Wages and Unequal Access to Organizational Power: An Empirical Test of Gender Discrimination

Administrative Science Quarterly,  Sept, 1999  by Mia Hultin,  Ryszard Szulkin

• E-mail
• Print
• Link

One basic assumption in discrimination theories is that ascribed characteristics are of great importance for how employees are treated and for how scarce rewards are distributed in the labor market. Several researchers have pointed out that studies of labor market inequality ought to take into consideration the impact of those actors who have direct influence over organizational procedures and policies, that is, managers and supervisors (Marini, 1989; Baron, 1991). According to several scholars, work organizations can be treated as arenas on which social conflicts between management and labor as well as between different groups of employees take place (Kalleberg, Wallace, and Althauser, 1981; Acker, 1987; Baron, 1991; Tomaskovic-Devey, 1993). One implication of such a perspective on organizations is that the distribution of rewards can be seen as a process in which conflicting interests are manifested. Thus, the reward structure to some extent reflects different groups' relative power within the work organization. This perspective is clearly stated by Pfeffer (1989: 389), who argued that "wages are a resource and, like other resources, are allocated at least in part on the basis of the power of various interests." These interests may be defined along various lines, such as social class or demographic characteristics. According to the literature on labor market discrimination, gender is one of the most influential lines of cleavages in distributive conflicts over scarce resources.

Earlier studies of gender discrimination have demonstrated unequivocally the existence and persistence of gender-based wage differences in the labor market (e.g., Rosenfeld and Kalleberg, 1990; Blau and Kahn, 1992; for Sweden, see le Grand, 1994). Several scholars have tried to explain how these differences are influenced by gender composition at the occupational level (e.g., England, 1992), at the organizational level (e.g., Groshen, 1991a), and at the job level (e.g., Treiman and Hartmann, 1981). It has been found in general that wages are relatively low in labor market structures in which many women work. Pfeffer and Davis-Blake (1987) showed in their study of administrators in colleges and universities that the proportion of female incumbents depressed the wages for both male and female administrators. Huffman and Velasco (1997) presented empirical evidence indicating that relatively high numbers of women among the managers in an organization is associated with wage penalties for managers. Even though both of these studies pertain to incumbents of relatively high organizational positions, the results primarily corroborate the more general finding that a high representation of women in a specific category can serve as a negative signal about the relative value of the work performed by those in this category. The aim of the present study is different. Our aim is to move beyond this work, toward a more detailed understanding of the gender wage discrimination process, by studying how gender differences in access to organizational power structures influence the wages of male and female subordinates. We investigate the idea that women's limited access to organizational power structures is a constituent part of the explanation of gender wage inequality. The empirical question we address is whether and how the gender composition of the managerial and supervisory staff in organizations influences the earnings of female and male employees in subordinate positions. To our knowledge, there are no previous large-scale quantitative studies that address this specific research question. Our empirical analyses are based on individual-level data from the Swedish 1991 Level of Living Survey and organizational-level data from the Swedish 1991 Establishment Survey.(1) Given the relatively egalitarian character of the Swedish labor market, if we find discriminatory wage-setting practices in Sweden, they may be viewed as a conservative estimate of their importance in other developed countries.

Wage-setting Processes in Sweden

There are reasons to believe that the impact of organizational characteristics on earnings is smaller in Sweden than in many other countries. When analyzing the specific features of the Swedish labor market, observers have emphasized the high degree of institutionalization and regulation (Edin and Holmlund, 1995; Edin and Topel, 1997). The most prominent feature mentioned in this context is that wages from the mid 1950s to the early 1980s were determined to a large extent through centralized collective bargaining at the national level. A solidaristic wage policy aiming at equal pay for equal work, regardless of the profitability and productivity of the firm or industry, has been widely pursued. This policy had the explicit purpose of minimizing wage variation between equal jobs across firms and sectors. The very high union coverage rate in Sweden in combination with a well-organized and, until recently, highly centralized confederation of employers facilitated implementation of a solidarity wage policy.(2) This policy resulted in a decrease in the overall wage dispersion in the Swedish labor market (Hibbs, 1991). A number of recent empirical studies have demonstrated that decentralized wage bargaining produces relatively large wage inequality in general (Barth and Zweimuller, 1995; Blau and Kahn, 1996; Elliott and Bender, 1997) and large gender wage differentials in particular (Hammond and Harbridge, 1995). The centralized wage bargaining process in Sweden, aimed at raising the relative wages of low-wage workers, may indirectly have resulted in a relatively small wage gap between women and men. Empirical findings from international comparisons of gender wage gaps show that a society's overall wage inequality is positively related to the gender wage gap. The Swedish labor market is characterized by relatively low wage dispersion and, hence, comparatively small wage differentials between women and men (Rosenfeld and Kalleberg, 1990; Blau and Kahn, 1992), although even during the regime of centralized bargaining, a far from negligible adjustment of wages occurred at the firm level in the form of wage drift.