Idle Rights: Employees' Rights Consciousness and the Construction of Sexual Harassment Policies

Abstract

This article analyzes women's legal consciousness in responding to unwanted sexual attention in the workplace. By focusing on a particular social problem, this study is situated in the particular legal domain of sexual harassment laws and in a specific organizational context. Taking the perspective of the intended beneficiaries of sexual harassment policies and procedures—women with complaints about sexual conduct in the workplace—I show that the implementation of grievance procedures creates powerful obstacles to women's efforts to assert those rights. Moreover, the practices implementing the policies can alter the very definition of sexual harassment in that setting. Thus, in enacting grievance procedures, women and supervisors construct a legality in particular workplaces that offers only limited protection for women's rights.

Kimberly Ellerth was a salesperson for Burlington Industries in Chicago. Her supervisor, Ted Slowik, subjected Ellerth to a stream of offensive remarks and gestures and suggested several times that her job depended on complying with his sexual demands. Just before Ellerth quit, Slowik said, “Are you wearing shorter skirts yet, Kim, because it would make your job a whole lot easier” (Burlington Industries, Inc. v. Ellerth 1998:747–8). Ellerth resigned, but only informed Burlington three weeks later that her reason for leaving was Slowik's harassing behavior. Like many women confronting sexual harassment, Ellerth relied on a variety of strategies to cope with Slowik's conduct. She confided in her husband and her parents; she told colleagues and clients. Once, she confronted Slowik directly and told him that his comments were inappropriate. But Ellerth never complained to other supervisors, even though she knew that Burlington had an anti–sexual harassment policy. She claimed that she was afraid that pursuing a complaint would jeopardize her job (Burlington Industries, Inc. v. Ellerth 1998).

Ellerth's case presents a number of interesting empirical and theoretical questions about the role of law in everyday encounters with sexual harassment. For example, her experience illustrates the many different tactics that women use when confronting sexual conduct at work, most of which involve trying to avoid the harasser and seeking emotional support to cope with the situation. Ellerth's case also shows how the law on the books often promises more than law in action can actually deliver (Bumiller 1988; Quinn 2000). When women like Ellerth fear retaliation for exercising their rights, then the remedial policies and procedures may be inadequate to address the underlying problems.

Many significant law and society studies have documented patterns of legal mobilization among average people confronting conflict in their daily lives (Jacob 1969; Felstiner, Abel, & Sarat 1980–81; Kritzer, Vidmar, & Bogart 1991; Zemans 1983; Curran 1977; Bumiller 1988). These studies have demonstrated that although legal rights and entitlements may be formally available, they are rarely invoked. There can be considerable difference between what people are entitled to under law and what they actually receive (Zemans 1983; Jacob 1969). Thus, this article fits into a long tradition of exploring this gap between the law on the books and the law in action, but it builds on this tradition by including in the “law in action” the meaning-making activities of ordinary women confronting unwanted sexual attention.

In this study, I examine rights at work in a grievance procedure in a single workplace. I adopt a legal consciousness perspective that emphasizes the experiences of employees with complaints about unwanted sexual attention from co-workers and supervisors. Indeed, this study confirms that management interpretations shape the way supervisors implement the policy, although in this case, supervisors act to discourage complaints rather than to offer dispute resolution assistance to aggrieved employees (Edelman, Erlanger, & Lande 1993). But just as important are employee responses to these management practices. Women anticipate skeptical treatment by their supervisors and develop strategies to meet that skepticism. As a result, women complain about only the most serious or most troubling forms of sexual conduct, thus enacting a legal consciousness that reflects a narrow meaning for “sexual harassment.”

Building a Theory of Legal Consciousness in Action: Organizations and Women's Experience

Grievance procedures are the most common mechanism for enforcing employee rights in the workplace, but there are serious questions about their effectiveness. Implemented by supervisors and managers, grievance procedures may actually be more effective at protecting employers from liability than they are at protecting employee rights. In this section, I review studies grounded in institutional theories of organizations that critique internal grievance procedures and show that research about women's responses to sexual harassment offers some support for this critique. I also propose incorporating employees' perspectives using a model of legal consciousness in action, situated in the legal rules surrounding sexual harassment and a particular workplace.

Institutional Theories of Grievance Procedures and the Case of Sexual Harassment

The U.S. Supreme Court has great faith in the power of employers' grievance procedures to resolve women's problems with sexual harassment. Internal dispute resolution mechanisms, such as anti-harassment policies, were designed to provide much-needed dispute resolution resources to less-powerful members of organizations. These procedures are less costly than formal legal processes because they lack the barriers to access associated with the formal legal system; there are no rules of evidence, no burdens of proof, and lawyers are usually excluded (Kihnley 2000; Hunter & Leonard 1997; Goldberg, Green, & Sander 1986; Hill 1990). In addition, grievance procedures can build on legal protections by addressing employee needs rather than just employee rights (Menkel-Meadow 1985; Hill 1990).

Yet critics argue that grievance procedures are inadequate to protect employee rights or to dismantle structural inequality in the workplace because they are susceptible to the prejudices and power disparities that exist in organizations (Delgado et al. 1985; Bobo 1992; Edelman, Erlanger, & Lande 1993; Gutek 1992; Kihnley 2000). In the workplace, for example, employees are formally subordinate; it may be difficult to direct complaints against supervisors who are more powerful actors even when the right to complain exists. Moreover, because they are confined to individual organizations, such procedures do nothing to advance public rights (Harkavy 1999; Kihnley 2000).

Institutional theorists of organizations, such as Edelman and her colleagues, have shown that grievance procedures are as likely to protect organizational interests as employee rights. Organizations adopt such policies for symbolic purposes, to demonstrate organizational commitment to prevailing norms and values, such as fair dealing and equal opportunity, reflected in civil rights legislation and the expansion of due process protections (Edelman, Erlanger, & Lande 1993; Edelman 1990; see also Dobbin & Sutton 1998; Sutton et al. 1994). For example, such procedures give employees the right to challenge unfair supervisory decisions, a type of due process protection (Edelman 1990). But in administering grievance procedures, managers face competing obligations; they must shield the employer from liability even as they try to redress employee grievances (Edelman, Erlanger, & Lande 1993; Kihnley 2000).

In interviews with human resources professionals, Edelman, Erlanger, and Lande (1993) found that managers resolve this tension by protecting the organization rather than the employees. Specifically, human resource professionals reframe workplace disputes as management lapses or personality conflicts and deny that there are any problems with discrimination in the workplace. Thus, they interpret employee rights in terms of management interests rather than civil rights. While managers often try to settle disputes among employees, they rarely do so to vindicate the principle of equal opportunity, leading Edelman and her colleagues to observe, “The legal right to a nondiscriminatory workplace in effect becomes a ‘right’ to complaint resolution” (Edelman, Erlanger, & Lande 1993:529).

But how do employees react to these policies and procedures? The existing research on sexual harassment provides some evidence suggesting that management practices make employees cautious about pursuing complaints. Many studies show that women rarely report their experiences with sexual harassment to third parties (Gutek 1985; Fitzgerald, Swan, & Fischer 1995; Gruber & Smith 1995; Merit Systems Protection Board 1995; Welsh 1999).¹ Targets of harassment complain most often when the conduct is severe or pervasive (Gruber & Smith 1995; Cochran, Frazier, & Olson 1997) and when the harasser is a co-worker rather than a supervisor (Gruber & Smith 1995).² Complaints are also more common when the employer has a policy or procedure in place (Gruber & Smith 1995; Gruber 1998), but studies also reveal employees' ambivalence about these policies and the personnel who administer them. For example, when asked in a survey why they did not complain, federal employees said that they were worried that they would be blamed for the incident, that they would not be believed, or that the complaint would not be kept confidential (MSPB 1995). The MSPB survey respondents were also concerned that management's reaction to the complaint would be at best ineffectual and at worst threatening.³ These findings suggest that employees perceive grievance procedures to be adversarial and hostile processes.

Emerging qualitative research on sexual harassment has begun to illustrate the organizational and institutional underpinnings of employees' skepticism about anti-harassment policies and procedures. For example, in a recent study of a university's sexual harassment policy, Kihnley (2000) found that supervisors were more likely to take the side of the alleged harasser than the employee making the complaint. As one official reported, “‘[W]hen a complaint is made, often times [the] complainant becomes an outsider, a troublemaker, and the harasser becomes the institution’” (Kihnley 2000:80). Ethnographic studies of temporary workers and female firefighters find that resistance to sexual harassment is shaped by specific organizational settings characterized by particular power arrangements (Rogers & Henson 1997; Yoder & Aniakudo 1995). Of course, there are other obstacles to women's complaints, beyond the employer's organizational practices and routines, such as the embarrassment and psychological costs associated with such complaints (Morgan 1999; Quinn 2000). Morgan (1999) has demonstrated the way that women's familial roles are inconsistent with the aggressive pursuit of grievances. Still, the research on sexual harassment suggests that because they reflect the power dynamics at work in particular organizations, anti-harassment policies and procedures may dampen rather than promote employee complaints.

This study builds on institutional analyses of organizations by adopting the perspective of employees rather than managers. Managers implement grievance procedures by interacting with employees who have complaints. In fact, employees determine the agenda of the grievance procedures by choosing which incidents to report and which to ignore. In making these decisions, employees rely on their supervisors for signals about how their rights will be enforced. Thus, employees and their expectations should be included in analyses of how grievance procedures work.

Legal Consciousness in Action

To include employees' participation in the construction of meaning of sexual harassment policies, I rely on the framework for studying legal consciousness. In everyday locations, such as workplaces, ordinary people make sense of their experiences by relying on legal categories and concepts even though they may not be familiar with the details of formal rules and regulations (Ewick & Silbey 1998; Sarat & Kearns 1995; Marshall & Barclay 2003). Law provides cultural schemas that people use to understand their everyday experiences. Ewick and Silbey define these schemas broadly as “legality”: “the meanings, sources, authority and cultural practices that are commonly recognized as legal, regardless of who employs them or for what ends” (Ewick & Silbey 1998:22). Thus, legality embraces such widely familiar concepts as “property,”“evidence,” and “sexual harassment,” which provide commonsense categories for everyday experiences while at the same time implicating specific rights, claims, and legal procedures.

Like other schemas, legality does not reside exclusively in an individual's ideas and attitudes. Instead, to maintain its vitality, legality “must also be continually produced and worked on—invoked and deployed—by individual and group actors” (Ewick & Silbey 1998:43; Sewell 1992). Grounded in social and cultural practice, legal consciousness is the individual's “participation in this process of constructing legality” (Ewick & Silbey 1998:45). In this constitutive theory of law, then, even as legality shapes the meaning of everyday life, ordinary people reshape the meaning of legality as they deploy legal meanings in new settings. According to Ewick and Silbey,

Every time a person interprets some event in terms of legal concepts or terminology—whether to applaud or criticize, whether to appropriate or resist—legality is produced. The production may include innovations as well as faithful replication. Either way, repeated invocation of the law sustains its capacity to comprise social relations. (Ewick & Silbey 1998:45)

Yet legal consciousness is more than just the meaning that people assign experience; it must also include the social and cultural practices of enacting those meanings (Ewick & Silbey 1998; McCann 1994; Marshall & Barclay 2003). Legal consciousness is reflected in what people say and do, in addition to what they think (Ewick & Silbey 1998; Merry 1990). Indeed, the expansion of civil rights logic—from African Americans (Tushnet 1987) to women (MacKinnon 1979; McCann 1994) and most recently to gays and lesbians (Eskridge 2002)—emerged from litigation campaigns based on these new meanings. But law can also be reproduced—and transformed—in the context of everyday situations and conflicts, in conversations with supervisors and confrontations with co-workers. Indeed, it is in these interactions that the difference between constitutive and instrumental conceptions of law begins to collapse and gives law its expansive power to define widening realms of experience.

Yet legality may also lose its power “to comprise social relations” if it falls into disuse. When a person ignores or rejects some right or benefit authorized by law, those rights remain idle. Indeed, rights may contribute to the individual's understanding of events and experience (Engel & Munger 2003; Marshall 2003; Nielsen 2000). However, by declining to enact those rights in any context—by ignoring them or by rejecting their significance to remedy an injustice—individuals decline to participate in the continuing reconstruction of legality that gives vitality to law. When rights remain idle, law's ability to shape meanings and opportunities and practices is diminished (Quinn 2000; Bumiller 1988).

Many recent studies have provided “account[s] of how a cultural, constitutive theory of law actually works” (Mezey 2001:161). Situated in concrete locations where ordinary people confront specific problems,⁴ these studies have generated theoretical insights into such issues as the differences in legal consciousness along lines of social stratification (Nielsen 2000), the salience of law in particular settings (Levine & Mellema 2001), and the way law interacts with other frames to shape the meaning of everyday life (Marshall 2003). In this article, I further develop the theory of legal consciousness in action by locating the study in a specific problem, a set of legal rules, and a particular organizational setting, an employment grievance procedure.

First, the specific problem in this study—sexual harassment—is the subject of extensive legal regulation. By linking legal consciousness to specific legal rules, I can show that those rules, and the institutions that enforce them, can actually shape the meaning that ordinary people give everyday events. For example, equal employment laws confer rights that are supposed to redress structural inequalities, and employers' grievance procedures are supposed to assist employees in realizing those rights. However, workers face meaningful obstacles in exercising those rights (Edelman, Erlanger, & Lande 1993; Fitzgerald, Swan, & Fischer 1995). Thus, rights are an arena for struggles against inequality, or they can ratify existing conditions of domination and subordination (McCann 1994; Merry 1995; Quinn 2000; Bumiller 1988).⁵ In this study, by examining both employees who did not complain and those who did, I can analyze the sometimes contradictory consciousness of law where law provides oppositional meanings and opportunities for resistance, but mechanisms for enforcing the law—in this case, the sexual harassment policies—can undermine this resistance.

Second, by situating the study in a particular organizational setting—a grievance procedure in a single workplace—I can show that organizational practices may influence individuals' legal consciousness. Although they concentrate on the development of disputes in everyday life, studies of legal consciousness can be abstracted from the neighborhoods, schools and workplaces where everyday problems develop (Ewick & Silbey 1998). As a result, these accounts fail to consider ways that organizational practices reshape legal rules. Everyday disputes often develop in the context of organizations and institutions that have their own routines and practices that influence the meaning of the experience (Heimer 1999). Organizations themselves are an important source of messages about law, translating legal rules in ways that serve organizational interests (Edelman, Fuller, & Mara-Drita 2001; Edelman, Uggen, & Erlanger 1999; Edelman, Erlanger, & Lande 1993; Edelman 1990). When interacting with organizational actors, people absorb these translations and reinterpretations, adopt them, question them, and sometimes act on them, thus contributing to the continuing reconstruction of the meaning of legality (Merry 2003). In this context, rights cannot disrupt inequality when organizational routines re-enforce those inequalities.

Adding legal rules and organizational practices to the legal consciousness framework provides much needed context that can reveal how the constitutive theory of law actually works. The legal rules and rights surrounding sexual harassment shape women's responses to their experiences with unwanted sexual attention at work, but so do management practices in the shadow of those laws. After describing the design of the study, I analyze interactions between supervisors and women with complaints about sexual conduct in a single workplace. I show that these interactions narrowed the meaning of sexual harassment for women working there.

The Design of the Study

The analysis in this article is based on a multimethod approach to studying women's experiences with unwanted sexual attention at work. I relied on both in-depth interviews and a survey in a single workplace to understand working women's decisions about whether to complain about their encounters with unwanted sexual attention. The survey confirmed women's widespread reluctance to complain, while the interviews revealed the way the sexual harassment policy and procedures shaped women's reasoning process in evaluating the value of pursuing a complaint. By triangulating the study with these different methods, I have greater confidence in the findings.

I conducted the research in a single workplace—a large university in the Midwest (the “University”)—and restricted the subjects to female staff members in administrative and clerical positions.⁶ I excluded faculty members and students from the sample unless they occupied administrative or clerical positions so that the employees in the sample would resemble employees in work settings other than universities. Situating a study of legal consciousness in a single organization has advantages. All the employees participating in the study were subject to a single set of legal rules and, more important a single set of policies and procedures. Thus, differences in behavior and meaning cannot be attributed to differences in law or policy. However, as I will show, the way that those policies were administered did vary.

I solicited the in-depth interviews with female employees through an e-mail message posted to a listserv. The listserv was sponsored by an organization of female employees at the University. The organization provided social, cultural, and career programs for its members, arranging outings to museums and concerts and sponsoring other social gatherings, such as holiday parties. The organization also offered workshops to give women career advice and networking opportunities. Although not an advocacy group or a union, the organization sponsored talks and brown-bag discussions on such topics as managing stress, the glass ceiling, and violence in the workplace. Members of this organization might be expected to be more attentive to University policies, which might, in turn, have yielded a sample that was more likely to pursue complaints. However, 24% of the women interviewed pursued grievances with their supervisors, compared to 16% of the survey respondents, who were drawn from the entire population of female employees at the University.

In my e-mail message to the listserv, I asked women to contact me if they had experiences with “unwanted sexual attention in the workplace.”⁷ I used this phrasing to leave open the question of whether the women had been sexually harassed, a topic of the interviews. Twenty-five women contacted me in the two weeks following my e-mail, and I was able to interview all of them. They varied in income and occupational status. Five of the women were low-paid clerical workers who administered budgets and performed clerical tasks. The rest were middle-management employees performing a range of administrative tasks, including supervising employees and developing workplace policies. Of course, using a computer-based method of communication limits the type of employees I reached—only women who had access to computers, only women in pink- and white-collar occupations—but my sample did reflect variation across occupations. One of the women interviewed is Latina; the remainder are white. The women chose pseudonyms to preserve their confidentiality; they are identified by these pseudonyms in this article. Lasting from 45 minutes to an hour and a half, the interviews were tape-recorded and transcribed.

I used a semi-structured battery of questions to conduct the interviews. I asked the women about their experiences with unwanted sexual attention at work. I also asked them about all the things they did in response, including formal and informal methods to redress the situation. In asking these questions, I was also able to discern their experiences with and observations of the University's sexual harassment policies and procedures. At the end of the questions about their experiences, I asked them whether they thought the behavior was sexual harassment. I departed from the schedule when the subjects wanted to elaborate about a particular topic.

Conducted after the interviews, the survey was designed to confirm the general patterns detected in the in-depth interviews. In particular, I sought to document women's experiences with unwanted sexual attention at work, their use of the label sexual harassment to describe those experiences, and their responses to such behaviors. To conduct the survey, I obtained a list of all female support staff at the University from its Women's Center. After deleting the interview subjects from the list, I drew a random sample of 1,000 subjects. I sent a questionnaire to each respondent along with a cover letter briefly describing the goals of the research and a follow-up letter asking them to complete the questionnaire if they had not already done so.

The questionnaire elicited information about women's experiences with unwanted sexual attention at work and their labeling such experiences as sexual harassment. For the purposes of this article, the most relevant items asked respondents what they did in response to the incident that “made the greatest impression” on them, and if they did not complain, their reasons for declining to do so. In addition, survey respondents had an opportunity to provide more detailed accounts of the experiences. At the end of the questionnaire, women had space to write “their comments on any of the issues raised by [the] questionnaire.” About 20% of the respondents used that space to provide more detail about their experiences and to elaborate on their efforts to respond to their harassers. I have included these comments in this analysis.

The response rate for the survey was 35%, an acceptable rate for mail-back questionnaires, particularly on a topic as sensitive as sexual harassment, where response rates can be as low as 20% (Arvey & Cavanaugh 1995:46). Moreover, there were no obvious sources of bias among the women responding to the survey. The sample reflected the diversity of pink- and white-collar occupations and incomes across the University. It contained clerical workers, administrators, research technicians, librarians, security personnel, and housekeeping staff. While respondents reported a high level of educational attainment,⁸ this was consistent with the workforce at the rest of the University. Moreover, the sample was also racially diverse: 78% were Caucasian; 11.5% were African American; 4.4% were of Asian descent; and 4.1% were Latina. According to University officials, the racial composition of the sample was comparable to the rest of the University.⁹

In addition, there was no obvious bias among the responses in terms of women's experiences with or responses to unwanted sexual attention in the workplace. Indeed, it is hard to determine the direction of such bias. Some have suggested that women who have had such experiences are more eager than others to respond to surveys; others argue that the experience is so traumatic for some women that they do not want to share it with researchers and decline to provide responses (Arvey & Cavanaugh 1995). The former type of bias is probably not present in this study. Twenty-six percent of the sample reported an experience with unwanted sexual attention over the previous two years. This percentage is comparable to but lower than other major studies of the incidence of sexual harassment (Fitzgerald et al. 1988; MSPB 1995). It is possible, however, that the study suffers from the latter form of bias and that some women were so traumatized by their experience or their use of the grievance procedures that they did not want to participate in the study. If such is the case, I have understated the problems associated with grievance procedures.

Because my focus is on women's understandings of the complaint process and their reasoning process in responding to sexual encounters at work, the data for this article came exclusively from the in-depth interviews and the open-ended responses in the questionnaire. (There were too few survey respondents [n=16] who actually told a supervisor about their experience with unwanted sexual attention to conduct meaningful statistical analyses.) In analyzing the transcripts and survey responses, I identified several themes with respect to women's evaluation of their options when they experienced unwanted sexual attention at work. After identifying these themes, I analyzed the transcripts for commonalities among the respondents in both the meanings they attached to the grievance procedures and their behaviors.

The interpretive methods I relied on and the small sample size, of course, made it difficult to make generalizations about women's experiences with and responses to unwanted sexual attention at work. Thus, I offer no conclusions about the incidence of sexual harassment at the University or about the effectiveness of the University's anti-harassment procedure. Rather, I focus on the way that organizational practices and routines shape the meaning women give their experiences and the way they evaluate their options. Thus, this article represents an important opportunity to build theory on the significance of organizational, institutional, and legal context for the development of legal consciousness.

The Legal Domain: Sexual Harassment Law and University Policies

In the United States, sexual harassment law is embedded in civil rights and employment discrimination jurisprudence.¹⁰ Courts have reasoned that sexual harassment is discrimination because members of one sex are targeted for harmful conduct due to their sex. The courts currently recognize two types of sexual harassment. The first, quid pro quo harassment, consists of demands for sexual favors in exchange for favorable job treatment (MacKinnon 1979). The broader category is hostile working environment harassment, where conduct must be sexual in nature and must be sufficiently “severe or pervasive” to have a negative effect on the employee's working conditions (Schultz 1998; Saguy 2000; MacKinnon 1979).¹¹ A hostile working environment has both an objective component—the environment must be such that “a reasonable person would find [it] hostile or abusive”—and a subjective component—“if the victim does not subjectively perceive the environment to be abusive, the conduct has not actually altered the conditions of the victim's employment” (Harris v. Forklift Systems Inc. 1993:21–2).¹²

The legal remedy for sexual harassment consists of obtaining damages from the harasser and the employer (MacKinnon 1979). The Supreme Court clarified the rules of employer liability in Faragher v. City of Boca Raton (1998) and in Burlington Industries, Inc. v. Ellerth (1998). Employers would be responsible for harassment when plaintiffs show that their refusal to submit to sexual advances from supervisors resulted in “tangible employment action” (Burlington Industries, Inc. v. Ellerth 1998:753). In other cases, where there is no tangible injury, employers can nevertheless establish an affirmative defense against sexual harassment claims by showing that they have a grievance procedure to handle employee complaints and that the “employee unreasonably failed to take advantage of any preventive or corrective opportunities provided by the employer or to avoid harm otherwise” (Burlington Industries, Inc. v. Ellerth 1998:765).

These cases ratified the importance of grievance procedures in sexual harassment cases (Edelman, Uggen, & Erlanger 1999). In articulating the affirmative defense in Burlington, the Supreme Court emphasized “Congress's intention to promote conciliation rather than litigation in the Title VII context” (Burlington Industries, Inc. v. Ellerth 1998:764). But in addition to requiring employers to adopt grievance procedures, the Court also effectively requires women to come forward and use them. The Court reasoned that this requirement furthers Title VII's deterrent purpose by “encouraging employees to report harassing conduct before it becomes severe or pervasive” (Burlington Industries, Inc. v. Ellerth 1998:764). In this view, complaints deter discrimination because they force employers to reevaluate their employment practices and make changes to eliminate discrimination. But by absolving employers of liability in the absence of a complaint, the Court has effectively reduced an employer's incentive to be proactive in ferreting out sexual harassment and has correspondingly made women responsible for whatever deterrence is available under Title VII.

By allowing that employers can lose the defense if the plaintiff's failure to complain was “reasonable,” the Court implied that grievance procedures alone may not provide complete protection for employees. The Court has left it to lower courts to identify a good reason for an employee's failure to complain, but the lower courts have yet to find one. A recent analysis of judicial opinions applying the affirmative defense found twenty-eight reported cases where the employees did not report harassment. The authors found that “[E]mployees who failed to report were deemed to have acted unreasonably” (Sherwyn, Heise, & Eigen 2001:1986). According to the authors, employers who design policies to avoid liability for sexual harassment will derive a mixed message from the courts: “Specifically, employers … should exercise just enough reasonable care to satisfy a court, but not enough to make it easy or comfortable for employees to complain of workplace harassment” (Sherwyn, Heise, & Eigen 2001:1267).

At the time this study was conducted, the University had a written sexual harassment policy (“Written Policy”) that conformed to these basic requirements. According to University officials, both human resources professionals and lawyers for the University participated in drafting the Written Policy, but human resources professionals were chiefly responsible for training supervisors how to use it. It applied to all members of the University community, including faculty, administrators, and support staff. In the preamble to its policy, the University offered an expansive view of the harms of sexual harassment. It stated that the prohibition was part of its commitment “to the maintenance of an environment free of discrimination and all forms of coercion that impede the academic freedom or diminish the dignity of any member of the University committee” (emphasis added). In addition to this broad endorsement of worker dignity, the University's definition of prohibited conduct expanded on the legal definition of sexual harassment:

Sexual advances, requests for sexual favors, and other verbal or physical conduct of a sexual nature constitute harassment when: (1) submission to such conduct is made or threatened to be made either explicitly or implicitly a term or condition of an individual's employment or education; (2) submission to or rejection of such conduct by an individual is used or threatened to be used as the basis for academic or employment decisions affecting that individual; or (3) such conduct has the purpose or effect of substantially interfering with an individual's academic or professional performance or creating an intimidating, hostile or offensive employment, educational or living environment.

These prohibited behaviors offered broader protection than the legal definition of sexual harassment. Silent about how severe or frequent the behaviors must be to be covered, the Written Policy provided no guidance about how to measure whether the conduct has “the purpose or effect of substantially interfering with the working environment.” But there was evidence that the University intended an expansive reading of the policy. For example, by prohibiting conduct whose purpose is to harm other employees, the Written Policy expanded on the legal right requiring that employees actually perceive a harm (Harris v. Forklift Systems Inc. 1993).

The Written Policy also outlined a flexible process for resolving complaints. First, employees could bypass their supervisors and take their complaints to one of many specified University officials, including the deans of the colleges, the human resources department, and the Women's Center. In fact, supervisors were not even specified as an option, thus perhaps suggesting that the policy was directed mostly at preventing and punishing harassment by supervisors. It also explicitly promised employees that they would not be retaliated against if they complained, and it directed officials receiving complaints to “immediately seek to resolve the matter by informal discussions with the persons involved.” The policy therefore anticipated that managers would handle problems before the more formal grievance procedure was ever invoked.

Along with its flexible approach to resolving complaints, the Written Policy provided modest due process protections for both the complainant and the accused. For example, the Written Policy provided that if the informal process did not produce a satisfactory outcome, employees could also file “formal complaints” within the University's formal employment grievance procedure. But the policy also provided due process protections for the accused. For example, an official could initiate an investigation of the employee's charges but only after finding “probable cause” to believe that the policy had been violated. The Written Policy offered very little guidance on the breadth and depth of these investigations except to require the complainant to support the claims with “clear and convincing evidence.” Thus, managers and supervisors had wide discretion under the Written Policy to take complaints very seriously or very lightly as circumstances dictated. If a complaint was found to be “substantiated,” the University could discipline and even discharge the harasser.

The University made significant efforts to publicize its sexual harassment policies. All employees were given a copy of the policy when they were hired. The University also circulated a copy to all employees every two years and kept the policy posted on its Web site. In addition, the University offered workshops and training programs to supervisors to provide them with guidance about how to handle complaints.

Crafting an anti-harassment policy can be a challenging task for employers. Prohibited behaviors can be difficult to define: Sexual conduct can be subject to competing interpretations; the comments and behaviors that bother some people may amuse others. Moreover, some incidents may be resolved with a simple admonition to a harasser, while others may require more complex investigatory and adjudicatory processes. The University resolved these challenges by adopting a policy that had a broad definition of prohibited conduct, gave supervisors flexibility to investigate and resolve complaints, and offered basic due process protections to both the accused and the accusers. Yet, as I show in the next section, this flexible policy was in practice a relatively rigid and adversarial system that undermined the legal goal of preventive dispute resolution over issues of unwanted sexual attention.

Enacting the Grievance Procedure: Management Practices and Women's Responses

All the women in the study confronted behaviors that arguably met the definition of prohibited conduct in the Written Policy. While the incidents employees described were not especially severe or pervasive, the Written Policy did not require that behavior rise to this level before a supervisor could take action. Moreover, some women complained of behaviors that were not explicitly sexual in nature, but these incidents occurred in working environments otherwise replete with sexual conduct and therefore violating the Written Policy. Indeed, the women in the study may not have characterized the conduct as “sexual harassment,” a legal label that they reserved for only the most intrusive, frequent, and harmful behaviors (Marshall 2003). Still, the Written Policy extended its protection beyond the formal definitions of sexual harassment. On paper, then, sympathetic supervisors had ample authority to intervene in most of the situations reported by women in this study.

In practice, however, supervisors rarely exercised their broad authority on behalf of complainants. According to the women who encountered sexual conduct at work, managers interpreted the policy in ways that discouraged complaints, adopting an adversarial posture that challenged rather than supported women seeking relief. Women learned of this management posture either through their own interactions with supervisors or by observing their colleagues who navigated the anti-harassment policy. These interactions with supervisors shaped women's choice of strategies in responding to harassers, with the most common response being “lumping”—or abandoning—their complaints. Thus, women did not even receive the assistance with dispute resolution that supervisors are thought to provide (Edelman, Erlanger, & Lande 1993).

As a result, the harassing conduct addressed through the grievance procedure was considerably more serious than the prohibited conduct as defined in the Written Policy. Women were most likely to complain—and management most likely to act—when the complainant had proof of the harassing behavior, when the harassers bothered more than one person, or when the conduct was truly outrageous. But incidents that did not meet these rather narrow standards were likely to go unaddressed. Moreover, the Written Policy and the practices enforcing it shaped women's legal consciousness, leading them to distinguish between their rights in the abstract and rights that they could enforce. Thus, the law of sexual harassment lost much of its power to shape women's working lives.

In the analysis that follows, I examine the management practices that implemented the University's Written Policy. I show that managers rarely exercised their broad mandate to protect employee rights, and instead often shielded the harassers—and the University—from women's complaints. I also show that women anticipated these management practices in fashioning their responses to unwanted sexual attention. These interactions reshaped—and narrowed—the meaning of sexual harassment for working women.

Management Practice

As gatekeepers to the grievance procedure, managers often protected the University's interests, not by transforming unwanted sexual attention into “personality conflicts” (Edelman, Erlanger, & Lande 1993), but by deflecting employee grievances. Managers at the University frequently engaged in three sets of practices that discouraged women from complaining. First, upon hearing complaints, supervisors took sides. Rather than assuming the role of neutral problem-solver, supervisors often placed their organizational power behind the harassers, effectively becoming the harassers' representative in the grievance process (Kihnley 2000). When supervisors did take complaints seriously, it was often because the women had more power—either formal or informal—than the harassers. Second, supervisors read into the Written Policy nonexistent requirements that manufactured obstacles to women's pursuit of complaints. Finally, supervisors offered restrictive, legalistic interpretations that narrowly construed the Written Policy's protection and, in effect, dismissed employee complaints because the conduct did not violate the policy. Thus, in the eyes of the beneficiaries of the policies, supervisors often deployed their considerable organizational power to protect harassers and the University rather than to enforce complainants' rights. These practices created an adversarial process, inconsistent with the Written Policy's preference for informal solutions.

Taking Sides

According to the Written Policy, supervisors were supposed to receive and resolve complaints. To fulfill this responsibility, they should have tried to remain neutral. Yet several women reported that when they described examples of unwanted sexual attention, their supervisors immediately took the harassers' side by making excuses and condoning their behavior. In some cases, the supervisors taking the complaints were closer in organizational status to the harassers than to the complainants. These closer organizational ties seemed to make the supervisors more sympathetic to the harasser. For example, one survey respondent reported that a harasser was a “total jerk who can't refrain from insulting or bullying anyone, period,” yet his fellow supervisors, whom the respondent characterized as “male chauvinists,” trivialized his conduct:

His actions are constantly excused, and you are told simply, that “you are too sensitive.”“That's the way he is” is also a popular comment, although it's obviously not okay that I am a certain “way” as well—bothered by what he says.

After a series of disagreements with a faculty member for whom she did clerical work, Emma passed him and several of his colleagues in the hall. The faculty member “asked if I minded if he would pinch my rear end. What could I say?” After thinking about this incident in the context of other problems she was having with the faculty member, she complained to the chair of the department, who told her, “‘He's just going through a stage. You need to be very—you need to handle him with kid gloves, and you need to be very tolerant.’ This was, of course, not what I wanted to hear.”

Women reported that supervisors were more sympathetic to complaints about contrapower harassment, where the harassers had inferior organizational status to both the targets and the supervisors. Several survey respondents reported incidents with student athletes, but their prompt intervention—with the help of human resources personnel—solved the problems. One reported, “I specifically spoke to the students about the comments and how I felt about it. It was a teachable moment for them. They seemed remorseful and apologized. They seemed to understand the error of the display. It has not happened again. I think [the University] does a good job about informing its employees about the issue. …” Thus, in these cases, the relative organizational status of the supervisors, the complainants, and the harassers influenced the supervisors' gatekeeping, which is inconsistent with the policy's broad purpose of protecting employee rights.

Personal relationships may likewise affect the way that supervisors handle complaints. Women reported that their concerns were treated with less sympathy and seriousness when supervisors were friends with harassers. In such cases, women reported that supervisors did not treat the problem with the seriousness it deserved. For example, a project leader told one survey respondent that she would not work on the project unless she slept with him: “I was sexually harassed by a co-worker and my complaints to my female supervisor were not only laughed at but I was removed from the project for complaining … He was a friend of my married, female supervisor.” When Joanne went to the human resources director to complain about the chief financial officer at a previous job who was continually making crude jokes and comments, she found the two of them together sharing raunchy jokes. Their friendliness with each other made her suspect that her complaints were not being taken seriously.

On the other hand, complainants could sometimes use their friendships with supervisors to prompt a proactive management response. For example, Dallas was working as a clerical staff member in a doctor's office at the University. Although she often exchanged off-color jokes with her co-workers, she drew the line at physical contact. One doctor she worked for crossed the line by telling a joke where the punch line included pinching her nipple: “That was it. Now, at this point, now you're touching me. So then, I had a very good friend who was an administrator. I told her—I was livid. And I told her what happened. She just went at him. She said: ‘I want you to get him now.’ I heard so much screaming in there.” In this case, Dallas was able to use her friendship ties with a manager to resolve a workplace dispute informally, but ordinary working women rarely enjoy this informal access to organizational power.

These accounts suggest that the supervisors designated as complaint handlers can appear to be biased before they ever hear a complaint, thus compromising their ability to conduct an investigation or to solve problems. Indeed, women may sometimes derive benefits from supervisory favoritism, but the grievance procedure's capacity for protecting employee rights nevertheless depends on the vagaries of close organizational or personal ties between those employees and the complaint handlers.

Manufacturing Obstacles

The Written Policy gave complaint handlers a great deal of discretion about how to proceed upon hearing about an incident of sexual harassment. In practice, managers interpreted the grievance procedure to exempt certain categories of harassment and harassers and to introduce new steps into the complaint process. These interpretations created obstacles that were inconsistent with the Written Policy's broad mandate to protect employee rights and dignity.

For example, some supervisors created exemptions from the Written Policy for powerful organizational actors, such as tenured faculty members. When she got nowhere with the chair of the department where she worked, Emma decided to approach the human resources department. She remembered:

But I was unhappy enough about it that I went and I wrote up the circumstance … and had taken it to human resources, and complained about it. And they were very, uh, not tremendously sympathetic about it. They said, “Really, if the person is a faculty member, you know, there's nothing we can do about it.” Only in very egregious circumstances, when there's very obvious, overt abuse, if you will, or harassment, would they do something. So they basically said, you know, “Tell him to lay off, and hope that he does.”

The human resources representative here did not just abandon responsibility for anything other than “obvious, overt abuse,” but also claimed that the conduct of faculty was beyond the reach of any disciplinary process, although the Written Policy contained no such exemption.

On occasion, same-sex, contrapower harassment was also exempt, particularly when the conduct underlying the incident was ambiguous. Rose's problem involved a female receptionist who engaged in frequent sexually explicit conversations with the male coaches in the office where she worked. Rose's supervisors were reluctant to intervene because the harasser was a woman, because she was in the lowest-level position in the office, and because her behavior consisted only of conversations and not more intrusive conduct. Rose described the results of her “formal conversations” with her supervisors:

There was kind of this, “We hear what you're saying, and we'll do what we can do, but you know, our hands are kind of tied without there being a major incident.” Without her, like, grabbing me and throwing me against the wall, or. … Who knows what that could have been? You know, without finding her and a coach, naked, on top of the coffee table—who knows what would have been legitimate?

Her supervisors also argued that a violation of the Written Policy was a prerequisite to taking any steps to resolve the problem. Thus, they did not offer Rose their assistance in resolving the dispute informally, as suggested in the Written Policy.

In addition to creating exemptions, supervisors could use their extensive discretion by adding steps that made complaints a less-attractive option for aggrieved employees. For example, like most such procedures, the Written Policy did not require employees to confront their harassers. However, as an office manager, Megan imposed this requirement on a group of women who wanted to complain about a co-worker whose behavior was bothering them. Megan herself had been a target of his harassing conduct; she admitted that he had been “very sexually inappropriate throughout the entire year that he was here.” Yet when a group of other women wanted to file formal charges, she sent them away:

I put the kibosh on that because I said, “Have these people addressed him personally on these issues. Have they approached him?” Well, some did but mostly no. And I thought I'm going to interject here and say that's not fair. This person—much as I do not care for him—we owe him … the opportunity to change his behavior. … Before you start pushing paper through those channels, let him know he's a jerk, and [you] don't like what [he's] doing. Otherwise it's not fair. …

Although Megan admitted that the women may not have felt comfortable coming forward, she dismissed their concerns because the harasser had no formal authority or influence over their jobs. Thus, Megan's authority to administer the grievance procedure also gave her the authority to interpret its provisions in a manner in keeping with her conceptions of fairness, which created new obstacles for the women who sought to complain about the harasser.

Like many sexual harassment procedures, the breadth of the Written Policy required that complaint handlers exercise extensive discretion to resolve problems. In addition to their evaluations of employees' complaints, their interpretations of the Written Policy also reflected their own conceptions of fairness, their understanding of power in the organization, or their desire to protect their employer from liability. When they implemented these interpretations, they also created meaningful obstacles to women's access to dispute resolution.

Narrowing the Written Policy

Many supervisors narrowly interpreted the Written Policy's broad promise of protection for worker dignity and equality. These supervisors effectively dismissed women's complaints on the grounds that incidents were not sufficiently serious or offensive to constitute a violation of the Written Policy. Without a violation, they argued, they were powerless to intervene. For example, Siena had a co-worker who lingered near her desk, making subtly sexual comments. She said,

He started making comments about what I was wearing. … He said something about [my] perfume, and I explained to him that I didn't wear perfume. … It wasn't what he said; it was the way it was said. It was like a double entendre. It could be taken one way or another. He always did it while we were alone, and there was never anything that was blatantly physical.

After listening to Siena's account of her co-worker's behavior, her supervisor told her that the comments were too ambiguous to merit a complaint. The supervisor's reaction persuaded Siena that there was no point in pursuing a formal complaint. Siena observed,

After I had talked to my boss, I felt really let down and betrayed. And I didn't feel like I had any other place to go. I thought I'd gone to somebody that would help me, and it was somebody who had known me for three years. And I thought if she acted this way, that I wouldn't have a leg to stand on anywhere else. … I did the first step and got nowhere, so I stopped because I figured I needed her backing in order to go any further.

Siena's complaint about her co-worker fell within the terms of the Written Policy. Indeed, while his behavior may not have been severe or intrusive, his comments could have been interpreted as “verbal conduct of a sexual nature” under a proactive reading of the Written Policy. In any event, the supervisor had the authority to tell the co-worker his attention was making Siena uncomfortable and to ask him to stop, authority she did not use. Moreover, the supervisor's skepticism about the seriousness of the incidents strongly suggested to Siena that the conduct was subject to multiple interpretations and therefore would not be construed as sexual harassment by other decision makers further up the chain of command in the complaint process.

The human resources department also offered preliminary assessments of complainants' cases and concluded that the behaviors were too trivial to violate the Written Policy. For example, a survey respondent's supervisor frequently made crude sexual remarks—remarks that bothered her a great deal. Yet she was discouraged from pursuing a complaint: “I discussed it with the EEOC rep—she didn't feel the case was strong enough to bring a formal complaint.” For all these women, the conduct in question consisted of “sexual advances, requests for sexual favors, and other verbal … conduct of a sexual nature” and “created … an offensive employment environment” that could be construed as violating the terms of the Written Policy. Yet supervisors and human resources managers refused to pursue their complaints because in their view, the incidents did not amount to a violation.

Edelman, Erlanger, and Lande (1993) found that supervisors transformed problems with race and sex discrimination into personality conflicts but then tried to resolve these conflicts to preserve workplace peace. Adopting the perspective of complainants, however, reveals a different picture of this transformation and its consequences. In this view, the managers make grievance procedures an adversarial process that actively discourages complaints. Supervisors rehearse the harassers' defense by questioning the complainants' credibility or rationalizing the harassers' behavior. They impose burdens that make complaints seem threatening or causing conflict, and they would probably conclude that a “personality conflict” did not violate the anti-harassment policy. In the end, when supervisors decline to deploy their authority, formal or informal, to help employees having problems with unwanted sexual attention, they do nothing to resolve employee problems and also strongly suggest that organizational power protects harassers rather than the targets.

Women's Strategies Against Unwanted Sexual Attention

When confronting these adversarial management practices—either themselves or vicariously through the experiences of their co-workers—women developed strategies for dealing with unwanted sexual attention. First, women prepared for complaints as though they were preparing for litigation, gathering evidence and witnesses to support their claims. Second, women engaged in self-help by confronting harassers directly, sometimes in the shadow of the Written Policy but sometimes acting on their own authority. Finally, some women rejected the Written Policy and procedures and lumped their complaints by ignoring the incidents.¹³ Through these strategies, women enacted a double consciousness about the law of sexual harassment, where rights created expectations of better treatment but where those rights were not easily enforced through the grievance procedure.

“Litigating” Complaints

When managers narrowed the meaning of the Written Policy and tried to block access to the grievance procedure, one group of women responded by carefully preparing their cases to establish that the events did, in fact, occur, and that they were serious enough to warrant attention. Thus, women anticipated their supervisors' legalistic interpretation of the Written Policy, invoking it only when they felt confident they could meet the standards it imposed. Yet even with this level of preparation, women were often disappointed with the results, sometimes finding themselves the targets of retaliation.

Women's use of the grievance procedure depended first of all on the seriousness of the intrusion that they encountered. Of the women interviewed for the study who complained to a third party about their sexual encounter at work, all were convinced that the experience amounted to sexual harassment. Consisting mostly of extremely explicit sexual overtures or physical contact, these incidents probably met the behavioral tests for sexual harassment. Once that threshold was crossed, however, not every woman braved the grievance procedure.

These women only approached the grievance procedure when they were able to gather evidence to substantiate their claims in the face of aggressive management gatekeeping. One type of evidence was the support of co-workers to corroborate their stories about harassing incidents. When everyone in a working environment objects to some form of conduct, supervisors and managers take the situation much more seriously. But co-workers also provide social support through the complaint process. They act as sounding boards for the complainant, offering sympathy and advice about how to pursue the complaint. Thus, when an individual had the backing of her co-workers, she was more likely to be able to navigate the adversarial nature of the process.

For example, Matilda was a departmental assistant where the chair had a foul temper and offensive personal habits: “He'd come to talk to you and he'd have his hands down his pants, and he'd be playing with himself in the front, or he'd have it down the back and be scratching himself.” Although Matilda had complained about his various behaviors, her complaints had not been taken seriously by the dean or by human resources. Then one day, in a rage because a document had not been photocopied, he screamed at the staff using racist and sexist epithets, directed particularly at a Chinese woman:

And then he knocked everything off of her desk, and he went over to the other lady, the program assistant, and threw his letter at her, and then knocked everything off the desk, yelling and screaming, “You're all stupid.” And I came out of my office, and I was standing in the main office, and I was trying to get him, either to come into my office, or go out in the hall, or go in his office, or anything—to just—to break this up. And I said, “Could we talk about this?” At which case, he turned—he turned beet red. He came up to me and took his fist, and [gesture] came that close. One inch. From hitting me.

After this show of temper, all four staff members left the office together and went to the human resources department. With all four women standing in the office seeking transfers out of the department, a human resources representative contacted an assistant dean at the college who was sympathetic: “They sent one of the assistant deans back to the department, with us, to confront the [chair]. And they ended up fighting. And he was in there—they were yelling and screaming an hour or more.” Although this incident was not sexual, it nevertheless constituted a serious infringement on the employee's dignity that violated the express terms of the Written Policy. Moreover, the dean's office already had notice of the chair's more sexual behaviors. The unanimity and urgency of the complaint provided University officials with a compelling justification for intervention.

Women also preserved written incriminating evidence to bolster their credibility. For example, Jane worked with a professor from another unit in the University. On an almost daily basis, he sent her e-mails that critiqued her marriage, complained about his sex life, and asked her out on dates. When asked whether she had thought about filing a complaint, she said: “Yeah, I have, but … I kept the messages, and I, you know, made sure I had a backlog of things just in case I had to, but I didn't think that I would do it unless I really had to.” She showed the e-mails to her supervisor, who was sympathetic but advised her not to mention the incident to anyone else. Although she agreed, she kept hard copies of the messages in case she finally decided to file a complaint.

When there was no incriminating evidence, women tried to create it by documenting the harassing incidents. Notably, every woman I interviewed—even the ones who did not complain—had committed the event to writing. For example, Siena told a friend working in the office that her co-worker's lingering near her desk making suggestive comments was making her feel uncomfortable. Her friend suggested that she record each event. She said, “I'm pretty thorough at documentation. I wrote down word for word what happened and what I said in response to it as well.” She later produced this document when she went to her supervisor to complain about her co-worker's behavior.

Several women reported that their complaints seemed to attract little attention from their supervisors; nothing ever seemed to happen to the harassers. Often, the conduct continued just as it had before the complaint. When Matilda and her colleagues complained to human resources and to the dean's office, they got an initial response: An associate dean came to the department and yelled at the chair about his threat of violence against Matilda. But in the end, nothing happened to the chair, and he returned to his obnoxious personal habits and verbal abuse. She said,

That assistant dean, who I think was sympathetic to our problem, he left the University shortly after that. Nothing really was done. … What made us go back was them coming after us, and us believing that there would be some sort of a solution to the problem. And then there never was. There was never any mention of it after that.

Similarly, Rose's complaints about the receptionist's sexual conversations with the coaches never resulted in any supervisory response. She said, “Basically, I don't think anything happened. If there was a conversation with the woman, it was certainly not handled in such a way that had any impact on her whatsoever. Nothing happened.”

Worse than witnessing inaction, women also found themselves to be targets of retaliatory action. This retaliation took a number of different forms. Some women reported receiving negative performance evaluations when they complained about their experiences. One survey respondent whose supervisors and co-workers engaged in daily efforts to engage her in conversation about her sex life reported, “The men aware of the situation intimidated me every time I brought it up. I was led to believe there would be repercussions if I continued to bring it up—and there were—I received two consecutive performance evaluations indicating that I had an inability to get along with co-workers.”

Another form of retaliation was removing the complainant from job duties or projects. In another of Dallas's experiences with sexual harassment, she was a senior departmental assistant to a faculty member who was directing a profitable enterprise for the University. This official was frequently verbally abusive to Dallas and the rest of his staff. In addition, he frequently made sexual jokes and observations about his employees' physical appearances and sex lives. Dallas withstood much of this attention by playing along or fighting back, but one day, he screamed at her using profanity when she asked him for some information. In the aftermath of this incident, Dallas began to complain about his general conduct: She complained to the office manager, human resources, and the dean of the college where the project was being conducted. None of these complaints had any effect. In fact, the only thing that happened was that Dallas was demoted to a lower clerical position and moved to a location out of the director's presence. She said, “He moved me from my office to the secretarial pool. He didn't want me close to him. He then gave my job to this other woman there who was prettier than me.”

Even when women did not face direct, negative job consequences, several found that their daily work lives suffered when they complained about harassment. One survey respondent had a male boss who complained about the large amount of sexual materials displayed in the workplace. She reported, “When my (male) boss reported it to our supervisor, I was asked if I was offended. Despite that, action was taken by this supervisor. My boss and I have received a lot of comments (only half-joking) about our sensitivity to sexual materials” (emphasis in original). Another survey respondent reported, “The comments about me and behavior towards me were symptoms of a very large problem on the part of my ex-supervisor. When I involved HR, the problem worsened because I had called attention to my supervisor's bad behavior. It was two months after involving HR that I quit.”

The gap between the theory and practice in the implementation of the Written Policy placed the grievance procedure in an unexpectedly adversarial light. Women sought advocates and representatives to support their allegations. They also developed evidentiary records to repel challenges to their credibility and the validity of their complaints. Still, women found that complaints did little to resolve the problem and often made things worse, particularly when the harasser was a supervisor. Thus, even with an informal complaint process, managers and employees reproduce an adversarial system.

Self-Help: Confronting the Harasser

Some women in the study engaged in self-help by directly confronting the harassers rather than getting a third party involved. But women who chose this strategy were enacting the Written Policy in very different ways. In some cases, women relied on the Written Policy to challenge harassers' conduct; others engaged in self-help only when the policy had failed to provide any relief. For still others, the Written Policy was utterly irrelevant in their confrontations with harassers.

Several women reported relying on the Written Policy to “educate” their harassers about the limits of appropriate behavior in the workplace. Supervisors sometimes re-enforced these efforts through informal discussions with the harassers. One survey respondent reported, “There were a few faculty members who made inappropriate, sexually harassing comments. One left. The other has been educated—partly by me, partly by a firmer dean. This person thought jokes and cartoons of a sexual nature were appropriate for work but now knows they are not and has been ‘rehabilitated.’” Thus, both employees and University supervisors invoked the Written Policy as a basis to encourage harassers to adapt to their working environments even without pursuing formal complaints.

Although the Written Policy empowered some women to engage in self-help, others confronted their harassers only as a last resort when the grievance procedure had failed to solve their problems. In these cases, employees told harassers that the behavior had to stop, even without the support of managers or supervisors. For example, Rose concluded that the receptionist's sexual conversations were sexual harassment after reading the University's sexual harassment policy. But none of the designated complaint handlers was willing to intervene because the accused harasser was a woman and a co-worker. Finally, Rose decided to take matters into her own hands:

So basically I finally said, “Screw the system.” Because I went through the system. And I said, “I'm going to confront her on my own.” And all four of us, all four of the secretary staff got together, and I just said it right to her face. I said, “I'm completely uncomfortable with the way you behave around the office. I think you're inappropriate in your behavior with the coaches. I don't need to be told that I'm a prude, or that I'm being irrational, but I've talked to some of the staff about it; they see it as well. If nothing else, you need to do your job. And you're not doing your job. I'm doing your job for you. I've had enough of it, and I'm not going to do it anymore.”

Beyond providing her a label for the conduct, the Written Policy did little to help Rose in her efforts to solve the underlying problem. In the end, she did not invoke her rights; rather, she rested her demands that the behavior stop on the personal comfort of her co-workers and workplace efficiency. Indeed, this confrontation was successful; the receptionist drastically curtailed her conversations with the coaches. But this success was only indirectly related to the Written Policy.

Finally, at least one woman in the study engaged in self-help where the Written Policy seemed utterly irrelevant. A survey respondent worked in close quarters with a male colleague who physically assaulted her several times. She asked him to stop and threatened to tell a supervisor, but it had no effect on his behavior. She knew about the Written Policy and thought about complaining but never did because she was worried that she might be retaliated against and that the complaint would make things worse. Finally, she wrote, after he “grabbed me and tried to force himself upon me, I promptly slugged him and almost threw him on the conveyer belt in the cage wash area. Needless to say, he apologized. …” She admitted that violence was “not a good idea, but it seems the only action that works.” In this case, the employee rejected the very premise of the Written Policy—dispute resolution that preserves peace in the workplace.

Women may be acting in the shadow of the Written Policy when they engage in self-help. In those cases, the Written Policy empowers individuals to enforce the policy's limits on tolerable conduct with nothing more than the threat of sanction implied by invoking the policy. Yet self-help can also reflect the failure of the Written Policy to place meaningful constraints on harassers or to protect employees, requiring them to act alone.

Lumping It: Rejecting the Complaint Process

Most women in the study did not complain to supervisors about their experiences with unwanted sexual attention at work. This decision was sometimes based on a judgment that the conduct was not sufficiently serious to merit a complaint. Other women, however, experienced incidents that were disruptive and distracting to their work performance and would have liked some assistance in resolving the problem. When considering whether to make a complaint, however, they anticipated their supervisors' response. Many believed that this response would be adversarial and hostile to their complaints; some believed that their supervisors would be ineffectual in solving the problem, while others were concerned about retaliation, so they decided not to complain. Thus, they acted as their own gatekeepers, incorporating the adversarial response into their own evaluation of the situation.

When the harassing conduct was neither severe nor harmful, women were not inclined to complain about the behavior. Infrequent jokes, occasional physical contact, and sexist comments were mostly shrugged off or ignored. For example, Nora concluded that occasional sexual jokes told within her hearing by some faculty members were not sexual harassment; the faculty members did not make a habit of telling such jokes, nor was Nora a target of such attention on a regular basis. Another employee, Ann, was asked out several times by a co-worker, but he stopped bothering her when she explained that she was not interested in a personal relationship. One of her faculty supervisors even hugged her once. Because these incidents did not constitute part of their everyday working environments, had no effect on their job performance, and were easily resolved informally, neither Ann nor Nora considered the behaviors serious enough to warrant a complaint.

Other women in the study, however, might have invoked the grievance procedure but were apprehensive about the reception they might get if they did. These women expected supervisors to tell them to handle the situation themselves or to accuse them of leading on their harassers. For example, Erna worked in an office with a middle-aged man whose relentless sexual comments and innuendo were a constant annoyance in an office made up mostly of women. Erna described him as

always making sexual innuendos; and with every word you said, he found something to make an innuendo about it. … To me, it is almost worse than having somebody come up and grab you. Because it's a constant barrage of innuendo. And it just gets really annoying. And then you don't know how to handle it. And if you would say, “Look, you're always making this innuendo,” then he would start to say, “Are you one of those dykes too?” or whatever the case was. That sort of thing. He just didn't understand that that was not appropriate.

Although her contact with the harasser was a daily irritant, Erna did not complain about his conduct to her supervisor. She anticipated that her supervisor would suggest that she was at fault for failing to handle the harasser in an appropriate manner. She said, “I think when you go to tell your supervisor, it always comes back: ‘Well, what did you do?’ You know, ‘Just tell him no …,’‘Well, I haven't heard this from anybody else,’ type of thing.”

Similarly, Lily was concerned that her boss would use her complaint about a harasser to find fault with her conduct. She was employed in a unit that worked with personnel across the University, including the medical school. One of the doctors she worked with invited her out to lunch to discuss work-related issues, but after their lunch together, he began calling her to ask her out on dates, and eventually he suggested that they have sex. Although she admitted that her fears were not necessarily “realistic,” she nevertheless identified a specific list of the negative conclusions her boss might reach about a possible complaint: “I think they're fears … my boss will believe him because he'll never defend me, so my value to my boss will decrease. … Or maybe my boss would think that I was leading him on, or whatever, and that it was inappropriate to go for a professional lunch with him—you never do that.” Thus, women expected that if they made complaints, supervisors would adopt an adversarial posture and interrogate the complainant's behavior. Such challenges to their credibility were better avoided.

For other women, their reluctance to complain was based on their skepticism that their supervisors could solve problems. Like other women I interviewed, Erna was pessimistic about the University's ability to resolve any kind of personnel problem, let alone an issue as sensitive as an allegation of sexual harassment. She observed,

The procedures here don't work … If I had wanted to complain about [the harasser] I really had two choices. I could have gone to my boss. Or I could have gone to the office manager. And my boss … might have told me to go to the office manager, or he might have said to me, “Why don't you just leave it?” And that was the whole chain of this process that was going on [with a different personnel problem], and people trying to take it up the chain of command, and it didn't work. In our particular case, our office manager just doesn't deal with this stuff. And I don't know if she doesn't deal with it, or if she just gets no reinforcement from her point of view. … And then you have another choice: you could go to human resources. And that seems to have been a very negative thing because human resources … has never come through with any other problems that people have had. So people don't do that.

In the course of her employment, Erna had watched the supervisors and human resources professionals as they tried to settle employee conflict with very little success. Their ineffectual responses to employee problems undermined Erna's confidence that they would ever be able to get the harasser to stop his behavior.

Finally, although the University's sexual harassment policy specially promised to protect employees from retaliation, women still feared the effects of a complaint. While they acknowledged that their jobs might not have been in danger, they feared that by coming forward, their more powerful harassers would make subtle but consequential changes in the working environment. At a previous job, Joanne had noticed such changes when she complained about a high-ranking company official. When she entered a room, he would either stop talking or loudly comment on her humorlessness to colleagues. This experience made her reluctant to complain at the University when her female supervisor grabbed her breast: “If I didn't lose my job, it would make for a more hostile environment than it was working with her, being how she is.”

These fears about the grievance procedures were sometimes magnified by co-workers and colleagues. Megan reported that early in her career, a manager high above her in the corporate chain cornered her in a conference room, making suggestive remarks about her appearance. After that, he made inappropriate comments every time he saw her. She turned to her fellow employees for advice:

I remember talking to a few other women in the company that I trusted and told them what happened. And they had told me, “You know what? It happened to me a million times from that exact same individual.” And it was kind of like “You want to make it in this company? You learn to deal with it, and you suck it up because you're going to get blackballed if you fight this.”

Thus, while they might offer emotional support to women planning to lodge complaints, social networks can also provide powerful reminders of the costs of complaint.

Because they feared antagonistic or skeptical treatment from supervisors, many women were discouraged from making complaints, even in circumstances that made complaints most likely—persistent and intrusive conduct by co-workers. But when women declined to participate in the routines and practices of the grievance procedure, some made that choice because they perceived the procedure to be ineffectual at best and, at worst, hostile to their needs. Their apprehension about using the grievance procedure rendered it irrelevant when they encountered sexual harassment, thus limiting its usefulness in protecting their rights at work (Quinn 2000; Bumiller 1988).

Social Practice Shaping Legal Consciousness

Management and employees enacted a very different sexual harassment procedure than the one enshrined in the Written Policy. While the Written Policy was designed to offer employees expansive protection from the indignities of harassing behaviors, the management practices enforcing it significantly shrank that protection. And as a result, women's understanding of sexual harassment reflected the adversarial nature of the complaint process, shrinking to include only those behaviors they could prove. For these women, supervisors did not need to dismiss complaints for failing to violate the policy. Women's narrow interpretation of sexual harassment accomplished this task by censoring complaints before they ever formed.

Women incorporated their skepticism of the grievance procedure into their definitions of sexual harassment itself. To several women in the study, an incident could not be characterized as sexual harassment without evidence that it had occurred and that it was a serious affront to the woman's working life. If the incident could not be documented or if it did not meet some external standard of offensiveness, then women argued that it was not sexual harassment and would not meet the threshold of behavior required to file a complaint. Lily, for example, distinguished between blatant forms of sexual harassment, which included unwanted physical contact and explicitly derogatory remarks, and

the subtle kind of thing, where you're not really sure. … Not being really sure, not having anything to grab onto—look, here's the definitive proof. I mean, all the doubting—it's gray; it's too gray. So I think that must happen on many occasions that you suspect motives, but you have no concrete evidence to support it, so then you doubt yourself.

Even though her harasser repeatedly asked her to go to bed with him, Lily was not sure his behavior constituted harassment because she could not prove to a third party that he had made these advances or that they were unwelcome.

These evidentiary concerns created ambiguity in women's understandings of the standards for sexual harassment. A survey respondent was repeatedly propositioned by some employees—both supervisors and co-workers—but ignored and avoided her harassers. She did not complain largely out of confusion about what constituted harassment—a confusion that was tied to her anxiety about proving any allegations she might make: “Mainly, it is hard to define the boundaries of what is considered harassment and what is not. E.g., is staring at my chest repeatedly harassment? If so, how can I prove he was doing it? Will I be believed? What if he is friends with the management.” To this woman, the confusing rules surrounding sexual harassment were aggravated by the power imbalances reproduced by a grievance procedure administered by managers.

Thus, the women in the study also re-framed the meaning of the Written Policy itself. Rather than associating the policy with the protection of their rights, some women offered their own institutionalist critique of the grievance procedure and argued that the Written Policy was ineffective because the University's main priority was to protect itself from liability. When Rose complained about the receptionist's sexually explicit conversations with the coaches, she found herself bouncing back and forth among several different lines of authority—an office administrator who was not in Rose's office, the human resources department, and the supervisors in the office itself. Yet no one took responsibility for the working conditions:

If you get involved with Employee Relations, invariably everything is colored by the question of risk management. Everything is colored by the question of the University's liability, and will we ever come to suit on this. And so all of the information that you want as a manager, in my experience at least, has been CYA—cover your ass—make sure you put it in writing, and not nearly enough of, and a much less clear emphasis on dealing with the practicality of the situation. … I think if you get Employee Relations involved, it just seems like you instantaneously feel like you're in a court of law. You feel like you have to get all your ducks in a row, to be able to put together the prosecution for this case.

Having been both a clerical worker and a manager at the University gave Rose insight into the management concerns behind the policies and procedures. She detected that such policies placed managers in the position of having to act as lawyers for the University, charged with “putting together the prosecution for this case.”

In addition, many women in the study believed that the Written Policy was actually a mechanism that protected the most powerful employees in the organization. When a harasser was a powerful person, women believed the University would take whatever action necessary to safeguard the harasser's position. For example, when Siena was considering pursuing a complaint, she thought about her observations of the grievance procedure in a previous position. These observations supported her view that employee complaints would be resolved in favor of the University. She said, “I would say that I wouldn't have gotten very far, and that I think the University would have taken the University's side and not the employee's side.” Similarly, Matilda argued that the chair of her department suffered no real consequences because the University would protect the faculty over the staff. She said,

I think they normally don't stand up for the staff anyhow. If a faculty member wants something, they usually get it. So are you going to go to them with this problem? I don't think so. I don't think they're impartial. I don't at all. I think there's too much interest in “Well, he makes a lot of money for the University, we have to appease him.” I see it every day.

In these accounts, sexual harassment policies offered employees little protection from the excesses of powerful harassers. When Dallas could no longer tolerate her abusive supervisor, she asked for a transfer. In all her exit interviews—with the human resources department and other administrators—she made it clear that she was making this request because of her supervisor's behavior, but she never filed a formal complaint because she thought it would be futile:

Staff are peons. [The University] is not going to get much money out of us. He is generating money for the University so whatever he does [they] are going to overlook. Besides … he was created out of a plan that the Provost had created and that the President loved, and so they're not going to lose money and get embarrassed by someone saying he's doing this and this, and he makes work intolerable, and he “f___'s” everybody, and he is prejudiced against everybody except himself. … He's bringing in revenue, and the University is not going to take the word of one female who is going to make him trouble.

Women felt that complaints in these situations were futile. When one of the assistant deans suggested that she might be preparing a case against the harasser, Dallas was skeptical that anything would ever come of it. Dallas reported, “She said, ‘If I ever prepared something, if something were to happen, would you be willing to testify?’ I said, ‘Yeah, I'd be willing to testify.’ But nothing's going to happen. I just said, ‘You're up against the system, and it's very nice that you're willing to in case something happens, but it's not going to happen.’”

The Written Policy articulated a set of legal schema that broadly defined sexual harassment and that prescribed a mechanism ostensibly protecting employee rights. Yet in the shadow of the practices implementing the Written Policy, women's definitions of sexual harassment narrowed considerably. And to these women, the Written Policy became an instrument through which the University protected the powerful rather than a process for enforcing their rights. Thus, the employees and supervisors enacted a Written Policy whose meaning in practice was very different than its symbolic purpose.

Discussion

This study has analyzed sexual harassment policies from an important but often missing perspective—that of the employees, the intended beneficiaries of such procedures. By adopting the Written Policy, the University fulfilled its legal obligations to protect its employees from sexual harassment. It broadly defined sexual harassment to include behaviors that created a hostile working environment and undermined a worker's dignity. It also authorized supervisors to seek out informal solutions to employee problems. Yet the policy provided far less protection in practice than it did on paper.

In keeping with institutional theories of organizations, managers often interpreted the Written Policy in a way that protected University interests. But rather than offering dispute resolution to aggrieved employees, University supervisors frequently deflected employee complaints about sexual harassment. By dismissing complaints as being groundless or by making excuses for the harassers, these supervisors turned the grievance procedure into an adversarial process that seemed inimical to informal problem-solving promised by the Written Policy. In turn, these management practices shaped women's responses to their own experiences with unwanted sexual attention. Women who pursued complaints—even at the most informal levels—prepared for the process as though they were preparing for litigation. They arranged for witnesses; they preserved incriminating evidence; they kept logs of events to document the incidents, all to prove to a supervisor or other third party that the events actually happened. But most often, women just lumped their complaints and tried to handle the situation themselves. Thus, managers and employees together enacted a much narrower sexual harassment policy than the one on paper— one where only the most intrusive forms of sexual harassment got addressed.

But these problems in implementation do not mean that sexual harassment policies should be abandoned. Rather, employers should designate different complaint handlers to remove some of the risks of bias. One possibility is the appointment of an ombudsperson outside of the traditional lines of organizational hierarchy who reports directly to the chief executive officer. This ombudsperson would be responsible for outreach, providing education and training as well as dispute resolution to employees. In some workplaces, the ombudsperson acts as an advocate for women with allegations of sexual harassment, so that employees feel more comfortable coming forward with their complaints.

Another possibility is to decentralize dispute resolution resources in the workplace. For example, instead of designating a few officials as complaint handlers, employers such as DuPont train dozens of ordinary employees on what to do when someone tells them about an experience with sexual harassment (Flynn 1997). Spread widely through all levels of management in the organization, these employees interact with almost everyone in the company on a daily basis and are plugged into the social networks where the common knowledge about harassers circulates. Their training emphasizes informal solutions to sexual harassment, including low-key conversations with harassers asking them to stop their behaviors. Strategies such as these mobilize already existing gossip networks that disseminate valuable information. Moreover, management does not assume that every problem will end in litigation, and is therefore willing to seek out commonsense solutions. (Flynn 1997).

This study also demonstrates the theoretical value of conducting constitutive studies of law situated in specific problems. Such studies can reveal the mechanisms through which law shapes meaning and practice, even as meaning and practice reshape the law. In this study, the problem of sexual harassment implicates a specific legal domain with a complex web of laws that defines inappropriate behaviors, establishes conditions of holding employers liable, and specifies remedies for rights violations. These rules provide women with the opportunity to resist unwanted sexual attention at work—interpretive frames through which women understand that a certain conduct is wrong and should be resisted (see also Marshall 2003). More important, the legal rules mandate grievance procedures to process complaints before major problems erupt in the workplace. These grievance procedures provide women with opportunities to protect their rights and to resolve their problems. Thus, their evaluation of these procedures—the meaning they assign the process—will shape their responses to incidents of sexual harassment.

But this study has also demonstrated that laws and legal institutions should be evaluated in the context of particular organizational settings. Institutional theories of organizations have already shown that organizations do not implement legal rules in straightforward ways. Rather, organizational actors imbue legal rules and institutions—such as grievance procedures—with organizational interests. Thus, many supervisors in this study interpreted and implemented the Written Policy in way that protected the University from employee complaints by creating an adversarial process. Their strategies worked, as many women who might have complained declined to do so, preferring to handle the situation on their own. Thus, studies of the legal consciousness of employees, for example, should “bring institutions back in” and account for the influence of specific organizational practices.

Finally, studies showing how constitutive theories of law actually work should also embrace not just meaning but also emphasize the importance of understanding legal consciousness as a form of social and cultural practice. Judicial opinions and EEOC regulations articulate rights, but those rights depend heavily on the initiative of ordinary individuals to invoke them—not just in the courtroom, but also in the context of their daily lives. This initiative, in turn, depends on the availability and the relevance of legal schema to people confronting problems in the workplace. Beyond these commonplace understandings, the meaning of rights also depends on what people do. They may invoke rights to make demands and to seek the intervention of third parties to resolve disputes, and as this study demonstrates, the context of those practices shapes the meaning of those rights. But in analyzing legal consciousness as a social practice, it is also worth noting the times when rights are ignored. When employees reject a grievance procedure, when they say that the policy is not relevant to their dispute, when they let their rights remain idle, they diminish the power of law to constitute their everyday relationships at work.