or the organization of legislatures, three possible explanations might be offered. First, if leadership norms are essentially norms defined by a predominantly male culture or by historically male institutions, these data may be explained as Stivers suggests: Women face more pressure than men to adapt, modify, or conform their styles to the dominant norms. Because institutional conditions of professionalism and the presence or absence of women and women's power vary, women respond to more complex and conflicting pressures, which in turn leads to greater variability of leadership styles.
By focusing on the smaller variability in men's scores, a second explanation might be offered: Perhaps institutional and cultural norms provide more constraints on men's leadership behavior than on women's behaviors. In an institution or organizational role that is historically male, it might be the case that organizational conformity exerts greater pressure on men. To be sure, women face the same institutional norms, but perhaps because they are pioneers in a sense, they are in effect freer to set their own course. Contrary to the findings of the authors of Breaking the Glass Ceiling, it may be the men in legislatures, rather than the women, who must conform to a narrow band of acceptable leadership behaviors. 70
A third explanation might be that legislatures are different from other organizations and thus findings from other organizations are not always applicable. Leadership in public organizations tends to flow to those who embody the norms and values of the organization. Legislatures differ from other organizations in terms of the higher turnover of members and a less hierarchical structure that recognizes the presumptive equality of each member. Instability and greater equality might negate the powerful socializing norms evident in other organizations. Because legislatures have less stable organizational norms, different styles of leadership brought by women may offer new possibilities for influence and change.
This analysis underscores the centrality of gender as a defining element of leadership style. Gender shows its impact in three ways. First, sex is a significant predictor of aspects of integrative leadership behavior even when controlling for different situational variables. The differences are not biological but presumably result from different life experiences and socialization. Second, and just as important, gender manifests itself in feminine and masculine personality traits and, therefore, demonstrates a significant influence on leadership styles. Third, gender shows its influence as an organizational phenomenon when we specify the effects of more women in a legislature or more power held by women in a legislature.
Integrative leadership behaviors and motivations are also quite variable given the demands of place, institution, and time. Contextual circumstances matter. Although aggregate-level analysis cannot begin to reveal or fully ex-