The Importance of Role Models and Demographic Context for Senior Women's Work Identity Development
Ruth H.V. Sealy,
International Centre for Women Leaders, Cranfield School of Management, Cranfield University, Bedfordshire MK43 0AL, UK
Ruth H.V. Sealy, Senior Research Fellow, International Centre for Women Leaders, Cranfield School of Management, Cranfield University, Bedfordshire MK43 0AL, UK. Mobile: 07711 477109; e-mail: firstname.lastname@example.org
Ruth H.V. Sealy, Senior Research Fellow, International Centre for Women Leaders, Cranfield School of Management, Cranfield University, Bedfordshire MK43 0AL, UK. Mobile: 07711 477109; e-mail: email@example.com
The lack of senior female role models continues to be cited as a key barrier to women's career success. Yet there is little academic research into the gendered aspects of role modelling in organizations, or the utility of role models at a senior level. The paper starts with a review of papers examining the construction of role models in organizational settings. This leads to the inclusion of two related areas – organizational demographics as the contextual factor affecting the availability of role models and how they are perceived, and work identity formation as a possible key explanatory factor behind the link between the lack of senior female role models and the lack of career progression to top organizational levels. The literature looking at social theories of identity formation is then considered from a gender perspective. The key gaps identified are that while the behavioural value of role models has been well documented, a better understanding is needed of how gender and organizational demography influence the role modelling process. Importantly, the symbolic value and possibly other values of female role models in the identity construction of senior women require further in-depth investigation. Finally, this review calls for a more integrated approach to the study of role models and work identity formation, pulling together literatures on organizational demography, the cognitive construal of role models and their importance for successful work identity formation in senior women.
Women at the top of today's largest companies are not yet much in evidence. After 30 years of equality legislation, only 4.8% of executive directors in the UK's top 100 companies in 2008 were women (Sealy et al. 2008). The organizational reality in many countries is that women still struggle with rigid and male-dominated hierarchies, promotion based on uninterrupted linear career paths with little flexibility, lack of credibility in a masculine culture, isolation from other women and transactional rather than transformational management styles (Rosener 1990). There are very few women at the top to act as examples, or role models, of how these challenges can be overcome. While it can be argued that time will eventually provide more female role models for senior women as increasing numbers of women achieve top positions, the fact that there were only 17 female FTSE 1001 executive directors in 2008 (up from 11 in 2000) indicates that it will take many years before there is a substantial pool of available female role models at board director level.
This issue is important for several reasons. The lack of female role models is cited by academics as contributing to women's propensity to resign (Rosin and Korabik 1995). Women often leave to join other, more synergistic organizations, where they feel their leadership qualities will be recognized, or set up their own businesses (White 1992). The issue has been highlighted in a study by Catalyst/Conference Board (2003) of a large number of European managers who cite the lack of female role models as the second biggest barrier (after sex-role stereotyping of leadership) to women's career success. Other recent surveys also highlight the lack of appropriate role models as an important barrier for women to achieve senior positions (Catalyst and Opportunity Now 2000; DDI/CIPD Leadership Forecast Survey 2005; Eve.lution Ltd 2005). From psychology, we would expect role models to be part of identity development processes in early career (Gibson 2003). But these European practitioner surveys do not explain why women approaching senior levels consistently report that they lack and need senior female role models.
However, despite this high profile and topical issue, initial examination of the literature on role models revealed few recent academic studies apart from that of Gibson (2003, 2004) who called for a re-examination of this construct outside the mentoring field, and Ibarra (1999) who identified role models as vital for the successful development of young professionals. Very few studies appear to have addressed the gendered aspects of this phenomenon in business settings and, prior to Gibson's work, the importance of role models for people at senior levels seems to have raised little interest.
While we recognize that senior women are not the only minority group in the workplace to have few similar role models, and that much of this literature review is relevant for other minority groups, our focus on women is derived from the desire to understand why senior women reportedly value female role models so highly, given their career maturity and success to date. What is the value of role models to senior women? If this can be established, perhaps interventions can be designed to address that need.
Therefore, a systematic literature review has been conducted to establish what is known and what more needs to be understood about role models for women as they establish their senior organizational roles. Our research question is: why are female role models reportedly so important for senior women? We begin with definitions of role models and the role modelling process, drawing on Gibson (2003, 2004). As the demographic context affects not only the availability of role models but also how those individuals are perceived, we next examine the context of organizational demography as it relates to role models. As Ibarra identified possible gender differences in the use of role models in professional identity formation, this literature review was extended to include women's work identity development. We then draw on social theories of identity to illuminate the process of role modelling as it impacts identity work in its organizational demographic context. Earlier research on role modelling has examined these processes in isolation from the male-dominated context that most senior women find themselves in, while organizational demographers mention the important phenomenon of the lack of female role models only at the surface level. We discuss the findings in relation to the paradox of the present cohort of well-educated professional and successful women managers and their expressed needs for female role models.
The contribution of this paper is threefold. First, it brings together the literature on the micro-processes of role modelling and on identity construction for senior women in their macro-level organizational demographic context, underpinned by insight from social theories of identification, to identify important gaps in knowledge about the importance of role models that will be pointers for further research. Second, in bringing these literatures together, we also reach understanding of why some senior women are rejected as role models. Third, the review identifies the need to increase the conceptual understanding of the symbolic value of role models.
The concept of role models
We begin by examining how the term ‘role model’ has been defined. Shapiro et al. (1978) defined role models as ‘individuals whose behaviours, personal styles and specific attributes are emulated by others’ (p. 52) and showed that modelling contributes to identity construction. A role model can be a symbolic entity, an inspirational and/or motivational individual, someone from whom one can learn and model desired behaviours (Lockwood and Kunda 1997). Similarly Gibson (2004) defines a role model as ‘a cognitive construction based on the attributes of people in social roles an individual perceives to be similar to himself or herself to some extent and desires to increase perceived similarity by emulating those attributes (p. 137). Such definitions exclude learning how not to do things, the negative role model. However, Gibson (2004) moves the definition forward to include this aspect in his definition of role modelling as ‘a cognitive process in which individuals actively observe, adapt and reject attributes of multiple role models’ (p. 136).
Gibson distinguishes the use of role models from mentors, which has more recently dominated the literature. Mentors provide advice and support through an interactive relationship, and behavioural role modelling is one of the functions of mentoring (Ragins and Cotton 1999). Other distinctions are those of permission and involvement – individuals can choose role models without involvement or permission from the model, whereas mentors have to agree to participate.
The dimensions of role models
Cognitive dimensions of role models
The focus of the Gibson papers (2003, 2004) is on the cognitive construal of role models. Gibson (2003) identified two cognitive dimensions: positive/negative role model constructions and global/specific constructions. Negative constructions refer to observation of a role model in behaviours that are examples of how not to behave in a particular situation. Global constructions relate to a variety of attributes in a role model to be emulated, while specific constructions refer to a single attribute that can be drawn on in a particular context. Role models are seen as cognitive constructions based on attributes of people in social roles. The individual perceives the role model to be similar to themselves to some extent and desires to increase perceived similarity by emulating those attributes, resulting in a behavioural change.
Gibson (2003) found that the basis of construing positive role model attributes was perceived availability, defined as ‘the degree to which the individuals think they are sufficiently similar in their environment’ (p. 599). We should understand here that similarity could be actual or perceived or desired similarity, and could cover a number of attributes, e.g. similar background, style or demographics. Gibson does not distinguish between these.
This is consistent with theories that link positive affect to similarity, e.g. social comparison theory (Collins 1996). Individuals seek others with some similarities, as they are informative for making accurate self-assessments and can be inspirational for self-improvement (Lockwood and Kunda 1997). This is an essential part of the identification process. In other words, if it is not possible to find someone sufficiently similar to emulate, the individual loses out on potential benefits. Hence, in an organizational context, if women do not see similar women above them, they may perceive that there is no one to emulate. Thus the recent surveys in Europe (Catalyst/Conference Board 2003; Catalyst and Opportunity Now 2000; DDI/CIPD 2005; Eve.lution Ltd 2005) citing a lack of available female role models may not be referring to just the low numbers of actual women, but suggests an even lower number of women perceived to be similar to those seeking role models, or a lack of female role models to whom they would desire to be similar.
Gibson's interest is in the cognitive construal of role models, with the emphasis on the individual's cognitive processes, rather than the behavioural outcomes, but already he has introduced the issue that this construal cannot be independent from the social context in which individuals find themselves. The organizations in which individuals work are social contexts such as organizational demography and culture. This overlap of theoretical areas (individual and societal, cognition and behaviour) indicates part of the challenge of studying this topic. These need to be considered carefully when looking at how and why individuals construe role models as they do.
Structural dimensions of role models
Gibson (2003) also identified two structural dimensions of role models: close/distant and up/across or down. The close dimension refers to someone well known to the individual, while the distant is a role model outside the normal interactions. Up refers to a role model in a higher position, across is a peer, and down is a subordinate, and across or down might also relate to others with undefined status. The gender demographic context is important in terms of availability of senior role models, as men will have many more possible ‘up’ candidates to select from in male-dominated hierarchies. However, the disadvantage for senior women of fewer available role models may be present at similar and lower levels too. Women at senior levels in male-dominated organizations are likely to have far fewer female peers than male, and may find female subordinates to have only limited possibilities as role models, owing to their different hierarchical status and experience.
Role models and career stages
We found only one research study (Gibson 2003) that investigated role modelling at different career stages. Important to the present study are the mid and late career stages in which senior women are situated. Gibson examined career stage differences in the dimensions of role model formation. Traditional career theories (Erikson 1950; Super 1957) suggest that as individuals get older, confidence in self-concept is increased and hence the requirement for role models diminishes. However, Gibson's findings suggest that the tendency to observe role models does not diminish with age, but rather the individual changes the emphasis placed on various dimensions. In Early-career, which Gibson called the acquiring stage, individuals work on a viable self-concept – emulating others, using positive, close role models and a range of attributes in the construction of the professional identity.
In Mid-career (the refining stage), individuals seek to refine the self-concept as confidence begins to grow, selecting specific and generally still positive attributes from role models. The individuals in Gibson's study also emphasized the importance of having role models for task transitions but perceived that few were available. Mid-career was often a state of ambiguity and uncertainty, where individuals felt they lacked guidance, particularly through visible role models, of what the next career step should look like. Respondents in mid-career also felt that, as time passed, career choices became limited with further restricted availability of relevant role models. Role models were now different people for different things, as the individual created composite models. In addition, the respondents’ ability to pick out useful attributes increased, giving them a better sense of their own self and style.
In Late-career (the affirming stage) individuals sought to enhance and affirm the self-concept, learning specific skills tied to specific goals. They often construed not just positive but also negative role models to help affirm their own sense of uniqueness.
The tendency of those in middle and late career stages was to integrate specific attributes of role models into composites that approximated a global role model. Creating a global composite is consistent with social learning theory, modelling being a cognitive process in which the person creates an image of how a behavioural model does a particular task and then generalizes that to different situations (Bandura 1977b).
Gibson and Barron (2003) found that older employees in a large US engineering organization who perceived they had multiple organizational role models available and identified with them showed increased organizational commitment and job/career satisfaction. These late career stage employees created composite role models, based on their own cognitions rather than the actions of the role model. If losing valued female employees is an issue for organizations, particularly at senior level, then the Gibson and Barron study is significant in showing that availability of role models increases commitment and job/career satisfaction. The important factor is that the employees believed the role models had similar values and goals to themselves. Thus the recent European surveys may suggest a lack of role models (female or male) perceived to have similar or desirable values or goals to the women completing the questionnaires.
The implications of Gibson's empirical studies, with the first steps to developing a career-stage framework of role models, are valuable for organizations. In mid-career, individuals felt they lacked role models. While organizations may feel that individuals should find their own feet, they risk losing or misdirecting experienced talent by not addressing this developmental need. This may be congruent with the perceived exiting of women at mid-stage career currently occurring (Hewlett and Luce 2005). Gibson's study suggests that organizations should recognize the growth needs of mid-stage individuals, by emphasizing exposure to exemplary peers and superiors. In this era of more rapid organizational and career changes, individuals need to establish their own ‘network’ or constellation of developmental relationships, varying in strength and variety (Higgins and Kram 2001). Gibson believes role models are an important part of this relationship portfolio.
While Gibson's research is undoubtedly important in terms of raising the profile of research into role models, there are some limitations of his studies and questions left unanswered. First, he does admit that, in searching for patterns of difference across the stages, there was a risk of assuming more homogeneity within the stages than was actually the case. This is particularly an issue in the mid-stage, where his sample has 10 out of 15 workers who have not yet been promoted to partner level, and only five who have made this important transition. We suggest that these two groups may have very different expectations and attitudes regarding their current career success or lack thereof. In addition, in his empirical paper (Gibson 2003), the boundaries between role models, mentors and behavioural models become blurred as Gibson talks about close role models and watching response and feedback. Moreover, while he defines role models as cognitive construals, he does not overtly take into account the symbolic or inspirational effect or emotive value that role models have. Finally, this study of career stages is undertaken in two professional services firms, a sector that is particularly competitive; hence, the need for more research into role modelling and career stages in other environments is indicated. In the next section, we examine research on organizational demography for a deeper insight into the influence of environmental context on role modelling, given that women in management are frequently seeking female role models within male-dominated hierarchies and organizations.
The significance of context: Organizational demography
Tokens and minority groups
Research on organizational demography and power addresses the context of the working environment that influences the availability of role models, and the processes of gender identity formation. In her ground-breaking work, Kanter (1977) observed the asymmetric power within an organization, part of a culture where the majority dominated and marginalized the minority, and where structures emerged to preserve this situation. The concepts of ‘homophily’ and ‘tokenism’ were based on the premise that people prefer to work with similar others. Kanter showed that women became ‘tokens’ when in a numerical minority of less than 15%. Through processes including assimilation, polarization and exaggeration, stereotypes were used to heighten boundaries, and women became both highly visible and isolated. In Kanter's view, only when the proportion of women passes the 15% threshold to become a minority, rather than a token, could they begin to overcome these pressures. It is frequently assumed that, by hiring more minorities and women, the power balance will improve, but identity groups need to be equal in their access to power resources (Kanter 1983) for an improvement of attitudes (Kossek et al. 2003). This suggests that organizations may need demographic earthquakes to take them to the tipping point of 35% (Kanter 1977) to prevent the perpetuation of stereotypical negative dynamics.
Impact of unbalanced sex composition
In her seminal work, Ely (1994) looked at the impact of women's proportional representation at the top of organizations and what effect it had on the relationships between other women in that firm. She found that, in firms with few senior women, women were
less likely to experience gender as a positive basis for identification with women, less likely to perceive senior women as role models with legitimate authority, more likely to perceive competition in relationships with women peers, and less likely to find support in these relationships. (p. 203)
She also found that, in firms with more balanced gender representation at the top, the opposite was true on all counts.
The presence of executive women signals the level, as well as likelihood, of possible promotion to other women. Women in executive positions may also be able to influence the organizational policies and culture and make it more attractive for women to stay with the organization. Elvira and Cohen (2001) proposed that organizational sex composition at senior levels explains turnover differences for men and women in a female-dominated organization. They found that the proportion of executive women above them directly affected the turnover of women, who were more likely to leave if they were in the lower ranks, but not if they were in middle and top positions. Elvira and Cohen speculate that lower rank women may perceive such a distance between themselves and the executive women that this limits their view of possibilities for change. In contrast, higher-level women are more similar and closer to the executive women who have the power and resources to affect their working conditions, and hence were less likely to leave the organization as the proportion of women above them increased.
Ely's work goes beyond that of Kanter (1977), as the latter suggested that balanced representation at peer level would reduce sex-role stereotyping and promote a greater sense of belonging for women. But Ely says that this is not the case unless there are women in positions of authority in that organization, as ‘sex may persist as a salient category with negative consequences for women lower down in the organization’ (1995, p. 590). Gender cannot be treated as an objective property, synonymous with biological sex or universal across organizational settings. Gender is an ongoing social construction. Both Ely's 1994 and 1995 papers show how women's presence in positions of power positively affects the social construction of gender definition and the processes that create gender identity at work.
Women in male-dominated firms will perceive greater psychological and behavioural differences between men and women and will define these differences along sex-role stereotypes (Ely 1995). Following social identity and self-categorization theories (Chattopadhyay et al. 2004), such women would evaluate women's attributes less favourably to the firm's criteria of success. Ely found that sex-integrated firms (those with considerably more than token numbers of women in management at all levels) had greater latitude in gender roles, with the women consciously enacting masculine and feminine roles as they saw fit. In male-dominated firms, women's discomfort with sex-roles and rating themselves less favourably in relation to the firm's requirements for success would explain lower levels of job satisfaction, lower expectations and desire for promotion. In sex-integrated firms, biological sex was less tightly linked to bipolar construction of gender. Women had a greater sense of acceptance, higher satisfaction with firms and optimism about their careers.
Liff and Ward (2001) examined the under-representation of women in senior management positions within a UK high street bank. Junior and middle male and female managers were asked their perceptions of the personality and behaviour characteristics associated with success within their organization. In many cases, men and women identified the same issues, but the significance of them for their own decision-making and the way others interpreted their behaviours varied – particularly in relation to the perceived incompatibility between active parenting and senior roles. The uncertainties around succeeding as a female in this UK bank were considered to be reinforced by the very small number of visible senior women who could act as role models. Those women who had made it to senior roles were described by participants as having ‘lost their femininity’. Characteristics and behaviours required for career success were reportedly more ‘male’ than ‘female’, and descriptions of what was required were of either a paragon or someone ‘not very nice’ or ‘unnatural’. This is a recurring theme in rationalist or voluntarist explanations of why women choose not to pursue high-level careers. Paraphrasing Schein et al.'s (1996) well-known phrase, Liff and Ward suggest ‘think female manager, think childless superwoman!’
Gibson and Cordova (1999) found that the proportion of women at the various hierarchical levels within an organization made a difference to men's but not women's cross-sex role modelling patterns, and women in sex-balanced firms were less likely to place importance on same-sex role models. They found that, across organizations, women were less likely to have specific role models for success and more likely to have negative role models. In firms with a greater number of female partners, women were more likely to say they had good role models available to them for career success, though their role models were not necessarily senior females. These findings are consistent with many of those of Ely (1994, 1995).
Some practical implications are clear from these findings. Increasing the number of women in senior positions may help organizations both reduce turnover and draw on a wider range of female talent. These studies also suggest that visible role models of women in authority could be associated with an increase in women's ambitions, not just because the exemplars prompt the individual women to increase their career aspirations, but because, by their presence, they start to change the old gender schema of status and power (Ragins and Sundstrom 1989).
Role model research – relevance to women managers
So far in this review, we have examined the process of role modelling and the significance of gender demographic context. We are interested in identifying work that pertains to women managers at senior levels, who we expect to be in their 30s to 50s. Gibson's (2003) work illustrates that there are career stage differences within the professional services context. Gibson's mid-career group average age was 38, while the late career group average age was 47 years, so findings related to both those groups are of interest. From organizational demography research, the work of Kanter on minority group interactions with the majority, and research by Ely on the influence of different sex compositions of organizations on women's career outcomes are particularly useful in helping us understand the context.
There are a few studies on role modelling and sex differences, but they are limited by undergraduate or junior samples and experimental designs (e.g. Murrell and Zagenczyk (2006) who allocated role model status to those who scored highly on performance measures rated by their peers.). Lockwood (2006) examined same-sex and opposite-sex role models with undergraduate students, revealing that women are inspired by outstanding women in their field, although not by outstanding men. Lockwood suggests that female role models are particularly inspiring in situations where they are in the minority. They provide evidence that barriers can be overcome despite discrimination, and may be an important means of undermining negative gender stereotypes. However, we suggest that the efficacy of the role models would depend on how ‘similar’ they are perceived to be. Lockwood suggests that, for women, exposure to another successful woman can make them rate themselves more positively. As she predicted, however, gender did not affect male participants’ identification with a role model, and their self-evaluations were not affected by exposure to male role models. Lockwood suggests this is because men do not face the same career barriers, and have no need for exemplars of success. Lockwood suggests that future research should look at the long-term effects of role models on women's career performance.
Although Gibson (2003, 2004) does not take a gendered perspective, he does address sex differences in his discussion. Stating that women typically have fewer available same-sex role models, he suggests that women face an arduous cognitive task of translating male role model behaviour into behaviour that works for them. The pool of role model material for women is constrained, providing them with lower-quality information than that available to men. Women must adapt the types of behaviours that work for men in order to make them work for them. Women's role modelling requires greater cognitive processing. Whereas men can take the attributes that the organization has recognized and rewarded in their male role models and add those behaviours to their own repertoire, women have to make such images from role models that come from more diverse and fragmented sources.
This would suggest a structuralist rather than developmental standpoint – in other words, that patterns of modelling are dependent on the gender context that individuals find themselves in, rather than on inherent differences in modelling tendencies between men and women.
Why are role models important in the identity construction of senior women?
Developing ‘Provisional Selves’ through Role Models
To gain a better understanding of the importance of role models, we need to examine the utility of role modelling in the development of work identity. Work identity has been defined as an enduring set of attributes, beliefs, values, motives and experiences by which people define themselves in a given work role (Schein 1978). However, as Gibson has shown above, work identity is not a fixed concept, but changes with work role changes. Career transitions are an opportunity for renegotiating one's work identity through the mechanism of ‘possible selves’, defined as who one might become, would like to become, or fears becoming (Markus and Nurius 1987). Ibarra has conducted extensive research into professionals in investment banking and consulting firms making transitions from middle to senior management positions (1999, 2000) and from senior management to leadership roles (2003, 2007). In her Adaptation Process, she revealed that successful transitions required three basic tasks: observing role models to identify potential identities; experimenting with ‘provisional selves’; and evaluating experiments against internal standards and external feedback.
In her 1999 study, over 90% of participants described how role models displayed the role identity they were attempting to assume. By observing successful role models, the subjects built a store of tacit knowledge, attitudes and impression management routines. They created the idea of a ‘possible self’– the role identity they wanted to assume. They also learned about acceptable variation in how to enact the role by comparing different role models. They then experimented with these ideas, using a process of observe, practice/test and evaluate against internal standards and external feedback. Participants also matched role models based on their attractiveness, i.e. to what extent did they admire or share the traits underlying the role model's behaviour. As with Gibson's studies, Ibarra found that participants used role models to define negative behaviours or characteristics, i.e. the role model represented a feared possible self, a negative role model. This suggests a value of more senior role models even if they are not attractive.
The process of acquiring behavioural skills, such as a work style, is different from learning knowledge in that the learning must be refined through personal experimentation, not just vicariously through observation (Bandura 1977b). In Ibarra's study, the most prevalent form of experimentation was imitation. This was done on either a wholesale or a partial basis (mimicking global or individual traits). Selective imitation, a ‘mosaic of different people’ was a more sophisticated form of mimicry, combining facets from multiple role models to craft a more self-tailored persona. Those using this tactic suffered less concern regarding authenticity, i.e. the degree of congruence between what one feels and what one communicates in public.
In contrast, participants who used true-to-self strategies (e.g. staying with their present style, focusing on their present strengths and searching for a very closely matched single role model) in making the transition made several references to caution, modesty, being acutely aware of their own limitations, avoiding exaggerated displays of confidence, and being more concerned with client credibility in the long term rather than creating a good first impression, focusing on substance over form. However, as they clung to their old identities, they struggled to transfer some of the new styles and skills required, experiencing longer-term dissonance between their current and ideal selves. Their actions also limited the growth of their repertoires, providing a restricted store of material from which to select and grow. Ibarra's study did not set out to give a gendered perspective, so it is not until the discussion part of the paper, that she makes it clear that the true-to-self subgroup who used this limiting tactic was almost entirely women. She does not, however, attempt to explain why this might be. How much of this is related to individual and how much to situational factors? There was a scarcity of senior females, and Ibarra found that this did constrain the attractive identity matches for women, making it harder to learn.
Identification with role models infuses behaviours with meaning and purpose, providing more motivation to change. By identifying with role models, people move from compliance to assimilating role requirements (O’Reilly and Chatman 1986). As participants evaluated their provisional selves, they became aware of the need to find more appealing, feasible role models and so extended their role model set. Ibarra's study highlights the value of role models, increasing repertoire variety and therefore the likelihood of successful adaptation.
Ibarra proposes a model whereby the Adaptation Process she described mediates between situational and individual influences on the one hand, and the identity construction processes, on the other. If successful identity construction is essential to career success, the availability and successful use of role models become key antecedents to this. Her findings on the women's experiences of career transition show consistency with well-established findings on minorities and majorities in organizations (Kanter 1977). Women were more likely to experience difficulty finding suitable role model matches, to use true-to-self strategies and to perpetuate provisional selves that they described as inadequate. Combining Ibarra's findings with those on organizational demography above furthers our understanding of the processes affecting women's career transitions and the importance of role models.
Many of the issues concerning women in the workplace today are around the lack of a sense of authentic identity. Pratt et al. (2006) conducted a six-year qualitative study of medical residents in the US to examine identity construction after work–identity integrity violation: a mismatch between what they did and who they were. Like Ibarra, they found that role models were critical to learning. However, unlike Ibarra they did not find that the junior doctors tried on ‘possible selves’ based on multiple role models. Rather they found that ‘role model choice was based more on a justification or validation of an existing or emerging identity’ (p. 255). But that presupposes a role model that can be aligned with the individual's authentic ‘true self’ identity. As Kahn (1990) noted, work becomes meaningful when one's ‘preferred self’ can be expressed through one's work and one's membership in an organization. Pratt et al. (2006) argue that ‘achieving alignment between identity and work is a fundamental motivator in identity construction’ (p. 255). This is a complicated and challenging task for women aspiring to leadership in unbalanced demographic contexts.
Developing managerial identities through role models
In a qualitative study of young women managers, Singh et al. (2003) found that they tended to use a selection of role models from a variety of domains, many from outside the workplace, to help them build appropriate identities. This is encouraging after Ibarra's study, which suggested women tended to search for a single global role model rather than plural, and is in line with Gibson's findings. However, they did not create a composite global role model, but rather drew on inspiration from any relevant role model for a particular task or situation. The role model sets crossed the business world, family and popular celebrities. Although 60% of the role models were from the business world, very few top businesswomen were mentioned, and women reported very few acceptable role models available within their own work environment. Often, those women at the top who did not have families were rejected as role models. They were seen to have sacrificed too much of their social and emotional capital in their quest to succeed in the masculine workplace, echoing Liff and Ward (2001).
As in Ibarra's study, when asked what they learnt from their role models, the women talked about personal characteristics and style. Various ‘masculine’ style traits emerged, such as control and determination. However, for others what was important was seeing their role models use their ‘feminine’ traits in their work style, as they wished to emulate these in their own working lives. Role models were used to develop ‘ideal selves’ for their future career. But such role models were rarely available. This study is useful in establishing that role models are an important influence in work identity development.
Sheppard (1989) highlights problems for organizations without senior female role models. Additional time and effort is spent by female managers ascertaining how they should behave and present themselves at work. Women end up taking on an inauthentic work identity, like ‘wearing a mask’. Men may sexualize and objectify women as a method of control, and women find themselves between a rock and a hard place of being either the unprofessional objectified sex object or the ‘not very nice’, too masculine woman (Liff and Ward 2001). Women without female role models tend to desexualize themselves as a coping strategy. Leadership and management styles are self-perpetuating in their masculinity, as femininity is not valued by either men or women (Schein et al. 1996). Wahl (1998) comments that men get confirmation of themselves and their identity as leaders through a shared maleness, whereas, unless there are sufficient female leaders, women have no such resource.
Social theories of identity
From the literature, we have shown that role models are used for professional and managerial identity development, and that organizational demographics have a major impact on the availability of female role models and the less favourable context within which women managers develop their work identities in male-dominated organizations. We now move to explore whether social theories of identity can shed light upon the process of role modelling within identity development to explain why role models are so important for senior women.
Socialization is a negotiated adaptation through which people aim to improve the fit between themselves and their work environment, by refining their emerging ideas of who they want to be in that role (Bandura 1977a). Identity refers to the meanings or self-conceptions that are attached to an individual either by him/herself or others. These meanings are usually based on social roles (social identities) as well as idiosyncratic character traits or personal identities (Ashforth and Mael 1989). Identities have been regarded as socially constructed (Haslam 2004) and negotiated, as people convey images of themselves about how they would like to be regarded by others.
Ibarra's work concludes that having a clear and consistent professional identity is fundamental to career success, although it could be argued that, while helpful, many other factors come into the equation. Nonetheless, many of the issues concerning women today, causing some of them to leave corporate life, are around the lack of a sense of authentic identity, a lack of clarity around ‘who am I?’ Inside and outside the work environment, identity development is not just an individualistic concept but a socially constructed one (Ibarra 1999).
Social identity theory. Organizations are social environments and how people familiarize and define themselves within and in relation to these social structures will help explain how they will think, feel and behave at work. Social identity theory, developed in the 1970s by Tajfel (1972) and Turner (1975), concerns inter-group relations. Its fundamental psychological idea is that ‘where people make social comparisons between groups, they seek positive distinctness for their in-groups compared to out-groups in order to achieve a positive social identity’ (Haslam 2004, xix).
Personal identity refers to self-knowledge about one's own attributes. Social identity is the knowledge of the sense of who one is, defined in terms of we rather than I, as part of a social group or clustering. Social identity theory shows how social and psychological factors combine to determine the courses of action that individuals take in order to achieve a positive social identity. If they feel they are in a lower group, their response to this negative assessment and emotion is to endeavour to dissociate and assimilate culturally and psychologically into a higher-status group.
There is some evidence, however, contrary to this, that individuals may internalize both the psychological and behavioural attributions of their in-group as well as the wider social evaluation of selves as inferior and less-deserving (Ashforth and Mael 1989; Ely 1995; Tajfel 1982). An explanation for this is that people have a general cognitive preference to have their expectations about reality supported, as opposed to experiencing cognitive dissonance. So if they accept the negatives about their own group, for example, if women expect men to hold higher positions, some will prefer to engage in low-level tasks and behaviours corresponding to low performance and low status. This is in line with expectancy theory (Vroom 1964), which states that an individual will act in a certain way based on the need to have their expectations met. This will clearly affect women's aspirations and beliefs in terms of their career potential. Psychologically and behaviourally, they will emulate characteristics associated with women as a group, and not with men. What remains unclear is why, with a shared understanding of status relations, some individuals will take this route and others will challenge the status quo – the individual idiosyncrasies that make each individual's interpretation different. Key to this is the individual's belief structures, in particular with regard to social mobility. Tajfel and Turner (1979) proposed strategies that individuals use for self-enhancement. Social/individual mobility beliefs state that anyone can rise to the top. Social change beliefs state that the only prospect for improving oneself lies in action as a group member. Social change beliefs are likely to be dominant when an individual believes themselves to be ‘locked into their group membership’ (Haslam 2004, 25) and feel they must act either to improve or defend their group's status.
Individual mobility is most likely to happen when a group has relatively low status but boundaries are seen to be permeable. The individual can disassociate themselves from other in-group members and work to improve their personal outcomes (rather than work collectively to improve the outcomes for the group). For example, with clear status differences highlighted by a predominance of men in positions of power, in attempting to join her higher-status counterparts, a woman may provide more favourable attributions to the out-group. There is plenty of evidence in scholarly work (Kanter 1977) and popular literature of women ‘acting like men’, as it allows women to feel favourable about themselves, despite the unfavourable evaluation they may give their sex in-group. Women in male-dominated firms will often evaluate other women less favourably in relation to the organization's requirements for success than their counterparts in more sex-integrated firms (Ely 1995). Some women have found that the glass ceiling can be broken by acting as an individual, defecting to become ‘one of the boys’, rather than trying to improve status and treatment of women in general.
Social and gender role theory. According to social role theory of sex and gender differences (Eagly 1987), women are expected to behave in a manner consistent with societal gender roles. There are general beliefs held that men have a higher level of agentic attributes, whereas women have a higher level of communal attributes, although whether men and women have higher levels of these attributes or simply use them more often is another question. Therefore, because leadership is still construed in masculine agentic terms, this presents a substantial barrier to women. Women are experiencing the double bind of incongruity between their gender role and leadership stereotypes. Eagly and Karau's (2002) theoretical paper on gender roles supported the notion that attitudes are less positive towards female than male leaders and potential leaders.
Another effect of the imposition of gender roles on behaviour is on self-regulatory processes. As Ely (1994, 1995) showed, women's social identities in their workplaces reflect the current gender stereotypes, particularly in organizations with low representation of women in senior positions. Women may behave gender stereotypically because of having internalized aspects of gender roles, especially if situational cues make these aspects particularly appropriate. Self-regulatory processes can induce gender roles, and may actually cause the women to become less attracted to top management positions (van Vianen and Fischer 2002). Many women struggle with the issue of balancing the feminine/masculine styles – they need to be sufficiently businesslike and professional to be considered a credible manager and sufficiently feminine not to challenge prevailing assumptions about gender. They compromise their career progression because they appear less confident or powerful, and do not ascend to executive leadership because they are perceived to lack sufficiently agentic behaviours. This is likely to be the result of stereotyping and prejudice. As Eagly and Karau (2002) state, none of this can change without a wider variety of role models.
Supporting Eagly and Karau's work, Heilman and Okimoto (2007) replicate previous work showing that, when performance levels are ambiguous, women are perceived to be less competent than men in male gender-typed work, thus leading to discriminatory practices. And, even when performance is unambiguous, women are then disliked and seen as undesirable bosses. Women who are successful in male domains violate gender stereotypes, and their perceived lack of feminine attributes causes negative reactions. However, if there is clear evidence of the woman's communal traits, which need not be displayed, but can be inferred by roles such as motherhood, these negative reactions can be abated. This study shows that it is possible for successful women to be seen as both agentic and communal, previously considered mutually exclusive, and that it is ‘the women's perceived violation of feminine “shoulds”, not their taking on of masculine “should nots” that underlies and fuels the penalties these women incur for their success’ (p. 91).
There is substantial literature suggesting that women perceived as communal are also presumed incompetent at supposedly male tasks, yet this study shows the double-edge sword, that women in the workplace have felt for years that, without showing those feminine qualities, they are disliked by their colleagues (male and female). Perhaps the call for senior female role models reflects women's search for exemplars of how to demonstrate both the agentic and the communal traits successfully. The current demographic context in many organizations means exemplars are few and far between.
Relational identity theory. The concept of relational identity is useful in explaining how individuals define themselves and construct identities through work relationships. This builds on models such as role theory and social identity theory by focusing on the interpersonal level, in an attempt to integrate the personal and collective levels of identity. Sluss and Ashforth (2007) propose that interpersonal relationships are simultaneously informed and influenced by both person-based and role-based identities. However, they do not address gender. But Collinson (2003) mentions how men also work to construct, negotiate and achieve their masculine identity. The challenges of identity formation at work are not exclusive to women, but men do it in an environment that is predominantly masculine and therefore not so alien.
Sluss and Ashforth's paper theorizes about relational identity and relational identification. The former looks at the nature of one's role relationship, e.g. manager–subordinate. It is the web of relational identities, roles and role incumbents, which form the social system of the organization. Relational identification is the extent to which one defines oneself in terms of that relationship, i.e. how much does one internalize that identity as a partial definition of oneself? This is clearly relevant to how an individual perceives and uses role models. Role-based identity is about the goals, values, beliefs typically associated with that role (Ashforth 2001), regardless of who is enacting the role. The person-based identity is concerned with the traits that define how that individual is enacting that role (e.g. considerate, fair, honest). The relational identity is the interaction of both person-based and role-based identities of two people within a relationship, and therefore draws on the interpersonal level. This brings more depth to the concept of role models as proposed in this review, and we suggest that it could be useful to consider on what level role model relationships are working.
Lewin's field theory. Kreiner et al. (2006) use qualitative studies to look at how individuals in a very demanding occupation (Episcopal priests) actively negotiate the construction of their identities. They discuss the processes used to achieve an optimal balance between over-identification and individuation or under-identification, by differentiating or integrating individual and social identities. This interesting paper creates a conceptual model using Lewin's Field Theory to describe the processes of identity construction. In essence, the paper explicates what the authors see as the various forces that interact to create a ‘field’ or context in which individuals and groups operate. By identifying these forces, Lewin believed we could understand why individuals behave and react as they do (Papanek 1973). This may be a useful model in which to consider the behavioural and other values of role models: the forces that help us understand better why individuals behave/react as they do and what the implications of strengthening or weakening these forces would be in terms of changed behaviour.
Identity management. One of the challenges with trying to understand identity formation for senior women in terms of collective theories, such as social identity theory, is that these women are often so isolated that they do not have a collective of which to feel a member – they are visibly very different from their male colleagues and very removed from any other female colleagues. In such circumstances, it is argued that individuals need to actively engage in identity management in order to deal with ‘identity dissonance’ (Gioia 1998). Identity management is not always just an individual process. As Alvesson and Wilmott (2002) argue, today's organizations attempt to exert power and discipline over individuals by shaping their identities and relationships through ‘identity regulation’. The role of the organization in the construction of identities through role models would seem to be another important area for in-depth research.
Explanations from identity theories. This section has examined several identity theories for further insight into why role models are important in identity construction. Social identity theory explains how individuals seek out role models by comparison that leads to a desire to join the higher status group by emulating its group characteristics. This explains how and why many highly successful women are seen to use masculine-typed behaviours. Social and gender role theory explains that such violation of traditional feminine roles leads to censure from both men and women, and often rejection of the masculine-style women as role models. Relational identity theory highlights the interpersonal elements, as the individual seeks to emulate the higher-status individual rather than the typical group characteristics. Field theory offers a new avenue for research into the forces that influence role model construction, including the organization.
After a decade with little interest, organizational researchers are again beginning to treat role models as an important developmental relationship. However, as this review shows, the extant literature on role models is limited.
Figure 1 shows that the symbolic value of role models needs much deeper level research, as the outcomes box has relied on mainly experimental research with hypothetical ‘role models’ looking at how various characteristics might affect performance, but using undergraduate samples or poor conceptualization of the role model construct. In addition, outcomes for senior women have not been identified. With the exception of Gibson's and Ibarra's studies, there is little empirical research in organizational settings concerning the criteria for choosing role models and how this impacts identity construction, which we argue from this review is likely to affect successful role modelling. Apart from Gibson's career stage study, we found no research on role models at a senior organizational level. More research is needed to develop a better understanding of the process and the relevance of role models and work identity in demographic context.
One of the problems often cited with the psychological study of identity is that it is often treated from a Functionalist perspective (Gioia 1998), i.e. as something that exists and can be tested, as opposed to something socially and symbolically constructed. Gibson (2003) explored the cognitive processes of role modelling but even in his own definitions (e.g. the basis of construing positive role model attributes was ‘the degree to which the individuals think they are sufficiently similar in their environment’ (p. 599)) he explicitly includes the social context. This review suggests that role models play a part in the continuous development of identity transitions in social contexts. Such issues ‘that question the definitional dimensions of identity are a healthy sign of an underdeveloped but high-potential concept’ (Gioia 1998, 24).
From the literature on social theories of identity, identification with a group is an important factor that affects the person's readiness to accept that social category as part of their definition of self. People organize and construe the world around them in ways that reflect the social groups to which they belong. For women at management levels in token or minority demographic contexts, the lack of availability of senior female role models highlights the limited access of women as a social group to top levels of that organization. If an individual can identify strongly with the organization, they may interpret the world and their place within it, in a manner consistent with the organization's values, ideology and culture. This may be a key point when looking at women's common disillusionment with their organization (e.g. the off-ramping described by Hewlett and Luce (2005)), explaining how they do not comfortably share the social identity of their work environment. If they cannot easily identify with the organization's (masculine) culture, perhaps it is at this point that the need for an individual role model becomes more salient. Alvesson and Wilmott (2002) talk about the emotional labour of counteracting imposed identities – which they describe as micro-emancipation (p. 637). Women may find social and emotional support in networks and groups, but similar to Gibson's arduous task of cognitive processing, it is hard work to push back continually against an identity that others assign, if it is not congruent with one's own self-concept.
Social identity theory explains that the individual's belief structures regarding social mobility are the key to why some individuals take action to achieve a more positive social identity, whereas others assimilate the attributes of their social group to their personal detriment. But where do these beliefs stem from and who or what determines whether the desired group boundaries are permeable or not? As hinted in Gibson and Cordova's (1999) paper, perhaps the importance of the role models is mainly symbolic, to help both genders change their cognitive schema of what is possible. The presence of the senior female proves that boundaries are permeable. Gibson would argue that there needs to be some degree of attraction/similarity for the role model to be perceived as ‘available’ and, Ibarra would suggest, for the role model to be used successfully. But perhaps the symbol is sufficient to provide permission to change the gender stereotype role.
Kossek et al. (2003) show the challenges of taking a purely structuralist approach of just recruiting more women, ‘throwing women at the situation’, as the result was that the male majority group started to acknowledge the competition as a threat to their own careers. In the UK, underlying structural or attitudinal factors may still tend to go unnoticed and unaddressed by equal opportunities legislation and relevant equality organizations such as Opportunity Now. Such initiatives are still based very much on the liberal feminist idea that, if opportunities are made available and if women try hard enough, inequalities in the workplace can be largely overcome. This can have negative connotations, as it places the responsibility squarely on the women's shoulders, emphasizing deficiencies in their own behaviour or attributes. Men feel comfortable with these prevailing attitudes and norms, which they perceive as gender neutral (Simpson 1997). Women are more aware of barriers, but often find them hard to define, as individual successful women are held up as examples of why the system is not at fault. But all too often the example is not an attractive one for women aspiring to get to the top, in terms of either the style or behaviours used, the similarity to them, or the sacrifices made to get there – hence the requirement for more relevant and attractive role models.
Having established that role models are used in identity formation, one of the biggest challenges facing the authors in trying to establish why role models are important for senior women is the paucity of literature that links or integrates the various findings and levels of identity work. This is something that Cornelissen et al. (2007) noted in a recent article looking at the disparate areas of social identity, organizational identity and corporate identity. In the emerging field of work identity, there are a number of competing literature areas: the social identity literature which focuses on the cognitive psychological aspects of identity formation; the literature on human assets, social capital and networks; and the more sociological investigations of structures, demographics and institutional behaviour. The recent research on relational identity attempts to integrate personal and social aspects. But we feel that work on role models could go further in combining these literatures and show how micro-level cognitive processes, affected by macro-level demographics, feed into larger systems and the underlying mechanisms of organizational structures.
A number of surveys recently cited the lack of senior female role models as a major barrier to women's career success. This led the authors of this review to one main line of questioning:
• Are female role models important for senior women, and if so why?
In addressing this question, we also added to understanding of a related question:
• Why are many of the few senior women rejected as role models by other women?
A comprehensive literature search of research on role models, as distinct from mentors or coaches, conducted in organizational settings produced a very small sample of studies. The behavioural value of role models has been well documented (Bandura 1977b; Gibson 2003; Ibarra 1999). This review has indicated how role models are used, drawing on Gibson's work on the cognitive and structural aspects of role modelling, and Ibarra's work on how behavioural role models guide the development of provisional selves. These studies in particular provide a partial answer to why role models are important: they are needed for identity construction. Our initial review of that small body of role model literature led to the inclusion of two related areas – organizational demographics as the contextual factor affecting the availability of role models, and work identity formation as a key explanatory factor behind the link between the lack of senior female role models and the lack of career progression. Organizational demography research has highlighted that women struggle to find suitable role models in organizations where the demographic profile is not sex balanced, and we have argued that this difficulty is likely to become stronger as women reach senior levels.
Work environments in the UK have changed in the past 15 years, and a new generation of female middle managers have senior and executive management positions within their sights. They are armed with the qualifications and experiences required, and hence there should be few barriers to their success. So why are their plans thwarted and why do they cite the lack of senior female role models as critical? Both men and women need to move towards changing their stereotypical cognitive schema of gender in the workplace, and this may be the key value of role models.
Building on the literature above, research is needed to ascertain further the value of female role models for women managers. As well as behavioural value, symbolic value has been mentioned but not explored in depth, but there may be other values that help explain why female role models are important for senior women. We need to understand whether the availability, proximity and successful use of role models are key antecedents to the cognitive processes critical to identity formation and crucial to the career success of senior female managers. The role of the organization in the identification and promotion of female role models is also new and interesting – what, if any, influence do they have? New technology such as corporate web pages and new interventions such as women's corporate networks allow for role models to be highlighted, but how are they selected?
The contribution of this paper is on three levels. First, it highlights the need to take a more integrated approach to the study of role models and work identity formation, pulling together literatures on organizational demography, the cognitive construal of role models and the importance of behavioural and symbolic role models for successful work identity formation in senior women. Second, it provides insight into why the few women at the top of organizations are often rejected as role models. Finally, it recognizes the need to increase the conceptual understanding of the symbolic value of role models.
1FTSE 100 is the Financial Times/London Stock Exchange listing of the 100 largest firms.