against quotas. This is why it is important to further consider the plight of individual women who have had to overcome gender bias to be successful in their career.

Some have argued that women need to display a superior performance at work to achieve the same career outcomes as men. However, empirical support for this position is mixed, if anything (Ellemers et al., 2004). At the same time, the available evidence does suggest that the conditions under which they are expected to perform are less favorable for women than for men. First, individuals who represent an undervalued minority have to contend with implicit performance undermining mechanisms, such as stereotype threat (Betz, Ramsey, & Sekaquaptewa, 2013), whereas their minority position at work communicates that their contributions are likely to be valued less than those of majority group members (Derks, Van Laar, & Ellemers, 2006, 2007b). Second, there is, by now, converging evidence showing that the leadership positions offered to women contain less resources to perform well than those offered to men (Ellemers, 2014;  Ellemers et al., 2012). This implies that women are expected to deliver the same results under less favorable circumstances and have to overcome greater risks of failure than men. Third, the demands of effective leadership behaviors are not naturally compatible with expectations of the way women should behave (Schein, Mueller, Lituchy, & Liu, 1996). As a result, it is more difficult for women than for men to find out how to be effective as a leader in a way that is compatible with one’s gender (Eagly & Carli, 2007). At the same time, there is a relative lack of role models or other senior women who might support them in this process (Derks, Van Laar, & Ellemers, 2007a). Finally, due to the greater demands they have to meet to show an equal performance at work, professional women have to prioritize their work above all else to be successful. However, when they do this, they are seen as less attractive dates (Badgett & Folbre, 2003) because they go against norms expecting them to prioritize family care above work (Folbre, 2012). As a result, women who invest in career advancement less often than men manage to succeed in finding work–life balance (Hochschild & Machung, 2012). All these conditions that tend to make it more difficult for women than for men to succeed in leadership positions imply that on one hand, women who aim to achieve career success typically need to invest more time and (mental) energy in their work; on the other hand, they are more likely to lack important alternative sources of social support that men often do have.

Prior accounts of QB-effects have explained this phenomenon as being primarily driven by the explicit desire of successful women to emphasize—and exaggerate—how they are different from other women, as a way to advance their own career. This motivation to present the self as distinct from the group stereotype thus can be seen as a deliberate strategy to overcome sexist expectations in the workplace by putting down other women. This explanation relies on the assumption that there is competition between individual women in who is most likely to be seen as meeting organizational demands, which should emerge regardless of the position or identity of other women.

To complement this explanation, the present research explores the possibility that less deliberate and strategic considerations may also play a role in the emergence of self-group distancing effects. Such a more cognitive explanation would relate to the—more motivationally neutral—observation of having made considerable sacrifices for one’s career, while having the impression that not all women are willing to make similar sacrifices. Here, self-group distancing should only emerge when considering women who have not made similar sacrifices or seem unwilling to do so. To the extent that self-group distancing stems from the available information about the willingness of the self and others to make personal sacrifices to advance one’s career, the realization that others are unlikely to prioritize their career in the same way is more motivationally neutral. Hence, we refer to this tendency as a more “cognitive” explanation for self-group distancing.

We argue that the level of personal sacrifices successful women feel they have made to overcome gender bias at work is a crucial factor in the emergence of the QB-response. These sacrifices help explain why women managers are likely to see themselves as more committed to their career than other women. If this is the case, then, self-group distancing and opposition to quotas should mainly emerge when comparing the self to more junior women who have not been tested in the same way as they have.

By contrast, less evidence of self-group distancing and opposition toward gender quotas should be observed when women managers compare themselves to other senior women who are likely to have made similar sacrifices for career success. This is why we can expect that women managers would consider other senior women as worthy of support and be favorable toward gender quotas targeting this group of women. However, if the QB-phenomenon reflects a general sense of competitiveness toward other women, then, we should find a more generic pattern of self-group distancing among successful women regardless of whether they compare themselves to junior women or other successful women and regardless of the sacrifices others have made. If anything, successful women then should be even more inclined to distance themselves from women who are equally successful and committed to their career, as these are most relevant as their direct competitors at work and oppose gender quotas for both groups of women.

The Present Research

Our current aim is to investigate whether the QB-phenomenon stems from a general sense of competitiveness toward all women or only reflects a negative attitude toward more junior women. We examined this in two correlational studies performed in different national contexts where we compared women who differ in the career investments they made. This was indicated by their hierarchical position in the organization, as well as the personal sacrifices they reported to have made. We adapted standard indicators of the QB-phenomenon to test their responses to specific subgroups of women. Specifically, in both studies, we compared the extent to which women managers were likely to support or oppose quotas targeting different subgroups of women in the organization. We additionally conducted mediation analyses to test the validity of our reasoning that self-group distancing and reluctance to support affirmative action policies relates to personal career experiences.

Study 1

Method

Participants and design

The sample consisted of 222 women employed in public (52%) and private (48%) sectors in the French part of Switzerland.[2] Participants were contacted at their workplace and were asked to complete an online survey. On average, participants had more than 15 years of work experience. The participants’ mean age was 37.55 (SD = 10.81) and did not correlate with other measures.

We examined support for AAPs in a 2 (participant’s hierarchical position: managerial vs. subordinate) × 3 (target of AAP: self vs. more junior vs. same-level other) between-participants design.[3] To assess the impact of women’s hierarchical position at the workplace, we compared 87 participants who indicated having a managerial position with 135 participants in a subordinate position. The sample size was driven by the size of the population we examined, and more specifically, the number of women managers it contained. The asymmetry in numbers of women managers versus women subordinates examined reflects the reality that there were many fewer women managers than subordinates that we could recruit for this study. Even though the overall number of women subordinates was higher, both women managers and subordinates were randomly distributed in roughly equal proportions across the three experimental conditions, χ2(2, N = 222) = 1.28, p = .53.

Procedure

The questionnaire consisted of a series of items assessing personal sacrifices for career success, self-reported career commitment, perceived career commitment of junior women at an early career stage,[4] self-descriptions in terms of masculinity and femininity, and that of junior women. Then, as an introduction to the experimental manipulation, all the participants first read general information about the intention of several organizations to implement a hiring policy to increase the representation of women in managerial positions in Switzerland. Next, the information provided to participants about the group targeted by the quota varied according to experimental conditions. In the target self condition, the proposed policy would support the respondent to obtain a position that would allow herself to make progress in her career. In the target junior other condition, the policy would support junior female colleagues at an early career stage to make progress in their career. In the target same-level other condition, the policy would support female colleagues at the same hierarchical position as the responded to make progress in their career (see supplementary material for experimental procedure).

Measures

Personal sacrifices for career success

On the basis of interviews conducted with women managers (from which we also recorded the opening quote), we developed a measure to assess personal sacrifices participants had made in different domains (family, personal convictions, vacation) to achieve career success. This resulted in six items (α = .81) such as “I adapted my decision whether or when to have children to the requirements of my career” (see supplementary material for all items used in Study 1).

QB-indicators

To indicate the emergence of the QB-phenomenon, we examined whether women managers described their own career commitment and masculinity as higher than that of junior women.

Career commitment

We announced to the participants that they were first asked to report on their personal career commitment and next on that of junior female colleagues at the beginning of their career. We used six items such as “My career is one of the most important things in my life” (α = .90 for the self, α = .93 for junior women; Ellemers, De Gilder, & Van den Heuvel, 1998).

Masculinity

To assess masculinity, we extracted nine items from Bem’s Sex Role Inventory (Bem, 1974; see also Ellemers et al., 2004) and asked participants to indicate to what degree stereotypically masculine (e.g., defends own beliefs, α = .80) and feminine (e.g., tolerant, α = .79) traits described themselves. Next, participants indicated to what degree the same traits characterized junior women (α = .88 for masculine traits, α = .92 for feminine traits).

Support for gender quotas

After reading about one of the three hiring policies that were proposed (depending on experimental condition), support for the hiring policy that had been described was measured with four items such as “This is a good hiring policy” (α = .86); see Faniko et al. (2012). Agreement with all questionnaire items was solicited on 7-point rating scales (ranging from 1 = strongly disagree to 7 = strongly agree). (See supplementary material for intercorrelations between variables, Table 1.)

Demographic variables

Participants reported their age, educational level, the organizational level they had reached (subordinate vs. managerial position), the number of people they supervised, their profession, the years of work experience, and the sector of their activity. They indicated a variety of different job types and professions in marketing, commerce, retail, banking, consulting, legal services, public services and administration, teaching and education, healthcare, pharmaceutics, and so on.

Results

Consistent with our predictions, the results of ANOVA showed that women managers indicated having made more personal sacrifices for their career success than women subordinates (M = 4.06, SD = 1.57; M = 3.01, SD = 1.33), F(1, 220) = 28.13, p < .001, = .11.

QB-responses

Two repeated measures MANOVAs were conducted to examine the extent to which women managers and subordinates considered themselves as different from junior women. Self- versus Other-ratings of career commitment and masculinity were included as within-participants variable and participant’s hierarchical position (managerial vs. subordinate) as between-participants variable.

As anticipated, participant’s hierarchical position predicted self-distancing from junior women, F(1, 220) = 8.27, p < .01, Wilks’s Λ = .96. Women managers rated themselves as more career-committed than women subordinates rated themselves, F(1, 220) = 14.05, p < .001, = .06 (see Table 1). Furthermore, women managers considered their own career commitment as higher than that of junior women, F(1, 220) = 4.11, p = .04, = .02. Women subordinates considered junior women as more career-committed than themselves, F(1, 220) = 4.28, p = .04, = .02.

Table 1.

Results of MANOVAs Examining Statistical Differences of QB-Responses as Predicted by Women’s Hierarchical Position (Managerial vs. Subordinate), Study 1.

	Women managers	Women subordinates
Self-reported career commitment	4.85a (1.32)	4.17b (1.33)
Perceived career commitment junior women	4.47c (1.29)	4.48c (1.21)
Self-reported masculinity	5.29a (1.11)	4.48b (0.97)
Perceived masculinity of junior women	4.33b (1.39)	4.35b (1.13)

Note. Means with different superscripts differ significantly in a between- and within-participants comparison (ps < .05). All comparisons made with pairwise comparisons tests. QB = Queen Bee.

As anticipated, participant’s hierarchical position predicted self-distancing from junior women, F(1, 220) = 14.27, p < .01, Wilks’s Λ = .94. Women managers described themselves as more masculine than did women subordinates, F(1, 220) = 33.12, p < .001, = .13 (see Table 1). Women managers also considered themselves as different from junior women by rating their own masculinity as higher than that of this subgroup of women, F(1, 220) = 31.03, p < .001, = .12. No significant difference between self-ratings and other ratings was observed among women in subordinate positions, F(1, 220) = .82, p = .37, = .003.[5]

Support for gender quotas that had been described to participants was examined with a 2 (participant’s hierarchical position: managerial vs. subordinate position) × 3 (target of AAP: self vs. junior vs. same-level other) ANOVA. This only revealed a significant two-way interaction,[6] F(2, 216) = 5.33, p < .01, = .05, which we examined with post hoc contrasts (see Table 2). Women managers were less supportive of quotas that would target junior women than were women subordinates, F(1, 216) = 4.95, p = .03, = .02. However, women managers were more supportive of quotas that would target someone in the same position as themselves than were women subordinates, F(1, 216) = 4.89, p = .03, = .02. Finally, women managers and subordinates showed equal levels of support for quotas that might benefit their own career, F(1, 216) = 1.26, p = .26, = .006. We also established that the tendency to show more support for different quotas described—depending on the target of this policy—was only visible among women managers. They were significantly more inclined to support quotas targeting someone at the same level as the self than quotas targeting junior women or the self (see supplementary material for more details, Table 2).

Table 2.

Results of ANOVA Examining Statistical Differences of Support Toward Gender Quotas as Predicted by Women’s Hierarchical Position (Managerial vs. Subordinate) and Target of Quotas (Self vs. Junior vs. Same Level Women), Study 1.

	Women managers	Women subordinates
Target of quotas
Self	4.09a,b (1.62)	4.47a (1.35)
Junior women	3.50b (1.89)	4.35a (1.57)
Women at the same level	4.94c (1.47)	4.12a (1.37)

Note. Means with different superscripts differ significantly in a between and within-participants comparison (ps < .05). All comparisons made with pairwise comparisons tests.

Model testing

We hypothesized that compared with women subordinates, those managers would have made more personal sacrifices for their career, and this would be related to their self-group distancing. As a result, compared with women subordinate, they should be less supportive of junior women. We additionally argued that the reluctance to support quotas should be particularly visible for quotas targeting junior women but emerge less for quotas targeting women at the same organizational level. To simultaneously test the multiple mediations, as well as the moderating effect of the target of the quotas proposed, we constructed a path model in AMOS. To this end, self-group distancing scores were calculated by subtracting the perceived career commitment and masculinity of junior women from self-rated career commitment and masculinity (i.e., higher values indicate more self-group distancing).[7]

The hypothesized model showed good fit to the data with χ2(df = 12, N = 222) = 7.958, p = .788, root mean square error of approximation (RMSEA) = .00, 90% confidence interval (CI) = [.000, .046], Akaike information criterion (AIC) = 73.958. Further testing of model parameters revealed evidence in line with predictions (see Figure 1). The relation between participant’s hierarchical position and self-group distancing in terms of career commitment was fully mediated by personal sacrifices made; personal sacrifices partially mediated the relation between hierarchical position and self-group distancing in perceived masculinity. Furthermore, as predicted, the differential support shown by women in managerial versus subordinate positions for quotas targeting more junior women depended on their personal sacrifices for career success, which accounted for self-group distancing in terms of career commitment. Personal career experiences or self-group distancing did not explain differential levels of support for quotas targeting women at the same organizational level or quotas that would benefit the self.

Figure 1.

Results of the hypothesized model predicting support for gender quotas of different targets (self vs. junior vs. same-level women).

Note. Only continuous paths are significant.

*Subordinate position coded 1, managerial position coded 2. p < .05.

**Difference score of self-reported career commitment–perceived career commitment of junior women. p < .01.

***Difference score of self-reported masculinity–perceived masculinity of junior women. p < .001.

We then examined four alternative causal order models (see supplementary material for alternatives models, Table 3). We first tested a model in which all paths were constrained to be equal across the three quota conditions (Alternative Model 1). The chi-square test showed that this alternative model has inferior fit in comparison with our hypothesized model, χ2(df = 8, N = 222) = 29.26, p < .001. Second, we addressed the possibility that participant’s hierarchical position predicts self-group distancing, that self-group distancing predicts personal sacrifices, and that personal sacrifices predict quota support (Alternative Model 2). The chi-square test showed that this alternative model fit the data less well than our hypothesized model, χ2(df = 10, N = 222) = 45.75, p < .001. Then, because of the correlations observed, we tested whether participant’s hierarchical position predicts self-group distancing, which in turn predicts quota support (Alternative Model 3). This model does not show acceptable fit, and fits the data significantly less well than our hypothesized model, χ2(df = 5, N = 222) = 53.27, p < .001). Finally, we examined a reversed causal order model to address the possibility that participant’s hierarchical position predicts quota support, which predicts personal sacrifices, whereas personal sacrifices predict self-group distancing (Alternative Model 4). This model does not show acceptable fit, and fits the data significantly less well than our hypothesized model, χ2(df = 9, N = 222) = 52.86, p < .001. Thus, our hypothesized path model emerges as the most parsimonious model that best fits the observed data.

Discussion

This study provided correlational evidence that participants’ own career experiences—rather than a more general sense of competitiveness toward other women—relate to self-group distancing tendencies, as well as to the reluctance to support gender quotas. That is, the inclination to see themselves as different from junior women was accounted for by the personal sacrifices for career success. Indeed, self-group distancing among women managers specifically emerged in their perceptions of career commitment and masculinity of junior women. Likewise, the reluctance to support quotas—which was mediated by personal sacrifices made and self-distancing in terms of career commitment—only emerged when these policies targeted more junior women in the organization. Women managers were quite happy to support quotas targeting women at the same hierarchical level in the organization, even though these would be their direct competitors. Thus, this study helps rule out that QB-responses reflect more general competitiveness or the operation of self-interested motives.

Nevertheless, this study suffers from some limitations, which we aim to address in a second correlational study. In Study 1, we decided to focus the self-group distancing measure on the tendency of participants to see themselves as different from more junior women in the organization. On one hand, this most accurately reflects our reasoning, as personal career experiences can explain why senior women perceive themselves as different from junior women who have not undergone similar experiences. However, to be able to draw this conclusion, it is important to examine whether this self-group distancing is indeed limited to junior women and does not occur to the same extent with regard to women at the same organizational level. Therefore, in Study 2, we also measured the degree to which women showed self-group distancing responses toward women at the same organizational level.

Study 2 also further examined the reasons why women managers would be more willing to support other women at the same organizational level. For this purpose, we added some further measures to explore the viability of different considerations that possibly play a role. Specifically, we assessed whether support for quotas might relate to participants’ perceptions of whether beneficiaries of these measures have made sacrifices for career success and/or relate to concern about being in competition with beneficiaries of these policies.

Study 2

One hundred fifty-six women (Mage = 32.12, SD = 8.82) employed in the public sector in Tirana (Albania) took part in this study; age did not correlate with other measures. Participants were contacted at their workplace and were asked to complete an online survey. The participants had on average more than 10 years of work experience.

Participants were randomly assigned to one of three experimental conditions, in which the nature of the hiring policy that was described to them differed. Specifically, participants were asked whether they would support quotas targeting the self, other women in more junior positions at the organization, or other women at the same level as the self. As in Study 1, the sample size—and the asymmetry in women managers versus subordinates—was driven by the characteristics and size of the population we examined, and resulted in 57 women managers and 99 women subordinates to be distributed in similar proportions across the three experimental conditions, χ2(2, N = 156) = .86, p = .65.[8]

The questionnaire was similar to the one used in Study 1. In addition, in the first part of the questionnaire, we included a series of items assessing the perceived career commitment of women at the same hierarchical level.

Subsequently, participants received information to introduce different types of hiring policies, depending on experimental condition identical to the procedure used in Study 1. After reading these instructions, participants were asked to report their opinions about the beneficiaries and support for the hiring procedure.

As in Study 1, respondents were asked to indicate their age, educational level, the organizational level they had reached (subordinate vs. managerial position), profession pursued in the public sector, and years of work experience. They indicated different job types and professions in public services and administration, engineering, teaching and education, healthcare, and so on.

We assessed personal sacrifices for career success with the same six items that were used in Study 1 (α = .87).

In this study, we focused on career commitment and observed whether women managers described their own career commitment as higher than that of other junior women and that of women in the same hierarchical positions.

We used the four items as in Study 1 to examine participants’ self-reported career commitment (α = .90). In addition, to examine whether self-group distancing was specific to junior women, we asked participants to report their perceptions of the career commitment of junior female colleagues (α = .93) and those at the same hierarchical level (α = .94).

Participants’ support for the hiring policy was assessed with the same four items that were used in Study 1 (α = .87).

Possible explanations for reduced quotas support

We assessed two possible explanations for the differential support for gender quotas targeting specific groups of women. First, we asked participants with three items whether they thought the beneficiaries of these measures had made personal sacrifices for career success (e.g., “These women have shown that they are willing to make sacrifices for their career,” α = .82). Second, we examined participants’ perceptions of being in direct competition with beneficiaries of the proposed policy with two items (e.g., “I have to compete with these women for future promotions,” r = .34, p < .001). (See supplementary material for intercorrelations between variables, Table 4.)

As in Study 1, women managers (M = 4.99, SD = 1.41) reported to have made more personal sacrifices for their career success than women subordinates (M = 3.53, SD = 1.46), F(1, 154) = 36.59, p < .001, = .19.

We conducted a repeated measure MANOVAs to examine the extent to which women managers and subordinates considered their career commitment as different from that of junior women and women at the same hierarchical level. As expected, participant’s hierarchical position predicted self-distancing from junior women, F(1, 154) = 12.86, p < .001, Wilks’s Λ = .92. Women managers considered themselves to be more career-committed than women subordinates did, F(1, 154) = 6.74, p = .01, = .04 (see Table 3). Women managers also considered their career commitment as higher than that of junior women, F(1, 154) = 7.63, p < .005, = .05. Conversely, as in Study 1, women subordinates considered junior women as more career-committed than themselves, F(1, 154) = 5.26, p = .02, = .03. Both groups of women did not consider themselves as different from other women holding the same rank; managers, F(1, 154) = .02, p = .89, = .00; subordinates, F(1, 154) = 1.37, p = .24, = .01.

Table 3.

Results of MANOVAs Examining Statistical Differences of QB-Responses as Predicted by Women’s Hierarchical Position (Managerial vs. Subordinate), Study 2.

	Women managers	Women subordinates
Self-reported career commitment	5.07a (1.35)	4.41b (1.65)
Perceived career commitment of junior women	4.42c (1.50)	4.82c (1.46)
Perceived career commitment of women at the same level	5.05a (1.31)	4.58b (1.54)

Note. Means with different superscripts differ significantly in a between and within-participants comparison (ps < .05). QB = Queen Bee.

ANOVA revealed the predicted interaction between the participant’s hierarchical position and the target of gender quotas, F(2, 150) = 6.84, p < .005, = .08.[9] Replicating the results of Study 1, compared with women subordinates, women managers were less supportive of quotas targeting junior women, F(1, 150) = 10.44, p < .005, = .07 (see Table 4). Furthermore, women managers were more supportive of quotas targeting other women at the same hierarchical level as the self than were women subordinates, F(1, 150) = 3.89, p = .05, = .03. Finally, both women managers and subordinates did not indicate different levels of support for quotas which might benefit their personal career, F(1, 150) = .57, p = .45, = .04. The tendency to show different levels of support for the quota policy described—depending on the target of this policy —was observed among two groups of women (see supplementary material for more details, Table 5). Women managers expressed similar level of support toward gender quotas regardless of their target. However, they were significantly less inclined to support quotas targeting junior women, rather than targeting women in the same position as the self. Women subordinates expressed similar level of support toward gender quotas that might benefit their own career and that of junior women. They were significantly more inclined to support quotas targeting junior women, rather than quota targeting women in the same position.

Table 4.

Results of ANOVA Examining Statistical Differences of Support Toward Gender Quotas as Predicted by Women’s Hierarchical Position (Managerial vs. Subordinate) and Target of Quotas (Self vs. Junior vs. Same-Level Women), Study 2.

	Women managers	Women subordinates
Target of quota
Self	5.06a,b (1.60)	5.37a,c (1.21)
Junior women	4.44b (1.83)	5.74a (1.26)
Women at the same level	5.51a (1.15)	4.75c (1.42)

Note. Means with different superscripts differ significantly in a between and within-participants comparison (ps < .05).

We used the same statistical models and procedures as in Study 1, to examine whether our proposed relations between the different model variables fit the observed data. The hypothesized model (see Figure 2) showed good fit to the data with χ2(df = 6, N = 156) = 7.23, p = .30, RMSEA = .04, CI = [.00; 12], AIC = 55.23. Further examination of indirect relations as specified in this model revealed evidence consistent with our predictions and the results of Study 1 (see Figure 2). That is, the relation between participant’s hierarchical position and self-distancing in terms of career commitment was mediated by personal sacrifices, and self-group distancing in terms of career commitment in turn predicts support for the hiring policy targeting more junior women. The willingness to support the policy targeting women at the same level as the self directly relates to participant’s hierarchical position, and is not mediated by personal sacrifices or self-group distancing tendency in terms of career commitment. Personal sacrifices and self-group distancing in terms of career commitment also do not explain support for policies benefiting the self.

Figure 2.

Results of the hypothesized model predicting support for gender quotas of different targets (self vs. junior vs. same-level women).

Note. Only continuous paths are significant.

*Subordinate position coded 1, managerial position coded 2. p < .05.

**Difference score of rated personal career commitment–junior women career commitment. p < .01.

***Difference score of self-reported masculinity–perceived masculinity of junior women. p < .001.

As in Study 1, we also tested the fit of the hypothesized model against four alternative models (see supplementary material for alternatives models, Table 6). The first model does not fit the data, and shows significantly less good fit than our hypothesized model, χ2(df = 6, N = 156) = 22.20, p < .001 (Alternative Model 110). The second alternative model does not show good fit to the data and fits the data significantly less well than our hypothesized model, χ2(df = 7, N = 156) = 37.40, p < .001 (Alternative Model 2). The third alternative model also does not show acceptable fit and fits the data significantly less well than our hypothesized model, χ2(df = 4, N = 156) = 23.65, p < .001 (Alternative Model 3). Forth, the reversed causal order model also does not fit the data and shows significantly less good fit than our hypothesized model, χ2(df = 5, N = 156) = 39.94, p < .001 (Alternative Model 4). Thus, the hypothesized model seems to offer the most parsimonious way to represent the relations between these different variables.

Additional measures

To explore why women managers were less supportive of quotas when these targeted junior women than when targeted women in the same level, we explored two possible explanations: (a) whether they considered that the beneficiaries of gender quotas have made personal sacrifices for career success and (b) whether they felt they were in competition with the beneficiaries. For these measures, we focused on the two experimental conditions offering preferential treatment to other women and conducted a 2 (participant’s hierarchical position) × 2 (quotas target: junior women vs. other women at the same level as the self) between-participants ANOVAs. These analyses only revealed significant effects for the perceived sacrifices for career success of quotas’ beneficiaries, as evidenced by a reliable two-way interaction, F(1, 102) = 12.01, p < .005, = .11.[11] The relevant means revealed that women managers saw quotas’ beneficiaries at the same level as the self as more willing to make sacrifices for career success (M = 5.55, SD = .92) than junior female beneficiaries (M = 4.35, SD = 1.96), F(1, 102) = 8.04, p < .01, = .07. By contrast, women subordinates perceived junior beneficiaries as more willing to make sacrifices (M = 5.44, SD = 1.14) than beneficiaries at the same level as the self (M = 4.78, SD = 1.31), F(1, 102) = 3.99, p = .05, = .04.[12]

We then explored whether perceived sacrifices for career success of quotas’ beneficiaries might mediate the differential willingness of women in managerial versus subordinate positions to support a preferential hiring policy targeting women at the same organizational level. We followed the procedure advocated by Preacher and Hayes (2004), with bootstrap estimates based on 2,000 bootstrap samples. The direct significant effect of hierarchical position on policy support was reduced to nonsignificance, b = .22, SE = .18, t(53) = 1.20, p = .23, after including perceived sacrifices for career success of quotas’ beneficiaries as a mediator, with the bias-corrected confidence interval not including zero (b = .16, 90% CI = [.01, .39]). Thus, the results of these additional measures help us exclude the possibility that a general sense of increased competitiveness toward other women explains the reluctance of those managers to support AAPs. Instead, they suggest that women at high (vs. low) hierarchical levels adapt their willingness to support quotas of specific target groups depending on how willing to make sacrifices for career success they consider this particular target group to be. These results are fully in line with our reasoning and offer additional support for our argument explaining reluctance to support AAPs as stemming from the tendency to emphasize differences between the self and more junior women in particular.

In Study 2, we replicated the main findings of Study 1, offering conclusive evidence for our reasoning that the QB-phenomenon should not be seen as indicating an overall increase of competiveness toward other women. That is, our correlational findings again established that whereas women managers are reluctant to endorse quotas of more junior women, they are quite supportive of women at the same organizational level. Likewise, the self-group distancing they show (in terms of career commitment) is directed at junior but not senior women. Furthermore, their reluctance to support junior women is accounted for by this self-group distancing tendency. We additionally showed—as in Study 1—that personal career experiences play a crucial role in this process. That is, participants’ perceptions of the personal sacrifices they made to achieve career success explained self-distancing tendencies from junior women, which in turn negatively predicted their support for AAP targeting these women.

Crucially, these results show no evidence of a generalized sense of competitiveness toward other women, as women managers did not display self-group distancing in relation to other women at the same organizational level. Indeed, they were more likely than women subordinates to support preferential treatment of other women at the same organizational level because they considered these women to have made similar sacrifices for career success.

Consistent with these results, participants’ hierarchical position or personal career experiences were unrelated to their tendency to support quotas that they might benefit from themselves. These findings speak against an account in terms of self-interested competition to explain the reluctance to support junior women, and rather is in line with our analysis that explains QB-responses as coping with gender stereotypes.

General Discussion

Whereas previous studies focuses on attitudes of women and men toward AAPs, this set of studies focuses more specifically on attitudes of women in different hierarchical positions and documents an additional source of resistance toward AAP’s among successful women. Even though these women tend to be seen as important change agents, we have now showed this is not necessarily the case, as they are rarely particularly supportive to the career of junior women. The studies reported here help explain why these women manifest these behaviors and clarify how this can affect the career opportunities of other women.

Across two correlational studies, we examined whether the opposition of women managers toward quotas is related to self-group distancing tendencies that have been documented as the QB-phenomenon, instead of indicating overall competitiveness among women. To address this, we compared responses of women managers with those of women subordinates. We found that women managers distanced themselves from junior women and were reluctant to support policies that would improve professional opportunities for junior women. By contrast, women managers did not distance themselves from women at the same rank and were supportive of quotas that would benefit women they would be directly competing with. These results showed evidence that QB-phenomenon does not reflect increased competitiveness among women managers and QBs do not perceive themselves as different from all women.

These results also help explain why successful women start distancing themselves from other women. We observed that women managers reported having made personal sacrifices to achieve career success, and realizing this was the case, made them see themselves as different from junior women. We think it is unlikely that this reflects an overall bias of senior professionals against junior colleagues because in another data-set where we examined men as well as women, we found that for senior men, the sacrifices made for career success do not relate to the way they perceive junior men (Faniko, Ellemers, & Derks, in preparation).

The current investigation shows an additional—more cognitive—route to the QB-phenomenon. We consider the observation of differences between the self and other women in personal sacrifices made for career success as a relatively “neutral” observation indicating how perceptions of actual differences between one’s own situation and that of other women may offer a “cognitive” explanation for the QB-phenomenon. As exemplified by the opening quote recorded from one participant in this study, and supported by our quantitative findings, the way senior women perceive junior women and react toward quotas targeting this group of women may also reflect motivation for fairness. That is, some senior women who have made difficult choices on their way up the organizational ladder may expect that junior women should make similar sacrifices for their career success. This does not diminish our conclusion that organizations in which women face less difficulties in realizing their career ambitions should be less likely to elicit QB-responses.

To the extent that these observed differences between the self and junior women reflect actual barriers that women have to overcome to be successful—and result from a process in which only those women who are willing to make additional sacrifices can be successful—this also points to external factors in the organization that contribute to the QB-effect. These findings defuse the common understanding that QB-effects simply reflect the general “cattiness” of women who compete with each other at work—in this sense, countering a purely motivational account.

Results from Study 2 add to this picture by showing that women managers support measures aiming to help women at their own organizational level because they feel that these women have also made sacrifices for career success. This finding helps to rule out the possibility that senior women are favorable toward quota policies targeting women at their level because they consider quota policies as a way to discredit other women, as these policies may help build the case that other women cannot succeed on their own merit. Not only this alternative interpretation is refuted by our mediation analysis, but previous work has also showed that women managers identify with women who are similar to them in terms of career success (Faniko et al., in press).

Our findings resonate with system justification theory (Jost & Banaji, 1994; Kay et al., 2009), but at the same time, extend this explanation. That is, although we document that women can resist measures aiming to enhance gender equality, we also observe distinct response patterns for specific subgroups of women and relate these responses to their personal career experiences. That is, we show that women managers who have stereotypical views of junior women and oppose policies supporting this group contribute to maintaining the status quo in which men hold the majority of managerial positions and their own position is exceptional. In addition, women subordinates may tend to internalize their low position and blame themselves for the lack of career advancement by considering themselves as less career-committed than those women who have been successful. Our research sheds more light on the origin of these tendencies by showing that the career experiences of senior women—in which they felt compelled to make difficult choices for career success—contribute to these effects.

Practical Implications

These results have clear practical implications, as they reveal that the tendency of successful women to resist AAPs stems from their own career experiences. The current results indicate that organizations can help reduce the self-group distancing tendencies and opposition toward gender quotas for junior women. They can do this by promoting a work environment where women are less inclined to think they have to make difficult life choices to achieve career success. For instance, supervisors can make a strong commitment to providing a family-friendly work environment. This has been documented as an effective strategy to enhance the well-being, health and productivity of female (as well as male) workers over time (Van Steenbergen & Ellemers, 2009). Furthermore, our results suggest that women managers were supportive of gender quotas intending to benefit the career of other women at the same hierarchical level. This was the case because these colleagues were perceived as similar to the self in terms of career commitment and having made difficult choices for career success. This suggests that organizations might enhance support for programs benefiting junior women by emphasizing the fact that junior women are generally quite committed to their career.

Limitations and Future Directions

This research is not without limitations. An obvious limitation is that these studies rely on correlational data, which makes it difficult to determine causality. Research based on experimental designs is currently under way in our research group to address this limitation and cross-validate the mechanisms observed in this correlational research. Given the mediating role of personal sacrifice for career success—found in Study 1 and 2—we anticipate that when women managers are invited to think about situations in which they have made personal sacrifices for their career, they should manifest more self-group distancing and less support for gender quotas targeting junior women than when they are asked to think about situations in which they are supported by others for their career success. An experimental design will also allow us to follow up on the results of Study 2 showing that the willingness of women managers to support gender quotas for women at the same level depended on how willing these colleagues were to make sacrifices for career success. To further examine the causality of this relation, we are examining whether women managers show less QB-responses and more support for gender quotas targeting junior women, after we emphasize the personal sacrifices for career success made by junior women.

A second limitation is the absence of counterbalancing the order of the measures of self-group distancing and perceived personal sacrifices for career success with the measure of support for the gender quotas. In this research, we always assessed self-group distancing and perceived personal sacrifices for career success before assessing support for gender quotas with different targets (self, junior women, and women at the same level). In analyzing these results, we have shown that our proposed explanation better fits these data than alternative models. Yet, we cannot rule out that the introduction of an explicit comparison between self and other women in the initial measures affected the willingness to support these different types of gender quotas. Future research might address this possibility by counterbalancing the order in which self-group distancing and support for quotas are assessed.

A third limitation of the present research is that we only assessed how research participants responded to junior women, without comparing these to their responses to junior men. This is a limitation we addressed in follow-up research. Specifically, we compared responses of senior men and women to junior male and female academics (Faniko et al., in preparation). Results of this additional study are in line with our interpretation of the current results. That is, senior women estimated the commitment of female junior academics to their scientific career to be less than the career commitment of male junior academics. This supports the general notion that senior women in male-dominated organizations are more biased against junior women than against junior men.