Although there has been a development toward gender equality in the last decades with women making around 50 % of the workforce and attaining equal or higher educational degrees (e.g., Catalyst 2012, 2014; Eurostat 2013; Rubery 2002) in western societies, they are still underrepresented in management positions, especially in higher management positions in business and politics (Catalyst 2012, 2014; European Commission 2013; United Nations 2010). In the United States, for instance, 46.9 % of the workforce is female, but women hold only the 14.6 % of the Fortune 500 executive officer positions (Catalyst 2014), a pattern that is seen in other western societies. In Germany, women compose circa 46 % of the overall workforce, but only 12.9 % of board directorships, while as in the UK, female representation among the workforce is 46 %, but only 10.7 % among the board directorships (Catalyst 2012).

One might be able to explain this pattern by combining and comparing expectations for gender and leadership roles. Gender roles are socially shared beliefs about male and female attributes (Biddle 1979), which consist of descriptive and prescriptive components. The former focuses on what women and men actually do, and the latter includes norms of appropriate conduct for women and men, i.e., how women and men should behave (Cialdini and Trost 1998; Eagly and Karau 2002; Glick and Fiske 1996). As Eagly (1987, p. 13) stated, “Many of these expectations are normative in the sense that they describe qualities or behavioral tendencies believed to be desirable for each gender.” Through socialization, gender roles are integrated into individuals’ self-concept and personalities (Feingold 1994), conforming an individuals’ self-standards, preferences, and influencing their behavior (Eagly 1987). Hence, people try to act in line with their gender roles and are motivated to fulfill their roles, respectively (Diekman and Eagly 2008; Eagly et al. 2000; Evans and Diekman 2009). The guiding principle behind this process is a maximization of utilities in social interactions (Eagly and Wood 1999), as role-congruent behavior generally goes along with positive affect, enhanced self-esteem and positive reactions by interaction partners (Guerrero Witt and Wood 2010; Wood et al. 1997). Role-incongruent actions, on the other hand, are socially sanctioned by disapproval (Eagly and Karau 2002; Diekman and Eagly 2008). Although one might think that gender roles are changing, recent research shows that gender role expectations for females and males still differ considerably (Bosak and Sczesny 2011; Diekman et al. 2005). Women are still more associated with being concerned about the wellbeing of others and thereby with communal attributes such as being supportive, gentle, empathetic, and caring, whereas men are still more associated with agentic attributes such as being assertive, controlling, dominant, and competitive (e.g., Bakan 1966; Deaux and Lewis 1983; Williams and Best 1990).

On the other hand, leader stereotypes are—as already stated in Schein’s “think manager–think male” phenomenon (2001; Schein et al. 1996)—mainly associated with agentic characteristics (Koenig et al. 2011). Therefore, there is an incongruence between the female gender role (communal) and the perceived requirements for a leader (agentic; Eagly and Karau 2002; Heilman 1983; Heilman et al. 1995; Lyness and Heilman 2006). As stated in role congruity theory (Eagly and Karau 2002), this incongruity leads to two forms of prejudice: (1) women are perceived as less suited for leadership roles and—if women comply with the leader role expectations—they are (2) evaluated less favorably than male leaders acting in the same way (Eagly and Karau 2002). Thus, female leaders are faced with a double standard (when they want to be perceived as competent, women have to perform better than their male counterparts do) and a double bind (they have to be tough and nice at the same time; Eagly and Karau 2002).

However, in the last years, the emergence of uplifting forms of leadership is shifting the traditional leader-centric and influence-based focus into a follower-centered and growth-oriented perspective (Avolio 2007), where authentic (Gardner et al. 2005), ethical (Brown et al. 2005), or servant (Liden et al. 2008) leaders elevate followers into higher states of personal and professional development. From all these emerging leadership styles, “arguably the most developed theory within this category of ‘value-driven and uplifting’ leadership is authentic leadership theory” (Hernandez Bark et al. 2011, p. 1174). On the one hand, while directive or autocratic leadership behaviors are more congruent with male gender role expectations (Eagly et al. 1992), one distinctive characteristic of authentic leaders is that they are highly sensible to followers’ developmental needs, and can easily adjust their behaviors to fulfill them (Leroy et al. 2012). This seems congruent with the implicit nurturing nature associated with the female gender role (Williams and Best 1990).

Further, a leader’s prototypicality, meaning that the leader represents the main characteristics of his or her group, might reduce the prejudice toward female leaders’ evaluation as their followers are more tolerant to leaders’ shortcomings (Giessner and van Knippenberg 2008; Ullrich et al. 2009). Thus, if female leaders are high prototypical and represent the main characteristics of the group, this might reduce the prejudice against female leaders, as prototypical leaders are liked and trusted more and successes are attributed to them (van Knippenberg and Hogg 2003). Hence, prototypical female leaders’ benefit of being prototypical might counterbalance the prejudice associated with the incongruence between the female gender role and the leader role. However, our literature review shows that up to now the relation between biological gender and authentic leadership has not been prominent in the focus of research, even though it might have relevant implications for gender equality. The same is true for prototypicality and biological gender in the workplace.

Eagly (2005) makes a case that being either an authentic or prototypical leader may be not enough to breach the gender-leadership divide. She agrees with authentic leadership scholars that before leaders can influence followers identification and its associated outcomes (e.g., trust and commitment; Avolio et al. 2004), they must first earn their followers’ acceptance and respect, for example, by being true representatives of the groups they lead. In other words, if leaders want to be taken as exemplary role models at the workplace, they may need to be perceived as prototypical and act in favor of their collectives (e.g., their organizations). Unfortunately, neither authentic leadership nor social identity theories directly address how authentic leaders identify with their organization, or how those with an “outsider” status, such as women in a male-oriented role, can become prototypical.

Therefore, examining how both a leader’s authenticity and prototypicality combined influence their organizational identification (OID; van Dick 2001) might reveal important insights for organizations and (female) managers. Besides the many positive work outcomes for organization associated to highly identified managers (e.g., reduced turnover, higher work commitment, satisfaction or extra-role behaviors; Riketta 2005), we believe that authenticity and prototypicality not only empower these identified managers to develop their followers authenticity but also re-shape their followers’ leader role expectations to be more congruent with female role expectation. Our rationale being the fact that prototypicality can be enhanced by leaders themselves by becoming entrepreneurs of identity (Reicher et al. 2005) or by exemplary role modeling (Gardner et al. 2005), a characteristic of authentic leaders by which female managers not only can become “one of us”, but also the “best of us”.

In consequence, in this study, we contribute to move the gender-leader congruency discussion forward by proposing and testing a model that considers a synergistic effect of both authentic leadership and prototypicality over a leader’s OID.

In consequence, the goals of this study were (1) exploring if biological gender differences exist in authentic leadership behaviors and if the latter acts as a mediating mechanism that leads to higher leader OID and (2) testing if being prototypical could reduce gender-leader role incongruency, fostering higher authentic leadership in female managers.

Theoretical Background

On the one hand, some authentic leadership behaviors seem to be congruent with the female gender role. For example, self-awareness, one key component of authentic leadership, implies being sensible to followers’ developmental needs. Further, highly integrated behaviors of their authentic functioning, such as establishing and maintaining transparent, growth-enhancing relations, enable leaders to attend to their followers’ developmental needs (Leroy et al. 2012). These behaviors are congruent with the female gender role’s nurturing connotation (Williams and Best 1990). Therefore, authentic leadership—similar to transformational leadership—holds the potential for women to reduce the incongruity of the expectations toward them as women and as leader (Eagly and Karau 2002; Eagly et al. 2003). Thus, authentic leadership might enable women to cope with the double bind, becoming an ideal leadership style for female managers. On the other hand, one might argue that due to the still existing incongruity of the female gender role and leader role (Eagly and Karau 2002; Koenig et al. 2011), it might be harder for female leaders to display relational authenticity in the first place (Eagly 2005). Female managers are faced with two confronting role expectations: The expectation toward their behavior as a woman is to be nice, nurturing, etc., whereas the expectation toward them as a leader is to be agentic, dominant, and assertive when needed. Additionally, as Eagly (2005) highlights in her paper

Because authenticity emerges in the transactions between leaders and followers, followers must identify with their leader and perceive the leader’s values as suitable for the community within which the leader has authority. It is not sufficient that the followers become aware of a leader’s deep value commitments. They must also trust that these values will serve the community in which they are joined to the leader. Moreover, even when outsider leaders advocate consensual values, they may find it difficult to garner support for their agenda if they are not perceived as appropriate spokespersons for the community. (p. 463).

Further, in those organizations where female managers are scarce, their biological gender is more salient to their followers and co-workers. As individuals use categories like gender to reduce the complexity in information processing, tend to ignore incongruent information, and even show biased memory regarding incongruent information (Stangor 2000), even if female leaders in such environments endorse the values of the community, it should be more difficult for women to gain their employees’ trust and identification (Eagly 2005). Both perspectives seem to be reasonable at the first glance, but this important topic has not received much attention in leadership research.

Authentic leaders excel in two self-based psychological mechanisms, self-awareness and self-regulation, expressing a highly integrated form of behavioral regulation in their leadership role (Gardner et al. 2005). Self-awareness indicates an awareness of goals, emotions, and needs of self and others. In turn, self-regulation implies a balanced processing of information (e.g., considering different viewpoints before making decisions), relational transparency, by establishing open and clear relations with others and maintaining an internalized moral perspective by acting coherently with their inner values even in adverse contexts. At the workplace, based on the female gender role, women are expected to show concern for others by (1) being highly aware of their needs and values (self-awareness), (2) to be relationship-oriented and developing open relations with others, and (3) to be emphatic and to consider different viewpoints (balanced processing of information), but as managers they are expected to be more agentic and act coherently with inner values (internalized moral perspective dimension of authentic leadership) even in adverse contexts. Because of this role conflict, female managers should attribute their self-awareness, balanced processing of information, and relational transparency to their gender role and not to their leadership role, perceiving themselves less authentic as leaders. On the other hand, if men display these communal behaviors, they should attribute them to their leadership style and not their gender role, perceiving themselves as authentic leaders. Thus, we assume that

Hypothesis 1

Women will report less authentic leadership behaviors than men.

Authentic Leadership and Leader’s OID

When leaders achieve an integrated form of regulation, they usually become exemplary role models to their followers, who tend to identify with them and imitate their behaviors (Avolio et al. 2004; Gardner et al. 2005). While a recent study found that this process mediates the link between authentic leadership and followers’ social (organizational) identification (Wong et al. 2010), we still do not know what drives authentic leaders to identify with their organizations in the first place.

From a relational authenticity approach (Eagly 2005; Ilies et al. 2005; Spitzmuller and Ilies 2010), we propose that exemplary role modeling not only drives followers’ identification, but may also explain why leaders’ identify with their organizations. If leaders become aware that their followers perceive them as a representative of their groups, they might assume that they are “doing things right,” and feel as valuable members of their organizations. According to the social identity theory (SIT; Tajfel and Turner 1978), this realization should trigger des-individuation and self-categorization processes, after which, the fact of belonging to their organization has become an important aspect of one’s self (Turner et al. 1987). Therefore, we hypothesize that

Hypothesis 2

High levels of self-awareness (H2a), relational transparency (H2b), balanced processing of information (H2c), and internalized moral perspective (H2d) will positively relate to a leader’s organizational identification.

Biological Gender and Leaders’ OID

Based on the higher congruence between the male gender role and the leader role, male managers should feel more comfortable in leadership positions as they are not prone to experience backlash effect like female managers (Rudman and Glick 2001). Further, as the male gender role comprises agentic characteristics and ascribes men to be dominant, competitive, and assertive, possessing power is congruent to their gender role (Bosak and Sczesny 2011; Williams and Best 1990). Therefore, men should not only be more motivated to attain leadership positions and power than women should (Hernandez Bark et al. 2014a; Schuh et al. 2014), but also once they assumed leadership positions, being a leader might be an (important) aspect of their self-concept. Moreover, being a manager in their organizations might become a relevant part of their self, and thereby strengthen their identification with their organizations. In contrast, one might argue that for women who have made it into managerial positions, being a leader becomes also a relevant aspect of their self (even more when becoming a leader was associated to a lot of effort). However, because female managers are more prone to role conflict and often need to perform better than their male counterparts (Eagly and Karau 2002), even if being a leader is an relevant aspect of their self, they might feel less welcomed or less appreciated by the organization than their male counterparts. Therefore, we assume:

Hypothesis 3

Male leaders will identify more with the organization than female leaders.

Further, based on the above hypotheses, we also propose that authentic leadership behaviors mediate the relation between biological gender and OID. Our main rationale is that because male gender and leader expectations are more congruent, men in managerial roles may feel more as they are acting according to their implicit expectations of what a leader should be compared to women. In consequence, men feel more authentic, and as result, male leaders should feel more appreciative of a context that allows them to behave authentic as leader. This appreciation should influence their perception of the organization and lead to a higher OID in men. Hence, we hypothesize that

Hypothesis 4

Authentic leadership, through self-awareness (H4a), relational transparency (H4b), balanced processing of information (H4c) and internalized moral perspective (H4d) will mediate the relation between biological gender and OID.

Team Prototypicality as Moderator

As an extension of SIT, the social identity model of leadership (SIMOL; Hogg 2001; van Knippenberg et al. 2004) may also be informative of when authentic leadership works better. Complementing other leadership research, the SIMOL focuses on the fact that leaders work in social groups, and not in a social vacuum. Thereby, SIMOL emphasizes the characteristics of the leader as a group member and the ability and motivation of the leader to act in the interest of the group (van Knippenberg and Hogg 2003). Further, it assumes that prototypical leaders will be more effective as it enhances the leader’s influence on the follower and followers’ trust in the leader. Additionally, prototypical leaders—as they are seen by their followers as one of them and acting in their interest—are liked more and successes are attributed to the leader’s capabilities (Van Knippenberg and Hogg 2003). Moreover, prototypical leaders have also more freedom of action (e.g., Ullrich et al. 2009). Combining these arguments with the role congruity theory, we assume if female managers feel and are perceived as prototypical for the group, they should experience less role conflict and incongruence between the female gender role and the leader role, enhancing their authentic leadership behaviors. Therefore, we assume:

Hypothesis 5

Team prototypicality will moderate the relation between biological gender and self-awareness (H5a), relational transparency (H5b), balanced processing of information (H5c) and internalized moral perspective (H5d). Female participants in the high team prototypical condition shall report higher scores in these four dimensions of authentic leadership.

In this line, if women experience that they are perceived as prototypical, this should facilitate their OID, because they should experience less conflict between the expectations toward them as women (female gender role, communal) and as managers (leader role, agentic). Hence, we hypothesize that

Hypothesis 6

Team prototypicality will moderate the relation between gender and OI. Prototypical female leaders are able to identify more with organization because of less incongruity.

In order to test our assumptions, and as part of larger project, we conducted an online experiment targeted at managers, in which we measured authentic leadership and manipulated participants perceptions of team prototypicality.

Method

Sample

Our sample consisted of 149 participants (43 % female). 71.8 % were German, 12.1 % from the US, and 16.1 % from other western European countries. At the time of our survey, mean age was 43.42 years (SD = 11.41). 76 % worked in the private sector and 18.1 % worked in public sector. Several work sectors were present in our sample, human resources (HR) being the largest sector (13.4 %), while 6 % of our sample did not indicate their work sector. 112 (42.3 % female) of our participants were upper managers with a mean of 24 employees (SD = 46.5) at their charge, whose mean age was 44.59, ranging from 25 to 64 years. In turn, 37 participants were employees (55.3 % female), but were asked to imagine themselves in a managerial role when completing our survey. Their mean age was 39.86, ranging from 19 to 86 years.

Procedure

Participants were recruited within western European countries from existing contact networks, a method of recruiting participants used in previous research (Escartín et al. 2013). All participants were invited to take part in an online survey, in which after filling their demographic information, they self-rated the frequency of their authentic leadership behaviors. On the following screen, they were presented with our team prototypicality manipulation, and a two-item manipulation check (see below). Immediately after, on the following two screens, two scenarios containing a moral dilemma were shown. The scenarios differed in the intensity of the standard violations and the scope of who would be harmed.

After reading each scenario, participants were asked to score how likely they would display 12 reaction possibilities provided. After this, participants completed out a measure of OID with the fictitious company portrayed in the scenarios. Previous studies show that using scenarios is an effective way to foster both leaders and followers OID in experimental research (e.g., van Dick and Schuh 2010).

Team Prototypicality Manipulation

We manipulated the participants’ beliefs of their team prototypically using a computer-based random function to assign participants to either a high vs. low team prototypicality condition. We showed participants a vignette text and six pie chart figures. In the vignette text, we asked participants to imagine that they were high-level managers in a multinational organization, leading a small team of direct reports. Furthermore, the text mentioned that they just received an email with the results of a recent HR internal survey exploring the match between how both upper and middle managers (their direct reports) on six key elements of their organization (vision, mission, organizational values and culture, strategy, work processes, and career development opportunities). Immediately below, all figures displayed either high (from 65 to 91 %; high prototypicality condition) or low (12–32 %; low prototypicality condition) levels of match (Fig. 1).

Fig. 1
figure 1

Theoretical model for conditional indirect effects of biological gender on OID, through authentic leadership dimensions using team prototypicality as moderator

Full size image

Measures

Cronbach’s αs for all measures are shown in the diagonal of Table 1.

Authentic Leadership (AL)

We used the Authentic Leadership Questionnaire (Walumbwa et al. 2008) to measure how frequently participants perceive themselves displaying authentic leadership behaviors. All items were rated on 5-point Likert-type scales, with values ranging from “1 = Not at all” to “5 = frequently, if not always”. Sample items are “Seeks feedback to improve interactions with others” (self-awareness), “Says exactly what he or she means” (relational transparency), “Makes decisions based on his/her core beliefs” (internalized moral perspective), and “Listens carefully to different points of view before coming to conclusions” (balanced processing of information).

Organizational Identification (OID)

We used three items of Kreiner and Ashforth’s (2004) measure. All items were rated on 5-point Likert-type scales ranging from “1 = strongly disagree” to “5 = strongly agree”. Sample items are “When I talk about this organization, I usually say we rather than they” and “the organization successes are my successes.”

Data Analysis

First, in order to ensure whether the different constructs and sub-constructs can be analyzed separately, we conducted a Hartman test, using confirmatory factor analysis to compare three models (Podsakoff et al. 2003). In the first model, we set all items of authentic leadership, OID and perceived team prototypicality to saturate a single latent factor. In the second model, we specified six latent factors, being OID, team prototypicality, and the four dimensions of authentic leadership. Finally, the third model includes three latent factors, OID, team prototypicality, and authentic leadership as higher order construct consisting of four dimensions. For this, we used LISREL 8.80 specifying a robust maximum likelihood (ML) estimation method, using Satorra–Bentler’s scaled χ 2 (S–B χ 2), the χ 2/DF ratio, Root Mean Square Error of Approximation (RMSEA), the Non-Normed Fit Index (NNFI), and the Comparative Fit Index (CFI) as goodness-of-fit indicators. Our rationale for this choice being that in spite the fact that χ 2 difference tests are frequently used for testing differences between factorial models, these tests are sensitive to sample size or minor departures form multivariate normality. Hence, it is generally recommended to use corrections and additional indices, such as the S–B χ 2 as goodness-of-fit indicator in combination with the χ 2/DF ratio to compare parsimony between models and evaluating differences in practical fit indices between models. Even though there are no generally accepted standards, some criteria have been proposed in the literature to interpret differences in practical fit indices. For example, for the NNFI, Widaman (1985) considered differences lower than 0.01 between NNFI values as an indication of negligible practical differences. Similarly, regarding differences between CFI values, based on the results of a simulation work, Cheung and Rensvold (2002) suggested that decreases in fit greater than 0.01 might be important. Finally, Chen (2007) suggests that when the RMSEA increases by less than .015 (a cut-off value similar to that of the CFI) one can also claim support for the more constrained (parsimonious) model.

Second, we tested our hypotheses using a conditional indirect effect approach. For this, we (1) first examined the proposed mediation model (Hypotheses 1–4) and then we evaluated (2) if our proposed moderator affected this mediation (Hypotheses 5 and 6). Furthermore, we tested the conditional effect on the remaining effect of biological gender over OID.

Tests of Moderated Mediation

Collectively, Hypotheses 1–4a to 4d suggest a mediation model, whereby the effect of leaders gender differences on their level of organization identification is transmitted by their authentic leadership behaviors. Concerning Hypotheses 5a–5d, we predicted that high levels of team prototypicality would substitute for gender differences in AL behaviors. Assuming this moderation hypothesis would receive support, it is plausible that the strength of the hypothesized indirect effect is conditional on the value of the moderator (low or high levels of team prototypicality), or what has been termed conditional indirect effects (Preacher et al. 2007). Testing conditional indirect effects through bootstrapped confidence intervals (CIs) is becoming an increasingly used practice in empiric research (Hayes 2009). Among its many advantages, this technique avoids power problems introduced by asymmetric and other non-normal sampling distributions of an indirect effect (MacKinnon et al. 2002). We used the PROCESS application provided by Hayes (2012) for testing our model.

Control Variables

Extant research shows that contextual variables such as the type of work contract can affect how people identify and commit to their organization (Guest 2004; Millward and Hopkins 1998). Hence, statically controlling for this contextual factor seems highly relevant to our study. Furthermore, in order to control for potential sample effects, we controlled for the effect of participants formal role as managers (by dummy coding into 0 = “manager” and 1 = “non-manager”) and possible nationality effects by entering it as control variables in our analyses. For this last, we created a dummy variable named “cultural cluster” which aggregates countries as suggested by researchers of the GLOBE project (House et al. 2001), where 0 = “Germanic cluster” included participants from Germany, Switzerland, Netherlands, and Austria, 1 = “Anglo-Saxon cluster” included participants from the US and UK and 3 = “Others” which included participants from the remaining western countries of our sample.

Results

Our CFA results show that the three-factor model (S–B χ 2 = 175.9; p. < 0.5; χ 2/DF = 2.71; RMSEA = .025; NNFI = 1; CFI = 1;) fits our data significantly better that the six factor model (S–B χ 2 = 208.28; p. < 0.5; χ 2/DF = 3.16; RMSEA = .047; NNFI = .96; CFI = .99;) or the single-factor model (S–B χ 2 = 349.75; p. < 0.5; χ 2/DF = 4.09; RMSEA = .078; NNFI = .96; CFI = 0.97). In addition, before testing our hypothesis, we explored if all assumptions for linear regression models were met. After testing for normality and exploring our model’s regression residuals, we identified and excluded nine outliers, whose regression residuals deviated more than 3 SD. After this, we recalculated our analyses to find that our model did not violate this or any other assumptions of linear models.

Manipulation Checks

To ensure that our manipulation had the intended effect, we performed a manipulation check of participants’ perceived prototypicality using two items, (1) “I represent what is characteristic about my team” and (2) “I represent what my team have in common” using a 7-items Likert scale ranging from 1 = “Do not agree with at all” to 7 = “Fully agree”. Cronbach’s α for these items was .91. We performed one-way Analysis of Variance, for the overall score of and for each item of our manipulation check. Our results indicate that participants in the high prototypical condition (M = 5.25, 95 % CI [4.97, 5.52]) had significantly higher scores than those in the low prototypicality condition (M = 2.48, 95 % CI [2.04, 2.91]), with p < .0001.

Testing of Hypotheses

Table 1 shows means, standard deviations, and correlation coefficients. Biological gender was only significantly related to self-awareness and team prototypicality did not show any significant associations. In turn, OID was positively related to contract types, and the four dimensions of AL.

Table 1 Means, standard deviations and correlation coefficients for continuous variables
Full size table

Table 2 and Fig. 2 show that having a permanent contract type (β = .19, p. < .01) and team prototypicality (β = .24, p. < .01) had positive effects over OID. Biological gender related to the four dimensions of authentic leadership, but did not have a main effect on OID. In this line, self-awareness did not relate to OID, but the three dimensions of self-regulation were significant related to this construct. These results support Hypotheses 1, H2b–H2d, but not Hypotheses H2a and 3.

Table 2 Conditional indirect effect of gender through authentic leadership dimensions with team prototypicality as moderator
Full size table
Fig. 2
figure 2

Conditional Indirect effects of biological gender on OID, through authentic leadership dimensions using team prototypicality as moderator

Full size image

Our Hypotheses 4a–4d suggested that the four dimensions of AL would partially mediate the effect of biological gender over a manager’s OID. Ordinary Least Squares (OLS) regressions paths (A-B1–A-B4) from biological gender to each AL dimension were significant, yet only paths from the self-regulation dimensions to OID (B2-C–B4-C) were statistically significant. Table 3 shows the indirect effect of each AL dimension at high and low values of team prototypicality, and its bootstrapped 95 % CI (using 10000 subsamples).

Table 3 Conditional indirect effect(s) of biological gender on organizational identification at high and low values of team prototypicality
Full size table

Self-Awareness

Both the OLS model (β = .01, n. s.) and the bootstrapped 95 % CI indicate that self-awareness did not mediate the effect of biological gender on OID at any level of our team prototypicality manipulation, so these results do not support hypothesis H4a. However, the OLS regression shows a significant moderator effect of team prototypicality in the relation between biological gender and self-awareness (β = −.32, p < .05). Following recommendations by Aiken and West (1991) and Dawson (2013), we performed single slope analysis to test the direction of these moderation effects, in terms of biological gender differences. The slope gradient for female participants between low (−1 SD) and high (+1 SD) levels of team prototypicality shows that highly prototypical female participants reported higher levels of self-awareness (β = .35, t (5, 133) = 4.30, p < .001), while male participants did not differ in their reports of self-awareness between low or high levels of team prototypicality (β = −.01, t (5, 133) = −.19 n. s.).

Self-Regulation

In turn, both OLS regressions and bootstrapped 95 % CI for the three AL’s self-regulation dimensions did not include zero at low levels of team prototypicality. In other words, these results indicate that the hypothesized mediation effect for these three dimensions is conditional of the manager’s team prototypicality levels, which acted as a first stage moderator. Therefore, hypotheses H4b–H4d are supported by our results.

The OLS model shows significant moderator effects team prototypicality in the relation between biological gender and relational transparency (β = −.30, p < .05) and a marginally significant moderator effect for balanced processing of information (β = −.26, p < .10),Footnote 1 but not for moral perspective. Figure 3 shows graphical representations of these interactions.Footnote 2

Fig. 3
figure 3

Interactive effect between biological gender differences and team prototypicality on self-awareness, relational transparency and balanced processing of information

Full size image

Single slope analysis for relational transparency (β = .22, t (5, 133) = 1.96, p < .05) shows significant slope differences between women low and high in team prototypicality, but no significant differences in slope gradients were found for male managers. In turn, no significant slope differences for balanced processing of information were found. Taken as whole, our results support Hypotheses 5a and 5b, but not 5c or 5d.

Biological Gender and OID

Our model also shows that team prototypicality (β = −.27, p < .05) moderates the remaining partial effect of biological gender over OID (path C) that is not mediated by AL dimensions. Figure 4 shows a graphical representation.

Fig. 4
figure 4

Interactive effect between biological gender differences and team prototypicality on Leaders’ OID (not mediated by authentic leadership dimensions)

Full size image

Slope gradient for female participants was significant, (β = .80, t (5, 137) = 2.23, p < .05), meaning that highly prototypical female participants reported higher levels of OID than low prototypical female participants. Similarly, a significant slope gradient for male participants indicates that highly prototypical male participants (β = .34, t (5, 129) = 2.26, p < .05) reported higher levels of OID that low prototypical male participants. Yet, from the absolute values of the gradients, without considering the mediation effect AL dimensions, our results show that if female managers perceive themselves as highly prototypical, they will identify more with their organizations than highly prototypical male managers. In consequence, our data support Hypothesis 6.

Discussion

This study had two objectives, first testing biological gender differences in managers' authenticity and its mediating role in the relation between biological gender differences and OID. Our second objective was to determine whether team prototypicality could facilitate this identification process by reducing the gender-leader role incongruency. Our results show significant differences in authentic leadership between men and women, with male managers being more authentic (Hypothesis 1), but no differences for OID (Hypothesis 3). In turn, while self-awareness did not relate to OID (Hypothesis 2a), the three self-regulation dimension mediates the effect of biological gender differences on OID (Hypotheses H2b–H2d). This suggests that male managers will only identify more with their organizations than female managers if they enact more often the regulatory dimensions of AL.

As hypothesized, team prototypicality moderated biological gender differences in self-awareness, relational transparency, and balanced processing of information, but not on internalized moral perspective. Overall, our results show no differences in these dimensions between male managers and female managers who perceive themselves as prototypical of their teams, but significant differences for those who do not feel prototypical. This not only supports previous findings (Eagly et al. 2004), as our data show that as leaders, men will act more accordingly to their inner values (whatever their content might be) than women, but also that social factors, such as feeling a prototypical leader does not reduce this biological gender difference.

As hypothesized, our results show that leaders’ team prototypicality also substitutes for biological gender differences in OID. Our findings suggest that although female managers are faced with the double bind and double standard based in the incongruity between the female gender role (communal) and the leadership role (agentic) which might lead to lower OID and lower authentic leadership behavior, if female leaders perceive themselves as prototypical these detrimental effects are reduced.

Theoretical Implications

Our study has three theoretical implications. First, in this study, we extended the authentic and prototypicality leadership literature into the gender-leader role congruity arena. While extant literature suggested that authentic leadership would be congruent with female role expectations as other uplifting forms of leadership (e.g., transformational leadership; Eagly et al. 2003), our results suggest that in fact, authentic leadership is more in line with androgynous approaches (Hernandez Bark et al. 2014b). Androgynous leadership styles are believed to combine the best of both worlds, as they englobe both communal and agentic behaviors, using positive modeling behaviors as one of an essential mechanism to reduce the gender-leader role incongruence. We proposed self-awareness, balanced processing of information and relational transparency to be dimensions of a more communal nature, while acting according to an internalized moral perspective is a more agentic dimension of authentic leadership. In line with role congruity theory, our findings show that only the relations between biological gender and the communal dimension of authentic leadership were moderated by team prototypicality, where the incongruence between gender and leader role expectations is higher. Furthermore, this suggests that it requires thinking beyond roles to tackle the gender-leader role dilemma, and that incorporating social attributions to the equation can help bridge this divide. To date, the primary focus of female leadership development efforts has been on either how to train women to overcome obstacles while by going through the labyrinth of leadership (Eagly and Carli 2007), or on how to reduce gender inequality in personnel selection, e.g., via transparent and standardized criteria and procedures. Our data show that a social-oriented approach can also be effective to facilitate women to achieve higher authenticity in their managerial roles and in consequence enable them to identify more with their organizations.

A second implication of our study is that authentic leadership can also contribute to expand our understanding of SIT. Our findings suggest that “outsiders”, (e.g., female managers) will identify more with their groups (organizations) if they perceive themselves in a context that enables an authentic functioning in their leadership roles, for example, when they perceive themselves as prototypical. In this direction, a higher identification should result in exemplary role modeling, understood as a higher form of prototypicality in which the leader not only represents what is common for their group, but also actively moves the group to higher state by becoming a positive model to which group members can not only identify with, but can also actually imitate, achieving a more integrated form of regulation. This is congruent with the social identity analysis of leader prototypicality, where leaders are not only passively prototypical or not but they can actively shape their teams’ identities (see Reicher et al. 2011).

Finally, a third contribution of our study is to authentic leadership theory. Our findings further support this theory’s assumption that authentic leadership is not only about the leaders, but also about the growth-enhancing relations that leaders and followers establish (Gardner et al. 2005; Ilies et al. 2005; Leroy et al. 2012). Our findings provide evidence that as some scholars suggest (Avolio et al. 2004; Ilies et al. 2005) followers are not mere passive targets of leader influence can facilitate leaders’ authentic behaviors, for example, through their prototypicality attributions, especially for those leaders facing additional challenges (e.g., females managers).

Practical Implications

Our study has several practical implications, especially for gender equality. Organizations can increase both the self-perception of female leaders, and the followers’ perception of leaders’ prototypicality, an essential aspect of leaders’ effectiveness (e.g., Van Knippenberg and Hogg 2003). First, by integrating communal and agentic competencies in competence models and leadership trainings, organizations can reduce the incongruity between the female gender role and the leader role (Hernandez Bark et al. 2014c). Second, as already stated in the SIMOL, prototypicality is not a fixed and stable characteristic, but leaders can actively influence their prototypicality by approval-seeking out-group behaviors or by reconstructing the social context (Van Knippenberg and Hogg 2003), an aspect of which especially female leaders might benefit. Therefore, future training programs, targeted at female managers, should focus on strategies and behaviors that individuals can use to increase their perceptions of prototypicality. Third, as highly identified employees stay longer in the organization and report lower turnover intentions, organizations might benefit from establishing and embracing authentic leadership.

Furthermore, managers can also benefit from our findings regarding the relevance of prototypicality particularly when they are female and would like to maintain or even increase their authenticity as leaders. In this sense, as mentioned, female managers can act as entrepreneurs of identity (Reicher et al. 2005) and themselves at a more prototypical position. For example, the first German female secretary of defense Ursula von der Leyen, right after taking office, started creating a new vision for the German army to become more family friendly and of supporting female soldiers in their careers. By putting such topics on the agenda, she as a woman made herself more prototypical as a female minister in an otherwise very male-dominated organization.

Strengths, Limitations, and Future Research

As any other study, our work has several strengths and some limitations that future studies could address. A first strength is that while it targets actual leaders in real organizations, our online survey allowed us to perform an experiment, by effectively manipulating a social attribution such as team prototypicality. This is important, because as this variable is context-based, we could not ensure that our participants would perceive themselves as prototypical unless we experimentally manipulated it. Second, a large part of our sample consisted of middle to upper managers, making our conclusions highly relevant for this target group. In this line, our study has important implications for current issues in organizations, such as gender equality, effective managerial socialization, and its associated positive organizational outcomes (Bauer et al. 2007).

Among the weakness of our study, first our sample size is relatively small and only involves participants of western societies (e.g., Germany or the United States). While we agree and encourage future research to replicate our study in other contexts with larger samples, our non-parametric approach for testing our model (95 % CI bootstrapping) has partially addressed this issue. A second weakness is that some may argue that we are testing a mediation model in a cross-sectional sample. In turn, we may reply that in fact, in order to avoid common source bias and support the causal relation implied in mediations, following Frazier et al. (2004) we measured our relevant criterion variables at two different points in time, before and after our team prototypicality manipulation. Because this is just a partial solution, we encourage future research to replicate our study using a full longitudinal design, and several data points. Further, future research might rely on follower’s perspective, and examine how followers’ perceived team prototypicality relates to both leader’s and followers biological gender, authentic leadership, and its related outcomes such as trust, job satisfaction, and loyalty.

Finally, because many scholars consider authentic leadership as root construct underlying other positive forms of leadership (e.g., ethical or spiritual leadership; Avolio and Gardner 2005; Avolio and Mhate 2011), our findings for the relation between biological gender differences and OID could be expanded by combining other uplifting leadership styles and different forms of identification. For example, based on Bandura’s (1986) social learning theory, ethical leadership agrees with authentic leadership on the importance of (ethical) role modeling as one its central influence mechanism to shape followers positive and pro-social attitudes and behaviors. Furthermore, this theory also suggests that ethical leaders cognitively evaluate the ethical context of their organizations using several indicators, such as its ethical climate, culture or reward systems (Brown and Treviño 2006), which seem analog to AL’s balanced processing of information mechanism. If there is a high fit between the internalized values of the ethical leader and the cultural values of the organization he or she belongs, then a high organizational commitment and identification could be expected. However, if there is a low fit between the manager's internalized values and the cultural values of the organization, other forms of identification may be triggered (ambiguous, neutral or des-identification; Kreiner and Ashforth’s 2004).

From a gender role perspective, we proposed that acting according to internalized values independently of contextual pressures could be classified as an agentic expectation. Therefore, male managers should perceive themselves as more moral managers than female managers, reporting higher frequency of these behaviors. Therefore, our findings seem to support the idea that ethical leadership could mediate the relation between biological gender and organizational identification, being this effect conditional of the ethical context, but independent of other social aspects such as team prototypicality.

As ethical leadership, spiritual leaders also uses positive role modeling to influence followers, by embodying “spiritual values such as integrity, honesty and humility, creating the self as an example of who can be trusted, relied upon and admired” (Fry 2003; Reave 2005, p. 663). However, from a gender role perspective, typical spiritual leader behaviors such as communicating a shared vision and a sense of membership to satisfy followers’ spiritual needs are clearly communal behaviors. Therefore, it also seems plausible that biological gender differences in identification in leaders displaying spiritual behaviors should be reduced if female spiritual leaders feel as prototypical, trusted, and admired members of their communities.

Concluding Remarks

To conclude, we have shown that authentic leadership is an important determinant of leader’s OID, and that women are disadvantaged for showing authentic leadership due to the gender-leader role incongruence. Prototypicality can reduce this disadvantage for female leaders and may be a route toward more equality in leadership, and promoting higher authenticity in the workplace.