STRATXGY USING DATA FROM COURT CASES AND EMPLOYEE SURVEYS TO DESIGN SEXUAL HARASSMENT POLICIES Toni P. Lestero Babson College ABSTRACT Study afier study shows that sexual harassment is a problem that threatens to undermine many of thei nroads made by working womenin the last three decades. Small business managers who fail to adopt policies to prevent harassment risk losing valuable employees, the goodwill of the public, and expensive lawsuits. Data from leading federal couri cases and two land- mark federal employee surveys can be used ioformulate policies rhat pre vera sexual harassment from occurring. INTRODUCTION After Anita Hill accused Clarence Thomas of sexual harassment during his Supreme Court nomination hearings in 1991, the Equal Opportunity Commission (EEOC) reported that the number of harassment claims in 1992increased by 50 percent over the number of claims that were filed for a similar period in 1991 (Adams, 1992). In 1993, there were 1,500 sexual harassment claims filed with the agency, with claimants winning $25.2 million in damages from employers ("A Special Report", 1994). In addition to the monetary costs associated with these claims, employers run the risk of losing valuable employees and the public's goodwill. Plaintiffs risk having their credibility questioned and their career prospects handicapped. Current federal law allows successful harassment claimants to receive back pay and up to $50,000 in combined compensatory (i.e., for such things as emotional pain and suffering) and punitive damages from small- and medium-sized companies who employ 15 to 100 employees. Larger companies are subject to higher statutory caps. For example, companies with 100 to 200 employees are at risk for up to $ 100,000, and companies with 200 to 500 employees are at risk for $200,000 (Civil Rights Act, 1992). Pto vis or This article isisuended to provide general legs! information and does not cons ti rute the gi ving ofspecific legal advice to individual readers. Notes Thc aurhor wishes to thonk litt La Rocca for her assistance wirh the collection of the research data, my cotteagtws, Donna La roti k, Lydiiia Pot urni k, and Ross Pe ay for their arsis isnca with the interpretation ofthe research daht, and the Babson Baard of Research for irs suppon. 15 Alternatively, harassment victims can sometimes bring their claims under state laws, many of which do not limit damage awards to the same extent that they are limited under federal law. For example, in one California case, a jury awarded a male harassment victim who was harassed by his female supervisor, $82,000 for economic losses, $375,000 for emotional distress, and $550,000 in puniuve damages (Gurierrez v. California, 1987). Small business managers can avoid many of the above risks by adopting sexual harassment policies that make it clear to their employees that certain types of conduct will not be tolerated. One of the most important sources of information that can be used to help formulate such policies is data from federal courts of appeal cases. Courts of appeal are the highest level of courts to review harassment claims, next to the U.S. Supreme Court. They have a strong impact on how sexual harassment law develops. Appeals court cases are particularly informative because they tell us what type of behavior judges consider sufficiently egregious to constitute illegal sexual harassment. Companies can use this information to prohibit specific conduct in their own harassment policies. An effective harassment policy, however, will do more than simply comply with the minimum requirements of the law. Even when no litigation is involved, the costs to harassment victims in poor physical and mental health and low morale are great. Employers can also lose millions of dollars a year because of high job turnover rates, lost productivity, and the sick leave used by employees who try to cope with harassment (U.S. Merit Systems, 1988). One study indicated that approximately $300 million was lost by the federal government because of sexual harassment (US Merit Systems, 1988). From a legal, economic and ethical perspective, it is therefore imperative that small business managers formulate sexual harassment policies that accommodate legal developments and thc actual experiences of harassment victims. Although both men and women can be harassed, women are by far the most common victims of harassment (U.S. Merit Systems, 1988; 1981), An extremely valuable source of information about the experiences of women harassment victims are two landmark studies on sexual harassment that were conducted by the U.S. Merit Systems Protection Board in 1981 and 1988 ("1981 Merit Study" and "1988 Merit Study", respectively). As the discussion below will show, data from thc court of apfieal cases and the two Merit Studies indicate that companies can reduce their risk of legal liability by adopdng and implement- ing policies that prohibit sexual harassment perpetrated by coworkers and supervisors that consists of a physical, verbal or visual conduct. This article will analyze the Merit Study and court case data to explain how this conclusion was reached. First, this article will describe a database that was developed to objectively characterize the types of harassment that have been reviewed by the federal courts in the past few years and explore the extent to which a preliminary study of the database reveals any relationship between particular types of harassment and a company's chances of winning or losing a lawsuit. Next, this article will discuss the results of the two Merit Studies and compare these results with the court cases reviewed to determine the extent to which court-prohibited harassment did or did not reflect the type of harassment reported by the women in the studies. Finally, suggestions will be made about what specific types of conduct should be prohibited in harassment policies and the type of language that can be included in policies that scck to describe that conduct. 16 TYPES OF SEXUAL HARASSMENT PROHIBITED BY THE FEDERAL COURTS The Supreme Court and EEOC Definitions of Sexual Harassment The Supreme Court first recognized that sexual harassment was a legitimate cause of action under TheCivilRightsActof1964in1986,inMeriiorv.Vinson. Commentingontheemployee's claim that she was coerced into having sex with her supervisor for several years, the Court said that "a requirement that ...a woman run a gauntlet of sexual abuse in return for the privilege of being allowed to work and make a living can be as demeaning and disconcerting as the harshest of racial epithets" (Meriior v. Vinson, 1986, p. 67). In rendering its decision, the Supreme Court relied on the EEOC's guidelines on sexual harassment, which said that harassing conduct included "unwelcome sexual advances, requests for sexual favors, and other verbal or physical conduct of a sexual nature" (Code, 1993). The guidelines further state that for the conduct to constitute illegal sexual harassment, one of the following circumstances has to have taken place: (I) the conduct is made either explicitly or implicitly a condition of a person's employment, (2) decisions about the employee's job status are influenced by the extent to which the employee accepted or rejected the harassing conduct, or (3) the conduct either unreasonably interferes with the employee's job performance or creates a hostile, offensive, or intimidating work environment (Code, 1993). The first of the above two sets of circumstances are usually characterized as "quid pro quo" harassment (i.e., sexual favors are required in exchange for tangible job benefits). Most legal experts believe that supervisors and other people in positions of authority over an employee are capable of engaging in quid pm quo harassmenL The third circumstance is usually called "hostile environment harassment," and can be caused by coworkers or supervisors. After this research project was completed, the Supreme Court made a ruling in another sexual harassment case, Harris v. Forklift Systems, inc. In the decision, the Court realfirmed its condemnation of sexual harassment and concluded that harassment victims need not suffer severe psychological harm in order to pursue their claims in court (Harris v. Forklifi, 1993). All federal appeals courts are bound to interpret and follow the rules set down in Supreme Court decisions like Vinson and Harris. Judges in these courts choose cases on appeal from the lower district courts that focus on issues that involve controversial legal questions or that address new and evolving areas of the law. Thus, appeals courts play an important role in the way in which harassment law develops. Once an appeals court renders its decision, the lower district court must rely on the principles articulated in that decision to decide new cases. This research was undertaken to discover (I) what types of harassment prompted the courts of appeal to decide to review sexual harassment cases on appeal, and (2) if there was any relationship between the type of harassment reviewed on appeal and the employer's chances of winning or losing a sexual harassment lawsuit. Creation of the Database During the first stage of the project, over 50 federal sexual harassment court of appeals cases that were decided since the Supreme Court's decision in Vinson were reviewed. A content 17 analysis of those cases for the purpose of categorizing types of harassment was then conducted. Finally, a database was developed that described as objectively as possible the range of harassing behavior that was covered in those cases and the range of possible decisions that could be rendered with respect to that conduct (i.e., Did the judge decide that the behavior did or did not constitute illegal sexual harassment?). A graduate student assistant also analyzed cases to ensure that conduct was categorized consistently. The attempt to classify conduct was influenced by the two Merit Studies, both of which showed that a wide variety of conduct constituted sexual harassment. Approximately 20,000 employees were surveyed in the 1981study and 8,500employees were surveyed in the 1988 study (U.S. Merit Systems, 1988, 1981). Developers of the 1981 Merit Study said that harassment ranged from uninvited sexual remarks to actual or attempted assauli or rape. While studies like the two Merit Studies have been used to categorize harassment from the victim's perspective, no similar approach has been used by legal scholars to develop a system for categorizing the types of harassment that are contained in disputes that actually make it to court. Since the United States is a common law country that follows the legal principle of stare decisis. this type of information would be extremely useful to legal scholars and practitioners, and business managers. "Stare decisis" is a doctrine that requires couns to render decisions by following the rulings of previously decided similar cases (Metzger, 1989). Lawyers advocate on behalf of their clients by finding past rulings in cases with similar fact scenarios and using those rulings to support their current clients in courL A review of a significant portion of the cases eventually revealed that one or more of the following five categories of harassment were present in each of the court cases reviewed: I) Physical harassment: unwelcome physical conduct ranging from lewd gestures to sexual intercourse with the alleged harassment victim, 2) Verbal harossmenn unwelcome oral or written comments ranging from requesting dates to threatening the alleged victim with physical harm, 3) Visual itarassmeiu: unwelcome visual displays, such as pornographic material on the walls in company common areas, 4) Co-worker harassment i any of the first three categories of harassment that could also be attributed to a co-worker of the alleged victim, and 5) Supervisor harassmeiui any of the first three categories of h uassment that could also be attributed to someone occupying a hierarchical position over the alleged victim. Finally, in addition to auempting to characterize types of conduct, the content analysis revealed that there were two general types of rulings that could be rcn&lcred by the courts: (I) that the facts alleged by the employee amounted to sexual harassment, or (2) that the facts alleged by the employee did not amount to sexual harassment. it should be noted that some companies may escape ultimate liability in sexual harassment cases even when the courts find that the particular conduct under review constitutes sexual harassment. For instance, sometimes courts are reluctant to make a company automatically liable 18 for the wrongful actions of lower level employees, like co-workers, especially when a company had no knowledge of the harasser's actions prior to the commencement of the lawsuit. This research, however, limited itself to determining the extent to which the conduct in question was condemned by the courts and not how the courts addressed this issue of vicarious liability. An initial study of recent federal court of appeal's cases was then conducted to determine if any preliminary inferences could be drawn about the relationship between types of harassment and the final decisions in those cases. Description of and Problems with the Court Cases Studied Of the 50 court cases that were initially reviewed, only 25 of those cases described the harassment under review sufficiently enough to be included in the study. These 25 cases, however, do include at least one case from each of the 11 federal appeals court circuits. Thus, the database reflects the broad range of decisions that have been rendered by all the courts of appeal in the country. Although the more numerous lower disuict court decisions theoretically might have produced a larger sample, it is difficult to get access to all district court decisions because they are not reported consistently. Furthermore, from a legal and practical perspective, district court decisions are not as influential as court of appeal's decisions. Usually, district courts only have local impact because they are required to follow decisions rendered by couns of appeal located in their respective circuits. It is common, however, for a court of appeal in one circuit to follow a decision rendered by a court of appeal in another circuit. Thus, appeals court decisions have a strong impact on the way in which the law develops around the country. From the very beginning of this research project, it was difficult to derive objective descriptions of harassing conduct from the court reported decisions. Court decisions, which can range from 10to 40 pages, are written in narrative form, and the judges who deliver these decisions are not always the most organized or coherent writers. As a result, many cases could not be used because the decisions did not describe the nature of the harassment sufficiently or because the issues being addressed on appeal were not relevant to this siudy. Because of the high number of new harassment claims that have been filed with the EEOC, it is likely that over the next few years, many of these new claims will work their way up to the appeals couns. This larger number of appeals cases can then be added to the sample and be the subject of a future study. Despite the problems that were encountered with the court cases, the preliminary study did reveal some interesting information about the nature of sexual harassment and its possible relationship to final court decisions. Analysis of Court of Cases Of the 25 court cases that were reviewed, 5 were won by the employer and 20 were lost by the employer. All of the cases studied involved female plaintiffs. Table I shows the number of all cases lost or won by the employer, by type of harasser. Table 2 shows the number of all cases lost or won by the employer, by types of harassing conduct. An analysis of the information in the two tables will be discussed immediately below. 19 Table 1 Cases Won Or Lost By Type of Harasser ¹(%) Of All ¹ (9o) ¹ (9o) Cases Reviewed» Won Lost Coworkers 3 (12%) 0 3 (100%) Supervisors 13 (529o) 2 (15%) 11 (859») Coworkersf Supervisors 9 (369o) 3 (33%) 6 (67%) » 25 cases were reviewed Table 2 Cases Won Or Lost By Type of Harassment ¹(9o) Of All ¹ (9o) ¹ (%) Cases Reviewed'on Lost Physical 17 (689b) 2 (129») 15 (88%) Verbal 24 (969o) 5 (21') 19 (79%) Visual 9 (36%) 2 (229») 7 (78%) » 25 cases were reviewed Companies tended to lose lawsuus when the employee was harassed exclusively by a supervisor. In 13 (52 percent) of all the cases in the study, the employee was allegedly harassed by only one type of harasser —a supervisor. This figure is understandable in light of the Vinson decision's strong condemnation against supervisor harassment. It suggests that a majority of the courts followed the Supreme Court's lead and decided that an employee should be entitled to have her day in court if she could produce preliminary evidence that she was harassed by her supervisor, Furthermore, of the cases involving supervisor-only harassment, 11 (84 percent) were lost by the employers and two (15 percent) were won by the employers (!lee Table 1). The difference between these two figures indicates that supervisor harassment —on its own —strongly influenced the court's final determination that harassment from a supervisor constituted illegal sexual harassmenL 20 Companies tended to lose lawsuits when the employee was harassed exclusively by a co-worker. The research on this point indicated that the appeals courts have yet to address the issue of co-worker harassment to any significant degree. Only three (12 percent) of all the cases in the sampleinvolvedharassmentfromaco-workerexclusively. This small numberofcasesmayrelate to the fact that the Vinson case did not speciTically involve co-worker harassment. As a result, lawyers representing clients in post-Vinson harassment cases may have been reluctant to file appeals in cases where their clients were subjected only to co-worker harassment. This is because they would have had no Supreme Court level precedents to rely upon that supported their position. It is interesting to note, however, that three (100 percent) of all of the co-worker-only harassment cases were lost by the employers, notwithstanding the facts in the Vinson case. While this figure may suggest that the courts consider co-worker harassment to be sufficiently egregious to constitute illegal sexual harassment, it could also be argued that the number of cases is too small to produce any significant results. The three cases, however, may represent a small but growing trend in the courts to view co- worker harassment as a form ofhosule environment harassment. For example, Ellison v. Brady, one of the three cases in the sample that involved co-worker only harassment, is a landmark 1991 appeals court decision from the influential 9th circuit in California. Ellison broke with tradition by declaring that love leuers and verbal overtures (i.e., non-physical conduct) from a co-worker constituted hostile environment sexual harassment (Ellison v. Brady, 1991). The Elli son case received a great deal of media attention when it was decided. It has already influenced decisionsincertainU.S.courtcases(Harris v. Forklift, 1993;Robinson v.Jacksonville Shipyards, 1991)and has been cited in important international treatises on sexual harassment law as well(int'I Labour Office, 1992). Since the sample contains no similar post-Ellison cases, it will be important to examine co-worker harassment in future studies. Cases in which a woman was harassed by both a supervisor and a co-worker were also reviewed. Table I shows that nine (36 percent) of all the cases in the pilot study involved a combination of co-worker and supervisor harassers. Six (67 percent) of those cases were lost by the employers and three(33 percent) were won by them. In view of the fact that there were so few cases involving co-worker-only harassment that made it to court, these figures may indicate that co-worker harassment had to occur in combination with supervisor harassment before a signifi- cant number of courts would consider co-worker harassment serious enough to prohibit it. Companies tended to lose lawsuits when the employee was physically harassed. The sample results indicated that the appearance of physical harassment was very important to the courts. Seventeen (68 percent) of all the cases in the sample involved some kind of alleged physical harassment. This may indicate that the occurrence of physical harassment was the minimum fact scenario required to convince the courts to review a case in the first place. Table 2 also shows that 15 (88 percent) of all the cases involving physical harassment were lost by the employers and two (12 percent) were won by the employers. This may suggest that an employee who was physically harassed may have had a good chance of both getting a court to review her case initially and to render a final decision against the employer. 21 Companies tended to win lawsuits when the employee was subyected to non-physical harass- ment fi.e., verbal or visual). (1)Verbol Harassment. Almostall of the courtcascs studied(i e.,24(96 percent)) included some form of alleged verbal harassment. Verbal harassment also seemed to have an impact on the final decisions in those cases. For instance, Table 2 shows that there was a large difference between the percentage of verbal harassment cases that were lost by companies (i.e., 19 (79 percent)) and the percentage of cases that were won by them (i.e., 5 (20.8 percent)). (2) Visual Harassment. Visual harassment was less prevalent in the case sample than verbal harassment. Of all of the court cases examined, only nine (36percent) involved visual harassment. There was a sizable difference, however, between the percentage of all the visual harassment cases that were lost (i.e., 7 (78 percent)) and the percentage of cases that were won by the employers (i.e., 2 (22 percent)) (See Table 2). Thus, visual harassmem may also have been an important measure of liability for the judges who reviewed the cases in the study. A 1991 federal district court case from Florida, Robinson v, Jacksonville Shipyards, may provide some insight into the way in which the courts will view visual harassment in the future. ln this case, the judge determined that the display of pornography in the work place did in fact constitute illegal sexual harassment. The court said that pornography "comm unicates a message about the way [the employer] views women, a view strikingly at odds with the way women wish to be viewed in the workplace" (Robinson v. Jacksonville Shipyards, 1991,p. 1509). While the Jacksonville case was only decided by a district court, it reccivcd a great deal of attention in the legal and business communities because thc judge relied on the expert testimony of a psychologist and a management consultant to show how visual harassment negatively affects women at work. As such, this unusual case has the potential to influcncc other courts across the country, despite the fact that it was only decided by a district court. WHAT HARASSMENT VICTIMS HAVE TC) SAY ABOUT THE TYPES OF HARASSMENT THAT THEY ARE SUBJECTED TO AT WORK Women in the morc recent 1988 Merit Study made it clear that sexual harassment continues to be a big problem for them at work. Forty percent of the women in the 1981 Merit Study and 42 percent in the 1988 Merit Study said that they had experienced some form of sexual harassment. A discussion of the extent to which the women surveyed said that they experienced each of the six categories of harassment and a comparison of the survey results with the court cases reviewed will be discussed below. Supervisor Harassment is Very Serious and Occurs Less Often than Co-worker Harassment The survey respondents viewed supervisor harassment as one of the most egregious forms of sexual harassment. For example, 99 percent of the women in the 1988 Merit Study believed that pressure for sex from a supervisor should be equated with sexual harassment, and 87 percent said that a supervisor's pressure for dates should consutute illegal sexual harassment. 22 The women also said that they were harassed a lot less frequently by their supervisors than the court case data indicated. Although 52 percent of all the court cases discussed above involved supervisor-only harassment, only 29 percent of the women in the more recent 198& Merit Study reported that they had been harassed by an immediate supervisor or someone holding a position higher than that. It should not be inferred, however, that supervisor harassment should not be taken seriously by employers. Because of the unequal power relationship that exists between supervisors and their subordinates, supervisors have the capacity to threaten the job prospects of employees who shun their advances. The Supreme Court acknowledged this in the Vinson decision when it made quid pro harassment (i e.,supervisor harassment that either explicitly or implicitly ties itself to an employee's job status) illegaL That the courts take this matter seriously is also evidenced by the fact that 84 percent of all the supervisor-only cases reviewed were lost by the employers. Co-worker Harassment is Also Serious and it Occurs More Frequently than Supervisor Harassment Although the respondents seemed to view co-worker harassment as a slightly less serious form of harassment than supervisor harassment, a significant number of women still believed that co-worker-only harassment was an inappropriate form of work place behavior. For instance, in the more recent 1988 Merit Study, 87 percent of the women said that they believed that supervisor pressure for dates should constitute sexual harassment, while 76 percent said that co-worfrer pressure for dates should constitute sexual harassment. One of the most imponant differences between the court case data and two Merit Studies is what they have to say about the pervasiveness of co-worker harassment. As was stated earlier, only 12percent of all the couitcases in the sample involved co-worker harassment exclusively. However, 65 percent of the women in the 1981 Merit Study and 69 percent of the women in the 1988 Merit Study said that they had been harassed by a co-worker or someone who had no supervisory authority over them. These results may suggest that even though women were generally subjected to co-worker harassment much more often than they were to supervisor harassment, they may have been reluctant to file lawsuits alleging co-worker harassment because they feared that their claims would not be taken seriously by the courts. Physical Harassment is Serious but Occurs Less Frequently than Non-physical Harassment As was discussed earlier, the court case data indicated that the courts found physical harassment to be particularly abhorrent. Eighty-four percent of all the court cases involving some form of physical harassment were lost by the employers. An even greater percentage of women in the 1988 Merit Study said that they believed that unwanted deliberate touching, pinching, or cornering should be considered a form of sexual harassment. Ninety-five percent said this should be the case when the harasser was a supervisor and 92 percent said this should be the case when the harasser was a co-worker. The overall percentage of women in the two Merit Studies who experienced some form of physical harassment, however, was much lower than it was for the women covered by the court case data. For instance, 68 percent of all the court cases studied involved some form of physical harassment. In contrast, only 26 percent of the women in the 1988 Merit Study said that they were subjected to physical harassment. Furthermore, only 0 8 percent said that they had been the victim of actual or attempted rape or assault. 23 Managers should be careful not to infer from this discussion that physical harassment should not be taken seriously, however. While the 0.8 percent figure in the 1988 Merit Study may appear to be small, if this percentage were applied to the entire work force at that time, this could have amounted to over 6,000 women who had been the victims of some form of actual or attempted assault. Physical harassment can have a long lasting and traumatic effect on the life of a harassment vicum. It is also a crime. Verbal Harassment is Serious and Occurs More Frequently than Physical Harassment In the 1988 Merit S wdy, 72 percent of the respondents believed that unwanted sexual comments should constitute sexual harassment if the harasser were a supervisor, and 64 percent said that this should be true if the harasser were a co-worker. While these numbers are lower than thc percentage of women who equated physical harassment with sexual harassment (i.e., 92 percent and 95 percent for co-worker and supervisor harassment, respectively), they still represent a significant number of women who believed that verbal harassment should bc taken seriously. Women in the studies also said that they were exposed to verbal harassment slightly more often than they were to physical harassment. Approximately 35 percent of the women in the 1988 Merit Study said that they were the objects of unwanted sexual remarks, such as teasing, jokes, or questions. Verbal harassment may have actually been more pervasive than these figures suggest because the sponsors of the 1988 Merit Study also asked respondents about other forms of verbal harassment. For instance, the 1988 Merit Study also asked employees if they received unwanted calls, pressure for dates or pressure for sexual favors. Twelve percent of the women said that they had received unwanted letters and calls; 15 percent said that they had received pressure for dates; and 9 percent said that they had received pressure for sex. Preliminary Results Indicate that Visual Harassment is Serious but Frequency is Not Clear As has already been discussed, the court case research suggested that the courts were inclined to decide that visual harassment should be prohibited by law. While the two Merit Studies did not ask respondents specifically about visual harassmeru, some of the respondents offered personal comments to describe their experiences with visual harassment. For example, the sponsors of the 1981 Merit Study said that many of the women found "materials of a sexual nature bothersome. One woman dislike[d] the way her male coworkers pass[ed] amund and put up pornographic cartoons in work spaces. When she object[ed], hcr Imss [told] her [she was] too sensitive" (US Merit Systems, 1981,p. 37). However, the responderus were not asked about the pervasiveness of visual harassment WHAT TYPE OF HARASSMENT SHOULD BE PROHIBITED IN COMPANY HARASSMENT POLICIES? Although the courtcase sample may be too small to produce any hard statisucal conclusions at this time, there are still some preliminary inferences that can be drawn from the study that can serve as a starting point for the formulation of company harassment policies. Since employees 24 and employers each share a signiTicant portion of the costs associated with harassment even when itdoesn't lead to litigation, an effective policy will cover the range of conduct that women in the surveys said that they were subjected to. Examples of how this can be accomplished are shown in Table 3 and discussed below. Table 3 Types of Harassment ro be Prohibited Supervisor Hamssment Co-worker Harassment Physical Harassment Verbal Harassment Visual Harassment Supervisor Harassment Companies should prohibit supervisor harassment because of the Vinson decision's strong pronouncements against it, and because the court case research indicates that supervisor harassment negatively affects a company's chances of winning a harassment lawsuit. Further- more, supervisor-only harassment should beprohibited because almost all of the women surveyed said that it was anathema to them. This strong response probably can be attributed to the fact that the respondents recognized that there is an inherent unequal power relationship that exists between supervisors and their subordinates. A sample harassment policy can be found in the Appendix, which lists the types of conduct that the Jacksonville court ordered the employer to prohibit. Co-worker Harassment A substantial number of the women in the more recent 1988 Merit Study (i.e.,70 percent) said that they had been subjected to co-worker only harassment. Co-worker harassment thus appears to be a widespread problem. Because it has the ability to reduce the emotional well-being and job productivity of targeted employees, managers need to make all employees aware that this type of conduct will not be tolerated. Furthermore, co-worker harassment and all other types of non-quid pro quo harassment, are legally prohibited forms of hostile work environment harass- ment. Thus, managers should take proactive steps to eliminate this type of harassment. Physical Harassment Company policies should also prohibit physical harassment, since the court case research clearly indicates that companies have a strong chance of losing cases if it can be proven that physical harassment took place. Physical harassment should also be condemned because the women surveyed almost uniformly found it to be offensive and because it can be considered a form of criminal behavior. 25 An example of the type of language that can be used in a policy to prohibit physical harassment, can be found in the Jacksonville Shipyards case. In this case, the court ordered the company to adopt a sexual harassment policy that prohibited the following: I) rape, sexual battery, molestation, or attempts to commit these assaults; 2) intentional physical conduct that is sexual in nature, such as touching, pinching, patting, grabbing, brushing against another employee's body, or poking another employee's body (Robinson v. Jacksonville Shipyards, 1991,p. 1542). More detailed language can be found in the Appendix. Verbal Harassment Verbal harassment also should be covered in harassment policies, since of all the types of harassment discussed in this article, verbal harassment appeared to occur most frequently and to be viewed seriously by the respondents the Merit Study respondents. For specific language that can be included in a policy to prohibit verbal harassment, managers can look to the Jacksonville Shipyards case. It prohibits., among other things, "sexually- oriented gestures, noises, remarks, jokes, or comments about a person's sexuality or sexual experience directed at or made in the presence of any employee who indicates or has indicated in any way that such conduct in his or her presence is unwelcoine" (Robinson v. Jacksonville Shipyards, 1991,p. 1542).(See the Appendix for the expanded version of this prohibition). Although the Jacksonville policy does not refer to written harassment, it also seems logical to conclude that written harassment should be prohibited as well, since both written and verbal harassment involve the unwanted communication of sexually-charged subject matter. Visual Harassment In light of the 1981 Merit Study, which showed that many of the women surveyed were offended by visual harassment, and the potential for the Jacksonville Shipyards case to influence future judicial attitudes about the impropriety of visual harassment, managers should also prohibit visual harassment in their policies. The court in the Jacksonville Shipyards endorsed a broad condemnation of visual harassment. It ordered the company to draft a policy that prohibited "displaying pictures, posters, calendars, graffiti, objects, promotional materials, reading materi- als or other materials that are sexually suggestive, sexually demeaning, or pornographic ..." (Robinson v. Jacksonville Shipyards, 1991,p. 1542). (See the Appendix for an expanded version of this prohibition.) CONCLUSION In addition to the information contained in this article, managers can consult local legal specialists in employment law to get more individualized advice about sexual harassment law. Since sexual harassment hw is constantly evolving, companies should maintain an ongoing relationship with a good attorney who has a sound background in this field. Management consultants who specialize in this field can also provide important advice on this topic. 26 Written policies that prohibit specific types of harassment, however, are only the first step that companies should take to eliminate sexual harassment. Such policies, if not genuinely implemented and endorsed by management, will have no real effect on the problem. By their own behavior, high-level managers should show other employees what appropriate work place behavior looks like. Companies also need to offer mandatory training programs to both supervisory and non- supervisory employees. These programs should educate employees about the range of conduct covered in company policies, the penalties that will be assessed against harassers, and information about company grievance procedures. Such programs should be offered to new employees and all employees on a periodic basis. Training programs on harassment can be conducted by a company's in house human resources department or duough the use of outside consultants. If the latter is economically prohibitive, small business managers can instead buy videos about harassment that are currently on the market or available in local business college or public libraries. Managers also need to respond to complaints about harassment swiftly, by investigating complaints and penalizing harassers if it appears that harassment has occurred. The first time an employee engages in harassment, the harasser should receive a formal warning and be told that, if the harassment continues, the ultimate penalty is firing. Of course, more egregious forms of harassment, like physical harassment, should warrant a more serious response, even if it only occurs once. Finally, companies should be careful to provide employees with the opportunity to report supervisory harassment to other employees who are outside of the supervisory chain of command. In addition to the positive impact that the adoption of the above measures can have on overall employee morale, they also serve the important function of reducing an employer's future risk of legal liability. Generally, if a company has a clearly communicated harassment policy, a training program, and a grievance process that is triggered as soon as the company knows or has reason to know that harassment is taking place, the company has a very good chance of winning a harassment lawsuiL It should be noted, however, that there recently has been a trend recently in some judicial circuits for judges to hold companies auroman'cally liable for supervisory harassment, even when those companies had no reason to know that the harassment was taking place (Plevan, 1994). In these cases, courts treat harassment claims like workmen's compensation claims and conclude that companies have an absolute duty to keep their workplaces safe and free from harassment. It remains to be seen whether or not this trend will continue. Even if the above trend does not continue to grow, some of its underlying premises may be good for managers to heed. Companies need to consider the backgrounds of the people that they hire, especially when they are considering hiring people who have a history of harassing others at their previous place of employment. Companies may also want to survey their employees confidentially to determine if sexual harassment is a problem in their organizations. Such precautions can help eliminate the problem of work place sexual harassment and the risk of potential legal liability as well. 27 IMPLICATIONS FOR FURTHER RESEARCH As was mentioned at the beginning of this ardcle, the 25 courts of appeal cases were reviewed to determine if any preliminary inferences could be drawn about the relationship between types of harassment and final court decisions. The preliminary inferences discussed above indicate that there are several issues that lend themselves to further research and review. First, courts of appeal cases, as they are decided, should be adde&1 to the sample to enrich the research sample. It also might be useful to sub-categorize the five different forms of harassment discussed here to see if the courts treat these subcategories differently. For example, physical harassment could be divided into two separate categories (i.e., lewd gestures vs. unwanted touching) to see which category is taken more seriously by the courts. Finally, it would be interesting to compare district court decisions with courts of appeal decisions or to do a study of district court cases exclusively. Those considering engaging in such resemch should be forewarned, however. Since district court cases are more numerous and more difficult to get access to, such a literature review would be extremely time-consuming and require significant staff and research support to accomplish effectively. Even if none of the above suggestions for further research are undertaken, companies still have a wealth of information to draw from the two Merit Studies. This information, on its own, is enough to justify implementing harassment policies that cover the types of conduct discussed in this article. REFERENCES Adams, Marilyn. (July, 1992). Sex harassment claims up sharply. Boston Globe, 3. A special mport on people and their jobs in offices, fields and factories. (April 19, 1994). Wall Street Journal, A 1. Civil Rights Act, 41 U.S. Code, Sec. 2000e (1992). Civil Rights Act Amendments, 42 U.S. Code 1981,Sec. 1981(a) g& (b) (1992). Code of Federal Regulations, Vol. 12 Sec. 1604.11(a)(1993). Ellison v. Brady, 924 F.2d 872 (9th Cir. 1991). Guii errer v. California Acrylic Industries, 832 F. 2d 194 (1st Cir. 1987). Harris v. Forklift Systems, Inc., 1993 U.S. Lexis 7155 (Nov. 9, 1993). International labour office. (1992). Combating Sexual Harassmeni Ai Work, 16. Interview with James Neely, Deputy General Counsel, U.S., Equal Opportunities Commission, Interview (July 21, 1994). Lester, Toni. (1993).The reasonable woman test in sexual harassment law —will it really make a difference? Indiana Law Review, 26 (2). Meritor Sav. Bank v. Vinson,477 U.S. 57 (1986). Metzger, Michael B.,et al (1989). Business Law And The Regulatory Environment, 1547. Plevan, Bettina B. &k Borg, Jennifer A. (August 8, 1994). Rulings by several courts of appeals may lead to expanded employer liability for supervisor's conduct in hostile work environ- ment sexual harassment cases. National Law Journal, B5, B8-9. Robinson v. Jacksonville Shipyards, 760 (N.D. Fla.) 1486 F. Supp. (1991). Twomey, DavidP. (1993) Equal Opporiuni&iesLaw. Cincinnati, W,Chicago: Southwestern Publishing. U.S. Merit Systems Protection Board. (1988). Sexual Harassment (n The Federal Government: An Update. U.S. Merit Systems Protection Board. (1981). Sexual Harassment In The Federal Workplace, Is Ii A Problem? 28 APPENDIX Sample Sexual Harassment Policy on Prohibited Conduct The following is taken from the court case, Robinson v. Jacksonville Shipyards, referenced in endnote [14]: STATEMENT OF POLICY Title VII of the Civil Rights Act of 1964 [(and its 1991amendments)] prohibits employment discrimination on the basis of race, color, sex, age, or national origin. Sexual harassment is includedin the prohibitions. Sexual harassment, according to the federal Equal Opportunity Commission (EEOC), consists of unwelcome sexual advances, requests for sexual favors or other verbal or physical acts of a sex-based nature where (I) submission to such conduct is made either explicitly or implicitly a term or condition of an individual's employment; (2) an employment decision is based on an individual's acceptance or rejection of such conduct; or (3) such conduct inter- feres with an individual's work performance or creates an intimidating, hostile or offensive working environment. It is also unlawful to retaliate or take reprisals in any way against anyone who has articulated concern about sexual hamssment or discrimination. Examples of conduct that would be considered sexual harassment ...are set forth in the Statement of Prohibited Conduct which follows. These examples are provided to illustrate the kind of conduct proscribed by this Policy; the list is not exhaustive. STATEMENT OF PROHIBITED CONDUCT The management of the company considers the following conduct to represent some of the types of acts that violate the company's sexual harassment policy: A. Physical assaults of a sexual nature, such as: I) rape, sexual battery, molestation, or anempts to commit these assaults; and 2) intentional physical conduct that is sexual in nature, such as touching, pinching, patting, grabbing, brushing against another employee's body, or poking another employee's body. 29 SAMPLE SEXUAL HARASSMENT POLICY ON PROHIBITED CONDUCT B. Unwanted sexual advances, propositions, or other sexual comments, such as: 1) sexually-oriented gestures, noises, remarks, jokes, or comments about a person' sexuality or sexual experience directed at or made in the presence of any employee who indicates or has indicated in any way that such conduct in his or her presence is unwelcome; 2) preferenual treatment or promise of preferential treaunent to an employee for submitting to sexual conduct, including soliciting or attempting to solicit any employee to engage in sexual activity for compensation or reward; and 3) subjecting, or threats of subjecting, an employee to unwelcome sexual attention or conduct or intentionally making performance of the employees job more difficult because of that employee's sex. C. Sexual or discriminatory displays or publications anywherein ...i(he company'si work place by ...lcompanyl employees, such as: l) displaying pictures, posters, calendars, graffiti, objects, promotional materials, reading materials, or other materials that are sexually suggestive, sexually demean- ing, or pornographic. A picture will be presumed to be sexually suggestive if it depicts a person of either sex who isnotfullyclothedorinclothesthatarenotsuitedtoorordinarilyacceptedfortheaccomplishment for routine work in and around the ...[companyl and who imposed for the obvious purpose of displaying drawing attention to private portions of his or her body. 30