Multiculturalism & pluralism
|Classification Codes:||9130: Experimental/theoretical|
2500: Organizational behavior
|Copyright Plenum Publishing Corporation Nov 1999|
Diversity has become a central element of the modern organizational landscape-so much so that 75% of the 50 largest companies in the U.S. now have diversity directors or managers (Economist, 1995). A whole new industry has even materialized-diversity training-to help organizations address the challenges that diversity creates (Ferguson, 1994). This interest in diversity should not be surprising. As firms move toward flatter organizational structures featuring groups and teams, and the global economy moves labor across national boundaries (e.g., Byrne, 1993), effective interaction among diverse workers becomes more critical to smooth organizational functioning (Jackson & Ruderman, 1995).
Through all of this, precise definitions of diversity remain elusive (Carrell & Mann, 1993). Diversity in organizations typically is discussed in terms of the benefits of variety in the characteristics of a workforce (e.g., Cox & Blake, 1991; Northcraft & Neale, 1993). However, McGrath, Berdahl, and Arrow (1995) note that the dialogue concerning diversity often confuses different types of workforce variety. One type of variety in the workforce is social category membership, i.e., membership in a group based on a shared characteristic-for example, a demographic characteristic such as race, gender, age, ethnicity, sexual orientation, or even family status. Legal and moral stances on diversity have emphasized creating representative social category diversity in an organization's workforce, i.e., workforce demographic heterogeneity which mirrors the demographic mix of the local labor market. While any personal characteristic (e.g., eye color, astrological sign) might provide the basis for social categorization, only some characteristic workforce variety may be functional, i.e., directly relevant to organizational performance (Northcraft, Polzer, Neale, & Kramer, 1995). McGrath et al. (1995) have noted several different types of diversity that fit under this "functional characteristics" umbrella, including differences within a workforce in knowledge, skills, and abilities (KSAs), values, beliefs, and attitudes (VBAs), and personality, cognitive style, and behavioral style (PCBs).
Researchers have decried the absence of good conceptual frameworks for understanding how to manage workforce diversity (Triandis et al., 1994)-an absence which may account for the indifferent success of diversity initiatives in organizations (Economist, 1995). This paper presents a conceptual framework for understanding diversity in organizations. The framework is grounded in two well-documented theories of social behavior: social identity theory (e.g., Tajfel & Turner, 1986) and the theory of social dilemmas (e.g., Messick & Brewer, 1983). We argue that the costs associated with workforce diversity in organizaions arise because managers and their employees perceive organizational conflicts (and organize their interactions) along social identity lines. The result is three social dilemmas of diversity: organizational, individual, and managerial. Social identity effects create the dilemmas of individual and managerial participation; the costs of managing individual and managerial dilemmas create the dilemma of organizational participation. Viewing organizational diversity in terms of social dilemmas and social identity theory can help organizations overcome the dilemmas and make the benefits of diversity more accessible to organizations. We begin by reviewing our two central theoretical perspectives: the theories of social dilemmas and social identity.
SOCIAL DILEMMAS AND SOCIAL IDENTITY
Organizational membership often represents a social dilemma because individuals can realize substantial benefits from organizational membership, even if they do not pay the costs of those benefits. However, if everyone tries to enjoy the benefits of organizational membership without paying the costs, the benefits disappear. The classic social dilemma is Hardin's (1968) "Tragedy of the Commons." In this dilemma, cattle herders must restrict their use of a commonly shared grazing land in order for the land to remain productive. However, it is in the short-term self-interest of any individual herder that all other herders limit the number of cattle they graze while that individual herder does not. When all of the herders rely on others to restrict usage but don't restrict their own usage, the grass becomes overgrazed, the cows die, and the benefit disappears. The dilemma occurs because no individual herder has any short-term incentive to forego the immediate benefits that the commons provides.
Messick and Brewer (1983) have identified several types and dimensions of social dilemmas which are related to diversity in organizations. "Temporal" dilemmas (Messick & McClelland, 1983) occur when the costs of some action (or inaction) are immediate, while the benefits are delayed. Performance appraisal represents a temporal dilemma for managers. Time spent on performance appraisal is time not spent on more immediately productive activities. But if time is never spent on performance appraisal, in the long run workgroup performance will suffer. "Collective" dilemmas occur when a benefit to one individual is costly to the collective. Supply hoarding is a collective dilemma; an individual's hoard of supplies provides that individual a benefit, but at the social cost of the collective shouldering higher inventory costs, or suffering from collective unavailability of supplies, particularly if everyone attempts to board.
Messick and Brewer (1983) also distinguish between traps and fences. Traps occur when there is a positive incentive to ignore the collective good. For example, individuals are "trapped" into hoarding office supplies because it is beneficial to have a personal cache of supplies, despite the increased inventory costs to the organization. Fences occur when there is a disincentive for individuals to take actions that promote the collective good. Undergraduate advising is an example of a collective fence. While it is in the best interests of everyone in an academic department that someone advise undergraduates, it is in the self-interest of any particular faculty member that some other faculty member undertake this obligation. If all professors shirk their responsibilities for undergraduate advising, department course enrollments (and perhaps even tuition revenue allocated to that department) may suffer. The hallmark of a collective fence is that it makes the most sense (in terms of economy of scale) for some individual to undertake the burden, but it never makes sense to be that individual (Murnighan, Kim, & Metzger, 1993).
Diversity initiatives in organizations present social dilemmas at three levels-organizational, individual, and managerial. First is the dilemma of organizational participation: Though it is in the interest of society-for both moral and practical reasons-that all organizations support diversity, it may be in the interest of any particular organization that other organizations assume this responsibility. Second is the dilemma of individual participation: Though it is in the interests of all members of an organization that everyone further the organization's pursuit of diversity, it may be in the interests of particular individuals to further only their own goals. Finally, there is the dilemma of managerial participation: Though it is in the interest of the organization that all managers have diverse work groups, it may be in the interest of any particular manager that other managers assume this responsibility. Significantly, the dilemma of organizational participation is a dilemma because of the other two dilemmas. An organization's pursuit of diversity poses a dilemma because it invokes the dilemmas of individual and managerial participation. Resolving the individual and managerial dilemmas removes the costs of organizational participation, and thereby removes the dilemma for organizations. Key to resolving the individual and managerial dilemmas is understanding the role of social identity processes (e.g., Tajfel & Turner, 1986; Turner, 1987) in diverse organizations,
According to social identity theory, individuals validate their social identity (and thereby accrue self-esteem) by showing favoritism for their own social category or "ingroup," at the expense of "outgroups" to which they do not belong (e.g., Billig & Tajfel, 1973; Tajfel, Billig, Bundy, & Flament, 1971).
Both demographic characteristics (such as age, gender, or race), and functional distinctions within the organization (e.g., professional or departmental affiliations), provide salient means for individuals to categorize themselves. An individual's social category memberships often are taken (perhaps incorrectly) as surrogates for functional characteristic variety (e.g., Eagly, 1987; Tsui, Egan, & Xin, 1994). Individuals sharing a social category (their "ingroup") often are assumed by others to share similar values and interests (McPherson & Smith-Lovin, 1987; Turner, 1987; Wilder, 1986), and ingroup members are often assumed by each other to be easier to communicate with, more predictable, more trustworthy, and more likely to reciprocate favors than outgroup members (Brewer, 1981; Brewer, 1979; Kanter, 1977; Komorita, Parks, & Hulbert, 1992; Lincoln & Miller, 1979).
Social dilemmas and social identity theory are inextricably connected in the dialogue of diversity because social identity provides dimensions along which the lines of social dilemma conflicts can be drawn. A shared social identity increases perceived differences between social categories (Messick & Mackie, 1989; Brewer, 1979), leading individuals to assume that there will be irreconcilable conflicts of interest between social categories where none necessarily exist (Northcraft et al., 1995). Thus, the success of diversity initiatives in organizations (and the realization of diversity benefits) rests on mitigating the assumption that the needs, desires, values, and perspectives of members of different social groups are mutually exclusive, rather than potentially congruent or complementary (Northcraft et al., 1995).
THE DILEMMA OF ORGANIZATIONAL PARTICIPATION
For diversity to constitute a social dilemma, diversity must have longterm benefits (for the organization or society), but there must be short-term costs to diversity which deter the organization from pursuing it (Dawes, 1980). While the benefits of functional diversity to organizations should not be much of a mystery, the benefits of social category diversity are perhaps less obvious.
Benefits of Functional Diversity
Solving complex problems is an activity where groups should benefit from functional diversity. Compared to individual problem solving, groups have more people to look at a problem, which increases the number of opportunities to find errors or discover key information (Marquart, 1955; Shaw, 1932; Triandis, Kurowski, & Gelfand, 1994). Functionally diverse groups do not simply provide more people to examine a problem; they are composed of people who offer different talents and perspectives. This enhances the probability that an adequate solution to the problem will be proposed. Functional diversity is also important to innovation, renewal, and creativity in organizations (Amabile, 1983). Creative solutions to problems are more likely to emerge when a diversity of theoretical perspectives can be brought to a problem (Maier, 1930; Souder, 1987; Moenaert & Souder, 1990). This may explain why scientists benefit most from contacts with dissimilar colleagues (Pelz, 1957), and why top management heterogeneity predicts innovativeness in banks (Jackson, 1992). Of course, whether increasing the number of perspectives actually improves problem solving depends on a number of factors such as whether and how members of functionally diverse groups interact (e.g., Stasser, Taylor, & Hanna, 1989; Miller & Hamblin, 1963).
Functional diversity also implies connections to a more varied external network (Lipnack & Stamps, 1993). This makes it more likely that group members can use outside contacts to gain access to information (Granovetter, 1972). For example, in a study of new product teams in high-technology companies, functionally diverse teams were more likely to communicate with others outside of the team, and such external communication led to greater innovativeness (Ancona & Caldwell, 1992). Access to more outside contacts also makes it easier for the group to monitor turbulence and uncertainty-and their possible causes-in the environment (Donnellon, 1993; Trist, 1977).
Functional diversity also provides the possibility of specialized division of labor. Portfolio theory (e.g., Sharpe, 1970) emphasizes the importance of a diversity of skills to the success of an organization. There are skills that every organization or group needs to survive; a more diverse workforce means it is that much more likely that needed skills are represented somewhere in the workforce. In the context of particular projects, greater diversity in backgrounds and skills makes it more likely that any required aspect of a problem or project can be attended to by an expert or specialist, and that any required aspect of a problem or project can be attended to by someone interested in that aspect (Northcraft & Neale, 1993). Functional diversity also makes it more likely that an organization will have represented in its workforce any skill, information, or contacts necessary to respond to any competitive challenges in the environment.
Benefits of Social Category Diversity
Workforce social category diversity is not so much a benefit to have as it is a problem if avoided. A strong argument for the benefits of social category diversity comes from statistical sampling theory and the "law of large numbers." In the U.S., workforce participation by Hispanics grew from 4.5% in 1976 to 7.4% in 1988, and is expected to exceed 10% by the year 2000; U.S. workforce participation by individuals of Asian descent also is on the rise (Coates, Jarratt, & Mahaffie, 1991). Further, the globalization of the world economy is repeating this picture in other countries as well (Byrne, 1993).
These changing workforce demographics mean that an organization that resists social category diversity in its workforce will be an organization that limits its opportunities to hire the most qualified applicants. Such organizations will be selecting deeper into a smaller labor pool, thus hiring less qualified individuals. Even assuming no correlation between an individual's social category and functional characteristics, the smaller the applicant pool that an organization is willing to entertain, the less likely that many applicants with a particular desirable functional characteristic (or any particularly desirable combination of functional characteristics) will turn up in that limited applicant pool. This does not mean that organizations will automatically become more talented by hiring individuals from diverse social categories; it does mean, however, that organizations that are not open to social category diversity are not availing themselves of the best available personnel.
The argument that constraining social category diversity de facto constrains functional diversity becomes even more compelling if one believes that social category membership in some cases correlates with differences in cultural background (e.g., Knight, Bernal, Garza, Cota, & Ocampo, 1993) and socialization experiences (e.g., Lever, 1978; Fox, 1990), and thereby probabilistically implies real differences in functional characteristics (e.g., Doktor, 1990; Eagly, 1990; Hayes & Allison, 1988). Currently there is little direct empirical evidence to support this assumption, though it is congruent with previous research (e.g., Eagly, 1987).
Paradoxically, even social category diversity not based on functional characteristics may prove functional. If (as suggested by social identity theory) birds of a feather flock together (e.g., Schneider, 1987), then perceived similarity to organization members may drive organization choice by customers, clients, prospective investors, and prospective employees. Social category diversity thus provides organizations a functional benefit: a better probability of being selected as employer or product/service provider, via access to external networks (Ibarra, 1993; Lipnack & Stamps, 1993).
Barriers to Organizational Diversity
While functional and social category diversity in an organization's workforce offers potential benefits, the fact is that these advantages often are not realized (Hoerr, 1989). Ironically, once workforce diversity is attained, that diversity both provides the possibility of constructive co-action, and also makes constructive co-action unlikely (Epton, Payne, & Pearson, 1985; Triandis et al., 1994). Homogeneous teams typically take less time to organize and complete tasks than do diverse ones (Moos & Speisman, 1962). Furthermore, the amount of turnover within an organization is often a function of the group's diversity, and the probability that an individual will leave the organization is a function of the difference between that individual and the rest of the group (Jackson, Brett, Sessa, Cooper, Julin, & Peyronnin, 1991). Participation in diversity initiatives represents a dilemma for organizations if such short-term costs of a diverse workforce keep organizations from enthusiastically pursuing diversity's potential long-term benefits.
Organizations may resist social category diversity because it represents a collective fence: there are lots of costs and few benefits, and the benefits mostly accrue to society rather than particular organizations. Organizations may resist functional diversity because it represents a temporal trap: the costs are certain and immediate, while the benefits are uncertain and in the future. In both cases, the costs that create the dilemmas reflect the other two dilemmas of diversity: the social dilemmas of individual and managerial participation. It is these two subordinate dilemmas that create the barriers that make diversity a dilemma for the organization.
THE DILEMMA OF INDIVIDUAL PARTICIPATION
Even if an organization formally embraces diversity, individuals within the organization may resist it. Social identity research has shown that individuals are likely to eschew contact with others they categorize as different (Brewer, 1986), whether those differences are functional (e.g., knowledge, skills, abilities), or not. In diverse organizations, informal coalitions may take shape around shared category membership, even without explicit collusion to exclude others. Information then will bypass the formal network of reporting relationships in favor of informal networks based on functional (e.g., professional affiliation) or social category (e.g., gender, race) distinctions. Individuals left out of such networks have difficulty succeeding in organizations (e.g., Brass, 1985.- Ibarra, 1993; Kanter, 1977; Lincoln & Miller, 1979). For example, women and minorities are often excluded from the informal networks of the majority (e.g., Bartol, 1978; Ibarra, 1993; Kanter, 1977; Lincoln & Miller, 1979; Morrison & Von Glinow, 1990; Northcraft & Gutek, 1993). As one female corporate vice-president observed, "The ways I am excluded are very subtle ... before meetings or during the break, I am often left by myself while the men chat with each other" (Watts, 1989, p. 32).
In response, it may seem sensible for members of diverse subgroups to develop their own, independent social networks (Ibarra, 1993; Watts, 1989). Yet, in the name of promoting their own subgroups, individuals will be turning the simplest transactions between individuals into opportunities for subgroup competition (Kramer, 1993). Worse, these sorts of separatism and self-segregation negate the benefits of functional diversity, as individuals no longer interact with diverse others.
Often, a key assumption driving self-segregation is that diverse individuals suspect that their preferences and values are mutually exclusive. When diverse individuals perceive such conflicts of interest, trust, communication, and mutual helping behaviors decline (Souder, 1987; Deutsch, 1949). Separatism and disruptive conflict are likely to follow (e.g., Ginn & Rubenstein, 1986).
This assumption of preference or interest incompatibility can be wrong in two important ways. First, diverse individuals may share similar underlying preferences, but distrust and miscommunication may lead them to misconstrue their commonalties of interest as conflicting (e.g., Thompson & Hastie, 1990). Second, even when diverse individuals do have real differences in interests, preferences, and values, there may be integrative resolutions (e.g, Neale & Bazerman, 1992) that serve the underlying interests of both sides-if they work together. For example, in a new product launch, the integration of apparently mutually exclusive preferences might lead to the development of a better product-one that appeals to a larger market segment by serving more diverse interests-than if only one perspective or the other guided the product's development. This illustrates the positive potential of conflict arising from diversity (e.g., Lewicki & Litterer, 1985).
It should be apparent that ultimately it is functional diversity which permits innovative adaptability in an organization. Social category diversity per se that does not offer some measure of functional diversity may provide an organization few real benefits. Nevertheless, either form of diversity can create the social identity crisis that puts a collective fence around diversity for individuals. Ironically, an organization that fulfills its labor market representativeness obligations without attaining some functional diversity may experience the social identity costs of a diverse workforce without the benefits.
THE DILEMMA OF MANAGERIAL PARTICIPATION
The dilemma of individual participation is precisely what creates the dilemma of managerial participation. Managers might appreciate the longterm benefits of having a more functionally diverse stable of subordinates (and in not precluding a demographically diverse one). However, diversity represents both a temporal trap and a collective fence for managers because the costs of diversity to managers are (apparently) certain and immediate, while the benefits are likely to take time to develop, if they are realized at all. Two categories of these costs to managers are: selection/retention processes, and co-action/cooperation processes.
Selection/Retention Processes in Organizations
Social identity theory suggests that managers Oust like their subordinates) are likely to use salient social categories (such as race, gender, and functional specialization) as indicators of value and preference similarity, and thereby the potential for disruptive conflict. Therefore, managers are likely to prefer (in hiring and mentoring) individuals with whom they share (and with whom their subordinates share) category membership on such dimensions (Kanter, 1977). Kanter (1977, p. 63) refers to the tendency of managers to hire their own kind as "homosocial reproduction." In effect, the dilemma for managers appears to be the choice between short-term harmony, vs. probable short-term disruption in exchange for potential longterm performance improvements. Given the short-term evaluation horizon of many organizational performance evaluation systems, short-term harmony is an understandable choice.
Schneider's (1987) "Attraction-Selection-Attrition" (ASA) theory suggests that not only are managers likely to select applicants similar to themselves and the current workforce; applicants similar to the current workforce are more likely to be attracted to the organization in the first place. Ultimately, those who do not fit in become alienated and leave, creating over time a more homogeneous organization by attrition (George, 1990). Schneider's theory suggests than any diversification of the workforce-whether functional or demographic-is likely to meet with resistance (Schneider, 1987). The current membership may fear that any new subgroups will create disruption and conflict in the organization.
Furthermore, when managers see themselves in their (similar) employees, they are more motivated to help them succeed (Kanter, 1977). Consequently, potential mentors are likely to prefer similar others, which limits the advancement opportunities of dissimilar others. This cyclical process serves to maintain the status quo, as the dominant subgroup retains the positions of power in the organization (Kanter, 1977). Because diversity is beneficial, this path to workforce homogenization presents a clear and present danger to an organization's long-term competitiveness.
Cooperation in Organizations
As an organization becomes more diverse, it must overcome the practical problems that arise from differences in background, values, perspectives, and preferences. For example, individuals with different socialization and life experiences are not only going to differ in what they are interested in talking about; they are also likely to differ in how they talk about shared interests (e.g., Donnellon, 1993; Tannen, 1990). Effective communication is necessary for successful co-action (e.g., Tjosvold & Deemer, 1980), and miscommunication can create substantial barriers for heterogeneous groups that are trying hard to work together.
Stasser et at. (1989) found that individuals working in groups are more likely to focus on information that is shared among several group members, rather than information that is uniquely held by just one individual. If new perspectives help a diverse group of individuals to develop unique solutions to organizational problems, but that group's natural tendency is to ignore unique information, then access to one of diversity's major benefits is jeopardized. Diversity affects a group's internal communication and interaction dynamics (Pfeffer, 1983; Hoffman, 1985), which may eventually lead to absenteeism, member attrition, or even the dissolution of the group (e.g., O'Reilly, Caldwell, & Barnett, 1989; Tsui, Egan, & O'Reilly, 1992).
More troubling is the prospect that members of heterogeneous groups may not be trying hard to work together. Groups work best when they have shared or common preferences and goals (Tjosvold, 1984; Deutsch, 1949), and especially when these goals cannot be attained independently, or by a subgroup of members (Sherif et al., 1961). When diversity leads individuals to suspect a lack of goal congruence between themselves and dissimilar others, competition, derogation, hostility, distrust, and misinformation are all likely to follow (Sherif et al., 1961).
For these reasons, managers probably believe they benefit in the short term when they supervise homogeneous work groups. They feel comfortable with such groups, the groups are likely to be more cohesive, and managers probably assume that the conflicts that arise between group members are less likely to be serious. On the other hand, managers of homogeneous groups forego the long-term benefits associated with having a diversity of backgrounds, perspectives, values, and preferences represented in the group.
Dilemmas of managerial participation are similar in structure to dilemmas of individual participation, but they are greater in magnitude. Managers have more opportunities (e.g., selection and retention of workers, organization of work) to choose between short-term conflict and long-term benefits, vs. short-term harmony and comfort. Even if managers do not believe that diversity necessarily means disruptive conflicts within the workgroup, if they believe that their subordinates think that conflict is inevitable, workgroup homogenization may seem a safer short-term choice.
ADDRESSING THE DILEMMAS OF DIVERSITY
The structure of the argument to this point is that: (1) functional and social category workforce diversity offer benefits to an organization, but (2) managers and their employees-members of the organization-see themselves shouldering the apparent (and immediate) costs, but not necessarily enjoying the benefits, and certainly not short term. Because individual employees see few benefits for themselves, there is a collective fence surrounding diversity, and individuals therefore may not support an organization's diversity initiatives. Lack of support for diversity from their subordinates in turn creates both a temporal trap and a collective fence for managers, which causes them to be unsupportive. Finally, lack of support by managers and employees becomes the short-term cost that creates both a temporal trap and a collective fence for the organization. The long-term benefits of diversity appear relatively undisputed, yet the dynamics of these nested dilemmas lead many organizations to resist diversity.
Reframing diversity as a hierarchy of three interrelated social dilemmas underscores the unwieldy nature of the problem, and helps to explain why organizations often have been unable to capitalize on the apparent benefits of having a diverse workforce (Galen & Palmer, 1994). In order for organizations to embrace diversity initiatives, the temporal traps and collective fences surrounding individual and managerial participation in diversity must be removed.
When discussing possible resolutions to the social dilemmas of organizational diversity, Messick and Brewer's (1983) distinction between structural and psychological solutions is important. Structural solutions constrain short-term choices in order to ensure long-term collective interests. With psychological solutions, individuals instead are led to voluntarily forego their own short-term interests for the long-term collective good. Ultimately, all effective solutions to social dilemmas probably need to be psychological, in order to encourage adaptation (beyond that requested) and commitment to the value of diversity, rather than just empty compliance.
The question of addressing the three dilemmas exists at two levels. First is the management problem: What can interested organizations (i.e., their top managers) who appreciate the potential benefits of diversity do to assure effective participation by their resistant managers and employees? Second is the public policy problem: What can society do to assure effective participation by resistant organizations?
The Management Problem
Addressing the management problem effectively mean addressing the two subordinate dilemmas of individual and managerial participation. As noted earlier, the dilemma of individual participation is primarily a collective fence. Individuals perceive little reason to support organizational diversity efforts because they suspect diverse others to have different, mutually exclusive agendas, making diversity immediately and personally costly. The structural solution to this dilemma seems obvious: change incentives to align individual interests with organizational diversity initiatives. The psychological solution, in turn, is to lead individuals see that their selfinterests ultimately are aligned with workforce diversification.
Incentives can dramatically enhance cooperation in social dilemmas (Kormorita, Sweeney, & Kravitz, 1980; Messick & Brewer, 1983). Rewards may help solve the dilemma of individual participation because they can change the perceived short-term costs and benefits associated with participation in an organization's diversity initiatives. To address the dilemma of individual participation by using rewards, organizations must create measurable objectives for diversity, and then reward employees for achieving them (Hamner & Hamner, 1976; Luthans, & Kreitner, 1985). For example, mentoring someone from a different background is one easily measurable goal which could be added to an individual's annual performance objectives (Kanter, 1977). Individuals also might be rewarded for attending diversity workshops, or participating in cross-functional teams.
At the group level, organizations can create superordinate goals that encourage (make it in the best interests of) diverse members of the workforce to cooperate. Superordinate goals are objectives that cannot be achieved without cooperation across group boundaries in the organization (Sherif et al., 1961). One possible application of superordinate goals in organizations is to create interdependent performance incentives in the organization that require cooperation between functionally diverse groups or social categories. A more subtle version of this suggestion is the implementation of a Total Quality Management (TOM) program (e.g., Lawler, Mohrman, & Ledford, 1992). Emphasizing and rewarding a superordinate goal like quality within a diverse organization may not only improve quality, but also reduce conflicts and misunderstandings among diverse employees by providing them a unifying purpose, language, and culture, where none previously existed.
Part of the reasoning behind superordinate goals is that they will encourage intergroup contact and communication among diverse members of the workforce-communication and contact which eventually will lead individuals to recognize and internalize the value of workforce diversity. Increased communication among diverse colleagues will reverse discriminatory social identity effects when it enhances feelings of togetherness (Van Lange, Liebrand, Messick, & Wilke, 1992) or shared fate (Campbell, 1958; Kramer & Brewer, 1984; Rabbie & Horwitz, 1969; Wit & Wilke, 1992). However, contact between diverse groups does not necessarily improve relations between them; competition and hostility may result instead (Sherif et al., 1961; Wallace & Rothaus, 1969).
In order for contact among diverse groups to improve intergroup relations, the minimum requirements are: (1) some group members must be equal status (i.e., peers), (2) group goals must be interdependent, and (3) group reward distribution must be egalitarian (Cook, 1984; Stephan & Brigham, 1985). Contact between groups should also allow for intimate personal associations, and for opportunities to dispel stereotypes (Brewer & Miller, 1984; Cook, 1984; Desforges et al., 1991). Aronson and his colleagues demonstrated many of these principles when they installed "jigsaw" classrooms in a recently desegregated Texas school (Aronson & Bridgeman, 1979). Students in small, integrated cooperative-learning groups were each given a unique piece of a lesson. If the students wanted to do well on the lesson, everyone had to participate in the group discussion. Compared to children in traditional classrooms, children in jigsaw classrooms showed increases in self-esteem and in liking for school, decreases in negative ethnic stereotypes, and either improved or stable performance (Aronson & Bridgeman, 1979).
The current trend of creating cross-functional teams embraces the concept of the jigsaw classroom. Each member of the team brings unique expertise to the problem at hand, and success of the team may depend upon integrating all of the diverse perspectives and information represented in the group. Cross-functional or otherwise heterogeneous teams have the potential to increase intergroup cooperation in the organization, but only if two other issues are addressed. First, if only one member of the group is different-female, Hispanic, or an engineer-other members of the group may decide that this particular person is not like other members of his or her social category (Kanter, 1977; Miller & Brewer, 1984; Weber & Crocker, 1983). While the group may learn to appreciate this member as an individual, these positive attitudes may not generalize to other members of the category (e.g., Hewstone & Brown, 1986; Wilder, 1986). Choosing members of work teams as "tokens" to represent their social categories often has little beneficial effect on intergroup relations (Kanter, 1977).
Second, if a heterogeneous team is to work together effectively as a unit, subgroup composition should be category-independent (cross-cutting) rather than category-related (Marcus-Newhall, Miller, Holtz, & Brewer, 1993). When a team goal is highly complex, the team assembled is likely to be large and fairly heterogeneous with regard to both social categories and functional specialties (i.e., engineering, marketing, manufacturing, etc.). It is also likely that the task will be broken into subtasks for subgroups of the team to undertake independently. Allowing working subgroups to form based on social category, while comfortable for the group members, will not enhance intergroup cooperation, Experimental work by Marcus-Newhall et al. (1993) found that members of subgroups whose membership cuts across social categories and functional roles were more likely to see similarity among team members, and less likely to show bias in allocating rewards.
This suggests that social identity effects are less likely to be problematic precisely when social category diversity is not highly correlated with functional characteristic variety. In organizations, when Blacks and Whites, or males and females, can readily find similarities among their functional characteristics, they should be more likely to engage in personal interactions with one another (Marcus-Newhall et al. 1993), which in turn should reduce the discrimination they engage in based on social category distinctions. However, when social category membership is correlated with career choice-for example, when all of the engineers in an organization are male and all of the human resource specialists are female (Gutek, 1988; Kanter, 1977)-it may be difficult to break down group stereotypes and enhance cooperation across social categories (Weber & Crocker, 1983).
An important moderating variable in addressing the dilemma of individual participation in social category and functional diversity is group size. Several studies have shown that as the number of individuals in a group increases, social dilemmas become increasingly difficult to solve (Bonacich, Shure, Kahan, & Meeker, 1976; Fox & Guyer, 1978; Olson, 1965). One reason for this may be that it is easier to shirk responsibility in a large group than in a small one (Latane, Williams, & Harkins, 1979; Stroebe & Frey, 1982). Another reason may be that trust is more likely to have a positive effect on cooperation in a social dilemma in smaller groups (Sato, 1988). One implication of these findings is that the success of diversity initiatives may be sensitive to the size of the business units which must achieve diversity. In smaller business units, even though there are fewer opportunities to hire, diversity may have a greater chance for success.
It would be tempting to suggest that once the dilemma of individual participation is solved, the dilemmas of managerial and organizational participation should dissolve on their own. Unfortunately, this ignores the critical support role that managers play in addressing the dilemma of individual participation. If an organization is to successfully achieve diversity, managers must play that support role effectively. For managers, there is a collective fence around diversity because diversity is a temporal trap. If organizations emphasize the long-term in manager's performance appraisal and rewards, managers should feel encouraged to take a longer-term perspective-to accept diversity's short-term risks in exchange for its long-term benefits. In the final analysis, however, the easier it is for managers to realize the benefits of diversity immediately, the less of a dilemma diversity will represent.
An important factor in helping managers realize the benefits of diversity faster is the creation of a strong corporate culture which embraces diversity. A strong corporate culture will assist members of a diverse workforce in establishing a superordinate organizational identity, and in identifying shared, superordinate goals. Such cultures can be difficult to cultivate. Leaders who clearly articulate their vision for their organization seem to have the most success in establishing a strong corporate culture (Bennis & Nanus, 1985). If organizations hope to achieve diversity, the leader's vision of the organization must address diversity specifically, and emphasize tolerance for and appreciation of differences among individuals.
Of course, without support programs that confirm an organization's commitment to diversifying its workforce, any attempts to change an organization's culture can quickly take on the look of empty rhetoric. At Levi Strauss, senior executives have developed programs designed to hire, train, and support a diverse workforce, because they believe that their organization's success in the fashion industry is dependent on its first-hand knowledge of the preferences of its diverse customers (Mitchell & Oneal, 1994). Such support programs make the benefits of diversity more accessible to managers, effectively removing the temporal trap and collective fence that diversity represents.
The Public Policy Problem
Legislation has been a primary vehicle used to encourage resistant organizations to participate in diversity initiatives in the U.S. and other countries. A variety of laws have established "protected classes" in order to protect members of particular social categories from employment discrimination, The purpose of these laws is to break the Attraction-Selection-Attrition cycle of preference for similar others by giving preference to qualified members of protected classes when an organization's workforce does not adequately represent the diversity of the applicable labor pool (Conry, Ferrera, & Fox, 1993).
The paradox in this approach to diversity is that structural constraints upon organizational actions may actually exacerbate the dilemmas of individual and managerial participation. By creating "protected classes," legislation raises the salience of social category distinctions and thereby increases the probability of the very social identity effects that the legislation is attempting to defeat. Worse still, some employees may infer that their diverse co-workers obtained employment because of their social category membership, and not because they were qualified for the job (Northcraft & Martin, 1982; Galen & Palmer, 1994). If this inference means that new employees are not given the support and cooperation that their positions require, they are less likely to succeed (e.g., Word, Zanna, & Cooper, 1974), thus reinforcing the attribution that they weren't qualified in the first place (Kanter, 1977; Rosenthal & Jacobson, 1968).
In assessing the appropriateness of structural remedies, it is worth remembering that organizational pursuit of diversity is a burdensome dilemma precisely because the dilemmas of individual and organizational participation make participation for the organization costly. Structural solutions to the dilemma of organizational participation (such as legislation) impose costs for non participation. However, structural solutions at this level typically neither alleviate the underlying costs of participation (thereby removing the temporal trap), nor shift benefits to individuals and managers (thereby removing the collective fence). Such approaches thereby cannot address the underlying dilemmas facing managers and employees. A likely outcome is nominal participation by managers and employees, without ownership of the problems or real commitment to solving them. In the worst case, structural interventions may present enough of an apparent solution (i.e., the workforce appears representative) to undermine an organization's motivation to find and implement more effective remedies (e.g., Lepper & Greene, 1978).
At a public policy level, this means that legislation focused only on attaining workforce diversity in organizations (i.e., requiring organizations to have a representative workforce) either needs to go farther with target organizations, or the effort needs to be refocused elsewhere. Going farther might mean requiring target organizations to commit to (and be accountable for) detailed implementation plans that will improve the likelihood that their diversity initiatives will have sustainable success. If organizations interested in diversity identify techniques which improve the probability that organizational diversity initiatives will succeed (such as mentoring programs, or the use of group-level incentives to encourage positive, collaborative contact among diverse members of the workforce), these techniques should be required as part of a resistant organization's plan for workforce diversification. In contrast to some current legislation, this casts public policy in the role of solving the diversity problem, not just making the symptoms go away, i.e., not just making organizations diverse but effectively diverse.
Alternatively, public policy might refocus its efforts away from directly attaining workforce diversity in organizations to instead addressing the root cause of disruptive social identity processes in workers. Diversity dilemmas begin because individuals recognize some characteristics (like race or gender) as important for self-categorization. Aronson and Bridgeman's "jigsaw classroom" experiments (1979) suggest that the recognition of some characteristics as important for self-categorizing can be minimized by taking the interventions recommended for interested organizations (e.g., grouplevel incentives that encourage positive, collaborative contact among diverse students) into early school training. Of course, more research is needed to identify techniques for schools that will successfully minimize social categorization processes, and to assess the longevity of the effects those techniques engender. However, if schools could play a more effective role in breaking down the social stereotypes that encourage self-categorization and self-segregation (for example, along racial lines) before workers ever join the workforce, the necessity for government intervention in work organizations would diminish.
To reap the potential benefits of diversity, organizations must first become diverse and then must maintain that diversity. To attain and maintain diversity, organizations must solve the dilemmas of individual and managerial participation. The primary implication for future research of casting problems with diversity in terms of two well-established theories-social dilemmas and social identity-is that such an approach provides opportunities to generate theory-driven solutions to remedy these problems. For example, these two theories, taken together, highlight the important role that perception plays in the successful diversification of an organization's workforce. An organization's members experience social dilemmas when they perceive collective fences and temporal traps arising from conflicts between their personal goals and those of the collective, Social identity theory suggests systematic ways in which individuals may perceive this goal conflict, but be incorrect. To be successful, solutions to the three dilemmas of organizational diversity must correctly align employees' perceptions. Employees must come to see workforce diversity as a vehicle for achieving their goals, rather than as an impediment. This alignment will occur when employees perceive workforce diversity-whether functional or social category-as not necessarily implying mutually-exclusive goals.
No one really knows what the future holds, but everyone should know that diversity could represent an important social and organizational resource for facing it. If the importance and immediacy of this benefit can be accepted by individuals and their managers, then for organizations there should be no diversity dilemmas-only participation.[Reference]