1Women are still scarce in top leadership positions in wealthy industrialized societies in spite of substantial gender equity in education and the entrance of an increasing number of women into high prestige professions and managerial jobs long dominated by men. Cross-national variations exist and women from ethnic and racial minorities are less well represented at the top than women from the dominant racial group, usually white, but the general pattern holds everywhere (e.g. Wright et al., 1995). Angela Merkel, the German prime minister, stands out as the only woman in pictures of gatherings of top world leaders. In 2005, eight (1.6%) Fortune 500 companies in the US had female Chief Executive Officers (CEO); this number had not varied substantially over at least 10 years (Catalyst, 2007). Of Fortune magazine’s Global 500 companies, just 1% had women CEOs (Eagly and Carli, 2007). Women of color filled 3.1% of Fortune 500 corporate board positions, while women in general filled 14.6% of such board positions (Catalyst, 2007). The pattern holds even for women predominate fields, such as nursing or elementary school teaching, in which men move more quickly than women into administrative and leadership positions (Williams, 1992; Budig, 2002).

2This phenomenon has been called the “Glass Ceiling.” Women seem to move up organizational ladders, but only a very few actually make it to the most powerful positions. This reality is important for several reasons. First, the difficulties in explaining the persistence of this pattern expose gaps in our understanding of the stubborn persistence of gender and often race/ethnic, inequalities. In the last 30 to 40 years in the United States, as well as in other countries, a growing number of women have been in the “pipe line” to higher positions, but the proportion of women still declines at each higher level of most organizations. Discriminatory practices and beliefs have been identified, discussed and litigated for at least that long. Is something else happening? What prevents the full success of efforts to eliminate these patterns of discrimination?

3The existence of a gender imbalance at the top of most organizations is important for another reason: the absence of women at the top probably makes work life more difficult for women further down in the organizational ranks. When women fill the top jobs, other women may benefit in some ways. As research is beginning to show (e.g. Cohen and Huffman, 2007; Hultin and Szulkin, 2003), when women occupy high managerial positions, the gender wage gap at lower hierarchical levels tends to be lower than when women are not present at the top. Other manifestations of gender discrimination may also be reduced when women hold managerial positions (Cohen and Huffman, 2007). Thus, breaking the “glass ceiling” can contribute to more general gender equality goals. This statement should be treated with caution: changes are apt to be modest unless women at the top can go beyond interventions in individual cases and also change larger economic and organizing processes that create inequalities.

1. Inequality regimes

8All organizations have inequality regimes, defined as loosely interrelated practices, processes, actions, and meanings that result in and maintain class, gender and racial inequalities within particular organizations. The ubiquity of inequality is obvious: managers, executives, leaders, department heads have much more power and higher pay than secretaries, production workers, students, or even professors. Even organizations that have explicit egalitarian goals develop inequality regimes over time, as considerable research on egalitarian feminist organizations has shown (e.g. Ferree and Martin, 1995; Scott, 2000).

9I define inequality in organizations as systematic disparities between participants in power and control over goals, resources, and outcomes; in work place decision-making such as how to organize work; in opportunities for promotion and interesting work; in security in employment and benefits; in pay and other monetary rewards; and in respect and pleasures in work and work relations. Organizations vary in the degree to which these disparities are present and in how severe they are. Equality rarely exists in control over goals and resources, while pay and other monetary rewards are usually unequal. Other disparities may be less evident or a high degree of equality might exist in particular areas, such as employment security and benefits, particularly in public sector organizations and organizations with strong labor unions.

10Inequality regimes are highly various in other ways; they also tend to be fluid and changing. Practices that at one time in one place limit the access of white women and people of color to high positions may be different at other times or in other places. These regimes are linked to inequality in the surrounding society, its politics, economic conditions, history and culture.

11In the following sections I discuss the characteristics of inequality regimes. An examination of inequality patterns in a particular organization can begin with an analysis of these characteristics and how they function in the particular case. These characteristics include the bases of inequality, the shape and degree of inequality, organizing processes that create and recreate inequalities, the invisibility of inequalities, and the legitimacy of inequalities. Inequality processes shape work organizations and the experiences of employees at every organizational level. In this discussion, I highlight the practices, processes and beliefs that influence women’s possibilities of moving into top jobs.

2. The components of inequality regimes

2.1. The bases of inequality

12The bases for inequality in organizations vary, although class, gender and race processes are usually present. Class, as I use the term, refers to enduring and systematic differences in access to and control over production of goods and services as well as the resources for provisioning and survival (Nelson, 1993; Acker, 2006a). Those resources, primarily money and property in wealthy industrial societies, are unequally distributed through wages, salaries and other forms of transfer. Much of this distribution occurs in and through organizations. Thus, class differences are produced as the essential processes of production and distribution take place; the production of class is intrinsic to employment and to most organizations. In large organizations, hierarchical positions are congruent with class divisions in the wider society. The CEO of the large corporation operates at the top of the national and often global class society. In smaller organizations, the class structure may not be so congruent with societal-wide class relations, but the owner or the boss still has class power in relations with employees. Class processes are gendered and racialized (Acker, 2006b). Efforts to increase the representation of white women and people of color in top management are, in effect, efforts to reduce the effects on class practices of gendered and racialized power and ideology. Race is often ignored in these efforts, which are usually efforts to make white women and white men equal in access to class power.

14Race refers to socially defined differences based on physical characteristics, culture, and historical domination and oppression, justified by entrenched beliefs. Ethnicity may accompany race, or stand alone, as a basis for inequality. Race, too, has often been integrated into class hierarchies, but in different patterns than gender. In the US, African Americans, Native Americans, Asians and Hispanics have long histories of exclusion from many jobs and from positions of power. While these exclusions have diminished, each racial/ethnic group has a different history and today faces different patterns of inequality, that vary across gender and class divisions in particular organizations (Glenn, 2002). Although total exclusion of racial and ethnic minorities from higher level jobs no longer exists, inclusion is still rare (Smith, 2002).

15Other differences are sometimes bases for inequality in organizations. The most important, I believe, is sexuality. Heterosexuality is assumed in many organizing processes and in the interactions necessary to these processes. The secretary is or was the “office wife” (Kanter, 1977). In most work places, women and men are expected to behave in heterosexually appropriate ways, although the definition of “appropriate” may differ. Homosexuality is disruptive of organizing processes because it flouts the assumptions of heterosexuality. It still carries a stigma that produces disadvantages for lesbians and gays. Other bases of inequality are religion, age and physical disability. Again, in the not too distant past, having the wrong religion such as being a Jew or a Catholic could activate discriminatory practices. Today, a Middle Eastern origin or being a Muslim may have similar consequences.

2.2. Shape and degree of inequality

2.2.1. The steepness of hierarchy

16The steepness of hierarchy is one dimension of variation in the shape and degree of inequality. The steepest hierarchies are found in traditional bureaucracies in contrast to flat organizations with team structures, in which most, or at least some, responsibilities and decision-making authority are distributed among participants. Between these polar types are organizations with varying degrees of hierarchy and shared decision-making. Hierarchies are usually gendered and racialized: both racial and gender disparities in authority are the greatest at the higher levels of organizational hierarchies (Smith, 2002). Some research shows that flat team structures provide professional women more equality and opportunity than hierarchical bureaucracies (Smith-Doerr, 2004), but only if the women function like men. One study of engineers in Norway (Kvande and Rasmussen, 1994) found that women in a small, collegial engineering firm gained recognition and advancement more easily than in an engineering department in a big bureaucracy. However, the women in the small firm were expected to put in the same long hours as their male colleagues and to put their work first before family responsibilities.

17Other research suggests (e.g. Barker, 1993) that team organized work may not reduce gender inequality. Racial inequality may also be maintained as teams are introduced in the workplace (e.g. Vallas, 2003). While the organization of teams is often accompanied by drastic reductions of supervisors’ roles, the power of higher managerial levels is usually not changed: class inequalities are only slightly reduced (Morgen et al., 2008).

2.2.2. The degree and pattern of segregation

18The degree and pattern of segregation by race and gender is another aspect of inequality that varies considerably between organizations. Gender and race segregation of jobs is complex because segregation is hierarchical across different class levels of an organization, across jobs at the same level, and within jobs (Charles and Grusky, 2004). Occupations should be distinguished from jobs: occupation is a type of work, a job is a particular cluster of tasks in a particular work organization. For example, Human Resources Director is an occupation; the Director of Human Resources at Boeing in Seattle is a job. More statistical data are available about occupations than about jobs, although “job” is the most relevant unit for examining segregation in organizations. However, research indicates that “gender segregation at the job level is more extensive than gender segregation at the level of occupations” (Wharton, 2005, p. 97). In addition, even when women and men “are members of the same occupation, they are likely to work in different jobs and firms” (Wharton, 2005, p. 97). Racial segregation also persists, is also complex, and varies by gender (Padavic and Reskin, 2002).

19Jobs and occupations may be internally segregated by both gender and race: what appears to be a reduction in segregation may only be its reconfiguration. Reconfiguration and differentiation have occurred as women have entered old male-dominated occupations (Reskin and Roos, 1990). For example, the occupational category “Managers” is very broad: in the US, the substantial progress toward gender equality in this category masks sex segregation internal to the category. For example, “line” positions managing central production functions are more often filled by men than by women, while “staff” positions, such as human relations managers, tend to be filled by women (Catalyst, 2007). In addition women’s statistical representation in managerial jobs may have been increased through a process of reclassification in which jobs previously designated as non-managerial have been converted to a managerial category. This seems to have happened, in particular, in clerical and service occupations, occupations with a high proportion of women. For example, a recent study of former welfare clients (Morgen et al., 2008), included interviews with a number of women working in retail establishments who were classified as “managers” although their management responsibilities were minimal and their pay was very low, on a par with pay in fast food restaurants. One conclusion from these disparate bits of evidence is that many women who are classified as “managers” never get close to the upward paths that lead to top positions.

2.2.3. The size of wage differences

2.2.4. The severity of power differences

3. Organizing processes that produce inequality

22Organizations vary in the practices and processes that are used to achieve their goals; these practices and processes also produce class, gender and racial inequalities. Considerable research exists exploring how class or gender inequalities are produced, both formally and informally, as work processes are carried out (e.g. Willis, 1977; Burawoy, 1979; Acker, 1989, 1990; Cockburn, 1985). Some research also examines the processes that result in continuing racial inequalities (e.g. Royster, 2003; Glenn, 2002; Bell et al., 2001). These practices are often guided by textual materials supplied by consultants or developed by managers influenced by information and/or demands from outside the organization. To understand exactly how inequalities are reproduced, it is necessary to examine the details of these textually-informed practices. Inequality producing practices are often guided by and justified by powerful gendered and racialized images and beliefs in which white, African American, Hispanic, Asian, Native American, and other categories of women and men are all seen as suited for some jobs and not for others. Included in these sets of images is the view that white women and women and men of color are often inappropriate for or unprepared for performing top managerial tasks (Ridgeway, 2001; Smith, 2002; Wajcman, 1998).

3.1. Organizing the general requirements of work

24“Family friendly” policies sometimes offer managers and professionals ways to accommodate work demands to home responsibilities. These policies are less available to lower level workers with family responsibilities, increasing the strains on low wage working mothers in particular. Family friendly policies may include part-time work, flexible hours, working at home and assistance with child care arrangements. However, such modifications of work expectations are often arbitrary, short-term or unavailable when needed (Stone, 2006). Thus, family friendly policies provide only temporary relief for some people from the male model of organizing.

25The lack of fit between paid work and the necessary but unpaid work of reproduction may stall the early careers of potential women managers, especially after they have children. As Pamela Stone reports, such women often find themselves marginalized and stigmatized by their managers and are likely to be defined by their superiors as not committed to a career, thus losing their opportunities for upward advancement (Stone, 2006). This is particularly the case for women who attempt to work part-time. The use of family friendly policies, primarily by women when they have young children, may actually increase gender inequalities in organizations (e.g. Glass, 2004). Such measures may reinforce, not undermine, the male model of organizing by defining those who conform to it as serious, committed workers and those who don’t (those who use family friendly programs) as rather peripheral and probably unworthy of promotions and pay increases (e.g. Hochschild, 1997; Stone, 2006).

26In summary, the organization of work and work expectations based on the assumption of an unencumbered worker are important in maintaining gender inequality in organizations and, thus, the unequal distribution of women and men in organizational class hierarchies, especially in upper management.

3.2. Organizing jobs and class hierarchies

27Techniques also vary for organizing class hierarchies inside work organizations. Bureaucratic, textual techniques for ordering positions and people are constructed to reproduce existing class, gender and racial inequalities (Acker, 1989). One example of such techniques, a large job classification system, comes from my study of a pay equity project (Acker, 1989). Job classification systems describe job tasks and responsibilities and rank jobs hierarchically. Jobs are then assigned to wage categories with jobs of similar rank in the same wage category. This study found that the bulk of sex-typed women’s jobs were in the clerical/secretarial area and included thousands of women workers, grouped into four large categories at the bottom of the ranking and assigned to the lowest wage ranges. A new evaluation of the clerical/secretarial categories showed that many different jobs with different tasks and responsibilities, some highly skilled and responsible, had been lumped together. In contrast, sex-typed men’s jobs were spread over a wider range of categories, reflecting wide differences in skill and responsibility. One result was that women’s skilled jobs, hidden in these large categories, were paid much less than men’s skilled jobs.

28Structural barriers to the movement of women into management are embedded in job and wage classification systems that reproduce hierarchy. An instance of this process was revealed in the comparable worth study described above as researchers discussed the project with representatives of the large consulting firm that provided the job evaluation system (Acker, 1989). These representatives would not let the job evaluation committees alter their system to compare the responsibilities of managers’ jobs with the responsibilities of the jobs of their administrative assistants. Often, it was observed, managers were credited with responsibility for tasks actually carried out by their assistants. The assistants did not get credit for these tasks in the job evaluation system because their job descriptions did not include the tasks. This definition of their job contributed to their relatively low status and low wages. But, if managers’ and assistants’ actual tasks could never be compared, no adjustments for inequities could ever be made. Moreover, the assistant job was never seen as a stepping-stone to management, even when many of the tasks of these assistant jobs were managers’ tasks. The hierarchy was inviolated in this system: a women’s job as a subordinate could never lead to promotion to an entirely different category, management.

30In sum, class hierarchies in organizations, with their embedded gender and racial patterns, are constantly created, renewed, and sometimes reduced, through organizing practices. Gender and sometimes race, in the form of restricted opportunities and particular expectations for behavior, are reproduced as different degrees of organizational class hierarchy are also reproduced in everyday interactions and bureaucratic decision-making.

3.3. Recruitment, hiring and promotion

31Recruitment, hiring and promotion are processes of finding the workers most suited for particular positions. From the perspectives of employers, the gender and race of existing job holders at least partially define who is suitable, although prospective coworkers may also do such defining (e.g. Enarson, 1984). Images of appropriate gendered and racialized bodies influence perceptions and hiring. White bodies are often preferred, as a great deal of research shows (e.g. Royster, 2003). Female bodies are appropriate for some jobs; male bodies for other jobs.

32At the managerial level, recruitment, hiring and promotion processes are informed by images of the successful manager. This image is stereotypically masculine; the successful organization and the successful leader share many of the same characteristics, such as strength, aggressiveness and competitiveness (Kanter, 1977; Wajcman, 1998). Such stereotyping has been documented for over 30 years and continues into the 21st century, constituting a major barrier to women’s entry into top level management, particularly in the private sector (Catalyst, 2007; Ridgeway, 2001), but also in academia and in the public sector. Women do not fit the image of the (masculine) leader. Thus, women managers face a gendered double bind: they are either too masculine and assertive or they are too feminine and soft (e.g. Kanter, 1977). If women’s behavior seems too assertive and masculine they may be seen as competent but not likeable; if their behavior is too feminine, they may be seen as likeable but incompetent (Eagly and Carli, 2007). Women enacting power violate conventions of relative subordination to men, risking the label of “witches” or “bitches”. The 2007 Catalyst study of US and European business leaders, cited here, also found that women leaders have to work harder than men at the same corporate levels to be perceived as equally competent and to receive the same levels of rewards. This stereotyping is often unrecognized, even as it enters into assessments of candidates for hiring or promotion.

33Gendered and racialized stereotyped images contribute to and help to perpetuate gender and race typing and segregation of jobs at all levels of class hierarchy (see, e.g. Brown and Misra, 2003). At management levels, there are consequences for women’s opportunities to move into top positions. For example, the gendered practice of assigning men to line positions and women and some minority men to staff positions, discussed above, is one reason that few white women or people of color reach the top. Line positions manage the core processes of a company and top managers are usually recruited from those who have had line experience. In contrast, staff positions, such as Human Resources Director, facilitate and assist the core processes. These positions are not stepping stones to the top.

34Elites tend to reproduce themselves through inclusion of those like themselves and exclusion of others (Reskin, 2003; Smith, 2002). Certain hiring and promotion practices maximize the possibility that those chosen will be similar to those doing the choosing. Hiring through social networks is one of the ways in which gender and racial inequalities are maintained in organizations, and top managerial positions are disproportionately allocated to white men (Reskin and McBrier, 2000). Social networks are usually composed of people with similar characteristics and interests. Affirmative Action programs altered hiring practices in many organizations, requiring open advertising for positions and selection based on gender and race neutral criteria of competence, rather than selection based on an old boy (white) network. These changes in hiring practices in the US contributed to the increasing proportions of white women and people of color in a variety of occupations, including lower and mid-level managers and professionals, particularly during the 1970s (Reskin, 1998; Tomaskovic-Devey et al., 2006). However, criteria of competence do not automatically translate into gender and race neutral selection decisions. “Competence” involves judgment: the race and gender of both the applicant and the decision-makers can affect that judgment, resulting in decisions that white males are the more competent, “natural leaders” more suited to the managerial job than are others (Catalyst, 2007). In addition, enforcement of Affirmative Action almost ended during the 1980s, reducing the pressure for more open and equal hiring. Thus, gender and race as a basis for hiring or a basis for exclusion have not been eliminated in many organizations, as continuing patterns of segregation attest.

35At top management levels, hiring is usually done by corporate boards or by other top managers. Given the gender and race composition of these top corporate levels, it is not surprising that newly selected leaders tend to have the gender and race characteristics of the decision-makers.

3.4. Wage setting and supervisory practices

36Wage setting and supervision are class practices. They determine the division of surplus between employees and management and control the work process and workers. Gender and race affect assumptions about skill, responsibility, and a fair wage for jobs and workers, helping to produce wage differences (Figart et al., 2002). Considerable evidence exists that white women and people of color earn less than white men at all levels of organizational hierarchies, including management jobs, with discrepancies increasing as the level of authority rises (Smith, 2002).

37Wage setting is often a bureaucratic organizational process, integrated into the processes of creating organizational class hierarchies, as I described above. Many different wage setting systems exist, many of them producing gender and race differences in pay. Differential gender-based evaluations may be embedded in even the most egalitarian appearing systems. Individual wage setting may result in greater inequality than group based wage setting that uses formalized criteria (Acker, 1991). The salaries of managers, especially those in top jobs, are usually set through individual negotiations, processes that can be heavily influenced by images of women managers as less competent than men.

38Supervisory practices also vary across organizations. Supervisory relations may be affected by the gender and race of both supervisor and subordinate, in some cases preserving or reproducing gender or race inequalities. For example, in Swedish banks in the 1980s, women and men in the same entry level job classification were assigned to different duties by their supervisors: men had duties that led to promotion, women did not (Acker, 2006c [1994]). At that time, this supervisory practice was a major step in the sequence of practices that kept the management of Swedish banks almost entirely in the hands of men. Supervisors probably shape their behaviors with subordinates in terms of race and gender in many other work situations, influencing in subtle ways the existing patterns of inequality. Much of this can be observed in the informal interactions of workplaces.

3.5. Informal interactions while “doing the work”

40As women and men go about their every day work, they routinely use gender, race, and class based assumptions about those with whom they interact, as I discussed above. Body differences provide clues to the appropriate assumptions, followed by appropriate behaviors. What is appropriate varies, of course, in relation to the situation, the organizational culture and history, and the standpoints of the people judging appropriateness. For example, managers may expect a certain class deference or respect for authority that varies with the race and gender of the subordinate; subordinates may assume that such positions require deference and respect, but also find these demands demeaning or oppressive. Jennifer Pierce in a study of two law firms showed how both gendered and racialized interactions shaped the organizations’ class relations (Pierce, 1995): women paralegals were put in the role of supportive, mothering aides, while men paralegals were cast as junior partners in the firms’ business. African American employees, primarily women in secretarial positions, were acutely aware of the ways in which they were routinely categorized and subordinated in interactions with both paralegals and attorneys.

41Black women managers are subjected to both racism and sexism on the job, facing many more, and more complex, negative situations in work place interactions than those faced by white women managers (Bell et al., 2001). The interaction practices that recreate gender and racial inequalities are often subtle and unspoken, thus difficult to document (Krefting, 2003). White males may devalue and exclude white women and people of color by not listening to them in meetings, by not inviting them to join a group going out for a drink after work, by not seeking their opinions on workplace problems. Other practices, such as explicit racism or sexual harassment, are open and obvious to the victim, but not always obvious to others. The subtle and repeated experiences of exclusion, denigration and indifference may accumulate in ways that lead some women to renounce managerial ambitions (Krefting, 2003).

42Masculine-stereotyped patterns of on-the-job behavior in team organized work may mean that women must make adaptations to expectations that interfere with family responsibilities and with which they are uncomfortable. In a study of high-level professional women in a computer development firm, Joanne Martin and Debra Meyerson found that the women saw the culture of their work group as highly masculine, aggressive, competitive and self-promoting (Martin and Meyerson, 1998). The women had invented ways to cope with this work culture, which they experienced as alien to their usual way of working; they felt that they were partly outsiders who did not belong.

4. Organizational impediments to change

4.1. The invisibility of systematic inequalities

43The invisibility of systematic inequalities often obstructs efforts to achieve greater equality. Visibility of inequality, defined as the degree of awareness of inequalities, varies in different organizations. Lack of awareness may be intentional or unintentional. Managers may intentionally hide some forms of inequality, as in Swedish banks in the late 1980s (Acker, 1991). Bank workers said that they had been told not to discuss their wages with their co-workers. Most seem to have complied, partly because they had strong feelings that their pay was part of their identity, reflecting their essential worth. Some said they would rather talk about the details of their sex lives than talk about their pay.

45Class also tends to be invisible. It is hidden by talk of “management, leadership, or supervision” among managers and those who write and teach about organizations from a management perspective. Workers in lower level, non-management positions may be very conscious of inequalities, although they might not identify these inequities as related to “class.” Race is usually evident, visible, but segregated, denied and avoided by members of the dominant group. Ella L.J. Bell and Stella M. Nkomo provide numerous examples of how Black women managers experienced racist insults that white colleagues and managers claimed were just mistakes or jokes (Bell et al., 2001). Another basis of inequality, sexual orientation, is almost always invisible to the majority who are heterosexual. Heterosexuality is simply assumed, not questioned.

4.2. The legitimacy of inequalities

46The legitimacy of inequalities also obstructs efforts at change. Legitimacy varies between types of inequality and between organizations. Some inequalities, such as power inequalities between upper and lower level managers or class inequalities between managers and non-managers, seem to be accepted as normal and legitimate aspects of organizational life. Beliefs in the efficacy of markets and the individual as an economic actor responsible for his/her own economic success tend to legitimate inequality as simply the outcome of individual choice. Thus, white women and people of color who rise to powerful positions can do so without questioning the legitimacy of the class inequalities involved. The legitimacy of gender and race inequalities is revealed by the ways in which they were taken for granted until strong social movements challenged that legitimacy in the US, leading to equal opportunity legislation and Affirmative Action programs. Uneven, even stalled, change processes attest to continuing and widespread belief in the legitimacy of inequality.

4.3. Mechanisms of control and compliance

47Mechanisms of control and compliance function to maintain inequalities. Organizational controls are, in the first instance, class controls, directed at maintaining the power of managers, ensuring that employees act to further the organization’s goals, and that they accept the system of inequality. Gendered and racialized assumptions and expectations are embedded in the form and content of controls and in the ways in which they are implemented. Controls are made possible by hierarchical organizational power, but they also draw on power derived from hierarchical gender and race relations. They are diverse and complex, and they impede changes in inequality regimes. Mechanisms for exerting control and achieving compliance with inequality vary. Organization theorists have identified many types of control, including direct controls, unobtrusive or indirect controls, and internalized controls. Direct controls include bureaucratic rules and various punishments for breaking the rules. Rewards are also direct controls. Wages, because they are essential for survival in completely monetized economies, are a powerful form of control (Perrow, 2002). Coercion and physical and verbal violence are also direct controls often used in organizations (Hearn and Parkin, 2001). Unobtrusive and indirect controls include control through technologies, such as monitoring telephone calls or time spent on-line, or restricting information flows. Selective recruitment of relatively powerless workers can be a form of control (Acker and Van Houten, 1974). Recruitment of illegal immigrants who are vulnerable to discovery and deportation, or recruitment of women of color who have few employment opportunities and thus will accept low wages are examples of this kind of control that preserves inequality.

5. Can inequality regimes change?

49Inequality regimes change as organizing goals, culture, technology, the pool of potential employees and government equality laws change. Thus, previously impenetrable glass ceilings have been breached, as the data cited earlier indicate. Inequality regimes can also be intentionally challenged through change projects. However, change is difficult and change efforts often fail. One reason is that owner and managerial interests and the power those interests can mobilize, usually outweigh the interests of those who suffer inequality. Even where no obvious economic interests are threatened by changes, men managers and lower level employees often insist on maintaining ongoing organizing patterns that perpetuate gender and race inequalities. White masculine identity may be tied to even small relative advantages in workplace power and income (Acker, 1989). Advantage is hard to give up: increasing equality with devalued groups can be seen and felt as an assault on dignity and masculinity. Several studies have shown that these complicated motives on the part of white men, in particular, can scuttle efforts at organizational change, even when top management is supporting such change. For example, Cynthia Cockburn analyzed the multiple ways that men, particularly middle managers, resisted equality efforts in four British organizations in spite of top-level support for these efforts (Cockburn, 1991). Opposition to the promotion of women at the middle management level can significantly reduce the pool of women in line for higher level jobs, increasing the “glass ceiling” effect.

50Other change projects illustrate difficulties in achieving greater representation of women in top management levels because of the resistance fostered by the fusion of gendered identities and workplace organizing practices. For example, Robin J. Ely and Debra E. Meyerson describe a change project aimed at discovering why a company had difficulty retaining high-level women managers and difficulty increasing the proportion of women in upper management (Ely and Meyerson, 2000). The researcher/change agents documented a culture and organizing practices at the executive level that rewarded stereotypical “heroic” male problem solving behaviors, tended to denigrate women who attempted to be heroes and failed to reward the mundane organization building most often done by women. Although members of the management group could see that these ways of behaving were dysfunctional for the organization, they did not comprehend the links between these organizing practices, gender, and the under representation of women. In their eyes, the low representation of women in top jobs was still due to the failure of individual women, not to system processes.

52These interventions also fail to address other underlying processes of inequality regimes, in particular the masculine unencumbered worker model of organizing.

6. Conclusion

53In this paper, I have suggested the idea of “Inequality Regimes”, interlinked organizing processes that produce patterns of complex gender, race and class inequalities. This is an effort to develop a conceptual strategy for analyzing how persistent inequalities actually come about through the everyday practices of organization participants. This approach can be used to identify where in a particular organization’s practices the “glass ceiling” effect is produced. An analysis of an inequality regime has a number of components:

• the bases of inequality such as gender, class and race;

• the shape and degree of inequalities including the steepness of hierarchy, the degree of segregation, the size of wage differences, and the severity of power and authority differences;

• the organizing processes that produce inequality, including the general organization of work based on the assumption of an unencumbered worker, organizing jobs and hierarchies, recruitment/hiring/promotion, wage setting/supervision, and informal interactions in doing the work;

• the organizational impediments to change, including the visibility of inequalities to organization participants, the legitimacy of inequalities to participants, and mechanisms of control and compliance that preserve inequalities.

54Beliefs, images and stereotypes based on gender, race and class shape actions, policies and practices throughout the components of inequality regimes.

55The processes and patterns that constitute inequality regimes vary in different organizations, but there are also commonalities that, across a national population of organizations, result in statistical pictures of gender, race, and class inequality. In the United States at the present time, almost all organizations have two characteristics that rarely vary: class inequality, inflected through gendered and racialized beliefs and practices, is the normal and natural bedrock of organizing and white males are the normal and natural top leaders.

56The “glass ceiling” is a consequence of varying practices that occur across organizational structures and across the career history of white women and women and men of color who do and who don’t make it into upper level management. These practices include the organization of work on an unencumbered worker model, selective hiring and promotion, gender and race segregation of jobs in which “women’s jobs” do not provide mobility channels or experience that prepares for upper management, and discrimination against women as they begin to bear children. These practices are supported by deeply embedded images of men as natural leaders and women as unsuitable for leadership, but unacceptable if they attempt to show stereotypical, masculine leadership behaviors. Thus, women managers face a double bind that places them in a more ambiguous situation than that experienced by men.

57Greater equality inside organizations is difficult to achieve during a period, such as the early years of the 21st century, in which employers are pushing for more inequality in pay, medical care, and retirement benefits and are using various tactics, such as downsizing and outsourcing to reduce labor costs. Another major impediment to change within inequality regimes is the absence of broad social movements outside organizations agitating for such changes. In spite of all these difficulties, efforts at reducing inequality continue. The under representation of white women and people of color in top jobs is still there, but assaults on the strength of the “glass ceiling” are also still strong: the outcome has not been decided.