Feminists argue that to a large extent what appear to be gender dif­ferences in personal style and interpersonal relations may be explained largely on the basis of differences in status or power that are embedded in gender (Hyde, 1995; Unger & Crawford, 1996). For example, experimental studies have demonstrated that what is colloquially understood as women’s intuition may be the effect of subordinate role on interpersonal sensitivity (Snodgrass, 1985). Although there are positive and legitimate ways to ex­ercise power and authority, often referred to as empowerment or the power to, we focus here on power as control over others (Yoder &. Kahn, 1992).

Power can be exercised in a variety of subtle ways (Hyde, 1995). Dominant people or groups can limit the opportunities of subordinate groups and are more likely to ignore or pay less attention to subordinates. Dominants also can deny the legitimacy of subordinates’ experience.

Furthermore, power can be exercised in the course of interaction epi­sodes and can be exercised by shaping the context in which transactions occur. For example, power can affect the availability of alternatives, and, more particularly, high-power people can prevent low-power people from developing alternatives. Social exchange theory suggests that when relation­ship alternatives are limited, a low cost-benefit ratio will be tolerated. That is, women may not protest sexual harassment if they perceive that they have few alternative options for employment. Recent exchange theorizing directs our attention to coercive power as well as reward power (Molm, 1997).

Power or standing may affect the interpersonal style used in negoti­ations. For example, high-status people often tend to negotiate directly, using forthright bargaining, direct asking, or reasoning; whereas low-status people are more likely to use indirect styles, such as, lying low, being eva­sive, hinting, being nice, withdrawing, and so on (Falbo & Peplau, 1980). Falbo and Peplau (1980) observed in a study of negotiating among couples that women were more likely to rely on these indirect styles. Other studies of couples (Gruber & White, 1986) have found related patterns, but ob­served that although men report using both male and female styles, women were more restricted in the types of strategies they adopted.

In recent decades the idea that interactants are aware that interaction is rule directed and that choices are made concerning which rules to follow has gained currency (Harre & Secord, 1973). Although commonalities in behavior exist, a female-male expressive-instrumental dichotomy appears for directness of approach, with men directly seeking to realize their ends and women wanting to develop rapport with men first (Midwinter, 1992).

One may be tempted to offer a rather simple recommendation that women could solve these problems if they would only act more like men. Besides problems with the implicit value system of such a recommendation, there are other practical difficulties. First, due to inertia as well as vested interest in maintaining the status quo, efforts to change dominance pat­terns are likely to be met with resistance. Furthermore, violation of ex­pected gender roles and styles are sometimes met with negative sanctions. Experimental studies of group dynamics that varied tentative versus asser­tive style by gender of speaker found that women who used assertive (tra­ditionally male styles) were liked less than women using traditional female (tentative) styles (Carli, 1989). And although tentative female speakers were judged to be less confident and less competent, they were at least more persuasive with a male audience than were women who exhibited a more expert style.

We suggest that such contrasting styles reflect differences in power (e. g., prestige, position power, access to resources, or ability to influence or orchestrate) that overlap. Thus, women as a class may generally appear more tentative, more uncertain, and more vulnerable, thereby increasing the likelihood for attempted exploitation and harassment. Serious problems arise when mechanisms for dealing with harassment impose an assertive (traditionally male) style of response on women who have been encultured with a style of low entitlement; we will discuss this in the following section on consent.

The power dimension ranges from complete equality at one end to total domination at the other (with one having complete authority over

another). The specific ways in which power leads to sexual harassment have been more the subject of speculation than of research (Cleveland & Kerst, 1993). Gender arrangements in society may generate power differ­entials. Specifically, men may have greater power based in heterosexuality as a political institution according to feminist Adrienne Rich (1980). Dif­ferences in behavior by gender may reflect power differences (Yoder & Kahn, 1992).

The formal organizational power a person has due to occupancy of a position is central to whether unwanted sexual gestures directed at a sub­ordinate are perceived as harassment. Researchers such as Bursik (1992) argue that this power imbalance is a crucial contextual variable shaping interpretations of situations as sexual harassment. In Bursik’s research, whether behavior was perceived as harassment was linked to how blatant it was and the power of the perpetrator; more blatant behavior was con­sistently seen as harassment, whereas less blatant behavior was defined as harassment when committed by a more powerful person.