Effects of team gender and racial composition on perceptions of team performance in cross-functional teams.

There is no question that the workforce in the United States is changing. Great interest in workforce diversity has been generated by the Workforce 2000 report (Johnston, 1987), which forecasts large increases in the number of women and minority group members entering the workforce by the year 2000. This report has generated renewed interest in the effects of gender and racial diversity on organizational behavior.

A second trend is an increasing reliance on work teams to solve the complex problems that organizations now face. Teams bring a diversity of functional skills and expertise to bear on complex problems. Teams also bring together a diversity of people. The increasing representation of women and minorities within organizations has naturally led to an increase in their representation on work teams. The purpose of the present study is to investigate the impact of limited diversity with respect to gender and racial composition on perceptions of team effectiveness among team members and among external evaluators relative to teams that are homogeneous with respect to gender or ethnicity.

EFFECTS OF GENDER COMPOSITION

The effects of gender composition in work groups have received some empirical attention. Ely (1994, 1995) found that a higher proportion of female partners in a law firm (15%) influenced beliefs and attitudes among female associates. Female associates viewed female partners more positively, evaluated women's attributes more favorably relative to criteria for success, and perceived less psychological and behavioral difference between men and women in firms with 15% or more female partners compared with female associates in firms with less than 15% female partners. Women in more male-dominated firms experienced more conflict among women and found that permissible role behavior was more constrained and more stereotyped than in the gender-balanced firms, a finding that echoed Kanter's (1977) classic study.

Looking at a similar phenomenon at an earlier point in time, women's responses were influenced by the sex ratio in law school. Female students in a law school with a tilted (33% women) sex ratio were similar to their male peers in terms of achievement and social integration, whereas women in a law school with a skewed sex ratio (20% women) either over- or under-compensated in terms of achievement and social integration (Spangler, Gordon, & Pipkin, 1978). These results suggest that the stereotyped gender roles are found even in the early socialization process.

The proportion of women in an organization had a curvilinear effect on their salaries, with lower salaries being paid to women in organizations with low and high proportions of female employees. The proportion of women employed in the firm had no effect on men's salaries (Shenhav & Haberfeld, 1992). The proportion of women in management positions, however, had a negative effect on men's salaries, to the point where men's and women's salaries were roughly equal (Shenhav & Haberfeld, 1992).

Organizational commitment and intention to stay were negatively related to dissimilarity from the team with respect to gender, and absence was positively related (Tsui, Egan, & O'Reilly, 1992). Further, the effects of dissimilarity were stronger for men than for women. Moving to the dyadic level, mismatch on gender between superior and subordinate had a negative effect on the superior's ratings of the subordinate's performance and the superior's liking for the subordinate, as well as positive association with the subordinate's role ambiguity and role conflict (Tsui & O'Reilly, 1989). Gender composition at the team level and dissimilarity from the majority of the team with respect to gender at the individual level both influence important individual and organizational outcomes, especially among women. It also seems that the types of roles that women fill influence outcomes. It is less clear how gender diversity affects men's responses or how gender diversity within a team influences the team's process. It appears that variability in gender composition may have negative effects on both men and women.

EFFECTS OF RACIAL COMPOSITION

Racial composition has also been the subject of research. In a longitudinal study, variation with respect to race was found to affect both member-reported team process and performance on a team project among teams (four or five members) of college students. Homogeneous teams were all White, whereas racially heterogeneous teams included a White, Black, and Hispanic American, as well as an international student. Racially heterogeneous teams initially reported poorer team process and performed less well than homogeneous teams. Both types of teams improved over the 17-week semester, but the racially heterogeneous teams improved enough to catch up with the homogeneous teams by the end of the semester (Watson, Kumar, & Michaelsen, 1993).

Sessa (1993) assigned a case analysis project to temporary teams of hospital employees. Among these short-lived teams, those that varied in racial composition exhibited more conflict than racially homogeneous ones. Hoffman (1985) found a positive relationship between the proportion of Black supervisors in state government agencies and the amount of organizational communication in the form of formal meetings. These results suggest organizational communication may become more formalized when racial composition at the supervisory level becomes more variable. Although not suggested in the results, it is possible that increasingly formalized communication is employed in an attempt to control or manage conflict (Mintzberg, 1979) or as a substitute for informal communication (Shaw, 1981).

Using meta-analytic techniques, Kraiger and Ford (1985) found that supervisors provided higher performance ratings to subordinates of the same race as themselves. This effect became more pronounced for White supervisors as the percentage of Blacks in the work group increased. Minority managers also reported poorer fit with their work group after 9 months on the job than nonminority managers (Kirchmeyer, 1995).

Dissimilarity from the work group on race was negatively associated with organizational commitment and intent to remain on the job and was positively associated with absence. The dissimilarity effect was stronger for Whites than for racial minorities (Tsui et al., 1992). Minority graduate students reported better attitudinal responses in heterogeneous teams than in homogeneous teams, but team composition had no effect on attitudinal responses …