Amy S Wharton. Handbook of Social Problems: A Comparative International Perspective. Editor: George Ritzer. Sage Publication. 2004.
Gender is a central organizing principle of social life in virtually all cultures of the world. But what is gender and how exactly is it expressed? Answers to these questions have proliferated in recent years (Hawkesworth 1997). Early theorists treated gender as strictly an individual characteristic—as an attribute of people, acquired through socialization and embedded in personality or identity. This view also tended to recognize a clear distinction between sex, defined as the biological or genetic aspects of maleness and femaleness, and gender, which represented the socially defined meanings attached to sex categories (Acker 1992). Though some still adhere to these views, most contemporary gender scholars endorse a view of gender that is more multilayered and complex (Acker 1992; Hawkesworth 1997; Ridgeway and Smith-Lovin 1999; Risman 1998). For example, Ridgeway and Smith-Lovin (1999:192) conceive of gender as a “system of social practices”; this system creates and maintains gender distinctions and it “organizes relations of inequality on the basis of [these distinctions].” In this view, the gender system involves the creation of both differences and inequalities.
Three features of this more recent definition are important to keep in mind. First, gender is as much a process as a fixed state. This implies that gender is being continually produced and reproduced. Stated differently, we could say that gender is enacted or “done,” not merely expressed. Second, gender is not simply a characteristic of individuals, but occurs at all levels of the social structure. This is contained in the idea of gender as a “system” of practices that are far-reaching, interlocked, and exist independently of individuals. Gender thus is a multilevel phenomenon (Risman 1998). This insight enables us to explore how social processes, such as interaction, and social institutions, such as work, embody and reproduce gender. Third, this definition of gender refers to its importance in organizing relations of inequality. Whether gender differentiation must necessarily lead to gender inequality is a subject of debate. As a principle of social organization, however, gender is one critical dimension upon which social resources are distributed.
These shifts in the meanings of gender have had considerable impact on how gender inequality has been understood. While previous approaches tended to trace gender inequality to differences in the characteristics of women and men as individuals, more recent views emphasize the roles played by social interaction patterns and institutions. More important, however, by treating gender as a multilevel phenomenon, gender theory and research on inequality has had to address the ways that these levels interrelate. These interrelations play an important role in the reproduction of gender inequality and the possibilities for challenging the gender order.
One inadvertent consequence of an individualist view of gender is that women and men have often been portrayed as either villains or victims—oppressing, exploiting, or defending against each other. While inequality does not just happen, how it happens is more complex than this. Just as gender must be viewed as not simply a property of individuals, so too, gender inequality must be understood as the product of a more complex set of social forces. These may include the actions of individuals, but they are also to be found in the expectations that guide social interaction, the composition of social groups, and the structures and practices of the institutions. These forces are subject to human intervention and change, but are not always visible, known, or understood. They are subtle, may be unconscious, and are reproduced often without conscious intent or design.
This chapter begins with a brief discussion of the shift from individual to social relational and institutional accounts of gender. Next, I examine the implications of this shift for theory and research on gender inequality. In particular, I present several alternative “social relational” approaches to gender inequality, followed by a discussion of a gendered institutions perspective. In both sections I discuss some of the key research on gender inequality that has been informed by each perspective. In the final section, I briefly explore directions for future research and debate.
Individuals, Interaction, and Institutions: Changing Conceptions of Gender
Gender theory and research can be generally divided into three broad frameworks, each with a different view of where the “sociological action” is with respect to gender. For some, this action resides mainly in individuals—their personalities, traits, emotions, or identities. Proponents of these views thus focus their attention on the individual attributes of women and men, often seeking to identify areas of gender difference.
By contrast, others argue that students of gender should focus less on individuals and more on interaction and social relations. Rather than assuming that people possess a relatively stable set of traits and abilities that they bring to organizations, social relational approaches argue that people actively construct their reactions and behaviors in response to features of the social context. The social context includes the other participants in the setting, and it includes features of the environment where the interaction takes place. From this perspective, then, gender emerges and is sustained within social relations; hence, social context the groups and settings where people gather plays a much greater role in these perspectives than in those with a more individualist orientation.
Understanding the power of gender in social life without reducing it to individual traits is also the aim of a gendered institutions approach. This perspective directs us to the organization, structure, and practices of social institutions. From a gendered institutions perspective, these entrenched, powerful, and relatively taken-for-granted aspects of the social world play key roles in producing and reproducing gender inequality.
While research informed by an individualist approach to gender (i.e., one that locates gender in the traits and behaviors of individuals) continues to expand, social relational and institutional perspectives have become more popular in recent years. This popularity can be seen in almost all areas of gender scholarship, including research on inequality. Rather than viewing inequalities as arising from differences in the characteristics of women and men, researchers have grown more attentive to the ways in which inequalities are produced by features of the social context. Those influenced by social relational accounts of gender have tended to focus on how the structure of social interaction perpetuates gender distinctions and gender inequality. Gendered institutions approaches, by contrast, place greater emphasis on the practices, policies, and structures of organizations themselves.
Social Context and Social Relations: The Interactional Basis of Gender Inequality
The key feature of social relational explanations of gender inequality is their recognition of the fundamental interdependence that exists between individuals, groups, and larger social units. These perspectives’ roots can be found in the work of classical theorists like Marx ( 1973) and Simmel ( 1955), and later researchers like Blau (1977). Two kinds of social relational approaches have been especially relevant to understanding gender inequality. These include status characteristics theory and research on group composition and social networks.
While they differ in important respects, each approach views social categorization as essential to social interaction. Social categorization refers to the processes through which individuals classify others and themselves as members of particular groups. Virtually everyone agrees that sex category is an extremely important social category (Aries 1996). For some, it is the most important social category. There are many other social categories, however, including those based on racial or ethnic distinctions, age, ability, and so on. All of these social categories may be relevant for social interaction in particular situations and settings. Social categorization is important because it sets into motion the production of gender differences and inequality. The perspectives examined below differ somewhat in their understanding of how and why that occurs, however.
Status Characteristics Theory: The Importance of Expectations
How does social interaction help produce gender distinctions and inequalities? Status characteristics theory (also referred to as the theory of “expectation states”) offers a straightforward answer to this question: As Ridgeway (1997) explains, “[G]ender becomes an important component of interactional processes because the problems of organizing interaction evoke cultural schemas that reinforce continual sex categorization” (p. 219). Because interaction requires that people orient themselves to one another, it is necessary to have some basis for categorizing others vis-à-vis oneself. In Risman’s (1998) words: “Gender is something we do in order to make social life more manageable” (p. 33).
Sex categorization serves this purpose better than any other categorization system, according to Ridgeway and other status characteristics theorists. Continuing reliance on sex categorization as a way to organize interaction, however, tends to create gender expectations and stereotypes. People learn to expect certain kinds of behaviors and responses from others based on their sex category. These expectations serve as cognitive reminders of how we are supposed to behave in any given situation. Risman (1998) refers to them “as accurate folklore that must be considered in every interaction” (p. 32). People thus respond to others based on what they believe is expected of them and assume that others will do likewise.
To explain why and how categorizing others by sex produces gender expectations and inequality, these theorists introduce the idea of a status characteristic. A status characteristic is “an attribute on which individuals vary that is associated in a society with widely held beliefs according greater esteem and worthiness to some states of the attribute (e.g., being male) than others (being female)” (Ridgeway 1993:179). Gender is clearly a status characteristic, and this is true in virtually all societies. Once a characteristic like sex category has status value, it begins to shape expectations and form the basis for stereotypes.
Gender is not the only basis on which people differentially assign power and status, however. For example, age is also a status characteristic; adults are generally ascribed more status and power than children. Similarly, depending upon the particular society, racial or ethnic distinctions may also operate in this way. Furthermore, expectation states theory recognizes that multiple status characteristics may be activated in any given situation.
Status characteristics theory was developed to explain goal-oriented interaction, such as occurs in workplaces, classrooms, or in any group oriented toward a collective end. In these kinds of settings, the important expectations are those relating to performance. That is, group members assess how competent each is and how much value to attach to each other’s contributions. People form their expectations about others’ competence by weighing each status characteristic in terms of its relevance to the task at hand. This weighting process is not assumed to be conscious or precise; rather, expectation states theorists believe that people seek cues as to how others will perform in a particular situation and use status characteristics to assess this. These performance expectations tend to disadvantage those with lower status value (in the case of gender, women). Women are expected to be less competent than men, and their contributions are expected to be less valuable.
Status characteristics theory recognizes that the effects of gender on social interaction may vary from situation to situation. It assumes that status characteristics such as gender are more likely to be “activated” (i.e., central to people’s awareness) in some situations than others. Ridgeway expects gender to be most influential when two conditions hold: the interactants are members of different sex categories, and gender is relevant to the task or purpose of the interaction.
For status characteristics theorists, a group’s sex composition helps to determine how gender will shape the group’s interactions. The second approach focuses explicitly on the role of sex composition. From this perspective, the meaning and impact of one’s own sex category depends on the sex composition of the group. A person’s own sex category is less relevant to any particular interaction than the sex category memberships of those with whom she or he is interacting.
Group Composition, Homophily, and Social Networks
Much research suggests that social ties of all types tend to be organized according to the homophily principle: Social ties tend to be between people who are similar on salient sociodemographic dimensions (Popielarz 1999). Partly, this reflects people’s preferences. Homophily thus arises out of people’s preference for sameness, a preference that is expressed in their interpersonal relations. McPherson, Popielarz, and Drobnic (1992) argue that the homophilous social ties experienced in everyday life are further reinforced[.arrowhorizex]and developed[.arrowhorizex]in people’s group memberships. As they explain, “We argue that most homophily occurs because ties are shaped by the opportunities presented to people in groups. We do not encounter people who are seriously different enough from us frequently enough for them to become social network contacts” (p. 168). Thus, individuals’ access to others is mediated by the groups to which they belong; groups enable certain kinds of ties (namely, ties to similar others) while reducing the likelihood of ties to those who are dissimilar.
These ideas about the importance of homophily in social life have implications for understanding gender inequality. The social-psychological underpinnings of homophily lie in social categorization or social identity theory, perspectives that emphasize how people use social categories to classify self and others, and to guide interaction (Hogg and Abrams 1988; Tajfel 1982; Turner 1987). These theories assume that people define themselves in terms of one or more social identities, and as members of one group as compared to other groups. Categorization, in turn, engenders certain feelings and behaviors. “Otherness” (or difference) is generally associated with more negative outcomes for individuals and groups. Similarity, on the other hand, attracts; that is, people are drawn to those whose attitudes, values, and beliefs are similar to their own. Similarity may also make communication easier, thus contributing to a greater sense of trust and feeling of kinship. Conversely, difference may be threatening and communication more difficult.
Ascribed characteristics, such as sex, race, and age, often become “proxies” for similarity (or dis-similarity). Sex, race, and age are important ascribed characteristics in social life because they are easily observed and difficult to hide. The power of these characteristics also derives from the fact that sex, race, and age are highly institutionalized statuses and, hence, each is laden with layers of social meaning. This increases their value as proxies for similarity and dissimilarity, since they are believed to be reliably associated with particular characteristics.
The similarity-attraction connection implies that sex-segregated groups would be more satisfying to their members than more integrated groups. People should prefer to interact with others like themselves and feel uncomfortable, threatened, and less committed when they are in more heterogeneous groups. These issues have received significant attention from researchers and have been especially important in understanding women’s and men’s work experiences. For example, studies have focused on people’s experiences in groups of varying sex composition. They are interested in whether people have different experiences in mixed-sex groups than in groups that contain all men or all women. In general, the similarity-attraction hypothesis assumes that both women and men would prefer settings where they were in the majority to those where they were less well represented.
These dynamics were captured in a provocative study by Tsui, Egan, and O’Reilly (1992). These researchers examined the consequences of “being different” for workers’ attachment to their firms. They hypothesized that people who were more different from other members of their work groups would be less attached (e.g., less psychologically committed, more likely to be absent from work, and more likely to quit) than those who were more similar. Several forms of difference were examined, including sex, age, race, education, and tenure with the employer.
Consistent with the arguments presented above, Tsui, Egan, and O’Reilly found that being different from one’s coworkers on ascribed characteristics (i.e., age, race, and sex) had negative consequences on attachment, while being different with respect to education or tenure with the employer did not have these consequences. Moreover, these authors found that whites and men—that is, those who were members of the historically dominant categories—reacted more negatively to being different than nonwhites and women. This research thus suggests that being different is difficult for people, especially when it involves difference of an ascribed characteristic, like sex.
This study is part of a large body of literature on the consequences of similarity and difference (Allmendinger and Hackman 1995; Chemers, Oskamp, and Costanzo 1995; Williams and O’Reilly 1998). Though much of this research has been conducted in the United States, the effects of demographic diversity have also been studied in other societies, such as China (Tsui and Farh 1997; Xin and Pearce 1996). While the findings are not always consistent, the results suggest that a group’s demography affects both individual members and group outcomes, and that “at the micro level, increased diversity typically has negative effects on the ability of the group to meet its members’ needs and to function effectively over time” (Williams and O’Reilly 1998:116). More recent research has qualified this latter finding, as studies have begun to identify the conditions under which the negative effects of difference at the individual level, or diversity at the group level, may be eliminated or reduced (Chatman et al. 1998; Flynn, Chatman, and Spataro 2001; Jehn, Northcraft, and Neale 1999). Overall, however, this line of research suggests that the social relations of difference and similarity are important in themselves, regardless of the effects of specific kinds of difference (i.e., gender, race, age).
While Tsui and colleagues (1992) focused on the reactions of those who are different from others in the group, others have examined the majority’s reactions or looked at the interactions between the majority and the minority. Rosabeth Moss Kanter explored these issues in her 1977 classic, Men and Women of the Corporation. Kanter argued that the relative proportions of different “social types” in a group shape members’ social relations. “As proportions shift,” she suggests, “so do social experiences” (p. 207). Proportions have this effect because they influence how people perceive one another.
Kanter (1977) was particularly interested in what she called “skewed groups” (p. 208). In these groups, one social type is numerically dominant and the other is a very small numerical minority (e.g., 15 percent or less). Kanter’s focus on this type of group stemmed from the fact that this is likely to be the situation experienced by “newcomers” to a social setting. Women who enter jobs or workplaces historically dominated by men, for example, are apt to enter as a minority of this type, as are people of color who enter jobs historically dominated by whites. Because it is unlikely that an employer would hire large numbers of women or people of color at one time, sex (and race) integration happens slowly, one or two people at a time. Members of the numerical minority in skewed groups are called tokens. For Kanter, this term is not pejorative, nor does it refer to people who are assumed to have been hired because of their sex or race. Instead, the term token is a neutral label, referring to those whose “social type” constitutes 15 percent or less of a group.
Being a token can be a highly stressful experience. Even if successful in terms of their overall job performance, the conditions under which tokens work are different from those of the dominant group and may be psychologically burdensome. Of course, some tokens will not experience these stresses, and some may even derive self-esteem from successfully overcoming the challenges associated with token status. Nevertheless, Kanter’s point is that how people experience work is shaped in part by how many of their social type are present.
While some researchers focus on tokens, others are interested in how people’s experiences differ across the full range of group types. Allmendinger and Hackman’s (1995) study of symphony orchestras provides an example of this line of research. These researchers were interested in how the sex composition of a symphony orchestra affected its members’ attitudes. This study examines 78 orchestras in four geographical locations, including the United States, United Kingdom, the former East Germany, and the former West Germany. Historically, women have been only a small percentage of players in professional orchestras, and this is true worldwide (Allmendinger and Hackman 1995). In this study, the percentage of women ranged from 2 percent to 59 percent.
Allmendinger and Hackman’s (1995) findings are generally consistent with the similarity-attraction paradigm, though they show that it is more complicated than one might assume. For example, they found that while women were less satisfied when they were in orchestras dominated by men (i.e., 90 percent or more male) than those that were more balanced (i.e., between 40 percent and 60 percent women), they were especially dissatisfied in orchestras that contained between 10 percent and 40 percent women. Male orchestra members also were less satisfied when women were greater than 10 percent but less than 40 percent of members. These findings held true in all four countries, underscoring the power of group composition. Allmendinger and Hackman suggest that once women become a significant minority (i.e., greater than 10 percent), they gain power and cannot be as easily overlooked by their male counterparts. In their words: “Together, these processes result in tightened identity group boundaries for both genders, increased cross-group stereotyping and conflict, less social support across gender boundaries, and heightened personal tension for everyone” (p. 453).
These studies illustrate some of the dynamics that may help perpetuate gender inequality. When men or women enter an occupation, job, or work setting that has been previously dominated by the other sex, discomfort—even hostility—may ensue. Those already employed may resent the newcomer and be unsure about how to relate to him or her. Group norms may have to be renegotiated and miscommunication may occur. The newcomer is likely to feel equally uncomfortable, cautious, and unsure about how or where she or he fits in. The discomfort on both sides may produce conflict. The newcomer may not have much incentive or desire to stay.
The dynamics surrounding women who enter jobs traditionally held by men may be very different from those occurring when men enter predominantly female jobs. Men in predominantly male jobs may perceive women as a threat to their power and status and thus may be motivated to drive them out. The forms this resistance may take range from attempts to make women uncomfortable or to refuse assistance and support, to more serious expressions of hostility and harassment, including sexual harassment.
Williams’s (1989, 1995) research on men employed in predominantly female occupations, such as nursing and elementary school teaching, tells a different story. She shows that while relatively few men seek out predominantly female occupations, those who do are likely to be successful and more highly economically rewarded than their female coworkers. Williams attributes this to several factors: Because femaleness is less highly valued than maleness, women entering predominantly male occupations must struggle to fit in and demonstrate their competence. Men entering predominantly female occupations, on the other hand, carry no such burden. Maleness is positively regarded, in general, and thus men in predominantly female occupations may strive to demonstrate these qualities and preserve their distinctiveness from women.
Men are not necessarily strategic and conscious of these efforts to “do masculinity.” Indeed, Williams believes that men’s motives to preserve gender distinctions stem in part from deep-seated psychological processes. Nevertheless, men are likely to benefit from their token status in ways that women do not. While female tokens must prove themselves capable of doing “men’s work,” male tokens often find themselves on glass escalators; these are invisible and sometimes even unwanted pressures to move up in the workplace (Williams 1995).
The relations between group composition and gender inequality are also the focus of social network research. Network analyses are well represented in the literatures on organizations and careers (Granovetter 1974; Podolny and Baron 1997) and have been increasingly used to examine gender (and racial) differences in career dynamics and outcomes (Ibarra 1992, 1993; Ibarra and Smith-Lovin 1997; Smith-Lovin and McPherson 1993). In general, this research suggests that women and men tend to have different kinds of work-related networks and that these differences tend to be disadvantageous to women’s careers.
The network literature on gender and race has generated many useful substantive findings (see Ibarra and Smith-Lovin 1997 and Smith-Lovin and McPherson 1993 for reviews). More important, however, these analyses have helped illuminate forms of inequality that are often overlooked by other frameworks. By treating networks as social resources (or social capital) (Coleman 1988; Portes 1998), network research calls attention to the instrumental importance of social ties, including the ways that these ties provide resources that can be invested or mobilized to secure other valued ends. For example, researchers have studied the ways in which social ties, as represented by personal networks, can be useful in job searches and can positively influence other occupational outcomes, such as occupational status (Hanson and Pratt 1991). Both the range (diversity) and composition (members’ placement in social hierarchies) of personal networks can affect occupational attainment; more diverse networks provide greater access to information than less diverse networks, and networks that include more powerful or prestigious actors provide greater access to influence (Campbell 1988; Campbell, Marsden, and Hurlbert 1986).
By viewing networks as social resources (or “social capital”), this literature thus calls attention to the role of social relations in constraining and enabling access to power, opportunity, and other kinds of valued social goods. From this perspective, gender and other forms of inequality can be understood as a function of differences between the social relations within which individuals are embedded.
As the twenty-first century begins, the global labor force is more diverse than ever. Roughly 40 percent of the world’s paid workers are women, a figure that has been increasing since the 1970s (Padavic and Reskin 2002). In the United States, as Tsui and colleagues (1992) note, “more and more individuals are likely to work with people who are demographically different from them in terms of age, gender, race, and ethnicity” (p. 549). Given these changes, it has become more important than ever to understand the consequences of homophily, social networks, and “being different” for women and men in the workplace.
Gendered Organizations and Gendered Institutions
Much of social life is organized and routine. People are employed by organizations, such as business firms or the government. They attend school and may be members of churches or voluntary associations, such as neighborhood or political groups. In fact, many of the interactions people have take place within organizations. The social practices that are associated with organizations play an especially important role in the production and reproduction of gender and gender inequality.
Institution is a somewhat more abstract and more all-encompassing concept. In simplest terms, sociologists define an institution as “an organized, established pattern,” or even more simply, “the rules of the game” (Jepperson 1991:143). Institutions, then, are those features of social life that seem so regular, so ongoing, and so permanent that they are often accepted as just “the way things are.” Each major social institution is organized according to what Friedland and Alford (1991) call “a central logic—a set of material practices and symbolic constructions” (p. 248). These logics thus include structures, patterns, and routines, and they include the belief systems that supply these with meaning.
Two additional aspects of institutions are important to keep in mind. First, institutions tend to be self-perpetuating and thus to take on a life of their own. This implies that there need not be[.arrowhorizex] and often is not[.arrowhorizex]any conscious intent on the part of participants to create or reproduce gender. In addition, because they are taken for granted, institutions tend to produce a socially shared “account” of their existence and purpose: “Persons may not well comprehend an institution, but they typically have ready access to some functional or historical account of why the practice exists…. Institutions are taken-for-granted, then, in the sense that they are both treated as relative fixtures in a social environment and explicated (accounted for) as functional elements of that environment” (Jepperson 1991:147). The availability of these accounts helps explain why institutions are so rarely challenged or scrutinized: People believe that their purpose and functioning are self-evident.
As this discussion reveals, institutions incorporate more of the social landscape than organizations. In fact, many institutions contain several different types of organizations. Given this, I refer to this framework as the gendered institutions approach, recognizing that it includes aspects of organizations as well.
Acker (1992) observes that many of the institutions that constitute the “rules of the game” in American society and, indeed, most societies embody aspects of gender. As Acker (1992) defines it, to say that an institution is gendered means that “gender is present in the processes, practices, images and ideologies, and distributions of power in the various sectors of social life. Taken as more or less functioning wholes, the institutional structures of the United States and other societies are organized along the lines of gender…. [These institutions] have been historically developed by men, currently dominated by men, and symbolically interpreted from the standpoint of men in leading positions, both in the present and historically” (p. 467).
Drawing on these ideas, gender scholars suggest that the structures and practices of organizations are gendered at all levels and that these aspects of gender are analytically distinct from the effects of group composition (Britton 2000). In Steinberg’s (1992) words: “Masculine values are at the foundation of informal and formal organizational structures…. Images of masculinity and assumptions about the gendered division of labor organize institutional practices and expectations about work performance” (p. 576). As an example of this line of argument, some research has focused on how cultural beliefs about gender infuse people’s understandings of jobs, occupations, and particular work activities (Pierce 1996). By establishing certain work roles, jobs, and occupations as appropriate for one gender and off-limits to another, these cultural beliefs establish the “way things are” or a set of commonsense understandings of who should engage in what type of work.
Gendered Jobs and Gender Inequality
That jobs dominated by a particular gender come to be seen as most appropriate for that gender may seem unproblematic and inevitable, but this association is produced through a complex process of social construction. As Reskin and Roos (1990) note, virtually any occupation can be understood as being more appropriate for one sex or another “because most jobs contain both stereotypical male and stereotypical female elements” (p. 51). Nursing, for example, increasingly requires workers to be skilled in the use of complex medical technologies. Emphasizing the caring aspects of this occupation, however, allows it to be cast as an occupation particularly appropriate for women—in the United States, at least. Most jobs and occupations contain enough different kinds of characteristics that they can be construed as appropriate for either women or men. Indeed, jobs and occupations that are predominantly female in the West are not necessarily held by women in other parts of the world and vice versa. For example, in Senegal and Tunisia, approximately half of all nurses are men (Anker 1998).
The gendering of work can also be seen in gender-integrated positions or in jobs that contain a minority of the other gender (Hall 1993; Williams 1989, 1995), and it applies to jobs held by men as well as those that are predominantly female (Maier 1999; Pierce 1996). For example, as a growing literature on work and masculinity has shown, many predominantly male jobs implicitly and explicitly require incumbents to display traditionally masculine behaviors, such as aggressiveness (Pierce 1996). Maier (1999:71) argues that managerial practices and organizational cultures—not merely specific jobs—embody a “corporate masculinity” that privileges individualism, competitiveness, and technical rationality.
In some situations, the work tasks may be gendered as feminine, but the worker performing them is male. Hall’s (1993) study of table servers offers a useful example of these arrangements. Styles of table service are laden with gender meanings. A familial style of service, which Hall labels “waitressing,” has been historically associated with women working in coffee shops and family restaurants. By contrast, “waitering” is a more formal style, usually associated with male servers in high-prestige restaurants. Hall suggests that even in sex-integrated restaurants, “work roles, job tasks, and service styles” (p. 343) continue to be gendered, such that waitering whether performed by women or men is more highly valued by employers and customers than is waitressing.
The gender-typing of occupations, jobs, and work tasks is not a random process, however. In particular, low-status jobs containing low amounts of power and control over others are much more likely to be gendered female than high-status jobs requiring the exercise of authority. Anker (1998) shows that three-quarters of employed women worldwide work in seven occupations: nurses, secretaries/typists, housekeepers, bookkeepers and cashiers, building caretakers and cleaners, caregivers, and sewers and tailors. Padavic and Reskin (2002) suggest that “most of these occupations appear to conform to stereotypes of women as caring, patient, nimble-fingered, skilled at household tasks, and docile” (p. 71).
Deference—or the capacity to place oneself in a “one down” position vis-à-vis others—is a characteristic demanded of low-status social groups in many circumstances. This capacity may also be expressed as “niceness” or the ability to get along. It is not surprising that when this capacity is a job requirement, women will be viewed as better qualified than men. Moreover, even when deference may not be a formal job requirement, jobs containing large numbers of women are likely to contain an informal job requirement that encourages this behavior. Conversely, jobs involving the display of authority are more likely to be gendered as male, at least in part because authority in the context of the United States is seen as a masculine characteristic.
The Worth of Jobs
An important link between the gendering of jobs and gender inequality appears when we examine the relative values attached to different kinds of work. The higher societal value placed on males and masculinity over females and femininity is reproduced within the workplace. In this setting, the relative worth of activities can be assessed economically, in the form of wages, and symbolically, in the form of status and prestige. On both counts, men and masculine activities are more highly valued than women and feminine activities.
Economists have devoted considerable attention to understanding what determines the “worth” of jobs and why some jobs pay more than others. For the most part, these scholars have emphasized the role of human capital in wage determination (see, e.g., Becker 1985; Mincer and Polachek 1974). Human capital refers to the portfolio of skills that workers acquire through various kinds of “investments” in education, training, or experience. Applied to the gender wage gap, human capital theory implies that women earn less than men because of differences in the kind and amount of human capital each has accumulated.
Feminist scholars and others argue that this provides an incomplete understanding of the gender wage gap; they have critiqued economic arguments for overlooking the many ways that social factors shape wage-setting and wage inequality (see, e.g., England and Farkas 1986; Steinberg and Haignere 1987). In addition, researchers have suggested that institutional or structural factors, as well as individual-level factors, contribute to the gender wage gap (Roos and Gatta 1999). By focusing on these issues, the study of gender inequality and the gender wage gap has broadened from a narrow focus on individuals’ human capital to a concern with social structural and institutional factors. On a more general level, interest in understanding women’s earnings and work experiences has contributed to a more accurate view of labor markets. Researchers who focus only on men—especially white men—tend to “exaggerate the extent to which labor market processes are meritocratic” (Reskin and Charles 1999:384).
During the last few decades, research on the gender wage gap has proliferated. Men earn more than women in virtually every country in the world (Anker 1998; Blau and Kahn 1992, 2001; Padavic and Reskin 2002). Researchers’ understanding of how the gender-based wage inequality occurs, how it has changed over time, and its variations among women and among men have become increasingly more complex and nuanced (Bernhardt, Morris, and Handcock 1995; Blau and Kahn 2001; England 1992; Goldin 1990). Two major themes in this work illustrate recent developments.
First, research on the gender wage gap in the United States has become increasingly attentive to issues of race and ethnicity (Browne 1999; Kilbourne, England, and Beron 1994; Tomaskovic-Devey 1993). As Reskin and Charles (1999) note, the historical tendency has been a “balkanization of research on ascriptive bases of inequality” (p. 380). Studies of the gender wage gap, in particular, have often ignored racial and ethnic variations among women and men. Not only does this produce potentially inaccurate results, it has hindered efforts to understand the forces generating and maintaining wage inequality.
A second, emerging theme in recent studies of the gender wage gap are attempts to relate gender-based wage inequality to wage inequality more generally, as well as to other forces that are transforming the workplace, such as globalization. Wage inequality in the United States has increased in recent decades—a pattern that reflects industrial and occupational restructuring, changing labor force demographics, globalization, and political trends (Morris and Western 1999). This widening inequality not only reflects earnings differences between women and men, but also differences among women and among men (Mishel, Bernstein, and Schmitt 1999). A narrow focus on the gender wage gap misses these broader patterns of inequality. These trends are not confined to the United States, but are operating on a global scale as economic competition intensifies and employment is restructured (Blau and Kahn 2001; Walby 2000).
The Gender Segregation of Occupations and Jobs
Gender segregation is an entrenched and pervasive feature of the industrial workplace and, like other aspects of gender inequality, is a worldwide phenomenon. Women make up almost half of the paid labor force, but women and men are employed in different occupations, firms, and jobs (Bielby and Baron 1984; Cotter et al. 1995, 1997; Jacobs 1999; Reskin 1993, 1994; Reskin and Hartmann 1986; Reskin, McBrier, and Kmec 1999; Tomaskovic-Devey 1993). Large-scale, quantitative studies of occupational segregation in the United States have documented trends in segregation over time and identified factors associated with changes in segregation levels (e.g., Beller 1982; Beller and Han 1984; Cotter et al. 1995). These studies suggest that the declines in occupational segregation that took place in the 1970s appear to have leveled off in the 1990s (Jacobs 1999).
Other researchers have examined segregation cross-nationally, looking for clues as to the economic, political, and cultural factors that produce gender segregation and inequality (Charles 1992; Jacobs and Lim 1992; Roos 1985). Although occupational sex segregation is a feature of all industrial societies, its form varies widely. These variations have been traced to several factors. For example, Brinton (1993) argues that Japan’s relatively low levels of occupational sex segregation as compared to the United States’s reflects differences in the two countries’ occupational distributions. In particular, Japan employs more people in lesser-segregated areas of agriculture and manufacturing than in more highly sex-segregated white-collar positions. In her study of occupational sex segregation in 12 industrialized countries, Roos (1985) concluded that “the United States pattern is fairly typical—large numbers of women in high-prestige clerical occupations, low-prestige professional and technical positions, and low-prestige service jobs” (p. 48).
More recent studies extend Roos’s findings and suggest that a country’s level of occupational sex segregation depends upon a variety of economic, social, and cultural factors. Women generally have greater access to predominantly male occupations in countries with low birthrates and strong egalitarian belief systems, while sex segregation is increased when countries have large service sectors (Charles 1992). Governmental policies relating to gender also play a role in shaping a country’s level and pattern of occupational sex segregation. Chang (2000) distinguishes between “interventionist” and “non-interventionist” governments; interventionist governments actively attempt to influence women’s labor force participation by passing legislation guaranteeing equal opportunity in the workplace, or by providing direct benefits to families, such as state-subsidized child care or paid family leave. Depending upon the level and type of intervention they engage in, governments help to define their country’s “sex segregation regime” (Chang 2000).
This literature on occupational segregation has been supplemented in recent years by research at other levels of analysis. Bielby and Baron (1984), for example, moved segregation research from the occupation level to the job level and firm level of analysis; they showed that job segregation within firms was considerably higher than segregation at the occupational level. Tomaskovic-Devey (1993) also found extremely high levels of job segregation by gender in his study of North Carolina firms. Achatz, Allmendinger, and Hinz (2000) also reported high levels of job-level sex segregation in their study of West German organizations. Even if occupations appear gender-integrated, these studies demonstrate that women and men rarely work together, holding the same job in the same firm. A second extension of segregation research moves it in a more macro direction by focusing on labor markets—rather than jobs or occupations. Cotter et al. (1997) show that occupational integration at the local labor market level improves earnings for all women in that labor market, regardless of the gender composition of a woman’s own occupation.
A Gendered Institution’s View of Gender Inequality across the Globe
A gendered institutions perspective has been particularly useful for understanding cross-national variations in gender inequality. As Buchmann and Charles (1995) argue, “If we are to understand cross-national variability in women’s economic roles, it may therefore be useful to begin by identifying a set of macrolevel variables that systematically influence the perceptions and actions of individual men and women…. Organizational and institutional factors are important in this regard, because they set the context within which men and women make educational, labor market, and/or fertility decisions” (p. 66).
Researchers have conceived of these institutional contexts in different ways. For example, Chang (2000) contends that a country’s response to gender inequality generally adheres to one of four patterns: substantive egalitarian, formal-egalitarian, traditional family-centered, and economy-centered. Each pattern embodies a distinct “gender logic” that is reflected in that country’s norms and state policies. Hence, rather than cross-national convergence in patterns of gender stratification, Chang suggests that gender stratification assumes different forms under different institutional regimes. Esping-Andersen’s (1990, 1999) typology of Western welfare states provides another framework for an institutional analysis of gender inequality. In addition to factors such as the relative importance of the market versus the state in shaping employment outcomes, Esping-Andersen’s framework takes into account the state’s and market’s roles in easing family caregiving responsibilities. While gender inequality figures less explicitly into his typology than Chang’s, Esping-Andersen’s approach has been used extensively to examine aspects of gender inequality and women’s employment across societies (van der Lippe and van Dijk 2002).
Recall that an arrangement that is highly institutionalized is one that is so taken for granted, it seems to reproduce itself. In other words, the arrangement persists without conscious intervention and effort. It thus is much more difficult to alter something that is highly institutionalized than it is to perpetuate it. As a result, highly institutionalized arrangements do not require coercion to sustain them, making participation appear voluntary and easily justifiable. The gendered aspects of state policy and the economy described above clearly fit this description; they are often unintended, taken for granted, and may operate so subtly that they are rarely scrutinized. In this respect, gender distinctions and inequalities are highly institutionalized features of the modern state and economy.
Where do we go from Here? toward a Multilayered Conception of Gender and Gender Inequality
One of the most significant and ongoing contributions of feminist scholarship has been its intensive examination of the concept of gender. This examination has produced a conceptual shift over time in how gender is understood. Early feminist scholars treated gender as an attribute of individuals; gender was a role people acquired through socialization; it was “carried” into workplaces (or to schools or families), where it was then expressed in people’s behaviors and beliefs. Because gender was treated as a property of people, the possibility that institutions or features of the social context could play a role in the reproduction of gender distinctions and inequalities was overlooked. Although some contemporary feminist scholars continue to view gender solely as an aspect of the person, most believe that gender is a multifaceted system of practices and relations that operates at all levels of the social world (Hawkesworth 1997; Ridgeway and Smith-Lovin 1999).
This more multilayered view of gender can be seen in many areas of feminist research, but has been most powerfully put to use in research on gender inequality in the workplace. The broadening of feminist conceptions of gender has increased the reach of feminist scholarship and expanded its influence on the social sciences. Treating gender strictly as an individual-level characteristic limits our ability to examine how gender distinctions and inequality are produced at other levels of the social order. A gendered institutions approach makes clear, however, that social structure, institutions, and interaction cannot be understood without taking gender into account.
As a multilevel system affecting individuals’ identities and characteristics, patterns of social interaction, and social institutions, the gender system and the inequalities it produces shape social life in crucial ways. What are the key issues this raises for future research on gender inequality? First, as gender research becomes more social relational and institutional, social context becomes a more critical variable than in the past. This implies that we need to be less concerned with gender distinctions and inequalities per se and more concerned with understanding systematically how social contexts vary in the way these distinctions and inequalities are constituted and expressed. As this chapter shows, the relevant social contexts for understanding gender inequality range from the level of social groups engaged in interaction to organizations and institutions and societal regimes. Instead of asking whether gender differences and inequalities exist, we need to look at how their level and expression vary across social contexts (see Nelson and Bridges 1999).
One important line of research moving in this direction relates to studies exploring the effects of diversity on workers and work organizations. This research draws from Kanter’s work in an important respect, as it is concerned with the ways that the demographic composition of work groups shapes interaction and behavior (Chemers et al. 1995; Ruderman, Hughes-James, and Jackson 1996). Diversity researchers, however, are not exclusively or even primarily concerned with the sex composition of groups, but are also interested in how other kinds of differences shape people’s interactions on the job and their responses to work.
A significant finding emerging from these studies is that differences between people—such as those deriving from gender or race—are not always salient in the workplace. A salient characteristic is one that influences a person’s perceptions and behavior in a situation, and it is one that shapes how others respond to that person (Turner 1987). While sex category is probably more salient in more situations than many other attributes of a person, diversity research suggests that it is not always an important factor in workplace social relations. For example, Chatman et al. (1998) found that a more collectivistic organizational culture that emphasizes teamwork and encourages people’s sense of a shared fate can create cohesiveness even among diverse groups. Diversity researchers challenge employers in the twenty-first century to create workplaces where people who are different can work together.
Second, future research should attend to multiple levels of analysis and institutional contexts. For example, Nelson and Bridges (1999) combine the insights of a gendered organizations view with more conventional organizational theory. They use the latter to specify the conditions under which larger societal patterns of gender inequality—including those engendered by market forces—might be reduced or enhanced by particular organizational structures, policies, and practices. In this way, Nelson and Bridges argue that organizations vary in the ways that gender is institutionalized; organizations differ both in the degree to which they reinforce market-based gender inequality and the means through which they accomplish this.
Finally, researchers have become increasingly concerned with the relations between gender and other bases of distinction and stratification, such as age, race, or ethnicity, sexual orientation, or social class. A growing literature challenges the notion that women (or men) represent a homogeneous category, whose members share common interests and experiences. Theory and research seeking to describe the intersections between race, class, and gender, in particular, have proliferated, and there is increasing attention to these questions among those interested in the workplace. In the years ahead, it will be important to more systematically address how other forms of social differentiation besides gender are constructed and maintained in the workplace. Doing this requires that we overcome the “balkanization” between the literatures on racial, gender, and other forms of inequality (Reskin and Charles 1999:380) and begin to focus on “organizational systems of inequality” (Nelson and Bridges 1999:5).
As a system of social practices that produces distinctions between women and men and organizes inequality on the basis of those distinctions, gender is a powerful principle of social life. Gender thus is a complex, multilevel system that is visible throughout the social world. This complexity poses challenges for those interested in eliminating gender inequality. At the same time, gender scholars are continuing to develop new tools and insights to unravel this complexity. Their work will be as important during the twenty-first century as ever.